Download Phase variation mediated niche adaptation during prolonged

Microbiology (2005), 151, 917–923
DOI 10.1099/mic.0.27379-0
Phase variation mediated niche adaptation
during prolonged experimental murine infection
with Helicobacter pylori
Laurence Salaün,1 Sarah Ayraud2 and Nigel J. Saunders1
1
Correspondence
Nigel Saunders
[email protected].
Bacterial Pathogenesis and Functional Genomics Group, The Sir William Dunn School of
Pathology, University of Oxford, South Parks Road, Oxford, OX1 3RE, UK
2
Laboratoire de Microbiologie A, CHU La Milétrie, Université de Poitiers, France
ac.uk
Received 4 June 2004
Revised 22 November 2004
Accepted 1 December 2004
Changes in the repeats associated with the recently redefined repertoire of 31 phase-variable
genes in Helicobacter pylori were investigated following murine gastric colonization for up to
one year in three unrelated H. pylori strains. Between the beginning and end of the experimental
period, changes were seen in ten genes (32 %), which would alter gene expression in one or
more of the three strains studied. For those genes that showed repeat length changes at the
longest time points, intermediate time points showed differences between the rates of change for
different functional groups of genes. Genes most likely to be associated with immediate niche
fitting changed most rapidly, including phospholipase A (pldA) and LPS biosynthetic genes.
Other surface proteins, which may be under adaptive immune selection, changed more slowly.
Restriction-modification genes showed no particular temporal pattern. The number of genes that
phase varied during adaptation to the murine gastric environment correlated inversely with their
relative fitness as previously determined in this murine model of colonization. This suggests a
role for these genes in determining initial fitness for colonization as well as in subsequent niche
adaptation. In addition, a coding tandem repeat within a phase-variable gene which does not
control actual gene expression was also investigated. This repeat was found to vary in copy
number during colonization. This suggests that changes in the structures encoded by tandem
repeats may also play a role in altered protein functions and/or immune evasion during
H. pylori colonization.
INTRODUCTION
Phase variation is a mechanism of reversible ON/OFF
gene switching normally mediated by DNA mutation or
modification. It is strongly associated with genes involved in
adaptation to different environmental conditions, such as
adaptation to different sites of infection, different steps of
host interaction and immune evasion (reviewed by Salaün
et al., 2003; Saunders, 2003; van der Woude & Baumler,
2004). This process is frequently associated with virulence
and an organism’s capacity to establish and maintain
colonization and cause disease.
Helicobacter pylori is known to phase vary many of its
genes. Initial studies of the complete phase-variable gene
Follow-up of the changes in the number of tandem repeats in three
LPS biosynthesis genes of H. pylori during the course of infection in the
mouse model is shown in Supplementary Table S1, and follow-up of
variations in the length of the repeat associated with changes in status
of phase-variable genes of H. pylori during the course of infection in the
mouse model is shown in Supplementary Table S2 available with the
online version of this paper at http://mic.sgmjournals.org.
0002-7379 G 2005 SGM
repertoire were performed based upon analysis of the two
completed genome sequences (Tomb et al., 1997; Saunders
et al., 1998; Alm et al., 1999). Using a previously established
comparative approach (Snyder et al., 2001), this repertoire
was recently reassessed and extended, based upon the two
completed genome sequences. These candidate genes and
their associated repeat polymorphisms have been investigated in a representative sample of unrelated strains (Salaün
et al., 2004), based upon the multilocus sequence typing
(MLST)-based population structure associated with human
migration (Falush et al., 2003), on which basis a group of
30 genes has been established as phase variable, or very
likely to be phase variable, in this species. An additional
phase-variable gene, described by De Vries et al. (2002), has
been included in the study reported here.
Since the original genome sequence-based analyses of the
potential phase-variable genes (Tomb et al., 1997; Saunders
et al., 1998; Alm et al., 1999), there have been several studies
supporting their predictions (reviewed by Salaün et al.,
2003, 2004). These include genes associated with LPS biosynthesis (Appelmelk et al., 1998, 2000; Logan et al., 2000),
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
Printed in Great Britain
917
L. Salaün, S. Ayraud and N. J. Saunders
flagella expression and motility (Josenhans et al., 2000),
adhesion properties (Ilver et al., 1998; Peck et al., 1999;
Yamaoka et al., 2000; Mahdavi et al., 2002) and adaptation
to acidic environments (Tannaes et al., 2001).
