Download Rates and patterns of chromosome evolution in enteric bacteria

580
Rates and patterns of chromosome evolution in enteric bacteria
Howard Ochman* and Ulfar Bergthorsson†
Although several types of large-scale alterations potentially
affect the structure and organization of bacterial genomes,
recent analyses of physical maps and complete genomic
sequences reveal that chromosome heterogeneity in enteric
bacteria has resulted from the acquisition and deletion of large
segments of DNA. These acquired sequences can provide
novel functions immediately upon their introduction and play a
significant role in the diversification of bacterial species.
Addresses
*Department of Biology, University of Rochester, Rochester, NY
14627, USA; e-mail: [email protected]
†Department of Biology, University of Utah, Salt Lake City, UT 84112,
USA
Correspondence: H Ochman
Current Opinion in Microbiology 1998, 1:580–583
http://biomednet.com/elecref/1369527400100580
© Current Biology Ltd ISSN 1369-5274
Abbreviation
sv.
serovar
Introduction
How quickly, and in what way, do bacterial chromosomes
evolve? Comparisons of the complete genomic sequences
of some dozen microorganisms have revealed that certain
features of bacterial chromosomes have remained conserved over vast evolutionary periods [1•,2•]. For example, there are similarities in the arrangement of
sequences, and in the order of specific genes surrounding
the replication origin in diverse bacterial genomes; in fact,
in lieu of direct experimental localization, the same organization of these specific genes in other organisms has
allowed the preliminary positioning of replication origins
[3,4••]. In addition, the proteins encoded in each of the
completely sequenced genomes can be classified according to their structure, function and evolutionary relationships, yielding information about the ancestry of groups of
proteins, the origins of novel sequences, and the modification of gene function [5••,6].
Although the comparative analysis of whole chromosomes
is extremely useful for identifying the unique and conserved features of genomes, the organisms sequenced to
date are too distantly related, and their genomes too divergent, to allow for a reasonable assessment of the rates and
patterns of chromosome evolution in bacteria. This is not
surprising: organisms occupying very different habitats
and separated for hundreds of millions of years have sustained so many changes that the evolutionary histories
may have been erased [2•,7•,8•]. Even the congeneric
species Mycoplasma pneumoniae and M. genitalium (for
which complete genome sequences are available) are sufficiently different in chromosome size and organization to
impede characterization of the specific order, number, and
rate of genetic events that distinguish their genomes [9].
Therefore, to trace the course of chromosome evolution in
bacteria, it is most informative to compare the genomes of
very closely related organisms. In this paper, we review
several recent studies have detected unprecedented
amounts of chromosome variation within bacterial species
and have begun to identify the factors that influence
genome size and organization.
Comparative genomics of Escherichia coli
and Salmonella
Early alignments of the genetic maps of Escherichia coli
K12 and Salmonella enterica serovar (sv.) Typhimurium LT2
revealed a surprising amount of congruence and promoted
the view that bacterial chromosomes are evolutionarily
well conserved. But although the chromosomes of these
species are of similar size and gene order, there are several large regions unique to each of the species as well as differences in their gene arrangements. These findings have
led to investigations focusing on four general aspects of
chromosome evolution: first, the degree of chromosome
heterogeneity within species; second, the mechanisms
generating diversity in chromosome organization; third,
the role of variable regions; and, finally, the rate of chromosome evolution.
Despite the similarity in chromosome size of laboratory
isolates — the E. coli K12 chromosome is 4.6 Mb and the
Typhimurium LT2 chromosome is 4.8 Mb — strains from
natural populations of each of these species can differ by
as much as 1 Mb (i.e. more than 20% in total length).
Chromosome sizes in natural isolates of E. coli range from
4.5 to 5.5 Mb, whereas the chromosomes from strains of
S. enterica range from 3.9 to 4.9 Mb in length [10,11•,12•].
Though of equivalent size ranges, the apportionment of
chromosome diversity appears to differ in these two
species. Nearly all of the chromosome size variation in
Salmonella was detected among strains typed to a single
serovar, S. enterica sv. Typhi, and much of the size diversity in serovars other than Typhi maps to a single region of
the chromosome. In contrast, the chromosome size variation in E. coli is not restricted to a single group of strains,
or to a specific region of the chromosome, and there is a
large phylogenetic component to the variation, as evident
from the significant differences in chromosome size
between broad subspecific groups [13].
