Vol. 173, No. 15 JOURNAL OF BACTERIOLOGY, Aug. 1991, p. 4814-4819 0021-9193/91/154814-06$02.00/0 Copyright X 1991, American Society for Microbiology Relationships among the rfb Regions of Salmonella Serovars A, B, and D DAN LIU, NARESH K. VERMA, LAJWANT K. ROMANA, AND PETER R. REEVES* Department of Microbiology, Building G08, The University of Sydney, Sydney, New South Wales 2006, Australia Received 18 March 1991/Accepted 28 May 1991 The 0 antigens of Salmonella serogroups A, B, and D differ structurally in their side chain sugar residues. The genes encoding 0-antigen biosynthesis are clustered in the rib operon. The gene rjbj in strain LT2 (serovar typhimurium, group B) and the genes ribS and ribE in strain Ty2 (serovar typhi, group D) account for the known differences in the rfb gene clusters used for determination of group specificity. In this paper, we report the nucleotide sequence of 2.9 kb of DNA from the rjb gene cluster of strain Ty2 and the finding of two open reading frames which have limited similarity with the corresponding open reading frames of strain LT2. These two genes complete the sequence of the rib region of group D strain Ty2 if we use strain LT2 sequence where restriction site data show it to be extremely similar to the strain Ty2 sequence. The restriction map of the rib gene cluster in group A strain IMVS1316 (serovar paratyphi) is identical to that of the cluster in strain Ty2 except for a frameshift mutation in ribE and a triplicated region. The rib gene clusters of these three strains are compared, and the evolutionary origin of these genes is discussed. Lipopolysaccharide is a major component of the outer membrane of gram-negative bacteria that consists of three parts: lipid A, an oligosaccharide core, and oligosaccharide repeat units. In Salmonella enterica, the oligosaccharide repeat units comprise three to six sugar residues and are responsible for 0-antigen specificity (14). In S. enterica groups B, D, and A, the 0-antigen basic repeat units comprise four sugars, of which three form a backbone (mannosyl-rhamnosyl-galactose) common to the three groups. The fourth side chain sugar is a dideoxyhexose: abequose in group B, tyvelose in group D, and paratose in group A. The genes essential for the synthesis and assembly of the 0 antigen in S. enterica are located in the rfb gene cluster (9). In some strains, the basic structure is modified later by the addition of 0-acetyl or glucose residues, but this is not important for our study, as the genes do not map the rib locus. The three dideoxyhexose pathways diverge at the last step, with the only differences being the presence of CDP- abequose synthase in group B but CDP-paratose synthase in groups A and D and an additional enzyme, CDP-paratose-2- epimerase, which converts CDP-paratose to CDP-tyvelose, in group D (18, 20). These enzymes are encoded by the rfbJ gene in strain LT2 (serovar typhimurium group B) and the ribS and ribE genes in strains IMVS1316 (serovar paratyphi A group A) and Ty2 (serovar typhi group D), with rfbE functional only in strain Ty2 (18). We have shown previously (17) that the rib gene clusters of strains Ty2 and IMVS1316 have identical restriction maps, except that a 2.8-kb fragment (from positions 14.4 to 17.2) is triplicated in strain IMVS1316. In the same paper (17), we also compared the restriction maps of the rfb gene clusters of strains Ty2 and LT2 and showed that the two were closely related, with a region of limited similarity flanked by regions in which the DNA was almost identical. On the basis of similarity, the rib gene clusters could be divided into four regions (18). * Corresponding author. Region A, which comprises the ends of the cluster from positions 0 to 11.7 and 16.5 to 22.8, is almost identical in the two strains and includes the rhamnose and mannose path- ways and part of the common dideoxyhexose pathway (Fig. 1). In regions B, C, and D (from positions 11.7 to 16.5), there is only limited similarity (Fig. 1). Region B includes rfbS or ribJ in strains Ty2 and LT2, respectively. The nucleotide sequences of these genes have been determined (18, 20), and comparison of the sequence data suggests that the two genes have a common ancestor but diverged a very long time ago (18). Region C is present only in strain Ty2 and comprises the ribE gene. In region D (from positions 13.6 to 16.5), there is a low level of similarity between strains Ty2 and LT2, detectable only by low-stringency hybridization (18). The triplication in strain IMVS1316 referred to above was also located in this region. As discussed above, ribS and ribE in strain Ty2 (region B and C) and rfbJ in strain LT2 (region B) are the only group-specific determinants required to account for the difference in 0-antigen specificity. We do not expect differ- ences in the transferase, which attaches the dideoxyhexoses to the mannose residue, as the enzyme does not appear to discriminate between the three dideoxyhexoses (18). Thus, the additional differences between these strains in region D were quite unexpected. In this paper, we present the sequence within region D from strain Ty2, compare it with the corresponding region in strain LT2, and present sufficient sequence from strain IMVS1316 to establish the ends of the triplicated region. MATERIALS AND METHODS Media, enzymes, and reagents. Media were as described previously (6). The enzymes KpnI, BamHI, EcoRI, NruI, Bal3l, and DNA T4 ligase were from Boehringer Mannheim. Sequencing was done by using a kit from USB Cleveland, and oligonucleotides were synthesised by using an Applied Biosystems DNA synthesizer. Salmonella 04 antiserum was purchased from Wellcome. 4814 on February 16, 2021 by guest http://jb.asm.org/ Downloaded from
Transcript
Vol. 173, No. 15JOURNAL OF BACTERIOLOGY, Aug. 1991, p. 4814-48190021-9193/91/154814-06$02.00/0Copyright X 1991, American Society for Microbiology
Relationships among the rfb Regions of SalmonellaSerovars A, B, and D
DAN LIU, NARESH K. VERMA, LAJWANT K. ROMANA, AND PETER R. REEVES*Department of Microbiology, Building G08, The University of Sydney, Sydney,
New South Wales 2006, Australia
Received 18 March 1991/Accepted 28 May 1991
The 0 antigens of Salmonella serogroups A, B, and D differ structurally in their side chain sugar residues.The genes encoding 0-antigen biosynthesis are clustered in the rib operon. The gene rjbj in strain LT2 (serovartyphimurium, group B) and the genes ribS and ribE in strain Ty2 (serovar typhi, group D) account for theknown differences in the rfb gene clusters used for determination of group specificity. In this paper, we reportthe nucleotide sequence of 2.9 kb of DNA from the rjb gene cluster of strain Ty2 and the finding of two openreading frames which have limited similarity with the corresponding open reading frames of strain LT2. Thesetwo genes complete the sequence of the rib region of group D strain Ty2 if we use strain LT2 sequence whererestriction site data show it to be extremely similar to the strain Ty2 sequence. The restriction map of the ribgene cluster in group A strain IMVS1316 (serovar paratyphi) is identical to that of the cluster in strain Ty2except for a frameshift mutation in ribE and a triplicated region. The rib gene clusters of these three strainsare compared, and the evolutionary origin of these genes is discussed.