To date, no study has addressed the behaviour of a complete phase-variable gene repertoire in adaptation to a
complex infection model system in any species. This study
was performed to gain an overall picture of the rapidity of
phenotypic change, and to determine the relative importance of the different phase-variable genes, either singly or
in combination, in adaptation to this colonization niche.
We have assessed the repeat-associated changes in the 31
repeat-associated phase-variable genes in the mouse model
of gastric colonization, using three unrelated H. pylori
strains. We find that all the functional classes of phasevariable genes show changes over the course of the infection, that the time-course of changes differs between
different classes of genes, and that the number of changes
observed can be related to the relative fitness of the strains
within the model system. We also report variations in a
coding tandem repeat structure within one of the phasevariable genes that also shows variation indicative of
functional selection.
METHODS
The bacterially derived materials used in this study were derived
from the previously described study of Ayraud et al. (2003). It
should be noted that the samples used differ in one important
respect from those described in the former publication. For the purposes of this study, as described below, samples were deliberately
prepared from multiple colonies. This approach avoids sampling
single colonies, which might be unrepresentative of the larger colonizing population, and allows the presence of mixed populations
to be observed, thereby providing an indication of the degree and
time-course of changes.
H. pylori strains. Three H. pylori strains were used for this study:
the Sydney strain (SS1), previously adapted to the mouse stomach
(Lee et al., 1997), and two strains (Hp141 and Hp145) freshly isolated from patients (Ayraud et al., 2003). Growth conditions prior
to inoculation were as described by Ayraud et al. (2003). The strains
Hp141 and Hp145 were isolated from human stomach biopsies onto
Skirrow medium (BioMérieux), and the strain SS1 was grown on
blood agar. The three strains were subcultured once in Brucella
broth (BioMérieux) for 2 days to prepare the infecting inocula.
Experimental infections. Experimental infections were conducted
with 6-week-old female C57BL/6 inbred mice, as previously described (Ayraud et al., 2003). Animals were serially killed (two mice
for each sacrifice time point), and their stomachs entirely removed
after sacrifice and homogenized in 1 ml Brucella broth. The resulting
suspension was serially diluted and cultured on Skirrow medium.
Ten to fifteen colonies from the stomach culture of each of the two
mice were randomly selected from each sacrificial day for each
infecting strain. The colonies obtained from the two mice were
pooled together and DNA was extracted for further experiments.
Genomic DNA of the infecting strains and the emerging strains was
extracted with purification columns (QIAmp DNA minikit, Qiagen)
according to the manufacturer’s instructions.
Amplification and sequencing of the repeat-containing
regions. At first, only the infecting strain and the strain obtained at
918
day 150 for SS1 (the latest time point available for this strain) and
at day 360 for Hp141 and Hp145 were analysed; then, if a change in
phenotype was observed at this extended time point, the strains
emerging at day 3 and at day 21 were also analysed. PCRs were performed to amplify the regions of the putative phase-variable genes
containing the potentially variable repeats. Primers used in this
study for the amplification of phase-variable genes were described
previously (Salaün et al., 2004). In order to investigate variation in
the number of tandem repeats, primers were designed to amplify
the two full a-1,3-fucosyltransferase genes (HP0651 and HP0379):
HP0651WF (59-caacgcttgttgaatatcag-39) and HP651WR (59-ccatgctttagaggacatgc-39); HP0379WF (59-gcgtgctagggttttattcg-39) and
HP0379WR (59-agttttgaagtggtggatgc-39). The product amplified by
the existing primers for HP0619 already included the tandem repeat.
PCRs were carried out using Taq DNA polymerase (Invitrogen)
according to the manufacturer’s instructions. PCR products were
sequenced using the PCR primers. Sequencing was performed using
ABI Prism BigDye Terminator cycle sequencing, version 3.0 (Applied
Biosystems) and was resolved on an ABI Prism 3100 DNA sequencer
(Applied Biosystems). Sequences were read, edited and aligned using
Seqlab, from the Wisconsin Package, version 10.2 (Genetics Computer Group) through the Oxford University Bioinformatics Centre.
The ON/OFF status of the phase-variable genes was determined on
the basis of the repeat lengths.
The amplification and sequencing methods were exactly the same
as those previously extensively tested and validated for working and
interpreting repeat lengths (Salaün et al., 2004), in which genomic
templates and cloned templates were compared, and example electropherograms are presented.
RESULTS AND DISCUSSION
Phase variation is a process of stochastic reversible ON
and OFF switching of genes that typically provide adaptive advantages to different environmental conditions.