Source of chromosome diversity
Under laboratory conditions, enteric bacteria display a
high incidence of spontaneous gene inversions, duplications and translocations; however, these events do not
seem to account for much of the chromosome heterogeneity observed in natural populations. There are a few
Chromosome evolution in enteric bacteria Ochman and Bergthorsson
cases where strains are known to differ by chromosome
rearrangements: for example, many of the differences in
the organization of S. enterica sv. Typhi chromosome have
resulted from recombination between ribosomal RNA
operons [14], and E. coli K12 and S. enterica sv. Enteriditis
SSU7798 can each be distinguished from Typhimurium
LT2 on the basis of a large inversion spanning the replication terminus [15••]. Natural strains of E. coli and S.
enterica undoubtedly harbor additional rearrangements
that are not detected by large-scale physical mapping
techniques; however, there is overwhelming evidence
that the majority of variation in genome size and content
is generated through the acquisition and deletion of large
chromosomal segments.
Base composition is relatively homogeneous over the
entire bacterial chromosome and, therefore, regions
acquired through horizontal transfer from distantly related
organisms can often be identified by their atypical base
compositions or other sequence characteristics. Applying a
genomic subtraction procedure, Lan and Reeves [16] estimated the amount of DNA contained in S. enterica sv.
Typhimurium LT2 that was not present in the genomes of
four divergent strains of S. enterica. They established that
the Typhimurium LT2 genome contained as much as
1.3 Mb of unique DNA when compared to a very distantly
related strain; by analyzing the genetic content and base
composition of these strain-specific sequences, they concluded that nearly half of this DNA was gained through
horizontal transfer.
Considerable attention has been directed towards the
investigation of pathogenicity islands, which are chromosomal regions required for virulence in pathogenic strains
but absent from related strains that do not cause disease
[17,18]. Boyd and Hartl [19••] found that genes associated
with pathogenicity islands were confined to those subgroups of E. coli with larger genome sizes, indicating that
these regions contribute to the chromosome size variation
in natural populations of E. coli. Most notably, the phylogenetic distribution of these virulence-associated genes
within E. coli suggests that pathogenicity islands are
ancestral to these subgroups, despite the fact that most of
these present-day strains were originally isolated from
healthy hosts.
Chromosomal regions recognized as arising through horizontal transfer are commonly situated adjacent to tRNA
loci [20]. This distribution implicates bacteriophages as
vehicles for gene transfer, since several lysogenic coliphages target tRNA loci, presumably because tRNA
sequences are conserved across taxa. The same tRNA
locus has been used as the integration site for distinct
DNA segments in different strains or species. For example, the selC locus contains a 35 kb insert in enteropathogenic E. coli [21•], a 70 kb insert in uropathogenic E. coli
[18], and a 17 kb insert in S. enterica [22•], and each of these
inserted regions encodes unique sequences and was
581
acquired independently. Because of the variety of
sequences already detected at tRNA loci, it is certain that
further analysis of these sites will uncover new inserts in
other strains or species.
Functional aspects of chromosome
heterogeneity
Many changes in chromosome structure and organization
have deleterious effects on DNA replication and cell
growth. In laboratory populations of E. coli, there is a
reduction in cell fitness related to the degree of asymmetry in the location of the replication terminus relative to
the origin [23], and in Salmonella, large inversions encompassing certain regions of the chromosome are lethal to the
cell [24]. Although selection on the basis of chromosome
structure has certainly occurred in natural populations —
for example, each of the inversions distinguishing E. coli
sv. Enteritidis and sv. Typhimurium encompasses the terminus but maintains chromosome symmetry — there is
currently no evidence that any of the observed differences
in chromosome size and organization have an effect on the
fitness of strains recovered from natural populations. Since
bacterial chromosomes contain mostly coding regions, it
has been predicted that strains with smaller genomes
would lack scores of genes and would be adapted to rapid
growth under complex nutrient-rich conditions, whereas
strains with larger genomes would be at a selective advantage in nutrient-poor conditions [12•]. But among natural
isolates of E. coli, there is no association between growth
rate and genome size, and other factors, such as the translational efficiency of ribosomes, outweigh the effects of
total genome size on growth rates [25].
Because much of the heterogeneity in E. coli and S. enterica chromosomes is attributed to insertions and deletions,
the evolutionary consequences of this variation are best
judged from the specific functions of genes residing in
these regions. Although most of the acquired or deleted
regions probably have little effect on the organism, some
horizontally acquired sequences can provide a novel
function immediately upon their introduction and, in
effect, change the character of a species. The genetic
contents of several species-specific regions are known,
and many confer novel metabolic properties that distinguish E. coli and S. enterica. For example, the lac operon,
allowing for the degradation of β-galactosides, was
acquired by E. coli, and in S. enterica the cob and pdu operons, providing for vitamin B12 biosynthesis and the B12dependent degradation of propanediol, also arose through
horizontal transfer [26••].