Lipopolysaccharide is a major component of the outermembrane of gram-negative bacteria that consists of threeparts: lipid A, an oligosaccharide core, and oligosacchariderepeat units. In Salmonella enterica, the oligosacchariderepeat units comprise three to six sugar residues and areresponsible for 0-antigen specificity (14). In S. entericagroups B, D, and A, the 0-antigen basic repeat unitscomprise four sugars, of which three form a backbone(mannosyl-rhamnosyl-galactose) common to the threegroups. The fourth side chain sugar is a dideoxyhexose:abequose in group B, tyvelose in group D, and paratose ingroup A.The genes essential for the synthesis and assembly of the
0 antigen in S. enterica are located in the rfb gene cluster(9). In some strains, the basic structure is modified later bythe addition of 0-acetyl or glucose residues, but this is notimportant for our study, as the genes do not map the riblocus. The three dideoxyhexose pathways diverge at the laststep, with the only differences being the presence of CDP-abequose synthase in group B but CDP-paratose synthase ingroups A and D and an additional enzyme, CDP-paratose-2-epimerase, which converts CDP-paratose to CDP-tyvelose,in group D (18, 20). These enzymes are encoded by the rfbJgene in strain LT2 (serovar typhimurium group B) and theribS and ribE genes in strains IMVS1316 (serovar paratyphiA group A) and Ty2 (serovar typhi group D), with rfbEfunctional only in strain Ty2 (18).We have shown previously (17) that the rib gene clusters
of strains Ty2 and IMVS1316 have identical restrictionmaps, except that a 2.8-kb fragment (from positions 14.4 to17.2) is triplicated in strain IMVS1316.
In the same paper (17), we also compared the restrictionmaps of the rfb gene clusters of strains Ty2 and LT2 andshowed that the two were closely related, with a region oflimited similarity flanked by regions in which the DNA wasalmost identical. On the basis of similarity, the rib geneclusters could be divided into four regions (18).
* Corresponding author.
Region A, which comprises the ends of the cluster frompositions 0 to 11.7 and 16.5 to 22.8, is almost identical in thetwo strains and includes the rhamnose and mannose path-ways and part of the common dideoxyhexose pathway (Fig.1).
In regions B, C, and D (from positions 11.7 to 16.5), thereis only limited similarity (Fig. 1). Region B includes rfbS orribJ in strains Ty2 and LT2, respectively. The nucleotidesequences of these genes have been determined (18, 20), andcomparison of the sequence data suggests that the two geneshave a common ancestor but diverged a very long time ago(18). Region C is present only in strain Ty2 and comprisesthe ribE gene. In region D (from positions 13.6 to 16.5), thereis a low level of similarity between strains Ty2 and LT2,detectable only by low-stringency hybridization (18). Thetriplication in strain IMVS1316 referred to above was alsolocated in this region.As discussed above, ribS and ribE in strain Ty2 (region B
and C) and rfbJ in strain LT2 (region B) are the onlygroup-specific determinants required to account for thedifference in 0-antigen specificity. We do not expect differ-ences in the transferase, which attaches the dideoxyhexosesto the mannose residue, as the enzyme does not appear todiscriminate between the three dideoxyhexoses (18). Thus,the additional differences between these strains in region Dwere quite unexpected.
In this paper, we present the sequence within region Dfrom strain Ty2, compare it with the corresponding region instrain LT2, and present sufficient sequence from strainIMVS1316 to establish the ends of the triplicated region.
MATERIALS AND METHODS
Media, enzymes, and reagents. Media were as describedpreviously (6). The enzymes KpnI, BamHI, EcoRI, NruI,Bal3l, and DNA T4 ligase were from Boehringer Mannheim.Sequencing was done by using a kit from USB Cleveland,and oligonucleotides were synthesised by using an AppliedBiosystems DNA synthesizer. Salmonella 04 antiserum waspurchased from Wellcome.
