Antibodies to Borrelia burgdorferi OspA, OspC, OspF, and C6
Antigens as Markers for Early and Late Infection in Dogs
Bettina Wagner,a,bHeather Freer,a,bAlicia Rollins,a,bDavid Garcia-Tapia,cHollis N. Erb,aChristopher Earnhart,dRichard Marconi,dand
Department of Population Medicine and Diagnostic Sciencesaand Animal Health Diagnostic Center,bCollege of Veterinary Medicine, Cornell University, Ithaca, New York,
USA; Veterinary Medicine Research and Development, Pfizer Animal Health, Kalamazoo, Michigan, USAc; and Department of Microbiology and Immunology, Center for
the Study of Biological Complexity, Medical College of Virginia at Virginia Commonwealth University, Richmond, Virginia, USAd
Lyme disease in the United States is caused by Borrelia burgdorferi sensu stricto, which is transmitted to mammals by infected
and antibodies to OspC decline over time. The onset and decline of antibody responses to B. burgdorferi Osp antigens and C6
ria, which is transmitted to mammalian hosts by infected ticks
(Ixodes spp.) (9, 32). Clinical signs of Lyme disease in dogs are
fever, acute arthritis, arthralgia, lameness, and nephritis in some
cases. Central nervous system involvement, heart block, and uve-
itis are less frequently reported in dogs (2, 10, 12).
The diagnosis of Lyme disease is made on the basis of symp-
toms, including the animal living in an area where the disease is
endemic, ruling out other causes of clinical signs, and a high titer
accomplished by the detection of serum antibodies by a quantita-
tive but rather nonspecific enzyme-linked immunosorbent assay
ting (WB) (2, 11, 43). Other, more recent tests are based on the
detection of an invariable domain (IR6) of the variable surface
immunodominant in human patients with Lyme disease and also
in dogs infected with B. burgdorferi (14–16).
Fluorescent bead-based, multiplex analysis of antibodies to B.
burgdorferi is a novel approach of high analytical sensitivity and
allows for the simultaneous detection of immune responses to
several antigens in dogs and horses (39, 40). Several antigens of B.
burgdorferi, including outer surface protein A (OspA), OspC, and
OspF, are differentially expressed in ticks (OspA) (23, 24, 30, 36,
28) or later in the mammalian host (OspF) (1, 18, 21). Although
thoroughly investigated in ticks and during transmission, less in-
formation is available about the antigen expression of the spiro-
chetes in the mammalian host. Besides the studies on C6 men-
yme disease is the most common vector-borne disease in the
United States. It is caused by B. burgdorferi sensu stricto bacte-
tioned above, almost no data exist about the dynamics of
antibodies to various B. burgdorferi antigens after the infection of
dogs. Considering the ability of B. burgdorferi to regulate its sur-
it is likely that the differential expression of these antigens during
early or persistent infection results in a variation of the immune
of antibody to different B. burgdorferi antigens in dog serum will
likely provide us with greater insights into various stages of infec-
tion, could improve our understanding of this persistent patho-
gen, and is likely to influence prognosis and treatment decisions
for Lyme disease.
The aim of this study was to identify markers for early and late
infection by comparing antibody responses to different surface
antigens of B. burgdorferi in two sample sets. First, sera of experi-
mentally infected dogs were used to compare antibody responses
to OspA, OspC, OspF, flagellin B (FlaB), and two C6 peptides
during the first 3 months of infection and by using a novel multi-
based on C6 peptide. Second, we analyzed sera from canine pa-
Received 3 December 2011 Returned for modification 23 December 2011
Accepted 8 February 2012
Published ahead of print 15 February 2012
Address correspondence to Bettina Wagner, email@example.com.
Supplemental material for this article may be found at http://cvi.asm.org/.
Copyright © 2012, American Society for Microbiology. All Rights Reserved.
1556-6811/12/$12.00 Clinical and Vaccine Immunologyp. 527–535cvi.asm.org
ers OspC, OspF, and C6.
MATERIALS AND METHODS
or 50 (n ? 4) wild-caught ticks (Ixodes scapularis). A control group (n ?
