phylogenetic relationships of salmonella based on dna sequence comparison of atpd encoding the β...

Phylogenetic relationships of Salmonella based on DNA sequencecomparison of atpD encoding the L subunit of ATP synthase

Henrik Christensen *, John Elmerdahl OlsenDepartment of Veterinary Microbiology, 1-20, Buëlowsvej 13, Royal Veterinary and Agricultural University, 1870 Frederiksberg C, Denmark

Received 14 December 1997; revised 30 January 1998; accepted 2 February 1998

Abstract

DNA sequences covering 57% of atpD encoding the L subunit of ATP synthase were determined for 16 strains of Salmonellaenterica, two strains of S. bongori, and one strain each of Citrobacter freundii and Yersinia enterocolitica, and comparison wasmade with the published Escherichia coli and Enterobacter aerogenes sequences. The phylogenetic tree based on maximum-likelihood analysis showed separation of the subspecies of S. enterica except for two serotypes of subspecies II which wereunsupported by a common node. The two serotypes of S. bongori were separated from S. enterica and related to the serotypesof subspecies II. A tight relationship was found between S. enterica subspecies IIIa consisting of monophasic serotypes andsubspecies IIIb consisting of diphasic serotypes. This is in conflict with results obtained for most other housekeeping genes andthe 23S rRNA gene separating mono- from diphasic subspecies. z 1998 Federation of European Microbiological Societies.Published by Elsevier Science B.V.

Keywords: Salmonella enterica ; Salmonella bongori ; atpD ; ATP synthase; Phylogeny

1. Introduction

The genus Salmonella includes seven DNA-DNAsimilarity subgroups [1] six of which are categorizedas subspecies of S. enterica while the seventh group,S. bongori, has been given species rank [1,2]. Thiscategorization has been con¢rmed by sequencing ofhousekeeping and invasion-associated protein genes.Analysis of these genes further resulted in an evolu-tionary model where serotypes predominantly dipha-sic in £agellar expression (S. enterica subsp. I, II,

IIIb, VI) were separated from the monophasic sero-types (S. enterica subsp. IIIa, IV, S. bongori) [3].Analysis of 23S rRNA gene sequences con¢rmedthis model [4].

The DNA sequence of the L subunit of the ATPsynthase gene is suitable for phylogenetic analysisbecause of its ubiquitous distribution, functionalconstancy and sequence conservation. Phylogeneticrelations between Bacteria based on atpD gene se-quence comparison resembled phylogenetic relationsbased on 16S rRNA sequences [5,6]. Comparison ofphylogenetic models obtained by atpD and rRNAhas not been reported for closely related taxa at sub-species level.

The function of ATP synthase is to couple theelectrochemical potential di¡erence for H� across

0378-1097 / 98 / $19.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V.PII S 0 3 7 8 - 1 0 9 7 ( 9 8 ) 0 0 0 5 4 - 8

FEMSLE 8064 19-3-98

* Corresponding author. Tel. : +45 35282783;Fax: +45 35282757; E-mail: [email protected]

FEMS Microbiology Letters 161 (1998) 89^96

the inner membrane to synthesis of ATP from ADPand Pi. The enzyme is divided into the two subunitsFo and F1. Fo traverses the membrane facilitatingproton transport and F1 is located on the cytoplas-mic surface of the membrane responsible for ATPaseactivity. F1 is composed of the ¢ve polypeptides K, L,Q, N and O. The gene atpD (1383 bases long) is re-sponsible for the L subunit functioning in nucleotidebinding and catalysis [7].

The present study aims to reconstruct the phylog-eny of Salmonella based on DNA sequence compar-ison of atpD.

2. Materials and methods

2.1. Bacterial strains and media

Serotypes were selected to represent the S. entericasubspecies with six, two, two, three, two, and onestrain for the subspecies I, II, IIIa, IIIb, IV andVI, respectively. Two serotypes of S. bongori, one

Citrobacter freundii, and one Yersinia enterocoliticastrain were included (Table 1). Isolates were culturedovernight at 37³C in Luria-Bertani broth (10 g bactotryptone, 5 g bacto yeast extract, 10 g NaCl, 1 ldeionized water, pH 7.0).