COMPARISON OF rfb REGIONS OF SALMONELLA STRAINS 4815
Plasm ids
G+C content
< pPR6l16
0.44 0.40~~1. 0.32
2kb o~~~~~~~~~~~~~~~~~~~~~~ b h Ma
IT lII[IF.8;x 8 r ^ 0 3 D
0.0 1.3 2.8 B A 6.1 7.1 7.6 F G 10.4 S E 12.8 14.1 154 16.4 M K PIr 1r 1 I I J- 1 I I7
rfbE orfl2.8 orfl4,1 orfI5.Ty2
A B D ALT2
X
rfbJ orfl2.8 orfl4.1 orfl5.4
FIG. 1. rjb region of Salmonella Ty2 and comparison of centralregion with that of strain LT2. The ORFs are named for the startpositions in strain LT2 or for the LT2 gene names, if known (6). Theextent of r)b DNA, with selected restriction sites and their mappositions, is indicated. A plot of G+C content is given above themap; The region of pPR616 which was sequenced is indicated byheavy lines above this plot. Region A is common to strains Ty2 andLT2. Region B contains ribS in strain Ty2 and rJbJ in strain LT2.Region C comprises ribE, which is present in strain Ty2 only.Region D includes orfl2.8 and orfl4.1 in both strains.
Bacterial strains and plasmids. S. enterica Ty2 (Ty2la) (5)and IMVS1316 (17) were sources of the chromosomal DNAfor cloning. Plasmid pPR616 was made by cloning a 6.8-kbKpnI fragment (from positions 13.4 to 20.2) from the Ty2 ribgene cluster into pUC19 (Fig. 1). The 1.8-kb EcoRI fragment(from positions 13.7 to 15.5) from the IMVS1316 rfb genecluster was cloned into pBR325, and the restulting plasmidwas named pPR437 (see Fig. 3). Plasmid pPR848 was madeby cloning the 2.9-kb EcoRI fragment (from positions 15.5 to18.4) of strain IMVS1316 into pUC18 (see Fig. 3).DNA sequencing and DNA methods. The dideoxy chain
termination method of Sanger et al. (15) was used for DNAsequencing. The sequencing of region D of strain Ty2 wasbased on the use of synthetic primers to extend the knownsequence, as described below in Results. The deletion familyused for sequencing one of the junctions of the triplication instrain IMVS1316 was made from plasmid pPR848 by diges-tion with NruI followed by Bal3l nuclease treatment (10).The resulting fragments were cloned into M13mpl8 forsequencing with the universal primer. Other DNA methodswere as described by Maniatis et al. (10).Computer analysis. The programs PepPlot, based on the
method of Kyte and Doolittle (8), and ALOM, based on themethod of Klein et al. (7), in the GCG package were used toidentify potential transmembrane segments, and these trans-membrane segments were further checked by using thealgorithm described by Eisenberg et al. (3). PREDICT (4)was the program used to predict protein secondary struc-ture. All other programs used in this study were described inprevious papers (6, 18, 20).
Nucleotide sequence accession number. The sequence re-ported in this paper has been assigned the GenBank acces-sion number M65054.
RESULTSSequence of region D of strain Ty2. In region D, shown in
Fig. 1, the DNA of strain Ty2 exhibits only a low level of
similarity with the DNA of strain LT2 (17, 18). In order tosequence this region, the KpnI-BamHI fragment (from posi-tions 13.4 to 19.5) of pPR616 (Fig. 1) was cloned into vectorsM13mpl8 and M13mpl9. Sequencing was started at position13.4, using the universal primer. At the other end, we used asynthetic oligonucleotide primer (position 16.5) based on theDNA sequence of strain LT2 (6). Sequencing in both orien-tations was completed by using synthetic primers. Thesequence thus obtained overlapped the known sequence ofstrain Ty2 (18) upstream (position 13.6) and the knownsequence of strain LT2 (6) downstream (position 16.5).There were two open reading frames (ORFs) found in theregion sequenced (Fig. 2). These ORFs correspond to twoORFs in strain LT2 and are named after them as orfl2.8(Ty2) and orfl4.1 (Ty2).
orfl2.8 (Ty2) and orfl4.1 (Ty2). orfl2.8 (Ty2) has 1,296 bp,corresponding to a protein of 432 amino acids, which is 2amino acids longer than orfl2.8 (LT2). orfl4.1 (Ty2) is 999bp long, corresponding to a protein of 333 amino acids,which is the same length as orfl4.1 (LT2).
orfl2.8 and orfl4.1 are likely to have the same function instrains Ty2 and LT2, as discussed below. They are amongthe ORFs not allocated a specific function, so we are not yetable to confirm this possibility directly. However, in the caseof orfl4.1, we were able to demonstrate identity of functionby complementation. In another study (19), an insertionmutation had been made in orfl4.1 of plasmid pPR1009 (6),which contains the whole rjb region of strain LT2. Theinsertion sequence carries a promoter and gives transcrip-tion of downstream genes such that the insertion would nothave any polar effect. The loss of orfl4.1 results in loss of0-antigen expression in the strain carrying this plasmid withthe insertion, but when plasmid pPR616 was also present,function was restored, as detected by agglutination withSalmonella 04 antiserum. Thus, orfl4.1 of strain Ty2 hasbeen shown to complement a mutation in orfl4.1 of strainLT2.
Sequences of junctions of the triplication in strainIMVS1316. Three sequences are required to determine thesites of the triplication (sites I, II, and III in Fig. 3). The1.8-kb EcoRI fragment of strain IMVS1316 (from positions13.7 to 15.5) in pPR437 containing site I (Fig. 3) was clonedinto M13mpl9 with its orientation such that the universalprimer would allow sequencing through position 13.7. Thesequence was further extended by using a synthetic oligo-nucleotide, and the 872-bp sequence obtained was found tobe 100% identical to the sequence in orfl2.8 (Ty2).