4) was not exposed to ticks. Adult Ixodes scapularis ticks were collected in
the spring of 2008 in southern Rhode Island from an area in which B.
burgdorferi was endemic and were stored in the laboratory (in ventilated
brief period before infestation on dogs. Borrelia infection rates in ticks
were determined by dark-field assay to be 52% (20). On day 0, adult ticks
were placed in infestation chambers affixed to both sides of each dog.
Ticks were allowed to feed to repletion (between days 7 and 14) and were
removed on day 14. Starting on the day of tick challenge through to the
end of the study, 92 days later, animals were examined daily for clinical
signs associated with Borrelia infection, including, but not limited to,
lameness, ataxia, joint swelling, enlarged lymph nodes, depression, and
before exposure to ticks and on days 21, 35, 49, 63, 77, and 92 and were
analyzed for antibodies to Borrelia burgdorferi. The blood was allowed to
clot. Three drops of serum were used immediately in the SNAP 4Dx test,
using the Lyme Quant C6 test and Cornell’s multiplex Lyme assay (see
below). All tests were performed in a masked manner. Body skin punch
were taken from body regions close to the tick attachment sites at each
sampling day. Half of the biopsy specimens were used for culturing and
were placed directly into screw-cap tubes containing 6 ml Barbour-
and were frozen immediately and stored at ?70°C until analyzed. PCR
was performed with primers amplifying a fragment of the housekeeping
gene encoding FlaB (FlaB-F, 5= GACGACGACAAGATGATTATCAATC
TAAGCAATGACAAAACATA 3=). The dogs were euthanized at the end
of the study (day 92). A full necropsy was performed, and samples were
collected to investigate Borrelia-associated histopathology and tissue dis-
were collected: superficial lymph nodes, skin, the capsules of the left and
right carpi, elbows, shoulders, stifles, and tarsi. For each of these samples,
imal change; 2, dermatitis, superficial, mild; and 3, lymphoid prolifera-
tion, nodular, perineural. Histopathological findings in synovial mem-
branes were rated with the following scale: 0, no change to minimal
change; 1, focal to regional hyperplasia/hypertrophy of synovial lining
cells; 2, generalized hyperplasia of synovial lining cells; and 3, generalized
hyperplasia plus inflammatory cells plus fibrin deposition; lymph nodes
were determined to be either quiescent or reactive.
procedures (SOPs) and legal requirements in the conduct of the study,
national animal welfare regulations, and other applicable national regu-
latory requirements. The study protocol was reviewed and approved by
Pfizer’s Animal Use Committee (IACUC) prior to the start of the study.
Sera from canine patients. In addition to sera from experimentally
infected dogs, 125 serum samples from client-owned dogs were tested
using the multiplex assay to compare the performance of C6 and OspF as
serological Lyme testing to the Animal Health Diagnostic Center at Cor-
nell University between July 2008 and January 2009, and antibodies to B.
burgdorferi were analyzed by Western blotting using a whole-cell lysate
Information about clinical signs or a disease history was not available for
these dogs. The presence (positive) or absence (negative) of serum anti-
proteins on the blots was determined blindly by an observer who was not
aware of the multiplex assay results. All serum samples included in this
ysis, i.e., both tests were either negative or positive for OspF and also for
C6-P1, respectively. Patterns with positive C6 and negative OspF results
or the opposite were included in the comparison.
used to measure antibodies to B. burgdorferi OspA, OspC, and OspF an-
ies to six antigens of B. burgdorferi. Recombinant FlaB antigen was ex-
for OspA, OspC, and OspF (39). In brief, a 909-bp partial FlaB gene
corresponding to the extracellular part of the antigen was amplified from
Primers spanning bases 1 to 20 and 909 to 889 of the B. burgdorferi strain
gene into the pQE-30Xa expression vector (Qiagen Inc., Valencia, CA).