Adaptation mediated by this mechanism provides competitive advantages for colonizing subpopulations that are
more adapted to their environmental conditions than their
co-colonizing neighbours, so that they rise to dominance
through a process of clonal replacement. The rate of change
of the population composition is primarily determined by
the relative fitness differences between the alternate phenotypes (Saunders et al., 2003). We have used a redefined list
of phase-variable genes identified in H. pylori (Salaün et al.,
2004) to perform the first study of the changes within a
complete repertoire of phase-variable genes in adaptation
to a model infection system.
Although the murine experimental model of infection has
its limitations with respect to the presence of human
host-specific ligands, it is nevertheless one of the most
established and widely used tools for the investigation of
H. pylori infection. It has been shown that strains are stable
with regard to major chromosomal physical features, and
within this model system, using the same strains and mice
used for this study, the only change previously identified
was within the ppk gene [Ayraud et al., 2003; L. S., unpublished random amplified polymorphic DNA (RAPD) and
multilocus enzyme electrophoresis (MLEE) data]. Prolonged
colonization of the gastric environment, as well as exposure
to immunological responses, are common features of both
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
Microbiology 151
Phase variation in H. pylori during colonization
the model system and natural infection, and this and similar
model systems are used extensively in investigations of
colonization (Moran et al., 2000) and vaccine studies for
this pathogen (Nedrud, 1999; Raghavan et al., 2002;
Sommer et al., 2004).
The number of genes with altered repeat
lengths that would affect expression
Prolonged colonization is associated with alterations in
several phase-variable genes. Between the time of initial
colonization and the longest time point (150 or 360 days),
changes in the length of the repeats were seen in 15 of the
31 genes studied in one or more strains. Changes in five of
these repeats would not alter the expression of the associated gene, for example, changing between two lengths
associated with an OFF phenotype. Ten genes were associated with repeat length changes that resulted in alterations to in-frame shifting of the encoded proteins which
would alter their expression status in one or more strains
(Table 1). These included genes from each of the functional classes previously described (Saunders et al., 1998;
Salaün et al., 2004), with the exception of frxA (HP0642),
which is the only phase-varied gene associated with
metabolic responses. This gene was maintained as ON in
strains SS1 and Hp145, and OFF in strain Hp141, indicating an absence of a strong selective pressure for either
phenotype in this model system.
Differences in the time course of altering
different phenotypes
Differences in the rate of change indicate that different
levels of selective pressure act on different genes during the
course of infection. For each gene that showed altered
expression at the end of the study period, intermediate
genomic samples were investigated to determine the point
at which changes occurred during colonization. The time
points selected were at 3 days, to reflect adaptation to
immediately acting selective environmental conditions, and
at 21 days, to allow for the development of an adaptive
immune response, although changes at this time point
might also reflect weaker selection pressures acting over a
longer time.
The phospholipase A gene (pldA) was among the most
rapidly changing genes, with changes occurring within the
first 3 days of colonization (Table 2). The change in pldA
showed a phenotypic selection from an initial inoculum,
which comprised a mixture of ON and OFF phenotypes, to
an exclusively ON population. This is consistent with the
application of a purifying selection on the population by
the transition from in vitro conditions to conditions in the
stomach, and also with the involvement of pldA in the
adaptation to acid conditions that is necessary for residence
in the stomach, as previously proposed (Tannaes et al.,
2001). For what niche the alternate phenotype is adaptive
is currently unknown.
http://mic.sgmjournals.org
Early variations in the status of the LPS biosynthetic genes
(HP0379 and HP0217) were also observed. These changes
may reflect an early role for LPS structures in colonization
and acid resistance (Logan et al., 2000; Moran et al., 2002).
In contrast to the results of our study, Webb & Blaser (2002)
reported slower changes in the expression of these genes in
their mouse-model infection study of LPS biosynthesis
genes, in which the population reached its stable Lewis
phenotypes distribution in about 20 weeks. However, these
studies were done in C3H/HeJ mice, a lineage in which B
cells and macrophages do not respond significantly to LPS
(Wong et al., 1999), and in which the stomach ligands
available for adhesion may differ. Such differences might
account for the observed differences between these model
infections.
