A MOLECULAR
EPIDEMIOLOGICAL STUDY ON
HAEMORRHAGIC SEPTICAEMIA
DISEASE IN PAKISTAN
Ahmed Magdi Moustafa Mahmoud Moustafa
BPharm
This thesis is presented for the degree of Doctor of Philosophy of
Murdoch University
2014
I declare that this thesis is my own account of my research and
contains as its main content work which has not previously been
submitted for a degree at any tertiary education institution.
....................................................................................
Ahmed Magdi Moustafa, 2014
I dedicate this thesis to the soul of an inspiring friend and
colleague; Amr Kassem who lost his life in the incidences of the
Egyptian revolution in 2013. It's also dedicated to his strong wife
Mrs. Asmaa Hussein and his lovely daughter Ruqaya.
i | P a g e
Table of Contents
Table of Contents ........................................................................................................................................... i
Acknowledgements ...................................................................................................................................... vi
Abstract ......................................................................................................................................................... x
Conference Proceedings and Publications Arising from this Work............................................................ xiv
List of Abbreviations .................................................................................................................................... xv
List of Figures ............................................................................................................................................. xix
List of Tables .............................................................................................................................................. xxii
Acknowledgement of Work Not Performed by the Author ...................................................................... xxiv
Notes to Reader ........................................................................................................................................ xxv
Chapter 1 – Purpose of Project and Outline of Thesis .................................................................................. 1
1.1 Purpose of Project ............................................................................................................... 1
1.2 Outline of Thesis .................................................................................................................. 2
Chapter 2 – Literature Review ...................................................................................................................... 3
2.1 Introduction ........................................................................................................................ 3
2.2 Pasteurella multocida: Classification and Diseases ............................................................. 3
2.2.1 Morphology ................................................................................................................. 5
2.2.2 Serological classification .............................................................................................. 5
2.2.3 Designation of serotypes ............................................................................................. 6
ii | P a g e
2.2.4 Cellular components ................................................................................................... 7
2.2.5 Preferred media and colony morphology ................................................................. 10
2.2.6 Virulence in experimental animals ............................................................................ 11
2.2.7 Diseases caused by Pasteurella multocida ................................................................ 11
2.3 Haemorrhagic Septicaemia ............................................................................................... 13
2.3.1 Clinical signs .............................................................................................................. 16
2.3.2 Pathology and pathogenesis ..................................................................................... 17
2.4 Epidemiology of Haemorrhagic Septicaemia .................................................................... 18
2.4.1 Distribution of the disease ........................................................................................ 18
2.4.2 Morbidity, mortality and case fatality ....................................................................... 22
2.4.3 Carrier status ............................................................................................................. 26
2.5 Haemorrhagic Septicaemia Epidemiology in Pakistan ...................................................... 31
2.6 Diagnosis of Haemorrhagic Septicaemia ........................................................................... 33
2.6.1 Clinical diagnosis ....................................................................................................... 33
2.6.2 Differential diagnoses ............................................................................................... 34
2.6.3 Laboratory diagnosis ................................................................................................. 34
2.7 Treatment .......................................................................................................................... 56
2.7.1 Antibiotic therapy ...................................................................................................... 57
2.7.2 Antiserum therapy .................................................................................................... 59
2.7.3 Effects of hypertonic saline solution on survival rate of affected buffaloes............. 59
2.8 Prevention and Control ..................................................................................................... 60
iii | P a g e
2.8.1 Vaccine types ............................................................................................................. 60
2.8.2 Prevention and control in endemic areas ................................................................. 60
2.8.3 Prevention and control during outbreaks ................................................................. 60
2.9 Conclusion ......................................................................................................................... 61
Chapter 3 – A retrospective case-control study of haemorrhagic septicaemia in Karachi, Pakistan in
2012 ............................................................................................................................................................. 62
3.1 Abstract ............................................................................................................................. 62
3.2 Introduction ...................................................................................................................... 62
3.3 Materials and Methods ..................................................................................................... 64
3.4 Results ............................................................................................................................... 67
3.4.1 Univariable analysis ................................................................................................... 77
3.4.2 Multivariable analysis ................................................................................................ 78
3.5 Discussion .......................................................................................................................... 80
Chapter 4 – Molecular typing of haemorrhagic septicaemia-associated Pasteurella multocida
isolates from Pakistan and Thailand using multilocus sequence typing and pulsed-field gel
electrophoresis ............................................................................................................................................ 88
Chapter 5 – Comparative genomics analysis of Asian Haemorrhagic Septicaemia-associated strains
of Pasteurella multocida ............................................................................................................................. 95
5.1 Abstract ............................................................................................................................. 95
5.2 Introduction ...................................................................................................................... 95
5.3 Materials and Methods ..................................................................................................... 97
5.3.1 Whole genome sequencing ....................................................................................... 98
iv | P a g e
5.3.2 Sequence assembly ................................................................................................... 99
5.3.3 Genome alignments and feature analysis ............................................................... 101
5.3.4 Phylogenetic trees ................................................................................................... 102
5.4 Results and Discussion .................................................................................................... 102
Chapter 6 – Development of loop-mediated isothermal amplification (LAMP)-based diagnostic
tests for the specific detection of Pasteurella multocida and haemorrhagic septicaemia-associated
Pasteurella multocida serovar B:2 ............................................................................................................ 119
6.1 Abstract ........................................................................................................................... 119
6.2 Introduction .................................................................................................................... 120
6.3 Materials and Methods ................................................................................................... 121
6.3.1 Bacterial strains ....................................................................................................... 121
6.3.2 Pm-LAMP primer design .......................................................................................... 122
6.3.3 HS-LAMP primer design........................................................................................... 123
6.3.4 HS-1517-PCR ............................................................................................................ 124
6.3.5 Pm-LAMP reaction ................................................................................................... 124
6.3.6 HS-LAMP reaction ................................................................................................... 125
6.3.7 Loop primer ............................................................................................................. 125
6.3.8 Real time evaluation of LAMP reaction ................................................................... 126
6.3.9 Processing colonies and inoculated broth without DNA extraction for Pm-LAMP . 126
6.3.10 Evaluating the sensitivity and specificity of both the Pm-LAMP and HS-LAMP ...... 127
6.3.11 Visualisation of the HS-LAMP and Pm-LAMP products ........................................... 128
v | P a g e
6.3.12 Validation of both the Pm-LAMP and HS-LAMP using Receiver Operating
Characteristic (ROC) curves ..................................................................................................... 128
6.4 Results ............................................................................................................................. 129
6.5 Discussion ........................................................................................................................ 138
Chapter 7 – Thesis Summary and General Discussion .............................................................................. 143
7.1 Thesis Summary .............................................................................................................. 143
7.2 The Aims of the Project ................................................................................................... 146
7.3 Future Directions of Research ......................................................................................... 148
7.4 Conclusion ....................................................................................................................... 150
Bibliography .............................................................................................................................................. 151
Appendix ................................................................................................................................................... 170
vi | P a g e
Acknowledgements
It is with pleasure that I acknowledge the many people that have helped me with my research over
the last four years.
Firstly, I would like to express my deepest gratitude to my supervisors Dr Mark Bennett, Prof. John
Edwards, Dr Tim Hyndman and Prof. Stan Fenwick. Thank you for your patience and support
during this project. I came to you in 2011 with a passion to work on haemorrhagic septicaemia
disease in Southeast Asia. With your help, my knowledge in this area has improved greatly. Your
continuous support allowed me to get the best outcomes that were possible given the difficulties
that were faced at times during my work in the field. I sincerely thank you all for providing me with
the freedom and assistance to pursue this project despite the many challenges that it presented.
A special thank you also needs to go to the many others that helped me at various stages of this
project. To Dr Kirsty Townsend, you were the first one to teach me the basics about Pasteurella
and provided me with valuable books about haemorrhagic septicaemia. Without you, I would not
be at this stage now.
To Prof. Dr M. Iqbal Choudhary, Dr Kamran Azim and Dr Muhammed A. Mesaik, who hosted me at
the International Center for Chemical and Biological Sciences (ICCBS), University of Karachi,
Pakistan, thanks for being my co-supervisors during my stays in Karachi from June 2011 to April
2013, for facilitating my work in your fantastic research environment, for providing me with
scientific advice, and helping me in conducting my field work in Karachi. Also, thank you for
hosting me and for providing the private transportation needed for my field work.
vii | P a g e
I am most grateful to Mr Syed Noman Ali for helping me in conducting the epidemiological case-
control study in Pakistan. Without you, I would not have been able to conduct this study. Thanks
Dr Ghulam Sarwar Shaikh, former director general of livestock department, Government of Sindh,
Pakistan, for your support and help in the epidemiological study done in Karachi.
I would like to express my special appreciation to Dr John D. Boyce, Dr Torsten Seemann, Dr
Marina Harper, Dr Simon Gladman and Prof. Ben Adler for hosting me at their laboratories at the
Australian Research Council Centre of Excellence in Structural and Functional Microbial Genomics,
Monash University. You helped and taught me about the basics of bioinformatics and comparative
genomics for eight weeks. I really enjoyed working with you in the research environment you
have.
My sincere thanks also goes to Prof. Ian Robertson for teaching me the basics of epidemiology,
giving me a generous amount of your time and for conducting the multivariable analysis for the
epidemiological study we conducted in Pakistan. To Dr Rongchang Yang, thanks for providing us
with some bacterial DNAs that were used for evaluating our diagnostic test.
In addition, a thank you to Dr Qurban Ali, Mr Eid Nawaz, Mr Muhammad Abubakar and Dr Rehana
Anjum, who were working at the National Veterinary Laboratories in Islamabad, Pakistan, for
providing me with the Pakistani Pasteurella isolates.
I would like to thank Dr Pornpen Pathanasophon, Ms Apasara Worarach, Ms Naiyana Husrangsi,
Ms Wanwisa Satsomnuek and Mr Gumtorn Promto, who were working at the National Institute of
Animal Health (NIAH) at the Department of Livestock Development in Bangkok, Thailand, for
viii | P a g e
providing me with the Thai Pasteurella isolates. Thank you also for conducting the pulsed field gel
electrophoresis experiments in your laboratories which resulted in a published manuscript.
Thanks to the Veterinary Research Institute in Lahore, Pakistan, for providing a vaccine strain.
Thanks also to Mrs Farheena Iqbal Awan, from the DNA Core Facility at the Centre for Applied
Molecular Biology (CAMB), Ministry of Science and Technology, Lahore, Pakistan, for helping with
DNA sequencing.
Mr Peter Han, Ms Lu Gao, Mr Shengmao Liu and Dr Juanni Zhao from the Beijing Genomics
Institute (BGI), China, I greatly appreciated your help and assistance with my whole genome
sequencing work.
I want to express my deep thanks to Dr Nathan Tanner for his help and support in developing our
LAMP-based diagnostic tests.
Dr Henrik Christensen offered much advice and insight throughout Pasteurellaceae taxonomy. I
thank him for this.
To Asifullah Khan, Saifullah Afridi, Irfan Khan, Burhan Khan, M. Wassem, Mujeeb Rahman, Ishtiaq
Khan, Shan Zeb, Fazal Rahim, Said Nadeem, Akhtar Mohammed, Imad Uddin, Imran Khan, Achut
Adhikari, Rehan Imad, Rehan Siddiqui, Ali Raza, Junaid Kori, Imran Safdar, Nizakat Ali, M. Irfan, Ali
Hussain, Shaz Rashid, M. Saqib, Areeba Anwar, Munixa Shaikh, Sana Aamir and Bushra Bilal, thank
you for your support and hospitality in Pakistan. I really loved your country because of you.
ix | P a g e
Thanks to my department at the faculty of Pharmacy, Alexandria University. In particular, thank
you to my professors and doctors who taught me for five years and made me competent enough
to begin scientific research.
Thanks to my teachers at primary, preparatory and secondary schools who taught me how to love
science.
To El Ameen, E. Elnikity, M. Daoud, M. Sobhy, Badawy, Tunnina, Refaat, Ismail, Mishref, A. Gaber,
Elwy, Sally Mohsen (my childhood best inspiring friend) and Jill Austen, thanks for your support
and always being there for me. Without you in my life, I would not have been able to have
achieved the things that I have.
Finally, but most importantly of all, my family, mum, dad, sister, M. Gaber, Lolo, Khalo Ahmed,
Mama Monna and Teta Aisha, thank you for your unwavering personal and moral support
throughout this project. I thank you for bearing the difficulties of being away from you and not
being able to share with you our special moments. Thanks to my uncles, aunts and cousins for
their endless love and support.
x | P a g e
Abstract
Haemorrhagic septicaemia (HS) is an acute fatal septicaemic disease of cattle and buffaloes
associated with strains of serotypes B:2 and E:2 of the bacterium Pasteurella multocida. Asia and
Africa are currently the regions where HS occurs with the highest prevalence and has the greatest
economic importance. There is currently only limited information available on the diversity of P.
multocida strains that cause HS in these regions.
A retrospective epidemiological case-control study was performed in Karachi, Pakistan from
January to April 2013 looking at HS cases that occurred in the 2012 calendar year. The owners
from 217 dairy cattle and dairy buffalo farms from six different locations around Karachi were
interviewed. The study was based on a questionnaire that was designed to identify independent
variables that were statistically associated with the presence of HS on the farm. The final
multivariable logistic model contained five factors. Two protective factors were identified: HS
vaccination (Odds Ratio (OR) = 0.22) and the length of time cattle were kept on farm (months): for
every extra month cattle were kept, the odds of disease were reduced by a factor of 0.9. In
contrast, supplying underground water and the presence of foot and mouth disease on the farm
increased the risk of infection by 2.90 and 2.37, respectively. In addition, for every extra animal on
the farm, the risk of infection increased by a factor of 1.01. To understand the epidemiology of HS
in Karachi dairy herds more fully, further in-depth research is required to study the risk and
protective factors identified in this survey and to evaluate risk mitigation strategies where
possible. The study also showed the need for developing a point-of-care diagnostic test that can
be used by veterinarians to diagnose the disease on the farm.
Haemorrhagic septicaemia-associated strains of P. multocida were then analysed using molecular
methods. Haemorrhagic septicaemia vaccine strains and HS-associated field isolates were
xi | P a g e
obtained from different places in Pakistan and as a comparison group, vaccine strains and HS-
associated field strains were obtained from Thailand to investigate the genetic diversity of strains
associated with the disease from two endemic countries. Initially, 21 field isolates and three
vaccine strains from different regions within these countries were analysed by multilocus
sequence typing (MLST) and pulsed field gel electrophoresis (PFGE).
The MLST technique was not able to differentiate between the strains from Pakistan and Thailand
as all of the tested isolates (n = 21) were sequence type (ST) 122. The PFGE technique showed a
difference of one band between the Thai and Pakistani isolates. Neither technique was able to
show any variation between isolates from the same country.
Based on the MLST and PFGE results, 12 of the 24 HS-associated strains were selected for next
generation sequencing (NGS). Analyses of the genome data identified a core set of 1824 genes
that were shared by the 12 HS-associated Asian strains, M1404 (North American HS-associated
strain), and the available P. multocida complete genomes of strains Pm70, 3480, 36950 and HN06.
Four sets of unique genes were found. One set (96 genes) was shared by all HS-associated strains.
The second set (59 genes) was shared by the Asian HS-associated strains only. The third set (39
genes) was shared by seven out of nine Pakistani HS-associated strains. The last set (42 genes) was
shared by the remaining two Pakistani HS-associated strains that were studied.
The set of 96 unique genes, found in all HS-associated strains but absent from the non-HS-
associated strains Pm70, 3480, 36950 and HN06, was identified by the PHAST bacteriophage
detection algorithm to encode two putative temperate phages. It seems reasonable to suspect
that the presence of these putative prophages may provide virulence genes that contribute to the
pathogenesis of HS.
xii | P a g e
The set of 59 unique genes shared by the 12 Asian HS-associated strains but not the North
American HS-associated strain M1404, was also predicted to encode a temperate phage. Likewise,
another putative temperate phage was predicted in the seven Pakistani strains that accounted for
the set of 39 unique genes they shared. Interestingly, two Pakistani strains (BUKK and TX1) carried
acquired antimicrobial resistance genes as predicted by the Resfinder tool. BUKK and TX1 strains
shared 42 unique genes that were not present in any other HS strains. Some of these unique genes
may contribute to a putative integrative conjugative element (ICE). The putative ICE of BUKK and
TX1 is not identical to ICEPmu1, the first ICE that was found in P. multocida (strain 36950), and
therefore, it may be the second ICE to be discovered in P. multocida strains.
Phylogenetic analysis, predicted by analysis of core genome single nucleotide polymorphisms,
demonstrated a strong correlation between the individual strains and their countries of origin. The
Pakistani and Thai strains were more closely related to each other than the North American strain.
Similarly, the isolates from Pakistan clustered together, and were distinctly separate from Thai
isolates.
A specific rapid diagnostic test, based on loop-mediated isothermal amplification (LAMP), for HS-
associated B:2 strains, was developed using the previously-identified 96 unique genes obtained
from NGS. Duplicates of each reaction were run, and this allowed two different definitions of a
positive result to be applied to the experimental data. The first definition, “singlets”, treated each
reaction tube as an independent entity. The second definition, “duplicates”, required that both
tubes containing the same reactants must be positive for the overall reaction to be considered
positive. The best sensitivity and specificity for HS-LAMP singlets (96.7% and 92.5%) and duplicates
(100% and 100%) was achieved at 27 and 28 minutes, respectively. The detection limit for
xiii | P a g e
template DNA amount was 5 pg. The sensitivity and specificity have not yet been determined
using clinical specimens in the field.
This thesis looked at HS in a funnel-shape pattern, from theoretical research to application. We
first conducted a survey to understand the husbandry conditions in Pakistan and identify the
variables that were statistically associated with the disease. Molecular epidemiology was then
used to compare the genetics of different field and vaccine strains. Consequently, unique genes
found only in HS-associated strains were identified. This allowed the design of a rapid and cheap
diagnostic test suitable for the limited infrastructure in developing countries like Pakistan. Two
important future directions of research are identifying potential virulence genes and validating the
LAMP test under field conditions to evaluate its performance.
xiv | P a g e
Conference Proceedings and Publications Arising from this Work
Conference Proceedings
“A retrospective case-control study of haemorrhagic septicaemia disease in Karachi, Pakistan in 2012” International Pasteurellaceae Conference (IPC), Prato, Italy, 2014 (poster). “Comparative genomics analysis of Asian HS-associated strains of Pasteurella multocida” International Pasteurellaceae Conference (IPC), Prato, Italy, 2014 (oral presentation). “Development of rapid LAMP diagnostic test specific for haemorrhagic septicaemia disease” International Pasteurellaceae Conference (IPC), Prato, Italy, 2014 (poster, winner).
Papers for Publication
Moustafa, A.M., Bennett, M.D., Edwards, J., Azim, K., Mesaik, M.A., Choudhary, M.I., Pathanasophon P., Worarach A., Ali Q., Abubakar M., Anjum, R. Molecular typing of haemorrhagic septicaemia-associated Pasteurella multocida isolates from Pakistan and Thailand using multilocus sequence typing and pulsed-field gel electrophoresis. 2013. Research in Veterinary Science, 95 (3), p. 986-90
In this paper, a comparative genetic study of 23 field isolates and vaccine strains of Pasteurella multocida associated with haemorrhagic septicaemia cases from Pakistan and Thailand using pulsed field gel electrophoresis (PFGE) and multilocus sequence typing (MLST) is described.
Moustafa, A.M., Ali, S.N., Bennett, M.D., Hyndman, T.H., Edwards, J., Robertson, I. A case-control study of haemorrhagic septicaemia in buffaloes and cattle in Karachi, Pakistan in 2012. Accepted
In this paper, a retrospective case-control study that was performed in Karachi, Pakistan from January to April 2013 is described. The aim of the study was to identify independent variables that were statistically associated with the presence of haemorrhagic septicaemia (HS) on the farm.
Moustafa, A.M., Seemann, T., Gladman, S., Adler, B., Harper, M., Boyce, J.D., Bennett, M.D. Comparative genomics analysis of Asian haemorrhagic septicaemia-associated strains of Pasteurella multocida identifies more than 90 haemorrhagic septicaemia-specific genes. 2015. PLoS ONE, 10(7): e0130296
In this paper, whole genome sequencing of 12 HS-associated strains is described. This is followed by the comparative genomics analysis of the isolates, identifying unique genes shared by HS-associated strains and finally, the phylogenetic relatedness of the isolates is presented.
Moustafa, A.M., Bennett, M.D. Development of loop-mediated isothermal amplification (LAMP)-based diagnostic tests for the specific detection of Pasteurella multocida and haemorrhagic septicaemia (HS)-associated Pasteurella multocida serovar B:2. In preparation
In this paper, the results of two rapid specific loop-mediated isothermal amplification (LAMP)-based diagnostic tests for the detection of P. multocida and HS-associated serovar B:2 are described.
xv | P a g e
List of Abbreviations
For conciseness, only abbreviations that appear in isolation (i.e. are not always defined) are
listed here.
ATCC American Type Culture Collection
ATTK Attock isolate
B3 outer backward primer
BIP backward inner primer
BLAST Basic Local Alignment Search Tool
BLASTN Basic Local Alignment Search Tool for Nucleotides
bp base pairs
Bst Bacillus stearothermophilus
BUKK Bhakkar isolate
C Celsius
CDS coding sequences
CFU colony forming unit
ddH2O double-distilled (PCR grade) water
DNA deoxyribonucleic acid
dNTPs an equal mixture of the four deoxynucleotides: deoxyadenosine triphosphate
(dATP), deoxycytidine triphosphate (dCTP), deoxyguanosine triphosphate
(dGTP) and deoxythymidine triphosphate (dTTP)
ELISA enzyme-linked immunosorbent assay
ERIC enterobacterial repetitive intergenic consensus
F3 outer forward primer
Faisal Faisalabad isolate
FAO food and agriculture organisation
FIP forward inner primer
FL Florida
FMD foot and mouth disease
Fwd forward
GC guanine and cytosine
GDP gross domestic product
xvi | P a g e
GMT geometric mean titre
HCl hydrochloride
Hep heptose
HS haemorrhagic septicaemia
ICE integrative conjugative element
ICEPmu1 integrative conjugative element of Pasteurella multocida strain 36950
ICSP International Committee on Systematics of Prokaryotes
IHA indirect haemagglutination
IM intramuscular
Inj. Injection
Islm Islamabad 1 isolate
IV intravenous
Karachi Karachi 3 isolate
kbp kilo base pairs
kDa kilo Daltons
Kdo phosphorylated 3-deoxy-D-manno-octulosonic acid
kg kilo grams
L litre
LAMP loop-mediated isothermal amplification
LB loop backward primer
LD50 median lethal dose
LDC Landhi dairy colony
Leu leucine amino acid
LF loop forward primer
LPS lipopolysaccharides
µL microlitre
µm micrometre
µmol micromole
m milli
mg milligram
mL millilitre
xvii | P a g e
mm millimetre
mmol millimole
Met methionine amino acid
min minute
Mbp mega base pairs
MgSO4 magnesium sulphate
MLEE multilocus enzyme electrophoresis
MLST multilocus sequence typing
NaCl sodium chloride
NCBI National Center for Biotechnology Information
ng nanograms
NGS next generation sequencing
OIE Office International des Epizooties
OMPs outer membrane proteins
OR odds ratio
P probability
PBS phosphate-buffered saline
PCR polymerase chain reaction
Pesh Peshawar isolate
PEtn phosphoethanolamine
PFGE pulsed field gel electrophoresis
pg picograms
PHAST PHAge Search Tool
Pm Pasteurella multocida
PMT Pasteurella multocida toxin
PVAcc Peshawar vaccine strain
qPCR quantitative polymerase chain reaction
REA restriction endonuclease analysis
REP repetitive extragenic palindromic
Rev reverse
RIRDC Rural Industries Research and Development Corporation
xviii | P a g e
RNA ribonucleic acid
ROC receiver operating characteristic
rRNA ribosomal ribonucleic acid
Ser serine amino acid
SNPs single nucleotide polymorphisms
ST sequence type
THA Thailand A isolate
THD Thailand D isolate
THF Thailand F isolate
tRNA transfer ribonucleic acid
TX1 Taxilla 1 isolate
UK United Kingdom
USA United States of America
USD United States Dollar
UV ultraviolet
V1 Lahore Vaccine strain
Ver. Version
WA Washington
WGS whole genome sequencing
xix | P a g e
List of Figures
Figure 1.1 Overview of the seven chapters presented in this thesis. ................................................................ 2
Figure 2.1 Association between world haemorrhagic septicaemia (HS) distribution in 2012 and world
average temperature and annual rainfall. (A) Average temperature (assembled from 1961-1990 data) (from
the food and agriculture organisation (FAO)). (B) Mean annual rainfall (FAO) (1961-1990). (C) Countries with
HS (January to June 2012) (from the World Animal Health Information Database (WAHID) of World
Organisation for Animal Health (OIE)). ............................................................................................................. 20
Figure 2.2 Age distribution of deaths in a haemorrhagic septicaemia outbreak in Sri Lanka, adapted from De
Alwis et al. (1976). ............................................................................................................................................ 24
Figure 2.3 Presumptive epidemiological cycle for haemorrhagic septicaemia (De Alwis, 1999e). .................. 30
Figure 2.4 Routine laboratory procedures for isolating and characterising Pasteurella multocida associated
with haemorrhagic septicaemia, adapted from De Alwis (1999c). .................................................................. 36
Figure 2.5 Schematic representation of the mechanism of a loop-mediated isothermal amplification (LAMP)
reaction. This process starts once the forward inner primer (FIP) has annealed to the DNA template.
However, DNA synthesis can also begin from the backward inner primer (BIP). The sequences (typically 23-24
nucleotides) inside both ends of the target region for amplification in a DNA are designated F2c and B2,
respectively. Two inner sequences (typically 23-24 nucleotides) 40 nucleotides from the ends of F2c and B2
are designated F1c and B1 and two sequences (typically 17-21 nucleotides) outside the ends of F2c and B2
are designated F3c and B3. FIP contains F1c, a TTTT linker and the sequence F2, which is complementary to
F2c. BIP contains the sequence B1c, which is complementary to B1, a TTTT linker and B2. The two outer
primers consist of B3 and the sequence F3, which is complementary to F3c, obtained from Notomi et al.
(2000). .............................................................................................................................................................. 46
Figure 2.6 Illustration of multilocus sequence typing process, adapted from Urwin and Maiden (2003). ...... 55
Figure 3.1 A map of Karachi city showing the six different locations in the study, adapted from the Karachi
metropolitan corporation (2014). .................................................................................................................... 65
Figure 3.2 Local roads are prone to seasonal flooding and this can hinder access to farms. .......................... 68
Figure 3.3 A road where a farm is located showing difficulty in access. .......................................................... 69
Figure 3.4 Typical day in holding facilities. ....................................................................................................... 70
Figure 3.5 Foot wear is rarely worn and bare feet are often unclean. ............................................................ 71
Figure 3.6 Dead calf, which was suspected to have died with haemorrhagic septicaemia, had been left with
other animals (red arrow). ............................................................................................................................... 72
Figure 3.7 Manual milking. ............................................................................................................................... 73
Figure 3.8 Flies in milk due to open-air storage. .............................................................................................. 74
Figure 5.1 Unrooted neighbour-joining tree showing the phylogenetic relationship between various strains.
(A) Relationship between Gallibacterium anatis, Mannheimia haemolytica, Pasteurella dagmatis, P. bettyae
xx | P a g e
and the P. multocida strains Pm70, 36950, HN06, P3480, X73, VP161, Anand1C, Anand1B, Anand1P,
Anand1G, P1059, P52Vac, VTCCBAA264, M1404 and the nine Pakistani and three Thai isolates. (B)
Relationship between all the P. multocida strains. (C) Relationship between the HS-associated P. multocida
B:2 strains M1404 and the twelve Pakistani and Thai isolates. Phylogenetic relatedness for all comparisons
was determined by analysis of only the single nucleotide polymorphisms (SNPs) found at conserved
positions in all genomes of the comparison set; 789 shared positions for the tree in panel A, 7,829 shared
positions for the tree in panel B and 722 shared positions for the tree in panel C. Trees were rendered using
SplitsTree v4.11.3 (Huson and Bryant, 2006). The line segments above the trees with the number '0.01'
indicate the length of branch that represents an amount genetic change of 0.01. ....................................... 107
Figure 5.2 Flower plot showing core and unique genes across all strains. The central circle shows the number
of genes common to all strains while the petals show the number of strain-specific genes, and in
parentheses the number of unique genes compared to all other strains in the analysis. The strain names are
given outside each petal and the strain details are given in Table 5.1. The B:2 strains are shaded in colour
while the non-B:2 strains are shaded in grey. Of the B:2 strains, M1404 is shaded in orange, the Thai strains
are shaded in blue and the Pakistani strains are shaded in green. ................................................................ 109
Figure 5.3 Comparison of the genomes of 3480, 36950, HN06, Pm70, M1404, BUKK, TX1, THA, THD, THF,
ATTK, Faisal, Islm, Pesh, PVAcc, V1 and Karachi with the genome of the PVAcc strain. The three inner rings
show the DNA size, GC content and GC skew of the reference genome (PVAcc strain). The 17 outer rings
show regions of the comparison genomes that match the reference genome PVAcc and in the order (inside
to outside) 3480, 36950, HN06, Pm70, M1404, BUKK, TX1, THA, THD, THF, ATTK, Faisal, Islm, Pesh, PVAcc, V1
and Karachi. Regions 1-4 on the outside identify particular regions of difference between the strains that are
potential prophage elements. The position of the capsule B locus is also noted between regions 2 and 3. This
figure was drawn using BLAST ring image generator ..................................................................................... 111
Figure 5.4 Comparison of the genomes of 36950, Pm70, combined 3480 and HN06, combined 11 B:2 strains
(M1404, ATTK, BUKK, Faisal, Karachi, Islm, Pesh, PVAcc, THA, THD and THF) and combined TX1 and BUKK
with the sequence of the integrative conjugative element present in 36950 strain (ICEPmu1) (Michael et al.,
2012a). The three inner rings show the DNA size, GC content and GC skew of the reference element
(ICEPmu1). The five outer rings show regions of the comparison genomes which match the reference
ICEPmu1. This figure was drawn using BLAST ring image generator (Alikhan et al., 2011). .......................... 112
Figure 5.5 Four putative prophages were identified by PHAST (Zhou et al., 2011): a questionable prophage
(green); an incomplete prophage (grey); and two intact prophages (pink). The questionable prophage was
present in seven Pakistani strains (ATTK, Faisal, Pesh, PVAcc, Islm, Karachi and V1) (region 1 in Figure 5.3),
the incomplete prophage was present in all Asian B:2 strains (region 2 in Figure 5.3), and the two intact
prophages were found in all B:2 strains (regions 3 and 4 in Figure 5.3). ....................................................... 114
xxi | P a g e
Figure 6.1 Specificity of HS-1517-PCR. Lane 1: 100 bp ladder (Axygen). Lane 2-17: genomic DNA of
Heddleston types 1-16 of Pasteurella multocida (10 ng), respectively. The 400 nucleotide band in lane 3
represents the B:2 serotype positive control. ................................................................................................ 130
Figure 6.2 Ladder pattern of positive Pm-LAMP reaction. Lane 1: 1 kb ladder (Axygen). Lanes 2-6: genomic
DNA (1 ng) of Heddleston types 1, 2, 3, 4 and 5 of Pasteurella multocida, respectively. Lane 7: water
(negative control). .......................................................................................................................................... 131
Figure 6.3 Ladder pattern of positive HS-LAMP reaction. Lane 1: 1 kb ladder (Axygen). Lanes 2-4: genomic
DNA of B:2 serotype of Pasteurella multocida (100 ng, 1 ng and 10 pg, respectively). Lane 5: water (negative
control). .......................................................................................................................................................... 132
Figure 6.4 Specificity of Pm-LAMP. Lane 1: 1 kb ladder (Axygen). Lane 2-17: genomic DNA of Heddleston
types 1-16 of Pasteurella multocida (10 ng), respectively. Lane 18: water (negative control). ..................... 133
Figure 6.5 Specificity of HS-LAMP. Lane 1: 1 kb ladder (Axygen). Lane 2-17: genomic DNA of Heddleston
types 1-16 of Pasteurella multocida (10 ng), respectively. Lane 18: water (negative control). Lane 19: DNA
extracted from bovine blood. Lane 20: DNA extracted from bovine bone marrow. ..................................... 133
Figure 6.6 Visualisation of positive results of Pm-LAMP by comparing the fluorescence of positives and
negatives in the real time reactions. Fluorescence is plotted on the y-axis, and reaction cycle number (equal
to minutes of reaction) is plotted on the x-axis. Black lines at fluorescences of approximately 5 units are
negative results. Red traces are positive results. ........................................................................................... 135
Figure 6.7 Visualisation of positive results under UV light by adding SYTO® 9 dye before running the reaction
(A) or adding SYBR® Safe once the reaction had completed (B). Left tube: positive result. Right tube:
negative result. ............................................................................................................................................... 135
Figure 6.8 (A) HS-LAMP Receiver Operating Characteristic (ROC) curve with at least 5 pg of target DNA per
reaction. (B) Pm-LAMP ROC curve with at least 10 pg of target DNA per reaction. Singlets and duplicates are
represented by green and blue lines, respectively. Points on each curve represent the incubation durations.
For Pm-LAMP, the incubation times were 20, 25, 27, 28, 29, 30, 31, 32, 33 and 34 minutes and for HS-LAMP,
the incubation times were 20, 25, 26, 27, 28, 29 and 30 minutes. Arrows point to the incubation time that
represents the highest sensitivity and specificity. .......................................................................................... 137
Chapter 4 Figure:
Fig. 1. PFGE results for fourteen isolates. Lane M: Lambda PFG marker. Lane 1: Thailand A. Lane 2: Thailand
B. Lane 3: Thailand C. Lane 4: Thailand D Lane 5: Thailand E Lane 6: Thailand F. Lane 7: Thailand G. Lane 8:
Thailand H. Lane 9: Thai HS vaccine strain. Lane 10: Attock. Lane 11: Karachi 3. Lane 12: Peshawar. Lane 13:
Peshawar vaccine. Lane 14: Lahore vaccine…………………….……………………………………………………………………. 92
xxii | P a g e
List of Tables
Table 2.1 The serotype designation for the African haemorrhagic septicaemia (HS) serotype, Asian
haemorrhagic septicaemia serotype and non-HS type B strain of Australia using the Carter-Heddleston and
Namioka-Carter systems, adapted from De Alwis (1993a). ............................................................................... 7
Table 2.2 Morbidity, mortality and case fatality rate due to haemorrhagic septicaemia in buffaloes and cattle
from ten randomly selected villages/towns from within the district of Malakand, Pakistan, obtained from
Khan et al. (2006). ............................................................................................................................................ 33
Table 2.3 Results of biochemical reactions for Pasteurella multocida, adapted from De Alwis (1999d). ........ 37
Table 2.4 Results of sugar fermentation patterns of Pasteurella multocida, adapted from De Alwis (1999d). 38
Table 2.5 Complete genomes of Pasteurella multocida and haemorrhagic septicaemia-associated draft
genomes. .......................................................................................................................................................... 56
Table 3.1 General results of the study. ............................................................................................................ 75
Table 3.2 Feeding pattern in the 100 controls and 66 cases farms. ................................................................. 75
Table 3.3 Common diseases in the 100 controls and 66 cases farms. ............................................................. 75
Table 3.4 Signs and course of disease seen in haemorrhagic septicaemia cases at 66 farms. ........................ 77
Table 3.5 Discrete risk factors for farms with (n=66) and without (n=100) haemorrhagic septicaemia cases in
2012 in Karachi, Pakistan. ................................................................................................................................. 79
Table 3.6 Continuous risk factors for 66 farms where haemorrhagic septicaemia cases were seen in 2012 in
Karachi, Pakistan. ............................................................................................................................................. 80
Table 3.7 Variables included in the final logistic regression model ................................................................. 80
Table 5.1 Haemorrhagic septicaemia-associated strains of Pasteurella multocida used in this study. ........... 99
Table 5.2 Sequencing and assembly statistics for the genomes of the 12 Asian HS-associated strains ........ 100
Table 5.3 Genomic features of the 12 Asian strains ...................................................................................... 103
Table 6.1 The 16 Heddleston types of Pasteurella multocida used in developing the diagnostic tests. ....... 122
Table 6.2 The nine bacterial species used for checking the specificity of Pasteurella multocida (Pm) LAMP
and haemorrhagic septicaemia (HS)-specific LAMP. ...................................................................................... 122
Table 6.3 Primer sequences used in developing the diagnostic tests. ........................................................... 123
Table 6.4 Eighteen DNA samples used for validation of Pm-LAMP and HS-LAMP. ........................................ 129
Table A.1 Unique genes shared by haemorrhagic septicaemia-associated strains of Pasteurella multocida
serovar B:2. ..................................................................................................................................................... 170
Table A.2 Unique genes of the TX1 strain of Pasteurella multocida serovar B:2. .......................................... 173
Table A.3 Unique genes of the Karachi strain of Pasteurella multocida serovar B:2. .................................... 174
Table A.4 The unique gene of the BUKK strain of Pasteurella multocida serovar B:2. .................................. 174
Table A.5 The unique gene of the PVAcc strain of Pasteurella multocida serovar B:2. ................................. 174
xxiii | P a g e
Chapter 4 Tables:
Table 1 Isolates used in the study………………………………………………………………………………………………….…………... 90
Table 2 Primer sequences for molecular identification of the bacterial isolates (multiplex PCR), Pasteurella multocida capsular typing PCR reactions and MLST scheme………………………………………………………………………. 91
xxiv | P a g e
Acknowledgement of Work Not Performed by the Author
The author acknowledges that Mr Syed Noman Ali conducted the interviews of the
epidemiological case-control study that was done in Karachi, Pakistan, mentioned in chapter 3.
The multivariable analysis for the epidemiological study, mentioned in chapter 3, was performed
by Prof. Ian Robertson. The pulsed field gel electrophoresis, mentioned in chapter 4, was done by
our collaborators at the National Institute of Animal Health (NIAH) at the Department of Livestock
Development in Bangkok, Thailand. Purified amplicons of the MLST technique, mentioned in
chapter 4, were sequenced at the DNA Core Facility at the Centre for Applied Molecular Biology
(CAMB), Ministry of Science and Technology, Lahore, Pakistan. Whole genome sequencing of the
12 HS-associated isolates, mentioned in chapter 5, was done at the Beijing Genomics Institute,
Beijing, China. The rest of the work presented in this thesis was performed by the author.
xxv | P a g e
Notes to Reader
The name of genera, species, and subspecies are generally printed in italics (or
underlined), but for higher categories, such as families and orders, conventions vary: in
Britain, they are written in ordinary roman type, but in America and France, they are
written in italics. The Bacteriological Code (1990 Revision) sets no binding standard in this
respect, as typography is a matter of editorial style and tradition, not of nomenclature.