As noted, many of the genes implicated in pathogenesis —
such as the genes required for host cell invasion and intramacrophage survival by Salmonella [17,21•,27,28] — occur on
unique segments of the chromosome. In contrast to pathogenicity islands, which can confer virulence upon their acquisition, the absence of certain chromosomal regions may also
contribute to pathogenesis. For example, Shigella and
582
Genomics
enteroinvasive E. coli each harbor a chromosomal deletion of
up to 90 kb relative to E. coli K12. This region inhibits
enterotoxin activity and its deletion facilitates the expression
of plasmid-borne virulence genes in these pathogens [29•].
Rates of chromosome evolution
Despite the relatively high frequency of spontaneous
alterations in bacterial chromosomes, evidence from natural isolates of E. coli and S. enterica suggests that largescale changes in the size and organization of bacterial
genomes are relatively rare events on an evolutionary
timescale. Short-term rates of chromosome evolution
have been analyzed in experimental populations of
E. coli B propagated for 2000 generations under four different thermal regimes: 32°C, 37°C, 42°C, and alternating 32°C and 42°C. In an analysis of these lines,
chromosome alterations were monitored by pulsed-field
gel electrophoresis and a total of 12 changes (formed
through deletions, duplications, inversions or point
mutations) were detected, with none occurring in strains
cultured at 37°C (U Bergthorsson, unpublished data).
These results suggest that strains propagated at temperatures other than 37°C either have higher mutation rates
or have incurred changes that are adaptive under nonstandard growth conditions.
How does this rate of chromosomal evolution in bacteria
compare to that in eukaryotes? Differences in genome size
and organization make such comparisons difficult; but for
experimental populations of Saccharomyces cerevisiae, the
number of chromosome alterations, as detected by pulsedfield gel electrophoresis [30], was about twofold higher
than in E. coli.
Conclusions
Despite their longer evolutionary history, bacteria display far less variation in total genome size than do
eukaryotes, but they are certainly more adept at obtaining new chromosomal regions that confer unique phenotypic properties. Based on the sequence characteristics
of horizontally transferred genes, it has been estimated
that E. coli has acquired new sequences at a rate of nearly 30 kb per million years since diverging from the
Salmonella lineage 100 million years ago. Although many
of the acquired genes were subsequently deleted, those
sequences that have persisted, which constitute about
18% of the current genome, include genes that distinguish E. coli from other enteric species [31••]. Such findings promote the view that bacterial speciation is driven
by horizontal transfer, which introduces genes permitting the rapid exploitation of new environments. The
identification of the full complement of sequences that
distinguish closely-related bacteria will undoubtedly
come from comparisons of complete genomic sequences.
Such studies have, in fact, already begun with the recent
completion of the nucleotide sequence of enterohemorrhagic E. coli 0157:H7, a pathogen whose chromosome is
some 30% larger than that of E. coli K12.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
1.
•
Tamames J, Casari G, Ouzounis C, Valencia A: Conserved clusters
of functionally related genes in two bacterial genomes. J Mol Evol
1997, 44:66-73.
This paper, along with [2•,8•], compares some of the completely sequenced
genomes and describes the manner in which gene order and chromosome
organization have evolved in bacteria. By first classifying genes into one of
several functional classes and then examining their relative positions, the
authors find that in both Escherichia coli and Haemophilus influenzae
functionally related genes tend to be neighbors more often than functionally
unrelated genes.
2. Watanabe H, Mori H, Itoh T, Gojobori T: Genome plasticity as a
•
paradigm of eubacterial evolution. J Mol Evol 1997, 44:S57–S64.
By examining the locations of homologous genes in several sequenced
genomes, these authors detect very few evolutionarily conserved gene
clusters and conclude that bacterial genomes are subject to repeated
events that alter the chromosome structure.
3.
Ogasawara N, Yoshikawa H: Genes and their organization in the
replication region of the bacterial chromosome. Mol Microbiol
1992, 6:629-634.
4. Mrázek J, Karlin S: Strand asymmetry in bacterial and large viral
•• genomes. Proc Natl Acad Sci USA 1998, 95:3720-3725.
Most bacterial chromosomes are polarized around their origins of
replication, both in terms of base composition and gene distribution. This
paper examines the extent of GC skew — the excess of G over C in the
leading strand of replication — in several sequenced genomes and
discusses the potential forces generating the strand asymmetry in
base composition.
