Serum Antibodies to Porphyromonas gingivalis
Chaperone HtpG Predict Health in Periodontitis
Charles E. Shelburne1*, P. Sandra Shelburne1, Vishnu M. Dhople1¤, Domenica G. Sweier2, William V.
Giannobile3, Janet S. Kinney3, Wilson A. Coulter4, Brian H. Mullally4, Dennis E. Lopatin1
1Department of Biologic and Materials Sciences, The University of Michigan School of Dentistry, Ann Arbor, Michigan, United States of America, 2Department of
Cariology, Restorative Sciences and Endodontics, The University of Michigan School of Dentistry, Ann Arbor, Michigan, United States of America, 3Department of
Periodontics and Oral Medicine and The Michigan Center for Oral Health Research, The University of Michigan School of Dentistry, Ann Arbor, Michigan, United States of
America, 4Centre for Oral Research, Queen’s University, Belfast, Northern Ireland
Background: Chaperones are ubiquitous conserved proteins critical in stabilization of new proteins, repair/removal of
defective proteins and immunodominant antigens in innate and adaptive immunity. Periodontal disease is a chronic
inflammatory infection associated with infection by Porphyromonas gingivalis that culminates in the destruction of the
supporting structures of the teeth. We previously reported studies of serum antibodies reactive with the human chaperone
Hsp90 in gingivitis, a reversible form of gingival disease confined to the oral soft tissues. In those studies, antibodies were at
their highest levels in subjects with the best oral health. We hypothesized that antibodies to the HSP90 homologue of P.
gingivalis (HtpG) might be associated with protection/resistance against destructive periodontitis.
Methodology/Principal Findings: ELISA assays using cloned HtpG and peptide antigens confirmed gingivitis subjects
colonized with P. gingivalis had higher serum levels of anti-HtpG and, concomitantly, lower levels of attachment
loss. Additionally, serum antibody levels to P. gingivalis HtpG protein were higher in healthy subjects compared to
patients with either chronic or aggressive periodontitis. We found a negative association between tooth attachment
loss and anti-P. gingivalis HtpG (p=0.043) but not anti-Fusobacterium nucleatum (an oral opportunistic commensal)
HtpG levels. Furthermore, response to periodontal therapy was more successful in subjects having higher levels of
anti-P. gingivalis HtpG before treatment (p=0.018). There was no similar relationship to anti-F. nucleatum HtpG
levels. Similar results were obtained when these experiments were repeated with a synthetic peptide of a region of
P. gingivalis HtpG.
Conclusions/Significance: Our results suggest: 1) anti-P. gingivalis HtpG antibodies are protective and therefore predict
health periodontitis-susceptable patients; 2) may augment the host defence to periodontitis and 3) a unique peptide of P.
gingivalis HtpG offers significant potential as an effective diagnostic target and vaccine candidate. These results are
compatible with a novel immune control mechanism unrelated to direct binding of bacteria.
Citation: Shelburne CE, Shelburne PS, Dhople VM, Sweier DG, Giannobile WV, et al. (2008) Serum Antibodies to Porphyromonas gingivalis Chaperone HtpG
Predict Health in Periodontitis Susceptible Patients. PLoS ONE 3(4): e1984. doi:10.1371/journal.pone.0001984
Editor: Debbie Fox, The Research Institute for Children at Children’s Hospital New Orleans, United States of America
Received January 15, 2008; Accepted February 28, 2008; Published April 23, 2008
Copyright: ? 2008 Shelburne et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Supported by NIH/NIDCR DE11117 (DEL) and patient sample collections were supported by NIH grants NIH/NIDCR U01-DE-14961 and NIH/NCRR M01-
RR000042 (WVG). The sponsors of this work had no role in the design and conduct of the study, in the collection, analysis, and interpretation of the data and in
the preparation, review, or approval of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
¤ Current address: The Centre for Cellular and Molecular Biology, Hyderabad, India
Porphyromonas gingivalis is a gram negative obligate anaerobe that
has a major etiological roleinhuman periodontitis. The bacterium is
found with high frequency in persons with periodontitis where it
participates in the initiation and establishment of chronic, infectious
biofilms [1,2]. These biofilms facilitate the long term survival of P.
gingivalis and induces an inflammatory reaction that is responsible for
the destruction of the hard and softtissuesupporting structures of the
teeth. In addition P. gingivalis can invade and persist in the cells of the
gingival tissue. It also can escape the oral cavity and has been
found in atherosclerotic plaque where it may have a role in the
pathophysiology of cardiovascular disease .
P. gingivalis produces a number of chaperones as essential tools in
normal cellular processes and in response to environmental stresses
. In addition, the role of chaperones, like the P. gingivalis HSP90
homologue HtpG, in immune response dynamics has become an
area of intense investigation because immunomodulation by
chaperones has been demonstrated [6,7]. HtpG, like most
chaperones tested , induces a strong humoral response that
PLoS ONE | www.plosone.org1 April 2008 | Volume 3 | Issue 4 | e1984
may have consequences in the pathogenesis of periodontitis .
These functions are important in the establishment and perpet-
uation of chronic inflammatory diseases.
Recent studies have shown that antibodies to periodontal
disease associated pathogens may have protective effects although
the exact mechanism is still unclear. Rams et al  demonstrated
that serum levels of IgG antibodies against A. actinomycetemcomitans
or P. gingivalis in periodontitis-stable patients were higher than
those in patients with active periodontitis suggesting that these
antibodies have a protective effect against periodontal infections.
Yamasaki et al have shown that antibodies to P. gingivalis Hsp60
homologue increases with successful treatment . We have
recently described experiments that indicate that antibodies to
HtpG may mitigate some of the induction of inflammatory
chemokines through TLR4 and CD91 , a receptor expressed
in human atherosclerotic lesions . Taken together, these
findings suggest a role for antibodies to P. gingivalis chaperones in
both periodontal and cardiovascular disease.
The possible protective role of antibodies to chaperones in
periodontitis is controversial. It has been suggested that these
antibodies simply reflected the high level of homology between
human and bacterial proteins, a hallmark of these evolutionarily
conserved molecules . Other results suggest that the conserved
nature of the chaperones might lead to autoimmune phenomenon
due to ‘‘immune mimicry’’. Earlier results from this laboratory
suggested that high levels of anti-Hsp90 antibodies could have
protective qualities . However, that study utilized a group of
individuals with minimal periodontal disease. Here we describe
findings from a study of periodontitis subjects with more extensive
disease that are similar to those reported earlier and support the
notion of a protective role for anti-HtpG antibodies in untreated
subjects. In addition, these studies suggest that the levels of these
antibodies also may predict the response to treatment. These
results may also reflect an immune control mechanism in chronic
infections unrelated to direct binding of bacteria.
