Tara C. Smith, PhD and Brenda Boese

1
Tara C. Smith, PhD and Brenda Boese Center for
Emerging Infectious Diseases, Department of
Epidemiology, University of Iowa College of
Public Health, Iowa City, IA
Background Streptococcus agalactiae (group B
streptococcus, GBS) is a gram-positive bacterium
and a leading infectious cause of neonatal
morbidity and mortality. More recently, the role
of this organism as a cause of infection in
nonpregnant adults has been described. GBS is a
frequent inhabitant of the gastrointestinal tract
in both males and females, and women can also be
colonized in the genital tract.Methods An in
silico analysis was performed comparing two
sequenced isolates of Streptococcus agalactiae.
BAA-611 (2603 V/R) is a serotype V isolate and
12043 (NEM316) is a serotype III isolate. Both
strains were taken from invasive infections and
their entire genome sequences are available at
TIGR (www.tigr.org). An area of the S. agalactiae
genome was identified which contained a stretch
of 24 hypervariable genes (termed the GBS
hypervariable region, HVR). DNA and predicted
protein sequences were compared between the two
sequenced isolates in order to estimate the ratio
of non-synonymous (Ka) to synonymous (Ks) base
pair changes. This ratio (Ka/Ks), termed w,
provides a quantitative measure of selection the
higher the ratio, the more likely it is that
positive (diversifying) selection has affected
the evolution of the locus in question.Results
One of these genes (SAG1293/GBS1366 collectively
termed hvy1) was chosen for further analysis.
This gene codes for a putative protease
approximately 200 amino acids in length. The hvy1
genes in BAA-611 and 12043 were approximately 50
identical at the sequence level. Preliminary
sequencing of the gene in 10 isolates (5 serotype
III, 5 serotype V) revealed 2 new alleles in
addition to the alleles in the sequenced
isolates. A calculation of w in each case gave a
Ka/Ks ratio greater than 2, indicating positive
selection was a force driving the evolution of
the locusConclusions This hypervariable region
identified in S. agalactiae has undergone
diversifying selection, and may potentially play
a role in virulence. Additional studies are
underway to test that hypothesis.

• A hypervariable genetic region is present in the
  group B streptococcus
• Several genes in this region appear to have
  undergone positive selection

• Does S. agalactiae contain a region of
  hypervariable genes?
• What is the genetic diversity in this region?
• Are any alleles associated with invasive
  disease?