Valencia, CA). The bacteria were lysed, and the His-tagged FlaB protein
tography instrument (both from GE Healthcare, Piscataway, NJ). The
(Pierce, Rockford, IL). In addition, two 26-amino-acid C6 peptides from
the B31 and 297 strains of B. burgdorferi sensu stricto (15) were used as
antigens for the multiplex assay. One C6 peptide sequence (C6-P1) orig-
inated from strain B31 (MKKDDQIAAAIALRGMAKDGKFAVKD), and
the other C6 peptide (C6-P2) was from strain 297 (MKKNDQIAAAIVL
RGMAKDGEFALKD). The two C6 peptides differed in amino acids in
positions 4, 12, 21, and 24 (in boldface). The peptides were synthesized,
desalted by high-performance liquid chromatography (HPLC), and con-
Before coupling to the multiplex beads, the FlaB protein and the C6 pep-
tides coupled to BSA were run on an SDS gel and tested by Western
blotting with positive and negative canine sera as previously described
(39). The analysis confirmed the specific detection of the FlaB protein or
the supplemental material). BSA alone was not detected.
Fluorescent bead-based multiplex assay. Six B. burgdorferi antigens
were coupled to fluorescent beads (Luminex, Austin, TX), and the multi-
coupled to bead 33, OspC to bead 34, OspF to bead 37, FlaB to bead 36,
C6-P1 to bead 35, and C6-P2 to bead 38. For each antigen, 5 ? 103beads
were used per microtiter plate well (Multiscreen HTS plates, Millipore,
fer (phosphate-buffered saline [PBS] with 1% [wt/vol] BSA and 0.05%
[wt/vol] sodium acid) and incubated with the beads for 30 min on a
shaker at room temperature and in the dark. The plate was washed with
0.05% [wt/vol] sodium acid), and 50 ?l of biotinylated rabbit anti-dog
luted 1:5,000 in blocking buffer was added to each well and incubated for
30 min as described above. After washing, 50 ?l of streptavidin-phyco-
erythrin (Invitrogen, Carlsbad, CA) diluted 1:100 in blocking buffer was
added. Plates were incubated for 30 min as described above and washed.
The beads were resuspended in 100 ?l of blocking buffer, and each plate
Wagner et al.
cvi.asm.org Clinical and Vaccine Immunology
analyzed in a Luminex IS 100 instrument (Luminex, Austin, TX). The
data were reported as median fluorescent intensities (MFI). The positive
cutoff values for the multiplex assay were established by comparing MFI
values and Western blotting results of each antigen by receiver-operating
curve (ROC) analyses using 188 canine patient sera as described previ-
ously (39) or as shown in Fig. S1B in the supplemental material for FlaB
and the C6 peptides. Positive cutoff values were ?1,000 MFI (OspC and
C6-P2) and ?1,500 MFI (OspA, OspF, C6-P1, and FlaB).
Antibody testing by SNAP 4Dx test and Lyme QuantC6 ELISA.
Three drops of serum from experimentally infected dogs were tested im-
mediately after blood collection and processing using SNAP 4Dx tests as
described in the test kit (IDEXX Laboratories, Westbrook, ME). Samples
were considered positive if any color developed at the site of the B. burg-
Laboratories to perform the Lyme Quant C6 ELISA (C6 ELISA) testing.
Sera were considered positive for antibodies to C6 if the ELISA result was
Data and statistical analysis. For the C6 and FlaB bead assays within
canine serum samples (39). The Western blotting result (positive/nega-
tive) of the corresponding proteins (39 kDa for C6 and 41 kDa for FlaB)
served as relative gold standards for the comparison to MFI values ob-
tained by multiplex analysis. The procedure was previously described for
were generated using the MedCalc program, version 18.104.22.168 (F.
Schoonjans, Mariakerke, Belgium).
Serum samples from experimentally infected and control dogs were
coded on collection, and all antibody testing was performed masked by
ed-measures analysis of variance (RM ANOVA). The model was run for
ELISA, and a 2-sided P value of 0.05 was considered significant.
For the analysis of antibodies to the infection markers OspC, OspF,
and C6 in canine patient sera, we used 125 serum samples that were sub-
mitted for Lyme antibody testing to Cornell University. All of these sera
positive or negative for the respective antigens in both assays. The spear-
based on multiplex assay results. In addition, antibodies to OspC, OspF,
and C6 antigens were compared by a one-way RM ANOVA with Bonfer-
roni post tests to compare the responses between the three antigens. All
analyses were performed with 95% confidence intervals and P ? 0.05 for
using the GraphPad Prism program, version 5.01.