Colonization of plaque by P. gingivalis and F. nucleatum
Initially we reported  that in gingivitis subjects probing depth
and the levels of P. gingivalis were both associated with serum anti-
Hsp90 antibody levels. Subjects from the gingivitis group were
tested for the presence of P. gingivalis by slot immunoblot; 82 % of
the subjects were positive for P. gingivalis [15,16]. Genomic DNA of
pooled plaque samples from all teeth for each subject in the CP
and AP groups were tested; 71% of the healthy subjects and 86%
of individuals with periodontitis were positive for P. gingivalis 16S
rRNA genes by PCR. The low end sensitivity of the assay was
between 10 and 100 bacterial cells per sample. The mean
percentage of P. gingivalis was 0.11 and 0.98 (p,0.001, t-test) in the
controls and periodontitis subjects respectively. Both CP groups
were 96% positive for F. nucleatum; the mean percentage of F.
nucleatum was 2.77 and 3.78 (p=0.07; t-test) in the controls and
periodontitis subjects respectively (Table 1). Plaque samples from
the normal controls for the AP group were not available for F.
nucleatum testing and as reported previously 17% of the AP subjects
were positive for P. gingivalis 16S rRNA .
Antibodies to P. gingivalis and F. nucleatum total cell
The level of antibodies to total cell lysates of P. gingivalis and F.
nucleatum was determined as a measure of the long term
colonization experience that each subject had with these bacteria.
These two bacteria have different roles in periodontitis, P. gingivalis
as a pathogen while F. nucleatum is an opportunistic commensal in
the subgingival biofilm. Anti-P. gingivalis antibodies were detected
in 38 of 45 samples obtained from gingivitis subjects. All CP and
AP subjects had levels of anti-P. gingivalis and anti-F. nucleatum more
than 3 standard deviations above the background of the assay. In
the CP group there was a statistically significant elevated levels of
antibodies to P. gingivalis (p=0.01, t-test) and F. nucleatum (p=0.02,
t-test) in the diseased subjects compared to the control group; there
was no difference between levels of antibody to total P. gingivalis in
the AP group (Table 2). Similar results were found in CP subjects
that were colonized by P. gingivalis (Table 3) and CP subjects not
colonized by P. gingivalis (Table 4) at the time the serum samples
were obtained. These results suggest that most subjects in the study
were likely to have been colonized with P. gingivalis prior to their
participation in this study.
Total IgG Antibodies to P. gingivalis HtpG and F.
Levels of IgG-class antibodies that specifically reacted with
recombinant HtpG proteins of P. gingivalis and F. nucleatum were
measured. Anti-HtpG antibody levels were measured in gingivitis
subjects; subjects with more CAL (more tissue loss) had lower
levels of anti-P. gingivalis HtpG.
In the CP subjects we hypothesized that the pattern of anti-
HtpG would be different for the two bacteria. However, there was
a trend, both in CP individuals colonized (Table 3) or un-
colonized (Table 4) by P. gingivalis, to higher antibody levels to
HtpG from both species in healthy subjects than in subjects with
disease even though this trend did not reach statistical significance.
There were substantial and significant differences between the
healthy and diseased AP subjects in all the variables tested. The
three clinical measurements were all significantly higher in the
Table 1. Research Subject Demographics
Group (n) Gingvitis (50)CP-Healthy (49)CP-Diseased (50)AP-Healthy (21) AP-Diseased (24)
% Female 5854 586268
Age Range in years (mean)18-66 (31.6)20-78(43.0)30-77(51.3)18-35 (34.4) 22-40(32.0)
Attachment Loss (mm, mean6SD)1.5660.700.6860.522.2461.070.7260.523.2560.63
Probing Depth (mm, mean6SD)2.6460.54 1.6060.21 2.6060.711.5060.222.8361.32
Bleeding on Probing (%, mean6SD)ND0.2460.120.5760.17 0.1360.060.3360.16
Gingival Index (%, mean6SD )0.260.17 0.1860.19 0.5260.23ND ND
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org2April 2008 | Volume 3 | Issue 4 | e1984
diseased groups (Table 5) while the anti-HtpG levels were
significantly lower in the diseased subjects (p,0.002, ANOVA).
Antibodies to P. gingivalis HtpG peptide p18
Previous studies  suggested that a peptide antigen, p18, of
the HtpG molecule was responsible for eliciting the apparent
protective effect found in the serum of healthy subjects. Therefore
the serum response to a synthetic peptide antigen of this region
was investigated in the CP and AP subjects. We expected that the
levels of antibody to the peptide would be higher in control
subjects compared to periodontitis subjects of both groups and that
elimination of the majority of the epitopes in the HtpG molecules,
most of which are highly conserved between bacterial species,
would clarify the trend observed using the whole HtpG proteins as
ELISA target antigens. As expected, subjects in both the CP
(Table 3 & 4) and AP (Table 5) groups diagnosed with
periodontitis had lower levels of anti-p18 than their respective
control groups. Although the same trend was observed in both
colonized and un-colonized individuals using the HtpG molecule
only the differences in p18 antibody levels in CP subjects colonized
by P. gingivalis reached statistical significance (Table 3 & 4).
Cluster Analysis by Disease Subgroups
There were trends supporting our hypothesis when the results of
the assays were compared between groups as they were recruited
into the studies, basically health or disease status. However,
examination of those groupings showed that there was overlap
between the groups based on clinical measures, especially in the
CP subjects, the most complete study. Therefore the subjects were
regrouped in line with their clinical measures by K-means
clustering, irrespective of recruitment groups. The clusters were
used as a basis for ANOVA analysis of differences in the clinical
measures and antibody levels.
Gingivitis subjects were regrouped into three clusters and
differences in four clinical measures (PD, CAL, GI, PL) and two
antibody levels (anti-P. gingivalis HtpG, anti-P. gingivalis whole cell
lysate) sought. Cluster 1 had the greatest PD and CAL; there were
significant differences between the clusters for PD and CAL.
Antibody levels against P. gingivalis HtpG were lowest in the
subjects with the highest disease measures, similar to earlier
findings  for human Hsp90, but did not reach statistical
significance (Table 6).