In contrast, genes encoding outer-membrane proteins
(babB and hopZ) changed status later in the course of
infection (following 21 days infection), a time frame
consistent with a role for phase variation in adaptation to
avoid specific immune responses mounted against these
exposed surface proteins. Notably, across all the genes
showing alterations that would affect expression, the
changes that were observed at earlier time points are
consistent with those observed at one year. This suggests a
relatively constant selective pressure on these phenotypes
over the duration of the experiment and between the
different individual mice, which contrasts with the progressive changes seen in some other model systems of LPS
phase variation (for example, Inzana et al., 1992). This
may be more consistent with a model in which some or all
of these changes are associated with other interactions,
although less strongly selected for than the most rapid
changes, such as altered adhesion properties, as has recently
been suggested by others (Solnick et al., 2004).
The status of the genes encoding enzymes of the restrictionmodification system (e.g. HP1353-4 in Hp145) showed no
particular pattern of changes over time, and did not show
consistent changes in independently colonized mice. This
suggests that there may be changes related to the relative
fitness of the phase variants, but that they are not necessarily
adaptive to specific constant environmental conditions
present in this model system. This would be consistent
with a model in which the alternate phenotypes are either
associated with competition between subpopulations of
colonizing bacteria, or are randomly co-selected within
bacteria in which other genes that are under direct selection
have phase varied.
It is impossible to determine the rate of phase variation
or, therefore, to measure the actual selective pressure within
a model infection system such as this, because there is
no means by which to measure the number of bacterial
divisions during colonization or accurately determine the
exact composition of the initial inocula. Equally, the nature
of this model requires serial sacrifice at the sampled time
points, which precludes performing studies over time in
individual mice. However, if some basic assumptions are
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
919
L. Salaün, S. Ayraud and N. J. Saunders
Table 1. Repertoire of the 31 phase-variable genes of H. pylori and their status in the three mouse-adapted strains
Absent, the gene is absent in the strain under study; (pro), the repeat is located within the promoter region; Change?, the repeat is located
within the promoter so that the change in its length in the strain hp141 cannot be interpreted as a change in status. Entries in bold type
show where change has occurred.
Function encoded by the phase-variable gene
LPS biosynthesis
a-1,3-Fucosyltransferase
a-1,3-Fucosyltransferase
a-1,2-Fucosyltransferase
Lex2B
RfaJ (a-1,2-glycosyltransferase)
RfaJ (a-1,2-glycosyltransferase)
b-1,4-N-Acetylgalactoamyl transferase
Cell-surface-associated proteins
FliP (flagellar protein)
OM adherence associated protein
Streptococcal M protein
OipA (proinflamatory outer-membrane protein)
TlpB (methyl-accepting chemotaxis protein)
BabB (Lewis b antigen-binding adhesin)
SabB (sialic acid-binding adhesin)
SabA (sialic acid-binding adhesin)
HopZ (adhesin)
Oxaloglutarate/malate translocator
PldA (phospholipase A)
HcpB (cysteine-rich protein B with b-lactamase activity)
DNA restriction/modification system
Adenine-specific methyltransferase
Type III restriction enzyme M protein
Type II R/M enzyme b-subunit
Type III R/M modification enzyme
HsdR (type I restriction enzyme R protein)
MboIIR (type II restriction enzyme R protein)
Type III restriction enzyme M protein
Metabolic proteins/other proteins
FrxA (NAD(P)H flavin oxidoreductase)
Hypothetical ORFs without identified homologies
Hypothetical protein
Hypothetical protein
Hypothetical protein
Hypothetical protein
Gene
26695 CDS
J99 CDS
SS1
Hp141
Hp145
HP0651
HP0379
HP0093-4
HP0619
HP0208
No homologue
HP0217
jhp0596
jhp1002
jhp0086
jhp0563
jhp0194
jhp0820
jhp0203
ON
OFF
ON
ON
OFF
OFF
OFF
ON
ON
ON
Absent
OFF
OFF
OFF
OFF
ONROFF
OFF
ON
OFF
OFF
ONROFF
HP0684-5
HP1417
HP0058
HP0638
HP0103
HP0896
HP0722
HP0725
HP0009
HP0143
HP0499
HP0335
jhp0625
jhp1312
jhp0050
jhp0581
jhp0095
jhp1164
jhp0659
jhp0662
jhp0007
jhp0131
jhp0451
jhp0318
ON
ON
Absent
ON
(pro)
OFF
Absent
Absent
ON
ON
MixRON
OFF
ON
OFF
OFF
ON
Change?
