1
Rapid detection of Panton Valentine leukocidin in Rapid detection of Panton Valentine leukocidin in Staphylococcus aureus Staphylococcus aureus cultures by monoclonal cultures by monoclonal antibodies using a lateral flow assay antibodies using a lateral flow assay Stefan Monecke 1/2 , Joseph Buechler 3 , John Rejman 4 , Ralf Ehricht 1 1. Alere Technologies GmbH, Jena, Germany 2. Institute for Medical Microbiology and Hygiene, Technical University of Dresden, Dresden, Germany 3. Alere San Diego, Inc., San Diego, CA, USA 4. Alere Scarborough Inc., Scarborough, ME, USA Objectives: Panton Valentine leukocidin (PVL) is a phage born virulence factor of Staphylococcus aureus, which is associated with chronic/recurrent skin and soft tissue infections (SSTI) and necrotising pneumonia. Because of its clinical relevance, the detection of S. aureus which carry PVL genes warrants more aggressive therapy and infection control measures than PVL-negative strains (see www.hpa.org.uk ). However, PVL detection is currently essentially limited to reference centres and specialised laboratories as it is performed by molecular methods. Methods: Recombinant Panton Valentine leukocidin (F-component) was used to generate a set of monoclonal antibodies by phage display. These antibodies were purified after over- expression in E. coli, characterised initially by ELISA and spotted in different dilutions in microtiterstrip-mounted protein microarrays. Results from these microarray assays assisted in the identification of a suitable pair of antibodies which then were used to establish a lateral flow assay. Two different versions of this assay were used to detect PVL in overnight cultures of S. aureus from different growth media within 10 minutes. Isolates were genotyped by microarray hybridisation in parallel for confirmation and for assignment to clonal complexes. Results: The detection limit for the lateral flow test was determined to be around 1 ng/ml. Overnight cultures from Columbia blood agar, Mueller Hinton agar and a commercial MRSA selective growth medium as well as liquid cultures (in a broth described by Kato&Noda) after as few as 3 hrs incubation proved suitable for PVL detection. For evaluation, 450 clinical isolates from patients with skin and soft tissue infections from America, Europe, Australia, Africa and the Middle East were tested. 258 isolates belonging to 37 distinct strains were PVL-positive. 192 isolates from 47 strains were PVL-negative. This included methicillin-susceptible as well as -resistant S. aureus. The sensitivity of the assay in these initial trials was 99.7%, the specificity was 95.3%. The positive predictive value was found to be 96.3%, the negative predictive value 99.7%. Conclusion: This test allows the rapid detection of PVL under conditions of a routine bacteriological laboratory. As it utilises cultures from standard media and as it does not require sophisticated equipment, it can easily be integrated into a laboratory´s workflow. This might contribute to timely therapeutic interventions in cases of PVL-associated infections. Contact and enquiries: [email protected] ; [email protected] Download at: http://alere-technologies.com/fileadmin/Media/Paper/Poster/ECCMID_2012.pdf Introduction: Panton Valentine leukocidin (PVL) is a phage born virulence factor of Staphylococcus aureus. It comprises two units (S and F components) that are encoded by two separate, although co-localised and co-expressed genes. Polymeres of these molecules form pores in human leukocyte membranes leading to cell death. PVL is associated with chronic/recurrent skin and soft tissue infections (SSTI), especially in young and previously healthy adults, and necrotising pneumonia. Because of its clinical relevance, the detection of S. aureus which carry PVL genes warrants aggressive therapy and infection control measures (see http://www.hpa.org.uk/webc/HPAwebFile/HPAweb_C/1218699411960 ). However, PVL detection is currently essentially limited to reference centres and specialised laboratories as it is performed by molecular methods. In order to facilitate a rapid, non-molecular detection in clinical laboratories, monoclonal antibodies were raised and a lateral flow test was developed. Methods: Over-expressed PVL, F-component, was used to generate monoclonal antibodies via phage display. Following immunisation of mice, mRNA from B-cells was isolated and amplified. Resulting cDNA, specific for the antigen-binding parts of antibodies, was ligated into bacteriophages and then transformed into E. coli. Resulting antibodies were purified, characterised initially by ELISA and spotted in different dilutions in microtiterstrip-mounted protein microarrays. This allowed to rapidly determine the optimal combination of capture and detection antibodies (see below). These antibodies were used to design a lateral flow test, i.e., an immunochromatographic test in which gold-labelled detection antibodies are mixed with sample material (S. aureus cultures) flow by capillary action towards a zone of immobilised detection antibody. In positive cases, the formation of a visible line can be observed. Two differently manufactured test formats (dipstick and Binax cards) were used in