CP Subjects were regrouped into 4 clusters and differences
sought in four clinical measures (PD, CAL, BOP, GI) and three
antibody levels (anti-P. gingivalis HtpG, anti-Fn HtpG and anti-Pg
p18). There were significant differences between the groups in all
clinical measures (Table 7). Cluster 2 had the highest levels of all
clinical measures and the lowest levels of anti-P. gingivalis HtpG
and anti-P. gingivalis HtpG p18, the latter being significant
AP subjects were regrouped into 3 clusters and differences
sought in three clinical measures (PD, CAL, BOP) and three
antibody levels (anti-Pg whole bacteria, anti-P. gingivalis HtpG and
anti-Pg P18). There were significant differences in all the clinical
measures between groups and in anti-P. gingivalis HtpG and anti-P.
gingivalis HtpG p18, but not anti-Pg whole bacteria (Table 8).
Cluster 2 had the highest levels of clinical measures and lowest
levels of anti-P. gingivalis HtpG p18. Both clusters 1 and 2 had
substantially lower levels of anti-P. gingivalis HtpG compared to
Table 2. Antibody responses to P. gingivalis and F. nucleatum in subjects.
Group (n) Gingivitis (50)CP-Healthy (49)CP-Diseased (50) AP-Healthy (21)AP-Diseased (24)
% positive for P. gingivalis 8271 86 ND17
% positive for F. nucleatum ND96 96 NDND
Antibody to total P. gingivalis (Log mean ELISA Units6S.D.) 4.3960.173.6860.23 3.9260.26 4.4760 .07 4.3960.15
Antibody to total F. nucleatum (Log mean ELISA Units6S.D.) ND3.9560.624.8660.81 NDND
Table 3. Serum antibody (IgGc) levels to bacterial antigens in
CP subjects colonized by P. gingivalis.
Valid N (Health/
Anti-P.gingivalis HtpG p187,138 6,2540.047 35/40
Anti-P.gingivalis W835,65410,0760.002 35/40
Anti-F.nucleatum 22586 27,73513,7010.009 35/41
Table 4. Serum antibody (IgGc) levels to bacterial antigens in
CP subjects not colonized by P. gingivalis.
Valid N (Health/
Anti-P.gingivalis HtpG 9,224 8,594NS13/6
Anti-P.gingivalis HtpG p18 6,8126,477NS13/7
Anti-F.nucleatum 22586 17,9178,4730.009 13/7
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org3 April 2008 | Volume 3 | Issue 4 | e1984
Correlation of Anti-P. gingivalis HtpG antibodies to
clinical measures in Chronic and Aggressive Periodontitis
The supposition that antibodies to P. gingivalis HtpG may be
protective is based on the hypothesis that high levels of antibodies
should be found in subjects exhibiting healthier clinical signs and
lower levels in subjects exhibiting more periodontitis related
damage. Clinical measurements from each site in all the subjects
were obtained and the average value determined for each subject.
Relationships between those measurements and the levels of P.
gingivalis and F. nucleatum anti-HtpG were sought using Pearson’s R
to compare baseline antibody levels in both AP and CP groups
with the clinical signs obtained at the baseline were compared.
When anti-P. gingivalis HtpG levels from CP subjects were
analyzed with four clinical measurements in each of the same
patients there was a trend for pocket depth and attachment loss
measurements to inversely correlate with antibody levels, but the
correlations were small and not statistically significant. Similar
results were found with anti-F. nucleatum HtpG values. However, in
the AP subjects there were substantial negative associations three
Table 5. Serum antibody (IgGc) levels to bacterial antigens in AP subjects.
Health (mean6S.D) Disease (mean6S.D.)p-value (ANOVA)Valid N
PD (mm) 1.5160.223.3160.67
AL (mm) 0.71960.52 2.8561.45
BOP (%) 0.12560.070.2960.17
Log anti-HtpG 4.2560.04 3.9660.13
Log anti-W83 cells 4.4760.07 4.3960.150.032 45
Table 6. Analysis of clustered subjects in the Gingivitis group.
Cluster 1 Cluster 2 Cluster 3
Mild perio. n=22 Gingivitis n=10 Healthy n=18
AL 2.260.61.060.0 1.1160.3
Pl 45.565.98.066.3 27.864.6NS
Anti-P. gingivalis HtpG982564720 149736101781267666928 NS
Anti-P. gingivalis W83 310066898 130161777488611008NS
Subjects were clustered by K-means clustering and differences sought in means (6S.D.) of variables by ANOVA. PD and AL measured in mm; GI is the % sites with
gingival redness; PI values are % sites with accumulated plaque. Antibody values are expressed in ELISA units.
Table 7. Cluster analysis of CP subjects.
Cluster 1Cluster 2 Cluster 3Cluster 4
Moderate CP n=31Severe CP n=10Gingivitis n=29 Healthy n=27
BOP56.8616.2 68.4616.6 188.8.131.524.6614.6
Anti-P. gingivalis HtpG 10531647917877639069311638331035864443 NS
Anti-F. nucleatum HtpG200536806185306934618231657522074866540 NS
Anti-P. gingivalis HtpG p186485619204793621207171617826957610680.006
Subjects were clustered by K-means clustering and differences sought in means (6S.D.) of variables by ANOVA. PD and AL measured in mm; GI is the % sites with
gingival redness. Antibody values are expressed in ELISA units.
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org4April 2008 | Volume 3 | Issue 4 | e1984
clinical measurements and anti-HtpG levels, all of which were
statistically significant (Table 9). Similar experiments using HtpG
cloned from F. nucleatum gave no significant correlations to the
clinical signs (data not shown).
Correlation of Anti-P. gingivalis HtpG p18 antibodies to
clinical measures in Chronic and Aggressive Periodontitis
We then sought relationships between the same measurements
and the levels of P. gingivalis anti-HtpG p18 by comparing antibody
levels in both AP and CP groups with the clinical measurement
using Pearson’s R analysis. Sera from the CP group and controls
showed that levels of anti-p18 were significantly (p=0.047)
inversely correlated to CAL but not correlated to the other
indices. In the AP group antibodies to p18 were significantly
inversely correlated to PD (p=0.008), CAL (p=0.018) and BOP
(p=0.046), Table 10. Lastly, associations were sought between
changes in clinical measurements 6 months after SRP treatment
and the anti-p18 levels 6 months before treatment to determine if
they might predict treatment success or failure. There was a
significant positive correlation (p=0.05) between the pre-treat-
ment levels of anti-HtpG p18 and reduction in CAL.