2.2. PCR ampli¢cation

DNA was extracted as described by Vogel et al.[8]. 1^4 Wl of DNA solution containing 0.3 Wg DNAwas added to 95 Wl of a PCR mixture consisting of10U mix (USB, Cleveland, OH, USA), 2.5 mMMgCl2, 60 ng of each oligonucleotide (see below),0.20 nmol of each of the four nucleotides, and 2.5U Taq DNA polymerase (USB).

Ampli¢cation was carried out in a thermocycler(GeneAmp PCR System 2400, Perkin Elmer,Norwalk, CT, USA) and initiated by 1.5 min dena-turation at 94³C followed by 35 cycles consistingof 30 s denaturation at 94³C, 1 min annealing at55³C and 2 min extension at 72³C. The solutionwas left at 72³C for 10 min after the last cycle.

FEMSLE 8064 19-3-98

Table 1Bacteria investigated and sequences obtained for atpD

Species Subspecies Serotype Strain (accession numbera)

Salmonella enterica enterica I Dublin K771 (JEO11), 241/82 (JEO71)b

Dublin K228 (JEO2)Typhimurium LT2 (JEO14)Infantis Div36-86 (JEO334)Tennessee Div95-86 (JEO338)

salamae II 1,9,12:1,w:e,n,x NSC72 (JEO349)40:d:3 S114655 (JEO1845)

arizonae IIIa 51:z4,z23:3 S83769 (JEO1689)(Arizonae) u24 (JEO792)

diarizonae IIIb 61:i :z JEO30748:r:z DS210/89 (JEO823)60:r:z S109671 (JEO1863)

houtenae IV 43:z4,z23:3 S84366 (JEO1674)16:z4,z32:3 S84098 (JEO1672)

indica VI 1,6,14,25:a:e,n,x BR2047 (JEO1935)S. bongori 66:z41:3 BR1859 (JEO1946)

66:z65:3 JEO4162Citrobacter freundii JEO503Yersinia enterocolitica NCTC10460 (JEO2341)

aPositions 99^891 of the E. coli sequence with accession number J01594.bThe strain designation is given for the culture collection from which the strain was obtained. Other strain designations are given inparentheses. BR, Div and S: Laboratory for Enteric Pathogens, Central Public Health Laboratory, Colindale, London, UK. DS: theFederation of Danish Pig Producers and Slaugtherhouses, Copenhagen, Denmark. JEO: Department of Veterinary Microbiology, RoyalVeterinary and Agricultural University, Copenhagen, Denmark. K: the State Serum Institute, Copenhagen, Denmark. NCTC: the NationalCollection of Type Cultures, Central Public Health Laboratory, Colindale, London, UK.

H. Christensen, J.E. Olsen / FEMS Microbiology Letters 161 (1998) 89^9690

The e¤ciency of PCR ampli¢cation was veri¢ed byagarose gel electrophoresis using ethidium bromidestaining.

In each PCR reaction, one of the oligonucleotideprimers was labelled with biotin (Biodite, biotin ami-dite, Pharmacia Biotech, Uppsala). This allowedDNA strand separation and puri¢cation of ampli¢edDNA using streptavidin-coated magnetic beads (Dy-nabeads M-280, Dynal, Oslo) to trap the biotin-la-belled PCR fragments. DNA strand separation wasperformed in NaOH and both strands were used astemplates for DNA sequencing.

2.3. Oligonucleotides

Oligonucleotides for PCR ampli¢cation and forsequencing of atpD were synthesized identical andcomplementary to the Escherichia coli sequencebased on the GenBank sequence J01594. Oligonu-cleotides for forward ampli¢cation were 5737 5P-TAGTTGACGTCGAATTCCCTCAGG, 6247 5P-GTACTCGTGAGGGTAACGACTTC, and for re-verse ampli¢cation 6025 5P-GCCCAACGCTC-TTCTTCACC, 6390 5P-ACGTCACGACCTTCGT-CACGGAA, and 6625 5P-GGAGACGGGTCAGT-CAAGTCATC.