5. Huynen MA, Bork P: Measuring genome evolution. Proc Natl Acad
•• Sci USA 1998, 95:5849-5856.
This is a superb analysis of the organization and evolution of bacterial
chromosomes as reconstructed from complete genome sequences. The
authors describe the relative rates at which differ features of the genome,
ranging from protein sequence to overall chromosome content and
structure, have changed over an evolutionary timescale.
6.
Tatsunov RL, Koonin EV, Lipman DJ: A genomic perspective on
protein families. Science 1997, 278:631-637.
7.
•
Koonin EV, Galperin MY: Prokaryotic genomes: the emerging
paradigm of genome-based microbiology. Curr Opin Genet Dev
1997, 7:757-763.
A comparison of the complete sequences of several eubacterial and archael
genomes shows that while families of proteins are conserved among taxa,
gene order and gene families are not. On the basis of the heterogeneity
among genomes, the authors conclude that horizontal transfer plays a
significant role in the evolution of genomes.
8.
•
De Rosa R, Labedan B: The evolutionary relationships between the
two bacteria Escherichia coli and Haemophilus influenzae and
their putative last common ancestor. Mol Biol Evol 1998, 15:17-27.
On the basis of an analysis of the genes common to E. coli and H.
influenzae, these authors hypothesize that the common ancestor of these
species had a genome size similar to that of present day E. coli. Several
gene duplications are ancestral and common to both species; except for
few small regions, gene order has not been conserved since these
species diverged.
9.
Himmelreich R, Plagens H, Hilbert H, Reiner B, Herrman R:
Comparative analysis of the bacteria Mycoplasma pneumoniae
and Mycoplasma genitalium. Nucleic Acids Res 1997, 25:701-712.
10. Liu S-L, Hessel A, Sanderson KE: Genomic mapping with I-Ceu I,
an intron-encoded endonuclease specific for genes for ribosomal
RNA, in Salmonella spp., Escherichia coli, and other bacteria. Proc
Natl Acad Sci USA 1993, 90:6874-6878.
11. Thong KL, Puthucheary SD, Pang T: Genome size variation among
•
recent human isolates of Salmonella typhi. Res Microbiol 1997,
148:229-235.
On the basis of multilocus enzyme electrophoresis data, natural isolates of
Typhi are thought to be genetically homogeneous; however, this study
shows that strains of Typhi exhibit extraordinary plasticity in chromosome
size and organization, with genomes that vary from 3.9 to 4.9 Mb in length.
12. Bergthorsson U, Ochman H: Distribution of chromosome length
•
variation in natural isolates of Escherichia coli. Mol Biol Evol 1998,
15:9-16.
Chromosome evolution in enteric bacteria Ochman and Bergthorsson
Among natural isolates of E. coli, chromosome sizes range from 4.5 to
5.5 Mb. The distribution of this length variation is not random: strains with
larger chromosomes are primarily found in certain subspecific groups and
the chromosome size variation has a symmetric distribution with respect to
the replication origin.
13. Bergthorsson U, Ochman H: Heterogeneity of genome sizes
among natural isolates of Escherichia coli. J Bacteriol 1995,
177:5784-5789.
14. Liu S-L, Sanderson KE: Highly plastic chromosomal organization in
Salmonella typhi. Proc Natl Acad Sci USA 1996, 93:10303-10308.
15. Sanderson KE, Liu S-L: Chromosomal rearrangements in enteric
•• bacteria. Electrophoresis 1998, 19:569-572.
A review on the variation in chromosome organization found within
Salmonella enterica which describes physical mapping techniques and the
types of structural changes that have been detected by these procedures.
Serovars of S. enterica showing little or no host specificity appear to have
more conserved chromosome structure than do serovars adapted to a
single host.
16. Lan R, Reeves PR: Gene transfer is a major factor in bacterial
evolution. Mol Biol Evol 1996, 13:47-55.
17.
Groisman EA, Ochman H: Pathogenicity islands: bacterial evolution
in quantum leaps. Cell 1996, 87:791-794.
18. Hacker J, Blum-Oehler G, Mühldorfer I, Tschäpe H: Pathogenicity
islands of virulent bacteria: structure, function and impact on
microbial evolution. Mol Microbiol 1997, 23:1089-1097.
19. Boyd EF, Hartl DL: Chromosomal regions specific to pathogenic
•• isolates of Escherichia coli have a phylogenetically clustered
distribution. J Bacteriol 1998, 5:1159-1165.