Chaperones are simultaneously highly conserved and immuno-
dominant antigens which may have important roles in the
pathogenesis of numerous human diseases. Although a subject of
intense investigation, the role of the immune response to
chaperones in human disease is currently not well understood.
Periodontal diseases present a unique opportunity to examine the
Table 8. Cluster analysis of AP subjects.
Cluster 1Cluster 2Cluster 3p-value (ANOVA)
Localized AP n=11Generalized AP n=12Healthy n=21
AL 2.261.23.461.5 0.760.5
BOP21.4611.2 37.9618.1 12.565.7
Anti-P. gingivalis HtpG p18 1018262437837162114 12390653180.0283
Anti-P. gingivalis HtpG8853624931016563479 1774261349
Anti-P. gingivalis W83 2689969509 26040612191 3004164977NS
Subjects were clustered by K-means clustering and differences sought in means (6S.D.) of variables by ANOVA. PD and AL measured in mm; BOP is the % sites bleeding
on probing. Antibody values are expressed in ELISA units. NS–not significant.
Table 9. Correlation of Anti-P. gingivalis HtpG antibody levels with clinical measurements.
A. PretreatmentB. Post-treatment
Anti-P. gingivalis HtpGAnti-F. nucleatum HtpGAnti-P. gingivalis HtpG Anti-F. nucleatum HtpG
p-value Pearson’s Rp-valuePearson’s Rp-valuePearson’s Rp-value
20.019NS 0.3610.002 0.076 NS
20.113 NS0.2730.020 0.144NS
BOP0.144NS 0.057NS 0.3810.0010.113NS
GI 0.093NS0.001 NS0.2630.026 0.140NS
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org5 April 2008 | Volume 3 | Issue 4 | e1984
immune response to chaperones in a distinctive and readily
accessible human environment. A few reports of antibody levels to
P. gingivalis chaperones (particularly GroEL) have been published
but this is the first comprehensive analysis of anti-P. gingivalis HtpG
in multiple periodontal disease states. Results reported here and
earlier [9,12] suggest that anti-P. gingivalis HtpG antibodies predict
health in patients susceptible to periodontal disease and are
protective in the untreated periodontal disease patient. These
results may also reflect a previously uncharacterized immune
control mechanism unrelated to direct binding of bacteria.
Further, these antibodies may augment periodontitis treatment
and P. gingivalis HtpG might be an attractive vaccine candidate.
These results appear to be unique to P. gingivalis as they are
directed at a segment of HtpG unique to P. gingivalis; parallel
experiments with the HtpG homologue of F. nucleatum did not
manifest these same qualities. In addition, there is reason to
believe that these results may be extended to other chronic
Earlier studies of a population of subjects with minimal
periodontal disease suggested that antibodies to human Hsp90
related chaperones might have a protective effect. This report
expands on that notion by examining the response of periodontitis
subjects to the Hsp90 homologue in P. gingivalis, HtpG. In
gingivitis subjects as a whole it was found that there was a
distinctive, but not statistically significant, trend in support of the
hypothesis. Interestingly, when anti-P. gingivalis HtpG levels and
clinical measures are compared on an individual basis there is a
discernable correlation with CAL. At the opposite end of the
periodontitis disease spectrum, the AP subjects, there is consider-
able support for the hypothesis. At the group level, there are
substantial and significant differences between both anti-P.
gingivalis HtpG and anti-P. gingivalis HtpG p18 between clusters
based on clinical measures, but no differences in the response to
the whole bacterium. On an individual basis, both anti-P. gingivalis
HtpG and anti-P. gingivalis HtpG p18 are inversely and
significantly correlated to the clinical measures.
Support of the hypothesis was somewhat tenuous in the initial
examination of the CP subjects. Those recruited as either healthy
or with CP and colonized by P. gingivalis possessed a humoral
response to Pg HtpG p18 higher in subject groups with less
periodontal destruction and inversely related to the level of
attachment loss (p,0.05). A similar relationship was not found in
un-colonized individuals. This was the opposite of antibody levels
to the whole bacterium which were significantly higher is subjects
with disease than controls (p#0.01), a finding reported by many
other laboratories. Similar trends were found when we examined
the humoral response to HtpG from F. nucleatum but there was no
relationship to disease status. In addition, correlations between the
same antibodies and clinical measures when considered in the
context of individual subjects were not found, except for a
correlation between ant-P. gingivalis HtpG p18 and CAL
(R=20.205, p,0.05). This dilemma was resolved when the
subjects were clustered by clinical measures. In a cluster of 10
subjects with the most severe disease there was a substantial and
significantly lower level of anti-P. gingivalis HtpG and anti-P.
gingivalis HtpG p18 compared to the other clusters in the CP
subjects. As might be expected, the clinical measures of these
subjects resemble those of the AP subjects. In summary, there is a
trend for both anti-P. gingivalis HtpG and anti-P. gingivalis HtpG
p18 to be lower in subjects with the more serious disease in each of
the 3 periodontal disease groups. Neither trend reaches statistical
significance in the gingivitis group. In the CP group anti-P.
gingivalis HtpG p18 is significantly lower in the most diseased
cluster; anti-P. gingivalis HtpG trends in the same direction but
does not reach statistical significance. In the AP group both anti-P.
gingivalis HtpG and anti-P. gingivalis HtpG p18 are significantly
lower in the diseased clusters of subjects. Our initial observations
(16) were probably due to the response of the most diseased
individuals in that group of subjects. However, we believe that the
current results are not fortuitous because 1) the relationship
between the antibody levels and disease is evident not only in the
original group but in groups with more serious disease in an almost
dose dependent manner; 2) there is a relationship between these
antibody levels and response to periodontal treatment; 3) there is a
link through these antibodies to cellular receptors involved in
antigen recognition and inflammation; 4) similar results have been
noted in other chronic bacterial infections. In addition, there is no
evidence that the lower levels of anti-P. gingivalis HtpG are due to
adsorption by high circulating levels of either human Hsp90.