Site II was within orflS.4, which is expected to beidentical in strains LT2, Ty2, and IMVS1316, and in thiscase the sequence of site II in strain IMVS1316 was taken tobe the same as the sequence of orflS.4 (LT2) (6).The 2.9-kb EcoRI fragment of strain IMVS1316 (from
positions 15.5 to 18.4) in plasmid pPR848 contains site III(Fig. 3). A deletion family was made from this plasmid, withthe deletions starting from the NruI site, and the resultinginserts with deletions were transferred to M13mpl8 asdescribed above in Materials and Methods. The 718-bpsequence thus obtained included 341 bp identical to thesequence in orflS.4 followed by 377 bp identical to thesequence in orfl2.8 (Ty2). We have sequenced only one ofthe two copies of site III, as we expect them to be identical.
Triplication is common in serovar paratyphi A. Chromo-somal DNAs from several group A strains were digestedwith KpnI and probed by Southern hybridization with a1.8-kb EcoRI fragment (positions 13.7 to 15.5 of strainIMVS1316) to determine if they carried the same triplication.
orfl2.8V R X L R L V R I P R H L I I A A S s M L S K I I I A G V Q L V S V
1 GAGGAA&CTGAGGTTGGTTAGAATTCCAAGACATCTTATTATTGCCGCTTCCTCTTGGCTTTCAAAGATAATA&TTGCCGGTGTTCAGTTAGTAAGTGGT.. ..G.TCAA ... T.A.A ..G..G .T.As.T.. ..i.A.G.. .A.G .. ..T.A. .C..A.. . . ...GAV V Q L K S V G A I
K F L L E I L G r E S Y A V F T L L T G L L V N F S I A D I G I G
101 TTAAATTTCTAGAAATTCTTGGCGAAGAATCATACGCTGTATTTACwCTTAACTGGATTATTGGTCTGGTTTAGCATTGCAGATATTGGGATTGG..TC. .A....A.TTCT .G..A..T....TGAA ..T. .AA.C... G.T.G. T.A. .A... .... GC. .TT...T .C. C..A..S Y IS tl K I S C A V F
S S L Q N Y IS3 L K A D R K S Y D A Y I K A A V H I L F A S L I201 TAGTAGTCTACAAAATTATATATCTGAGTGAGCTGATAGAAaATCATATGATGCATATAAGGCCGCAGTTCATATTCTATTCGCATCCTTAATC
..CAG.A.G........ A. ..A.GC.G.. CA.A.AC .. AGT.T...... .AT.A . T.A . C.AAGC. .TAT.G.TA.T. .TT G C R K N S L L S I A I
I L S S T L F F L S D K L S S L Y L T S F S D E L K N N S G S Y F F301 AT TC TACATTATTCTTCTTATCAGATAAATTATCGTCACTATATCTTACTTCATTTAGCGAT G AACAA GAAGTTATT
T.... T.TTG. .TT.. .T.ATA.T ..T..T.GGGT.A.T ..CG.TAA.....T.T ....T .CAT..G.T.. AC.GG... .AA.CA...TGCTC....F F I A L F Y I F G V I A K s H E V Q D K T R N L
I a S I L F I F I G V G S V V Y K I LF AE L L G N K A N I I N A401 TTATAGCAAGTATATTATATATTCATCGGCGTTGGGAGTGTGGTCTATAAAATATTATGCGGAacTGTTAGGGTGGAAAGCTAATATAATAATGC
...CCT .T. .C.GG.T ..C.GT.CT ..T..AA.C. .AGC.A.T.CT. C.T. .. .T..G.C. C. .T.A. .C..T S C L V S S A I A V L L
L S Y L L G F L D V V A I H Y L N P D S S I T F A L V A L Y A P V
A I L P I I T I S F R Y I Y V L K A K V N F N T Y K L L L S R S S G601 GCAATaCTGCCCATTATATATATATCGTTTCGGTATATATATGTTCTTAAAGCGAAAGTAAACTT CCTATAAATTATTACTATCACGTTCATCAG
.GT..GA.TT.AT.GTGC.... TGTAA.A.A. .c.. A.GC. .TA..T.TT... AC. .CAA CTCAT....T.GC.A.TT..CGT&.A.a......G N I S L C V Y K L Y H V T T K S H I A I R
F L I F S S L S I I V L Q T D Y I V M S Q K L S A A D I I K Y T V701 GGTTTCTTATCGATAATAGTTTTACAACTGATTATATTGTGATGTCAGAAATTATCTGCAGATATTATAAAATATACTGT
..... T.TC..... A.T.TA......... G.G..GC.T.A............ G..C..T......GGC..A..C....T.... G.TC... ....a..F L T L V M I R T P V Q
T I K I F G L N F F I Y T A V L Q A L W P V C A E L R V K M Q W R801 AACGaTGAAAATamGGTTTAATGTCGGTATTACAAGCaTTATGGCCAGTATGTGCTGAATTACGAGTGAAAATGCaGTGGAGA
............T .. G.C.TA.T.G..TA.GA... C...CA ..a. .. AA.V I I Q K
K L H R I I F L N I IG G V FF I G L G T L F I Y V L K D Y I Y S I
L Q S M N Q L K I L N L I V P C QA L I G G V TQW Y F A E R Y G1101 ACTTCAATATGAACCAATTAAAATTCTGGCTCATAGTTCCGTGTCAGGCATTAATTGGTGGTGTGACTCAATGGTATTTTGCAGAGCATTATGGA
.T.G. T......ATa . .A.AT .... A..ACTA................AG..........A.T.TAGTACGCT....Y I L L I I A S S T L
I V G I L Y G L I L S FS L T VFN G L P V Y Y M Y K S K R L A1201 ATAGTTGGTATTTTATACGGACTAATTTTATCGTTCTCGCTAACTGTTTTGGGGATTGCCAGTGTATTATATGTATAAGAG GGCTAGCATAAT
..CAG...AG.GC.GCTT..CT.G....... .T. .TG.TT . ..GC.T .C.AAC....AAT.... .T.A.GG.TA .C..S V L I A L T L I A N K G*
orfl4. 1N K V S F C I P T Y N R V K F I E D L L E S I N N Q S S H S L I V E
1301 A:A GGTATCATTTTGTATCCCAACGTATAATCGAGTA AATTCATTGAAGACCTTCTTGAAAGTATTAATAATCAATCTTCTCACTCCTTAATTGTAG...CTTA..T. T. A.A.AA.C ...ATC....... GT.GT.GA.T. A. GGAAAAATTT .A&T.. GA.A.T.