However, according to Chapter 4 (Advisory Notes) of the Bacteriological Code (1990
Revision), scientific names of taxa should be preferably indicated by a different type face,
e.g., italic or by some other device to distinguish them from the rest of the text [1].
Consequently, the name of a taxon above the rank of genus, up to and including order, is
written in the entire thesis with an initial capital letter and the word is italicised.
Italics are not being used for the first three-letter acronym of the restriction enzymes’
names [2].
Chapter 4 is a publication. Its figures and tables are listed separately after the Lists of the
Figures and Tables. They do not follow the same numbering system as other figures and
tables. The bibliography for this chapter only is listed at page 86 and not at the end of the
thesis.
[1] Advisory Notes. In: Lapage SP, Sneath PHA, Lessel EF, Skerman VBD, Seeliger HPR, Clark WA (editors): International Code of Nomenclature of Bacteria (1990 Revision). Washington, DC: American Society for Microbiology, 1992, pp. 51-53. [2] Roberts RJ, Belfort M, Bestor T, Bhagwat AS, Bickle TA, Bitinaite J, Blumenthal RM, Degtyarev SKh, Dryden DT, Dybvig K, Firman K, Gromova ES, Gumport RI, Halford SE, Hattman S, Heitman J, Hornby DP, Janulaitis A, Jeltsch A, Josephsen J, Kiss A, Klaenhammer TR, Kobayashi I, Kong H, Krüger DH, Lacks S, Marinus MG, Miyahara M, Morgan RD, Murray NE, Nagaraja V, Piekarowicz A, Pingoud A, Raleigh E, Rao DN, Reich N, Repin VE, Selker EU, Shaw PC, Stein DC, Stoddard BL, Szybalski W, Trautner TA, Van Etten JL, Vitor JM, Wilson GG, Xu SY. A nomenclature for restriction enzymes, DNA methyltransferases, homing endonucleases and their genes. Nucleic Acids Research. 2003;13:1805–12.
1 | P a g e
Chapter 1 – Purpose of Project and Outline of Thesis
1.1 Purpose of Project
This thesis investigates an ongoing disease seen in cattle and buffaloes in Asia and discusses the
methods and results of an investigation into this disease. Haemorrhagic Septicaemia (HS) is a
significant disease caused by virulent strains of Pasteurella multocida and causes death and
production loss in ruminants in South Asia and other parts of the world. There is a need for further
work to characterise pathogenic strains of P. multocida in order to develop better diagnostic tests,
therapeutic methods and vaccines for HS.
The broad aims of this project are:
1. To observe husbandry systems and identify independent variables that are associated with
the presence of HS in farms in Karachi, Pakistan; and to collect and screen field samples
from HS suspected cases in Karachi so that the prevalence of HS can be determined in this
city.
2. To conduct molecular genetic studies that characterise the strains of P. multocida
associated with HS from South Asia and look for DNA sequences that could be used for the
development of a diagnostic test (or tests).
3. To produce a cheap, rapid, reliable and accurate point-of-care diagnostic test to diagnose
HS in clinical samples from buffaloes and cattle.
4. To apply this diagnostic test to the field in endemic countries to facilitate diagnosis of the
disease.
2 | P a g e
1.2 Outline of Thesis
Figure 1.1 Overview of the seven chapters presented in this thesis.
Chapter one
•Purpose of project and outline of thesis
Chapter two
•Literature review
•General information about Pasteurella multocida: Classification and diseases.
•Information about haemorrhagic septicaemia (HS) disease, its epidemiology, diagnosis, treatment, control and vaccines.
Chapter three
•A retrospective case-control study of haemorrhagic septicaemia in Karachi, Pakistan in 2012
•A study identifying risk factors associated with the presence of haemorrhagic septicaemia at farms in Karachi, Pakistan, is discussed.
Chapter four
•Molecular typing of haemorrhagic septicaemia-associated Pasteurella multocida isolates from Pakistan and Thailand using multilocus sequence typing and pulsed-field gel electrophoresis
•Comparative genetic analysis of 23 field strains of Pasteurella multocida associated with HS cases from Pakistan and Thailand was done using pulsed field gel electrophoresis (PFGE) and multilocus sequence typing (MLST).
Chapter five
•Comparative genomics analysis of Asian HS-associated strains of Pasteurella multocida
•Whole genome sequencing of 12 Asian HS isolates and identification of unique genes compared to non-HS genomes.
•Phylogenetic analysis.
•Identification of integrative conjugative element in two Pakistani isolates.
Chapter six
•Development of loop-mediated isothermal amplification (LAMP)-based diagnostic tests for the specific detection of Pasteurella multocida and haemorrhagic septicaemia (HS)-associated Pasteurella multocida serovar B:2
•Developing rapid diagnostic tests for detection of P. multocida and HS in clinical samples.
Chapter seven
•Thesis Summary and General Discussion
3 | P a g e
Chapter 2 – Literature Review
2.1 Introduction
Pasteurella multocida is a Gram-negative bacterium associated with a wide range of diseases in
both companion and agricultural species. The bacterium can cause disease as a primary pathogen
and one example is haemorrhagic septicaemia (HS): an acute and generally fatal septicaemic
disease which occurs mainly in cattle and buffaloes and is caused by B:2 and E:2 serovars (Carter
and De Alwis, 1989). For additional information on the diseases caused by the bacterium,
molecular biology, and diagnostic and typing options for investigating the diseases associated with
the bacterium, the interested reader is referred to a selection of available comprehensive reviews
(De Alwis, 1993a; De Alwis, 1999d; Dziva et al., 2008; Dziva and Christensen, 2011; Aktories et al.,
2012). This literature review will provide background information on the organism (2.2 -
Pasteurella multocida: Classification and Diseases); haemorrhagic septicaemia as a disease entity
(2.3 - Haemorrhagic Septicaemia); world haemorrhagic septicaemia epidemiology (2.4 -
Epidemiology of Haemorrhagic Septicaemia); its epidemiology in Pakistan (2.5 - Haemorrhagic
Septicaemia Epidemiology in Pakistan); the specific diagnostic tests that are available to diagnose
haemorrhagic septicaemia (2.6 - Diagnosis of Haemorrhagic Septicaemia); the treatment options
available (2.7 - Treatment) and lastly; preventative and control measures that can be taken (2.8 -
Prevention and Control).
2.2 Pasteurella multocida: Classification and Diseases
Trevisan (1887) named the genus Pasteurella to commemorate the eminent work by Louis Pasteur
on the aetiology of fowl cholera. Pasteurella multocida is a member of the family Pasteurellaceae
(Pohl, 1981; Garrity et al., 2005) within the order Pasteurellales (Garrity et al., 2005). Members of
Pasteurellaceae are Gram-negative bacteria which are facultatively anaerobic, fermentative and
chemo-organotrophic (Christensen and Bisgaard, 2008). As P. multocida can infect different animal
4 | P a g e
host species, names of species were originally assigned according to the affected species of
animal. Lignières (1900) introduced the names Pasteurella aviseptica, Pasteurella boviseptica, and
Pasteurella suiseptica to isolates from poultry, cattle and pigs, respectively. Later, Topley and
Wilson (1929) introduced a collective name Pasteurella septica to unite these various isolates from
different animal hosts into a single species. In 1939, Rosenbusch and Merchant replaced the taxon
P. septica with the currently-accepted scientific name, Pasteurella multocida (Rosenbusch and
Merchant, 1939).
The validly-published names of members of Pasteurellaceae are listed in the List of Prokaryotic
Names with Standing in Nomenclature (Euzéby, 1997), and a summary can be found from the
Pasteurellaceae Taxonomic Subcommittee of the International Committee on Systematics of
Prokaryotes (ICSP) (ICSP, 2012). In 2008, Pasteurellaceae contained thirteen genera:
Actinobacillus, Aggregatibacter, Avibacterium, Bibersteinia, Gallibacterium, Haemophilus,
Histophilus, Lonepinella, Mannheimia, Nicoletella, Pasteurella, Phocoenobacter and Volucribacter
(Christensen and Bisgaard, 2008). In 2009, Chelonobacter was introduced as a novel genus
(Gregersen et al., 2009). In 2010, Basfia was introduced (Kuhnert et al., 2010), and in 2011, two
novel genera, Bisgaardia and Necropsobacter, were also introduced (Christensen et al., 2011;
Foster et al., 2011). These 17 genera contained 41 properly classified species, as well as 28
misclassified species (ICSP, 2012). Eight of the misclassified species are now in the genus
Pasteurella (ICSP, 2012). In 2012, a new genus, Otariodibacter, with one new species,
Otariodibacter oris, was described (Hansen et al., 2012). And even more recently, in April, 2014,
two new genera, each with one species, were described: Vespertiliibacter pulmonis (Mühldorfer et
al., 2014) and Frederiksenia canicola (Korczak et al., 2014). At the time of this writing, the most
recent contribution to Pasteurellaceae taxonomy was the addition of two new genera in July 2014,
5 | P a g e
each with one species, were described: Mesocricetibacter intestinalis and Cricetibacter
osteomyelitidis (Christensen et al., 2014). Consequently, Pasteurellaceae now contains 22 genera.
The genus Pasteurella was reported to contain 13 species (Euzéby, 1997). Now, five of these
species are considered Pasteurella (Dziva and Christensen, 2011; Christensen et al., 2012) and
form the “sensu stricto” group of Pasteurella species, while the other eight await reclassification
(ICSP, 2012). The “sensu stricto” group of Pasteurella species are P. canis, P. dagmatis, P. stomatis,
P. multocida and P. oralis (previously known as the unnamed Pasteurella species B) (Dziva and
Christensen, 2011; Christensen et al., 2012). Pasteurella multocida contains three subspecies:
multocida, gallicida and septica (Euzéby, 1997; ICSP, 2012). The eight Pasteurella species which
await reclassification are P. aerogenes, P. bettyae, P. caballi, P. langaaensis, P. mairii, P.
pneumotropica, P. skyensis and P. testudinis (ICSP, 2012).
2.2.1 Morphology
Pasteurella multocida is a Gram-negative non-motile and non-sporogenous, short rod or
coccobacillus, 0.2-0.4 × 0.6-2.5 µm in size (Rimler and Rhoades, 1989). Pleomorphic, filamentous
and long rod forms may appear as a result of repeated subculturing from old cultures or growth
under unfavourable conditions (Rimler and Rhoades, 1989; De Alwis, 1999d). The typical
coccobacillary form can be demonstrated when the organism is obtained from tissues, exudates
and recently isolated cultures. Pasteurella multocida gives typical bipolar staining, especially with
Leishman, Giemsa or methylene blue stains (Rimler and Rhoades, 1989; De Alwis, 1999d).
2.2.2 Serological classification
Many attempts have been made to serologically classify P. multocida. The first successful attempt
was by Roberts who relied on passive protection tests in mice (Roberts, 1947). He identified four
types and designated them I, II, III and IV. All isolates of P. multocida that are associated with HS,
6 | P a g e
fell into Roberts type I (Roberts, 1947). In 1954, Hudson added serotype V (Rimler and Rhoades,
1989).
Carter was able to identify four serotypes designated A, B, D and E using the Carter indirect
haemagglutination test (IHA) in a number of trials (1952; 1955; 1961; 1963). Capsular type C was
deleted as it gave inconsistent results in the IHA tests (Carter, 1963). Rimler and Rhoades (1987)
isolated a fifth serogroup from turkeys, designated F.
A test based on identifying core bacterial components, known as somatic typing, was described by
Namioka and Murata (1961c; 1961a; 1964) and Namioka and Bruner (1963). This method
identified 11 somatic types (Namioka and Murata, 1961c; Namioka and Murata, 1961a; Namioka
and Bruner, 1963; Namioka and Murata, 1964). Currently, the most commonly used method for
somatic typing is agar gel precipitation and this method was developed by Heddleston et al.
(1972). Sixteen somatic types (1-16) have been identified by this technique (Heddleston et al.,
1972).
2.2.3 Designation of serotypes
There are two combinations for designation of P. multocida serotypes. Both of them use the
Carter capsular typing as one component (e.g. A, B, D, E and F) and the second component uses
either Heddleston somatic typing (e.g. 1-16) or Namioka somatic typing. The most commonly used
system is the Carter-Heddleston system (De Alwis, 1993a). The serotype designations for the
African serotype, Asian serotype and a non-HS type B strain from Australia are listed in Table 2.1.
7 | P a g e
Serotype Carter-Heddleston system Namioka-Carter system
African HS E:2 6:E Asian HS B:2 6:B
non-HS type B strain of Australia B:3,4 11:B Table 2.1 The serotype designation for the African haemorrhagic septicaemia (HS) serotype, Asian haemorrhagic septicaemia serotype and non-HS type B strain of Australia using the Carter-Heddleston and Namioka-Carter systems, adapted from De Alwis (1993a).
2.2.4 Cellular components
Capsule
The capsule is the outermost layer of the bacterial cell and is responsible for serogroup specificity
(Rimler and Rhoades, 1989). The crude capsule preparation includes polysaccharides,
lipopolysaccharides (LPS) and proteins. Incomplete separation of capsular constituents during
purification steps by different scientists resulted in conflicting results (Dhanda, 1960; Knox and
Bain, 1960; Bain and Knox, 1961; Rimler and Rhoades, 1989). The majority of
P. multocida serogroups have a polysaccharide capsule, similar to those commonly found on many
other bacterial species (Boyce et al., 2000b).
Non-toxic polysaccharide capsular antigen, separated by Penn and Nagy (1976), was compared
with crude capsular extracts and this polysaccharide was non-immunogenic in rabbits but in cattle,
a dose-dependent serological response was obtained as demonstrated by the mouse passive
protection test. Muniandy et al. (1993) used a modified method to prepare a purified extract of
3.5% polysaccharide that was free from LPS and contained only a minimum amount of
contaminating protein and nucleic acid. This extract was both non-toxic and non-immunogenic in
rabbits. In vitro antiphagocytic activity of the polysaccharide extracts of Penn and Nagy (1976) was
shown (Muniandy et al., 1993).
8 | P a g e
Endotoxins (LPS) of P. multocida are similar to those of other Gram-negative bacteria lacking
O-antigens, such as species of Neisseria and Haemophilus. Lipopolysaccharides play an important
role in the pathogenesis of HS (Rebers et al., 1967; Rhoades et al., 1967). Purified extracts of LPS
were shown in vitro to have antiphagocytic activity (Muniandy et al., 1993). The toxic effects of LPS
have been proven experimentally (Rebers et al., 1967; Rhoades et al., 1967). The virulence of LPS
depends on preventing phagocytosis. This allows increased multiplication of the organism (De
Alwis, 1999f). Interestingly, it was shown by Muniandy and Mukkur (1993) that mice treated with
purified LPS of capsular type B did not develop pasteurellosis when challenged with the bacterium.
This observed protection was abolished if the LPS extract was treated with phenol or digested with
Proteinase K and further purified by gel filtration prior to inoculation. Further, the protein
associated with LPS that is probably responsible for the observed protection, appeared to be of
outer membrane origin (Muniandy and Mukkur, 1993).
The immunogenic and protective role of outer membrane proteins (OMPs) has been hypothesised
(Muniandy and Mukkur, 1993). High homogeneity in protein profiles of 14 strains of different P.
multocida serogroups associated with HS, was illustrated using electrophoretic techniques
(Johnson et al., 1991). B:2 strains of Asian and North American origin, gave a major protein band
of 32 kDa. A corresponding band for E:2 strains occurred at 37 kDa. Other bands, shared by all
serotypes, occurred at 27, 45 and 47 kDa.
The OMPs expressed on P. multocida differ according to the environmental iron content. During in
vivo and in vitro growth under iron-restricted conditions, B:2 strains of P. multocida elaborated a
protein of high molecular mass (116 kDa). This protein was absent in the same strains cultured in
non-iron restricted media (De Alwis, 1999d). A similar protein of 84 kDa was expressed in
9 | P a g e
abundance in strains of B:2 (Malaysian C82) P. multocida when cultured under iron-restricted
conditions and was not expressed under iron-replete conditions (Kennett, 1993).
Enzymes
Pasteurella multocida strains of capsular types A, B, D and E produce neuraminidase (Rimler and
Rhoades, 1989). There is evidence that HS-associated B:2 strains also produce hyaluronidase and
chondroitinase (Carter and Chengappa, 1980; Rimler, 1993; Rimler and Rhoades, 1994) but other
somatic types of serogroup B, not associated with HS, such as B:3 and 4, do not produce
hyaluronidase. However, mutant strains of B:2, which have low virulence in mice and rabbits, and
are avirulent to cattle and buffaloes, can produce hyaluronidase (De Alwis et al., 1996; De Alwis,
1999d). Consequently, there is no absolute relation between production of hyaluronidase and
virulence.
Toxins and bacteriocins
Protein exotoxins are produced by P. multocida serogroups A and D (Frandsen et al., 1991; Orth
and Aktories, 2012). The dermonecrotic Pasteurella multocida toxin, PMT, which causes atrophic
rhinitis in pigs, leads to atrophy of nasal turbinates and prevents bone regeneration (Sterner-Kock
et al., 1995; Mullan and Lax, 1998). Strains of P. multocida serogroup B associated with HS do not
produce exotoxins (Rimler and Rhoades, 1989).
Bacteriocins are proteins produced by many bacterial species and are bactericidal against
members of their own species or closely related species. A study of 33 strains of P. multocida from
serogroups A, B and D, of bovine and bison origin, showed the presence of a bacteriocin
(Chengappa and Carter, 1977). It was found that 14 of these 33 strains were producers and 17
strains were susceptible (Chengappa and Carter, 1977). De Alwis (1999d) stated that only a few
10 | P a g e
studies have been made on the role of bacteriocin in the pathogenesis of HS, but these studies
were not specifically referenced.
2.2.5 Preferred media and colony morphology
Most laboratory media (e.g. nutrient agar) are adequate for the growth of P. multocida (Woolcock,
1993; De Alwis, 1999d). However, media like dextrose-starch agar, blood agar and casein-sucrose-
yeast (CSY) result in good growth (Rimler and Rhoades, 1989; De Alwis, 1999d). The optimum
growth temperature is 35-37 °C (Rimler and Rhoades, 1989).
The resultant colonies after 18-24 hours of incubation at 35-37 °C are 1-3 mm in diameter when
enriched media such as blood agar and brain heart infusion agar are used (Rimler and Rhoades,
1989). Normal cultures on agar media may develop into either one or both types of the two
principal colony forms: mucoid and smooth. The mucoid colonies are the largest and are
composed of capsulated cells, while the smooth colonies are composed of capsulated or non-
capsulated cells (Rimler and Rhoades, 1989). Rough colonies are sometimes encountered. They
consist of filamentous non-capsulated cells. The rough colonies are flat with irregularly serrated
edges and are slightly dry (Rimler and Rhoades, 1989).
Colonies consisting of capsulated cells are yellowish-green, bluish-green or have pearl-like
iridescence. On the other hand, colonies consisting of non-capsulated cells are non-iridescent and
appear blue, bluish-grey or grey. Intermediates occur between the iridescent and non-iridescent
forms (Rimler and Rhoades, 1989). Special lighting setups, including a stereomicroscope
illuminated by obliquely transmitted light, and transparent agar medium are required to see
colony iridescence (Heddleston, 1972).
11 | P a g e
Old cultures, or cultures stored in stock culture media or in lyophilised form, can lose their
capsular material in a process known as dissociation which is manifested by a change from
iridescent to non-iridescent colonies (Rimler and Rhoades, 1989; De Alwis, 1999d). This results in
rough colonies and the loss of virulence and antigenicity. Non-iridescent and non-capsulated
colonies can be reverted to their capsulated and iridescent form by passage in the natural host or
laboratory animals (Rimler and Rhoades, 1989).
The colony iridescence differs according to the capsular type. Colonies of capsular serogroups A, D
and F display pearl-like iridescence while groups B and E produce yellowish-green or bluish-green
iridescence in oblique transmitted light (Rimler and Rhoades, 1989). For groups B and E, the
colonies are smooth in texture and smaller.
2.2.6 Virulence in experimental animals
The virulence of isolates of P. multocida varies dramatically between laboratory animal species.
Strains of B:2 associated with HS are highly virulent to mice and rabbits: the LD50 is between 1-10
viable organisms following intraperitoneal inoculation in these hosts (De Alwis, 1999d). However,
in guinea-pigs, consistent killing was not achieved even with intraperitoneal administration of 0.1
mL of undiluted broth culture (De Alwis, 1999d).
2.2.7 Diseases caused by Pasteurella multocida
Bovine pasteurellosis
Bovine pasteurellosis is also sometimes called shipping fever, transit fever or bovine respiratory
disease. The term shipping fever had previously been mistakenly applied to HS (Frank, 1989).
Unlike HS, P. multocida here plays a secondary and opportunistic role after a respiratory viral
infection (e.g. bovine respiratory syncytial virus, bovine viral diarrhoea virus, parainfluenza-3 virus
(PI3V) and infectious bovine rhinotracheitis (IBR)) and/or stress (Frank, 1989). Mannheimia
12 | P a g e
haemolytica (previously known as P. haemolytica biovar A) and to a lesser extent
Histophilus somni and serotype A of P. multocida have also been associated with the disease (Kahn
and Line, 2010).
Other animals
Three different pasteurelloses in pigs (Sus scrofa) are defined according to the infecting serogroup.
Sporadic outbreaks of HS have occurred in pigs, associated with P. multocida B:2 strains.
Moreover, atrophic rhinitis has occurred in association with toxigenic strains of serogroup D and
occasionally A. In addition, pneumonia has been associated with P. multocida serogroup A (De
Alwis, 1999d).
Fowl cholera is one of the most economically important diseases in the poultry industry, especially
in developed countries and large-scale operations (De Alwis, 1999d). Fowl cholera is caused by P.
multocida serogroup A, most commonly A:1, A:3 and A:4 (De Alwis, 1999d).
Rhinitis “snuffles” is one of the most common diseases that affects domestic rabbits (Oryctolagus
cuniculus). It is a general term describing a group of upper respiratory signs. The most common
and generally accepted cause of rhinitis is infection with P. multocida (Manning et al., 1989).
Sporadic outbreaks have occurred in other host species such as deer (Dama dama), cats (Felis
catus), dogs (Canis lupus familiaris), horses (Equus ferus caballus) and mink (Neovison vison)
(Carter, 1959). Elephants (Elephas maximus) in Sri Lanka, bison (Bison bison) in USA, camels
(Camelus dromedarius) in Sudan and a snow leopard (Panthera uncia) in the Himalayas have been
reported to have pasteurelloses (Carter, 1957; De Alwis and Thambithurai, 1965; Bain et al., 1982;
De Alwis, 1982a; Wickremasuriya and Kendaragama, 1982; Chaudhuri et al., 1992). Various
serogroups had been associated with the disease in these cases. The diseases ranged from
13 | P a g e
septicaemia and respiratory infections most commonly, to wound infections, abscesses, mastitis,
peritonitis and encephalitis. Human infection usually occurs as a result of animal bites (De Alwis,
1993a).
2.3 Haemorrhagic Septicaemia
Haemorrhagic septicaemia is one of the most economically important pasteurelloses (Carter and
De Alwis, 1989; De Alwis, 1999d). Haemorrhagic septicaemia in cattle and buffaloes was previously
known to be associated with one of two serotypes of P. multocida: Asian B:2 and African E:2
according to the Carter-Heddleston system, or 6:B and 6:E using the Namioka-Carter system
(Carter and De Alwis, 1989).
The disease occurs mainly in cattle and buffaloes (Carter and De Alwis, 1989). However, some
cases of infection in other animals have been reported. Sporadic cases in goats (Capra aegagrus
hircus) have been reported in Malaysia and India (FAO, 1959; FAO, 1979; FAO, 1991). Kasali (1972)
presented evidence for the disease in the African buffalo (Syncerus nanus). In Sudan, HS was
reported in camels (Bain et al., 1982). However, it is interesting to note that Awad et al. (1976)
reported that camels are resistant to experimental infection with doses that are lethal to buffaloes.
Pigs have been infected by serogroup B in Malaysia and India (FAO, 1959; Murthy and Kaushik,
1965; Pillai et al., 1986). In Sri Lanka, some cases of HS have been reported in wild elephants
(Elephas maximus) during simultaneous outbreaks among cattle and buffaloes in the same locality
(De Alwis and Thambithurai, 1965; De Alwis, 1982a; Wickremasuriya and Kendaragama, 1982).
Serotypes B:1 and B:3,4 have been responsible for a septicaemic disease in the USA in antelope
(Antilocapra americana) and elk (Cervus canadensis), respectively. Serotype B:4 was reported in
Canada to be associated with a septicaemic disease in bison (Bison bison) (Rimler, 1993). Cases in
horses and donkeys (Equus africanus asinus) have been reported in India (Pavri and Apte, 1967).
14 | P a g e
Serotypes E:2 and B:2 were formerly associated with HS outbreaks in Africa and Asia respectively.
Serotype E:2 was reported in Senegal, Mali, Guinea, Ivory Coast, Nigeria, Cameroon, Central
African Republic and Zambia (Francis et al., 1980; Carter and De Alwis, 1989). However, it is now
inaccurate to associate outbreaks in Africa with serotype E:2 as many outbreaks of HS in Africa
have now been associated with serogroup B (Dziva et al., 2008). In the same manner, serogroup E
has been associated with outbreaks in Asia (Dziva et al., 2008). For instance, one record of “Asian
serotype” (B:2) was reported in Cameroon (Martrenchar, 1993). Some reports showed that
serotype B:2 may be present in some East African countries (De Alwis, 1999e). Both serogroups B
and E have been reported in Egypt and Sudan (Shigidi and Mustafa, 1979).
In North America, Rimler and Wilson (1994) re-examined isolates which were previously identified
as B:2 using Carter-Heddleston serotyping and DNA fingerprinting methods. They found that a
bison isolate from Montana (1965) and a dairy calf isolate (1969) were both B:3,4, not B:2 as
previously described. In the same study, beef calf isolates from California (1993) were confirmed as
B:2 (Rimler and Wilson, 1994).
Subcutaneous inoculations of serotypes B:2 and E:2 were able to produce consistent and
predictable HS in domestic buffaloes and cattle (De Alwis, 1999f). Serotypes B:1, B:3,4 and B:4
have occasionally been associated with the disease. B:3,4 failed to produce consistent disease even
following experimental transmission (De Alwis, 1999f). B:1 and B:4 serotypes caused occasional
sporadic outbreaks of a disease which was indistinguishable from classical HS. These outbreaks
occurred mainly in wild animals and to a lesser extent in domestic cattle (De Alwis, 1999f).
Outside the animal, P. multocida cannot be recovered from dry soil for more than a few hours
unless it is in moist conditions (Carter and De Alwis, 1989). Therefore, carcasses that are thrown
15 | P a g e
into rivers and other waterways represent an important route of HS dissemination. The survival of
P. multocida in animal tissues and decomposing carcasses in waterways is believed to be for a few
days (Carter and De Alwis, 1989). Additionally, freshly infected pasture and bedding can be a
source of infection. No permanent reservoirs of infection have been detected outside the animal
body (De Alwis, 1999f). When clinically-affected or carrier animals were introduced to a flock,
outbreaks sometimes started. A major source of infection, once an outbreak had started, was
incorrectly buried or burnt carcasses (De Alwis, 1999f).
Natural routes of infection are inhalation and/or ingestion (Carter and De Alwis, 1989).
Experimental transmission has succeeded using intranasal aerosol spray or oral drenching (Carter
and De Alwis, 1989). When subcutaneous inoculation is used experimentally, it results in rapid
onset of the disease, a shorter clinical course and less marked pathological lesions compared to
the longer course of disease and more profound lesions of oral drenching and the intranasal
infection by aerosols (De Alwis, 1999f).
During experimental infection trials, approximately 107- 1012 colony forming units (CFU) of
P. multocida organisms needed to be inoculated into animals such that they would transmit the
disease to exposed but uninoculated buffaloes (De Alwis et al., 1990). Inoculation volumes that
were less than this did not result in disease transmission to in-contact animals. How the bacteria
are transmitted naturally from an active carrier to in-contact animals is not known. It is likely that
the virulence of the organisms transmitted from a carrier is higher than that from cultures grown in
vitro for experimental transmission. Moreover, certain conditions, such as age, concurrent
infection, stress and pregnancy, could alter the susceptibility of animals to be infected by smaller
doses. In a separate study, no difference in virulence (quantified by median lethal dose (LD50) for
mice) was noticed between isolates from clinical cases and from latent carriers (Wijewardana et
16 | P a g e
al., 1986a). Lethality to mice could be helpful in laboratory diagnosis of P. multocida. Following the
subcutaneous injection of 0.1-0.2 mL of a saline suspension of suspected sample, if it is P.
multocida, the mouse will die within 24 hours (Carter and De Alwis, 1989).
Where husbandry is very basic and rearing conditions are free-range, observations of animals are
usually limited and so the only reported sign of an HS outbreak is sometimes only sudden death.
Consequently, under these conditions, the occurrence of HS is likely to be higher than officially
reported.
2.3.1 Clinical signs
A wide variety of clinical signs have been described for HS in cattle and buffaloes (Carter and De
Alwis, 1989). In Sri Lanka, some descriptions of the disease were derived from experimental
transmission studies (Horadagoda et al., 1991; De Alwis, 1999f). The incubation periods (the time
between exposure and observable disease) for buffalo calves 4-10 months of age varied according
to the route of infection (De Alwis, 1999f). The incubation period was 12-14 hours, approximately
30 hours and 46-80 hours for subcutaneous infection, oral infection and natural exposure,
respectively.
There is also variability in the duration of the clinical course of the disease. In the case of
experimental subcutaneous infection, the clinical course lasted only a few hours, while it persisted
for 2-5 days following oral infection and in buffaloes and cattle that had been exposed to naturally-
infected animals (De Alwis, 1999f). It has also been recorded from field observations of five HS
outbreaks in Malaysia that the clinical courses of per-acute and acute cases were 4-12 hours and 2-
3 days, respectively (Saharee and Salim, 1991).
Generally, progression of the disease in buffaloes and cattle is divided into three phases. Phase one
17 | P a g e
is characterised by temperature elevation (rectal temperature 40-41 °C), loss of appetite and
depression. Phase two is typified by increased respiration rate (40-50/minute), laboured breathing,
clear nasal discharge (turns opaque and mucopurulent as the disease progresses), salivation and
submandibular oedema spreading to the pectoral (brisket) region and even to the forelegs. Finally,
in phase three, there is typically recumbency, continued acute respiratory distress and terminal
septicaemia (Horadagoda et al., 1991). The three phases overlap when the course is short. In
general, buffaloes have a more acute onset of disease than cattle, and furthermore, the disease
typically has a shorter duration (Graydon et al., 1993).
Some atypical syndromes associated with the same HS-associated serotypes have been reported.
In Sri Lanka, an outbreak of pneumonia in buffalo calves (four-ten months of age) with terminal
septicaemia and a longer course of disease than usual (up to ten days) was reported (De Alwis et
al., 1975). Occurrence of paraplegia in 189 cattle in India after a vaccination campaign against
rinderpest was reported (Dhanda and Nilakanthan, 1961). Pasteurella multocida was isolated from
135 dead animals. They were typed as Roberts type I which includes serotypes B:2, E:2 and B:3,4.
It is therefore unclear if this atypical syndrome was caused by the B:2 serovar that is commonly
associated with HS in Asia. There was no terminal septicaemia, which is a characteristic of HS
infection by B:2 strains (Dhanda and Nilakanthan, 1961).
2.3.2 Pathology and pathogenesis
When investigating the carcass of an animal that died from HS, the most obvious gross lesion is
subcutaneous oedema in the submandibular and pectoral (brisket) regions (De Alwis, 1999f).
Petechial haemorrhages are found subcutaneously and in the thoracic cavity. In addition,
congestion and various degrees of consolidation of the lung may occur (De Alwis, 1999f). An
experimental transmission study showed that the duration of the illness affected the observed
pathological findings (De Alwis et al., 1975). It was shown that animals that died within 24-36
18 | P a g e
hours, had only few petechial haemorrhages on the heart and generalised congestion of the lung,
while in animals that died after 72 hours, petechial and ecchymotic haemorrhages were more
evident and lung consolidation was more extensive (De Alwis et al., 1975).
The histopathological lesions seen in a calf infected experimentally with a bison B:2 strain, included
mild haemorrhages involving the peritracheal adventitia and submucosa, diffuse interstitial
pneumonia, hyperaemia, oedema, and increased inflammatory cells, including macrophages and
lymphocytes, in thickened alveolar septa (Rhoades et al., 1967). There were strong similarities in
the histopathology observed in animals infected experimentally by the intranasal route and
naturally-infected animals (Horadagoda et al., 1991). This similarity makes biological sense as the
tonsils are thought to be the first sites of multiplication in an animal infected with P. multocida.
When an infection occurs, a conflict between two factors starts. The virulence of the organism and
its rate of multiplication in vivo compete with the animal’s immune system. There are two possible
outcomes resulting from this interaction. First, if the bacterial virulence and multiplication rate
dominate, clinical disease occurs which may lead to death. Second, if the host immune system
prevails, the infection is arrested and recovery ensues (De Alwis et al., 1986; De Alwis, 1999f).
2.4 Epidemiology of Haemorrhagic Septicaemia
2.4.1 Distribution of the disease
Three factors affect the global distribution of HS: climatic conditions, husbandry practices and the
species of animal (De Alwis, 1993b). For example, in 1981, Sri Lanka was a good example of
different distribution patterns because it had a variety of agroclimatic regions and different
husbandry practices. Consequently, Sri Lanka had distinct endemic and non-endemic areas for HS
(Carter and De Alwis, 1989). The disease was almost non-existent where there was a
predominance of hills. Here, the climatic conditions were mild and also temperate dairy breeds
19 | P a g e
were reared. In contrast to this, in the warmer dry plains, where there were seasonal heavy rains
and indigenous cattle, buffaloes and zebu cattle, the disease was endemic. Occasional sporadic
outbreaks happened in areas with topography, climate and animals that were between these
extremes (Carter and De Alwis, 1989).
Seasonal changes have affected outbreaks of HS. Haemorrhagic septicaemia has been seen more
frequently during wet (high precipitation) and humid weather. Outbreaks of HS have been
associated with heavy rains in Sri Lanka (Dassanayake, 1957; Perumalpillai and Thambiayah, 1957)
and Zambia (Francis et al., 1980) and also rainy seasons in Sudan (Mustafa et al., 1978). In India,
the disease had been reported to be more frequent in states with higher annual mean rainfall
(Dutta et al., 1990; Saini et al., 1991). Maps showing annual rainfall and average temperature
(assembled from 1961-1990 data) and countries having HS (January to June 2012), are shown in
Figure 2.1.
20 | P a g e
Figure 2.1 Association between world haemorrhagic septicaemia (HS) distribution in 2012 and world average temperature and annual rainfall. (A) Average temperature (assembled from 1961-1990 data) (from the food and agriculture organisation (FAO)). (B) Mean annual rainfall (FAO) (1961-1990). (C) Countries with HS (January to June 2012) (from the World Animal Health Information Database (WAHID) of World Organisation for Animal Health (OIE)).
21 | P a g e
Although there are associations between HS outbreaks and the humidity and precipitation, there
are also other reports where these associations were not observed. An active surveillance study in
Sri Lanka from 1978-1980 showed that disease outbreaks occurred throughout the year,
independent of rainfall (De Alwis and Vipulasiri, 1980; De Alwis, 1981). However, according to
these studies, outbreaks which occurred in the dry seasons were milder. In contrast, wet season
outbreaks were more likely to spread extensively (Carter and De Alwis, 1989). During these
surveillance studies it was found that 15% and 8% of buffalo and cattle deaths, respectively, were
attributed to HS annually in the HS endemic areas (De Alwis and Vipulasiri, 1980; Benkirane and De
Alwis, 2002). The passive reporting systems (routine reporting of the cases reaching health care
facilities for treatment) in Sri Lanka during the same period, recorded 1200-1500 deaths a year in a
national population of 2.5 million cattle and buffalo (De Alwis and Vipulasiri, 1980). The higher rate
of outbreaks in wet seasons was attributed to the longer survival of P. multocida outside the
animal body (Carter and De Alwis, 1989). Due to this method of reporting, these reported numbers
are likely under-estimates.
In India from 1974-1986, HS was responsible for the highest mortality rate of infectious diseases in
buffaloes and cattle, and was second in its morbidity rate in the same animals. When compared to
foot and mouth disease, rinderpest, anthrax and black leg (Dutta et al., 1990), HS accounted for
58.7% of the deaths due to these five endemic diseases (Dutta et al., 1990; Benkirane and De
Alwis, 2002). In India, HS has been the second most reported disease during the two decades from
1991 to 2010, as per the National Animal Diseases Referral Expert System (NADRES). The disease
had been reported in 25 (out of 29) Indian states (Rahman et al., 2012). It has emerged as the most
economically important bacterial disease that is prevalent throughout India and is responsible for
approximately 60% of all bovine mortalities (Rahman et al., 2012). There has been a gradual
22 | P a g e
decline in the number of HS outbreaks reported since 1998 which could be attributed to increased
vaccination rates (Rahman et al., 2012).
2.4.2 Morbidity, mortality and case fatality
Sporadic HS outbreaks in non-endemic areas, or when HS was introduced for the first time into a
geographic area, morbidity and mortality rates were high (De Alwis and Vipulasiri, 1980). In fact,
when the clinical disease first appeared and clinical signs started, the case fatality rate approached
100% unless animals were treated in the very early stages of disease e.g. once an animal that is in
contact with an HS case shows an increased rectal temperature (De Alwis, 1999e).
In India, between the years 1949 and 1987, the rate of recovery from HS was reported at 10-50%
(De Alwis, 1999e). In Nepal and Philippines, recovery rates have been reported as being 60-90%
and over 90%, respectively (FAO, 1979). The reason for this wide range and the higher recovery
rates seen in Nepal and Philippines is due to the reporting systems of these countries which report
the total number at risk and not the actual number of affected cases.
Sheikh et al. (1996) reported in a study, based on the results from nine districts of Punjab province
in Pakistan, that the morbidity and mortality rates were 11% and 9% in buffaloes and 4% and 2.5%
in cattle, respectively. The case fatality rate (number of deaths measured against the number of
animals that were clinically-affected) was 78% and 62% in buffaloes and cattle, respectively.
Lower mortality rates were observed in Sabah, Malaysia. A total of 23 outbreaks were recorded for
the period 1983-91 in a population of 65,000 buffaloes. The mortality rate was 1.4% (935 buffaloes
died) and the case fatality rate was 100% (Yeo and Mokhtar, 1993).