There was no significant correlation between individual serum
levels of human Hsp90 protein and anti-P. gingivalis HtpG, anti-F.
nucleatum HtpG or anti-P. gingivalis HtpG p18 (Table S1). In fact,
there was a trend to higher levels of anti-P. gingivalis HtpG in
subjects with higher levels of human Hsp90, the opposite of what
be expected if such absorption was taking place. There was also no
significant difference between levels of human Hsp90 in healthy
subjects with high levels of anti-P. gingivalis HtpG antibodies and
severe periodontitis subjects with low levels of anti-P. gingivalis
HtpG (Table S2). Lastly, while it is possible that a subset of anti-P.
gingivalis HtpG antibodies might bind other HSP90 homologues,
including human Hsp90, the P. gingivalis HtpG p18 epitope is
unique and antibodies to that peptide could not be adsorbed by
other HSP90 family proteins (see discussion below).
Subjects with CP who were given periodontal treatment
responded more effectively to that treatment the higher their
original pretreatment levels of anti-P. gingivalis HtpG. There was
no similar effect with antibodies to F. nucleatum HtpG. Notably, the
serum samples, which are time-averaged values, correlate best
Table 10. Correlation of Anti-P. gingivalis HtpG p18 antibody
levels with clinical measurements.
Anti-P. gingivalis HtpG p18Anti-P. gingivalis HtpG p18
Pearson’s Rp-valuePearson’s R p-value
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org6 April 2008 | Volume 3 | Issue 4 | e1984
with clinical indices related to tissue destruction and less with
indices related to inflammation. This suggests that these anti-
bodies are found in subjects predisposed to a better treatment
outcome or they may somehow facilitate healing after treatment.
Interestingly, drugs that inhibit Hsp90 have been shown to
prolong survival, attenuate inflammation, and reduce lung tissue
injury in murine sepsis .
These findings are also worthy of note because despite high
homology among chaperones in diverse organisms, the P. gingivalis
HtpG is dissimilar from A. actinomycetemcomitans and F. nucleatum
HtpG proteins: the C-terminus contains an extra inserted
sequence (Table 11. bold in alignment). This insert appears to
be exclusive to the Cytophaga-Flavobacterium-Bacteroides group
HtpG proteins sequenced to date, and contains both a portion
unique to each species (p18, underlined in alignment) and a
conserved portion (p19). In contrast, the DnaK and GroEL of P.
gingivalis, A. actinomycetemcomitans, and F. nucleatum align closely and
display high homology. Examination of an alignment of
Bacertoidacae HtpG molecules (Table 12) shows the insert can
be roughly divided into 2 sections of about 30 amino acids each.
The first-which contains p18-has relatively little sequence
homology to the other species, only 6 of 36 (17%) amino acids
are exact matches to the other Bacteroidacae HtpGs. The adjacent
segment, p19, has 13 of 25 (52%) amino acids that are identical,
similar to that between whole HtpG molecules from all 5 species
(60-63%). BLAST analysis of p18 against the entire non-
redundant protein database at GeneBank results in no hits except
P. gingivalis HtpG. Similar searches using p19 produce significant
hits with species in each of the Bacteroidacae genera (B. fragilis
group; Non-B. fragilis group, Prevotella, Porphyromonas, Taner-
ella and ‘‘other’’). Taken together we have concluded it is
reasonable to assume that p18 is or contains an epitope that is
unique to P. gingivalis HtpG and are currently investigating that
The mechanism that may connect anti-HtpG antibodies and
progression of periodontitis is not known, but we speculate that it
may well be that these antibodies block an interaction between
HtpG and cellular receptors on macrophage . Interactions
between HtpG and the TLR4 and CD91 receptors induce the
chemokine CXCL8, a chemoattractant for phagocytic cells in
periodontitis [20,21] and thus perpetuating the uncontrolled
inflammation characteristic of periodontitis. Additionally, interac-
tions between TLR4 and CD91 and HtpG induce inflammatory
cytokines, including TNFa, as part of the innate immune response.
Serum antibodies in 8 subjects with titers to P. gingivalis HtpG were
found to reduce CXCL8 production in human monocytes in a
dose dependent manner while serum antibodies from subjects
without anti-HtpG activity did not .
Table 11. Alignment of oral bacterial species HtpG amino acid sequences.
A. actinomycetemcom. HtpG
A. actinomycetemcom. HtpG
A. actinomycetemcom. HtpG
Alignment of oral bacterial species HtpG amino acid sequences in the region of the 65 amino acid ‘‘Bacteroidacae insert’’ (Bold, underlined characters). The insert has
not been described outside the Cytophaga, Flavobacteria and Bacteroides (CFB) group. Italic, bold–p18.
Table 12. Alignment of Bacteroidiacae HtpG C-terminal inserts
* . :** ::*. *::. : . . :::
***: ******** ..:*.* ** .*::***:
Alignment of 65 amino insert found in most CFB group bacteria, but not other bacterial groups. Bacteroidiacae HtpG C-terminal inserts shows p18 is much less
conserved than p19. Underline/bold–p18 sequence. Dashed underline–p19 sequence. Amino acid identities across all species-*.
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org7 April 2008 | Volume 3 | Issue 4 | e1984
The results reported here are similar to those of other chronic
infections including Heliobacter pylori. World wide as much as 75%
of the population has evidence of H. pylori infection but only
subsets of these individuals manifest the peptic ulcers and stomach
cancers associated with those chronic infections. The H. pylori
chaperone hsp60 induces CXCL8 production in monocytic cells
 and humoral immune response to a peptide epitope (pH9) of
hsp60 is unique and seems to be associated with protection against
H. pylori infection . Intriguingly, when subjects with mucosa-
associated lymphoid tissue lymphoma were examined for anti-
bodies to the H. pylori hsp60 chaperone it was found that pre-
treatment titers in patients whose tumors regressed after treatment
were significantly higher than in patients whose tumors did not
regress . Low levels of antibodies to chaperones in disease
subjects compared to controls has also been reported in
inflammatory bowel disease , fungal infections  and other
chronic bacterial infections .
The unique restriction of the P. gingivalis HtpG p18 peptide to
the most important pathogen in periodontitis may have applica-
tions in vaccine and diagnostic arenas. The use of chaperones as
vaccine candidates has been suggested by a number of
investigators. Antibody to the Hsp90 homologue of C. albicans
(MycograbH, Novartis) has been shown to be effective in the
treatment of disseminated fungal infections  in combination
with amphotericin B. Other chaperones have been suggested as
vaccine targets for diverse diseases in humans, cattle  and fish
. P. gingivalis GroEL has been used as a vaccine in a rat model
of periodontitis that resulted in prevention of attachment loss ,
similar to that described here in humans. However, the authors
cautioned that extensive homology between human and microbial
chaperones may require use of peptides that do not induce cross-
reactive antibodies with human hsp60 molecules. We believe this
is the case with p18, which appears unique to P. gingivalis even to
the exclusion of other Bacteroides species. The data presented
here also suggests that serum antibodies to p18 may be useful in
diagnosing periodontitis patients with extensive treatment require-
ments, identification of which would have substantial economic
and epidemiological impact on the practice of periodontology.