DNA and predicted protein sequences were
compared between the two sequenced isolates in
order to estimate the ratio of non-synonymous
(Ka) to synonymous (Ks) base pair changes. This
ratio (Ka/Ks), termed ?, provides a quantitative
measure of selection (7) the higher the ratio,
the more likely it is that positive selection has
affected the evolution of the locus in question.
One of these genes (SAG1273/GBS1345
collectively termed hvy1) was chosen for further
analysis. This gene codes for a hydrophilic
hypothetical protein approximately 100 amino
acids in length. The hvy1 genes in 2603V/R and
NEM316 were approximately 50 identical at the
sequence level. A calculation of ? in each case
gave a Ka/Ks ratio greater than 2, indicating
positive selection was a force driving the
evolution of the locus (see Figure 2).
We anticipate future research leading in several
directions. It is possible that this region may
be useful as a molecular typing scheme, similar
to that use for the emm gene in Streptococcus
pyogenes (1). The current method of typing
Streptococcus agalactiae is based on properties
of the capsule, and is of limited usefulness.
One reason for this is the relatively few number
of different serotypes it yields yet studies
have shown that organisms which group together
within these serotypes are far from homogeneous
(reviewed in 5). Additionally, capsular typing
is labor-intensive and few laboratories have the
reagents necessary to perform it (2). A typing
scheme based on one or more of these genes would
be reproducible, inexpensive, and portable from
laboratory to laboratory. It is predicted that
it will also provide far more than 8 groups of
GBS, elucidating more information on the
epidemiology of these organisms than the current
serotype grouping does. These genes could
also be further examined to determine whether
they play any role in bacterial pathogenesis.
Initial studies may focus on the expression of
these genes under various circumstances, both in
vitro (by cells grown in broth media or on agar
plates) or in vivo (using either a tissue culture
model or an animal model of infection).
Knockouts of these genes could be made and tested
in an animal infection model. Additionally, the
genes can be cloned and expressed in E. coli and
used as antigens to screen sera from patients who
have experienced a GBS infection in order to
determine whether these are indeed expressed in
vivo and are antigenic. Knowing the population
structure and diversity of these genes beforehand
will facilitate this last project, as it will
allow primers to be more efficiently designed.
An in silico analysis was performed comparing
two sequenced isolates of Streptococcus
agalactiae. 2603V/R is a serotype V isolate (6)
and NEM316 is a serotype III isolate (3). Both
strains were taken from invasive infections and
their entire genome sequences are available at
TIGR (www.tigr.org). An area of the S.
agalactiae genome was identified which contained
a stretch of 24 hypervariable genes (collectively
termed the GBS hypervariable region, HVR). This
region was found in cluster XIII of NEM316 and
and cluster 16 of 2603V/R.
One of the most challenging of host defenses is
adaptive immunity. Therefore, many species of
microbes have evolved a strategy that allows for
variation of antigens which are exposed to host
defenses (generally, proteins or portions of
proteins which are either exposed on the surface
of the pathogen, or secreted proteins). These
genes mutate at a high rate and undergo
positive or diversifying selectionthat is,
there is a selection for mutations that change
the amino acid sequence of the protein, and thus,
result in the generation of a novel antigen which
the host immune system has not targeted
previously. Therefore, at the population level,
the pathogen will show a large variability in
gene sequences for these antigens. In
Streptococcus agalactiae, a common cause of
sepsis and meningitis in newborns, a
hypervariable gene or region has not been
previously characterized. Most of the
epidemiological studies in this organism have
focused on the capsular serotype, of which 8
varieties have been identified (reviewed in 5).
The capsule has also been targeted as a possible
vaccine candidate, although as a polysaccharide,
it does not induce an immune response as well as
many protein antigens (4). A handful of other
putative virulence factors have been identified
in this organism (5), but all are fairly
conserved at the genetic level. Thus, this study
will be the first to systemically analyze a
potential hypervariable region in this bacterium.
It is hoped that this will lead to further
insights into not only the virulence of this
pathogen, but the natural history and evolution
of the organism as well. As this bacterium
causes severe invasive disease, particularly in
newborns and in the elderly, it merits further
study. The Streptococcus agalactiae hypervariable
region (HVR) is an area of genes which show this
diversifying selection, and as such, may be
targets of the human immune system.
1. Beall, B Facklam, R Thompson, T. 1996. J
Clin Microbiol. 34 953-8. 2. Borchardt SM,
Foxman B, Chaffin DO, Rubens CE, Tallman PA,
Manning SD, Baker CJ, Marrs CF. 2004. J. Clin.
M icrobiol. 42 146-50. 3. Glaser,P.,
Rusniok,C., Chevalier,F., et al. 2002. Mol.
Microbiol. 451499-1513. 4. Kelly, DF Moxon, ER
Pollard, AJ. 2004. Immunology. 113 163-74. 5.
Manning, SD. 2003. Front Biosci. 8s1-18. 6.
Tettelin H, Masignani V, Cieslewicz MJ, et al.
2002. PNAS. 99 12391-6. 7. Yang, Z Bielawski,
JP. 2000. TREE. 15496-502.
Table 1 summary of genes present in
HVR 1Conserved hypothetical protein 2First third
of protein is absent in type V no homology found
in 2603V/R. 3GBS 1359-1362 and 1364, and SAG
1286-1289 and 1291 correspond to proteins from
Tn5252. SAG1277 also some similarity to type
III 1132
This work was supported by a University of Iowa
New Investigator Grant (TCS).