.. I K Q Y L E K FN D I
V C I S D N A S T D G T E E S I N I N R D R F N FP I L Y Q R H N1401 AAGTATGT&TTTCAGATAATGCATCAACTGATGGAACTGAAGAATCAATAATATCTGGAGAGATAGATTTAATTTTCCAaTTTTATATCAAAGGCATAA
.GA.......A.CG.AT. G.T.T..T.A..G...ATG..TG..G.T. GA.C.AT.A. C. A. GGC.TA...GI N D V N N Y I R NS
E N I G P D R N Y L S A V N M G T G D Y C W I F G S D D I L T K N1501 CGAATAT2TGGGCCAGATAGGAATTATTTATCTGCAGTTAATATGGGGACTGGAGATTATTGTTGGATTTTCGGAAGTGATGATATTCTTACAAAAA&T
..TT. CC. T.C.TG..T....ATCCC.T.C..A..G.A..T..C. GC....G.G...G.CV L F A S S L A N A A D
S L A L M E D K L A A G s D I Y L C D R R E L D I SN T K I S N P H1601 TCCCTTGCATTGATGGAAGAT^AAATGGCTGCCGGAAGTGACATTATTATGTGACAGACGAGAATTAGATATTTCAATGACGAAAATATCTAATCCAC
. GT.A. .GA.AT.AC ..AC.T.TC.C.A.T.TCA.GCA ..T..A...............AA.R..aGACC.GGTG.GATT.AGTTG.G. .TAGA. .C. .T.I L Q T Y D S Q A K T G C D L V E R
R R N L N G G S R L F S F S N E A D L I E Y F S K C N S V G G L F1701 ATCGGCGATGGTTAAATGGTGGCAGTAATTATTTCTTTAGCAATGAGCTGATTTGAAGGTCAAATGTAACTCAGTC GGCGGACTTTT
....TTCT.. .C.GAACA.ATGA.A.C.T.A.TG.... .AT.. .A.TTTA...AG.GA.ATC. .TC.C....G...CTA..TA.T. .T. .TG.A..S R T D D TY V N N L R E I L R L I V
FIG. 2. DNA sequence and deduced amino acid sequence of region D of the rfb gene cluster in strain Ty2. The first 28 bp overlap withthe sequence of strain Ty2 published previously (18), and the last 26 bp are identical to the sequence of strain LT2 (6). The sequences of thecorresponding ORFs of strain LT2 are given underneath, with the nucleotide and amino acid differences presented. The start sites of ORFsare underlined.
COMPARISON OF rjb REGIONS OF SALMONELLA STRAINS 4817
1801S Y L S S I I V K R N K W S D V L F D E S Y I G T A Y A H V Y I LTAGCTACTTAAGCTCTATTATAGTGAAAAGGAATAAATGGTCTGACGTCTTATTTGACGAGTCATACATTGGTACAGCTTATGCGCATGTATATATTTTA
TC....T...TCATT .A ....AAG.ACG....GA..C.A.TGAT.. T.C...C..T.C..C.TTT.C...C.T .T...CA.GL K E R D A I D A S P F M
L R I I N N M N S T L Q Y I S L P L V D C R G D N D T F E S N G K A1901 TTAAGAATAATTAACAATATGAATTCGACCCTGCAGTACATATCACTGCCACTGGTTGATTGCAGAGGAGATAACGATACATTCGAAAGTAATGGGAAGG
A.G. .CG. .T .. .T.CGCCAGGG.GCCTTT .... ..T......AAA. .C.AATA...C.......... . GT. G.AG .A..A....M S V F T P G C L H K I S K K
RR I K I D F I G Y L K L R E D F Y N N N T K I Y I S F G R V L T2001 CTCGTCGAATAA TCATCGGTTATTTAAAACTGAGGGAGGATTTTTATAATAACAATACAAAAATA CTCATTTGGACGGGTTCTTAC
2101
2201
.CA.A.....TTT....... T.T.CA.... T.AGCTA.T........C.G ...A.... T.TCTT ..A.ACGAG. A.AAT.. .T.GCTL A A N S K I S L K R A E N L
K E R P W L Y T S L A M A C Y G D S T D R A E L A S F Y K K L G YAAAAGAGCGTCCTTGGCTTTATACTTCTTTAGCTATGGCATGTTATGGTGATTCTACCGATAGAGCTGAACTTGCATCATTTTATAAGAAGTTGGGATAT
....... .A.A... T.A. AA.. .G.CA. .AG.GAT ..A.A.AGA ..TT.AT.TGA... GCA ..CG.A. .T.G.T N D E K R D S E A N V C
P K I A T N L I F C L K G L A S Y T K K I K L A K M V I K K IFS *CCGAAAATAGCCACCAATTTAATATTTTGTCTTAAAGGGCTGGCTAGCTACACTAAA ATAAAATTAGCAAAAATGGTAATAAAAAAAATATTTTCATAAT . ..ATATG.T...CACTG. .C. .C.AT. .GGGAAA ..A..ATATGCAGTG.....T.T.CCG.GCTT ..G.ATT.T.CT ... CGG .. A. .AAG.N N M I T V L R F G K Y A V N T V L N F T R I K *
regonD ...";"....................region D region A
FIG. 2-Continued.