Materials and Methods
All work with human subjects was approved by the University of
Michigan Institutional Review Board or the Queen’s University,
Belfast Ethics Committee, respectively. Each subject gave written
individual informed consent and was advised that withdrawal from
the study was available at their discretion at any time. The clinical
condition of each subject was determined by examination of five
clinical measurements: Probing pocket depth (PD) was determined
to the nearest mm at six sites around each tooth and then averaged
for all sites in each subject . Bleeding on probing (BOP), was
reported as a dichotomous measure and recorded as a percentage
of the sites . Clinical attachment loss (CAL) was determined at
the same sites by measuring the distance between the cemento-
enamel junction and the bottom of each pocket to the nearest mm
and averaged . The gingival index (GI) value represents the
percentage of sites in each patient exhibiting redness associated
with inflammation. The plaque index value (PI) represents the
percentage of sites found to have plaque biofilm accumulated at
the gingival margin.
This group was examined from archived samples described in a
previous report and was included here to address the potential
differences between response to P. gingivalis HtpG and human
Hsp90. This retrospective study focused on a population of
individuals living in southwestern Michigan [15,16] with gingivitis,
a form of periodontal disease with no or minimal tissue destruction
that is reversible with active oral hygiene. Subjects were recruited
into the original study based on their membership in the rural
community, not their oral health status, and all essentially
presented with gingivitis or mild periodontitis. The clinical data
and archived serum samples collected from 50 subjects were used.
The average age was 31.6 years (range: 18–66) and 58% were
female. Other characteristics of these subjects are found in Table 1.
Chronic Periodontitis Subjects
Subjects were recruited at the Michigan Center for Oral Health
Research. Subject inclusion was based on: possession of at least 20
teeth, no periodontal treatment or antibiotic-related therapy for
medical or dental reasons for 3 months before study inclusion, no
history of long-term treatment with medications known to affect
periodontal status, and no history of metabolic bone diseases.
Healthy control subjects (n=49) were recruited who had less than
3 mm of CAL, no pocket depth (PD)s greater than 4 mm, no
radiographic bone loss, and less than 20 sites with bleeding on
probing. Chronic periodontitis (CP) subjects (n=50) exhibited at
least 4 sites with evidence of radiographic bone loss, mean CAL
.3mm, PD .4 mm and bleeding on probing (Table 1).
Aggressive Periodontitis Subjects
Studies were done with serum samples from an ongoing study of
Aggressive Periodontitis (AP) subjects (n=24) and age matched
control subjects (n=21) living in Northern Ireland. AP subjects
included in this investigation were 30.364.0 years of age at the
time of clinical and radiographic examination, diagnosed with
severe periodontitis, and had a minimum of 4 sites with a probing
depth of at least 5 mm and CAL loss of at least 2 mm. Age
matched controls without periodontitis were recruited from
regular attendees at the Queen’s University School of Dentistry.
Colonization of AP subjects by P. gingivalis was determined by PCR
using primers specific for the P. gingivalis 16S rRNA gene .
Bacterial strains and culture conditions
Porphyromonas gingivalis (ATCC 33277) and Fusobacterium nucleatum
(ATCC 25586) were obtained from the American Type Culture
Collection. Porphyromonas gingivalis strain W83 was a gift from Dr.
Donald Clewell, University of Michigan School of Dentistry. The
bacteria were maintained by weekly transfer in an anaerobe
chamber (Coy Manufacturing, Grass Lake, MI) at 37uC on PRAS
Brucella agar plates (Anaerobe Systems, Morgan Hill, CA) in a 5%
hydrogen, 10% carbon dioxide, 85% nitrogen atmosphere. Broth
cultures were grown in a mixture of 50% Brain Heart Infusion
Broth, 50% Trypticase Soy Broth and 5 gram/L Yeast Extract
supplemented with 0.01 gm/L Sodium Bisulfite, 5 mg/L hemin
and 5ug/L Vitamin K.
Cloning and purification of P. gingivalis and F. nucleatum
The full length sequence of P. gingivalis HtpG (GenBank
ascension number AF176245) and F. nucleatum HtpG (GenBank
ascension number EDK88176) were obtained from the National
Center for Biotechnology Information (www,ncbi.nlm.nih.gov).
PCR primers were designed to produce full-length products that
were subsequently inserted into pCRHT7 TOPOH cloning vector
following the manufacturer’s instructions (Invitrogen, Carlsbad,
CA.). Clones of One ShotH Chemically Competent E. coli
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org8April 2008 | Volume 3 | Issue 4 | e1984
(Invitrogen) transformedwith thevector were ampicillin selected and
then screened by PCR. Inserts that produced amplicons of the
correct size were sequenced to verify the full-length insert
(Biomedical Research Core Facilities, University of Michigan, Ann
Arbor, MI). Plasmids with in-frame inserts were used to transform
TOP10F BL21(DE3)pLysS E.coli (Invitrogen) cells that were
thiogalactopyranoside for 4 hours to produce a fusion protein with
6 consecutive histidine (6xhis) residues preceding the N-terminal of
the cloned proteins. The protein was purified to .95%, as
chromatography (Ni-NTATMAgarose, Qiagen, Valencia, CA).
Preparation of HtpG peptide (p18).
KYA) was synthesized on an Applied Biosystem 433A peptide
synthesizer at the University of Michigan department of
Chemistry using Fmoc amino acids from AnaSpec (San Jose,
CA). The crude peptide was purified by HPLC and the purity
(,98%) confirmed by MALDI-TOF analysis.