RESULTS AND DISCUSSION
Infection with B. burgdorferi after exposure of experimental
dogs to ticks. The successful transmission of Borrelia spirochetes
from the ticks to the dogs was confirmed using both PCR and
culture of skin samples taken close to the tick attachment site. On
obtained from the dogs and used to amplify the FlaB gene of B.
burgdorferi by PCR (Table 1). Bacterial DNA was detected in all
samples from dogs in the 25-ticks/dog group at both time points.
positive and two out of four samples were positive on day 92. All
PCRs from control dog samples were negative for B. burgdorferi.
ticks was further confirmed by bacterial cultures from the same
skin biopsy specimens (Table 1). B. burgdorferi could not be cul-
tured from any of the tissues of the control group. Spirochetes
were present in all noncontaminated cultures from dogs exposed
to 25 ticks and in all but one (day 92) culture from dogs in the
50-tick group (Table 1). Bacteria were found in cultured samples
from all dogs in the 50-tick group, i.e., in at least one of the skin
samples. PCR and bacterial culture confirmed that all dogs ex-
posed to wild-caught ticks were infected with B. burgdorferi. Be-
cause infection with B. burgdorferi was confirmed in all dogs that
were exposed to ticks, we will refer to day 0 as the day of infection
and to the ensuing days as days postinfection (p.i.).
Similar experimental infection models for dogs using wild-
caught ticks were previously reported (2, 34). In these models,
dogs were also successfully infected by tick exposure as indicated
needle injection of the spirochetes resulted in considerably lower
antibody titers (2).
Clinical signs and tissue histopathology. Sporadic fever
(body temperature of ?39.5°C) was observed in individual dogs
of the infected groups between days 9 and 21 (see Fig. S2A in the
supplemental material) and on days 83 and 92 in two dogs of the
control group. Clinical signs, such as weight loss, ataxia, joint
swelling, or depression, were not observed in any of the dogs dur-
days 90 to 92 p.i. in one dog that had been exposed to 50 ticks.
Enlarged lymph nodes were frequently found from day 15 p.i. on
end of the study, histology was performed on skin, lymph nodes,
and several synovial membranes, and a score was applied. Lymph
node tissues were reactive in all tick-exposed dogs and in two of
slight increase in histopathological changes was found in the in-
fected groups (see Fig. S2B). Increased inflammation in skin re-
gions close to the tick attachment sites was observed in both in-
fected groups but not in the control dogs (see Fig. S2C). Overall,
with B. burgdorferi were sporadic, and pathological tissue damage
was rather mild in the infected dogs compared to that of the con-
trol group. The clinical findings are in agreement with previous
studies of dogs reporting lameness as the predominant clinical
sign after experimental infection with B. burgdorferi. Lameness
was observed 2 to 5 months after tick exposure, and the most
severe histopathological changes were found at the time when
clinical signs occurred (2, 34).
Multiplex analysis of sera from experimentally infected
dogs. Multiplex analysis for the simultaneous detection of anti-
bodies to OspC, OspF, C6-P1,C6-P2, FlaB, and OspA antigens of
B. burgdorferi was performed on serum samples from all experi-
TABLE 1 Number of positive results obtained by amplification of the
FlaB gene by PCR and bacterial culture of B. burgdorferi using skin
biopsy specimens from experimentally infected dogs
and day p.i.
No. positive/total no. tested by exptl group
Control25 ticks/dog50 ticks/dog
aThe fourth culture could not be evaluated due to contamination.
B. burgdorferi Early and Late Infection Markers in Dog
April 2012 Volume 19 Number 4cvi.asm.org 529
mentally infected dogs on day 13 prior to tick exposure and on
days 21, 35, 49, 63, 77, and 92 p.i. (Fig. 1). Antibody values were
negative in sera from all dogs in the control group at all time
points and in sera from all infected dogs on day 13 prior to tick
exposure. Positive values for antibodies to B. burgdorferi were de-
tected by multiplex analysis on the first sampling day after infec-
tion (day 21) for selected antigens (see below). Between days 63
and 92 p.i., all infected dogs were positive for antibodies to five
antigens, OspC, OspF, C6-P1, C6-P2, and FlaB. The multiplex
data were further evaluated for two parameters: (i) the earliest
ues) and the time point when all infected dogs in a group were
positive for that antigen; and (ii) the dynamics of antibody values
to different antigens during the first 3 months after infection.