The fragment hybridized to bands of 6.8, 9.6, 12.4, and 15.2kb, which correspond to the sizes expected if the 2.8-kbsegment is repeated zero, two, three, or four times, respec-tively. The strains of Salmonella serovars nitra and kiel donot show evidence of repeats, but of nine patatyphi Aisolates, four had two repeats, three had three repeats, andone had four repeats. Only one had only one copy. Most ofthe paratyphi A strains tested were Australian isolates, butthree phage types were included. One strain was a vardurazzo strain sent by Le Minor from the Salmonellareference collection. The same probe was used to hybridizeKpnI digests of DNA from six group D strains, and in eachcase, only the 6.8-kb fragment was detected.
Sequence of region D. We have sequenced a 2,879-bpregion (essentially region D of Fig. 1) of the rJb gene clusterof strain Ty2. This, together with the published sequences ofregions B and C (18), gives us the sequence of all of the rJbDNA which differentiates group D strain Ty2 from group Bstrain LT2.Region A is common to both strains, and the segment from
positions 0 to 11.7 differs in 2 of 52 6-bp restriction enzymesites (17), indicating two changes in 312 bp, or 99.3%identity. For the other part of region A (from positions 16.5to 22.8), a 1,000-bp region was sequenced (from positions16.5 to 17.5). This region showed 20 base changes (98%identity), and only a few restriction sites differed in thewhole segment (1, 17). We can thus take the LT2 sequenceto represent the Ty2 sequence for region A and in effect nowhave the whole rfb sequence for group D strain Ty2, whichcomprises 18.7 kb from the start of rfbB to the end of rfbPcompared with 17.6 kb in the rfb sequence in strain LT2.
I I Itriplicated region
Junctions in triplication, showing sequences and reading frames12.8/12.8 TTIT.IIC TII/JIIIMQIC
11 15.4/15.4 6G6QAT/666/= .QJ=
III 15.4/12.8 GA Ta/I=ICJ.MJFlG. 3. Triplicated region of strain lMVS1316, showing the gene
order in strains Ty2 and IMVS1316, some of the restriction sites, theplasmids used in the analysis, and the sequences through thejunctions formed by triplication (sites I, II, and III). The sequencesof sites I and III are the original sequences, and site II is a knownsequence from strain LT2 (6). 15::12 is the chimeric geneorfl5.4::orfl2.8, and rfbE does not function in strain IMVS1316.The regions which were sequenced in the junctions are indicated byheavy lines.
4 'II P1\.91r.%I F f h1 \1-RP4zu - , .nV- - , I r y1 tMM
-3-
vli ~ II v
1 100 200 300 400
FIG. 4. Comparison of hydrophobicity of amino acid sequencesof orf72.8 (Ty2) and or112.8 (LT2), as indicated by the programPepPlot. The curve above the line indicates hydrophobic regions,and the curve below the line indicates hydrophilic regions.
Regions B, C, and D of strain Ty2 occupy 4,825 bp, andregions B and D of strain LT2 occupy 3,626 bp. Note thatalthough differences between rfbS and rfbJ and the presenceor absence of rfbE both affect the positions of all DNA distalto these genes, we have used the position in strain LT2 fornaming ORFs in strain Ty2 or92.8 (Ty2) and orfl4.1 (Ty2)and so on.Genes of region D. or92.8 and or94.1 of strain Ty2 on
average have 54.4% identity with orfl2.8 (LT2) and or94.1(LT2) at the amino acid level, with no insertion or deletiondifferences other than one additional amino acid at each endof or92.8 (Ty2). orfl2.8 in each strain starts with GTG andhas a poor Shine-Dalgamo region, whereas or94.1 in eachstrain starts with ATG and has a good Shine-Dalgarnoregion.The hydrophobicity profiles of the deduced amino acid
sequences of the two or92.8 proteins are very similar.or92.8 (Ty2) has 12 possible transmembrane segments, ashas or92.8 (LT2), and all segments occur in correspondingpositions in the two proteins (Fig. 4). There are 193 nonsyn-onymous base substitutions, a large number of which substi-tute one hydrophobic amino acid for another. For example,there are 19 Leu-Ile substitutions, 17 Ile-Val substitutions,and 12 Leu-Val substitutions. This suggests strongly thator92.8 has the same structure and function in both strains,but we have not yet been able to demonstrate this. We haveargued previously (6) that or92.8 (LT2) does not encode anyof the known rib functions but is the one gene surplus to theknown minimum requirements. We also suggested that itencodes a transmembrane protein which may be involved inoligosaccharide transport or may act as membrane anchorfor the other transferases.
or94.1 (Ty2) clearly resembles orf94.1 (LT2), as, again,the amino acid substitutions do not affect charge distribu-tion, predicted secondary structure, or hydrophobicity pro-files of the protein. In this case, we have been able to showthat or94.1 (Ty2) complements a mutation in or94.1 (LT2),and thus, on the grounds of great similarities of the predom-inant protein properties and of genetic complementation,or94.1 is believed to have the same function in both strains.Noncoding parts of region D. We found no homology
between the two untranslated regions, i.e., 39 bp upstreamof or92.8 (Ty2) (18) and 84 bp upstream of or92.8 (LT2) (6).The termination codon of rflS and the start codon of rfbEoverlap, so there is no sequence between them which couldbe homologous with the untranslated sequence of LT2.At the other end of region D, there are 318 bp between
TABLE 1. Comparison of DNA and amino acid similarities inrfbSlrflbJ, orfl2.8, and orfl4.1 of strains Ty2 and LT2'
Amino acidsSequence Total no. % DNA(base pairs) of bp homology Total % Simi- % Iden-
a Data for rfbS and ribJ are from a previous paper (18).