23 | P a g e
Factors affecting morbidity and mortality
Host susceptibility Cattle and buffaloes are the main host species and numerous authors have suggested that
buffaloes are more susceptible to HS than cattle (FAO, 1979; De Alwis and Vipulasiri, 1980; De
Alwis, 1981; FAO, 1991; Ramarao et al., 1991; De Alwis, 1999e; Khan et al., 2006; Farooq et al.,
2011). In Malaysia, 73% and 90% of all deaths due to HS during the periods 1970-79 and 1980-89,
respectively, were among buffaloes, even though the buffalo population was half that of cattle
(FAO, 1979; FAO, 1991). Both Khan et al. (2006) and Farooq et al. (2011) reported that the average
geometric mean anti-P. multocida antibody titre (GMT) of affected buffaloes in Malakand and
Southern Punjab in Pakistan, respectively, was lower than that of affected cattle which they
inferred made buffaloes more susceptible to the disease than cattle. However differences in host
susceptibility were not proven statistically. These results are consistent with the findings of De
Alwis et al. (1986) who reported higher GMTs for cattle compared to those for buffaloes. In India,
a higher incidence of HS in districts where the buffalo population was higher was recorded
(Ramarao et al., 1991).
Age
It was observed in Asia that young animals were more susceptible to HS than adults (De Alwis and
Vipulasiri, 1980; De Alwis, 1999e). In Sri Lanka, a study showed that 77% and 65% of HS buffalo
deaths and cattle deaths, respectively, were among animals under two years old (De Alwis,
1999e). Buffaloes and cattle under two years old represented 23% and 20% of the population in Sri
Lanka, respectively.
In another study, it was reported that the most commonly affected age group was six months to
two years (De Alwis et al., 1976). From this study, the age distribution of deaths in one Sri Lankan
farm is shown in Figure 2.2. It was concluded that the disease rarely occurred in very young calves
24 | P a g e
(under six months) and this could be because of the protective effect of maternal antibodies
and/or the oil adjuvant vaccine that was routinely given at four months of age. It was postulated in
the study that the percentage of deaths among animals over two years was negligible and this was
because of repeated vaccination. Unfortunately, these data cannot be critiqued any further
because important information was not provided. For example, the total number of animals in
each age group (affected and unaffected) was not provided so it is not known whether the
distribution of affected animals simply mirrored the distribution of all animals on this farm (De
Alwis et al., 1976). However, assuming typical age ranges on Sri Lankan cattle farms, it would seem
that the authors’ conclusion (that younger animals are more commonly affected) was reasonable.
Figure 2.2 Age distribution of deaths in a haemorrhagic septicaemia outbreak in Sri Lanka, adapted from De Alwis et al. (1976).
A study was held in Pakistan on thirty randomly-selected Sahiwal cows and their calves (Mahmood
et al., 2007). These cows were vaccinated against HS using alum-precipitated formalin-killed HS
bacterin. The sera of the thirty cows and their calves were tested for anti-P. multocida antibodies
using an indirect haemagglutination test (IHA). The sera from the dams had an antibody titre
0.70%
21%
41%
19.20%
0.70% 0% 0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
0-6 months 6-12 months 12-18 months 18-24 months 2-4 years >4 years
De
ath
s
Age group
Age distribution of deaths in an HS outbreak
25 | P a g e
(GMT) of 213 at 15 days before parturition; however, it was 182 days just after parturition. This
drop in antibody titre might have been due to the transfer of serum antibodies to the udder
(Moraes et al., 1997).
The antibody titres (GMT) of the calves’ sera examined at birth; just after birth but before
colostrum feeding; and six hours, three days, 15 days, 30 days, 45 days and 60 days after
colostrum feeding were 44, 69, 70, 97, 96, 72 and zero respectively. These results indicate that the
duration of maternal immunity for the newly born calves was approximately 45 days.
Consequently, in an endemic area, vaccination must start before the end of the first sixty days, i.e.
before the GMT of the calves falls to zero (Mahmood et al., 2007). Interference could happen
when vaccinating calves too early (within 45 days) while the maternally-derived antibody titre is
still high. From these data, it would appear that the ideal time to vaccinate calves would be after
day 45 but before the drop in the maternally derived antibody titre in day 60. In contrast, there
was a reported death of a young calf (two months age) due to HS in this Sri Lankan study. This calf
was born to a cow imported from an HS-free country (De Alwis, 1999e). In conclusion, the most
affected age group is 6-24 months and calves from cows from HS-free countries should be
vaccinated once born.
Immunological factors
The immunity of the herd influences the morbidity and mortality rates. The immunity of a herd
depends on the naturally- and artificially-acquired immunity within the herd, which in turn,
depends on the number of outbreaks that happen in that area, and the efficacy of vaccination (De
Alwis, 1999e). Moreover, immunity will depend on a number of other factors such as concurrent
diseases and nutrition.
26 | P a g e
Endemic or non-endemic areas
Haemorrhagic septicaemia-associated morbidity and mortality in non-endemic areas is higher than
that in endemic areas. This phenomenon is related to herd immunity (De Alwis, 1999e). The
percentage of animals with naturally-acquired immunity in endemic areas is significantly higher
than that in non-endemic areas (De Alwis and Sumanadasa, 1982). This was attributed to direct
exposure to the disease and was confirmed when a sharp increase in the antibody titres was found
among surviving animals after an outbreak (De Alwis, 1982b; De Alwis et al., 1986). Morbidity was
low when regular seasonal outbreaks occurred in endemic areas (De Alwis, 1981). Morbidity was
restricted only to young adults and older calves that were born after the last annual vaccination
program (De Alwis, 1981).
Husbandry methods
In contrast to large, nomadic, free-roaming herds, where HS commonly existed, it was generally
rare in smaller, well-managed, stall-fed herds (De Alwis and Vipulasiri, 1980; De Alwis, 1999e). In
Sri Lanka, it was observed that the incidence of HS in herds with over 50 animals was 4-5 times
greater than that of herds with less than ten animals (De Alwis and Vipulasiri, 1980). The number
of animals in a herd was likely to be merely a surrogate marker for the suboptimal, nomadic
husbandry practices found in herds with over fifty animals. Those herds with fewer than ten
animals were, in most cases, better managed (De Alwis and Vipulasiri, 1980; De Alwis, 1999e).
2.4.3 Carrier status
The presence of P. multocida in the nasopharynx of healthy cattle and buffaloes has been well
reported (Singh, 1948; Mohan et al., 1968). It was believed that the percentage of healthy carriers
of P. multocida was only 1-2% in a herd. However, Wijewantha and Karunaratne (1968) reported
that 15% of animals that came into an abattoir were carriers. Biofilm formation is a consideration
in contributing to the carrier state and chronic infections. There are two types of carriers: latent
and active. Latent carriers are animals surviving an HS outbreak and the organism becomes
27 | P a g e
harboured in their tonsils. Once active multiplication occurs in their tonsils, and the organism
sheds into nasal secretions, they became active carriers (De Alwis, 1999e).
The carrier rate in 500 buffaloes in a study done in an endemic area in Pakistan was 4.6% (Sheikh
et al., 1994). It was reported in an outbreak that 7.5% of clinically-normal animals were actually
nasopharyngeal carriers. However, none of these carriers could be detected 40 days later (Gupta,
1962).
It is important to note that in all of these studies, swabbing was done through the external nares
of live animals except for the study of Wijewantha and Karunaratne (1968) where swabbing was
done at post-mortem. This could account for the higher percentage of carriers they found (15%). It
was reported also by Singh (1948) that the detection rate doubled with swabbing post-mortem
compared with just swabbing the external nares in live animals. These findings could be attributed
to the post-mortem release of the organism in the body fluids (Carter and De Alwis, 1989) and
better access to nasopharynx during necropsy compared to access in the live animals.
It has been reported that there is a correlation between the length of time since an outbreak
occurred and the proportion of animals that are nasopharyngeal carriers of P. multocida
(Hiramune and De Alwis, 1982; De Alwis, 1999e). The percentages of nasopharyngeal carrier
animals detected were 22.7% and 1.9% in herds examined one and six weeks after an outbreak,
respectively (Hiramune and De Alwis, 1982; De Alwis, 1999e).
De Alwis et al. (1990) experimentally produced carrier animals following inoculation or natural
exposure to HS in 57 young buffaloes. Only 32 became carriers. These 32 were divided into several
groups for further investigation. One group was monitored for 360 days to investigate the
28 | P a g e
duration of the carrier status. Carrier status was defined as animals that showed positive
nasopharyngeal isolation of HS-associated P. multocida. It was found that in most animals, P.
multocida appeared for a short time (10-48 days) in the nasopharynx and then could not be
detected again, except in one case. In this one case, the bacterium could be detected in the
nasopharynx 215 days after initial exposure.
In this same study, another group of young buffaloes was slaughtered at varying intervals after
they had become carrier animals. Attempts were made to isolate the organism from 14 different
sites: the nasopharynx (swabbed directly after slaughter); tonsils; the mandibular,
retropharyngeal, cervical, axillary, mediastinal, bronchial, mesenteric and hepatic lymph nodes;
haemolymph nodes; spleen; parotid gland and mandibular salivary glands. Organisms were found
in the tonsils from 20 of the 27 animals. The longest period after which P. multocida serovar B:2
was isolated from the tonsils of a carrier animal was 229 days after initial exposure. This means
that the most consistent site to isolate P. multocida after death was the tonsils (De Alwis et al.,
1990). It was reported also that Pasteurella can be isolated from lymph nodes of slaughtered
animals (Wijewardana et al., 1986b; Dartini and Ekaputra, 1996). Recently, a study localised P.
multocida serotype B:2 to the gastrointestinal and urinary tracts of carrier buffaloes that survived
experimental outbreaks of HS, suggesting their involvement in the transmission of HS (Annas et
al., 2014). Pasteurella multocida was detected in the swabs by morphological criteria and PCR
(Annas et al., 2014).
Figure 2.3 (obtained from De Alwis (1999e)) shows a presumptive epidemiological cycle of HS.
After an HS outbreak in an endemic area, a large number of surviving animals become latent
carriers. These carriers, for unknown reasons, may become active carriers and then shed the
organism intermittently. No new clinical cases occur until a susceptible animal, which is usually
29 | P a g e
one born after the last outbreak or otherwise newly introduced into the herd, comes in contact
with shedders. This susceptible animal becomes a clinical case and will then die or, if treated
effectively or naturally recovered, it may survive and become an active carrier (De Alwis, 1999e). It
is unclear why many clinical cases occur despite surviving previous outbreaks and being
vaccinated. The role of stress has not been adequately investigated and so its role in disease
pathogenesis is unclear.
30 | P a g e
Figure 2.3 Presumptive epidemiological cycle for haemorrhagic septicaemia (De Alwis, 1999e).
Susceptible
animal
Clinical case
Shedding and
transmission of
the bacterium
Active carrier
Latent carrier
Contamination
of water and
pasture
Indirectly Directly
Infects Becomes
31 | P a g e
2.5 Haemorrhagic Septicaemia Epidemiology in Pakistan
Generally, South Asia is the area of highest prevalence and incidence of HS (Bain et al., 1982). This
is attributed to radical changes in weather between seasons, animal debilitation caused by
seasonal scarcities of fodder and the pressures of the work that animals do e.g. draught animals
(De Alwis, 1999b). The disease also occurs, but to a lesser extent, in the Middle East and Africa.
Predisposing conditions are not as clearly defined as in South Asia (De Alwis, 1999b).
In Pakistan, livestock contributes 11.8% to the nation’s gross domestic product (GDP). The total
numbers of cattle and buffaloes in 2013-2014 were 39.7 and 34.6 million, respectively (Farooq,
2014). The rural human population in Pakistan is 30-35 million and 30-40% of rural family income
depends on grazing small numbers (less than ten) of cattle, buffaloes, sheep and/or goats (Anjum
et al., 2006).
Haemorrhagic septicaemia has been reported as the most important bacterial disease of cattle and
buffaloes in Pakistan (Munir et al., 1994). In Pakistan, HS is considered as a disease of great
economic importance. In the Punjab province alone, the financial losses due to HS were estimated
to be more than 2.17 billion Pakistani rupees (equivalent to 58 million USD) in 1996 (Anonymous,
1996; Imran et al., 2007). According to farmers’ opinions in a participatory disease surveillance
(PDS) done in Karachi, HS is more important than foot and mouth disease (FMD) and this is due to
the higher mortality rate and the greater economic impact of HS (Ali et al., 2006).
Participatory Disease Surveillance was introduced to Pakistan for the first time by the Food and
Agriculture Organisation (FAO) under the project “Support for Emergency Prevention and Control
of Main Trans-boundary Diseases in Pakistan”. Fifty-one veterinarians, divided into 17 teams,
investigated trans-boundary animal diseases and other important animal diseases in six Pakistani
32 | P a g e
provinces. Data were collected from 2000-2005 (Farooq et al., 2007). The study showed that,
according to farmers’ opinions, the percentage of respondents who felt that the economic
importance of HS was “high” was always higher than the percentage of respondents who felt that
the prevalence of HS disease was high in all districts, due to the high case fatality and losses
associated with the disease (Farooq et al., 2007).
Another study was done in Malakand, Khyber Pakhtunkhwa, Pakistan in order to investigate the
sero-prevalence of HS in cattle and buffaloes (Khan et al., 2006). Antibodies against P. multocida
were detected by the indirect haemagglutination test. Regions within Malakand were categorised
as towns, big villages or small villages. The number of cattle and buffaloes included in this study
was 2,525 and 1,846, respectively (4,371 total) from ten randomly selected villages and towns. The
selected number of both cattle and buffaloes were further divided into two age groups, adult and
young. It was not stated what the cut-off age for each group was (Khan et al., 2006).
The morbidity, mortality and case fatality rates for buffaloes and cattle from the Malakand region
are shown in Table 2.2. The authors concluded that the occurrence of HS in buffaloes was higher
compared to cattle (Khan et al., 2006). This claim was not supported with a P value and confidence
intervals were not calculated. The authors compared their morbidity, mortality and case fatality
rate results with the results of Sheikh et al. (1996) who had carried out a similar study on HS in
nine districts of Punjab province in Pakistan. However the morbidity, mortality and case fatality
rates were not mentioned anywhere in the study by Khan et al. (2006). Another paper reported
exactly the same study, and obtained similar results, but the study was from Southern Punjab
(Farooq et al., 2011).
33 | P a g e
Animal Group Population Morbidity% Mortality% Case fatality rate%
Buffaloes
Young male 313 13.42 13.42 100 Young female 442 28.5 26.69 93.65
Adult male 182 6.59 - - Adult female 909 5.28 1.98 37.5
Combined Young 755 22.25 21.19 95.23 Adult 1091 5.49 1.65 30
Cattle
Young male 418 4.31 1.67 38.8 Young female 595 3.69 1.85 50
Adult male 234 0.85 - - Adult female 1278 2.82 0.47 16.6
Combined Young 1013 3.94 1.77 45 Adult 1512 2.51 0.39 15.79
Table 2.2 Morbidity, mortality and case fatality rate due to haemorrhagic septicaemia in buffaloes and cattle from ten randomly selected villages/towns from within the district of Malakand, Pakistan, obtained from Khan et al. (2006).
2.6 Diagnosis of Haemorrhagic Septicaemia
A definitive diagnosis of HS must be achieved very quickly in order to implement remedial and
preventive measures as soon as possible. First of all, a provisional clinical diagnosis can be used
when a suspected outbreak of HS is discovered. Later, laboratory diagnostics and molecular
techniques can be used to confirm the provisional diagnosis and the serotype involved (De Alwis,
1999c).
2.6.1 Clinical diagnosis
The clinical diagnosis of HS does not depend on any one individual clinical sign. A combination of
clinical signs, pathological lesions and local epidemiology should be used to arrive at a clinical
diagnosis. Haemorrhagic septicaemia tends to occur in animals reared in places with poor animal
husbandry conditions, poor management and large animal populations (De Alwis, 1999c).
Consequently, clinical signs are frequently not noticed in the first cases of an outbreak. Or put
another way, under these conditions, sudden death is often the only sign of HS (De Alwis, 1999c).
For clinical signs of HS disease, refer to the earlier relevant section (2.3.1-Clinical signs).
34 | P a g e
2.6.2 Differential diagnoses
When investigating sudden death cases in a known HS-endemic area, conditions other than HS
must also be considered. Infectious agents such as Bacillus anthracis (anthrax) and clostridial
diseases (e.g. blackleg, enterotoxaemia, botulism) should be investigated. Moreover, non-
infectious agents such as lightning strike, snake envenomation and poisonings are also potential
causes of sudden death in cattle and buffaloes (Carter and De Alwis, 1989).
2.6.3 Laboratory diagnosis
Sampling
An overview of laboratory-based diagnostic services is illustrated in Figure 2.4. Sample collection is
the first and the most important step in the chain of biological, biochemical, microbiological,
serological and molecular diagnostic techniques (Carter and De Alwis, 1989).The tissues, where
the specific organism is abundant, and is in relative purity, are the best places to collect diagnostic
samples. Fresh carcass tissues and blood drawn directly from the heart are often recommended
samples (Carter and De Alwis, 1989). Under tropical conditions, decomposition of the carcass will
occur quickly and after a few hours, the blood will likely be contaminated by large quantities of a
wide variety of microbes. For this reason, the bone marrow of a long bone is often a more
appropriate sample. It can be collected for culturing and biological screening in mice and this can
be done even after a carcass has been buried for a few days (Carter and De Alwis, 1989).
Sampling clinically-affected animals before death may not be able to determine whether the
animal was infected. However, blood taken during terminal septicaemia will typically give positive
results (Carter and De Alwis, 1989). It is preferable to test material from more than one animal in
an outbreak. The sample must be sent to the laboratory with minimal delay using transport media
that supports the growth of P. multocida while simultaneously inhibiting the growth of other
35 | P a g e
microbes (De Alwis, 1999c). Examples of suitable transport media include modified Stuart’s
transport medium, Amie’s transport medium (De Alwis, 1999c), transport enrichment medium
(TEM) (Warner, 1996) and the transport medium formulated by De Alwis (De Alwis, 1973). Once at
the laboratory, two procedures can then be applied.
First, direct culturing of uncontaminated material on enriched blood agar plates allows colonies to
be obtained within approximately 24 hours. The second procedure, which is more common in
diagnostic laboratories due to contamination in most samples, utilises mouse inoculation and
culture of mouse blood for obtaining a pure culture (Carter and De Alwis, 1989).
Mouse inoculation is a traditional and selective approach for isolation of P. multocida ssp.
multocida (Muhairwa et al., 2001). It is useful when the sample reaching the laboratory has
contaminants and extraneous bacteria. Subcutaneous inoculation of 0.1-0.2 mL of a saline
suspension of blood or bone marrow sample into laboratory mice is the best way to isolate P.
multocida from contaminants (De Alwis, 1999c). It has been reported that if Pasteurella is present,
the inoculated mouse will die within 24 hours (Carter and De Alwis, 1989; De Alwis, 1999c; OIE,
2012). Smears of heart blood from the dead mouse will show bipolar staining of coccobacilli when
stained with Gram’s stain, methylene blue or Leishman stain (De Alwis, 1999c). The disadvantages
of this technique are maintaining a mouse colony and there are animal ethics issues.
36 | P a g e
Figure 2.4 Routine laboratory procedures for isolating and characterising Pasteurella multocida associated with haemorrhagic septicaemia, adapted from De Alwis (1999c).
Animal sample
(such as long
bone, spleen
and lung tissues) Uncontaminated Contaminated
Mouse
inoculation
Mouse blood
Direct
examination of
stained smear Culture
Molecular
techniques Non-serological
tests (such as
acriflavine
flocculation test)
Serological tests Biochemical test
Direct
examination of
stained smear
37 | P a g e
Biochemical tests
No clear relationship has been established between (1) clinical disease, epidemiology, virulence or
serotype and (2) a pattern of acid production from substrates in an in vitro environment (Carter
and De Alwis, 1989). Some biochemical characteristics are reliably repeatable within the species
while others are variable. The biochemical properties of P. multocida are shown in Table 2.3. Sugar
fermentation capabilities of P. multocida are shown in Table 2.4 (De Alwis, 1999d).
Biochemical reaction Result
Indole + Hydrogen sulfide production +
Nitrate reduction + Growth on potassium cyanide +
Catalase production + Oxidase production +
Ornithine decarboxylase + Haemolysis -
Growth on MacConkey medium - Gelatin hydrolysis -
Methyl red reaction - Voges-Proskauer reaction -
Urease production - Citrate utilisation -
Malonate utilisation - Lysine decarboxylase -
Arginine decarboxylase -
Table 2.3 Results of biochemical reactions for Pasteurella multocida, adapted from De Alwis (1999d).
38 | P a g e
Sugar Result
Glucose + Fructose +
Galactose + Mannose + Sucrose + Maltose Variable (usually positive)a
Trehalose Variable (usually positive)a
Arabinose Variable (usually positive)a
Xylose Variable (usually positive)a
Mannitol Variable (usually positive)a
Sorbitol Variable (usually positive)a
Lactose Variable (usually negative)a
Inositol Variable (usually negative)a
Dulcitol Variable (usually negative)a
Salicin Variable (usually negative)a
a The percentage of positives were not given in the reference.
Table 2.4 Results of sugar fermentation patterns of Pasteurella multocida, adapted from De Alwis (1999d).
Serological tests
Serological tests are used to detect and categorise the somatic and capsular antigens of the P.
multocida bacteria. The three conventional tests used most often in laboratory diagnosis are rapid
slide agglutination (capsular), indirect haemagglutination (capsular) and agar gel precipitation
(somatic). First, rapid slide agglutination is a convenient and routine laboratory test and is based
on the technique described by Namioka and Murata (Namioka and Murata, 1961b). A positive
result is indicated by a rapid, flaky agglutination appearing in a few seconds with a clear
background. Second, in indirect haemagglutination, the agglutination of erythrocytes coated with
the surface antigen of P. multocida, in the presence of rabbit hyperimmune antiserum, is the
underlying principle (Carter, 1955). Several modifications were done to accommodate various
erythrocyte types being used in the test (Carter and Rappay, 1962; Sawada et al., 1982;
Wijewardana et al., 1986a). Third, agar gel precipitation identifies the 16 Heddleston somatic types
of P. multocida; a positive result is marked by bands of antibody-antigen precipitate (Heddleston et
al., 1972).
39 | P a g e
Several other serological tests for somatic and capsular typing have been developed and used such
as an agar gel precipitation test to detect capsular (as opposed to somatic) antigen (Anon, 1981;
Wijewardana et al., 1982); counter-immunoelectrophoresis (Carter and Chengappa, 1981);
agglutination test (somatic) (Namioka and Murata, 1961c; Namioka and Murata, 1961a; Namioka
and Bruner, 1963; Namioka and Murata, 1964); and enzyme-linked immunosorbent assay (ELISA)
(Dawkins et al., 1990).
Non-serological tests
Non-serological tests include the acriflavine flocculation test for identification of serogroup D
(Carter and Subronto, 1973), a hyaluronidase decapsulation test for serogroup A (Carter and
Rundell, 1975) and hyaluronidase production to test for serogroup B (Carter and Chengappa,
1981).
Molecular techniques
Techniques for the identification of P. multocida and its capsular serogroups using PCR have been
developed. Molecular techniques are also used for the diagnosis of HS (De Alwis, 1999c). Not only
can molecular techniques differentiate between serotypes associated with HS, but also advanced
molecular techniques can differentiate between strains within serotypes (De Alwis, 1999c).
Molecular techniques can be combined with epidemiological studies to enable phylogenetic
disease mapping.
1. Polymerase chain reaction (PCR) PCR tests amplify minute quantities of a targeted region of DNA from organisms and organs’ cells
and tissues by specific primers designed for the target DNA. Many PCR methods have been
developed, based on different techniques and targets, for the diagnosis of HS (Brickell, 1996;
Natalia, 1996; Thomas, 1996; Townsend et al., 1998).
40 | P a g e
It is important to clarify that PCR as a technique is not a diagnostic test but it is an amplification
technique. It is followed by gel electrophoresis and visualisation of the amplicons. Other
techniques like sequencing are needed to confirm the identity of the amplicon.
Direct testing of a clinical sample that has a low number of bacteria can be achieved by
appropriate design of the PCR test. Different levels of specificity for the pathogen, such as strain,
serotype or species can be achieved with well-designed PCR tests (De Alwis, 1999c). Advantages of
PCRs used for HS diagnosis over conventional methods are rapidity, sensitivity and specificity (De
Alwis, 1999c).
The following section will discuss PCRs used for identification and characterisation of P. multocida.
A. PCRs specific for Pasteurella multocida A polymerase chain reaction was designed by Kasten et al. (1997) to amplify 453 bp of psI gene
encoding a P6-like protein that is unique to P. multocida and H. influenzae. Limitation of the test
was the cross reactivity with H. influenzae which restricted the use of this PCR for veterinary (and
not human) specimens. The other limitation was the need of southern hybridisation to achieve a
detection limit of 10 P. multocida organisms per reaction (Dziva et al., 2008).
Another P. multocida-specific PCR by Townsend et al. (1998) was based on primers that were
designed using the sequence of a P. multocida specific clone that was isolated by genomic
subtractive hybridisation. This primer set KMT1SP6-KMT1T7 amplifies a template to produce a
product of approximately 460 bp. At the time of this test’s development, a “false” positive
occurred with Pasteurella canis biovar 2. The closely related P. avium biovar 2 was not tested with
this PCR. Since this publication, P. canis biovar 2 and P. avium biovar 2 have been renamed
Pasteurella multocida (Christensen et al., 2004). Consequently, this test could detect all known
41 | P a g e
subspecies of P. multocida and this PCR is considered specific for P. multocida.
A third P. multocida-specific PCR targets the 23S rRNA gene (Miflin and Blackall, 2001). It gave
positive results with all subspecies of P. multocida as well as P. canis biovar 2 and P. avium biovar 2.
However, as mentioned before, these two species were renamed P. multocida(Christensen et al.,
2004).
Two PCR tests based on two transcriptional regulator gene sequences (pm0762 and pm1231) were
developed to detect P. multocida. The results showed that these two genes appeared to be
species-specific. The two PCR tests gave products of 567 bp and 601 bp, respectively (Liu et al.,
2004).
Corney et al. (2007) designed a 5′ Taq nuclease assay targeting the 16S rRNA gene to detect P.
multocida. This test was initially described for detecting P. multocida in swabs from cases of fowl
cholera. However, the potential for this test to detect P. multocida in other host species has been
reported. In addition to reference strains of P. multocida, 17, 11 and 12 Australian field isolates
from avian, porcine and bovine hosts were all detected by this test. A minimum of approximately
10 CFU of P. multocida per reaction was detected in this test. Specificity was confirmed by
negative results obtained with nine other Pasteurella species, 26 other bacterial species (18 being
members of the family Pasteurellaceae) and four poultry virus isolates (Corney et al., 2007).
A recent study compared four main PCR methods, based on kmt1, (Townsend et al., 1998); 23S
rRNA, (Miflin and Blackall, 2001); and transcriptional regulator genes pm0762 and pm1231, (Liu et
al., 2004), for specific detection of P. multocida (Adhikary et al., 2013). There were 85 strains of
P. multocida and 13 strains of other taxa (such as Gallibacterium, Actinobacillus, Mannheimia,
42 | P a g e
Avibacterium, Salmonella and Bisgaard taxa 14, 16 and 45) used for this comparison. The two
PCRs, based on transcriptional regulator genes pm0762 and pm1231, were concluded to be
specific for P. multocida. The sensitivity and specificity of the 23S rRNA PCR was 80% and 23%,
respectively. On the contrary, the kmt1 method had the expected sensitivity (100%) and specificity
(92%), and authors recommended it for identification of the species (Adhikary et al., 2013).
B. Multiplex PCR for capsular typing of Pasteurella multocida The nucleotide sequences of the genes dcbE-dcbB, ecbA-ecbI and fcbE-fcbB from the cap loci of
serogroups D, E, and F, respectively, were determined by Townsend et al. (2001) using primers that
were designed using the information derived from sequencing the biosynthetic loci of the capsules
of serogroups A and B of P. multocida A:1 (Chung et al., 1998) and P. multocida B:2 (Boyce et al.,
2000a), respectively. Consequently, serogroup-specific oligonucleotide primers were identified and
used in a multiplex PCR for capsular typing of P. multocida (Townsend et al., 2001). In this PCR,
pan-P. multocida primers (KMT1SP6 and KMT1T7) can be used as an internal control. The amplicon
sizes of capsular types A, B, D, E and F are 1044, 760, 657, 511 and 851 bp, respectively.
C. PCR specific for serogroup B associated with HS The sequence of a clone named 6b was obtained by genomic subtractive hybridisation from an HS-
associated P. multocida serogroup B. This sequence was used to design primers that could be used
in a PCR specific for serogroup B (Townsend et al., 1998). This primer pair, KTSP61-KTT72 amplifies
a fragment of 560 bp in all P. multocida serogroup B of serotypes 2 or 5.
The 956 bp clone, 6b, had shown identity to a region in the genome of Haemophilus influenzae Rd
strain, specifically to bacteriophage Mu genes. This 956 bp clone is close to the gene fragment
amplified in the PCR technique of Brickell et al. (1998).
A multiplex PCR composed of the P. multocida diagnostic PCR and an HS diagnostic PCR has been
used for rapid and specific diagnosis (Townsend et al., 1998).
43 | P a g e
D. PCR specific for Pasteurella multocida serotype B:2 A gene fragment which codes for a protein associated with the pathogenicity of P. multocida
serotype B:2 formed the basis for a PCR designed to detect this agent in Asia (Brickell et al., 1998).
Primers were designed from a 16S-23S rRNA intergenic spacer region unique to serotype B:2
(Brickell et al., 1998). The primers used are IPFWD and IPREV producing an amplicon of 334 bp
(Brickell et al., 1998). Other authors have subsequently concluded that this PCR is neither serotype
B:2-specific nor HS-specific (Brickell et al., 1998; De Alwis, 1999c). While the former is a result of
an isolate 0350 (E:2) being detected with this PCR, the latter resulted from the inability of this PCR
to detect all known HS-associated serotypes (Brickell et al., 1998; De Alwis, 1999c).
Due to the genetic similarity of this unique fragment of P. multocida with that of H. influenzae Rd
strain, it was claimed that the sequences neighbouring this unique fragment of P. multocida will
recognise the phage insertion site and phage-associated virulence determinants shared with that
of H. influenzae Rd (Brickell et al., 1998).
2. Real-time PCR specific for HS-associated strains In a recent study, Petersen et al. (2014) typed 64 isolates of P. multocida using multilocus sequence
typing (MLST). Fifty five of these isolates were associated with HS (serotypes B:2 and E:2). They
found that the majority of isolates belonged to sequence type (ST) 122 (n = 50) and only one, two
and two isolates belonged to ST63, ST147 and ST162, respectively (Petersen et al., 2014). The four
isolates belonging to ST147 and ST162 were capsular type E. Single-nucleotide polymorphisms in
the est gene suitable for detection of HS-associated STs were found. A new real time PCR was
designed and it was named HS-est-RT-PCR. This PCR was able to specifically detect ST122, ST63,
ST147 and ST162 which have all been associated with HS. It should be recognised that
Petersen et al. (2014) used the abbreviation “RT-PCR” to denote “real-time PCR”, not reverse
transcription PCR.
44 | P a g e
The new HS-est-RT PCR is claimed to be more specific in detecting HS-associated isolates than
other PCR detection methods previously published (Petersen et al., 2014). In contrast, the PCR
developed by Brickell et al. (1998) left one capsular type B strain and two capsular type E strains
undetected and the HSB-PCR of Townsend et al. (1998) could not detect strains of capsular type E.
3. Loop-mediated isothermal amplification (LAMP) Loop-mediated isothermal amplification is an autocycling DNA synthesis reaction that is performed
under isothermal conditions in the presence of a thermophilic strand-displacing DNA polymerase
such as Bst polymerase (Notomi et al., 2000). The technique uses four to six primers which target
six to eight sequences on the nucleic acid. The four primers are two inner primers, forward inner
primer (FIP) and backward inner primer (BIP), and two outer primers, forward primer (F3) and
backward primer (B3). The outer primers are a few bases shorter and are lower in concentration
than the inner primers (Notomi et al., 2000). Two loop primers, loop forward (LF) and loop
backward (LB), can be added to increase the rate of the reaction, decreasing overall amplification
times by approximately 50% (Nagamine et al., 2002).
The mechanism and expected reaction steps have been illustrated previously by Notomi et al.
(2000) (Figure 2.5). The first strand synthesis is initiated when the forward inner primer (FIP)
anneals to the complementary region (F2c) in the target DNA (step 1). The outer forward primer
(F3) then hybridises and displaces the first strand (steps 2 and 3), forming a loop structure at one
end (step 4). This single-stranded DNA acts as template for DNA synthesis initiated by the
backward inner primer (BIP) and subsequent outer backward (B3)-primed strand displacement
DNA synthesis (steps 4 and 5), leading to the formation of dumb-bell shaped DNA structures (step
6). This dumb-bell form is quickly converted to a stem-loop DNA by self-primed DNA synthesis
(step 7), which acts then as a template for the second stage of the LAMP reaction. One inner
45 | P a g e
primer hybridises to the loop on the product and initiates the displacement DNA synthesis,
forming the original stem loop and a new stem loop that is twice as long. The final products are a
mixture of stem-loop DNA with several inverted repeats of the target DNA, and cauliflower-like
structures having multiple loops (Notomi et al., 2000).
46 | P a g e
Figure 2.5 Schematic representation of the mechanism of a loop-mediated isothermal amplification (LAMP) reaction. This process starts once the forward inner primer (FIP) has annealed to the DNA template. However, DNA synthesis can also begin from the backward inner primer (BIP). The sequences (typically 23-24 nucleotides) inside both ends of the target region for amplification in a DNA are designated F2c and B2, respectively. Two inner sequences (typically 23-24 nucleotides) 40 nucleotides from the ends of F2c and B2 are designated F1c and B1 and two sequences (typically 17-21 nucleotides) outside the ends of F2c and B2 are designated F3c and B3. FIP contains F1c, a TTTT linker and the sequence F2, which is complementary to F2c. BIP contains the sequence B1c, which is complementary to B1, a TTTT linker and B2. The two outer primers consist of B3 and the sequence F3, which is complementary to F3c, obtained from Notomi et al. (2000).
47 | P a g e
There are different methods to detect LAMP amplification products such as gel electrophoresis
(Notomi et al., 2000), real time monitoring of turbidity with a turbidimeter (Mori et al., 2004), and
visual detection using intercalating fluorescent dye (Parida et al., 2008).
There is one previously described LAMP reaction for detection of P. multocida in clinical samples
from swine (Sun et al., 2010). To the author’s knowledge, there is no LAMP reaction that had been
previously developed for the specific detection of HS-associated isolates of P. multocida in any
animal species.
Compared to real-time PCR, LAMP uses cheaper reagents and results can be obtained more
quickly. In addition, there is no need for sophisticated laboratory infrastructure or highly trained
personnel. Consequently, real-time PCR is unsuitable in a laboratory with limited resources; which
is the case in most of the HS-endemic countries.
4. DNA fingerprinting methods
This section will review the molecular techniques that can be used for DNA fingerprinting and
epidemiological studies.
I. PCR fingerprinting of Pasteurella multocida isolates that are associated with HS
a. REP-PCR Repetitive extragenic palindromic (REP) sequences were identified as the first family of conserved,
repetitive sequences of DNA in the prokaryotic genome (Higgins et al., 1982; Lupski and
Weinstock, 1992). These sequences, approximately 35 bp in length, are commonly found in
intercistronic regions from multicistronic operons of Salmonella enterica serovar Typhimurium and
E. coli (Higgins et al., 1982). These sequences have been reported to be found in eubacteria, more
commonly in Gram-negative enteric bacteria and related species (Versalovic et al., 1991). The
distribution of repetitive extragenic sequences throughout the prokaryotes’ genomes was the basis
48 | P a g e
for the REP-PCR for DNA fingerprinting purposes (Versalovic et al., 1991).
Primers in this PCR are designed to be directed outwardly from each half of the consensus REP
sequence in order to amplify DNA fragments located between neighbouring REP sequences (Gilson
et al., 1987; Versalovic et al., 1991). This method has a high discriminatory power between strains
of the same species depending on their different fingerprinting profiles. Furthermore, it can
identify virulence-associated elements (Versalovic et al., 1991; Go et al., 1995). Repetitive
extragenic palindromic (REP) primers will amplify DNA fragments which contain the gene between
the adjacent REP elements.
Townsend et al. (1997b) used REP-PCR to analyse the DNA fingerprint of different strains of P.
multocida associated with HS in order to determine the discriminatory power of this method. The
goal was to correlate REP profiles with the previously reported virulences of HS-associated strains
of P. multocida (Townsend et al., 1997b). They examined different isolates from different
geographical areas (Townsend et al., 1997b). The primers used were REP1R-IDt and REP2-IDt
(Townsend et al., 1997b). Results showed an increase in the amplification products (number of
bands) of the profiles of non-HS isolates compared to HS isolates (Townsend et al., 1997b). These
findings support the contention that there are fewer REP sequences in the genomes of HS isolates
and therefore, there is an association between the number of REP sequences and the bacterium’s
pathogenicity. Homogeneity of REP fingerprints of isolates associated with HS supported the idea
of a disease-associated REP profile (Townsend et al., 1997b). The role of REP sequences is poorly
understood, but it has been suggested that they stabilise upstream mRNA in order to regulate gene
expression (Newbury et al., 1987).
49 | P a g e
In a comparison between vaccine strain P52 (serotype B:2) and five isolates from an HS outbreak in
Uttar Pradesh, India (Biswas et al., 2004), it was found that the five isolates showed three different
profiles (Biswas et al., 2004). The decreased number of REP sequences in the profile of these three
strains may be related to the pathogenicity of these isolates. The author concluded that the
movement of animals in and out of the herd resulted in this molecular diversity of the strains
responsible for the outbreak (Biswas et al., 2004).
In contrast to the reports by Townsend et al. (1997b) about the homogeneity profiles of HS
isolates, a marked variation was noticed in the three profiles found for HS isolates from the study
by Biswas et al. (2004).
b. ERIC-PCR Enterobacterial Repetitive Intergenic Consensus (ERIC) sequences, also known as intergenic repeat
units (IRUs) (Sharples and Lloyd, 1990), are 126 bp long in eubacteria (Hulton et al., 1991;
Versalovic et al., 1991). These sequences were identified using the genetic information of E. coli
and Salmonella enterica serovar Typhimurium and appeared to be found either in the intergenic
regions of polycistronic operons or in untranslated regions upstream or downstream of open
reading frames (Hulton et al., 1991). The nucleotide sequence of ERIC is highly conserved among
bacterial species. However, their chromosomal locations differ according to the species (Hulton et
al., 1991).
ERIC-PCR has been conducted using primer sequences targeting the palindromic sequences of ERIC
(Versalovic et al., 1991). The primers used were ERIC1R and ERIC2. Biswas et al. (2004) used this
PCR to investigate the five strains obtained in an outbreak of HS in India. The results were similar
to the REP-PCR results in that three different profiles were obtained, compared to the vaccine
strain P52.
50 | P a g e
c. Single-primer PCR This PCR fingerprinting method uses a microsatellite primer which targets variable repetitive
regions (Dabo et al., 2000). Biswas et al. (2004) used this method and confirmed the results of
REP-PCR and ERIC-PCR. The primer sequence used was (GTG)5.