parallel for optimisation of handling and protocols. This test was applied to isolates of S. aureus from skin and soft tissue infections (see below) that in parallel were genotyped by microarray hybridisation in order to determine strain and clonal complex affiliation as well as their PVL-status. Results: For the selection of the optimal combination of capture and labelling antibodies, four different concentrations of each antibody was spotted onto protein microarrays. These arrays were tested with recombinant PVL F-component, native PVL (in two different concentrations, from strain ATCC25923) or “bovine leukocidin” lukM/lukF-P83 from a veterinary CC705 isolate as well as with all biotin-labelled preparations of all antibodies. Based on the results as simplified shown in Figure 2, a combination of Antibody 5 and Antibody 10 was selected for establishing a lateral flow assay that can detect PVL (F- component) as well as the gene product of lukF-P83. Figure 2: Matrix for the array-based determination of the optimal combination of antibodies. Summarised results are colour-coded from green (negative) to purple (strongly positive). In a first series of experiments, known strains cultured on different growth media were tested. Detectable PVL production was noted in a broth as described by Kato & Noda or by Schaedler, in Brain Heart infusion as well as in colony material harvested from Plain Agar, Mueller Hinton agar with and without blood, MRSA ID agar (BioMerieux), Columbia Blood, C.A.P. and “chocolate“ agar. False negative results were occasionally observed with glucose broth as well as false positives with clonal complex CC8 strains from Kato & Noda broth or blood agar. These lateral flow tests were used to screen a total of 450 clinical isolates obtained from diagnostic specimens from SSTI. These isolates originated from Australia, Trinidad & Tobago, the United States, the UK, Germany, Sweden, Spain, Norway, Japan, Uganda and Saudi-Arabia. 258 isolates proved to be positive. They belonged to 37 different strains from 20 clonal complexes (Table 2). 192 PVL-negative isolates have been tested belonging to belonged to 47 different strains from 29 clonal complexes. The proportion of PVL-positive isolates among all SSTI isolates tested ranged between 10.5% (Swedish samples) and 81.4% (Australian samples). Positive cases Negative cases Positive test results True positive 387 False positive 15 PPV TP / (TP + FP) 96.27 Negative test results False Negative 1 True Negative 306 NPV TN / (FN + TN) 99.67 Sensitivity TP / (TP + FN) 99.74 Specifity TN / (FP + TN) 95.33 Table 1: Sensitivity, specifity, positive and negative prediction values. Most isolates were tested twice using two differently manufactured tests (dipstick and Binax cards), hence the number of experiments is higher than the number of isolates. False positives proved to be negative when using cultures from Mueller Hinton rather than from Columbia Blood agar. Clonal Complex Strains CC1 CC1-MSSA, CC1-MSSA-SCCfus, CC1-MRSA-IV=USA400, ST573/772-MSSA, ST772-MRSA-V=Bengal Bay Clone CC5 CC5-MSSA, CC5-MRSA-IV=Paediatric clone CC8 CC8-MSSA, ST8-MRSA-IV [PVL+/ACME+]=USA300, ST8-MRSA-IV [PVL+/ACME-], CC8-MRSA-IV=WA MRSA-62, ST72-MSSA CC15 CC15-MSSA CC22 CC22-MSSA, CC22-MSSA-SCCfus, CC22-MRSA-IV CC25 CC25-MSSA CC30 CC30-MSSA, CC30-MRSA-IV=Southwest Pacific Clone CC45 CC45-MSSA CC49 ST49-MSSA CC59 CC59-MSSA, ST59/952-MRSA-V(T)=Taiwan Clone CC80 CC80-MSSA, CC80-MRSA-IV=European caMRSA Clone CC88 CC88-MRSA-IV CC93 ST93-MSSA, ST93-MRSA-IV=Queensland Clone CC98/154 CC96/154-MSSA CC121 CC121-MSSA CC152 CC152-MSSA, CC152-MRSA-V CC188 CC188-MSSA CC398 ST291/813-MSSA 2 unident. CCs 2 unknown strains (MLST pending) lukF-P83-positive strains from veterinary sources, not part of the SSTI study) CC133 CC133-MSSA CC479 CC479-MSSA CC705 CC705-MSSA < Table 2: Strains and clonal complexes from which PVL-positive isolates were identified. Discussion: This test allows the rapid detection of PVL under conditions of a routine bacteriological laboratory that is not able to perform molecular assays. As it utilises pure overnight cultures from standard media (including a chromogenic agar for MRSA screening), it can easily be integrated into such a laboratory´s workflow. Thus it might contribute to timely therapeutic interventions in cases of PVL-associated infections, and it also might help to select isolates that are to be submitted for further typing in reference centres. Acknowledgments: We thank P. Akpaka (St. Augustine, Trinidad & Tobago), G. Coombs (Perth, Australia), L. Skakni (Riyadh, Saudi Arabia), B. Söderquist (Örebro, Sweden), S. Molinos Abós (Barcelona, Spain), R. Burris (Kampala, Uganda), D. Bandt (Frankfurt/Oder, Germany) and the staff at the Institute for Medical Microbiology and Hygiene for collecting and donating isolates as well as A. Ruppelt, E. Müller and B. Stieber (Dresden/Jena, Germany) for excellent technical assistance. Figure 1: Test procedure for the Binax card format. For the dipstick format, the cultures are harvested with an inoculation loop and stirred in a tube containing a buffer with the labelled antibodies. Then, the dipstick is placed into the tube. The result is read after 10 min, too.