2.4. DNA sequencing

Sequencing was performed by the dideoxy chaintermination method [9] using Autoread Sequencingkit (Pharmacia Biotech) set up with 10 ng £uores-cein-labelled primer using 3 U of T7 DNA polymer-ase (Pharmacia Biotech) and 1^2 Wg of DNA tem-plate. Sequence reactions were analyzed on the ALFAutomatic DNA Sequencer (Pharmacia Biotech). Se-quencing was performed on both strands.

2.5. Sequence and phylogenetic analysis

Bases 99^891 (E. coli position) were aligned. Thepublished E. coli sequence (GenBank J01595) andEnterobacter aerogenes sequences were included forcomparison [5]. Analysis was performed by maxi-mum likelihood including bootstrap using PHYLIPand fastDNAml [10,11]. The 19 sequences of partialatpD are deposited in GenBank under the accessionnumbers listed in Table 1.

3. Results

3.1. Amino acid and nucleic acid sequence variation

Identical amino acid sequences were translatedfrom the 793 nucleic acid positions in 11 serotypesof S. enterica, and the two serotypes of S. bongori(Fig. 1). For ¢ve serotypes of S. enterica subspeciesI, IIIa, and IIIb one or two amino acid substitutionswere observed at codons 183 and 195. Comparisonwith related species of Enterobacteriaceae showedone amino acid substitution between Salmonellaand E. coli. Between Salmonella and E. aerogenes,C. freundii, and Y. enterocolitica, 10^23 amino acidsubstitutions were determined (Fig. 1).

Identical nucleic acid sequences were determinedin two out of three Salmonella serotype Dublinstrains. In the deviating strain K228, T was substi-tuted for G at the third position of codon 181. Thevariation in nucleic acid sequence was between 0.1and 1.5% within individual S. enterica subspecies and0.9^3.7% between subspecies. The variation was 3.0^4.5% between S. enterica and S. bongori, and 4.4^17% between Salmonella and the other members ofEnterobacteriaceae investigated (Table 2).

3.2. Phylogenetic analysis

The phylogenetic analysis identi¢ed Salmonella asa monophyletic unit. The subspecies of S. entericawere separated. The exception was the two serotypesof subspecies II without a common node. The twoserotypes of S. bongori formed a monophyletic unitbut the origin of S. bongori was within the branchingof S. enterica close to S. enterica subsp. II. There-fore, in a cladistic sense S. enterica formed a para-phyletic group (Fig. 2). The two subspecies IIIa andIIIb of S. enterica were joined by a common nodeand formed a larger group with subsp. II, IV and S.bongori.

Salmonella was related most closely to E. coli andmore distantly to E. aerogenes, C. freundii and Y.enterocolitica.

4. Discussion

This phylogenetic analysis of the atpD gene se-

FEMSLE 8064 19-3-98

H. Christensen, J.E. Olsen / FEMS Microbiology Letters 161 (1998) 89^96 91

FEMSLE 8064 19-3-98


quences enabled resolution of S. enterica and S. bon-gori at species and subspecies levels. The two sero-types of S. enterica subspecies II branched o¡ sepa-rately. The origin of S. bongori was within S. entericaclose to subsp. II and S. enterica was therefore para-phyletic. S. enterica was also found to be paraphy-letic by 23S rRNA analysis but not by 16S rRNAanalysis [4]. The analysis con¢rmed the division be-tween S. enterica subspecies and separation of S.enterica from S. bongori de¢ned by DNA hybridiza-tion [1] as well as the phylogenetic analysis of house-keeping genes, invasion-associated protein genes [3],and rRNA [4].