II. Restriction endonuclease analysis (REA) Restriction endonuclease analysis depends on digestion of DNA with a restriction enzyme at
specific nucleotide sequences. The DNA fragments that are produced can then be examined by
agarose gel electrophoresis. Different profiles result from different banding patterns. This
technique has been used to investigate different P. multocida strains associated with HS (Wilson et
al., 1992; Rimler, 2000).
Wilson et al. (1992) used REA and somatic serotyping to compare different serotypes of
serogroups B and E of P. multocida. One hundred cultures were used: 71 from serogroup B, 16
somatic reference strains and 13 from serogroup E. Preliminary studies investigating the
restriction endonuclease digestion of P. multocida DNA used EcoRI, KpnI, HindIII, AluI, BglII, DraI,
DpnI, EcoRV, HaeIII, HincII, HinfI, HpaII, MboI, PstI, RsaI, SmaI and HhaI. HhaI and HpaII were
ultimately selected because they were the most informative restriction enzymes for fingerprinting
because they yielded fingerprint profiles that were able to be distinguished the most easily and
clearly (Wilson et al., 1992). Twenty HhaI profiles were obtained for the 71 isolates of serogroup B
with the ability to discriminate different serotypes (Wilson et al., 1992). Thirteen of the 20 HhaI
profiles were obtained from the 54 isolates that were associated with HS serotypes (B:2 or B:2,5)
(Wilson et al., 1992). Only one of the 54 isolates had an identical profile to the reference somatic
serotype 2 strain (Wilson et al., 1992). All 13 serogroup E isolates shared the same HhaI profile.
Further discrimination of these 13 isolates was done by HpaII restriction enzyme analysis which
resulted in 5 different DNA fingerprint profiles (Wilson et al., 1992). In short, it can be concluded
that HhaI was capable of differentiating serogroup B isolates. However, the use of both HhaI and
51 | P a g e
HpaII was required to differentiate all the isolates of P. multocida.
In another study, the use of HhaI only resulted in 48 different profiles among 222 strains of
serogroup B, while the use of two enzymes (HhaI and HpaII) resulted in 88 profiles. Consequently,
a better discrimination among 222 strains of serogroup B was attained by the use of both HhaI and
HpaII (Rimler, 2000).
III. Analysis of genomic restriction patterns: The use of ribotyping and pulsed field gel electrophoresis (PFGE) to analyse the restriction
patterns, has shown a high discriminating power between different strains that are phenotypically
similar (Hector et al., 1992; Wei et al., 1992).
a. Ribotyping Ribotyping was previously shown to work well with REA (Snipes et al., 1989), and has the benefit
of allowing easier interpretation by highlighting rRNA gene polymorphisms.
Townsend et al. (1997a) used ribotyping to discriminate isolates of P. multocida associated with HS.
Seven different restriction enzymes were used in this study and each enzyme produced complex
fragment patterns which were difficult to analyse. The use of ribotyping made the interpretation
of these patterns easier (Hector et al., 1992).
b. Pulsed field gel electrophoresis (PFGE) This technique enables the analysis of large chromosomal DNA fragments (up to 10 megabase
pairs (Mbp) in length). Combining PFGE and restriction endonuclease digestion of genomic DNA of
different strains, a picture of the genetic variation between these strains was determined
(Townsend et al., 1997a). The advantage of this technique is its higher discrimination power, more
than REA and ribotyping, to investigate the clonality of outbreak strains and compare their
geographical distribution.
52 | P a g e
5. Multilocus sequence typing (MLST) Molecular typing methods can be used to investigate the local epidemiology of a disease outbreak
by determining the degree of similarity between the various isolates recovered from an outbreak.
Moreover, the global epidemiology of an agent can be checked for differences between strains
causing disease in one geographic area to those isolated from other areas. The most important
aspect about the methods used in local and global epidemiological investigations is to be able to
discriminate between closely and distantly related isolates. Isolates assigned to the same
molecular type should be descended from a recent and common ancestor while isolates that share
a more distant common ancestor should not be assigned to the same type (Maiden et al., 1998).
High levels of discrimination between different isolates can be achieved in two different ways. The
first way is that individual loci or uncharacterised regions of the genome, that are highly variable
within the bacterial population, are targeted. For bacterial pathogens, several methods based on
this approach are currently popular: ribotyping, PFGE, and REP-PCR or conventional PCR using
arbitrary primers. Methods that index rapidly evolving variation are useful for short term
epidemiology but may be misleading for global epidemiology. The second way is multilocus
enzyme electrophoresis (MLEE) which uses slowly accumulated variations in the population (likely
to be selectively neutral). Only a small number of alleles can be identified within the population by
using this type of variation. However, high levels of discrimination are achieved by analysing many
loci (Maiden et al., 1998).
Multilocus enzyme electrophoresis is an appropriate technique for long term (global)
epidemiology. This approach also has contributed the most to our understanding of the global
epidemiology and population structure of infectious agents (Maiden et al., 1998). A major problem
with all current typing methods, including MLEE, is the difficulty of comparing results of different
53 | P a g e
laboratories. To overcome this problem, MLST has emerged. Multilocus sequence typing methods
sequence the internal fragments of housekeeping genes rather than comparing the
electrophoretic mobilities of encoded enzymes (Maiden et al., 1998).
One advantage of this modification is the detection of more variations, and consequently more
alleles per locus result, than that obtained with MLEE. Another advantage is that it solved the
problem of comparing the results between laboratories as data are fully portable and stored in an
expanding central multilocus sequence database on the internet which can be used for global
epidemiology studies (Maiden et al., 1998).
Three elements are important for the design of a new MLST scheme: the choice of the isolates to
be used in the initial evaluation; the choice of the genetic loci to be characterised; and the design
of primers for gene amplification and nucleotide sequencing (Urwin and Maiden, 2003). For MLST
analysis, all unique sequences for a given locus are assigned an allele number in order of
discovery; this is equivalent to the designation of ‘electromorphs’ in MLEE. The alleles present at
each of the MLST loci for a given isolate are combined into an allelic profile and assigned an ST
designation. The different processes in a MLST scheme are shown in Figure 2.6 (Urwin and
Maiden, 2003).
An MLST scheme has been established for P. multocida and it is named the Pasteurella multocida
Rural Industries Research and Development Corporation (RIRDC) scheme (Subaaharan et al.,
2010). This scheme is based on seven loci: adk, est, pmi, zwf, mdh, gdh and pgi. As of July, 2015,
the database had the sequence information of the seven loci of 748 different isolates. The scheme
was originally designed to type avian isolates, but it has since been used by the international
research community to submit data relating to several other host species such as pigs, turkeys,
54 | P a g e
humans, camels and kangaroos (Pasteurella multocida RIRDC MLST Database). Cattle and
buffaloes are represented by 199 (the largest host source) and 55 isolates, respectively. There are
67 isolates (48 from buffaloes and 19 from cattle) from HS cases.
An alternative scheme, the Pasteurella multocida multi-host MLST scheme, hereafter referred to
as “the alternative MLST scheme” (http://pubmlst.org/pmultocida_multihost/), is also available
but at the current time it is not possible to submit isolates into this database (the alternative MLST
scheme).
55 | P a g e
Figure 2.6 Illustration of multilocus sequence typing process, adapted from Urwin and Maiden (2003).
6. Sequenced P. multocida genomes
Currently (September, 2014), there are 25 complete and draft genomes of P. multocida available on
Genbank, collected from a range of different hosts and representing a variety of diseases. Six of
them are fully sequenced and annotated P. multocida genomes: strains Pm70, 36950, HN06, 3480,
HB03 and ATCC 43137 (Table 2.5). Five of the 22 genomes, available as contigs, are HS-associated:
Collection of
isolates and DNA
extraction
Amplification of
target loci by
PCR
DNA sequencing
Matching allele
sequence and
sequence type
(ST)
identification
Assignment of
STs to clonal
complexes
Population
studies
Epidemiological
studies
Novel allele
sequences and
STs verified and
added by curator
to the database
Data
collection
Data
analysis
MLST
analysis
56 | P a g e
strains P52VAC, Anand1C, Anand1B and VTCCBAA264 are from India and strain PMTB is from
Malaysia (Table 2.5). All of the currently available P. multocida genomes are between 2 and 2.4
Mbp in length and comprise a single circular double-stranded DNA genome with a G+C content of
between 40 and 41%.
Recently, four incomplete genome sequences, derived from Illumina GAIIx technology, of the two
fowl cholera isolates X73 and VP161 (both A:1), the bovine HS isolate M1404 (B:2) and the swine
isolate P903 (D:11) were determined (Boyce et al., 2012). To our knowledge, there were no
sequenced HS-associated isolates from other regions of Asia such as Pakistan and Thailand, or from
Africa, that were currently available in publically accessible international archives.
Strain Genbank Accession
number Host
Status of genome sequence
Reference
Pm70 AE004439 Avian Complete (May et al., 2001)
36950 CP003022 Bovine Complete (Michael et al., 2012a)
HN06 CP003313 Swine Complete (Liu et al., 2012)
3480 CP001409 Swine Complete Direct submission
HB03 CP003328 Swine Complete Direct submission
ATCC 43137 CP008918 Swine Complete Direct submission
P52VAC ALBZ00000000 Buffalo
calf Draft Direct submission
Anand1C ALBY00000000 Buffalo Draft Direct submission
Anand1B ALBX00000000 Cattle Draft Direct submission
VTCCBAA264 ALYC00000000 Buffalo
calf Draft Direct submission
PMTB AWTD00000000 Buffalo Draft (Yap et al., 2013)
Table 2.5 Complete genomes of Pasteurella multocida and haemorrhagic septicaemia-associated draft genomes.
2.7 Treatment
The World Organisation for Animal Health (OIE) listed HS in the OIE-Listed diseases from 2006-
2014. Haemorrhagic septicaemia is a primary bacterial infection and because no other biological
57 | P a g e
agents are thought to be involved, it was found that clinical cure was possible if effective treatment
was implemented in the very early stages of disease (Chandrasekaran, 1993; Benkirane and De
Alwis, 2002).
The best way to detect the disease once an outbreak starts is by checking the rectal temperature
regularly of in-contact animals (Benkirane and De Alwis, 2002). Isolation and treatment should be
applied to any animal showing elevated rectal temperature (Benkirane and De Alwis, 2002). There
are some reports from field outbreaks that support the idea that antibiotic treatment in the final
stages of the clinical illness actually accelerates death (De Alwis, 1999a). However, these reports
are anecdotal. One reason that may help to explain this anecdotal observation is that antibiotic
therapy may precipitate endotoxic shock due to the release of the endotoxin from killed circulating
bacteria (De Alwis, 1999a).
2.7.1 Antibiotic therapy
Sulphonamides administered by intravenous (IV) infusion is the oldest recommended modern
therapy (Carter and De Alwis, 1989). A dose of 1 mL of 33.33% sulfadimidine sodium per 2.3
kilograms body weight IV has been a recommended therapy (De Alwis, 1999a). The convenience of
the intravenous route depends on the animal species involved, the temperament of the individual
animal (fractious animals make IV access difficult) and the skill of the person administering the
antibiotic (De Alwis, 1984). Using the intramuscular route is often more practical. Two other
antibiotics, streptomycin or oxytetracycline, administered intramuscularly, have been found to be
effective (Benkirane and De Alwis, 2002).
De Alwis (1984) reported that resistance to antibiotics in P. multocida strains associated with HS is
not a major problem. However, this information is now 30 years old and several years after this
publication, 19 different antibiotics were tested against 16 isolates obtained from different
58 | P a g e
outbreaks in Gujarat state in India (Bandopadhyay et al., 1991). It was found that 87.5%, 75%, 50%
and 37.5% of these isolates were resistant to sulphonamides, vancomycin, tetracycline and
colistin, respectively (Bandopadhyay et al., 1991). It was also found that 25% were resistant to
streptomycin and novobiocin, while 6.25% were resistant to nitrofurantoin and carbenicillin
(Bandopadhyay et al., 1991). The sixteen isolates were sensitive to penicillin, ampicillin,
chloramphenicol, erythromycin, gentamicin, cephalothin, polymyxin B and tilmicosin
(Bandopadhyay et al., 1991).
De Alwis (1999a) concluded that penicillin, ampicillin and oxytetracycline are recommended for
convenient and economical administration in cattle and buffaloes to treat P. multocida infections.
Whether this use is “on-label” or “off-label”, will depend on the jurisdiction in which the HS-
affected animals are treated. In contrast, the use of amoxicillin and ampicillin was shown not to be
effective when tested in vitro against 16 HS-associated isolates from Lahore (Naz et al., 2012).
The sensitivities to 15 antibiotics were tested against 16 field isolates of HS-associated P.
multocida. The 16 isolates were collected from buffaloes during 2005-2006 from different
locations around Lahore, Punjab, Pakistan. The highest percentage of isolates (87.5%) was found
to be sensitive to ciprofloxacin, ofloxacin, enrofloxacin and gentamicin, followed by 81.25% of
isolates to norfloxacin and amikacin. For kanamycin and tetracycline, 75% and 56.25% of isolates,
respectively, were sensitive, while 50% of isolates were sensitive to chloramphenicol, doxycycline
and vancomycin. Twenty five percent of isolates were sensitive to erythromycin, amoxicillin and
ampicillin. The lowest sensitivity (12.5%) was observed for sulfadiazine. The authors suggested
using ciprofloxacin, ofloxacin and enrofloxacin (Naz et al., 2012).
59 | P a g e
2.7.2 Antiserum therapy
Kheng and Phay (1963) reported no significant therapeutic effect when they administered 60-100
mL of hyperimmune serum to two year old (180 kg) buffaloes at varying periods from six hours
before to 18 hours after experimental inoculation of an HS-associated strain by nasal spray.
2.7.3 Effects of hypertonic saline solution on survival rate of affected buffaloes
Clinical data showed that small volume resuscitation with hypertonic saline solution (HSS; 7-7.5%
NaCl) promoted immediate plasma volume expansion due to mobilisation of fluids from the
intracellular compartments, thus restoring blood pressure, cardiac output and regional blood flow
and improving microcirculation, thereby decreasing inflammatory responses triggered by
endotoxins (Kramer, 2003; Rocha-e-Silva and Poli-de-Figueiredo, 2005).
A clinical trial was performed to investigate the effect of concurrent use of hypertonic saline
solution with an antibiotic (ceftiofur HCl) and a non-steroidal anti-inflammatory drug (NSAID,
ketoprofen) in the treatment of HS in buffaloes and its effect on the survival rate (Zafar et al.,
2010). One group received the antibiotic and the NSAID, while the second group was additionally
treated with the hypertonic saline solution. The trial showed that the group of animals not treated
with hypertonic saline solution was more likely to die within the first 24 hours (13 out of 25
animals = 52%) than the group treated with it (5/25 = 20%). The difference between the two
groups was statistically significant (P < 0.05) (Zafar et al., 2010). The duration of monitoring in this
study was only for one day after starting the treatment while the therapy continued for 5 days. It
is not mentioned what the survival rate was in each group after the five days of treatment.
60 | P a g e
2.8 Prevention and Control
2.8.1 Vaccine types
There are different types of vaccines for HS (De Alwis, 1999a). First, broth bacterins are the
simplest form of vaccine for HS as they are constituted of whole cell broth suspensions (Carter and
De Alwis, 1989). Immunity only lasts for approximately six weeks (Chandrasekaran et al., 1993).
Second, alum-precipitated vaccines are the most commonly used vaccines in Asia and Africa (De
Alwis, 1999a). They produce protective immunity for a period of four to six months (Bain et al.,
1982). Third, aluminium hydroxide gel vaccine is the main type of vaccine used in Thailand
(Neramitmansook, 1993). It shares common properties with the alum-precipitated vaccine and
produces the same immunity period (De Alwis, 1999a). Fourth, oil-adjuvant vaccines use oil in
order to increase the antibody-inducing effect of antigens, delay absorption and they produce
immunity for one year (De Alwis, 1999a).
2.8.2 Prevention and control in endemic areas
Annual vaccination is important in endemic areas, especially 2-3 months before the high risk
season. Moreover, good reporting, husbandry and monitoring systems for at-risk animals are
crucial, especially when an outbreak starts. Awareness campaigns for farmers and quarantine
areas for newly arrived animals can help in controlling the disease (Benkirane and De Alwis, 2002).
2.8.3 Prevention and control during outbreaks
It is recommended to continue vaccination programs, irrespective of previous vaccination history,
by administrating the broth bacterin vaccine or the oil adjuvant vaccine (Benkirane and De Alwis,
2002). Furthermore, a quarantine area for infected animals should be provided and rectal
temperature should be checked regularly in all animals that are at risk. In addition, infected
premises should be cleaned and disinfected and carcasses should be buried correctly or burned in
order to decrease dissemination of the disease (Benkirane and De Alwis, 2002).
61 | P a g e
2.9 Conclusion
In this chapter, the literature that concerns itself with the organism, HS disease, world HS
epidemiology, HS epidemiology in Pakistan, diagnosis, treatment and prevention has been
reviewed. From this review, a series of knowledge gaps can be identified. The current
epidemiological situation in the field in Pakistan is unknown and any independent variables that
are statistically associated with the occurrence of HS outbreaks should be determined. It would be
of interest to compare the level of genomic relatedness of different field and vaccine strains from
Pakistan and another endemic country from the same region, like Thailand. Developing a rapid
diagnostic test for HS-associated strains that will not need sophisticated equipment may help in
controlling the disease. These specific areas of research formed the basis for the aims of this
thesis. The following chapters outline the formal investigations that were performed in the field
and on HS-associated isolates to try and address the aims of this thesis.
62 | P a g e
Chapter 3 – A retrospective case-control study of haemorrhagic septicaemia in Karachi, Pakistan in 2012
3.1 Abstract
A retrospective epidemiological case-control study was performed in Karachi, Pakistan from
January to April 2013. The owners of 217 dairy cattle and buffalo farms from six different locations
in Karachi were interviewed. The aim of the study was to identify risk factors associated with the
presence of HS. Farms with a history of at least one instance of sudden death in a dairy animal
during 2012 and a positive clinical HS diagnosis (made by local veterinarians) were defined as
cases. Farms having no history of sudden deaths in 2012 were defined as controls. Univariable
analyses were initially conducted and factors with a P ≤ 0.25 were offered to a multivariable
logistic regression model to identify putative risk factors. The final multivariable logistic model
contained five factors. Vaccination was found to be a protective factor (OR = 0.22) along with the
length of time cattle were kept on farm. For every extra month cattle were kept, the odds of
disease were reduced by a factor of 0.9. In contrast, for every extra animal, the risk of infection
increased by a factor of 1.01. Supplying underground water and the presence of foot and mouth
disease on the farm increased the risk by a factor of 2.90 and 2.37, respectively. To understand the
epidemiology of HS in Karachi dairy herds more fully, further in-depth research is required to study
the risk and protective factors identified in this survey and to evaluate risk mitigation strategies
where possible. The study also showed the need for developing a point-of-care diagnostic test that
can be used by veterinarians to diagnose the disease on the farm.
3.2 Introduction
According to the Pakistan Economic Survey (2013-2014), the agriculture sector in Pakistan
accounts for 21% of gross domestic product (GDP) and employs 43.7 percent of the nation’s labour
force. The agriculture sector has four sub sectors: crops, livestock, fisheries and forestry. The
livestock sector contributed 55.9% of the value of the agriculture sector and almost 11.8% to the
63 | P a g e
Pakistani GDP during 2013-2014. Livestock are an important source of livelihood in rural
communities, as they provide security in case of crop failure (Farooq, 2014).
The total numbers of cattle and buffaloes during 2013-2014 were estimated as 39.7 and 34.6
million, respectively (Farooq, 2014). According to the Statistics Division of the Government of
Pakistan, 23% and 25% of Pakistan’s cattle and buffaloes, respectively, were in the province of
Sindh (Statistics Division, 2010). Karachi is the largest city in this province and in 2010, the
livestock population was estimated to contain 287,000 cattle and 439,000 buffaloes (Mari, 2013).
Pakistan is the world’s fourth largest milk producer (FAO, 2014) and the gross production of buffalo
and cattle milk in Pakistan during 2013-2014 was estimated at 31 and 18 million tonnes,
respectively (Farooq, 2014).
The risk factors and impact of husbandry practices on the occurrence and spread of HS have not
been well defined. Only a few studies have described HS under field conditions. In a retrospective
analysis of HS outbreaks that occurred in Haryana, India, between 1995 and 1999, some risk
factors for spreading the disease were observed (Jindal et al., 2002). Unvaccinated animals or
those vaccinated with alum-precipitated vaccines, which confer immunity for short periods of
time, were more susceptible to disease. Using a single needle for vaccination, without sterilisation
between animals, during seven outbreaks, was a reason for iatrogenic spread of the disease (Jindal
et al., 2002). Concurrent foot and mouth disease (FMD) viral infection during an outbreak was
thought to produce immunosuppression and was associated with higher levels of mortality
following infection with P. multocida (Jindal et al., 2002). In another HS outbreak in Perak,
Malaysia, in 2003, several risk factors were considered during the outbreak. These included the
immune status of the herd, seasonal variations, free range versus confined housing systems, and
improper burial of carcasses (being left in the communal pond where animals were wallowing and
64 | P a g e
drinking water) (Bisht et al., 2004). Recently, a retrospective HS study in Cambodia showed that
buffaloes had a higher morbidity rate (47.4%) than cattle (28.4%) (P = 0.003) (Kawasaki et al.,
2013). In addition, unvaccinated animals were 2.9 times more susceptible to HS than vaccinated
animals (P = 0.001). It was also shown in this study that the mean cost in an affected household
(having five large ruminants) was USD 952.50 (Kawasaki et al., 2013).
In a participatory study on surveillance for livestock diseases between 2002-2005 in Karachi,
farmers ranked HS as the most important animal health hazard because affected animals usually
died within a very short period of time (Ali et al., 2006). Because of the importance of livestock in
poverty alleviation and the impact of HS on the livestock sector in Pakistan, a retrospective
epidemiological case-control study was performed in Karachi, Pakistan from January to April 2013.
The aim of the study was to identify risk factors that were significantly associated with the
occurrence of HS on farms during 2012. To our knowledge, this is the only case-control study in
this area and the HS-associated risk factors identified in this manuscript are the first in this
production system of dairy colonies in Karachi.
3.3 Materials and Methods
The owners of 217 cattle and buffalo farms were interviewed in six different locations in Karachi:
Bilal colony, Al-Momin society, Nagori society, Surjani dairy colony, Landhi dairy colony (LDC), and
along the national highway (Figure 3.1). Of these farms, most (190) kept both cattle and buffaloes,
with 27 farms only running buffaloes.
65 | P a g e
Figure 3.1 A map of Karachi city showing the six different locations in the study, adapted from the Karachi metropolitan corporation (2014).
A questionnaire was used to collect data on farm practices; the number of buffaloes and cattle
kept; husbandry and management factors, including diet, method of feeding and source of
drinking water; where the animals were purchased; vaccination history; important diseases
affecting the farm; and the history of presumptive cases of HS on the farm. Animals were not
examined in this study. Bias was limited by having a fixed questionnaire with ordered and specific
procedures to be followed. The questionnaire included open and closed questions. It was pre-
tested on a group of ten farm owners, modified slightly and then administered to the randomly
selected farms. Sampling was done by proportional stratified random sampling where farms were
assigned into six districts: Bilal colony, Al-Momin society, Nagori society, Surjani dairy colony,
66 | P a g e
Landhi dairy colony, and along the national highway. Farms were then selected from each district
by systematic random sampling (a farm was selected every two farms). Where owners refused to
be interviewed, the farm was skipped and the next farm was selected. The interviewer (S. Noman
Ali) was a governmental veterinarian who had experience with participatory disease surveillance
(Ali et al., 2006). The questionnaire was approved by the Murdoch University Human Ethics
Committee (2012/223). The questionnaires were only administered over a three month period
due to the political unrest in the city at the time of the surveying. Photos were taken to show the
husbandry conditions in cattle and buffalo herds in Karachi.
For the analysis, farms with a history of at least one instance of sudden death in a dairy animal
during 2012, and a presumed positive clinical diagnosis of HS (made by local veterinarians), were
considered cases. Farms having no history of sudden death in dairy livestock in 2012 were
considered controls.
Data from the questionnaires were entered into Excel 2010 (Microsoft, Redmond, WA, USA) and
analysed with Statistix for Windows (Analytical Software, Tallahassee, FL, USA) and Egret for
Windows (Cytel Software, Ver. 2.0.31). A range of putative risk factors was checked. The
percentage of farms with and without HS cases, and the risk factors and their 95% confidence
intervals, were calculated. Univariable analyses were then conducted using Pearson’s chi-square
test for independence, Fisher’s exact test and odds ratios and their 95% confidence intervals for
discrete variables, and an analysis of variance for continuous variables.
After performing univariable analyses, a multivariable logistic regression model was generated.
Only variables significant at P ≤ 0.25 in the univariable analyses were considered eligible for
inclusion in the logistic multiple regression (Hosmer and Lemeshow, 1989; Frankena and Graat,
67 | P a g e
1997). Backward elimination was used to determine which factors could be dropped from the
multivariable model (Hosmer and Lemeshow, 1989). The likelihood-ratio test statistic was
calculated to determine the significance at each step of the model building. Interaction factors
were checked. The Hosmer-Lemeshow statistic was also calculated to determine the suitability of
the model. Odds ratios with 95% confidence intervals were calculated for the final model that
included variables with significance (P ≤ 0.05).
3.4 Results
Of the 217 farms included in the study, four (1.8%) were excluded in the data entry stage due to
missing data, so analyses were restricted to 213 farms (98.2%). According to the case definition,
there were 66 (31%) and 100 (47%) farms considered as cases and controls, respectively. There
were 47 (22%) farms that did not fit in either the case or control categories so they were called
suspects and excluded from the univariable and multivariable analyses. Figure 3.2 to Figure 3.8
show the husbandry conditions in cattle and buffalo herds in Karachi.
68 | P a g e
Figure 3.2 Local roads are prone to seasonal flooding and this can hinder access to farms.
69 | P a g e
Figure 3.3 A road where a farm is located showing difficulty in access.
70 | P a g e
Figure 3.4 Typical day in holding facilities.
71 | P a g e
Figure 3.5 Foot wear is rarely worn and bare feet are often unclean.
72 | P a g e
Figure 3.6 Dead calf, which was suspected to have died with haemorrhagic septicaemia, had been left with other animals (red arrow).
73 | P a g e
Figure 3.7 Manual milking.
74 | P a g e
Figure 3.8 Flies in milk due to open-air storage.
The 166 farms included in the case-control analyses were from Bilal colony, (27 of the 166 farms,
16.3%); Al-Momin society, (19, 11.5%); Nagori society, (15, 9%); Surjani dairy colony, (12, 7.2%);
LDC, (80, 48.2%); and along the national highway, (13, 7.8%). General results of the study are
shown in Table 3.1. The feeding pattern was similar in both controls and cases (Table 3.2).
75 | P a g e
Factor Number
Average number of holding sheds per farm 3
Number of buffaloes in the 166 farms 22,489
Number of cattle in the 166 farms 4,229
Average number of animals per farm 161
Average number of animals bought per year per farm 163
Number of farms (of 166 farms) visited by governmental veterinarians 24 (14.5%)
Number of farms (of 166 farms) visited by private veterinarian 130 (78.3%)
Number of farms (of 166 farms) that were not visited by any veterinarians 12 (7.2%)
Table 3.1 General results of the study.
Type of feed Controls Cases
Green fodder 87 (87%) 58 (88%) Wheat bran 88 (88%) 52 (79%) Wheat straw 86 (86%) 54 (82%) Wheat grain 61 (61%) 47 (71%)
Cotton seed cake 64 (64%) 45 (68%) Palm seed cake 31 (31%) 13 (20%)
Table 3.2 Feeding pattern in the 100 controls and 66 cases farms.
Common diseases in the 166 farms are shown in Table 3.3. Owners of farms that had experienced
HS reported that FMD was common in their farms. This was compared to control farms, where
FMD was reported by owners to be less common (P = 0.0028). Vaccination against FMD was
regularly done in 70% and 64% of controls and cases farms, respectively.
Disease Controls Cases P value
Foot and mouth disease 36 (36%) 39 (59%) 0.0028
Mastitis 99 (99%) 62 (94%) 0.1
Pneumonia 92 (92%) 54 (82%) 0.054
Toxaemia 80 (80%) 53 (80%) 0.96
Buffalo pox 13 (13%) 6 (9%) 0.44
Table 3.3 Common diseases in the 100 controls and 66 cases farms.
76 | P a g e
Eighty five (51.2%) and 36 (21.7%) farmers purchased their animals from Sindh and Punjab
provinces, respectively. Forty five (27.1%) farmers purchased animals from both provinces. The
two places where farmers were purchasing most of their animals in Sindh and Punjab were
Tando Saindad (45, 27.1%) and Chichawatni (41, 24.7%), respectively. There were 82 (49.4%)
farmers that sent high milk-yielding animals back to their villages (after the finish of lactation) for
recycling.
Regarding awareness about HS, all farmers knew the disease by its local name “Gal Ghotu” and
were able to mention the most common signs of the disease. The most economically important
diseases for the 66 farms defined by us as cases were considered to be mastitis (62%), HS (56%),
pneumonia (27%), toxaemia (6%), foot and mouth disease (FMD) (4.5%) and buffalo pox (1.5%).
For the 100 control farms, the most economically important diseases were considered to be
mastitis (84%), pneumonia (24%), buffalo pox (4%), toxaemia (3%) and FMD (2%). There were 111
farms (79 controls and 32 cases) vaccinating their animals against HS disease; of these, the most
commonly used HS vaccine (41.4%) was the aluminium hydroxide gel vaccine (Neo-bacterina
vaccine, Syva Company, Spain).
The owners of the 66 farms that had HS cases were asked about the months of highest prevalence,
62 (94%) answered December, 54 (81.8%) answered November, 53 (80.3%) answered January and
47 (71.2%) mentioned the three months together. Farmers’ actions if they suspected that their
animals had HS included: calling a veterinarian (85%), separation of affected animals from
apparently-unaffected animals (27.3%), starting antibiotic treatment (10.6%), did nothing (7.6%),
and slaughtering the animals (7.6%). Owners of all 66 farms which had HS cases said that no
samples from dead animals were sent to diagnostic laboratories. The signs and duration of the HS
disease are listed in Table 3.4.
77 | P a g e
Sign Number of farms
Course of disease (2-5 days) 20 (30.3%)
Nasal discharge and salivation 27 (40.9%)
Swelling in neck 33 (50.0%)
Course of disease (24 hours) 46 (69.7%)
Unable to move 54 (81.8%)
Loss of appetite 62 (93.9%)
High temperature 63 (95.5%)
Laboured breathing 65 (98.5%)
Table 3.4 Signs and course of disease seen in haemorrhagic septicaemia cases at 66 farms.
Haemorrhagic septicaemia risk factors
3.4.1 Univariable analysis
A range of putative risk factors such as number of animals on the farm, the use of underground
water, the origin of purchased animals, the presence of FMD, the use of bakery wastes as feed, the
number of buffaloes or cattle bought per year, and the geographical location of the farm, were all
identified. Moreover, putative protective factors including vaccination of animals against HS and
standard period of keeping cattle and buffaloes on farm (in months), were identified.
Statistically significant risk factors (P < 0.05) for farms experiencing cases of HS in the calendar
year 2012 identified in the univariable analysis included using underground water, using bakery
wastes as animal feed, buying animals from local areas in Karachi, and livestock number per farm.
Vaccinating animals for HS, farms located along the national highway and retaining cattle on the
farm for longer periods were identified as protective factors.
78 | P a g e
3.4.2 Multivariable analysis
Seventeen variables (P ≤ 0.25) were included in the logistic regression model to determine risk
factors for experiencing HS cases (Table 3.5 and Table 3.6). The final multivariable logistic model
contained five factors (Table 3.7). No interaction factors were found. Vaccination was found to be a
protective factor (OR = 0.22) along with the length of time cattle were kept on farm. For every
extra month cattle were kept, the odds of disease were reduced by a factor of 0.9. In contrast, for
every extra animal, the risk of infection increased by a factor of 1.01. Supplying underground
water and the presence of FMD on the farm increased the risk by a factor of 2.90 and 2.37,
respectively. The P value for Hosmer and Lemeshow’s goodness-of-fit statistic was 0.6604, so the
model was a good fit.
79 | P a g e
Variable n Farms with HS
cases Farms without
HS cases Odds ratio
(95% CI) P
value
Using underground watera 111 52, 46.85% 59, 53.15% 2.58 (1.27, 5.26) 0.01 Using bakery wastes as feeda 29 17, 58.62% 12, 41.38% 2.54 (1.12, 5.76) 0.04
Vaccinating animals against HSa 111 32, 28.83% 79, 71.17% 0.25 (0.13, 0.49) 0.0001 Presence of FMD in the farma 75 39, 52% 36, 48% 2.67 (1.40, 5.07) 0.0028
Presence of pneumonia in the farma 146 54, 36.99% 92, 63.01% 0.39 (0.15-1.02) 0.054 Using HS vaccine (Neo-bacterina) from Syva Company, Spain 46 15, 32.61% 31, 67.39% 0.65 (0.32, 1.34) 0.29
Using HS vaccine from veterinary research institute (VRI), Lahore, Pakistana 17 4, 23.53% 13, 76.47% 0.43 (0.13, 1.39) 0.19 Using HS vaccine from veterinary research institute (VRI), Tando Jam,
Pakistan 9 3, 33.33% 6, 66.67% 0.75 (0.18, 3.09) 1.0
Using HS vaccine from Nuclear Institute for Agriculture and Biology (NIAB), Faisalabad, Pakistana 17 3, 17.65% 14, 82.35% 0.29 (0.08, 1.06) 0.07
Farms located at Landhi dairy colony (LDC) 80 33, 41.25% 47, 58.75% 1.13 (0.61, 2.10) 0.75 Farms located at Bilal colony 27 13, 48.15% 14, 51.85% 1.51 (0.66, 3.45) 0.39
Farms located at Al-Momin society 19 8, 42.11% 11, 57.89% 1.12 (0.42, 2.94) 0.81 Farms located at Nagori societya 15 9, 60% 6, 40% 2.47 (0.84, 7.31) 0.10
Farms located at Surjani dairy colonya 12 2, 16.67% 10, 83.33% 0.28 (0.06, 1.33) 0.13 Farms located along the national highwaya 13 1, 7.69% 12, 92.31% 0.11 (0.01-0.89) 0.02
Buying animals from Sindh Province 130 50, 38.46% 80, 61.54% 0.78 (0.37, 1.65) 0.57 Buying animals from Punjab provincea 81 28, 34.57% 53, 65.43% 0.65 (0.35, 1.22) 0.21
Buying animals from Tando Saindad area 45 18, 40% 27, 60% 1.01 (0.50, 2.04) 1.0 Buying animals from Arifwala area 29 11, 37.93% 18, 62.07% 0.91 (0.40, 2.08) 1.0
Buying animals from local areas of Karachia 40 22, 55% 18, 45% 2.28 (1.11, 4.69) 0.03 Buying animals from Badin area 19 5, 26.32% 14, 73.68% 0.50 (0.17, 1.47) 0.30
Buying animals from Chichawatni area 41 14, 34.15% 27, 65.85% 0.73 (0.35, 1.52) 0.46 Buying animals from Okara area 32 10, 31.25% 22, 68.75% 0.63 (0.28, 1.44) 0.32 Buying animals from Thatta area 10 4, 40% 6, 60% 1.01 (0.27, 3.73) 1.0 Buying animals from Shahdadpur 22 7, 31.82% 15, 68.18% 0.67 (0.26, 1.75) 0.49
a P ≤ 0.25; risk factor offered to the logistic model.
Table 3.5 Discrete risk factors for farms with (n=66) and without (n=100) haemorrhagic septicaemia cases in 2012 in Karachi, Pakistan.
80 | P a g e
a P ≤ 0.25; risk factor offered to the logistic model.
b This variable was dropped from the logistic model because of high correlations with other variable(s).
Table 3.6 Continuous risk factors for 66 farms where haemorrhagic septicaemia cases were seen in 2012 in Karachi, Pakistan.
Variable Β Coefficient/SE P
value Odds ratios
95% Confidence intervals
Constant 0.60775 0.87435 - - -
Using HS vaccine -1.52285 0.45181 0.0008 0.2181 0.0899-0.5287
Period of keeping cattle on farm (months)
-0.11046 0.03428 0.0013 0.8954 0.8372-0.9577
Livestock number per farm
6.31E-03 1.86E-03 0.0007 1.0063 1.0027-1.0100
Using underground water 1.06489 0.45128 0.0183 2.9005 1.1977-7.0245
Presence of FMD in the farm
0.86212 0.41947 0.0399 2.3682 1.0408-5.3886
Table 3.7 Variables included in the final logistic regression model
3.5 Discussion
Information collected in this study should be interpreted and extended to rural and nomadic
conditions with caution as this study was performed in Karachi where the animals are reared
intensively and for commercial purposes.
Variable Mean in
cases Mean in controls
P value
The length of time cattle were kept on farm (months)a 18.74 22.17 0.003
The length of time buffaloes were kept on farm (months)a 11.54 10.65 0.12
Number of animals per farma 187 144 0.03 Number of buffaloes bought per farm per yeara 164 136 0.13
Number of cattle bought per farm per yearb 24 12 0.02 Number of animals bought per farm per yeara 187 147 0.07
81 | P a g e
This study was done in the unique system of dairy colonies in Karachi (Mari, 2013). Karachi has
many large and small dairy cattle colonies with an aggregate buffalo and cattle population of
roughly 800,000. The LDC alone has a cattle and buffalo population of roughly 400,000 at the one
location (> 95% buffaloes) (Khan et al., 2008; Klein et al., 2008; Afzal et al., 2012). LDC is the
largest buffalo colony in the world (Klein et al., 2008). It produces four million litres of milk daily
(Khan et al., 2008). The other dairy colonies, like Bilal colony, Al-Momin society, Nagori society and
Surjani dairy colony, have an aggregate of 400,000 cattle and buffaloes between them (Khan et al.,
2008). In each of these colonies, farms directly neighbour each other, thus, epidemiologically and
practically, they are considered one production unit (Afzal et al., 2012). The system of rearing
animals in these colonies is different from conventional systems. In this production system, herd
size is relatively large; mostly 50-500 in a farm with more than 90% lactating adult buffaloes (Afzal
et al., 2012). The reason for the dominance of buffalo milk in this production system is that
customers prefer the higher butterfat it contains compared to cows’ milk (Khan et al., 2008). High
yielding animals are kept for milk production and the turn-over rate of animals is high (low-
yielding animals are removed from the farms) at 10-12% per month (Khan et al., 2008; Klein et al.,
2008; Afzal et al., 2012). Most animals stay for only one lactation. Freshly-calved animals are
brought in at the start of lactation while at the end of lactation they are sold for slaughter and only
a few are kept for re-breeding (Klein et al., 2008; Afzal et al., 2012). The continuous influx of
animals with unknown disease status, the stress of over-crowding in farms, and animals coming
from different sources (from Sindh and Punjab) all play important roles in spreading diseases and
hindering their control.