ONROFF
Absent
OFF
ONROFF
ON
ON
Absent
ON
OFF
OFF
OFF
(pro)
OFF
Absent
ON
ONROFF
ON
MixRON
Absent
HP1353-4
HP1369-70
HP1471
HP1522
HP0464
HP1366
No homologue
jhp1272
jhp1284
jhp1364
jhp1411
jhp0416
jhp1442
jhp1297
OFF
OFF
Absent
OFF
OFF
Absent
ON
OFF
OFF
OFF
Absent
OFF
Absent
ONROFF
ONROFF
OFF
OFF
ONROFF
OFF
ON
ON
HP0642
jhp0586
ON
OFF
ON
jhp0681
jhp1326
Not annotated
jhp0540
ON
Absent
ON
(pro)
OFF
OFF
OFF
(pro)
OFFRON
OFFRON
OFF
(pro)
HP0744
HP1433
HP0767
No homologue
made, it is possible, on the basis of our published model
system (Saunders et al., 2003), to determine the selective
pressures or fitness differences between the strains that
would account for changes in the time frames studied. If
one assumes that the bacteria can divide once per hour
and that there is a typical phase variation mutation rate of
1/10 000 per generation per cell, then 3 days of culture
represents approximately 75 generations. A difference in
fitness of 10 % or greater between the alternate and original
phenotype would normally result in complete replacement
of the original phenotype in this time frame. This may well
be an overestimate of the replication rate and if, for example,
920
the rate of division were only once in 4 h, then a fitness
difference of 50 % or greater would be required to effect a
similar change. By 3 weeks, fitness differences of only 2 %
or 6 %, for 1 h and 4 h division rates, respectively, would
bring about a complete replacement of one phenotype by
another. This indicates that changes seen in the population
may reflect quite subtle niche-selective pressures, and that
even rapidly changing phenotypes, such as for pldA, may
reflect moderate differences in the fitness of the associated
phenotypes. It should be noted that these models and this
experimental system do not distinguish between the positive
selection for an adaptive change and the selection against a
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
Microbiology 151
Phase variation in H. pylori during colonization
Table 2. Follow-up of the changes in status of phase-variable genes of H. pylori during the course of infection in the mouse
model
Absent, the gene is absent in the strain under study; (pro), the repeat is located within the promoter region; ND, not determined. Entries in
bold type show where change has occurred. The repeat lengths associated with these phenotypes are shown in Supplementary Table S2 with
the online version of this paper at http://mic.sgmjournals.org.
HP no.
Gene [time point (day)]
Repeat
SS1
[0]
HP0379*
HP0217*
HP0103
jhp1164
HP0009
HP0499
HP1353-4*
HP1522
jhp1297
HP0744
HP1433
A
C
G
G
G-pro
[TC]
[CT]
G
G
G
G
C
[AG]
C
ON
OFF
OFF
OFF
(pro)
OFF
ON
ONROFF
ON
OFF
OFF
ON
ON
Absent
SS1
[3]
ON
OFF
ON
SS1
[21]
ON
OFF
ON
SS1
[150]
Hp141
[0]
ON
OFF
ON
OFF
(pro)
OFF
ON
ON
ON
OFF
OFF
ON
ON
Absent
ON
ON
OFF
ON
(pro)
ON
ON
ON
ON
OFF
Absent
ON
OFF
OFF
Hp141
[3]
ON
Hp141
[21]
ND
ON
ON
ON
ON
Hp141
[360]
Hp145
[0]
ON
ON
OFF
ON
(pro)
OFF
OFF
ON
ON
OFF
Absent
OFF
OFF
OFF
ON
ON
ON
ON
(pro)
OFF
ON
ONROFF
ON
OFF
ON
ON
OFF
OFF
Hp145
[3]
Hp145
[21]
OFF
OFF
OFF
OFF
ON
ON
OFF
OFF
ON
ON
ON
OFF
ND
ND
ND
ND
OFF
ON
Hp145
[360]
ON
OFF
OFF
ON
(pro)
OFF
OFF
ON
OFF
OFF
OFF
ON
ON
ON
*The ON/OFF status of genes HP0379, HP0217 and HP1353-4 is directed by two repeats.
phenotype that is inherently less fit under the particular
conditions investigated. In essence, it does not specifically
address adaptation to a new niche, although this is one
possible basis for change; rather, it measures the relative
fitness of the available alternative phenotypes to the selective pressures exerted upon the population.