orf94.1 and or95.4 in strain LT2 (6) and 574 bp in strain Ty2.Again, we can find no homology.G+C content. The overall G+C contents and P1, P2, and
P3 (the corrected G+C contents of codon bases 1, 2, and 3[16]) of orfl2.8 (Ty2) are 0.32, 0.39, 0.32, and 0.24, respec-tively, and those for orfl4.1 (Ty2) are 0.33, 0.37, 0.36, and0.24, respectively. All these values are comparable to thosefor the corresponding ORFs in strain LT2 (6) and for thegenes rfbJ, rfbS, and rfbE. The G+C contents of untrans-lated sequences are 0.29 and 0.24 in strains Ty2 and LT2,respectively. However, the G+C content of the Salmonellachromosome is 0.51 (12), and the values for P1, P2, and P3for most Salmonella genes are quite different (16). The lowoverall G+C content of these regions shows that the DNAmust have been derived from a low-G+C-content organismand inserted into the chromosome by recombination, asargued previously (18). The values for P1, P2, and P3 for thegenes sequenced show that, as for r]bJ and rtbS, the DNAhas not been in Salmonella spp. long enough for significantgenetic drift to have occurred (18, 20). Since all of thesegenes in regions B, C, and D have DNA of similar propertiesand since, with the exception of region C, which is absent instrain LT2, the ORFs of strain Ty2 correspond on a one-to-one basis with those of strain LT2, it seems likely that theDNAs of regions B, C, and D were imported into groups Band D from a common ancestral species.
orfl2.8 and orf94.1 show essentially the same level ofdivergence when Ty2 and LT2 forms are compared (Table1), but rfbS and rfbJ show considerably greater divergence,and indeed the two halves of ribS and rfbJ are not the samein this regard (Table 1). The genes rfbS and rfbJ encoderelated enzymes (18), and we suggest that the greater diver-gence results from adaptation to different functions ratherthan from greater divergence time and that more changeshave been required in an amino-terminal domain than in acarboxy-terminal domain. This result is consistent with ourhypothesis (18), expressed earlier in terms of regions ofDNA, that the group of three genes (ribS/rfbJ, orfl2.8, andorf94.1) have remained together since the two forms di-verged, with either the addition of rfbE to group D or its lossfrom group B as the only macroevolutionary event. Diver-gence must have occurred long ago, as the similarity is farless than between homologous genes of Escherichia coli andSalmonella spp., which have about 20% difference in se-quence (2) mostly due to synonymous changes at base 3, andwhich diverged about 120 to 160 million years ago (13).Recombination events. We have put forward a model for
the recombination events involved in the evolution of the ribregion of strains Ty2 and LT2 (18). Now that we have the fullsequence of the rib region of both strains and know that theregion includes segments of different G+C contents (6), wehave to extend our earlier model (18).
One junction between identical and divergent sequences(position 11.7) coincides with a junction between sequencesof different G+C contents, whereas the other junction (po-sition 16.5) does not, as 0.32-G+C-content DNA extends a
further 1 kb to about position 17.4 (on the map of strain Ty2)towards the end of orfl5.4 (Fig. 1).The junction at position 11.7 must be the site of two
separate recombination events. One possibility is that a
recombination to generate either the LT2 or the Ty2 formthat involved the insertion of 0.32-G+C-content DNA (frompositions 11.7 to 17.4) between 0.44- and 0.40-G+C-contentsegments (6) was followed by a recombination event inwhich the Ty2 segment from positions 11.7 to 16.5 was
replaced by the LT2 segment from positions 11.7 to 15.4.The first event was probably an illegitimate recombination
event (11) not dependent on homology. The second eventpresumably did depend on homology, and at position 11.7,although the intergenic region is short, there is some simi-larity between strains LT2 and Ty2, since, with gaps in-serted, the sequences can be represented as CAAIT[-----------AlCTAICIGGTfIAflGCG (LT2) and ---'l'lGAGGGGGGGGGAflCCCTCIA-TGAhrlC- (Ty2) (homol-ogous bases are underlined). The second recombination atposition 16.5 probably also involved a short sequence ofperhaps only a few bases of limited similarity, which may
have been lost since recombination.We have no means at present of ordering the recombina-
tion events involved in assembling regions of different G+Ccontents common to both strains Ty2 and LT2 or of relatingthese events to those involving regions B, C, and D discussedabove. However, in view of the absence of any indication thatthe third base has adjusted to the Salmonella G+C content
(12), it seems reasonable to assume that assembly of these ribregions occurred in a low-G+C-content species. It is alsopossible that the Ty2 form of the central region of the rJb genecluster was imported first and had more time for the G+Ccontent to drift towards the higher Salmonella level, as
suggested by the G+C content of the noncoding DNA ofstrain Ty2, which is higher than that of strain LT2.