These factors provide a favourable environment for the spread of HS, as evidenced by the high
number of HS cases on the farms we surveyed (66 out of 166 farms, 39.8%) which were identified
by local veterinarians during the calendar year 2012.
82 | P a g e
A limitation in our sampling technique was that some owners refused to be interviewed because
they were busy or not interested in what we were doing. Another limitation in this study was that
the case definition was based on a clinical diagnosis made by local veterinarians. Although the
veterinarians were experienced, these diagnoses were lacking confirmation by laboratory testing.
It is therefore possible that some diseases were misclassified. Consequently, extrapolation of
these results should be done cautiously as there may have been selection and information bias.
Farmers reported that HS attack rates peaked in December, January and February, which is similar
to a study done in Karachi in 2006 (Ali et al., 2006). It was also reported in India that more
outbreaks occurred in winter (January to March) (Jindal et al., 2002). This is contrary to what has
been reported in Pakistan, where HS outbreaks were more common during the wet season (July-
September) (Sheikh et al., 1996). Farmers who had observed HS cases in their farms said that HS
and mastitis were the most economically important diseases in their farms. Control owners said
that mastitis was the most economically important disease in their farms. This reflects the effect
of experiencing the severity of the disease and is similar to the results from the participatory
disease surveillance study conducted in 2006 (Ali et al., 2006). Cases and controls were both
affected by the presence of mastitis, pneumonia, toxaemia and FMD. These diseases could
contribute, along with other stressors, to immunosuppression and therefore increase the
vulnerability to HS.
Univariable analysis revealed that using bakery wastes as animal feed increased the risk for HS by
a factor of 2.54. According to veterinarians in the field, this could be attributed to fungus
contamination and its effect on immunity. This should be further studied. The univariable analysis
also showed that farms located along the national highway (border of the city) were at lower risk
83 | P a g e
of having HS (OR = 0.11). It was interesting that farmers who purchased buffaloes and cattle from
local areas in Karachi were more likely to get the disease than other farms by a factor of 2.28.
However, we do not know if the prevalence of HS is higher in Karachi than other places in
Pakistan.
The logistic regression model revealed some potential risk and protective factors associated with
experiencing HS cases. Farms with HS cases were 2.90 times (95%CI 1.20-7.02) more likely to be
using underground water. Underground water is present under the soil (aquifer) and it is drawn by
a power motor from underground bores. The water quality varies in mineral content and
contamination according to the topography of the land.
Few studies have documented the impacts on groundwater quality of burying dead animals.
However, a recently published study showed that leachate from carcass burial sites represented a
potential source of nutrients, organics, and residues of biologically active micro-contaminants to
soil and groundwater (Yuan et al., 2013). The improper shallow burying of carcasses could work as
a potential risk factor for disseminating HS. Moreover, during our survey in Karachi, we noticed
that carcasses were thrown on the sides of canals and along the roads which could increase the
dissemination of HS. These actions could contribute to the higher risk of having HS cases at farms
using ground water. It is also worth mentioning that the emergence of an HS outbreak in
dromedary camels in the Greater Cholistan Desert in Pakistan at the end of 2010 was attributed to
drinking from rainwater reservoirs (Khan, 2012). In another study in the same area, the highest
prevalence of HS disease was in a canal-irrigated area of Rahim Yar Khan district in Punjab (Khan,
2010).
84 | P a g e
Practical solutions should be implemented to supply safe water to cattle and buffaloes in farms.
Using water supplied by the Water Board could be one solution as the water is treated and passed
through water filtration systems. The disadvantage of this solution is the cost of infrastructure
needed to put all the farms onto the national water supply network. Another solution could be
treating the underground water to kill off any microorganisms before it is supplied to the cattle.
Choosing an appropriate technique to treat water such as UV light or filtration systems could help.
Treating underground water for such a large number of animals may be impractical.
In our study, FMD was observed more in cases than controls (according to owners). The presence
of FMD in a farm increased the risk of HS by 2.37 (95% CI 1.04-5.39). In an epidemiological study of
HS in all districts of Haryana state, India, 67% of 237 surveyed veterinarians reported that HS
outbreaks generally followed FMD outbreaks in the same animals. The correlation between HS
and FMD in this study was statistically significant (P < 0.05) (Subash et al., 2004). This association
between FMD infection and HS was also reported in an outbreak in India in 1997 (Jindal et al.,
2002). This interaction should be further studied in the future to show if FMD biologically
predisposes for HS or is merely just a surrogate marker for poor hygiene and/or biosecurity.
Another factor that increased the risk of experiencing HS cases was the number of animals at the
farm. The odds of the disease increased by a factor of 1.01 (95% CI 1.0027-1.0100) for every extra
animal added at the farm. This means that for one extra animal, the new odds were 1.01 raised to
the power of one and for 100 animals, the new odds were 2.70 (1.01100). The explanation of why
this causes an increased risk could be that new infections are introduced. However, the number of
animals in a farm could be also a surrogate marker for the suboptimal husbandry practices in large
animal herds while herds with fewer animals were, in most cases, better managed. These results
agree with what has been published before by De Alwis and Vipulasiri (1980) that HS commonly
85 | P a g e
existed in large and free-roaming herds, where it was generally rare in smaller, well-managed and
stall-fed herds. For example, in Sri Lanka, it was observed that the incidence of HS in herds with
over 50 animals was 4-5 times greater than that of herds with less than ten animals. Generally, as
the number of animals increased in the farm, the density (animals per unit of area) became higher
and the probability of disseminating the disease was higher in the farm.
In our study, vaccinating animals was a protective factor for HS. Farms where vaccination was
carried out were at a lower risk of getting the disease (OR = 0.22, 95% CI 0.09-0.53). Vaccination is
important for prevention of HS disease as it is a primary infection (Carter and De Alwis, 1989; De
Alwis, 1993a; De Alwis, 1999a; Benkirane and De Alwis, 2002). This was also reported in HS
outbreaks in Haryana, India, and Cambodia, where the disease was observed more commonly in
unvaccinated animals (Jindal et al., 2002; Kawasaki et al., 2013). In our survey, we noticed that
although vaccination was carried out in 67% of farms (111 of 166), 40% of farmers (44 of 111)
vaccinating their animals were not satisfied about the effectiveness of vaccines. In this study,
antibody titres for P. multocida in vaccinated and unvaccinated animals were not measured. It was
reported previously that antibody titres in vaccinated buffaloes and cattle were 3 and 4 times,
respectively, more than those in unvaccinated animals (Khan et al., 2006). The dissatisfaction of
40% of farmers with effectiveness of HS vaccines should not be over-interpreted as immune status
of vaccinated and unvaccinated animals was not recorded. However, improper storage of
medicines and vaccines in medical stores in Karachi could be one reason for poor vaccine efficacy.
It is also possible that the discrepancy between the statistically significant efficacy and the
farmers’ opinions may be due to high expectations that some farmers had on how protective the
vaccine would be. Serotype B:2 vaccines were used by the majority of the 111 farms (58.6%) in our
survey. However, of all the vaccines that were used, the most commonly used vaccine was Neo-
bacterina vaccine (Syva Company) (41.4% of 111 farms vaccinated against HS). This vaccine
86 | P a g e
contains P. multocida serogroup A, strains P-1062 and 11-A and is marketed as being protective
against HS, despite not using a strain of serotype B:2. This was not necessarily a problem as it has
been shown in other host species that cross protection between different serotypes of P.
multocida can occur. For example, antiserum against a fowl cholera strain (serotype A:5) was
previously reported to protect against strains capable of causing HS (Rimler, 1996). And further to
this, our data showed that there were no statistically significant differences between any of the
individual vaccines that were used for HS prevention on the surveyed farms. It would be of
interest to compare the efficacy of this vaccine with local HS vaccines that are produced by
veterinary research institutes from different provinces.
Another protective factor was the duration of keeping cattle at the farm. It was found that, for
every extra month animals were kept, the odds of disease were reduced by a factor of 0.9 (95% CI
0.84-0.96). Keeping animals on a farm for a longer time means that the replacement rate is less
and the number of animals entering and leaving the farm is less, which may be protecting the farm
from getting diseases from newly arrived animals. Farms with more sick and low-yielding milk
animals are also more likely to turn over their animals more frequently. These animals are
removed from the farm and then new animals are brought in which could then be the source of
new infections.
To understand the epidemiology of HS in Karachi dairy herds more fully, further research is
required to study the risk and protective factors identified in this survey and to evaluate risk
mitigation strategies where possible. It is also necessary to compare the field strains to the vaccine
strains at a molecular level to see if any differences can explain the percentage of farmers using
vaccines that still experienced HS in their herds. Moreover, it is important to investigate the effect
of improper storage conditions on vaccine stability and efficacy. Vaccine administration techniques
87 | P a g e
should also be carefully observed. This is an important consideration that may explain the
percentage of farmers that were unsatisfied with their HS vaccines (13.25%). The study also
showed that on affected farms, clinical samples were never sent to diagnostic laboratories. In
these cases, diagnosis was only done by a veterinarian based on clinical signs. This raises the need
for developing a simple, reliable, point-of-care diagnostic test that can be used by veterinarians to
diagnose the disease on the farm.
88 | P a g e
Chapter 4 – Molecular typing of haemorrhagic septicaemia-
associated Pasteurella multocida isolates from Pakistan and
Thailand using multilocus sequence typing and pulsed-field gel
electrophoresis
After the results obtained from the case-control study and understanding the husbandry system at
farms in Karachi and identifying risk factors associated with the presence of the disease, there was
a need to study field and vaccine HS-associated strains at a molecular level to look for relatedness
and variations, if present. The relatedness of vaccine and field strains from Pakistan was compared
to field strains from Thailand, which represents another country in the Southeast of Asia where HS
is endemic.
89 | P a g e
90 | P a g e
91 | P a g e
92 | P a g e
93 | P a g e
94 | P a g e
As discussed in chapter 4, two different molecular typing techniques were used, but they were
unable to differentiate isolates from the same country. Whole genome sequencing was therefore
selected as a method that could potentially provide much more molecular data about these
isolates. Comparing the genomes of these isolates with non-HS strains could result in identifying
shared sets of genes between HS-associated strains. Consequently, a diagnostic test could be
developed, which was one of the aims of the project.
95 | P a g e
Chapter 5 – Comparative genomics analysis of Asian Haemorrhagic
Septicaemia-associated strains of Pasteurella multocida
5.1 Abstract
In this molecular epidemiological study, the genetic diversity of Pakistani and Thai strains
associated with HS was investigated. Initially, 24 field isolates and vaccine strains from different
regions of Pakistan and Thailand were differentiated by MLST and PFGE (Chapter 4). Based on
these analyses, 12 HS-associated strains were selected for complete genome sequencing. The
sequenced strains had draft genomes of between 2,298,035 and 2,410,300 bp in length. Analysis
of the genomes identified a core set of 1,824 genes that were also shared by the P. multocida
strains Pm70, 3480, 36950, HN06 and M1404. A set of 96 unique genes was found to be shared by
all HS-associated strains but absent from the four non-HS strains: Pm70, 3480, 36950 and HN06.
Moreover, 59 genes were found to be shared only by Asian B:2 strains. Some genetic regions with
high identity to the integrative and conjugative ICEPmu1 element were identified in two Pakistani
strains along with a range of putative antimicrobial resistance genes. Four different putative
temperate phages were identified using the phage search tool (PHAST). Phylogenetic analysis
indicated that there was a correlation between genetic relatedness and country of isolation.
Analysis of the 96 genes shared uniquely by HS-associated strains may aid the identification of
virulence genes responsible for HS disease and enable the development of specific diagnostic tests
for HS strains.
5.2 Introduction
The first complete genome sequence of a P. multocida strain (Pm70; GenBank accession number
AE004439) was determined in 2001 (May et al., 2001). Analysis of the Pm70 genome identified
more than 100 genes predicted to be involved in virulence.
96 | P a g e
There are currently 25 publically available complete or draft genomes of P. multocida. These
genomes are from strains collected from different hosts and cause a range of different diseases
(http://www.ncbi.nlm.nih.gov/genome/genomes/912). There have been only limited comparative
analyses of these different genomes. However, analysis of the Pm70, 36950, 3480, HN06, X73, and
P1059 genomes identified a unique 18 kbp region in the toxinogenic strain HN06 carrying 14
genes, including the toxA gene for P. multocida toxin (PMT) (the toxin responsible for atrophic
rhinitis) as well as several phage-related genes (Wilson and Ho, 2013). Further analyses showed
that the integrative conjugative element (ICE) ICEPmu1 was found in strain 36950 but not in the
other strains. This element carried 11 different antibiotic resistance genes. A similar ICE was found
in Histophilus somni strain 2336 and Mannheimia haemolytica strain PHL213. Interestingly, these
three strains with similar ICEs are all bovine respiratory pathogens (Wilson and Ho, 2013).
Analysis of five of the genomes (M1404, Pm70, 36950, X73, and P903) identified a core set of 1786
genes (88% of Pm70 gene content) common to all strains and the pan genome was more than
2,800 genes. Furthermore, each of these strains contained between 90 and 261 unique genes not
found in any of the other strains (Boyce et al., 2012). For strain 36950, more than 47% of unique
genes were within the ICEPmu1 element, whereas for M1404, 28% of unique genes were phage-
derived elements. Importantly, phylogenetic comparison of nine strains indicated little correlation
between phylogenetic relationship and serovar, diseases caused, host predilection or place of
isolation (Boyce et al., 2012). However, other work has shown that a single clonal complex of P.
multocida isolates were associated with bovine respiratory disease. The isolates were
characterised by MLST but this technique only compares seven housekeeping genes, the whole
genome and virulence genes are not considered (Hotchkiss et al., 2011).
97 | P a g e
In Chapter 4, the genotypes of 23 P. multocida isolates recovered from HS cases in cattle and
buffaloes from different geographical areas and climate zones of Pakistan and Thailand were
analysed. All isolates were indistinguishable by MLST and only a single band difference existed
between Pakistani and Thai isolates by PFGE (Moustafa et al., 2013). Therefore, we chose to
determine whole draft genomes of these strains using next generation sequencing (NGS) to
determine whether there was any diversity across these isolates.
Here we present the draft genomes and bioinformatic analyses of twelve HS-associated B:2 strains
of P. multocida collected from Thailand and Pakistan. We also compare the genomes of these
twelve isolates with M1404 (a North American HS-associated strain) and the four available
complete genomes (as of November 2013), Pm70 (accession number AE004439, (May et al.,
2001)), 36950 (CP003022, (Michael et al., 2012a)), 3480 (CP001409) and HN06 (CP003313, (Liu et
al., 2012)). To our knowledge, this is the first detailed genomic analysis of multiple HS-associated
isolates of P. multocida.
5.3 Materials and Methods
Ten P. multocida strains were collected from HS cases from different regions of Pakistan and
Thailand. In addition, two vaccine strains from Pakistan were used for this study (Table 5.1).
Except for the Faisalabad isolate (Table 5.1), all of the isolates had previously been identified as
P. multocida strains and typed using MLST (Moustafa et al., 2013). The Faisalabad isolate was
received later from the National Veterinary Laboratory, Islamabad, Pakistan. Its identity was
confirmed using P. multocida-specific and HS-specific multiplex PCR (data not shown) (Townsend
et al., 1998).
98 | P a g e
5.3.1 Whole genome sequencing
Genomic DNA was purified from each of the twelve P. multocida isolates using the Qiagen DNeasy
blood and tissue kit (Qiagen Cat# 69504) using 5 mL of overnight cultures grown at 37 ˚C in brain
heart infusion (BHI) broth (Oxoid, UK) and following the Gram-negative bacterial protocol outlined
in the manufacturer’s instructions. DNA quantification and purity analysis were performed by
agarose gel electrophoresis and Qubit Fluorometry (Life technologies, USA). The purified genomic
DNA was sequenced using the paired-end sequencing protocol on an Illumina HiSeq 2000
according to the manufacturer’s protocols (Illumina Inc., San Diego, USA). Sequencing was carried
out at the Beijing Genomics Institute (BGI), China. The resultant raw sequence data were filtered
to eliminate low quality data in the following manner; all reads with > 40% low quality (Q20) bases
(parameter setting at 36 bp), > 10% Ns (parameter setting at 9 bp), > 15 bp overlap with adapter
sequences (parameter setting at 15 bp), or any reads showing duplication contamination were
removed.
99 | P a g e
Isolate Strain
abbreviation Host Year Country Province District
Location coordinates
Thailand A THA Buffalo 2006 Thailand Nakhon Si
Thammarat Thung song
8.16N, 99.68E
Thailand D THD Buffalo 2009 Thailand Chonburi Phanat Nikhom
13.45N, 101.18E
Thailand F THF Buffalo 2011 Thailand Lamphun Mueang Lamphun
18.58N, 99.02E
Attock ATTK Cattle 2010 Pakistan Punjab Attock 33.91N, 72.31E
Bhakkar BUKK Cattle 2008 Pakistan Punjab Bhakkar 31.63N, 71.07E
Taxila 1 TX1 Buffalo 2012 Pakistan Punjab Rawalpindi 33.75N, 72.79E
Karachi 3 Karachi Buffalo 2011 Pakistan Sindh Karachi 24.86N, 67.01E
Islamabad 1
Islm Wild
Buffalo 2011 Pakistan
Islamabad capital
territory Islamabad
33.72N, 73.07E
Peshawar Pesh Buffalo 2011 Pakistan Khyber
Pakhtunkhwa Peshawar
34.02N, 71.58E
Peshawar vaccine
PVAcc NA 2011 Pakistan Khyber
Pakhtunkhwa Peshawar
34.02N, 71.58E
Lahore Vaccine
V1 Buffalo 2011 Pakistan Punjab Lahore 31.55N, 74.34E
Faisalabad Faisal Buffalo 2011 Pakistan Punjab Faisalabad 31.418N, 73.079E
Table 5.1 Haemorrhagic septicaemia-associated strains of Pasteurella multocida used in this study.
5.3.2 Sequence assembly
The genomes of ten of the strains were assembled using SPAdes (Bankevich et al., 2012) and the
remaining two strains were assembled using Velvet (Zerbino and Birney, 2008). For ten of the 12
strains, the SPAdes procedure generated acceptable assemblies. However, for the Faisal and ATTK
strains, SPAdes produced unexpectedly large genome assemblies (8,278,703 and 6,134,337 bp for
Faisal and ATTK, respectively) due to low level contamination with genomic DNA from other
bacterial species (Bacillus cereus and Bacillus subtilis, respectively). Examination of the Velvet
statistics showed that the contaminating sequences were represented by short contigs having
sequencing depth below 10x. To filter this contamination, we assembled the Faisal and ATTK
100 | P a g e
sequences using Velvet with a manual setting of 10 for the minimum required coverage (“velvetg -
cov_cutoff 10”) to remove these undesirable components of the assembly graph prior to repeat
resolution and contig extraction. Final genome assemblies of these strains were also checked for
contaminating sequences using BLASTN v2.2. 26 (Altschul et al., 1990). To evaluate the accuracy of
the generated contigs of the 12 assembled genomes, they were compared with the Pm70
reference genome (May et al., 2001) using QUAST v2.3 (Gurevich et al., 2013); sequence and
assembly statistics of the 12 genomes are given in Table 5.2. Scaffolds (or contigs for the Velvet
assembled genomes) of less than 200 bp in length were removed before the final scaffolds were
reordered using Mauve (Rissman et al., 2009) with Pm70 as the reference sequence. The final
reordered scaffold sequences were then annotated using the NCBI Prokaryotic Genome
Annotation Pipeline (Angiuoli et al., 2008). The 12 annotated genomes were submitted to
Genbank (Benson et al., 2013).
Strain Sequence
yield
(Mb)
Number
of
contigs
Largest
contig
(bp)
N50 N75 Sequence
coverage1
(%)
Average
sequence
depth
ATTK 251 44 373941 265021 106344 92.17 96
BUKK 251 42 589995 289440 106424 91.95 110
Faisal 250 52 374481 265074 106436 92.15 68
Karachi 251 77 647631 289467 111128 92.17 100
Islm 252 35 647631 289467 111128 92.17 100
Pesh 250 35 647631 289467 111128 92.17 100
PVAcc 252 41 647631 289467 111128 92.17 100
THA 251 33 594579 289958 111130 92.19 100
THD 252 35 594578 289367 111130 92.19 100
THF 250 32 549574 290056 111130 92.18 100
TX1 250 40 635052 289440 106424 92.16 100
V1 251 32 647631 289395 111128 92.17 100
1Sequence coverage of the 12 assembled genomes is given relative to Pm70 (reference genome).
Table 5.2 Sequencing and assembly statistics for the genomes of the 12 Asian HS-associated strains
101 | P a g e
5.3.3 Genome alignments and feature analysis
The annotated assembled genomes of the nine Pakistani and three Thai strains were aligned with
M1404, Pm70, HN06, 3480 and 36950 using Mauve progressive aligner using default settings
(Darling et al., 2010). The homologs table in Mauve was used to identify colinear orthologues
across each of the genomes using 50% identity and 50% gene coverage as the minimum criteria for
a match. All genes identified in only a single strain were then manually checked by BLASTN
(Altschul et al., 1990) to confirm that they were unique. Genes unique to the Pakistani or Thai
strains, or unique to M1404, were also identified by this method.
The PHAge Search Tool (PHAST) (Zhou et al., 2011) was used to identify the positions of putative
phage elements in all genomes. For PHAST analysis, the FASTA files containing all concatenated
contigs for each genome were uploaded to the public PHAST web server
(http://phast.wishartlab.com/). To avoid false positives caused by non phage-related mobile
genetic elements, PHAST filters out these mobile genetic elements using a two-step process. In the
first step, PHAST uses a customized mobile genetic element database to directly filter out some of
the most typical mobile genetic elements (Y. Zhou, personal communication). In the second step,
PHAST filters out the rest of the mobile genetic elements when it identifies potential prophages by
the density-based spatial clustering of applications with noise (DBSCAN) algorithm (Zhou et al.,
2011). Specifically, the DBSCAN algorithm marks out the random mobile genetic elements as noise
and clusters other gene elements into potential prophages (Y. Zhou, personal communication). In
addition, PHAST predicts potential prophages based on a number of factors including the relative
density of identified prophage-like genes, GC ratio, functional completeness and gene similarity to
already known phages. The 13 HS-associated B:2 genomes (the 12 strains from Table 5.1 and
M1404) were checked for the presence of antimicrobial resistance genes using the Resfinder tool
(Zankari et al., 2012).
102 | P a g e
The capsule biosynthesis and lipopolysaccharide outer core biosynthetic loci were identified by
BLASTN (Altschul et al., 1990) comparison against the previously reported M1404 capsule
biosynthesis (20,418 bp) and outer core lipopolysaccharide biosynthesis (4,887 bp) gene clusters.
The presence and integrity of the two heptosyltransferase genes, hptA and hptB was checked in
order to predict whether the two different, simultaneously expressed lipopolysaccharide inner
core structures (glycoform A and glycoform B) were present (Harper et al., 2007). In addition, the
twelve strains were presumptively classified into either Heddleston serotype 2 or 5 by analysis of
the lpt-3 gene, required for the addition of phosphoethanolamine (PEtn) to the 3 position of the
second heptose (HepII) (St Michael et al., 2009).
5.3.4 Phylogenetic trees
The phylogenetic relationship between the strains was predicted by analysis of core genome single
nucleotide polymorphisms (CG-SNPs). Identification of CG-SNPs and phylogenetic analysis was
assessed using the Wombac v1.2 (https://github.com/tseemann/wombac). The four closely
related species, Gallibacterium anatis, Mannheimia haemolytica, P. bettyae and P. dagmatis were
used as outliers. Different P. multocida strains were used also including X73 (accession number:
AMBP00000000) (Abrahante et al., 2013), Anand1C (ALBY00000000), Anand1B (ALBX00000000),
Anand1P (AFRR00000000) (Ahir et al., 2011), Anand1G (AFRS00000000), P1059 (AMBQ00000000),
P52VAC (ALBZ00000000), VTCCBAA264 (ALYC00000000) (Vaid et al., 2014), and VP161 (Boyce et
al., 2012). Unrooted neighbour-joining phylogenetic trees were rendered for visualisation using
SplitsTree program (Huson and Bryant, 2006).
5.4 Results and Discussion
Genome sequencing of 12 Pasteurella multocida haemorrhagic septicaemia strains
Ten P. multocida strains from HS cases from various regions of Pakistan and Thailand and two
vaccine strains from Pakistan were used for this study (Table 5.1). Eight of the disease-associated
103 | P a g e
isolates were from HS cases in buffalo and two were from HS cases in cattle. Except for the
Faisalabad isolate (Table 5.1), all of the isolates had previously been typed using MLST and all
belonged to sequence type 122 (Moustafa et al., 2013). Faisalabad isolate was subsequently typed
using MLST and it too belonged to ST 122. Therefore, in order to identify if there were differences
between the strains, whole genome draft sequences of each strain were needed. Sequences were
assembled de novo, resulting in between 32 (strains THF and V1) and 77 (strain Karachi) contigs of
> 200 bp.
The genomic features and Genbank accession numbers of the 12 sequenced strains are shown in
Table 5.3. The predicted genome sizes ranged from between 2,344 and 2,458 kbp and the number
of coding sequences (CDS) between 2,082 (for strains THA, THD and THF) and 2,179 (for strain
TX1). The GC content was highly conserved across all strains (40.31 to 40.41%). The number of
rRNA and tRNA features varied between the different strains. Each strain contained four rRNA
clusters with the exception of the Karachi, ATTK and Faisal strains which had five, ten and 13,
respectively. The strains encoded between 46 (Faisal) and 51 (Karachi, Islm, Pesh and THF) tRNA
genes.
Sample name Genome Size (kbp) Number of CDS tRNA rRNA Accession number
ATTK 2,399 2,141 47 10 JQEA00000000 BUKK 2,410 2,148 50 4 JQAO00000000 Faisal 2,405 2,143 52 13 JQEB00000000
Karachi 2,419 2,164 51 5 JPHI00000000 Islm 2,396 2,142 51 4 JQAB00000000 Pesh 2,397 2,144 51 4 JQAC00000000
PVAcc 2,399 2,140 50 4 JQAD00000000 THA 2,344 2,082 50 4 JQAE00000000 THD 2,344 2,082 50 4 JQAF00000000 THF 2,344 2,082 51 4 JQAG00000000 TX1 2,458 2,179 50 4 JQAH00000000 V1 2,396 2,137 50 4 JQAI00000000
Table 5.3 Genomic features of the 12 Asian strains
104 | P a g e
Genetics of capsule and LPS biosynthesis in the HS-associated strains
As expected, all strains contained the type B capsular biosynthetic locus. There were only three
nucleotide changes observed in the capsule biosynthetic locus of all the Asian strains compared to
the previously reported 20,418 bp M1404 locus (Boyce et al., 2000a). Two of these were silent
mutations, while the third encoded a missense mutation (D50Y) within the putative
glycosyltransferase EpsJ. An additional nucleotide change in a non-coding region was observed in
the Pakistani strains (position 1185274 in BUKK strain).
All strains also contained the LPS outer core biosynthesis locus that is shared by type 2 and 5
strains (St Michael et al., 2009). Only a single nucleotide change was observed across this region of
the genome compared with the reported M1404 locus (St Michael et al., 2009). Heddleston type 2
and 5 strains are differentiated serologically by the presence (type 5) or absence (type 2) of a PEtn
on the HepII of the LPS (St Michael et al., 2009). Addition of this residue to the LPS is dependent
on the presence of an active lpt-3 (dcaA in Pm70) gene. All strains contained an lpt-3 gene with a
nonsense mutation (position 499751 in BUKK strain), so would be predicted to produce LPS
without the addition of PEtn on the HepII. Thus, all strains in this study were genetically B:2, as
expected for Asian HS-associated isolates. All of the strains had been serologically characterised
previously as B:2, except for THD which had been reported as B:2,5 (P. Pathanasophon, personal
communication, June, 2012). However, our genetic analyses indicated that THD was type B:2.
It has been shown previously that most P. multocida strains produce two inner core LPS
glycoforms designated A and B (Harper et al., 2007). Production of these two glycoforms is
dependent on the presence of two active heptosyltransferases, HptA and HptB. HptA is specific for
the addition of the HepI to the inner core glycoform A which contains a single phosphorylated 3-
deoxy-D-manno-octulosonic acid (Kdo), while HptB is specific for the inner core glycoform B which
105 | P a g e
contains two Kdo residues (Harper et al., 2007). The hptB gene was identified in all of the strains
and encoded an intact HptB. The hptA gene was also identified and encoded an intact HptA in all
strains except 3480 where it contained a premature stop codon (mutation at 2232375 in 3480
strain). Therefore, we would predict that strain 3480 expresses LPS with a fully extended double
Kdo inner core glycoform, but with a truncated phosphorylated Kdo glycoform. Interestingly, a
P. multocida strain VP161 (serovar A:1) hptA mutant was significantly attenuated for virulence
(Harper et al., 2007).
Phylogeny
The relatedness of the different strains was determined by comparing all nucleotide changes in
conserved genetic positions (core genome SNPs). Firstly, all the P. multocida strains were
compared with the closely related species P. dagmatis, P. bettyae, Gallibacterium anatis and
Mannheimia haemolytica. This analysis, using 789 SNPs, clearly showed that all the P. multocida
strains form a monophyletic group which is most closely related to P. dagmatis (Figure 5.1A).
Secondly, comparison of the P. multocida strains using 7,892 core SNPs (Figure 5.1B), indicated
that the P. multocida HS-strains were very closely related and clearly separated from all of the
other P. multocida strains. This is in contrast to previous analyses using a smaller number of strains
that showed little or no correlation between the phylogenetic relatedness of the strains with
serovar, disease type or host predilection (Boyce et al., 2012). However, there was still no clear
correlation with strain relatedness and disease other than for the HS-associated isolates. However,
this may be due to the fact that each of the other diseases was represented by only a small
number of strains.
Finally, we compared just the HS-associated isolates using 722 core SNPs (Figure 5.1C). Strain
M1404 (the North American isolate) was clearly separated from the Thai and Pakistani isolates,
106 | P a g e
which were also clearly separated from each other. Therefore, this analysis of the thirteen B:2 HS-
associated P. multocida strains, clearly shows that there is a correlation between strain
relatedness and the country of isolation. This is in contrast to our initial analyses using MLST which
were not able to differentiate these strains as they were a single sequence type (ST 122)
(Moustafa et al., 2013). Thus, these strains have subtle genomic differences but these changes are
outside the genes used for MLST.
107 | P a g e
Figure 5.1 Unrooted neighbour-joining tree showing the phylogenetic relationship between various strains. (A) Relationship between Gallibacterium anatis, Mannheimia haemolytica, Pasteurella dagmatis, P. bettyae and the P. multocida strains Pm70, 36950, HN06, P3480, X73, VP161, Anand1C, Anand1B, Anand1P, Anand1G, P1059, P52Vac, VTCCBAA264, M1404 and the nine Pakistani and three Thai isolates. (B) Relationship between all the P. multocida strains. (C) Relationship between the HS-associated P. multocida B:2 strains M1404 and the twelve Pakistani and Thai isolates. Phylogenetic relatedness for all comparisons was determined by analysis of only the single nucleotide polymorphisms (SNPs) found at conserved positions in all genomes of the comparison set; 789 shared positions for the tree in panel A, 7,829 shared positions for the tree in panel B and 722 shared positions for the tree in panel C. Trees were rendered using SplitsTree v4.11.3 (Huson and Bryant, 2006). The line segments above the trees with the number '0.01' indicate the length of branch that represents an amount genetic change of 0.01.
108 | P a g e
Core and pan genome predictions
We analysed the gene content of each of the 12 sequenced HS-strains and compared these
predictions with the coding sequences predicted for the four complete and annotated
P. multocida genomes (36950, Pm70, 3480, HN06) and the M1404 draft genome. These analyses
identified a shared set of 1,824 genes (core) and a pan genome of more than 2,700 genes
(Figure 5.2). Furthermore, each of the strains contained between zero (Pesh, THA, THD, THF, Islm,
V1, Faisal, ATTK) and 116 unique genes (HN06) not found in any of the other strains (Figure 5.2).
The lists of the unique genes in each of the four strains, TX1, Karachi, BUKK and PVAcc, are
provided in the Appendix (Table A.2, Table A.3, Table A.4 and Table A.5).
109 | P a g e
Figure 5.2 Flower plot showing core and unique genes across all strains. The central circle shows the number of genes common to all strains while the petals show the number of strain-specific genes, and in parentheses the number of unique genes compared to all other strains in the analysis. The strain names are given outside each petal and the strain details are given in Table 5.1. The B:2 strains are shaded in colour while the non-B:2 strains are shaded in grey. Of the B:2 strains, M1404 is shaded in orange, the Thai strains are shaded in blue and the Pakistani strains are shaded in green.
110 | P a g e
These comparative analyses of all the genome sequences showed that all HS-associated B:2 strains
(M1404, Pakistani and Thai strains) share two large regions of unique sequence compared to the
other complete genomes. The first region is approximately 34,000 bp in length (region 3 in
Figure 5.3) while the second region is approximately 15,000 bp in length (region 4 in Figure 5.3).
Furthermore, there are a number of dispersed genes uniquely present in all of the B:2 strains.
Overall, the B:2 strains share 96 genes that are absent from the other genomes analysed. These
genes unique to the HS strains are provided as Table A.1 in the Appendix. In addition the twelve
Asian strains share approximately 44,000 bp that is absent from the American B:2 strain M1404;
this region contains 59 genes (region 2 in Figure 5.3) that mostly encode proteins with no
significant identity to proteins of known function. The seven Pakistani strains (ATTK, Faisal, Islm,
Karachi, Pesh, PVAcc and V1) share 39 genes that are not present in the other genomes (region 1
in Figure 5.3). This region contains genes encoding mostly phage elements and hypothetical
proteins. Strains 36950, TX1 and BUKK share 35 unique genes, encoding elements with similarity
to the integrative conjugative element (ICEPmu1) of 36950 (Figure 5.4). There are also 42 genes
shared by TX1 and BUKK. TX1 is the only Asian B:2 strain with a large number (32) of unique genes
and these predominantly encode phage elements and hypothetical proteins. The three Thai
isolates have just a single unique gene encoding an abortive infection-(Abi-) like protein of 226
amino acids. Abi-like genes are found in various bacterial species, and encode proteins involved in
bacteriophage resistance (Anba et al., 1995; Garvey et al., 1995).
111 | P a g e
Figure 5.3 Comparison of the genomes of 3480, 36950, HN06, Pm70, M1404, BUKK, TX1, THA, THD, THF, ATTK, Faisal, Islm, Pesh, PVAcc, V1 and Karachi with the genome of the PVAcc strain. The three inner rings show the DNA size, GC content and GC skew of the reference genome (PVAcc strain). The 17 outer rings show regions of the comparison genomes that match the reference genome PVAcc and in the order (inside to outside) 3480, 36950, HN06, Pm70, M1404, BUKK, TX1, THA, THD, THF, ATTK, Faisal, Islm, Pesh, PVAcc, V1 and Karachi. Regions 1-4 on the outside identify particular regions of difference between the strains that are potential prophage elements. The position of the capsule B locus is also noted between regions 2 and 3. This figure was drawn using BLAST ring image generator (Alikhan et al., 2011).
112 | P a g e
Figure 5.4 Comparison of the genomes of 36950, Pm70, combined 3480 and HN06, combined 11 B:2 strains (M1404, ATTK, BUKK, Faisal, Karachi, Islm, Pesh, PVAcc, THA, THD and THF) and combined TX1 and BUKK with the sequence of the integrative conjugative element present in 36950 strain (ICEPmu1) (Michael et al., 2012a). The three inner rings show the DNA size, GC content and GC skew of the reference element (ICEPmu1). The five outer rings show regions of the comparison genomes which match the reference ICEPmu1. This figure was drawn using BLAST ring image generator (Alikhan et al., 2011).
113 | P a g e
Phage identification
All of the B:2 genomes were analysed for the presence of phage elements using PHAST (Zhou et
al., 2011). This analysis identified four regions corresponding to putative temperate phage
elements (Figure 5.5). These regions correspond with the main genetic differences identified
between different groups of strains (regions 1-4; Figure 5.3). Regions 3 and 4 were identified as
intact phages and have been reported previously (Boyce et al., 2012). Region 2 was identified as
an incomplete phage and region 1 as a questionable phage (Figure 5.3 and Figure 5.5). The
questionable phage identified in region 1 is present in the seven Pakistani strains; the incomplete
phage identified in region 2 is shared by all the Asian strains (BUKK_04735-04880) and the intact
phages identified in region 3 and 4 are shared by all the HS-strains (BUKK_07250-07540). The
phage elements identified in regions 1, 2, 3 and 4 are situated at tRNALeu, tRNAMet, tRNASer and
tRNAMet, genes respectively. This correlates with the previous reports indicating that the F108
phage and the lysogenic phage (PMT toxin) integrate into the t33tRNALeu and the t3tRNALeu,
respectively (Campoy et al., 2006; Boyce et al., 2012). Temperate phages may contain important
virulence genes (Boyd and Brüssow, 2002). Indeed, the P. multocida toxin (PMT), which is the
primary virulence factor for porcine atrophic rhinitis, is carried on a lysogenic bacteriophage
(Pullinger et al., 2004). Therefore, the presence of different phage elements in different sets of
strains may impact the virulence of these strains. Further studies should assess the impact of these
different gene sets on virulence.
114 | P a g e
Figure 5.5 Four putative prophages were identified by PHAST (Zhou et al., 2011): a questionable prophage (green); an incomplete prophage (grey); and two intact prophages (pink). The questionable prophage was present in seven Pakistani strains (ATTK, Faisal, Pesh, PVAcc, Islm, Karachi and V1) (region 1 in Figure 5.3), the incomplete prophage was present in all Asian B:2 strains (region 2 in Figure 5.3), and the two intact prophages were found in all B:2 strains (regions 3 and 4 in Figure 5.3).
A P. multocida HS-specific diagnostic PCR has been developed previously (Townsend et al., 1998).
We searched for the DNA sequence recognised by this PCR and identified it within the putative
intact prophage within region 3 (Figure 5.3 and Figure 5.5), Thus, this sequence is indeed present
in all of the HS strains analysed here and not present in the other analysed genomes.
115 | P a g e
Antibiotic resistance genes
The 13 HS-associated B:2 genomes were analysed for the presence of acquired antimicrobial
resistance genes using the Resfinder tool (Zankari et al., 2012). It should be noted that Resfinder
searches only for acquired resistance genes and not for chromosomal mutations. Antimicrobial
resistance genes were only identified in the BUKK and TX1 isolates. These included three
aminoglycoside resistance genes (strA, strB and aph(3’)-lc), one beta lactam resistance gene
(blaTEM-1B), one chloramphenicol resistance gene (catA2), one sulphonamide resistance gene (sul2)
and one tetracycline resistance gene (tet(H)). Thus, these two strains may be resistant to
streptomycin, kanamycin/neomycin, beta-lactams, chloramphenicol, sulphonamides and
tetracycline. Indeed, this correlates with the clinical data on these isolates.