Lack of change in core LPS biosynthesis
Observed alterations in LPS biosynthetic gene expression
only involved those affecting terminal LPS structures
(HP0379 and HP0217). The phase-variable rfaJ homologue (HP0208; jhp0820), encoding a family 8 glycosyltransferase predicted to affect an inner-core alpha-linked
glucose to either glucose or galactose structure, remained in
the OFF configuration in all three strains during colonization. This is the most usual observed phenotype for this
gene (Salaün et al., 2004) and probably reflects the fact that
the conditions selective for the alternate structure were not
a feature of this model system.
(Ayraud et al., 2003). On the basis of competition for
colonization, strain SS1 is the most fit in the murine gastric
environment, followed by strain Hp141 and then strain
Hp145. The number of genes that were phase varied in each
strain was inversely proportional to the previously established degree of adaptation to mouse colonization (using
the same strains and the same mouse system). Strain SS1
exhibited only one phenotypic change, strain Hp141 underwent four changes, while strain Hp145 had eight phase
variation changes in phenotype. Notably, in the colonization
comparison studies (Ayraud et al., 2003), strain SS1 produced a much higher initial colonization load of the
stomach than the other two strains, while the other strains
achieved their peak colonization densities later, which
might also be related to the initial proximity of SS1 to a fit
phenotypic gene configuration for this niche. Perhaps
only a small subpopulation of the other strains was able
to survive initially. The association between phase-varied
gene phenotypes and mouse or other animal colonization
potential raises interesting questions with respect to the
basis of different strain properties.
An association between the numbers of
phase-varied genes and initial strain fitness
for the model system
The final combination of expressed genes
differs between the strains tested
The number of phase-varied genes decreases with the
degree of original strain adaptation for mouse colonization.
Previous studies have compared the relative fitness of the
strains used in this study in mouse colonization competition experiments, in both co-infections and superinfections
The final phenotypes of the varied phenotypes were not
the same for all genes across all three strains at the end of
the colonization period (Table 2). One explanation for this
could be that the different H. pylori strains used different
strategies to adapt to the mouse environment, and that the
http://mic.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
921
L. Salaün, S. Ayraud and N. J. Saunders
different strategies can be equally successful for colonization. The gene complements of each strain will inevitably
differ between these strains, which may also contribute
to their different behaviours. For example, interactions
between LPS and bacterially encoded LPS-binding proteins
might also be expected to influence population behaviour.
Similarly, since LPS structures are the product of interacting phase-variable genes, the order of gene switching and
existing phenotypes may influence the final phenotype
selected. However, the consistency between early changes
and final phenotypes for each individual strain strongly
suggests that for a given strain, with a common starting
expressed gene configuration, there is a particular phenotype which is most readily achieved and adaptive for a
given colonization niche.
Alterations in a coding tandem repeat
The tandem repeat within the lex2B homologue (HP0619)
in strain Hp145 showed length variation during colonization. Three of the LPS biosynthetic genes (HP0651, HP0379
and HP0619) contain a 21 nucleotide coding tandem repeat
(which encodes a repeated 7 amino acid motif) towards
the 39 end of the gene, for which the copy number differs
between strains. Based upon the behaviour of proteins in
other species, coding tandem repeats have the potential to
generate changes in the protein through alterations in the
number of repeated copies, which might lead to alterations in protein function and/or immunogenicity. A recent
survey in Neisseria has shown that repeats like these can
show several differences between strains, but their observed
differences have not yet been extended to studies of their
stability during infection (Jordan et al., 2003). The repeats
within the phase-variable LPS biosynthetic genes were
sequenced at each time point in all three strains (see
Supplementary Table S1, available with the online version
of this paper at http://mic.sgmjournals.org, for details of
repeat numbers for each strain and time point). During
the course of infection, the number of coding tandem
repeats located at the 39 end of the genes encoding the
two fucosyltransferases (HP0651 and HP0379) remained
unchanged. In contrast, changes occurred in the number of
tandem repeats within the gene encoding the glycosyltransferase (HP0619) in strain Hp145, but not in the other two
strains. The changes occurred in two different colonizing
populations, as detected at days 3 and 360. The rapid change
detected at day 3 may reflect either a functional advantage
or, potentially, co-selection with another altered phenotype, such as pldA. This type of variation might also be
involved in immune evasion strategies. It is premature to
invoke a mechanism for selection in this regard, but the
observed changes strongly suggest that this mechanism is
an additional source of functional flexibility in this species.
The different changes observed between the two altered
populations indicate that this is not simply the product of a
founder effect from a second population in the inoculum.