The distribution of four virulence determinants associated with pathogenicity
islands is clustered in the subspecific groups of E. coli having larger
genomes. The distribution suggests that these virulence genes were
present in the ancestors of these subgroups despite the fact that the
majority of these strains were originally isolated from healthy hosts.
20. Cheetam BF, Katz ME: A role for bacteriophages in the evolution
and transfer of bacterial virulence determinants. Mol Microbiol
1995, 18:201-208.
21. Elliott SJ, Wainwright LA, McDaniel TD, Jarvis KG, Deng YK, Lai LC,
•
McNamara BP, Donnenberg MS, Kaper JB: The complete sequence
of the locus of enterocyte effacement (LEE) from
enteropathogenic Escherichia coli E2348/69. Mol Microbiol 1998,
28:1-4.
The insertion of the LEE island can, in a single step, convert a benign strain
of E. coli into a pathogen. This paper provides the newly determined
sequence of the entire insert; and aside from containing genes that are
apparently specific to LEE, this region also encodes a type III secretion
apparatus homologous to those present in pathogenicity islands of other
enteric pathogens, including Salmonella, Shigella and Yersinia.
22. Blanc-Potard A, Groisman EA: The Salmonella selC locus contains
•
a pathogenicity island mediating intramacrophage survival.
EMBO J 1997, 16:5376-5385.
The selC locus has been shown to be a common target for pathogenicity
islands in Escherichia coli, and this study takes the clever approach of
583
surveying the corresponding site in the Salmonella genome. The authors
discover a 17 kb insert required for virulence, adding to the growing list of
pathogenicity islands in Salmonella.
23. Hill CW, Gray JA: Effects of chromosomal inversion on cell fitness
in Escherichia coli K-12. Genetics 1988, 119:771-778.
24. Miesel L, Segall A, Roth JR: Construction of chromosomal
rearrangements in Salmonella by transduction: inversions of nonpermissive segments are not lethal. Genetics 1994, 137:919-932.
25. Mikkola R, Kurland CG: Is there a unique ribosome phenotype for
naturally occurring Escherichia coli? Biochimie 1991,
73:1061-1066.
26. Lawrence JG, Roth JR: Selfish operons: horizontal transfer may
•• drive the evolution of gene clusters. Genetics 1996,
143:1843-1860.
This paper very convincingly makes the case that bacterial genes
specifying a single metabolic function are typically arranged into operons
because this organization often facilitates efficient horizontal transfer of
genes among organisms. Their model accounts for the mosaic structure of
some bacterial genomes whereby ancestral chromosomal material is
interspersed with horizontally transferred operons providing novel
metabolic functions.
27.
Ochman H, Soncini FC, Solomon F, Groisman EA: Identification of a
pathogenicity island required for Salmonella survival in host cells.
Proc Natl Acad Sci USA 1996, 93:7800-7804.
28. Shea JE, Hensel M, Gleeson C, Holden DW: Identification of a
virulence locus encoding a second type III secretion system in
Salmonella typhimurium. Proc Natl Acad Sci USA 1996,
93:2593-2597.
29. Maurelli AT, Fernandez RE, Bloch CA, Rode CK, Fasano A: “Black
•
holes” and bacterial pathogenicity: a large genomic deletion that
enhances the virulence of Shigella spp. and enteroinvasive
Escherichia coli. Proc Natl Acad Sci USA 1998, 95:3943-3948.
Strains of Shigella and enteroinvasive E. coli were found to have
chromosomal deletions of cadA and the surrounding region. The
presence of cadA inhibits enterotoxin activity and attenuates virulence in
Shigella and these deletions are likely to have played a large role in the
evolution of these pathogens. A similar situation exists for ompT, a
surface protease which is present in E. coli K12 but not in Shigella and
enteroinvasive E. coli.
30. Adams J, Puskas-Rozsa S, Simlar J, Wilke CM: Adaptation and
major chromosomal changes in populations of Saccharomyces
cerevisiae. Curr Genet 1992, 22:13-19.
31. Lawrence, JG, Ochman H: Amelioration of bacterial genomes:
•• rates of change and exchange. J Mol Evol 1997, 44:383-397.
Based on their GC contents, genes acquired through horizontal transfer
can be distinguished from ancestral DNA. At the time of introduction,
horizontally transferred genes reflect the base composition of the donor
genome; but, over time, these sequences will ameliorate to reflect the DNA
composition of the new genome. This paper develops a method to estimate
the time of acquisition of horizontally transferred genes, which makes it
possible to establish the age of transferred genes and the rate at which
DNA has been acquired on an evolutionary timescale.