S. enterica subspecies IIIa and IIIb were closelyrelated which was unexpected because analysis ofmost housekeeping genes and of 23S rRNA genesequences showed separate phylogenetic positionsof subspecies IIIa and IIIb [3,4]. A close relationshipbetween IIIa and IIIb has earlier been observed withsome biochemical characteristics. Subspecies III wasoriginally de¢ned as a L-galactosidase-positive bio-type [12]. Later the subspecies IIIa and IIIb wereformed and both were reported to share L-galactosi-dase activity with S. bongori [1]. In addition, S. en-terica subspecies IIIa and IIIb shared a positive ge-latinase reaction with subspecies II and IV [1]. Aclose relationship between IIIa and IIIb has beenobserved for some housekeeping gene sequences.Subspecies IIIa and IIIb were closely related toeach other in phylogenetic trees based on multilocusenzyme electrophoresis [3], analysis of malate dehy-drogenase (mdh) [13] as well as proline permease(putP, put) gene sequences [14]. Sequence compari-son of the icd gene showed S. enterica subsp. IIIb tobe distantly related to the other diphasic subsp. I, II,and VI [15]. The close relation between subspeciesIIIa and IIIb observed with atpD could be explainedby a later separation of subsp. IIIb from the mono-phasic group than the other diphasic subspecies.

The separate branching of two serotypes of S. en-terica subsp. II (see Fig. 2) was also observed whenmaximum-likelihood analysis was run on silent nu-

cleic acid positions only and by neighbor-joininganalysis (data not shown). The nucleic acid variationbetween these two serotypes of subsp. II was 1.5%(see Fig. 1) which was the highest within-subspeciesvariation observed between pairs of S. enterica sero-types. The separation of the two serotypes of S. en-terica subspecies II on the phylogenetic tree cantherefore be explained by exceptionally high geno-typic divergence.

The relations of Salmonella to E. coli, E. aero-genes, C. freundii, and Y. enterocolitica were weaklyde¢ned by the present analysis because only one se-quence was included for each of the other membersof the Enterobacteriaceae. The close relationship be-tween Salmonella and E. coli is in accordance withphylogenetic analysis of rRNA [4] and the genotypicrelatedness between these two taxa in general [16].

This phylogenetic analysis of Salmonella wasbased on 57% of the atpD gene and included thecatalytic site identi¢ed at amino acids 142^246 ofE. coli [17]. This is thought to consist of a highlyconserved sequence in order to preserve enzymefunction. In correspondence, amino acid changeswere only determined at six positions of this regionwhen Salmonella and related members of the Enter-obacteriaceae were compared. The higher level ofamino acid variation registered in the region up-stream of the catalytic region may account for the3.8% variation in amino acid sequence determinedbetween E. coli and E. aerogenes in the present studycompared to the whole gene with only 2% variation[5]. By including both variable and more conservedstretches the phylogenetic analysis based on the par-tial atpD sequence is assumed to have been represen-tative for the whole gene.

In conclusion, the analysis of atpD enabled us toresolve phylogenetic lineages of Salmonella at speciesand subspecies levels. Such separation has also beenachieved by analysis of other housekeeping genesand 23S rRNA. The most remarkable evolutionaryrelationship was the close connection between S. en-terica subspecies IIIa and IIIb.

FEMSLE 8064 19-3-98

6Fig. 1. Amino acid substitutions of the atpD DNA sequence between Salmonella, Escherichia coli, Citrobacter freundii, Enterobacter aero-genes and Yersinia enterocolitica. Nucleic and amino acids are numbered from the start of atpD and dots indicate identity to the sequenceof E. coli obtained from GenBank with accession number J01594. Identical amino acid sequences were deduced for Salmonella except forthe ¢ve serotypes of S. enterica subsp. I, IIIa, and IIIb listed. The sequence of E. aeorogenes was from Amann et al. [5].


FEMSLE 8064 19-3-98

Tab

le2

Var

iabi

lity

(%)

ofth

eat

pDse

quen

ce(E

.co

lipo

siti

ons

99^8

91)

betw

een

stra

ins

ofS

.en

teri

ca,

S.

bong

ori,

E.

coli,

E.

aero

gene

s,C

.fr

eund

iian

dY

.en

tero

colit

ica

S.

ente

rica

Spec

ies/

subs

peci

es/s

erot

ype/

stra

inI

Dub

.D

ub.

Typ

.In

f.T

enn.

IIII

IaII

IbIV

VI

S.

bong

ori

E.

coli

E.

aero

gene

sC

.fr

eund

iiY

.en

tero

-co

litic

a

S.

ente

rica

ID

ublin

K22

80.