This study addressed the whole repertoire of known phasevariable genes simultaneously in a model infection, and
922
reveals that many of these genes are phase varied in the
process of adaptation to colonization of the murine gastric
environment. It shows that these genes display different
temporal changes reflecting the relative importance of the
different associated selective pressures (Saunders et al.,
2003) or, alternatively, in some cases the time necessary to
mount an immune response. The correlation between the
adapted nature of the strains to the mouse environment
and the number of genes that change during colonization
suggests that the expression state of phase-varied genes in
the population may significantly influence the capacity for
initial colonization and hence exchange between hosts,
and suggests that diversity within the initial inoculum is
likely to influence colonization potential. This indicates that
phase variation, and the associated population diversity,
affects the population fitness in both initial colonization
and subsequent adaptation.
ACKNOWLEDGEMENTS
A Wellcome Trust Advanced Research Fellowship awarded to N. J. S.
supported N. J. S. and L. S.
REFERENCES
Alm, R. A., Ling, L. S., Moir, D. T. & 20 other authors (1999).
Genomic-sequence comparison of two unrelated isolates of the
human gastric pathogen Helicobacter pylori. Nature 397, 176–180.
Appelmelk, B. J., Shiberu, B., Trinks, C. & 10 other authors
(1998). Phase variation in Helicobacter pylori lipopolysaccharide.
Infect Immun 66, 70–76.
Appelmelk, B. J., Martino, M. C., Veenhof, E. & 7 other authors
(2000). Phase variation in H type I and Lewis a epitopes of
Helicobacter pylori lipopolysaccharide. Infect Immun 68, 5928–5932.
Ayraud, S., Janvier, B., Salaün, L. & Fauchère, J. L. (2003).
Modification in the ppk gene of Helicobacter pylori during single
and multiple experimental murine infections. Infect Immun 71,
1733–1739.
de Vries, N., Duinsbergen, D., Kuipers, E. J., Pot, R. G.,
Wiesenekker, P., Penn, C. W., Van Vliet, A. H., VandenbrouckeGrauls, C. M. & Kusters, J. G. (2002). Transcriptional phase variation
of a type III restriction-modification system in Helicobacter pylori.
J Bacteriol 184, 6615–6623.
Falush, D., Wirth, T., Linz, B. & 15 other authors (2003). Traces of
human migrations in Helicobacter pylori populations. Science 299,
1582–1585.
Ilver, D., Arnqvist, A., Ogren, J. & 7 other authors (1998).
Helicobacter pylori adhesin binding fucosylated histo-blood group
antigens revealed by retagging. Science 279, 373–377.
Inzana, T. J., Gogolewski, R. P. & Corbeil, L. B. (1992). Phenotypic
phase variation in Haemophilus somnus lipooligosaccharide during
bovine pneumonia and after in vitro passage. Infect Immun 60,
2943–2951.
Jordan, P., Snyder, L. A. & Saunders, N. J. (2003). Diversity in
coding tandem repeats in related Neisseria spp. BMC Microbiol 3, 23.
Josenhans, C., Eaton, K. A., Thevenot, T. & Suerbaum, S. (2000).
Switching of flagellar motility in Helicobacter pylori by reversible
length variation of a short homopolymeric sequence repeat in fliP, a
gene encoding a basal body protein. Infect Immun 68, 4598–4603.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
Microbiology 151
Phase variation in H. pylori during colonization
Lee, A., O’Rourke, J., De Ungria, M. C., Robertson, B.,
Daskalopoulos, G. & Dixon, M. F. (1997). A standardized mouse
pp. 103–124. Edited by B. Henderson & P. C. F. Oyston. Cambridge:
Cambridge University Press.
model of Helicobacter pylori infection: introducing the Sydney strain.
Gastroenterology 112, 1386–1397.
Saunders, N. J., Peden, J. F., Hood, D. W. & Moxon, E. R. (1998).
Logan, S. M., Conlan, J. W., Monteiro, M. A., Wakarchuk, W. W. &
Altman, E. (2000). Functional genomics of Helicobacter pylori:
identification of a beta-1,4-galactosyltransferase and generation of
mutants with altered lipopolysaccharide. Mol Microbiol 35, 1156–1167.
Mahdavi, J., Sonden, B., Hurtig, & 20 other authors (2002).
Helicobacter pylori SabA adhesin in persistent infection and chronic
inflammation. Science 297, 573–578.
Simple sequence repeats in the Helicobacter pylori genome. Mol
Microbiol 27, 1091–1098.