21 p.i. in one out of four dogs (Table 2). On day 35 (25 ticks),
FIG1 Multiplex antibody values (MFI) to six surface antigens of B. burgdorferi in serum samples of experimentally infected dogs. Dogs were divided into three
groups with four dogs per group: a noninfected control group and two infected groups exposed to 25 or 50 ticks, respectively. Serum samples were obtained on
day 13 prior to tick exposure (d-13) and on days 21, 35, 49, 63, 77, and 92 postexposure. All sera were analyzed in a multiplex assay to simultaneously measure
antibodies to six surface antigens of B. burgdorferi (A to F). The graphs show the mean values and standard errors for each time point and group of dogs. The
dotted lines indicated the positive cutoff values for individual markers. The arrow shows the time of exposure to infected ticks. Asterisks indicate significant
differences in antibody values between the 25- and 50-tick exposure groups: *, P ? 0.05; **, P ? 0.01; and ***, P ? 0.001.
TABLE 2 Number of dogs with positive antibody responses to B.
burgdorferi antigens in serum samples obtained between days 21 and 49
p.i. and measured by multiplex analysisa
No. positive/total no. tested by exptl group and day p.i.
25 ticks/dog 50 ticks/dog
21 35492135 49
aOn day 13 prior to infection, all antibody values were negative for all antigens. From
days 63 to 92 p.i., all antibody values were positive by multiplex analysis for both
infected groups and all infection markers, with the exception of OspA, which resulted
in positive values on day 21 p.i. only. The time points when antibodies to a particular
antigen were first detectable are shown in boldface.
Wagner et al.
cvi.asm.org Clinical and Vaccine Immunology
antibodies to OspC and FlaB were detected in all dogs, and those
C6-P2, and FlaB were observed in one or two of the four dogs
out of four dogs. Looking at sera from all experimentally infected
dogs (both tick loads), we found that 5/8 dogs had antibodies to
OspC in their serum on day 21 p.i. and 8/8 dogs were positive for
OspC antibodies on day 35 p.i. The OspF, C6-P1, or C6-P2 infec-
day 21 p.i., with 0/8, 2/8, or 1/8, respectively, and on day 35 p.i.,
with 5/8 or 7/8 positive samples. None of the three dogs that were
negative for OspC on day 21 p.i. had positive antibody values for
OspF, C6-P1, or C6-P2 (see Tables S1 and S2 in the supplemental
material). Based on these patterns obtained by simultaneous an-
tibody detection in samples from experimentally infected dogs, it
can be concluded that antibodies to OspC are the first indicators
of infection with B. burgdorferi, and antibodies to FlaB and C6
develop afterwards, followed by antibodies to OspF. The finding
agreement with the role of OspC during spirochete transmission
(9). OspC expression is required for the effective migration of the
bacteria from the tick intestine to the salivary glands (25). After
spirochetes enter the mammalian host, OspC also inhibits bacte-
rial killing by antibody-mediated cytotoxicity (28). As a conse-
it is one of the first antigens of B. burgdorferi that is recognized by
tibody responses. In humans, antibodies to OspC are also consid-
ered to be markers of early infection (1, 25).
In addition, we observed a small but clear increase in antibody
responses to OspA in all experimentally infected dogs on day 21
in the supplemental material). Some of the OspA antibody values
OspA is expressed in the midgut of infected ticks and becomes
downregulated during transmission (9, 23, 24, 30, 36, 44). The
immune system thus is not exposed to significant OspA expres-
sion after infection with B. burgdorferi. Nevertheless, OspA is
protect mice (6, 29) and dogs from infection (3, 4, 41). OspA is a
component of all currently available Lyme vaccines for dogs. In
serological diagnostics, antibodies to OspA therefore have been
interpreted as vaccination markers in dogs (8, 11, 37, 38). How-
ever, our current data show that OspA antibodies also form in
response to infection. Compared to the usually high and long-
lasting anti-OspA vaccination titers (39), OspA antibody re-
sponses in the experimentally infected dogs were low and tempo-
rary, as indicated by the small antibody peak on day 21 p.i. (Fig.
regulated by the spirochetes at the time of infection and can in-
in some dogs.