00.

130.

760.

500.

633.

01.

63.

23.

03.

33.

43.

22.

82.

51.

03.

23.

74.

57.

47.

216

Dub

linK

771,

241/

820.

00.

880.

630.

763.

21.

83.

33.

23.

43.

53.

32.

92.

71.

13.

33.

84.

77.

67.

316

Typ

him

uriu

mL

T2

0.0

0.76

1.4

3.4

2.1

3.4

3.2

3.3

3.4

3.3

2.8

2.5

1.5

3.8

4.3

4.9

7.7

7.3

16In

fant

isD

iv36

-86

0.0

0.88

3.4

2.0

3.0

2.9

3.2

3.3

3.0

2.8

2.5

1.3

3.3

3.8

4.4

7.7

6.9

16T

enne

ssee

Div

95-8

60.

03.

31.

93.

33.

23.

43.

53.

33.

02.

50.

883.

33.

84.

97.

77.

316

II1,

9,12

:l,w

:e,n

,xN

SC72

0.0

1.5

3.4

3.3

3.5

3.7

3.4

2.7

2.7

2.9

2.7

3.0

5.6

7.8

7.8

1640

:d:d

S114

655

0.0

2.1

2.0

2.3

2.4

2.1

1.8

1.5

1.5

3.4

3.9

5.2

7.4

7.6

16II

Ia51

:z4,

z23

:3S8

3769

0.0

0.38

1.1

1.3

1.0

2.9

2.7

2.9

3.8

4.0

6.2

8.3

8.3

17(A

rizo

nae)

u24

0.0

0.88

1.0

0.88

2.8

2.5

2.8

4.0

4.3

6.3

8.6

8.5

17II

Ib61

:i:z

JEO

307

0.0

0.13

0.63

2.9

2.7

3.0

4.2

4.4

6.2

8.7

8.5

1748

:r:z

DS2

10/8

90.

00.

763.

02.

83.

24.

34.

56.

38.

88.

617

60:r

:zS1

0967

10.

02.

82.

52.

94.

24.

46.

18.

68.

316

IV43

:z4,

z23

:3S8

4366

0.0

1.0

2.7

3.8

4.3

5.7

8.3

8.5

1716

:z4,

z32

:3S8

4098

0.0

2.1

3.8

4.3

5.7

8.1

8.2

16V

I1,

6,14

,25

:a:e

,n,x

BR

2047

0.0

3.0

3.5

7.8

7.4

7.4

15S

.bo

ngor

i66

:z41

:3B

R18

590.

00.

505.

67.

77.

916

66:z

65:3

JEO

4162

0.0

5.8

8.1

8.3

16E

.co

li0.

07.

16.

816

E.

aero

gene

s0.

06.

816

C.

freu

ndii

JEO

503

0.0

16Y

.en

tero

colit

ica

NC

TC

1046

00.

0


FEMSLE 8064 19-3-98

Fig. 2. Phylogenetic relationships of Salmonella and related members of Enterobacteriaceae determined by maximum-likelihood analysisof 793 nucleic acid positions. The numbers that particular nodes were determined in 100 bootstrap analyses are shown for values higherthan 50.


Acknowledgments

Gitte Frederiksen, Manaz Moradi, and Jan Peter-sen are thanked for technical assistance. KimHolmstrÖm is acknowledged for the design of pri-mers. This work was ¢nanced by the Danish Centrefor Advanced Food Studies.

References

[1] Le Minor, L., Vieèron, M. and Popo¡, M. (1982) Taxonomiedes Salmonella. Ann. Microbiol. (Inst. Pasteur) 133B, 224^243.

[2] Reeves, M.W., Evins, G.M., Heiba, A.A., Plikaytis, B.D. andFarmer, J.J. III (1989) Clonal nature of Salmonella typhi andits genetic relatedness to other salmonellae as shown by multi-locus enzyme electrophoresis, and proposal of Salmonella bon-gori comb. nov. J. Clin. Microbiol. 27, 313^320.