Saunders, N. J., Moxon, E. R. & Gravenor, M. B. (2003). Mutation
rates: estimating phase variation rates when fitness differences are
present and their impact on population structure. Microbiology 149,
485–495.
Snyder, L. A., Butcher, S. A. & Saunders, N. J. (2001). Com-
Moran, A. P., Sturegard, E., Sjunnesson, H., Wadstrom, T. & Hynes,
S. O. (2000). The relationship between O-chain expression and
parative whole-genome analyses reveal over 100 putative phasevariable genes in the pathogenic Neisseria spp. Microbiology 147,
2321–2332.
colonisation ability of Helicobacter pylori in a mouse model. FEMS
Immunol Med Microbiol 29, 263–270.
Solnick, J. V., Hansen, L. M., Salama, N. R., Boonjakuakul, J. K. &
Syvanen, M. (2004). Modifications of Helicobacter pylori outer
Moran, A. P., Knirel, Y. A., Senchenkova, S. N., Widmalm, G., Hynes,
S. O. & Jansson, P. E. (2002). Phenotypic variation in molecular
membrane protein expression during experimental infection of
rhesus macaques. Proc Natl Acad Sci U S A 101, 2106–2111.
mimicry between Helicobacter pylori lipopolysaccharides and human
gastric epithelial cell surface glycoforms. Acid-induced phase variation in Lewis(x) and Lewis(y) expression by H. pylori lipopolysaccharides. J Biol Chem 277, 5785–5795.
Sommer, F., Wilken, H., Faller, G. & Lohoff, M. (2004). Systemic Th1
Nedrud, J. G. (1999). Animal models for gastric Helicobacter
Tannaes, T., Dekker, N., Bukholm, G., Bijlsma, J. J. & Appelmelk,
B. J. (2001). Phase variation in the Helicobacter pylori phospho-
immunology and vaccine studies. FEMS Immunol Med Microbiol
24, 243–250.
Peck, B., Ortkamp, M., Diehl, K. D., Hundt, E. & Knapp, B. (1999).
immunization of mice against Helicobacter pylori infection with CpG
oligodeoxynucleotides as adjuvants does not protect from infection
but enhances gastritis. Infect Immun 72, 1029–1035.
lipase A gene and its role in acid adaptation. Infect Immun 69,
7334–7340.
Conservation, localization and expression of HopZ, a protein
involved in adhesion of Helicobacter pylori. Nucleic Acids Res 27,
3325–3333.
Tomb, J. F., White, O., Kerlavage, A. R. & 39 other authors (1997).
Raghavan, S., Hjulstrom, M., Holmgren, J. & Svennerholm, A. M.
(2002). Protection against experimental Helicobacter pylori infection
van der Woude, M. W. & Baumler, A. J. (2004). Phase and antigenic
after immunization with inactivated H. pylori whole-cell vaccines.
Infect Immun 70, 6383–6388.
Webb, G. F. & Blaser, M. J. (2002). Dynamics of bacterial pheno-
The complete genome sequence of the gastric pathogen Helicobacter
pylori. Nature 388, 539–547.
variation in bacteria. Clin Microbiol Rev 17, 581–611.
Salaün, L., Snyder, L. A. & Saunders, N. J. (2003). Adaptation
type selection in a colonized host. Proc Natl Acad Sci U S A 99,
3135–3140.
by phase variation in pathogenic bacteria. Adv Appl Microbiol 52,
263–301.
Wong, P. M., Kang, A., Chen, H., Yuan, Q., Fan, P., Sultzer, B. M.,
Kan, Y. W. & Chung, S. W. (1999). Lps(d)/Ran of endotoxin-resistant
Salaün, L., Linz, B., Suerbaum, S. & Saunders, N. J. (2004). The
C3H/HeJ mice is defective in mediating lipopolysaccharide endotoxin responses. Proc Natl Acad Sci U S A 96, 11543–11548.
diversity within an expanded and re-defined repertoire of phasevariable genes in Helicobacter pylori. Microbiology 150, 817–830.
Saunders, N. J. (2003). Evasion of antibody responses: bacterial
phase variation. In Bacterial Evasion of Host Immune Responses,
http://mic.sgmjournals.org
Yamaoka, Y., Kwon, D. H. & Graham, D. Y. (2000). A M(r) 34,000
proinflammatory outer membrane protein (oipA) of Helicobacter
pylori. Proc Natl Acad Sci U S A 97, 7533–7538.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 08 May 2017 09:52:09
923