Dynamics of antibodies to different antigens of B. burgdor-
feri after infection. During the first 3 months of infection with B.
burgdorferi, the dynamics of antibodies to OspC differed from
those to the other antigens. Antibodies to C6-P1, C6-P2, OspF,
and FlaB (25 ticks) increased constantly (Fig. 1B to E), indicating
the presence of B. burgdorferi in the host and suggesting a contin-
uous immune stimulus provided by these antigens. Antibodies to
OspF, C6, and FlaB all provided robust infection markers by 2
months after infection. Their kinetics also suggested a titer in-
crease beyond 3 months of infection which eventually may de-
velop into a constant titer in chronically infected dogs, as previ-
ously described in long-term infection studies for antibodies to
antibodies to OspC peaked on day 49 (25 ticks) or 77 (50 ticks)
differential expression pattern of the OspC antigen during trans-
mission and its downregulation after infecting mammalian hosts.
The missing antigenic stimulus likely leads to the decline of anti-
bodies to OspC a few weeks after infection. Our findings also
suggested that OspC antibodies further decline after day 92 p.i.
and may become undetectable in late infection stages. The only
and 92 in the 50-tick exposure groups. However, this was not
consistent between the two exposure groups (Fig. 1E).
Exposure to 50 ticks resulted in slightly earlier detection of
antibodies to most antigens compared to that for the 25-tick ex-
posure groups (Table 2). We also analyzed if significant differ-
ences in the magnitude of antibody responses were observed for
duction after exposure to 50 ticks was observed only for OspF on
and E). Overall, this suggests that the influence of infection dose
on the magnitude of antibody responses to B. burgdorferi is small,
and that antibody values instead are influenced by the individual
animal’s immune response (see Tables S1 and S2 in the supple-
indicators for infection, thus are unlikely to give a direct correla-
tion to spirochete loads. Nevertheless, this requires further con-
firmation, because the number of spirochetes per tick was un-
have influenced the antibody responses of these dogs.
After experimental infection, FlaB appeared to be a robust
Although FlaB may be a valuable infection marker under experi-
mental conditions, dogs are usually exposed to various other bac-
teria expressing flagellin. In canine patient sera, cross-reactivities
with other flagellins were described previously (17, 31). Antibod-
ies to FlaB thus should not be considered specific markers for
infection with B. burgdorferi in canine patients.
In summary, the multiplex analysis of sera from experimental
markers for early infection with B. burgdorferi, (ii) antibodies to
with antibodies to C6 slightly preceding those to OspF in some
dogs, and (iii) that antibodies to OspF and C6 are maintained at
similar levels during the first 3 months of infection.
Analysis of Lyme QuantC6 ELISA titers in experimentally
infected dogs. Serum samples were also analyzed by SNAP 4Dx
tests and QuantC6 ELISA. The SNAP test was negative for all
control animals during the course of the study. In infected dogs,
all animals, positive in two out of four dogs per group at day 35
p.i., and positive in all infected dogs from day 49 on (Table 3).