[3] Selander, R.K., Li, J. and Nelson, K. (1996) Evolutionarygenetics of Salmonella enterica. In: E. coli and Salmonella.Cellular and Molecular Biology (Neidhardt, F.C., Ed.), Vol.2, pp. 2691^2707. ASM Press, Washington, DC.

[4] Christensen, H., Nordentoft, S. and Olsen, J.E. (1998) Phylo-genetic relationship of Salmonella based on ribosomal RNAsequences. Int. J. Syst. Bacteriol. (in press).

[5] Amann, R., Sostak, P., Ludwig, W. and Schleifer, K.H. (1988)Cloning and sequencing of genes encoding the beta subunitsof the ATP-synthases for Enterobacter aerogenes and Flavo-bacterium ferrugineum. FEMS Microbiol. Lett. 50, 101^106.

[6] Ludwig, W., Neumaier, J., Klugbauer, N., Brockmann, E.,Roller, C., Jilg, S., Reetz, K., Schachtner, I., Ludvigsen, A.,Bachleitner, M., Fischer, U. and Schleifer, K.H. (1993) Phy-logenetic relationships of Bacteria based on comparative se-quence analysis of elongation factor Tu and ATP-synthasebeta-subunit genes. Antonie van Leeuwenhoek J. Microbiol.64, 285^305.

[7] Walker, J.E., Saraste, M. and Gay, N.J. (1984) The unc oper-

on. Nucleotide sequence, regulation and structure of ATP-synthase. Biochim. Biophys. Acta 768, 164^200.

[8] Vogel, B.F., JÖrgensen, K., Christensen, H., Olsen, J.E. andGram, L. (1997) Di¡erentiation of Shewanella putrefaciensand Shewanella alga on the basis of whole-cell protein pro¢les,ribotyping, phenotypic characterization, and 16S rRNA genesequence analysis. Appl. Environ. Microbiol. 63, 2189^2199.

[9] Sanger, F., Nicklen, S. and Coulson, A.R. (1977) DNA se-quencing with chain termination inhibition. Proc. Natl. Acad.Sci. USA 74, 5463^5467.

[10] Felsenstein, J. (1981) Evolutionary trees from DNA sequen-ces: A maximum likelihood approach. J. Mol. Evol. 17, 368^376.

[11] Olsen, G.J., Matsuda, H., Hagstrom, R. and Overbeek, R.(1994) fastDNAml: A tool for construction of phylogenetictrees of DNA sequences using maximum likelihood. Comput.Appl. Biosci. 10, 41^48.

[12] Kaufmann, F. (1963) Zur Di¡erentialdiagnose der Salmonel-la-Sub-Genera I, II und III. Acta Pathol. Microbiol. Scand.58, 109^113.

[13] Boyd, E.F., Nelson, K., Wang, F.-S., Whittam, T.S. and Se-lander, R.K. (1994) Molecular genetic basis of allelic poly-morphism in malate dehydrogenase (mdh) in natural popula-tions of Escherichia coli and Salmonella enterica. Proc. Natl.Acad. Sci. USA 91, 1280^1284.

[14] Nelson, K. and Selander, R.K. (1992) Evolutionary geneticsof the proline permease gene (putP) and the control region ofthe proline utilization operon in populations of Salmonellaand Escherichia coli. J. Bacteriol. 174, 6886^6895.

[15] Wang, F.-S., Whittam, T. and Selander, R.K. (1997) Evolu-tionary genetics of the isocitrate dehydrogenase gene (icd) inEscherichia coli and Salmonella enterica. J. Bacteriol. 179,6551^6559.

[16] Maynard Smith, J. (1996) Population genetics : an introduc-tion. In: E. coli and Salmonella. Cellular and Molecular Biol-ogy (Neidhardt, F.C., Ed.), Vol. 2, pp. 2685^2690. ASMPress, Washington, DC.

[17] Ida, K., Noumi, T., Maeda, M., Fukui, T. and Futai, M.(1991) Catalytic site of F1-ATPase of Escherichia coli. J. Biol.Chem. 266, 5424^5429.

FEMSLE 8064 19-3-98


phylogenetic relationships of salmonella based on dna sequence comparison of atpd encoding the β...

Documents