The 42 genes shared by BUKK and TX1 strains and the 35 genes shared by 36950, BUKK and TX1
strains are located in two regions (BUKK genes 06700-06910 and 11175-11375). All of the
identified antibiotic resistance genes are present in these two regions. We believe that these
regions are likely colinear within the genomes but that they are separated in two regions because
of contig breaks within our draft genome assemblies. Of these 77 genes, 51 show high level
similarity (more than 29%) to genes from within the ICEPmu1 of strain 36950 (Michael et al.,
2012a) (Figure 5.4). Forty six of these genes are located in five colinear groups comprising 21
genes (BUKK_11175-11275), nine genes (BUKK_06830-06905), eight genes (BUKK_06700-06740),
five genes (BUKK_06810-06820, BUKK_11360 and BUKK_11370) and three genes (BUKK_06745
and BUKK_06775-06780). The eight gene group (BUKK_06700-06740) is flanked by tRNALeu and
these genes represent the right end of the ICEPmu1 element; this end of the element is also
located next to tRNALeu in strain 36950 (Pmu_03540-03610). These eight genes encode proteins
involved in DNA replication, a single-stranded DNA-binding protein (BUKK_06740 which shows
72% similarity with Pmu_03540 in strain 36950) and an ATPase involved in chromosome
116 | P a g e
partitioning (BUKK_06700 which shows 74% similarity with Pmu_03610 in strain 36950). In
addition, genes encoding DnaB (BUKK_06705; which shows 59% similarity with Pmu_03600) and a
ParB family protein (BUKK_06710 which shows 60% similarity with Pmu_03590) with a predicted
DNA nuclease function were also identified. This set of genes (BUKK_06700-06740) has been
reported as the most conserved region among diverse proteobacterial ICEs (Mohd-Zain et al.,
2004; Michael et al., 2012a).
Within the putative ICE, 47 genes encoded proteins with predicted functions. These included
proteins predicted to be involved in ICE mobility, including excision/integration and conjugative
transfer proteins. A putative phage integrase was identified (BUKK_06905) with similarity to two
integrases found in the ICEPmu1 within strain 36950 (similarities of 51% and 57% with the first and
second integrases, respectively). BUKK_06905 shared significant similarity with tyrosine
recombinases of the Xer family, which mediate integration via site-specific recombination. A
putative relaxase gene (BUKK_06900) was identified downstream of this integrase gene; a similar
organisation is found in the ICEPmu1 (Michael et al., 2012a). Genes encoding proteins necessary
for a conjugative transfer were also present, including proteins predicted to be necessary for the
formation of a type IV pilus (BUKK_11185 which shows 53% similarity with Pmu_03230), TraD
(BUKK_11205 which shows 73% similarity with Pmu_03190), TraG (BUKK_11275 which shows 60%
similarity with Pmu_03040) and TraC (BUKK_11210 which shows 67% similarity with Pmu_03070)
(Michael et al., 2012a). Moreover, a gene encoding a protein with a lysozyme-like domain
(BUKK_11195 which shows 52% similarity with Pmu_03210) and a putative DNA topoisomerase III
(BUKK_06775) (66% similarity with Pmu_03290 in strain 36950) were also identified.
While the resistance genes strA (BUKK_06815) and strB (BUKK_06820), aph(3’)-lc (BUKK_11370),
sul2 (BUKK_06810) and tetR-tet(H) (BUKK_11315 and BUKK_11320) were present in strains BUKK,
117 | P a g e
TX1 and 36950 (Michael et al., 2012b; Michael et al., 2012a), the blaTEM-1B (BUKK_06875) and catA2
(BUKK_11355) genes were unique to strains BUKK and TX1. However, strain 36950 contained more
resistance genes than strains BUKK and TX1 as it also contained resistance genes for
streptomycin/spectinomycin (aadA25), gentamicin (aadB), kanamycin/neomycin (aphA1),
chloramphenicol/florfenicol (floR), tilmicosin/clindamycin (erm(42)) and tilmicosin/tulathromycin
(msr(E)-mph(E)) (Michael et al., 2012b). These differences indicate that the putative ICE present in
strains BUKK and TX1 is not identical to ICEPmu1 in strain 36950 and should therefore be
designated ICEPmu2 as the second ICE discovered in P. multocida strains. Further work should aim
to close the contig breaks in this region to confirm this is a single element and also to investigate
the mobility of the element.
In conclusion, the 96 identified unique genes of HS isolates should be further studied to
understand the role (if any) that these genes play in disease pathogenesis, virulence and host
specificity. Moreover, based on these 96 unique genes in HS strains, a specific, cheap, rapid,
reliable and accurate point-of-care HS diagnostic test should be developed. The integrative
conjugative element (ICE) that was found in two Pakistani isolates should also be isolated and
studied to check its mobility and determine its relatedness to ICEPmu1 of strain 36950.
118 | P a g e
The set of unique genes, which was found to be shared by HS-associated isolates, encouraged us
to develop a diagnostic test that can specifically and rapidly detect serovar B:2 in clinical samples.
The LAMP technique was chosen for this purpose because of its simplicity and therefore potential
application to field conditions. This will be discussed in the next chapter.
119 | P a g e
Chapter 6 – Development of loop-mediated isothermal amplification
(LAMP)-based diagnostic tests for the specific detection of
Pasteurella multocida and haemorrhagic septicaemia-associated
Pasteurella multocida serovar B:2
6.1 Abstract
Following the comparative genomics study of 13 HS-associated strains of P. multocida from Asia
and North America with other non-HS-associated strains (Chapter 5), 96 gene sequences were
identified that were unique to the HS-associated strains. One of these sequences, BUKK_07490,
was used to develop a rapid (28 minutes) HS-specific loop-mediated isothermal amplification
(LAMP)-based diagnostic test that may be suitable for deployment in resource and infrastructure-
limited countries. A second Pm-LAMP test was designed and optimised for the specific detection
of P. multocida. Test results can be visualised by fluorescence under UV light using either SYTO® 9
or SYBR® Safe dyes. Our Pm- and HS-LAMP tests were optimised for magnesium concentration,
primer ratio, primer concentration, reaction time and temperature. Performance characteristics of
these LAMP tests were investigated using 16 P. multocida strains and nine Gram-positive and
Gram-negative bacteria at different concentrations. Results of LAMP testing were validated
against conventional PCRs that were designed for the specific detection of Pasteurella multocida
and the B:2 serovar of HS-associated strains. Following an incubation time of 27 minutes, the
specificity and sensitivity of the HS-LAMP test was 92.5% and 96.5% respectively for template DNA
amounts as low as 5 pg. When duplicates of each sample were incubated for 28 minutes (a
positive result was when both reactions of a sample were positive), the specificity and sensitivity
were both 100% for template DNA amounts as low as 5 pg. The next step should be field
evaluation of the HS-LAMP test in HS-endemic countries like Pakistan and Thailand.
120 | P a g e
6.2 Introduction
At present, the detection technology for HS disease in Pakistan relies largely on clinical signs,
bacterium isolation, Gram stain, and to a lesser extent, biochemical characteristics (E. Nawaz,
personal communication, January, 2012). Although PCR has been shown to be a powerful tool in
the detection of P. multocida (Townsend et al., 1997b; Brickell et al., 1998; Townsend et al., 1998;
Townsend et al., 2001; Petersen et al., 2014), its requirement for a thermocycler limits its use in
the field as a routine diagnostic modality.
Notomi et al. (2000) showed that isothermal nucleic acid amplification methods, such as loop-
mediated isothermal amplification (LAMP), offer feasible platforms for rapid detection of target
nucleic acids (DNA and RNA) in limited infrastructure settings. This method provides a rapid and
specific amplification of DNA using four to six primers that target six to eight separate regions of
the target DNA. The DNA polymerase used in this technique, Bst polymerase, has strand
displacement activity and the test is performed under isothermal conditions (60–65 ˚C) for
approximately 60 minutes. LAMP requires only a simple heating device like a heat block or water
bath to produce a large amount of amplified DNA. The technique has a high tolerance to the
presence of biological products such as urine. This has allowed initial DNA purification steps to be
omitted in most LAMP assays (Kaneko et al., 2007; Francois et al., 2011).
LAMP amplification products can be detected by assessing the turbidity of magnesium
pyrophosphate; a by-product of the reaction. The turbidity can be detected visually or by a real-
time turbidimeter (Mori et al., 2004). Amplification products can also be assessed by agarose gel
electrophoresis, which reveals the typical ladder pattern of LAMP amplified products, or by using a
fluorescent intercalating dye illuminated with a UV lamp. This fluorescence can be detected either
121 | P a g e
in real-time or at the conclusion of the reaction, either visually or by fluorimetry (Parida et al.,
2008).
The LAMP technique has been utilised for the diagnosis of various human and animal bacterial,
protozoan and viral pathogens including Salmonella enterica (Ohtsuka et al., 2005), African
trypanosomes (Kuboki et al., 2003), Foot-and-mouth disease virus (Dukes et al., 2006) and many
other microbes (Mori and Notomi, 2009). It has also previously been used for the detection of P.
multocida in clinical samples from swine (Sun et al., 2010). The aim of the present study was to
develop two LAMP diagnostic tests, one for detection of P. multocida DNA and one specifically for
detecting HS-associated strains of P. multocida.
6.3 Materials and Methods
6.3.1 Bacterial strains
The genomic DNAs of 16 Heddleston types of P. multocida were kindly provided by the
department of microbiology, Monash University, Australia (Table 6.1). Moreover, nine genomic
DNA samples from other bacterial species were used as negative controls to check the specificity
of both the P. multocida LAMP test (Pm-LAMP) and the HS-specific LAMP test (HS-LAMP)
(Table 6.2).
122 | P a g e
Strain Host Capsular type Heddleston type
X73 (ATCC 11039) Chicken A 1 M1404 Bison B 2
P1059 (ATCC 15742) Turkey A 3 P1662 Turkey A 4 P1702 Turkey A 5 P2192 Chicken A 6 P1997 Herring gull n.a. 7 P1581 Pine skin A 8 P2095 Turkey A 9 P2100 Turkey A 10 P903 Pig D 11
P1573 Human A 12 P1591 Human B 13 P2225 Cattle A 14 P2237 Turkey D 15 P2723 Turkey A 16
Table 6.1 The 16 Heddleston types of Pasteurella multocida used in developing the diagnostic tests.
Strain Bacterial species
ATCC 11576 Enterococcus durans ATCC 12228 Staphylococcus epidermidis ATCC 33317 Streptococcus bovis NCTC 9750 Citrobacter freundii
ATCC 13047 Enterobacter cloacae ATCC 14756 Serratia marcescens ATCC 25922 Escherichia coli ATCC 35218 Escherichia coli
ATCC BAA-410 Mannheimia haemolytica Table 6.2 The nine bacterial species used for checking the specificity of Pasteurella multocida (Pm) LAMP and haemorrhagic septicaemia (HS)-specific LAMP.
6.3.2 Pm-LAMP primer design
We designed P. multocida specific LAMP primers using the conserved region of the kmt1 gene
(Townsend et al., 1998). The nucleotide sequences of different alleles from this gene were
downloaded from Genbank (accession numbers FJ986389, DQ233648, DQ233649, EU873317,
AE004439, AF016259, AY225341, AY225342, AY225343, AY225344, CP003313, CP001409 and
CP003022). A total of six specific primers were designed using default parameters in
PrimerExplorer V4 (http://primerexplorer.jp/e/). Sequences for the outer primer pair (Pm-AMM-
123 | P a g e
F3 and Pm-AMM-B3), inner primer pair (Pm-AMM-FIP and Pm-AMM-BIP), and loop primer pair
(Pm-AMM-LF and Pm-AMM-LB) are shown in Table 6.3.
Primer Sequence 5’-3’
Pm-AMM-F3 CAGCAATTTCGAGCAAACAATG Pm-AMM-B3 AGCCAAATAAAAGACTACCGAC Pm-AMM-FIP CACCGCCCCACTGGGTAAATAGTGGGGCTTTACGCTGAT Pm-AMM-BIP ATGAACCGATTGCCGCGAAATTGAAGCCCACTCACAACGAG Pm-AMM-LF CGGATAGAGCAGTAATGTCAGCA Pm-AMM-LB TGCCACTTGAAATGGGAAATGG
2154-AMM-F3 TCGGGGATTTTCTTTATCTCTT 2154-AMM-B3 GAGCGATAATACAAAAGATCGTA 2154-AMM-FIP CCGCAAAAGCGAGAGCAAAAG-GCTTTGCGCATTTCTTCC 2154-AMM-BIP ATCCTTTCTCCATAAATGGTCTTGC-CAGACCCGTTCTATTGGT 1517-AMM-F3 AGCAATGAATGTTGCTGC 1517-AMM-B3 TTGTTGTATCGGGTTGTTTG 1517-AMM-FIP ACTGTTGAACCGAGATAAAACTCAATTGCCAAAAGAAATGGGGT 1517-AMM-BIP CCATTGTAAAGTGGATGCTACTCGGTTTCCTGCACCACCTAA 1517-AMM-LB CGTAGGTCATAATGGCATTGGTGTG HS-1517-Fwd GCCTTTGAGTTTGGAGCTGT HS-1517-Rev TGAACTACCGACAGTGGCAT
Table 6.3 Primer sequences used in developing the diagnostic tests.
6.3.3 HS-LAMP primer design
After preliminary molecular typing of 23 HS-associated isolates from Pakistan and Thailand by
PFGE and MLST (Moustafa et al., 2013), we performed comparative genomics analysis of 13
genomes of HS-associated B:2 strains from North America, Pakistan and Thailand with non-HS-
associated strains (Pm70, 36950, 3480 and HN06). The comparative genomics analysis revealed
that 96 genes were shared uniquely across the HS-associated B:2 strains. The sequences which
were shared by these B:2 strains were analysed using BLASTN (Altschul et al., 1990) and the
regions which had the lowest sequence similarity compared to any other organism were chosen.
Two candidate regions of 1350 and 1300 nucleotides, which were found within genes
BUKK_10630-10635 and BUKK_07490, respectively, were used for designing HS-specific LAMP
primers using default parameters in PrimerExplorer V4. The outer primer pair (1517-AMM-F3 and
124 | P a g e
1517-AMM-B3), inner primer pair (1517-AMM-FIP and 1517-AMM-BIP), and loop primer (1517-
AMM-LB) sequences are provided in Table 6.3. An alternate set of primers: outer (2154-AMM-F3
and 2154-AMM-B3), and inner (2154-AMM-FIP and 2154-AMM-BIP) was also tested (Table 6.3).
6.3.4 HS-1517-PCR
The 1300 nucleotides of BUKK_07490 were used for designing conventional PCR primers to
develop an HS-specific PCR test without needing to perform a species-specific test. The primers
were designed using the Primer3 online tool (Koressaar and Remm, 2007; Untergasser et al., 2012)
to yield an expected amplicon of 400 nucleotides. The primers 1517-Fwd and 1517-Rev are
forward and reverse primers, respectively, (Table 6.3) for this PCR. The PCR contained 0.5 μmol/L
of each primer, 2 μL of 10× buffer (Fisher), 1.25U of Taq DNA polymerase (Fisher), 0.8 mmol/L of
dNTPs, 1 μL of template DNA (of extracted genomic DNA, Table 6.1 and Table 6.2), and ddH2O up
to a final volume of 20 μL. The amplification conditions were an initial denaturation at 95 °C for 5
min, followed by 30 PCR cycles of 95 °C for 1 min, 55 °C for 1 min, and 72 °C for 1 min and a final
extension step of 72 °C for 5 min. The positive control for this test was the genomic DNA of M1404
strain (HS-associated B:2 serotype). The specificity of the HS-1517-PCR was evaluated by using the
genomic DNA of nine different bacterial species (Table 6.2) and the genomic DNA of the other 15
Heddleston types of P. multocida (Table 6.1). Ten microlitres of PCR product were analysed by
electrophoresis in a 1% agarose gel stained with SYBR® Safe (10,000×, Life technologies) and
illuminated by UV transillumination (GelDoc EZ system, Bio-Rad). The limit of detection of the PCR
was evaluated using 10 ng, 1 ng, 100 pg, 10 pg and 5 pg amounts of genomic DNA of B:2 serotype
of P. multocida.
6.3.5 Pm-LAMP reaction
LAMP reactions of 25 µL were optimised for outer primer: inner primer: loop primer ratios, primer
concentrations, dNTPs and MgSO4 concentrations, temperature, and reaction time. Briefly, the
125 | P a g e
primers FIP and BIP, LF and LB, and F3 and B3 were varied from 0.17 to 2 µmol/L, 0.083 to 1
µmol/L, and 0.017 to 0.2 µmol/L, respectively; dNTPs from 1 to 4 mmol/L; and MgSO4 from 1 to 18
mmol/L. To find the optimum temperature for the LAMP test, the reaction was carried out for 1
hour at 61, 63, 65, 67 and 69 °C using a Mastercycler gradient thermocycler (Eppendorf). Other
reagent conditions were kept standard, including 8 units of Bst 2.0 WarmStart™ DNA polymerase,
and 1× Isothermal Amplification Buffer (catalogue# M0538L, New England Biolabs). The reaction
was terminated by increasing the temperature to 90 ˚C for 2 min, then 10 μL of the LAMP product
were analysed by electrophoresis in 1% agarose gels stained with SYBR® Safe (10,000×, Life
technologies) and illuminated by UV transillumination (GelDoc EZ system, Bio-Rad).
6.3.6 HS-LAMP reaction
We tested the performance of primer sets 1517 and 2154 across a wide range of temperatures
(61, 63, 65, 67 and 69 °C) and magnesium concentrations (2, 4, 6, 8 and 10 mmol/L) in order to
choose the best set to be a good candidate for further experimentation. Similarly to Pm-LAMP, the
FIP and BIP; F3 and B3 concentrations were varied from 0.2 to 2 µmol/L; and 0.02 to 0.2 µmol/L,
respectively. Both sets were checked for the best outer primer: inner primer ratios. Other
conditions were held constant, including 8 units of Bst 2.0 WarmStart™ DNA polymerase, 1×
Isothermal Amplification Buffer (catalogue# M0538L, New England Biolabs) and 2 mmol/L dNTPs
at 63 ˚C for 1 hour. As before, the reaction was terminated by increasing the temperature to 90 ˚C
for 2 min, then 10 μL of the LAMP product were analysed by electrophoresis in 1% agarose gels
stained with SYBR® Safe (10,000×, Life technologies) and illuminated by UV transillumination
(GelDoc EZ system, Bio-Rad).
6.3.7 Loop primer
A reaction was performed where loop primer was added. For this particular reaction, the
specificity of the HS-LAMP was evaluated by using the genomic DNA of the nine different strains
126 | P a g e
mentioned above (Table 6.2), the genomic DNA of the other 15 Heddleston types of P. multocida
(other than B:2 serotype) (Table 6.1), and water. Different amounts of the B:2 serotype (10 ng, 1
ng, 100 pg, 10 pg, 5 pg and 1 pg) per reaction (Table 6.1) were used to evaluate the lower limit of
detection.
6.3.8 Real time evaluation of LAMP reaction
Once reaction conditions had been partially optimised using the above methods, further
optimisation experiments were conducted using a real time platform (qPCR, Rotor Gene Q series,
Qiagen) and SYTO® 9 (Life technologies). This platform enabled determination of the optimum
time for which to perform the reaction, as well as to refine or confirm the other reaction
parameters: primer concentration, primer ratios, MgSO4 concentration, dNTPs concentration, and
reaction temperature. Briefly, the FIP and BIP; LB; and F3 and B3 concentrations were varied from
0.2 to 2 µmol/L, 0.1 to 1 µmol/L, and 0.02 to 0.2 µmol/L, respectively; dNTPs from 1 to 4 mmol/L;
and MgSO4 from 1 to 10 mmol/L. Other conditions were held constant, including 8 units of Bst 2.0
WarmStart™ DNA polymerase, 1× Isothermal Amplification Buffer (catalogue# M0538L, New
England Biolabs), 0.76 µmol/L SYTO® 9 fluorescence dye (catalogue # S-34854, Molecular Probes)
and a 25 µL total volume. The reaction was terminated by increasing the temperature to 90 ˚C for
2 min.
6.3.9 Processing colonies and inoculated broth without DNA extraction for Pm-LAMP
Two techniques were used to process colonies from cultured blood agar plates and inoculated
broth cultures after incubation overnight. The first method was heating an aliquot of the broth (25
µL) or colony in water (25 µL) first at 95 ˚C for 1, 3, 5, 10, 12, 15, 17 or 20 minutes and then adding
2 µL of this material to the reaction mixture. The second technique was adding 2 µL of the broth or
touching a colony with a sterile pipette tip and adding it directly to the reaction mixture and
heating at 95 ˚C for 1, 3, 5, 10, 12, 15, 17 or 20 minutes and then cooling before adding the Bst
127 | P a g e
enzyme. The uninoculated broth was tested to be sure that it was not interfering with the reaction
conditions. We only had the opportunity to test colonies and broth with the Pm-LAMP. However,
we did not test them with HS-LAMP as we did not have access to live HS-associated P. multocida
(strain M1404) at the time of this experiment.
6.3.10 Evaluating the sensitivity and specificity of both the Pm-LAMP and HS-LAMP
The specificity of the Pm-LAMP method was evaluated by using the genomic DNA of nine different
bacterial species (Table 6.2). The genomic DNA of each pathogen was tested using the Pm-LAMP
method, and the genomic DNA of 16 Heddleston types of P. multocida were used as positive
controls (Table 6.1). For the HS-LAMP specificity evaluation, the genomic DNA of the same nine
bacterial species (Table 6.2) and the other 15 Heddleston types of P. multocida (Table 6.1, types 1
and 3-16) were tested using the HS-LAMP, while the B:2 serotype of P. multocida was used as a
positive control. In addition to the previously mentioned bacterial isolates, both tests were further
evaluated in the presence of 1 µL of genomic DNA extracted from bovine blood and bone marrow.
All reactions were run first on the Rotor Gene Q series (Qiagen) and amplification products were
subsequently analysed by electrophoresis in 1% agarose gels. The percent sensitivity of a test is
calculated by
True positives
True positives + False negatives× 100
and similarly percent specificity is calculated by
True negatives
True negatives + False positives× 100
The limit of detection of both tests were evaluated by varying the amount of genomic DNA from
the B:2 serotype of P. multocida (10 ng, 1 ng, 100 pg, 10 pg, 5 pg and 1 pg) per reaction in positive
control experiments.
128 | P a g e
6.3.11 Visualisation of the HS-LAMP and Pm-LAMP products
Real time reactions were checked on the fluorescence trace graph and the tubes were visualised
under UV-illumination after the conclusion of the reaction. Additionally, amplification products
were examined by agarose gel electrophoresis. The results of LAMP reactions, which were run in a
water bath or heat block, were visualised by observation under UV-illumination after adding 1 μL
of the fluorescent dye, SYBR® Safe (1,000×), to a 5 μL aliquot of the total reaction mixture. The
green fluorescence apparent in positive tubes was further confirmed by visualising the typical
ladder banding pattern of LAMP products by electrophoresis.
6.3.12 Validation of both the Pm-LAMP and HS-LAMP using Receiver Operating
Characteristic (ROC) curves
After choosing the optimum conditions for both Pm-LAMP and HS-LAMP, 18 DNA samples were
used to evaluate both developed tests (Table 6.4). The author was blinded to the identity of these
DNA samples until the results had been determined. Duplicates for each of the 18 samples, for
both Pm-LAMP and HS-LAMP, were tested five times using real time PCR (Rotor Gene Q series,
Qiagen). This allowed inter- and intra-run variations to be assessed. As a duplicate of each reaction
was run, this allowed two different definitions of a positive result to be applied to the
experimental data. The first definition treated each reaction tube as an independent entity
(“singlets”) while the second definition required that both tubes containing the same reactants be
positive for the overall reaction to be considered positive (“duplicates”). The sensitivity and
specificity were calculated for singlets and duplicates for Pm-LAMP and HS-LAMP at various (cut-
off) time points during incubation: 20, 25, 27, 28, 29, 30, 31, 32, 33 and 34 minutes for
Pm-LAMP, and 20, 25, 26, 27, 28, 29 and 30 minutes for HS-LAMP. Receiver operating
characteristic (ROC) curves were established for both tests. In ROC curves, sensitivity is plotted
versus 100-specificity for different cut-off points. Each point on the curve represents a
sensitivity/specificity pair. When a diagnostic test has perfect discrimination power (100%
129 | P a g e
sensitivity and specificity), the ROC curve passes through the upper left corner. Consequently, the
closer the curve is to the upper left corner, the higher the accuracy of the test (Lusted, 1971; Metz,
1978; Zweig and Campbell, 1993).
DNA sample
Content Pm-
LAMP HS-
LAMP
Concentration of Pasteurella multocida
(pg/reaction)
1 P. multocida B:2 strain M1404 + + 40 2 P. multocida B:2 strain M1404 + + 300 3 P. multocida B:2 strain M1404 + + 200 4 P. multocida strain P1059 + - 20 5 P. multocida strain P2192 + - 50 6 P. multocida strain P1059 + - 30 7 P. multocida strain P2192 + - 80 8 Water - - 0
9 Bovine blood DNA and Escherichia
coli (ATCC 25922) - - 0
10 P. multocida strain P1662 + - 12.5 11 P. multocida B:2 strain M1404 + + 5
12 Mannheimia haemolytica and
Enterococcus durans (ATCC 11576) - - 0
13 P. multocida B:2 strain M1404 + + 10
14 Enterobacter cloacae (ATCC 13047) and Escherichia coli (ATCC 35218)
- - 0
15 Streptococcus bovis (ATCC 33317)
and Citrobacter freundii (NCTC 9750) - - 0
16 Staphylococcus epidermidis (ATCC 12228) and Serratia marcescens
(ATCC 14756) - - 0
17 P. multocida B:2 strain M1404 + + 5 18 P. multocida strain P1662 + - 100
Table 6.4 Eighteen DNA samples used for validation of Pm-LAMP and HS-LAMP.
6.4 Results
HS-1517-PCR
A 400 nucleotide band was present for the positive control of B:2 serotype while 15 out of 15
other Heddleston types of P. multocida showed no bands (Figure 6.1). The nine genomic DNAs of
130 | P a g e
bacteria from genera other than Pasteurella were also negative. The lowest concentrations of
template genomic DNA of B:2 serovar that could be detected was 10 pg per reaction.
Figure 6.1 Specificity of HS-1517-PCR. Lane 1: 100 bp ladder (Axygen). Lane 2-17: genomic DNA of Heddleston types 1-16 of Pasteurella multocida (10 ng), respectively. The 400 nucleotide band in lane 3 represents the B:2 serotype positive control.
Pm-LAMP reaction
The optimised reaction was 0.314 µmol/L for each of FIP and BIP, 0.157 µmol/L for each of LB and
LF, 0.029 µmol/L for each of F3 and B3, 8 units of Bst 2.0 WarmStart™ DNA polymerase, 1×
Isothermal Amplification Buffer, 2 mmol/L MgSO4 solution (in addition to 2 mmol/L in the buffer
already, giving an overall concentration of 4 mmol/L), 2 mmol/L dNTPs and 1 μL of template DNA
at 63 ˚C for 34 minutes (Figure 6.2). For real-time reactions, 0.76 µmol/L SYTO® 9 fluorescence dye
was added.
131 | P a g e
Figure 6.2 Ladder pattern of positive Pm-LAMP reaction. Lane 1: 1 kb ladder (Axygen). Lanes 2-6: genomic DNA (1 ng) of Heddleston types 1, 2, 3, 4 and 5 of Pasteurella multocida, respectively. Lane 7: water (negative control).
HS-LAMP reaction
During preliminary testing, the 1517 set showed better results in terms of specificity and tolerance
of different temperatures (61, 63, 65, 67 and 69 °C) and magnesium concentrations (2-10 mmol/L)
than the 2154 set, so optimisations were continued for the 1517 set only.
Optimal reaction conditions for the 1517 primer set were found to be 0.29 µmol/L for each of FIP
and BIP, 0.143 µmol/L for LB, 0.029 µmol/L for each of F3 and B3, 8 units of Bst 2.0 WarmStart™
DNA polymerase, 1× Isothermal Amplification Buffer, 1 mmol/L MgSO4 solution (in addition to 2
mmol/L in the buffer already, giving a total of 3 mmol/L), 2 mmol/L dNTPs and 1 μL of template
DNA at 63 ˚C for 30 minutes (Figure 6.3). For real-time reactions, 0.76 µmol/L SYTO® 9
fluorescence dye was added.
132 | P a g e
Figure 6.3 Ladder pattern of positive HS-LAMP reaction. Lane 1: 1 kb ladder (Axygen). Lanes 2-4: genomic DNA of B:2 serotype of Pasteurella multocida (100 ng, 1 ng and 10 pg, respectively). Lane 5: water (negative control).
Processing colonies and inoculated broth without DNA extraction
Both methods for processing colonies and broth gave positive results at all incubation times, even
one minute. Broth reactions produced results more quickly on qPCR analysis than the colony
reactions. The uninoculated broth did not interfere with the reaction conditions and gave negative
results.
Adding Loop primer (1517-AMM-LB) to HS-LAMP reaction
The PrimerExplorer V4 program was able to design only one loop primer for the 1517 set, which
was used to improve the reaction performance. The HS-LAMP without loop primer showed 100%
sensitivity (for template amounts ≥10 pg) and 76% specificity when the reaction was run for 1
hour, while after the addition of the single loop primer, the sensitivity was 100%, the detection
limit improved to be ≥5 pg, the specificity improved to 100%, and the minimum length of time
required was reduced to 28 minutes.
133 | P a g e
Evaluating the sensitivity and specificity of Pm-LAMP and HS-LAMP
Figure 6.4 shows that the Pm-LAMP reaction gave positive results with all 16 Heddleston types of
P. multocida. After being tested against the same 16 Heddleston types, the HS-LAMP reaction was
specific only for the serotype B:2 of P. multocida (Figure 6.5). Both the Pm-LAMP and HS-LAMP
tolerated the presence of DNA extracted from bovine blood, while only the HS-LAMP tolerated
DNA extracted from bovine bone marrow (Figure 6.5). Both tests gave uniformly negative results
against the nine genomic DNAs extracted from non-P. multocida bacteria (Table 6.2). The limits of
detection for Pm-LAMP and HS-LAMP were 10 pg and 5 pg, respectively (data not shown).
Figure 6.4 Specificity of Pm-LAMP. Lane 1: 1 kb ladder (Axygen). Lane 2-17: genomic DNA of Heddleston types 1-16 of Pasteurella multocida (10 ng), respectively. Lane 18: water (negative control).
Figure 6.5 Specificity of HS-LAMP. Lane 1: 1 kb ladder (Axygen). Lane 2-17: genomic DNA of Heddleston types 1-16 of Pasteurella multocida (10 ng), respectively. Lane 18: water (negative control). Lane 19: DNA extracted from bovine blood. Lane 20: DNA extracted from bovine bone marrow.
134 | P a g e
Visualisation of the HS-LAMP and Pm-LAMP products
The positive real time reaction could be detected in three ways: either by checking the
fluorescence in real time (Figure 6.6), illumination of the tubes under UV light (Figure 6.7A) or
following agarose gel electrophoresis of amplification products (Figure 6.4 and Figure 6.5).
Addition of SYBR® Safe dye after running the LAMP reaction in a water bath or heat block could be
used to reliably differentiate between positive and negative reactions (Figure 6.7B) by observing
the intense green fluorescence of positive tubes when illuminated by UV light.
135 | P a g e
Figure 6.6 Visualisation of positive results of Pm-LAMP by comparing the fluorescence of positives and negatives in the real time reactions. Fluorescence is plotted on the y-axis, and reaction cycle number (equal to minutes of reaction) is plotted on the x-axis. Black lines at fluorescences of approximately 5 units are negative results. Red traces are positive results.
Figure 6.7 Visualisation of positive results under UV light by adding SYTO® 9 dye before running the
reaction (A) or adding SYBR® Safe once the reaction had completed (B). Left tube: positive result. Right tube: negative result.
136 | P a g e
Validation of both the Pm-LAMP and HS-LAMP using ROC curves
The best sensitivity and specificity for HS-LAMP singlets (96.7% and 92.5%) and duplicates (100%
and 100%) reactions were achieved at 27 and 28 minutes, respectively (Figure 6.8A).
For Pm-LAMP, the best sensitivity and specificity for singlet (92.5% and 86.7%) and duplicate
(88.3% and 96.7%) reactions was achieved at 34 minutes. However, when the results from DNA
tubes 11 and 17 (Table 6.4) were excluded (as they had the lowest template DNA amount of 5 pg),
the best sensitivity and specificity for Pm-LAMP singlets (91% and 93.3%) and duplicates (98% and
96.7%) reactions were achieved at 33 and 34 minutes, respectively (Figure 6.8B).
137 | P a g e
Figure 6.8 (A) HS-LAMP Receiver Operating Characteristic (ROC) curve with at least 5 pg of target DNA per reaction. (B) Pm-LAMP ROC curve with at least 10 pg of target DNA per reaction. Singlets and duplicates are represented by green and blue lines, respectively. Points on each curve represent the incubation durations. For Pm-LAMP, the incubation times were 20, 25, 27, 28, 29, 30, 31, 32, 33 and 34 minutes and for HS-LAMP, the incubation times were 20, 25, 26, 27, 28, 29 and 30 minutes. Arrows point to the incubation time that represents the highest sensitivity and specificity.
138 | P a g e
6.5 Discussion
Haemorrhagic septicaemia is a serious infection that causes severe economic losses in affected
buffalo and cattle enterprises (De Alwis, 1993a). A traditional microbiological approach for
isolation and identification of P. multocida usually takes two to three days of laboratory time, and
in developing countries like Pakistan, where the disease is endemic, it can take days or weeks for
the samples to reach the government laboratories. This significantly delays definitive diagnosis (E.
Nawaz, personal communication, January, 2012). PCRs require expensive equipment which is
beyond the resources of many diagnostic laboratories in the developing world. For these reasons,
HS is typically diagnosed in resource limited settings on clinical signs alone (De Alwis, 1999c).
In this study, we describe two novel, rapid, and economical Pm-LAMP and HS-LAMP methods to
detect P. multocida and HS-associated serotype B:2 strains of P. multocida. For P. multocida, the
conserved region of the kmt1 gene was used for the specific primer amplification region and for
the HS-associated serotype B:2 strains of P. multocida, a 1,300 nucleotide region of the gene
BUKK_07490 was used (Genbank accession number JQAO00000000).
Compared with conventional PCR, the diagnosis of P. multocida and HS-associated P. multocida
using LAMP assays requires only a simple water bath or heating block to incubate the reaction
mixture at 63 °C for approximately 30 minutes before the results can be visualised. It is worthwhile
mentioning that LAMP reagents still need to be kept frozen until they are used. The isothermal
nature of the LAMP assay decreases the time lost in temperature cycling, which leads to rapid and
efficient testing compared with regular PCR (Notomi et al., 2000). The results of both our LAMP
assays, when either a water bath or a thermal cycler was used to maintain the temperature at
63 °C, were robust.
139 | P a g e
Studies have shown superior tolerance of LAMP tests for biological substances such as urine, stool,
plasma, serum, phosphate-buffered saline (PBS), saline and aqueous and vitreous humors (Kaneko
et al., 2007; Francois et al., 2011). Our HS-LAMP test showed tolerance to bovine bone marrow
and blood which could be advantageous in diagnosis, as usually the clinical samples arriving at
veterinary diagnostic laboratories in Pakistan are long bones (E. Nawaz, personal communication,
January, 2012). It is reported that when blood is overgrown by contaminants, the bone marrow
from long bones are a good source from which to recover P. multocida; even from animals
exhumed a few days after burial (De Alwis, 1993a; De Alwis, 1999c).
Four ways to visualise a positive result are available for our LAMP assays. Real-time monitoring of
the reaction is possible which decreases the time needed to get results and reduces the risk of
carry-over contamination during post-PCR processing. The real-time tubes can also be visualised
under UV-illumination as they already contain SYTO® 9 fluorescence dye. The other two methods
are detection by gel electrophoresis or by directly visualising the tube under UV light after adding
DNA intercalating dyes such as SYBR® Safe once the reaction has finished. The latter two
visualisation methods have the disadvantage of possibly enabling contamination as the tubes are
opened.
Colonies from blood agar dishes and overnight culture broths containing P. multocida were
successfully amplified following our Pm-LAMP protocol, without the need for prior DNA extraction.
Heating the colonies or the broth at 95 ˚C for one minute is sufficient to release bacterial DNA and
in our hands, produced results that were 100% sensitive and specific. This method facilitates
better compliance and will ease the roll-out into the field situation. Heat treating the broth or
colonies before adding them to the reaction mix was the easier way, as the alternate method of
140 | P a g e
adding broth and colonies and heating them in the reaction mix, cooling the mixture, then adding
Bst enzyme was inconvenient. The uninoculated broth was tested to be sure that it did not
interfere with the reaction conditions. Broth results outperformed colony results on qPCR, so we
recommend broth as a better starting material for this Pm-LAMP test. The reasons for this
difference were not investigated. We expect the HS-LAMP to work in the same way but this was
not tested.
Addition of a single loop primer to the HS-LAMP improved the detection limit from 10 pg to 5 pg.
Moreover, the time required decreased from 1 hour to 28 minutes. Marked improvements with
the addition of two loop primers is well documented (Nagamine et al., 2002) and has been
demonstrated in the detection of Mycobacterium sp. (Iwamoto et al., 2003), periodontal
pathogens (Yoshida et al., 2005), and Plasmodium falciparum (Poon et al., 2006). In our study, only
one loop primer could be designed but despite this, marked improvements in the test were
achieved.
We followed very strict protocols to avoid contamination but despite this, sporadic but non-
reproducible false positives were sometimes produced. LAMP tests in general should theoretically
not amplify non-target sequences, however, non-template amplification can occur (Tanner and
Evans, 2014).
Specificity was confirmed for the Pm-LAMP test by giving positive results for the 16 Heddleston
types of P. multocida tested strains while it did not show any cross-reactivity with a panel of DNAs
from another nine bacterial species. Similarly, the HS-LAMP assay gave positive results with the
HS-associated B:2 serotype of P. multocida only, and it did not show reactivity with other DNAs
tested from the other 15 Heddleston types of P. multocida or the other nine different bacterial
141 | P a g e
species tested. Our specificity results can only be viewed in the context of what we tested. Only
with more testing, could the specificity of this test be clarified further.
From analysis of the ROC curve for HS-LAMP, the highest sensitivity (96.7%) and specificity (92.5%)
was achieved at an incubation time of 27 minutes. However, if practical and monetary constraints
allow, we recommend running two tubes per sample (duplicates), in which case both the
sensitivity and specificity were 100% after 28 minutes of incubation. Although we found that
running the reaction in duplicate produced higher sensitivity and specificity, compared to running
single tube reactions, the reagent cost per test will be doubled, which is an important factor in
developing countries. It must be remembered that for duplicates, our definition of a positive is
when both tubes are positive. A negative result is when one or both tubes are negative.