B. burgdorferi Early and Late Infection Markers in Dog
April 2012 Volume 19 Number 4cvi.asm.org 531
identified one control dog as being positive for antibodies to C6
this study. All other samples from control dogs and from infected
dogs at day 13 before exposure were negative using the C6 ELISA
(Table 3). In tick-exposed groups, antibodies were identified at
day 21 p.i. in two out of four (25 ticks) or all dogs (50 ticks). The
antibodies to C6 increased until day 49 p.i. and decreased after-
wards in both infected groups (Fig. 2). The decrease in antibody
means between days 49 and 92 as identified by the C6 ELISA was
tively. The assay did not result in significant differences in anti-
bodies to C6 between the 25- and 50-tick exposure groups. Indi-
vidual dog antibody results for the C6 ELISA are shown in Table
S3 in the supplemental material. Antibodies to C6 identified by
ELISA or SNAP tests are valuable markers to identify infection
with B. burgdorferi in dogs (5, 13, 16, 26, 33). Here, the C6 ELISA
similar to antibodies to OspC in the multiplex analysis using the
depending on the dog, indicating the improved analytical sensi-
tivity of quantitative tests compared to the qualitative SNAP 4Dx
The Quant C6 ELISA is frequently used to make treatment
decisions in dogs and to monitor treatment success by the quan-
tification of antibodies (http://www.idexx.com/pubweb
quant-c6-white-paper.pdf). In an experimental study, a decrease
of C6 antibodies to baseline values was reported at 6 months after
antibiotic treatment with ceftriaxone (26). The antibiotic treat-
ment of C6 ELISA-positive client-owned dogs showed a drop in
the same study, a decline in C6 antibody units of up to 31% was
observed in C6 antibody-positive, untreated control dogs. Here,
we observed a decline of antibody values of 23 to 29% in the two
infected groups between days 49 and 92 p.i. in the absence of any
treatment. This finding is somewhat unexpected, considering the
persistent nature of infection with B. burgdorferi and the previ-
ously reported constant increase of antibodies during the first 3
months p.i. using whole-cell lysate ELISA (2, 34). It also was not
confirmed by using two C6 peptides in multiplex analysis. In the
same serum samples, both C6 multiplex values increased con-
63 on may reflect the limitations of this ELISA compared to mul-
tiplex technology. Multiplex assays have a much wider dynamic
range. Samples with antibody values between 0.001 and 1,000
ng/ml will fall into the linear range of the assay and allow the
quantification range (about 0.1 to 500 ng/ml). Thus, samples will
more frequently fall into the lower and upper plateau ranges, i.e.,
quantification is not accurate. Thus, one explanation for the de-
creasing values observed in Fig. 2 is that the samples have chal-
lenged the upper plateau of the C6 ELISA, resulting in inaccurate
antibody values. However, the technical details about the C6
ELISA are unknown, and it is difficult to speculate about the rea-
sons for its performance.
Comparison of antibodies to OspC, OspF, and C6 in canine
confirmed that antibody titers to B. burgdorferi whole-cell lysate
antibody kinetics obtained here for antibodies to OspF and C6
during experimental infection suggested that antibodies to both
interval of the study. Our data also suggested that antibodies to
OspC decrease after 7 to 11 weeks of infection and eventually
using an ELISA, antibodies to OspF and C6 (39 kDa) were found
to be the most important indicators of natural exposure to B.
burgdorferi in dogs, while antibodies to OspC were identified in
less than 10% of the clinical samples, leading to the conclusion
the disease is endemic and had a history of clinical signs of Lyme
disease and positive antibody titers by ELISA. Thus, the majority
of these samples likely represented chronic cases.
To compare antibodies to OspF and C6 as long-lasting and
robust markers for chronic infection with B. burgdorferi, and to
confirm that antibodies to OspC can become undetectable in late
infection stages, we analyzed sera from 125 canine patients. For
these samples, the Spearman’s rank correlations for antibodies to
OspF and C6 measured by multiplex analysis was 0.85 (P ?
0.0001). Identical (either negative or positive) multiplex results
confirming the high level of agreement between antibodies to
FIG 2 Antibody titers in sera from experimentally infected dogs identified by
Lyme Quant C6 ELISA. Serum samples were obtained on day 13 prior to
infection (d-13) and on days 21, 35, 49, 63, 77, and 92 p.i. The graph shows
mean values and standard errors for C6 antibody values of the control group
and two experimentally infected groups by exposure to 25 or 50 ticks per dog.
TABLE 3 Detection of antibodies to C6 by SNAP4Dx test and Lyme
Quant C6 ELISA in sera from experimentally infected dogs
Assay and group
No. positive/total no. tested by day p.i.a
?13 2135 49637792
Lyme Quant C6 ELISA
aThe time points when antibodies to C6 were first detectable by each of the assays are
shown in boldface.