A 36 amino aid
All assays were carried out in 384-well microtiter plates
(NUNCTMblack MaxiSorp, Rochester, NY) using 4-methyumbli-
ferone phosphate as a substrate for alkaline phosphatase-labeled
tracers. Antigens were coated onto the plates in 25 ml/well volumes
in sodium carbonate/bicarbonate buffer (0.05M, pH 9.5) and
incubated overnight at 4uC. Plates were washed 3 times with
0.02 M phosphate buffered saline (PBS, pH 7.4) and PBS with 1%
bovine serum albumin (PBS-BSA) added to block unoccupied
protein binding sites (100 mL/well). After an additional hour of
incubation at room temperature plates were washed with PBS with
0.125% NP40 three times and human serum (or control rabbit
serum) diluted 1:100 in PBS-BSA (25 uL/well) added to the plates in
triplicate and incubated overnight at 4uC. Plates were then washed 3
times with PBS with 0.125% NP40 and detection reagents added as
described below. Blocked wells not coated with antigen were used as
negative controls for each individual serum sample. Data is
expressed as net relative florescent units (RFU) calculated by
subtracting the average of 3 control wells from the average of 3
antigen coated wells and were repeated at least 3 times each.
Antibodies to P. gingivalis Whole Cell Lysate
P. gingivalis cells (strain W83) were grown to mid-log phase,
washed by centrifugation at 10,0006g 3 times with sterile PBS and
resuspended at an OD600of 1.0 in water. The cells were sonicated
for 2 minutes on ice three times and centrifuged at 10,0006g for
20 minutes. Total protein was determined and the clarified lysate
diluted to 10 mg/mL for coating. Serum IgG binding was
determined as described above using alkaline phophatase-labeled
anti-human IgG(c) antibodies and 4-methylubelliferyl phosphate
(1 mg/mL in 0.2 M TRIS, pH 9.5) (Sigma, St. Louis).
Antibodies to P. gingivalis HtpG and F. nucleatum HtpG
Recombinant HtpG proteins were dissolved in carbonate buffer
to 1 mg/mL for plate coating. Total IgG(c) binding was
determined using the same second antibodies as described for
the whole cell lysate assay above.
Antibodies to P. gingivalis HtpG peptide 18
The peptide was dissolved in carbonate buffer to 10 mg/mL for
plate coating. Total serum IgG(c) binding was determined using
the same second antibodies as above.
Detection of P. gingivalis and F. nucleatum colonization in
The detection of P. gingivalis in pooled plaque samples from the
gingivitis subjects was done using a slot immunoblotting method as
described previously . Colonization of plaque samples collected
simultaneously with the serum in the CP groups were evaluated by
real-time PCR as described  using primers specificfor P. gingivalis
verse 59-AAACTGTTAGCAACTACCGATGTGG-39) and F.
nucleatum (forward: 59-AAATATGTTGAATTCTGGAAAGAGT-
TTG-39; reverse: 59-TGAACTCCAGCTTTTATACTTCTAC-
CAA-39). Percentage of the total flora for each species was calculated
by dividing the number of target organisms by the total number of
bacteria as determined by realtime PCR using 16S rRNA primers
that reacted with all bacterial species (forward: 59-CCATGAA-
GTCGGAATCGCTAG-39; reverse: 59-GCTTGACGGGCGG-
TGT-39). The presence of P. gingivalis in plaque from the AP
subjects was determined by PCR as described earlier  using the
Statistical Analysis of anti-HtpG levels
Statistical computations were done using STATISTICATMv.
6.0 (StatSoft, Omaha, NE). Comparison of two means was
performed using t-tests. The relationships between specific indices
of periodontal disease, antibody levels, and colonization were
assessed by ANOVA. Fisher’s method for multiple comparisons
was used for group comparisons. Log transformation of data was
performed where appropriate. Results with p-values of # 0.05
were considered significant.
serum levels in CP subjects.
Found at: doi:10.1371/journal.pone.0001984.s001 (0.03 MB DOC)
Antibody to P. gingivalis HtpG and human Hsp90
levels in 95 CP patients and healthy controls.
Found at: doi:10.1371/journal.pone.0001984.s002 (0.03 MB DOC)
Correlation of human Hsp90 levels to Anti-HtpG
The authors wish to acknowledge the technical assistance of Florence Y-P
An in cloning the F. nucleatum HtpG. The authors thank Dr. Vincent L.
Pecoraro (University of Michigan Department of Chemistry) for use of the
ABI 433A Peptide Synthesizer.
Conceived and designed the experiments: DL CS. Performed the
experiments: PS CS. Analyzed the data: WG CS. Contributed reagents/
materials/analysis tools: WG VD DS JK WC BM DL CS. Wrote the
1. Socransky SS, Haffajee AD (1997) The nature of periodontal diseases.
AnnPeriodontol 2: 3–10.
2. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent RL Jr (1998) Microbial
complexes in subgingival plaque. JClinPeriodontol 25: 134–144.
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org9 April 2008 | Volume 3 | Issue 4 | e1984
3. Progulske-Fox A, Kozarov E, Dorn B, Dunn W Jr, Burks J, et al. (1999)
Porphyromonas gingivalis virulence factors and invasion of cells of the
cardiovascular system. J Periodontal Res 34: 393–399.
4. Kozarov E, Sweier D, Shelburne C, Progulske-Fox A, Lopatin D (2006)
Detection of bacterial DNA in atheromatous plaques by quantitative PCR.
Microbes Infect 8: 687–693.
5. Goulhen F, Grenier D, Mayrand D (2003) Oral microbial heat-shock proteins and
their potential contributions to infections. Crit Rev Oral Biol Med 14: 399–412.
6. Prohaszka Z, Fust G (2004) Immunological aspects of heat-shock proteins-the
optimum stress of life. Mol Immunol 41: 29–44.
7. Shamaei-Tousi A, D’Aiuto F, Nibali L, Steptoe A, Coates AR, et al. (2007)
Differential regulation of circulating levels of molecular chaperones in patients
undergoing treatment for periodontal disease. PLoS ONE 2: e1198.
8. Young D, Lathigra R, Hendrix R, Sweetser D, Young RA (1988) Stress proteins
are immune targets in leprosy and tuberculosis. ProcNatlAcadSciUSA 85:
9. Lopatin DE, Shelburne CE, Van Poperin N, Kowalski CJ, Bagramian RA
(1999) Humoral immunity to stress proteins and periodontal disease.
J Periodontol 70: 1185–1193.
10. Rams TE, Listgarten MA, Slots J (2006) Actinobacillus actinomycetemcomitans
and Porphyromonas gingivalis subgingival presence, species-specific serum
immunoglobulin G antibody levels, and periodontitis disease recurrence.
J Periodontal Res 41: 228–234.