To compare our two tests with conventional PCR for P. multocida species and HS-associated
strains of P. multocida (Townsend et al., 1998), 18 “unknown” DNAs were checked. Results
showed that both PCRs had a detection limit of 10 pg per reaction. This was the same detection
limit for Pm-LAMP while the HS-LAMP had a better detection limit of 5 pg per reaction. This
confirmed that the LAMP tests described in this manuscript had detection limits at least as low as
currently available molecular tests.
Recently, a real-time PCR was developed by Petersen et al. (2014). In this study, single nucleotide
polymorphisms in the est target gene were identified. This gene is one of the seven genes used in
the RIRDC MLST scheme (Subaaharan et al., 2010) that is suitable for detection of sequence types
(ST) 122, 63, 147 and 162 that are all associated with HS disease. The test is able to detect both
the so-called “Asian” (B:2) and “African” (E:2) serotypes of HS. Compared to the test described by
Petersen et al. (2014), our LAMP tests were not designed to detect E:2 serotypes and these
142 | P a g e
serotypes have not yet been tested. However, our LAMP test does not need sophisticated
instrumentation like qPCR, and the non-primer reagent costs are less: 0.40 USD compared to 0.46
USD. Consideration should be given to whether the HS-LAMP reactions should be run in duplicate
for the sake of diagnostic accuracy. If that is the case, the non-primer cost of reagents mentioned
above would be doubled: 0.80 USD compared to 0.40 USD. In addition to the reagent cost
comparison, the time needed to obtain a positive result is substantially less in the case of the HS-
LAMP test: 28 minutes, compared to 82 minutes (Petersen et al., 2014). We were not able to
compare the limits of detection of the two tests as Petersen et al. (2014) did not report their limit
of detection.
It is important to mention that this is a preliminary trial to optimise the parameters for the two
LAMP assays. The number of strains used is too low for accurate calculation of the sensitivity and
specificity of the assays. The sensitivities and specificities included in this chapter should only be
viewed as preliminary estimates of the true values. Consequently, our developed LAMP reactions
should be further evaluated in the field through collaborations with veterinary diagnostic
laboratories, governmental and private veterinarians in affected endemic countries such as
Pakistan, India and Thailand during outbreaks of HS. Some improvements in processing DNA from
clinical samples will make the test a step closer to becoming a point-of-care test. Ultimately, we
hope that these tests will reach the stage of allowing the attending veterinarian to diagnose HS
using only a heating block, UV-illumination, a battery and ice bricks to keep reagents cool prior to
use.
143 | P a g e
Chapter 7 – Thesis Summary and General Discussion
7.1 Thesis Summary
Haemorrhagic septicaemia (HS) is the most economically important production animal disease in
Pakistan (Munir et al., 1994; Anonymous, 1996; Ali et al., 2006; Imran et al., 2007). In Pakistan in
the year 2013-2014, the livestock sector contributed 55.9% of the value of the agriculture sector
and almost 11.8% to the Pakistani GDP (Farooq, 2014). Pakistan also produces the fourth largest
volume of buffalo and cow milk in the world (6.5% of total global volume in 2012) (FAO, 2014). For
these reasons, great consideration should be given to economically-significant diseases that affect
Pakistani agriculture; especially the health of buffaloes and cattle.
According to the World Animal Health Information Database (WAHID) of the Office International
des Epizooties (OIE), HS is enzootic in Pakistan. However, no recent quantitative data were given.
The quantitative data were current until June 2007 and in the latest data presented that referred
to the first half of 2007, eight outbreaks were reported that represented 75,640 susceptible cattle,
3,334 cases and 652 deaths (OIE, 2014). Our study started with an epidemiological study to
identify factors that were associated with the spread of HS in Pakistan. Despite the importance of
the disease and its effect on the economy in Pakistan, no such studies existed before our
investigations.
To begin with, 217 owners of buffalo and cattle dairy farms from six different locations in Karachi,
Pakistan were interviewed. The information obtained from these interviews was used to
determine if there were any independent variables that were statistically associated with the
presence of HS. Univariable analyses and multivariable logistic regression were used to identify
putative risk and protective factors statistically associated with HS-affected (cases) and unaffected
144 | P a g e
(controls) farms, respectively. Five independent variables consisting of two protective factors and
three risk factors were contained in the final multivariable logistic model. The two protective
factors identified were vaccination and the length of time (in months) that cattle were kept on
farm. In addition, three risk factors were identified: using underground water, presence of FMD in
the farm and having more cattle and buffaloes on the farm. The problems identified during this
study showed that some farmers were not satisfied with their vaccines and for this reason we
used molecular methods to compare vaccine strains to field samples. A second problem was that it
was extremely uncommon for samples to be sent to diagnostic laboratories. This was due to cost
and extremely limited diagnostic laboratory capabilities. Because of this, the disease was typically
diagnosed based on clinical signs and local veterinary knowledge (E. Nawaz, personal
communication, January, 2012). Therefore, developing a rapid point-of-care diagnostic test was
seen as a priority that could help to facilitate the control of the disease.
Field isolates from two HS-endemic countries, Pakistan and Thailand, were then studied. These
two countries were chosen so that any geographical variations between B:2 HS-associated isolates
could be identified and then a test could be developed to diagnose infected animals from this
endemic region. The molecular epidemiological study utilised 21 field isolates and three vaccine
strains of HS-associated P. multocida from Pakistan and Thailand. Two molecular techniques, PFGE
and MLST, were used to compare these strains. This comparative genetic study showed that the
ST, determined by MLST, for 21 of the 24 isolates tested, was 122. In contrast, the PFGE results
showed only one band difference between the Pakistani and the Thai isolates. The results of this
study showed that these two techniques were ineffective at differentiating between isolates of
the same country. Next generation sequencing was then used to see if it could meet this need.
145 | P a g e
Ten HS-associated field isolates and two vaccine strains were selected from 24 isolates for WGS.
These 12 strains were compared with the North American HS-associated strain M1404 and the
four complete genomes of non HS-associated P. multocida: Pm70, 3480, 36950 and HN06. Analysis
of these genomes identified a set of 96 unique genes which was found to be shared by all HS-
associated strains, and the genes were located on two putative temperate phages. These genes
were absent from the four non HS-associated strains that were tested. Moreover, 59 genes were
found to be shared only by Asian B:2 strains. These genes were located on one putative temperate
phage. The presence of different phage elements in different sets of strains may impact the
virulence of these strains. Consequently, it is reasonable to suspect that the presence of these
putative prophages may provide virulence genes that contribute to the pathogenesis of HS.
Further studies should investigate the influence that these different gene sets have on virulence.
Some genetic regions with high similarity to the integrative and conjugative ICEPmu1 element
were identified in two Pakistani strains (BUKK and TX1). The putative ICE of BUKK and TX1 is not
identical to ICEPmu1 of strain 36950, and therefore, it may be the second ICE to be discovered in
P. multocida strains.
At the time of writing (2014), HS disease is diagnosed in Pakistan based largely on clinical signs,
bacterium isolation, Gram stain, and to a lesser extent, the biochemical characteristics of isolates.
The requirement for a thermocycler limits the use of PCR in the field. Loop-mediated isothermal
amplification (LAMP) offers a more feasible platform for rapid detection of target nucleic acids.
To develop a specific test for the detection of HS-associated P. multocida serovar B:2, one of the
96 genes that are unique to HS-associated strains of P. multocida, BUKK_07490, was used.
Ultimately, a rapid (28 minutes), sensitive and specific LAMP test was created. Following an
146 | P a g e
incubation time of 27 minutes, the specificity and sensitivity of the HS-LAMP test were 92.5% and
96.5% respectively for template DNA amounts as low as 5 pg. When duplicates of each sample
were incubated for 28 minutes (a positive result was when both reactions of a sample were
positive), the specificity and sensitivity were both 100% for template DNA amounts as low as 5 pg.
The sensitivity and specificity were increased by using duplicates but the cost of consumable
reagents was doubled.
7.2 The Aims of the Project
The most challenging part of this project was working in Pakistan. It was not easy to work there as
the political situation was unstable and there was civil unrest during much of the time we spent
working in the field and the laboratory. Unfortunately, this adversely impacted on the samples we
were able to retrieve and analyse from Pakistan. However, given the difficult circumstances that
were beyond our control, the best outcomes that were possible were achieved.
The broad aims of this thesis, listed in the first chapter (section: Purpose of Project) can now be
reviewed.
1. To observe husbandry systems and identify independent variables that are associated
with the presence of HS in farms in Karachi, Pakistan; and to collect and screen field
samples from HS suspected cases in Karachi so that the prevalence of HS can be
determined in this city.
To achieve this aim, an epidemiological case-control study was conducted in six different locations
in Karachi, Pakistan. The data were then analysed using univariable and multivariable analyses.
Putative risk and protective factors were identified. At the time of would-be sampling, there was
political unrest in Karachi and so samples were not collected due to safety concerns. Overall, it can
only be stated that this aim was partially achieved.
147 | P a g e
2. To conduct molecular genetic studies that characterise the strains of P. multocida
associated with HS from South Asia and look for DNA sequences that could be used for
the development of a diagnostic test (or tests).
In this project, three different molecular techniques, MLST, PFGE and WGS, were used on 21 field
isolates and three vaccine strains from Pakistan and Thailand. Initially, MLST and PFGE were used
on 24 isolates. While PFGE showed that Pakistani isolates differed slightly from the Thai isolates by
lacking one band, MLST showed that all of the 24 isolates were ST 122. These results lent support
to the use of WGS to compare some of these isolates to obtain a holistic view of their genomes.
Twelve isolates were selected for genome sequencing and after this was done, the 12 genomes
were submitted to Genbank. They were also compared to the North American M1404 strain and
the four available complete P. multocida genomes that are available on Genbank. This comparison
identified a set of 96 unique genes to HS-associated isolates. However, the comparison did not
include the genome of African serotype E:2 which is also associated with HS. There is no currently
available genome sequence of any strain of this serovar. The 96 unique genes were not studied
further. This aim was partially achieved and the work done in this study should form the basis of
future studies in this important area.
3. To produce a cheap, rapid, reliable and accurate point-of-care diagnostic test to
diagnose HS in clinical samples from buffaloes and cattle.
Following the comparative genomics study done on different field and vaccine strains associated
with HS disease from Pakistan and Thailand, and after identifying 96 unique genes that were found
only in HS isolates, it became clear that testing for the presence of one of these sequences may be
adequate to detect the presence of P. multocida serovar B:2 (HS disease). A LAMP test that could
rapidly and specifically identify the BUKK_07490 gene was designed, produced and validated on
positive control material. However, clinical samples were not available during this project and so
overall, this aim was only partially achieved.
148 | P a g e
4. To apply this diagnostic test to the field in endemic countries to facilitate diagnosis of
the disease.
To meet this aim, a period of time had to be spent in Pakistan to collect and then test field
samples. This was not performed due to time restraints and the political situation in Pakistan at
the only time this study could have been done; in 2013, during the Pakistani national elections.
Moreover, sending samples out of the country was not an option because there were only very
few samples collected, that had been inappropriately stored and also the importation of these
samples into Australia requires permits that cannot always be obtained. This aim was not achieved
in this project but we are working now on collaboration possibilities with a governmental institute
in Pakistan and it is anticipated that the field evaluation of the test will start soon.
There is still a large amount of important research that should follow this project to extend the
findings discovered so far. Some directions that this research may take are discussed.
7.3 Future Directions of Research
Epidemiological work
Investigations should be pursued on the five factors contained in the final multivariable logistic
regression model: vaccination, using underground water, presence of FMD in the farm, the period
of keeping cattle at the farm and the number of animals at the farm. Further studies on
vaccination types being used, their storage and their efficacy should be done. Moreover, studying
the effect of improving the quality of drinking water to the animals on alleviating the burden of HS
should be done. In addition, the interaction between HS and FMD should be further studied in the
future to show if FMD biologically predisposes for HS or is merely just a surrogate marker for poor
hygiene and/or biosecurity.
149 | P a g e
Further investigation of the twelve genomes sequenced
The identified unique genes of HS isolates should be further studied to understand the role (if any)
that these genes play in disease pathogenesis, virulence and host specificity. Unique genes found
in Asian isolates, and not in the North American isolates, should be further investigated to check
their effect on virulence. The four identified putative phages should be further studied. The
integrative conjugative element (ICE) that was found in two Pakistani isolates should also be
isolated and studied to check its mobility and determine its relatedness to ICEPmu1 of strain
36950.
Genome sequence of African serotype E:2 associated with HS
Currently, genomes of P. multocida associated with HS from Pakistan, India and Thailand are
available as well as a genome of an M1404 strain from North America. All these isolates are
serovar B:2. Genome sequencing of isolates from the E:2 serovar will allow us to compare these
genomes with the B:2 genome and see if there are shared genes; genes that potentially would be
responsible for pathogenicity.
Reviewing the HS-LAMP test and evaluation of performance in the field
It was shown in this thesis that DNA purification was not needed when using cultured bacteria as a
template for LAMP testing but work that is focussed on processing clinical samples (e.g. blood)
should investigate whether DNA extraction methods can be avoided for these sample types as
well. The HS-LAMP test should also be evaluated under field conditions. The ultimate goal of these
investigations is to produce a genuine “point-of-care” test. The E:2 serovar should be tested to see
if the HS-LAMP test described in this thesis, will be able to detect it. This could then broaden the
applicability of this test to a global context.
150 | P a g e
Due to time limitations and restricted access to strains at the time of validating the test, the
number of samples tested was only enough for optimisation. To accurately calculate the sensitivity
and specificity, more taxa should be tested and the test should be validated in a field setting. A
communication has already started with the Animal Production and Health Section (Joint
FAO/IAEA Division of Nuclear Techniques in Food and Agriculture) to evaluate the test under field
conditions.
It is hoped that the ability to detect the disease will allow the control of HS. It is hoped that in the
not-too-distant future, the HS-LAMP test will be adopted by Pakistani veterinarians in the field.
This test could facilitate the decision making processes that veterinarians need to make so that
they can save more animals and limit the spread of disease.
Vaccination
Screening the 96 identified unique genes for potential candidates that can be used for vaccine
development should form a research priority. Bacterial mutants could be developed that are
protective and not virulent. Alternatively, a DNA-based vaccine could be developed that would
have the advantage that it cannot revert to virulence (van Drunen Littel-van den Hurk et al., 1998;
Harpin et al., 1999; van Drunen Littel-van den Hurk et al., 2001; Hu et al., 2009).
7.4 Conclusion
This thesis has described the first genome sequencing of HS-associated field and vaccine isolates
from Pakistan and Thailand. A rapid and specific LAMP test was subsequently designed to facilitate
the detection of the particular serovar in limited infrastructure settings. Future research directions
have been clearly identified that will hopefully continue to contribute to the collective knowledge
of HS disease for many years to come.
151 | P a g e
Bibliography
Abrahante JE, Johnson TJ, Hunter SS, Maheswaran SK, Hauglund MJ, Bayles DO, Tatum FM, Briggs RE. Draft genome sequences of two virulent serotypes of avian Pasteurella multocida. Genome Announcements. 2013;1(1).
Adhikary S, Bisgaard M, Foster G, Kiessling N, R. FA, E. OJ, Christensen H. Comparative study of PCR methods to detect Pasteurella multocida. Berliner und Munchener Tierarztliche Wochenschrift. 2013;126(9-10):415-22.
Afzal MM, Hussain M, Khan E, Ferrari G. FMD control in dairy colonies milk production system in Pakistan. Scientific Developments and Technical Challenges in the Progressive Control of Foot-and-Mouth Disease (FMD) in South Asia Conference, (13-15 February 2012) in New Delhi, India [online] Available from: http://wwwfaoorg/docs/eims/upload/299852/an372e00pdf [Accessed July, 2014]; 2012.
Ahir VB, Roy A, Jhala MK, Bhanderi BB, Mathakiya RA, Bhatt VD, Padiya KB, Jakhesara SJ, Koringa PG, Joshi CG. Genome sequence of Pasteurella multocida subsp. gallicida Anand1_poultry. Journal of Bacteriology. 2011;193(19):5604.
Aktories K, Orth JHC, Adler B. Pasteurella multocida: Molecular Biology, Toxins and Infection: Springer; 2012. 157 p.
Ali SN, Asif M, Rehman A, Jat LA, Ali Q, Hussain M. Participatory Surveillance of Livestock Diseases in District Karachi–Pakistan. International Journal of Agriculture & Biology. 2006;8(5):652-6.
Alikhan N-F, Petty N, Ben Zakour N, Beatson S. BLAST Ring Image Generator (BRIG): Simple prokaryote genome comparisons. BMC Genomics. 2011;12(1):402.
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. Journal of Molecular Biology. 1990;215(3):403-10.
Anba J, Bidnenko E, Hillier A, Ehrlich D, Chopin MC. Characterization of the lactococcal abiD1 gene coding for phage abortive infection. Journal of Bacteriology. 1995;177(13):3818-23.
Angiuoli SV, Gussman A, Klimke W, Cochrane G, Field D, Garrity GM, Kodira CD, Kyrpides N, Madupu R, Markowitz V, Tatusova T, Thomson N, White O. Toward an online repository of standard operating procedures (SOPs) for (meta)genomic annotation. OMICS: A Journal of Integrative Biology. 2008;12(2):137-41.
Anjum R, Manzoor H, Bin Zahoor A, Irshad H, Farooq U. Epidemiological analyses of foot and mouth disease in Pakistan. International Journal of Agriculture and Biology. 2006;5:684-51.
Annas S, Zamri-Saad M, Jesse FF, Zunita Z. New sites of localisation of Pasteurella multocida B:2 in buffalo surviving experimental haemorrhagic septicaemia. BMC Veterinary Research. 2014;10(1):88.
Anon. Simple serological technique recommended for haemorrhagic septicaemia diagnosis. Asian Livestock. 1981;6(5):41-2.
152 | P a g e
Anonymous. Economic analysis and survey planning: Epidemiology, Punjab, 1994. Pakistan directorate of planning and evaluation department of livestock and dairy development, Punjab; 1996.
Awad FI, Salem AA, Fayed AA. Studies on clinical signs observed on experimentally infected animals with Pasteurella multocida type I. Egyptian Journal of Veterinary Science. 1976;13:53-6.
Bain RVS, Knox KW. The antigens of Pasteurella multocida type I. II: Lipopolysaccharides. Immunology. 1961;4:122-9.
Bain RVS, De Alwis MCL, Carter GR, Gupta BK. Haemorrhagic septicaemia. FAO animal production and health paper No 33. Rome, 1982.
Bandopadhyay PK, Tonganokar SS, Singh DK. Characterisation and antibiotic sensitivity of Pasteurella multocida isolates from cases of haemorrhagic septicaemia. Proceedings of the fourth international workshop on haemorrhagic septicaemia; Kandy, Sri Lanka: FAO-APHCA Publication No. 1991/13; 1991. p. 65-8.
Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, Lesin VM, Nikolenko SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV, Vyahhi N, Tesler G, Alekseyev MA, Pevzner PA. SPAdes: A new genome assembly algorithm and its applications to single-cell sequencing. Journal of Computational Biology. 2012;19(5):455-77.
Benkirane A, De Alwis MCL. Haemorrhagic septicaemia, its significance, prevention and control in Asia. Veterinarni Medicina. 2002;47(8):234–40.
Benson DA, Cavanaugh M, Clark K, Karsch-Mizrachi I, Lipman DJ, Ostell J, Sayers EW. GenBank. Nucleic Acids Research. 2013;41(Database issue):D36-D42.
Bisht KS, Hassan L, Saharee AA, Zunita Z, Nadzri S. Outbreak investigation of haemorrhagic septicaemia in buffalo and cattle in Perak, Malaysia. In: Iskandar FCTN, Hassan L, Dhaliwal GK, Yusoff R, Omar A, Khan MKG, Meng GY, Abu A, Arshad SS, Salim N, Abdul Aziz S, Daud HM, Abdul Rahman SO, editors. Animal health : a breakpoint in economic development : 11th International for Tropical Veterinary Medicine & 16th Veterinary Association Malaysia Congress, 23-27 August 2004. Petaling Jaya, Malaysia: Universiti Putra Malaysia Press; 2004. p. 263-5.
Biswas A, Kumar AA, Saxena MK, Srivastava SK, Singh VP, Shivachandra SB. Molecular variability among strains of Pasteurella multocida isolated from an outbreak of haemorrhagic septicaemia in India. Veterinary Research Communications. 2004;28(4):287-98.
Boyce J, Chung JY, Adler B. Genetic organisation of the capsule biosynthetic locus of Pasteurella multocida M1404 (B:2). Veterinary Microbiology. 2000a;72(1-2):121-34.
Boyce JD, Chung JY, Adler B. Pasteurella multocida capsule: composition, function and genetics. Journal of Biotechnology. 2000b;83(1-2):153-60.
Boyce JD, Seemann T, Adler B, Harper M. Pathogenomics of Pasteurella multocida. In: Aktories K, Orth JHC, Adler B, editors. Pasteurella multocida: Molecular Biology, Toxins and Infection. Current Topics in Microbiology and Immunology: Springer Berlin Heidelberg; 2012. p. 23-38.
153 | P a g e
Boyd EF, Brüssow H. Common themes among bacteriophage-encoded virulence factors and diversity among the bacteriophages involved. Trends in Microbiology. 2002;10(11):521-9.
Brickell S. Development of a polymerase chain reaction (PCR) test specific for isolates of P. multocida causing HS. The international workshop on diagnosis and control of haemorrhagic septicaemia; Bali, Indonesia: ACIAR; 1996.
Brickell SK, Thomas LM, Long KA, Panaccio M, Widders PR. Development of a PCR test based on a gene region associated with the pathogenicity of Pasteurella multocida serotype B:2, the causal agent of haemorrhagic septicaemia in Asia. Veterinary Microbiology. 1998;59(4):295-307.
Campoy S, Aranda J, Àlvarez G, Barbé J, Llagostera M. Isolation and sequencing of a temperate transducing phage for Pasteurella multocida. Applied and Environmental Microbiology. 2006;72(5):3154-60.
Carter GR. Type specific capsular antigens of Pasteurella multocida. Canadian Journal of Medical Science. 1952;30:48-53.
Carter GR. A haemagglutination test for identification of serological types. American Journal of Veterinary Research. 1955;16:481-4.
Carter GR. Studies on Pasteurella multocida. III: A serological survey of bovine and porcine strains from various parts of the world. American Journal of Veterinary Research. 1957;18:437-40.
Carter GR. Studies on Pasteurella multocida. IV: Serological types from species other than cattle and swine. American Journal of Veterinary Research. 1959;21:173-5.
Carter GR. A new serological type of Pasteurella multocida from Central Africa. Veterinary Record. 1961;73:1052.
Carter GR, Rappay DE. Formalinised erythrocytes in the haemagglutination test for typing Pasteurella multocida. British Veterinary Journal. 1962;118:289-92.
Carter GR. Proposed modification to the serological classification of Pasteurella multocida. Veterinary Record. 1963;72, 1264.
Carter GR, Subronto P. Identification of type D strains of Pasteurella multocida with acriflavine. American Journal of Veterinary Research. 1973;34:293-4.
Carter GR, Rundell SW. Identification of type A strains of Pasteurella multocida using staphylococcal hyaluronidase. Veterinary Record. 1975;96:343.
Carter GR, Chengappa MM. Hyaluronidase production by type B Pasteurella multocida from cases of hemorrhagic septicemia. Journal of Clinical Microbiology. 1980;11(1):94-6.
Carter GR, Chengappa MM. Identification of types B and E Pasteurella multocida by counterimmunoelectrophoresis. Veterinary Record. 1981;108:145-6.
Carter GR, De Alwis MCL. Haemorrhagic septicaemia. In: Adlam C, Rutter JM, editors. Pasteurella and Pasteurellosis: Academic Press, London; 1989. p. 131-60.
154 | P a g e
Chandrasekaran S. The use of ceftiofur in the treatment of cattle experimentally infected with Pasteurella multocida type B. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 188-90.
Chandrasekaran S, Kennett L, Muniandy N, Yeap PC, Rani B, Mukkur TKS. Characterisation of the immune response and duration of immunity in buffalo vaccinated with cellular haemorrhagic septicaemia vaccines. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 165-9.
Chaudhuri S, Mukherjee SK, Chatterjee A, Gauguli JL. Isolation of Pasteurella multocida F:3,4 from a stillborn snow leopard. Veterinary Record. 1992;130:36.
Chengappa MM, Carter GR. Demonstration of bacteriocin activity in bovine and bison strains of Pasteurella multocida. American Journal of Veterinary Research. 1977;38:1183-5.
Christensen H, Angen Ø, Olsen JE, Bisgaard M. Revised description and classification of atypical isolates of Pasteurella multocida from bovine lungs based on genotypic characterization to include variants previously classified as biovar 2 of Pasteurella canis and Pasteurella avium. Microbiology. 2004;150(6):1757-67.
Christensen H, Bisgaard M. Taxonomy and biodiversity of members of Pasteurellaceae. In: Kuhnert P, Christensen H, editors. Pasteurellaceae: Biology, Genomics and Molecular Aspects. Norfolk, UK: Caister Academic Press; 2008. p. 1-26.
Christensen H, Korczak BM, Bojesen AM, Kuhnert P, Frederiksen W, Bisgaard M. Classification of organisms previously reported as the SP and Stewart–Letscher groups, with descriptions of Necropsobacter gen. nov. and of Necropsobacter rosorum sp. nov. for organisms of the SP group. International Journal of Systematic and Evolutionary Microbiology. 2011;61(8):1829-36.
Christensen H, Bertelsen MF, Bojesen AM, Bisgaard M. Classification of Pasteurella species B as Pasteurella oralis sp. nov. International Journal of Systematic and Evolutionary Microbiology. 2012;62(Pt 6):1396-401.
Christensen H, Nicklas W, Bisgaard M. Investigation of taxa of Pasteurellaceae isolated from Syrian and European hamsters and proposal of Mesocricetibacter intestinalis gen. nov., sp. nov. and Cricetibacter osteomyelitidis gen. nov., sp. nov. International Journal of Systematic and Evolutionary Microbiology. 2014.
Chung JY, Zhang YM, Adler B. The capsule biosynthetic locus of Pasteurella multocida A:1. FEMS Microbiology Letters. 1998;166(2):289-96.
Corney BG, Diallo IS, Wright LL, Hewitson GR, De Jong AJ, Burrell PC, Duffy PF, Stephens CP, Rodwell BJ, Boyle DB, Blackall PJ. Pasteurella multocida detection by 5' Taq nuclease assay: a new tool for use in diagnosing fowl cholera. Journal of Microbiological Methods. 2007;69(2):376-80.
Dabo SM, Confer AW, Lu YS. Single primer polymerase chain reaction fingerprinting for Pasteurella multocida isolates from laboratory rabbits. American Journal of Veterinary Research. 2000;61(3):305-9.
155 | P a g e
Darling AE, Mau B, Perna NT. progressiveMauve: Multiple genome alignment with gene gain, loss and rearrangement. Plos One. 2010;5(6):e11147.
Dartini NL, Ekaputra IGM. Abattoir survey for isolation of Pasteurella multocida in the eastern region of Indonesia. International Workshop on Diagnosis and Control of Haemorrhagic Septicaemia; Bali, Indonesia: ACIAR; 1996.
Dassanayake L. The haemorrhagic septicaemia outbreak of 1955-56. Ceylon Veterinary Journal. 1957;5:56-9.
Dawkins HJS, Spencer TL, Patten BE, Johnson RB. Rapid identification of Pasteurella multocida organisms responsible for haemorrhagic septicaemia using an enzyme-linked immunosorbent assay (ELISA). Research in Veterinary Science. 1990;49(3):261-7.
De Alwis MCL, Thambithurai V. A case of haemorrhagic septicaemia in a wild elephant in Ceylon. Ceylon Veterinary Journal. 1965;13:17-9.
De Alwis MCL. Bacteriological changes in specimens during transport. II: Effect of using a suitable transport medium. Ceylon Veterinary Journal. 1973;21:2-6.
De Alwis MCL, Jayasekera MU, Balasunderam P. Pneumonic pasteurellosis in buffalo calves associated with Pasteurella multocida serotype 6:B. Ceylon Veterinary Journal. 1975;23:58-60.
De Alwis MCL, Kodituwakku AO, Kodituwakku S. Haemorrhagic septicaemia: An analysis of two outbreaks of disease among buffaloes. Ceylon Veterinary Journal. 1976;24:18-21.
De Alwis MCL, Vipulasiri AA. An epizootiological study of haemorrhagic septicaemia in Sri Lanka. Ceylon Veterinary Journal. 1980;28:24-35.
De Alwis MCL. Mortality among cattle and buffaloes in Sri Lanka due to haemorrhagic septicaemia. Tropical Animal Health and Production. 1981;13:195-202.
De Alwis MCL. Pasteurella multocida serotype 6:B from an elephant. Sri Lanka Veterinary Journal. 1982a;30:28.
De Alwis MCL. The immune status of buffalo calves exposed to natural infection with haemorrhagic septicaemia. Tropical Animal Health and Production. 1982b;14:29-30.
De Alwis MCL, Sumanadasa MA. Naturally acquired immunity to haemorrhagic septicaemia among cattle and buffaloes in Sri Lanka. Tropical Animal Health and Production. 1982;14:27-8.
De Alwis MCL. Haemorrhagic septicaemia in cattle and buffaloes. Revue Scientifique Technique Office International des Epizooties. 1984;3(4):707-30.
De Alwis MCL, Wijewardana TG, Sivaram A, Vipulasiri AA. The carrier and antibody status of cattle and buffaloes exposed to haemorrhagic septicaemia: Investigations on survivors following natural outbreaks. Sri Lanka Veterinary Journal. 1986;34:33-42.
De Alwis MCL, Wijewardana TG, Gomis AIU, Vipulasiri AA. Persistence of the carrier status in haemorrhagic septicaemia (Pasteurella multocida serotype 6:B infection) in buffaloes. Tropical Animal Health and Production. 1990;22:185-94.
156 | P a g e
De Alwis MCL. Pasteurellosis in production animals: A review. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993a. p. 11-22.
De Alwis MCL. Sri Lanka. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993b. p. 243-5.
De Alwis MCL, Wijewardana TG, Vipulasiri AA. A live vaccine against haemorrhagic septicaemia using a serotype B:2 mutant-preliminary communication. International workshop on diagnosis and control of haemorrhagic septicaemia; Indonesia: ACIAR; 1996.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999a. p. 53-65.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999b. p. 1-10.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999c. p. 43-52.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999d. p. 11-24.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999e. p. 33-42.
De Alwis MCL. Haemorrhagic septicaemia. ACIAR Monograph No. 57. Canberra: Australian Centre for International Agricultural Research; 1999f. p. 25-32.
Dhanda MR. Purification and properties of the soluble antigen of Pasteurella septica type I. Indian Journal of Pathology and Bacteriology. 1960;2:59-62.
Dhanda MR, Nilakanthan PR. Isolation of Pasteurella multocida of low virulenece of cases of Paraplegia in cattle in the Andhra Pradesh. Indian Veterinary Journal. 1961;38:339-43.
Dukes JP, King DP, Alexandersen S. Novel reverse transcription loop-mediated isothermal amplification for rapid detection of foot-and-mouth disease virus. Archives of Virology. 2006;151(6):1093-106.
Dutta J, Rathore BS, Mullick SG, Singh R, Sharma GC. Epidemiological studies on occurance of haemorrhagic septicaemia in India. Indian Veterinary Journal. 1990;67:893-9.
Dziva F, Muhairwa AP, Christensen H, Bisgaard M. Diagnostic and typing options for investigating diseases associated with Pasteurella multocida. Veterinary Microbiology. 2008;128(1-2):1-22.
Dziva F, Christensen H. Pasteurella. In: Liu D, editor. Molecular detection of human bacterial pathogens: Taylor & Francis; 2011. p. 945-58.
157 | P a g e
Euzéby JP. List of bacterial names with standing in nomenclature: A folder available on the Internet. International Journal of Systematic Bacteriology. 1997;47(2):590-2.
FAO. Report of the FAO meeting on haemorrhagic septicaemia. Manila, Philippines: 1959 Nov-Dec.
FAO, editor. Proceedings of the third international workshop on haemorrhagic septicaemia; 1979; Colombo, Sri Lanka: FAO-APHCA.
FAO, editor. Proceedings of the fourth international workshop on haemorrhagic septicaemia; 1991; Kandy, Sri Lanka: FAO-APHCA Publication No. 1991/13.
FAO. Milk production [online]. Available from: http://www.fao.org/agriculture/dairy-gateway/milk-production/en/#.U89wLBCSzSl [Accessed October 1, 2014] 2014.
Farooq O. Pakistan Economic Survey: Agriculture (2013-2014) [document on the Internet]. Available from: http://finance.gov.pk/survey/chapters_14/02_Agriculture.pdf [Accessed: July 13, 2014]; 2014.
Farooq U, Hussain M, Irshad H, Badar N, Munir R, Ali Q. Status of haemorrhagic septicaemia based on epidemiology in Pakistan. Pakistan Veterinary Journal. 2007;27(2):67-72.
Farooq U, Saeed Z, Khan MA, Ali I, Qamar MF. Sero-surveillance of hemorrhagic septicemia in buffaloes and cattle in Southern Punjab, Pakistan. Pakistan Veterinary Journal. 2011;31(3):254-6.
Foster G, Higgins R, Leclair D, Korczak BM, Mikaelian I, Patterson IAP, Kuhnert P. Proposal of Bisgaardia hudsonensis gen. nov., sp. nov. and an additional genomospecies, isolated from seals, as new members of the family Pasteurellaceae. International Journal of Systematic and Evolutionary Microbiology. 2011;61(12):3016-22.
Francis BKT, Schels HF, Carter GR. Type E Pasteurella multocida associated with haemorrhagic septicaemia in Zambia. Veterinary Record. 1980;107:135.
Francois P, Tangomo M, Hibbs J, Bonetti E-J, Boehme CC, Notomi T, Perkins MD, Schrenzel J. Robustness of a loop-mediated isothermal amplification reaction for diagnostic applications. FEMS Immunology and Medical Microbiology. 2011;62(1):41-8.
Frandsen PL, Foged NT, Petersen SK, Bording A. Characterization of toxin from different strains of Pasteurella multocida serotype A and D. Zentralbl Veterinarmed B. 1991;38(5):345-52.
Frank GH. Pasteurellosis of cattle. In: Adlam C, Rutter JM, editors. Pasteurella and Pasteurellosis: Academic Press, London; 1989. p. 197-222.
Frankena K, Graat EAM. Multivariate analysis: Logistic regression. In: Noordhuizen JPTM, Frankena K, van der Hoofd CM, Graat EAM, editors. Application of Quantitative Methods in Veterinary Epidemiology. Wageningen: Wageningen Pers; 1997. p. 137-78.
Garrity GM, Bell JA, Lilburn T. Order XIV. Pasteurellales ord. nov. In: Brenner DJ, Krieg NR, Staley JT, Garrity GM, editors. Bergey's Manual of Systematic Bacteriology. 2 (The Proteobacteria), part B (The Gammaproteobacteria). second ed: Springer, New York; 2005. p. 850-1.
158 | P a g e
Garvey P, Fitzgerald GF, Hill C. Cloning and DNA sequence analysis of two abortive infection phage resistance determinants from the lactococcal plasmid pNP40. Applied and Environmental Microbiology. 1995;61(12):4321-8.
Gilson E, Clement JM, Perrin S, Hofnung M. Palindromic units: A case of highly repetitive DNA sequences in bacteria. Trends in Genetics. 1987;3:226-30.
Go MF, Chan KY, Versalovic J, Koeuth T, Graham DY, Lupski JR. Cluster analysis of Helicobacter pylori genomic DNA fingerprints suggests gastroduodenal disease-specific associations. Scandinavian Journal of Gastroenterology. 1995;30(7):640-6.
Graydon RJ, Patten BE, Hamid H. The pathology of experimental haemorrhagic septicaemia in cattle and buffalo. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 105-7.
Gregersen RH, Neubauer C, Christensen H, Bojesen AM, Hess M, Bisgaard M. Comparative studies on [Pasteurella] testudinis and [P.] testudinis-like bacteria and proposal of Chelonobacter oris gen. nov., sp. nov. as a new member of the family Pasteurellaceae. International Journal of Systematic and Evolutionary Microbiology. 2009;59(7):1583-8.
Gupta BK. Studies on the carrier problem in haemorrhagic septicaemia. Thesis, Punjab University, Chandigar, India1962.
Gurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013;29(8):1072-5.
Hansen MJ, Bertelsen MF, Christensen H, Bojesen AM, Bisgaard M. Otariodibacter oris gen. nov., sp. nov., a member of the family Pasteurellaceae isolated from the oral cavity of pinnipeds. International Journal of Systematic and Evolutionary Microbiology. 2012;62:2572-8.
Harper M, Boyce JD, Cox AD, St. Michael F, Wilkie IW, Blackall PJ, Adler B. Pasteurella multocida expresses two lipopolysaccharide glycoforms simultaneously, but only a single form is required for virulence: Identification of two acceptor-specific heptosyl I transferases. Infection and Immunity. 2007;75(8):3885-93.
Harpin S, Hurley DJ, Mbikay M, Talbot B, Elazhary Y. Vaccination of cattle with a DNA plasmid encoding the bovine viral diarrhoea virus major glycoprotein E2. Journal of General Virology. 1999;80 ( Pt 12):3137-44.
Hector JSR, Pang YJ, Wallace RJ, Mazurek GH, Zhang YS, Brown BA. Large restriction fragment patterns of genomic Mycobacterium fortuitum DNA as strain-specific markers and their use in epidemiologic investigation of 4 nosocomial outbreaks. Journal of Clinical Microbiology. 1992;30(5):1250-5.
Heddleston KL. Fowl cholera. In: Hofstad MS, Calnek BW, Helmboldt CF, Reid WM, Yoder HW, editors. Diseases of poultry. 6th edition ed. Ames: Iowa state university press; 1972. p. 219-41.
Heddleston KL, Gallagher JE, Rebers PA. Fowl cholera: Gel diffusion precipitin test for serotyping Pasteruella multocida from avian species. Avian Diseases. 1972;16:925-36.
159 | P a g e
Higgins CF, Ames GFL, Barnes WM, Clement JM, Hofnung M. A novel intercistronic regulatory element of prokaryotic origin. Nature. 1982;298:760-2.
Hiramune T, De Alwis MCL. Haemorrhagic septicaemia carrier status among cattle and buffaloes in Sri Lanka. Tropical Animal Health and Production. 1982;14:91-2.
Horadagoda NU, De Alwis MCL, Wijewardana TG, Belak K, Gomis AIU, Vipulasiri AA. Experimental haemorrhagic septicaemia in buffalo calves. FAO; 1991. p. 73-80.