Triplication in strain IMVS1316. The triplicated region inthe rib gene cluster of strain IMVS1316 starts in orfl2.8 (posi-tion 14.4) and ends in orfJS.4 (position 23.0 of strain IMVS1316 and position 17.2 of strain Ty2) (Fig. 3). Each of the three2,885-bp segments contains the last part of orfl2.8, all oforfl4.1, and the first part of orfpS.4 (Fig. 3). The sequences atthe ends of the triplicated segment were determined, andthere is no evidence of any transposition or structural repeatsat the junctions. Presumably the region was first duplicated byan illegitimate recombination event and then triplicated by a
homologous recombination between the two segments. Notethat the triplicated segment starts in region D and extends intoregion A, indicating that the duplication happened afterstrains Ty2 and LT2 diverged. The junction between orflS.4and orfl2.8 generated by the triplication is in frame fortranslation, and the overall effect is to give two additionalcopies of orfl4.1 and two copies of a chimeric gene,
orflS.4: :orfl2.8 (Fig. 3). We have no reason to believe that thechimeric protein is functional or that the triplication has any
effect on function of the cluster, as no gene is lost. Thetriplication seems to be widely present in paratyphi A strainsbut absent in the two other group A strains tested.
In summary, we have completed the sequencing of theTy2 (Salmonella group D) rifb gene cluster by using thesequence of strain LT2 where it appears identical by restric-tion mapping, and we conclude that a central region whichdifferentiates groups B and D diverged long before the
divergence of Escherichia and Salmonella and in morerecent times was transferred as part of a complex assemblyto Salmonella. During this long period, one enzyme haschanged function and another (now present in group D only)has been lost or gained, and these two events account for thedifference in 0-antigen specificity of the two strains.
REFERENCES1. Brahmbatt, H. N., P. J. Wyk, N. B. Quigley, and P. R. Reeves.
1988. A complete physical map of the rfb gene locus encodingbiosynthetic enzymes for the 0 antigen of Salmonella typhimu-rium LT2. J. Bacteriol. 170:98-102.
2. Cossart, P., E. A. Groisman, M. C. Serre, M. J. Casadaban, andB. Gicquel-Sanzey. 1986. crp genes of Shigella flexneri, Salmo-nella typhimurium, and Escherichia coli. J. Bacteriol. 167:639-646.
3. Eisenberg, D., E. Schwarz, M. Komaromy, and R. Wail. 1984.Analysis of membrane and surface protein sequences with thehydrophobic moment plot. J. Mol. Biol. 179:125-142.
4. Eliopoulos, E. E., A. J. Geddes, M. Brett, D. J. C. Pappin, andJ. B. C. Findlay. 1982. A structural model for the chromophorebinding domain of ovine rhodopsin. Int. J. Biol. Macromol. 4:263.
5. Germanier, R., and E. Furer. 1975. Isolation and characteriza-tion of galE mutant Ty2la of Salmonella typhi. A candidatestrain for a live, oral typhoid vaccine. J. Infect. Dis. 131:553-558.
6. Jiang, X. M., B. Neal, F. Santiago, S. J. Lee, L. K. Romana, andP. R. Reeves. 1991. Structure and sequence of the rib (O antigen)gene cluster of Salmonella serovar typhimurium (strain LT2).Mol. Microbiol. 5:695-713.
7. Klein, P., M. Kanehisa, and C. Delisi. 1985. The detection andclassification of membrane spanning proteins. Biochem.Biophys. Acta 815:468-476.
8. Kyte, J., and R. F. Doolittle. 1982. A simple method fordisplaying the hydropathic character of a protein. J. Mol. Biol.157:105-132.
9. Makela, P. H., and B. A. D. Stocker. 1984. Genetics of lipopoly-saccharide, p. 59-137. In E. T. Rietschel (ed.), Handbook ofendotoxin, vol. 1. Elsevier Science Publishers B.V., Amsterdam.
10. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.
11. Marvo, S. L., S. R. King, and S. R. Jaskunas. 1983. Role of shortregions of homology in intermolecular illegitimate recombina-tion events. Proc. Natl. Acad. Sci. USA 80:2452-2456.
12. Normore, W. M. 1973. Guanine-plus-cytosine (GC) compositionof the DNA of bacteria, fungi, algae and protozoa, p. 585-689.In A. I. Laskin and H. A. Lechevalier (ed.), Handbook ofmicrobiology, vol. II. CRC Press, Cleveland, Ohio.
13. Ochman, H., and A. C. WiLson. 1987. Evolution in bacteria:evidence for a universal substitution rate in cellular genomes. J.Mol. Evol. 26:74-86.
14. Rick, P. D. 1987. Lipopolysaccharide biosynthesis, p. 648-662.In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik,M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli andSalmonella typhimurium: cellular and molecular biology, vol. 1.American Society for Microbiology, Washington, D.C.
15. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.
16. Sueoka, N. 1988. Directional mutation pressure and neutralmolecular evolution. Proc. Natl. Acad. Sci. USA 85:2653-2657.
17. Verma, N. K., N. B. Quigley, and P. R. Reeves. 1988. 0-antigenvariation in Salmonella spp.: rdb gene clusters of three strains.J. Bacteriol. 170:103-107.
18. Verma, N. K., and P. R. Reeves. 1989. Identification andsequence of ribS and ribE, which determine antigenic specificityof group A and group D salmonellae. J. Bacteriol. 171:5694-5701.
19. Wang, L., and P. R. Reeves. Unpublished data.20. Wyk, P. J., and P. R. Reeves. 1969. Identification and sequence
of the gene for abequose synthase, which confers antigenicspecificity on group B salmonellae: homology with galactoseepimerase. J. Bacteriol. 171:5687-5693.
![Zydus Wellness · 2019-07-09 · ofCentral Depository Services [India] Limited [CDSL] from a place other than venue ofAGM [remote e-voting]. All members are informed that: i. The](https://static.documents.pub/doc/80x56/5e9dbb2f32b2fb64d1525b6a/zydus-wellness-2019-07-09-ofcentral-depository-services-india-limited-cdsl.jpg)