11. Yamazaki K, Ueki-Maruayama K, Honda T, Nakajima T, Seymour GJ (2004)
Effect of periodontal treatment on the serum antibody levels to heat shock
proteins. Clin Exp Immunol 135: 478–482.
12. Shelburne CE, Coopamah MD, Sweier DG, An FY, Lopatin DE (2007) HtpG,
the Porphyromonas gingivalis HSP-90 homologue, induces the chemokine
CXCL8 in human monocytic and microvascular vein endothelial cells. Cell
Microbiol 9: 1611–1619.
13. Luoma J, Hiltunen T, Sarkioja T, Moestrup SK, Gliemann J, et al. (1994)
Expression of alpha 2-macroglobulin receptor/low density lipoprotein receptor-
related protein and scavenger receptor in human atherosclerotic lesions. J Clin
Invest 93: 2014–2021.
14. Kinane DF, Mooney J, Ebersole JL (1999) Humoral immune response to
Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis in
periodontal disease. Periodontol 2000 20: 289–340.
15. Bagramian RA, Farghaly MM, Lopatin D, Sowers M, Syed SA, et al. (1993)
Periodontal disease in an Amish population. JClinPeriodontol 20: 269–272.
16. Bagramian RA, Farghaly MM, Lopatin D, Sowers M, Syed SA, et al. (1994) A
comparison of periodontal disease among rural Amish and non-Amish adults.
JClinPeriodontol 21: 386–390.
17. Mullally BH, Dace B, Shelburne CE, Wolff LF, Coulter WA (2000) Prevalence
of periodontal pathogens in localized and generalized forms of early-onset
periodontitis. JPeriodontal Res 35: 232–241.
18. Sweier DG, Shelburne CE, Cameron J, Lopatin DE (2004) Localizing antibody-
defined immunoreactivity in Porphyromonas gingivalis HtpG recognized by
human serum utilizing selective protein expression. J Immunol Methods 285:
19. Chatterjee A, Dimitropoulou C, Drakopanayiotakis F, Antonova G, Snead C, et
al. (2007) Heat shock protein 90 inhibitors prolong survival, attenuate
inflammation, and reduce lung injury in murine sepsis. Am J Respir Crit Care
Med 176: 667–675.
20. Hosokawa Y, Hosokawa I, Ozaki K, Nakae H, Matsuo T (2006) Proin-
flammatory effects of tumour necrosis factor-like weak inducer of apoptosis
(TWEAK) on human gingival fibroblasts. Clin Exp Immunol 146: 540–549.
21. Okada H, Murakami S (1998) Cytokine expression in periodontal health and
disease. Crit Rev Oral Biol Med 9: 248–266.
22. Lin SN, Ayada K, Zhao Y, Yokota K, Takenaka R, et al. (2005) Helicobacter
pylori heat-shock protein 60 induces production of the pro-inflammatory
cytokine IL8 in monocytic cells. J Med Microbiol 54: 225–233.
23. Yamaguchi H, Osaki T, Kai M, Taguchi H, Kamiya S (2000) Immune response
against a cross-reactive epitope on the heat shock protein 60 homologue of
Helicobacter pylori. InfectImmun 68: 3448–3454.
24. Takenaka R, Yokota K, Mizuno M, Okada H, Toyokawa T, et al. (2004) Serum
antibodies to Helicobacter pylori and its heat-shock protein 60 correlate with the
response of gastric mucosa-associated lymphoid tissue lymphoma to eradication
of H. pylori. Helicobacter 9: 194–200.
25. Huszti Z, Bene L, Kovacs A, Fekete B, Fust G, et al. (2004) Low levels of
antibodies against E. coli and mycobacterial 65kDa heat shock proteins in
patients with inflammatory bowel disease. Inflamm Res 53: 551–555.
26. Matthews RC, Burnie JP, Howat D, Rowland T, Walton F (1991) Autoantibody
to heat-shock protein 90 can mediate protection against systemic candidosis.
Immunology 74: 20–24.
27. Zugel U, Kaufmann SH (1999) Role of heat shock proteins in protection from
and pathogenesis of infectious diseases. ClinMicrobiolRev 12: 19–39.
28. Burnie JP, Carter TL, Hodgetts SJ, Matthews RC (2006) Fungal heat-shock
proteins in human disease. FEMS Microbiol Rev 30: 53–88.
29. Koets A, Hoek A, Langelaar M, Overdijk M, Santema W, et al. (2006)
Mycobacterial 70 kD heat-shock protein is an effective subunit vaccine against
bovine paratuberculosis. Vaccine 24: 2550–2559.
30. Sudheesh PS, LaFrentz BR, Call DR, Siems WF, LaPatra SE, et al. (2007)
Identification of potential vaccine target antigens by immunoproteomic analysis
of a virulent and a non-virulent strain of the fish pathogen Flavobacterium
psychrophilum. Dis Aquat Organ 74: 37–47.
31. Lee JY, Yi NN, Kim US, Choi JS, Kim SJ, et al. (2006) Porphyromonas
gingivalis heat shock protein vaccine reduces the alveolar bone loss induced by
multiple periodontopathogenic bacteria. J Periodontal Res 41: 10–14.
32. Loe H, Theilade E, Jensen SB (1965) Experimental gingivitis man. JPeriodontol
33. Offenbacher S (2005) Commentary: clinical implications of periodontal disease
assessments using probing depth and bleeding on probing to measure the status
of the periodontal-biofilm interface. J Int Acad Periodontol 7: 157–161.
34. Wolff LF, Liljemark WF, Pihlstrom BL, Schaffer EM, Aeppli DM, et al. (1988)
Dark-pigmented Bacteroides species in subgingival plaque of adult patients on a
rigorous recall program. J Periodontal Res 23: 170–174.
35. Silness J, Loe H (1966) Periodontal disease in pregnancy. 3. Response to local
treatment. Acta OdontolScand 24: 747–759.
36. Van Poperin N, Lopatin DE (1991) Slot immunoblot assay for detection and
quantitation of periodontal disease-associated microorganisms in dental plaque.
JClinMicrobiol 29: 2554–2558.
37. Shelburne CE, Prabhu A, Gleason RM, Mullally BH, Coulter WA (2000)
Quantitation of Bacteroides forsythus in subgingival plaque comparison of
immunoassay and quantitative polymerase chain reaction. J Microbiol Methods
Anti-HtpG in Periodontitis
PLoS ONE | www.plosone.org 10April 2008 | Volume 3 | Issue 4 | e1984