Hosmer DW, Lemeshow S. Applied Logistic Regression. New York: Wiley; 1989.
Hotchkiss EJ, Hodgson JC, Lainson FA, Zadoks RN. Multilocus sequence typing of a global collection of Pasteurella multocida isolates from cattle and other host species demonstrates niche association. BMC Microbiology. 2011;11:115.
Hu XD, Yu DH, Chen ST, Li SX, Cai H. A combined DNA vaccine provides protective immunity against Mycobacterium bovis and Brucella abortus in cattle. DNA and Cell Biology. 2009;28(4):191-9.
Hulton CSJ, Higgins CF, Sharp PM. ERIC sequences: A novel family of repetitive elements in the genomes of Escherichia coli, Salmonella typhimurium and other enterobacteria. Molecular Microbiology. 1991;5(4):825-34.
Huson DH, Bryant D. Application of Phylogenetic Networks in Evolutionary Studies. Molecular Biology and Evolution. 2006;23(2):254-67.
ICSP. Homepage of Pasteurellaceae taxonomic subcommittee, International Committee on Systematics of Prokaryotes (ICSP) [online]. Available from: http://www.the-icsp.org/taxa/Pasteurellaceaelist.htm [Accessed: October 01, 2014]; 2012.
Imran M, Irshad M, Shahid MA, Ashraf M. Studies on the carrier status of Pasteurella multocida in healthy cattle and buffalo in district Faisalabad. International Journal of Dairy Science. 2007;2(4):398-400.
Iwamoto T, Sonobe T, Hayashi K. Loop-mediated isothermal amplification for direct detection of Mycobacterium tuberculosis complex, M. avium, and M. intracellulare in sputum samples. Journal of Clinical Microbiology. 2003;41(6):2616-22.
Jindal N, Kumar S, Narang G, Chaturvedi GC, Tomer P, Garg DN. Some epidemiological observations on haemorrhagic septicaemia in buffaloes and cattle in Haryana state of India. Buffalo Journal. 2002;2:273-80.
Johnson RB, Dawkins HJS, Spencer TL. Electrophoretic profiles of Pasteurella multocida isolates from animals with hemorrhagic septicemia. American Journal of Veterinary Research. 1991;52(10):1644-8.
Kahn CM, Line S. The Merck veterinary manual. 10 ed. Whitehouse Station, N.J.: Merck & Co., INC.; 2010. 2945 p.
Kaneko H, Kawana T, Fukushima E, Suzutani T. Tolerance of loop-mediated isothermal amplification to a culture medium and biological substances. Journal of Biochemical and Biophysical Methods. 2007;70(3):499-501.
160 | P a g e
Karachi metropolitan corporation. Karachi map [online]. Available from: http://www.kmc.gos.pk/Contents.aspx?id=94 [Accessed October 1, 2014]; 2014.
Kasali OB. A case of haemorrhagic septicaemia in an African buffalo (Syncerus nanus). Bulletin of Epizootic Diseases of Africa. 1972;20:203-4.
Kasten RW, Carpenter TE, Snipes KP, Hirsh DC. Detection of Pasteurella multocida-specific DNA in turkey flocks by use of the polymerase chain reaction. Avian Diseases. 1997;41(3):676-82.
Kawasaki M, Young JR, Suon S, Bush RD, Windsor PA. The socioeconomic impacts of clinically diagnosed haemorrhagic septicaemia on smallholder large ruminant farmers in Cambodia. Transboundary and Emerging Diseases. 2013.
Kennett L. Comparative protective potential of non-living intact cells and purified outer membrane and associated proteins of Pasteurella multocida type 6:B grown under iron-regulated conditions In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 144-8.
Khan A, Saddique U, Ahmad R, Khan H, Mohammad Y, Zubair M. Sero-surveillance of hemorrhagic septicemia in cattle and buffaloes in district Malakand, NWFP. Journal of Agricultural and Biological Science. 2006;1(4):11-4.
Khan FM. Participatory appraisal and scanning surveillance based contagious diseases risk profile of district Rahim Yar Khan (Pakistan). Pakistan Veterinary Journal. 2010;30(4):198-202.
Khan FM. Field epidemiology of an outbreak of hemorrhagic septicemia in dromedary population of greater Cholistan desert (Pakistan). Pakistan Veterinary Journal. 2012;32(1):31-4.
Khan UN, Lund JA, Javaid S, Zia-ul-Hasan. Economic analysis of milk production in different cattle colonies of Karachi. Pakistan Journal of Agricultural Sciences. 2008;45(2):403-9.
Kheng CS, Phay CH. Haemorrhagic septicaemia: Sulphamethazine and immune serum therapy in buffaloes infected by the nasal spray method. Veterinary Record. 1963;75:155-9.
Klein J, Hussain M, Ahmad M, Afzal M, Alexandersen S. Epidemiology of foot-and-mouth disease in Landhi Dairy Colony, Pakistan, the world largest buffalo colony. Virology Journal. 2008;5(1):53.
Knox KW, Bain RVS. The antigens of Pasteurella multocida. I: Capsular polysaccharides. Immunology. 1960;3(352-362).
Korczak B, Bisgaard M, Christensen H, Kuhnert P. Frederiksenia canicola gen. nov., sp. nov. isolated from dogs and human dog-bite wounds. Antonie Van Leeuwenhoek. 2014;105(4):731-41.
Koressaar T, Remm M. Enhancements and modifications of primer design program Primer3. Bioinformatics. 2007;23(10):1289-91.
Kramer GC. Hypertonic resuscitation: Physiological mechanisms and recommendations for trauma care. Journal Trauma. 2003;54:89-99.
161 | P a g e
Kuboki N, Inoue N, Sakurai T, Di Cello F, Grab DJ, Suzuki H, Sugimoto C, Igarashi I. Loop-mediated isothermal amplification for detection of African trypanosomes. Journal of Clinical Microbiology. 2003;41(12):5517-24.
Kuhnert P, Scholten E, Haefner S, Mayor D, Frey J. Basfia succiniciproducens gen. nov., sp. nov., a new member of the family Pasteurellaceae isolated from bovine rumen. International Journal of Systematic and Evolutionary Microbiology. 2010;60(1):44-50.
Ligniéres J. Contribution à l’etude et à la classification des septicémies hémorrhagiques. Les Pasteurelloses. Annales Institut Pasteur. 1900;15:734-6. As cited by Henderson A. Pasteurella multocida infection in man; a review of the literature. Antonie Van Leeuwenhoek. 1963;29(1):359-67.
Liu D, Lawrence ML, Austin FW. Specific PCR identification of Pasteurella multocida based on putative transcriptional regulator genes. Journal of Microbiological Methods. 2004;58(2):263-7.
Liu W, Yang M, Xu Z, Zheng H, Liang W, Zhou R, Wu B, Chen H. Complete genome sequence of Pasteurella multocida HN06, a toxigenic strain of serogroup D. Journal of Bacteriology. 2012;194(12):3292-3.
Lupski JR, Weinstock GM. Short, interspersed repetitive DNA-sequences in prokaryotic genomes. Journal of Bacteriology. 1992;174(14):4525-9.
Lusted LB. Decision-making studies in patient management. New England Journal of Medicine. 1971;284(8):416-24.
Mahmood AK, Sheikh MA, Akhtar S, Nabi G, Rashid HB. Duration of maternally derived antibodies against Pasteurella multocida in cow calves. Pakistan Veterinary Journal. 2007;27(2):92-4.
Maiden MC, Bygraves JA, Feil E, Morelli G, Russell JE, Urwin R, Zhang Q, Zhou J, Zurth K, Caugant DA, Feavers IM, Achtman M, Spratt BG. Multilocus sequence typing: A portable approach to the identification of clones within populations of pathogenic microorganisms. PNAS. 1998;95(6):3140-5.
Manning PJ, DiGiacomo RF, DeLong D. Pasteurellosis in laboratory animals. In: Adlam C, Rutter JM, editors. Pasteurella and Pasteurellosis: Academic Press, London; 1989. p. 263-302.
Mari FM. Status and constraints of livestock production in Sindh province of Pakistan. International Journal of Advanced Rural Management. 2013;1(1):1-10.
Martrenchar A. Haemorrhagic septicaemia in Cameroon. Veterinary Record. 1993;133:25-6.
May BJ, Zhang Q, Li LL, Paustian ML, Whittam TS, Kapur V. Complete genomic sequence of Pasteurella multocida,Pm70. PNAS. 2001;98(6):3460-5.
Metz CE. Basic principles of ROC analysis. Seminars in Nuclear Medicine. 1978;8(4):283-98.
Michael GB, Kadlec K, Sweeney MT, Brzuszkiewicz E, Liesegang H, Daniel R, Murray RW, Watts JL, Schwarz S. ICEPmu1, an integrative conjugative element (ICE) of Pasteurella multocida: Structure and transfer. Journal of Antimicrobial Chemotherapy. 2012a;67(1):91-100.
162 | P a g e
Michael GB, Kadlec K, Sweeney MT, Brzuszkiewicz E, Liesegang H, Daniel R, Murray RW, Watts JL, Schwarz S. ICEPmu1, an integrative conjugative element (ICE) of Pasteurella multocida: Analysis of the regions that comprise 12 antimicrobial resistance genes. Journal of Antimicrobial Chemotherapy. 2012b;67(1):84-90.
Miflin JK, Blackall PJ. Development of a 23S rRNA‐based PCR assay for the identification of Pasteurella multocida. Letters in Applied Microbiology. 2001;33(3):216-21.
Mohan K, Singha MN, Singh RP, Gupta GM. A study of immunity against Pasteurella multocida in buffalo calves, and their carrier status. Veterinary Record. 1968;85:155-6.
Mohd-Zain Z, Turner SL, Cerdeño-Tárraga AM, Lilley AK, Inzana TJ, Duncan AJ, Harding RM, Hood DW, Peto TE, Crook DW. Transferable antibiotic resistance elements in Haemophilus influenzae share a common evolutionary origin with a diverse family of syntenic genomic islands. Journal of Bacteriology. 2004;186(23):8114-22.
Moraes MP, Weiblen R, Silva AMd, Tobias FL. Evolution of passive immunity in Holstein heifers. Ciência Rural. 1997;27:435-40.
Mori Y, Kitao M, Tomita N, Notomi T. Real-time turbidimetry of LAMP reaction for quantifying template DNA. Journal of Biochemical and Biophysical Methods. 2004;59(2):145-57.
Mori Y, Notomi T. Loop-mediated isothermal amplification (LAMP): A rapid, accurate, and cost-effective diagnostic method for infectious diseases. Journal of Infection and Chemotherapy. 2009;15(2):62-9.
Moustafa AM, Bennett MD, Edwards J, Azim K, Mesaik MA, Choudhary MI, Pathanasophon P, Worarach A, Ali Q, Abubakar M, Anjum R. Molecular typing of haemorrhagic septicaemia-associated Pasteurella multocida isolates from Pakistan and Thailand using multilocus sequence typing and pulsed-field gel electrophoresis. Research in Veterinary Science. 2013;95(3):986-90.
Muhairwa AP, Mtambo MM, Christensen JP, Bisgaard M. Occurrence of Pasteurella multocida and related species in village free ranging chickens and their animal contacts in Tanzania. Veterinary Microbiology. 2001;78(2):139-53.
Mühldorfer K, Speck S, Wibbelt G. Proposal of Vespertiliibacter pulmonis gen. nov., sp. nov. and two genomospecies as new members of the family Pasteurellaceae isolated from European bats. International Journal of Systematic and Evolutionary Microbiology. 2014;64:2424-30.
Mullan PB, Lax AJ. Pasteurella multocida toxin stimulates bone resorption by osteoclasts via interaction with osteoblasts. Calcified Tissue International. 1998;63(4):340-5.
Muniandy N, Edgar J, Woolcock JB, Mukkur TKS. Virulence, purification structure and protective potential of the putative capsular polysaccharide of Pasteurella multocida type 6:B. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 47-54.
Muniandy N, Mukkur TKS. Protective potential of purified lipopolysaccharide versus conjugated oligosaccharide of Pasteurella multocida type B in mice. In: Patten BE, Spencer TL, Johnson RB,
163 | P a g e
Hoffmann D, Lehane L, editors. Pasteurellosis in produaction animals, an international workshop held at Bali, Indonesia, 10-13 August 1992: ACIAR Proceedings No. 43; 1993. p. 149-55.
Munir R, Akhtar S, Afzal MM. Evaluation of three oil–adjuvant vaccines against Pasteurella multocida in buffalo calves. Revue Scientifique Technique Office International des Epizooties. 1994;13:837-43.
Murthy DK, Kaushik RK. Studies on an outbreak of acute swine pasteurellosis due to Pasteurella multocida type B (Carter 1955). Veterinary Record. 1965;77:411-6.
Mustafa AA, Ghalib HW, Shigidi MT. Carrier rate of Pasteurella multocida in cattle associated with an outbreak of haemorrhagic septicaemia in Sudan. British Veterinary Journal. 1978;134:375-8.
Nagamine K, Hase T, Notomi T. Accelerated reaction by loop-mediated isothermal amplification using loop primers. Molecular and Cellular Probes. 2002;16(3):223-9.
Namioka S, Murata M. Serological studies on Pasteurella multocida. III: 'O' antigen analysis of cultures isolated from various animals. Cornell Veterinarian. 1961a;51:522-8.
Namioka S, Murata M. Serological studies on Pasteurella multocida. I: A simplified method for capsular typing of the organisms. Cornell Veterinarian. 1961b;51:498-507.
Namioka S, Murata M. Serological studies on Pasteurella multocida. II: Characteristics of the somatic 'O' antigen of the organism. Cornell Veterinarian. 1961c;51:507-21.
Namioka S, Bruner DW. Serological studies on Pasteurella multocida. IV: Type distribution of organisms on the basis of their capsular and O groups. Cornell Veterinarian. 1963;53:41-3.
Namioka S, Murata M. Serological studies on Pasteurella multocida. V: Some epizootiological findings resulting from 'O' antigen analysis. Cornell Veterinarian. 1964;54:520-34.
Natalia L. Abattoir survey for Pasteurella multocida B:2 in Indonesia using conventional and PCR technology. The international workshop on diagnosis and control of haemorrhagic septicaemia; Bali, Indonesia: ACIAR; 1996.
Naz S, Hanif A, Maqbool A, Ahmed S, Muhammand K. Isolation, characterization and monitoring of antibiotic resistance in Pasteurella multocida isolates from buffalo (Bubalus bubalis) herds around Lahore. The Journal of Animal and Plant Sciences. 2012;22(3):242-5.
Neramitmansook P. Thailand. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop sponsored by ACIAR held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 234-7.
Newbury SF, Smith NH, Robinson EC, Hiles ID, Higgins CF. Stabilization of translationally active mRNA by prokaryotic REP sequences. Cell. 1987;48:2297-310.
Notomi T, Okayama H, Masubuchi H, Yonekawa T, Watanabe K, Amino N, Hase T. Loop-mediated isothermal amplification of DNA. Nucleic Acids Research. 2000;28(12):e63.
164 | P a g e
Ohtsuka K, Yanagawa K, Takatori K, Hara-Kudo Y. Detection of Salmonella enterica in naturally contaminated liquid eggs by loop-mediated isothermal amplification, and characterization of Salmonella isolates. Applied and Environmental Microbiology. 2005;71(11):6730-5.
OIE. Haemorrhagic septicaemia (Chapter 2.4.12), Manual of Diagnostic Tests and Vaccines for Terrestrial Animals [online]. Available from: http://www.oie.int/fileadmin/Home/eng/Health_standards/tahm/2.04.12_HAEMORRHAGIC_SEPTICAEMIA.pdf [Accessed October 1, 2014]; 2012.
OIE. Haemorrhagic septicaemia, detailed country(ies) disease incidence [online]. Available from: http://www.oie.int/wahis_2/public/wahid.php/Diseaseinformation/statusdetail [Accessed October 1, 2014]; 2014.
Orth JH, Aktories K. Molecular biology of Pasteurella multocida toxin. Current Topics in Microbiology and Immunology. 2012;361:73-92.
Parida M, Sannarangaiah S, Dash PK, Rao PVL, Morita K. Loop-mediated isothermal amplification (LAMP): A new generation of innovative gene amplification technique; perspectives in clinical diagnosis of infectious diseases. Reviews in Medical Virology. 2008;18(6):407-21.
Pavri KM, Apte VH. Isolation of Pasteurella multocida from a fatal disease of horses and donkeys in India. Veterinary Record. 1967;80:437-9.
Penn CW, Nagy LK. Isolation of a protective non toxic capsular antigen from Pasteurella multocida types B and E. Research in Veterinary Science. 1976;20:90-6.
Perumalpillai C, Thambiayah VS. Outbreak of haemorrhagic septicaemia (Pasteurella multocida type 1 Roberts) in an epidemic form in Ceylon. Ceylon Veterinary Journal. 1957;5:24-8.
Petersen A, Bisgaard M, Townsend K, Christensen H. MLST typing of Pasteurella multocida associated with haemorrhagic septicaemia and development of a real-time PCR specific for haemorrhagic septicaemia associated isolates. Veterinary Microbiology. 2014;170(3–4):335-41.
Pillai AGR, Katiyar AK, Awadhiya RP, Vegad JL. An outbreak of pasteurellosis in swine. Indian Veterinary Journal. 1986;63:527-9.
Pohl S. Family Pasteurellaceae fam. nov. In validation of the publication of new names and new combinations previously effectively published outside the IJSB, list no. 7. International Journal of Systematic Bacteriology. 1981;31:382-3.
Poon LLM, Wong BWY, Ma EHT, Chan KH, Chow LMC, Abeyewickreme W, Tangpukdee N, Yuen KY, Guan Y, Looareesuwan S, Peiris JSM. Sensitive and inexpensive molecular test for falciparum malaria: Detecting Plasmodium falciparum DNA directly from heat-treated blood by loop-mediated isothermal amplification. Clinical Chemistry. 2006;52(2):303-6.
Pullinger GD, Bevir T, Lax AJ. The Pasteurella multocida toxin is encoded within a lysogenic bacteriophage. Molecular Microbiology. 2004;51(1):255-69.
Rahman H, Gajendragad MR, Uma S. Status of haemorrhagic septicaemia in India [online]. Available from:
165 | P a g e
http://nadres.res.in/Nadres_Uploads/UploadedFiles//VetEpiReports/ThirdSlot/Tech%20HS%20final.pdf [Accessed: July 15, 2014]; 2012.
Ramarao D, Rao BU, Ramanathan S. Incidence of haemorrhagic septicaemia in Andhra Pradesh during 1976-85. Indian Journal of Animal Sciences. 1991;61:145-9.
Rebers PA, Heddleston KL, Rhoades KR. Isolation from Pasteurella multocida of a lipopolysaccharide antigen with immunizing and toxic properties. Journal of Bacteriology. 1967;93(1):7-14.
Rhoades KR, Heddleston KL, Rebers PA. Experimental haemorrhagic septicaemia: Gross and microscopic lesions resulting from acute infections and from endotoxin administration. Canadian Journal of Comparative Medicine. 1967;31:226-33.
Rimler RB, Rhoades KR. Serogroup F, a new capsule serogroup of Pasteurella multocida. Journal of Clinical Microbiology. 1987;25(4):615-8.
Rimler RB, Rhoades KR. Pasteurella multocida. In: Adlam C, Rutter JM, editors. Pasteurella and Pasteurellosis: Academic press, London; 1989. p. 37-73.
Rimler RB. Serology and virulence of haemorrhagic septicaemia Pasteurella multocida isolatd from domestic and feral ruminants. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992: ACIAR Proceedings No. 43; 1993. p. 44-6.
Rimler RB, Rhoades KR. Hyaluronidase and chondroitinase activity of Pasteurella multocida serotype B:2 involved in hemorrhagic septicemia. Veterinary Record. 1994;134(3):67-8.
Rimler RB, Wilson MA. Reexamination of Pasteurella multocida serotypes that caused haemorrhagic septicemia in North America. Veterinary Record. 1994;134(10):256.
Rimler RB. Passive immune cross-protection in mice produced by rabbit antisera against different serotypes of Pasteurella multocida. Journal of Comparative Pathology. 1996;114(4):347-60.
Rimler RB. Restriction endonuclease analysis using Hhal and Hpall to discriminate among group B Pasteurella multocida associated with haemorrhagic septicaemia. Journal of Medical Microbiology. 2000;49(1):81.
Rissman AI, Mau B, Biehl BS, Darling AE, Glasner JD, Perna NT. Reordering contigs of draft genomes using the Mauve Aligner. Bioinformatics. 2009;25(16):2071-3.
Roberts RS. An immunological study of Pasteurella septica. Journal of Comparative Pathology. 1947;57:261-78.
Rocha-e-Silva M, Poli-de-Figueiredo LF. Small volume hypertonic resuscitation of circulatory shock. Clinics. 2005;60:159-72.
Rosenbusch CT, Merchant IA. A study of the hemorrhagic septicemia Pasteurellae. Journal of Bacteriology. 1939;37(1):69-89.
166 | P a g e
Saharee AA, Salim N. The epidemiology of haemorrhagic septicaemia in cattle and buffaloes in Malaysia. FAO; 1991. p. 109-12.
Saini SS, Sharma DR, Gill BS, Kwatra MS, Singh J, Sharma JK, Dihillon SS, Ramneek R. Reemergence of haemorrhagic septicaemia in Punjab. Indian Journal of Animal Sciences. 1991;61:1178-80.
Sawada T, Rhoades KR, Rimler RB. Indirect hemagglutination test that uses glutaraldehyde-fixed sheep erythrocytes sensitized with extract antigens for detection of Pasteurella antibody. Journal of Clinical Microbiology. 1982;15(5):752-6.
Sharples GJ, Lloyd RG. A novel repeated DNA-sequence located in the intergenic regions of bacterial chromosomes. Nucleic Acids Research. 1990;18(22):6503-8.
Sheikh MA, Yaqoob T, Baig MS, Afzal MM, Shakoori AR. The epidemiology of haemorrhagic septicaemia of buffaloes in Pakistan. Buffalo Journal. 1994;10:229-36.
Sheikh MA, Anzam M, Shakoori AR. Observations on haemorragic septicaemia in Pakistan livestock. Journal of Veterinary Medicine. 1996;43:293-304.
Shigidi MT, Mustafa AA. Biochemical and serological studies on Pasteurella multocida isolated from cattle in Sundan. Cornell Veterinarian. 1979;69:77-84.
Singh N. Nasal carriers in bovine pasteurellosis. Indian Journal of Veterinary Science and Animal Husbandry. 1948;18:77-80.
Snipes KP, Hirsh DC, Kasten RW, Hansen LM, Hird DW, Carpenter TE, McCapes RH. Use of an rRNA probe and restriction endonuclease analysis to fingerprint Pasteurella multocida isolated from turkeys and wildlife. Journal of Clinical Microbiology. 1989;27:1847-53.
St Michael F, Harper M, Parnas H, John M, Stupak J, Vinogradov E, Adler B, Boyce JD, Cox AD. Structural and genetic basis for the serological differentiation of Pasteurella multocida Heddleston serotypes 2 and 5. Journal of Bacteriology. 2009;191(22):6950-9.
Statistics Division GoP. Agricultural census 2010 [online]. Available from: http://www.pbs.gov.pk/content/agricultural-census-2010-pakistan-report [Accessed: October 1, 2014]; 2010.
Sterner-Kock A, Lanske B, Uberschar S, Atkinson MJ. Effects of the Pasteurella multocida toxin on osteoblastic cells in vitro. Veterinary Pathology. 1995;32(3):274-9.
Subaaharan S, Blackall LL, Blackall PJ. Development of a multi-locus sequence typing scheme for avian isolates of Pasteurella multocida. Veterinary Microbiology. 2010;141(3–4):354-61.
Subash CV, Mahajan MK, Malik G, Dahiya JP. An epidemiological study on Bovine haemorrhagic septicaemia in Haryana. Indian Journal of Animal Research. 2004;38:14-9.
Sun D, Wang J, Wu R, Wang C, He X, Zheng J, Yang H. Development of a novel LAMP diagnostic method for visible detection of swine Pasteurella multocida. Veterinary Research Communications. 2010;34(8):649-57.
167 | P a g e
Tanner NA, Evans TC. Loop-mediated isothermal amplification for detection of nucleic acids. In: Evans TC, editor. Current Protocols in Molecular Biology: John Wiley & Sons, Inc.; 2014.
Thomas L. Analysis of gene sequences as targets for an HS specific PCR. The international workshop on diagnosis and control of haemorrhagic septicaemia; Bali, Indoensia: ACIAR; 1996.
Topley WWC, Wilson GS. The Principles of Bacteriology and Immunity. 1st ed. London: Edward Arnold and Co.; 1929.
Townsend KM, Dawkins HJS, J.M. P. Analysis of haemorrhagic septicaemia-causing isolates of Pasteurella multocida by ribotyping and field alternation gel electrophoresis (FAGE). Veterinary Microbiology. 1997a;57(4):383-95.
Townsend KM, Dawkins HJS, Papadimitriou JM. REP-PCR analysis of Pasteurella multocida isolates that cause haemorrhagic septicaemia. Research in Veterinary Science. 1997b;63(2):151-5.
Townsend KM, Frost AJ, Lee CW, Papadimitriou JM, Dawkins HJS. Development of PCR assays for species- and type-specific identification of Pasteurella multocida isolates. Journal of Clinical Microbiology. 1998;36(4):1096-100.
Townsend KM, Boyce JD, Chung JY, Frost AJ, Adler B. Genetic organization of Pasteurella multocida cap loci and development of a multiplex capsular PCR typing system. Journal of Clinical Microbiology. 2001;39(3):924-9.
Trevisan V. Sul micrococco della rabbia e sulla possibilita di riconoscere durante il periode d'incubazione, dall'esame del sangue della persona moricata, se ha contratta l'infezione rabbica. Rendiconti Reale Istituto Lombardo (Ser 2). 1887;20:88-105. As cited by Namioka S. Chapter X Pasteurella multocida: Biochemical characteristics and serotypes. In: Bergan T, Norris JR, editors. Methods in Microbiology. Volume 10: Academic Press; 1978. p. 271-92.
Untergasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M, Rozen SG. Primer3—new capabilities and interfaces. Nucleic Acids Research. 2012;40(15):e115.
Urwin R, Maiden MCJ. Multi-locus sequence typing: A tool for global epidemiology. Trends in Microbiology. 2003;11(10):479-87.
Vaid RK, Shanmugasundaram K, Boora A, Bera BC, Shukla BN, Anand T, Singha H, Riyesh T, Virmani N, Barua S, Ahir VB, Koringa PG, Sajnani MR, Bhat VD, Rana N, Singh KP, Malik P, Singh RK, Joshi CG. Draft genome sequence of Pasteurella multocida subsp. multocida B:2 strain VTCCBAA264 isolated from Bubalus bubalis in North India. Genome Announcements. 2014;2(4).
van Drunen Littel-van den Hurk S, Braun RP, Lewis PJ, Karvonen BC, Baca-Estrada ME, Snider M, McCartney D, Watts T, Babiuk LA. Intradermal immunization with a bovine herpesvirus-1 DNA vaccine induces protective immunity in cattle. Journal of General Virology. 1998;79 ( Pt 4):831-9.
van Drunen Littel-van den Hurk S, Loehr BI, Babiuk LA. Immunization of livestock with DNA vaccines: current studies and future prospects. Vaccine. 2001;19(17-19):2474-9.
Versalovic J, Koeuth T, Lupski JR. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Research. 1991;19(24):6823-31.
168 | P a g e
Warner S. Development of a transport enrichment medium for improved isolation of Pasteurella multocida from field samples. The international workshop on diagnosis and control of haemorrhagic septicaemia; Bali, Indonesia: ACIAR; 1996.
Wei MQ, Wang F, Grubb WB. Use of contour-clamped homogenous electric field (CHEF) electrophoresis to type methicillin-resistant Staphylococcal aureus. Medical Microbiology. 1992;36:172-6.
Wickremasuriya UGJS, Kendaragama KMT. A case of haemorrhagic septicaemia in a wild elephant. Sri Lanka Veterinary Journal. 1982;30:34.
Wijewantha EA, Karunaratne TG. Studies on the occurrence of Pasteurella multocida in the nasopharynx of healthy cattle. Cornell Veterinarian. 1968;58:462-5.
Wijewardana TG, De Alwis MCL, Vipulasiri AA. An agar gel diffusion test for the rapid identification of Pasteurella multocida serotype B (Carter). Sri Lanka Veterinary Journal. 1982;30:12-4.
Wijewardana TG, De Alwis MCL, Bastiansz HLG. Cultural, biochemical and pathogenicity studies on Pasteurella multocida isolated from carrier animals and from outbreaks of haemorrhagic septicaemia. Sri Lanka Veterinary Journal. 1986a;34:43-57.
Wijewardana TG, De Alwis MCL, Vipulasiri AA. An investigation into the possible role of the goat as a host in haemorrhagic septicaemia. Sri Lanka Veterinary Journal. 1986b;34:24-32.
Wilson BA, Ho M. Pasteurella multocida: From zoonosis to cellular microbiology. Clinical Microbiology Reviews. 2013;26(3):631-55.
Wilson MA, Rimler RB, Hoffman LJ. Comparison of DNA fingerprints and somatic serotypes of serogroup B and E Pasteurella multocida isolates. Journal of Clinical Microbiology. 1992;30(6):1518-24.
Woolcock JB. The biology of Pasteurella multocida and Pasteurella haemolytica. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 25-34.
Yap HY, Ghazali K, Wan Mohamad Nazarie WF, Mat Isa MN, Zakaria Z, Omar AR. Draft genome sequence of Pasteurella multocida subsp. multocida strain PMTB, isolated from a buffalo. Genome Announcements. 2013;1(5).
Yeo BK, Mokhtar I. Haemorrhagic septicaemia of buffalo in Sabah, Malaysia. In: Patten BE, Spencer TL, Johnson RB, Hoffmann D, Lehane L, editors. Pasteurellosis in production animals, an international workshop held at Bali, Indonesia, 10-13 August 1992. Canberra: ACIAR Proceedings No. 43; 1993. p. 112-5.
Yoshida A, Nagashima S, Ansai T, Tachibana M, Kato H, Watari H, Notomi T, Takehara T. Loop-mediated isothermal amplification method for rapid detection of the periodontopathic bacteria Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola. Journal of Clinical Microbiology. 2005;43(5):2418-24.
169 | P a g e
Yuan Q, Snow DD, Bartelt-Hunt SL. Potential water quality impacts originating from land burial of cattle carcasses. Science of the Total Environment. 2013;456–457:246-53.
Zafar MA, Muhammad G, Iqbal Z, Riaz M. Effects of hypertonic saline solution on clinical parameters, serum electrolytes and plasma volume in the treatment of haemorrhagic septicaemia in buffaloes. Pakistan Veterinary Journal. 2010;30(2):95-9.
Zankari E, Hasman H, Cosentino S, Vestergaard M, Rasmussen S, Lund O, Aarestrup FM, Larsen MV. Identification of acquired antimicrobial resistance genes. Journal of Antimicrobial Chemotherapy. 2012;67(11):2640-4.
Zerbino DR, Birney E. Velvet: Algorithms for de novo short read assembly using de Bruijn graphs. Genome Research. 2008;18(5):821-9.
Zhou Y, Liang Y, Lynch KH, Dennis JJ, Wishart DS. PHAST: A fast PHAge Search Tool. Nucleic Acids Research. 2011.
Zweig MH, Campbell G. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry. 1993;39(4):561-77.
170 | P a g e
Appendix
Table A.1 Unique genes shared by haemorrhagic septicaemia-associated strains of Pasteurella multocida
serovar B:2.
Number Locus_tag Product
1 BUKK_00110 SppA protein
2 BUKK_00115 hypothetical protein
3 BUKK_00870 Patatin
4 BUKK_00885 KAP family P-loop domain protein
5 BUKK_00890 hypothetical protein
6 BUKK_00895 DNA-binding protein
7 BUKK_00900 hydrolase TatD
8 BUKK_01775 hypothetical protein
9 BUKK_01785 hypothetical protein
10 BUKK_02975 glycosyl transferase
11 BUKK_03720 protein PhyA
12 BUKK_03740 hypothetical protein
13 BUKK_03745 hypothetical protein
14 BUKK_03760 heme utilization protein
15 BUKK_05050 uvrD/REP helicase
16 BUKK_05490 UDP-N-acetylglucosamine 2-epimerase
17 BUKK_05495 UDP-N-acetyl-D-mannosaminuronic acid dehydrogenase
18 BUKK_05500 hypothetical protein
19 BUKK_05505 hypothetical protein
20 BUKK_05510 capsular biosynthesis protein
21 BUKK_05515 capsular biosynthesis protein
22 BUKK_05520 EcbG
23 BUKK_05525 hypothetical protein
24 BUKK_05530 spore coat protein
25 BUKK_05535 hypothetical protein
26 BUKK_07275 hypothetical protein
27 BUKK_07280 DNA-binding protein
28 BUKK_07290 transposase
29 BUKK_07295 hypothetical protein
30 BUKK_07300 DNA transposition protein
31 BUKK_07305 hypothetical protein
32 BUKK_07310 host-nuclease inhibitor protein Gam
33 BUKK_07315 hypothetical protein
34 BUKK_07320 bacteriophage protein
35 BUKK_07325 hypothetical protein
36 BUKK_07330 Mu-like prophage FluMu protein gp16
37 BUKK_07335 hypothetical protein
38 BUKK_07340 transcriptional regulator
171 | P a g e
39 BUKK_07345 N-acetylmuramoyl-L-alanine amidase
40 BUKK_07350 hypothetical protein
41 BUKK_07355 hypothetical protein
42 BUKK_07365 hypothetical protein
43 BUKK_07370 hypothetical protein
44 BUKK_07375 hypothetical protein
45 BUKK_07380 Mu-like prophage FluMu protein gp27
46 BUKK_07385 Mu-like prophage FluMu protein gp28
47 BUKK_07390 Mu-like prophage FluMu protein gp29
48 BUKK_07400 Mu-like prophage FluMu G protein 1
49 BUKK_07405 Mu-like prophage FluMu I protein
50 BUKK_07410 head protein
51 BUKK_07415 hypothetical protein
52 BUKK_07420 Mu-like prophage FluMu protein gp36
53 BUKK_07425 Mu-like prophage FluMu protein gp37
54 BUKK_07430 Mu-like prophage FluMu protein gp38
55 BUKK_07435 tail protein
56 BUKK_07440 tail protein
57 BUKK_07445 Mu-like prophage FluMu protein gp41
58 BUKK_07450 Mu-like prophage FluMu protein gp42
59 BUKK_07455 DNA circularization protein
60 BUKK_07460 Tail protein
61 BUKK_07465 Tail protein
62 BUKK_07470 Phage baseplate protein
63 BUKK_07475 Mu-like prophage FluMu protein gp46
64 BUKK_07480 Mu-like prophage FluMu protein gp47
65 BUKK_07485 Mu-like prophage FluMu protein gp48
66 BUKK_07490 hypothetical protein
67 BUKK_07495 hypothetical protein
68 BUKK_07500 uvrD/REP helicase
69 BUKK_07515 hypothetical protein
70 BUKK_07625 transposase
71 BUKK_07630 hypothetical protein
72 BUKK_07915 hypothetical protein
73 BUKK_07920 hypothetical protein
74 BUKK_08195 hypothetical protein
75 BUKK_08215 hypothetical protein
76 BUKK_08220 adenylosuccinate synthetase
77 BUKK_08760 transcriptional regulator
78 BUKK_08765 phosphomethylpyrimidine kinase
79 BUKK_10590 preprotein translocase
80 BUKK_10595 DNA primase
81 BUKK_10600 hypothetical protein
82 BUKK_10610 hypothetical protein
172 | P a g e
83 BUKK_10615 hypothetical protein
84 BUKK_10620 hypothetical protein
85 BUKK_10625 AlpA family transcriptional regulator
86 BUKK_10630 hypothetical protein
87 BUKK_10635 hypothetical protein
88 BUKK_10645 terminase
89 BUKK_10650 tape measure domain protein
90 BUKK_10655 hypothetical protein
91 BUKK_10660 bacteriophage protein
92 BUKK_10665 phage head-tail adapter protein
93 BUKK_10670 HK97 family phage portal protein
94 BUKK_10675 peptidase
95 BUKK_10680 capsid protein
96 BUKK_10685 Minor tail protein U
173 | P a g e
Table A.2 Unique genes of the TX1 strain of Pasteurella multocida serovar B:2.
Number Locus_tag Product
1 PMTX1_01570 serine recombinase
2 PMTX1_01575 conjugal transfer protein
3 PMTX1_01580 sigma factor
4 PMTX1_01585 membrane protein
5 PMTX1_01590 cobalt transporter
6 PMTX1_01595 cobalt ABC transporter
7 PMTX1_01600 ABC transporter
8 PMTX1_01605 ABC transporter ATP-binding protein
9 PMTX1_01610 TetR family transcriptional regulator
10 PMTX1_01615 conjugal transfer protein
11 PMTX1_01620 endonuclease
12 PMTX1_01625 polyketide cyclase
13 PMTX1_01630 HxlR family transcriptional regulator
14 PMTX1_01635 DNA-binding protein
15 PMTX1_01640 helicase
16 PMTX1_01645 ATP-dependent DNA helicase RecG
17 PMTX1_01650 AbrB family transcriptional regulator
18 PMTX1_01655 DNA topoisomerase
19 PMTX1_01660 cell surface protein
20 PMTX1_01665 conjugal transfer protein
21 PMTX1_01670 cell wall hydrolase
22 PMTX1_01675 conjugal transfer protein TraE
23 PMTX1_01680 conjugal transfer protein
24 PMTX1_01685 conjugal transfer protein
25 PMTX1_01690 membrane protein
26 PMTX1_01695 conjugal transfer protein
27 PMTX1_01700 single-stranded DNA-binding protein
28 PMTX1_01705 conjugal transfer protein TraG
29 PMTX1_01710 conjugal transfer protein
30 PMTX1_01715 nucleoside triphosphate hydrolase
31 PMTX1_01720 replication initiation protein
32 PMTX1_01725 methyltransferase
174 | P a g e
Table A.3 Unique genes of the Karachi strain of Pasteurella multocida serovar B:2.
Number Locus_tag Product
1 KARACHI_11310 transposase
2 KARACHI_11315 transposase
3 KARACHI_11320 transposase
4 KARACHI_11325 transposase
5 KARACHI_11330 transposase
6 KARACHI_11340 hypothetical protein
7 KARACHI_11345 transposase
8 KARACHI_11355 hypothetical protein
9 KARACHI_11360 transposase
10 KARACHI_11365 transposase
11 KARACHI_11375 hypothetical protein
Table A.4 The unique gene of the BUKK strain of Pasteurella multocida serovar B:2.
Number Locus_tag Product
1 BUKK_11395 hypothetical protein
Table A.5 The unique gene of the PVAcc strain of Pasteurella multocida serovar B:2.
Number Locus_tag Product
1 PVACC_11385 hypothetical protein