Wagner et al.
cvi.asm.org Clinical and Vaccine Immunology
OspF and C6 as markers for infection with B. burgdorferi. Out of
these 115 samples, 51 sera were negative for antibodies to C6 and
OspF, and 64 sera were positive for both markers. All but 1 out of
51 negative samples also were negative for antibodies to OspC,
to 3 weeks after infection, before antibodies to C6 and OspF de-
tified dogs that were previously infected with B. burgdorferi. Dis-
agreement between OspF and C6 results was found for nine sera
that were positive for antibodies to C6 and negative for those to
OspF and one serum that showed the opposite result. Out of the
nine C6?/OspF?samples, six had positive antibody values to
OspC (Fig. 3A, bars 1 to 6), suggesting an early infection stage of
about 3 to 7 weeks p.i. The remaining four samples (three C6?/
OspF?and one C6?/OspF?) overall had very low positive anti-
body values for C6 and OspF, respectively, and were negative for
OspC. For these samples, infection with B. burgdorferi remained
questionable because of overall inconclusive antibody results and
only one low positive value that also could represent the high
background value of the respective serum.
positive for antibodies to OspC (Fig. 3B), while 41 sera were neg-
ative for OspC antibodies (Fig. 3C). In the 23 samples that were
positive for all three antigens, OspC antibody values were signifi-
cantly decreased compared to values obtained for antibodies to
that developed in experimentally infected dogs during the first 3
months after infection, sera from canine patients that were posi-
tive for antibodies to OspC, OspF, and C6 likely represented dogs
that were infected several weeks to a few months ago. Alterna-
tively, samples with antibodies to these three antigens also could
originate from chronically infected dogs that were recently rein-
fected with B. burgdorferi, resulting in a recent response to OspC
on top of previously existing antibodies to OspF and C6.
The results also confirmed that antibodies to OspC are not
detectable in many canine patient sera with robust OspF and C6
titers and suggested that these patterns are derived from long-
term, chronic infection with B. burgdorferi. Thus, the simultane-
ous appearance of antibodies to OspC, OspF, and C6 is likely
characteristic for an intermediate infection stage, and sera with
antibodies to OspF and C6 only suggest a late infection stage be-
cause antibodies to OspC disappear from the circulation over
time. The latter observation is in agreement with the downregu-
lation of OspC on the spirochete surface in the mammalian host
(7, 25, 36). The short availability of OspC directly after transmis-
sion for the induction of host immunity offers an explanation for
OspC antibodies over time because of the missing antigenic stim-
ulus. Figure 4 summarizes our results from the experimental in-
infection history. The projected data for OspF and C6 antibodies
after 3 months of infection closely resemble antibody titers to B.
burgdorferi that were described in previous long-term infection
studies after the experimental infection of untreated dogs (2, 16,
dogs and canine patient sera, confirmed that antibodies to OspC
are a sensitive marker for early infection with B. burgdorferi in
dogs and showed that antibodies to OspF and C6 are highly com-
FIG 3 Comparison of antibodies to OspC, OspF, and C6 obtained from ca-
nine patient sera. Antibodies were determined by multiplex analysis. Canine
University between July 2008 and January 2009. (A) OspC antibody values of
nine dogs (#1 to 9) with a C6?/OspF?and one dog (#10) with C6?/OspF?
detection pattern in serum. The antibodies to OspC, OspF, and C6 P1 were
analyzed in the multiplex assay. These 10 samples, with disagreement on the
Lyme antibody status interpretation based on C6 and OspF, were obtained
from a total of 125 canine patient sera. For the remaining 115 samples, the
multiplex assay interpretation based on OspF and C6 was in agreement. The
C6 values decrease from dog 1 to 10. The horizontal lines show the positive
antibodies to OspF and C6 that were negative for antibodies to OspC. The
horizontal lines in plots B and C indicate significant differences between the
antigens. The P value applies to both lines in each plot.
FIG 4 Summary of experimentally confirmed and projected values based on
canine patient sera for antibodies to OspC, OspF, and C6 antigens of B. burg-
dorferi during early and late infection. The data were obtained by multiplex
analysis. The lines for the first 3 months after infection are based on the mul-
tiplex results from experimentally infected dogs. After 3 months the lines are
projected from the data obtained from patient serum. The horizontal dotted
line shows the cutoff value for the multiplex assay. The vertical dotted line
indicates 3 months after infection.
B. burgdorferi Early and Late Infection Markers in Dog
April 2012 Volume 19 Number 4 cvi.asm.org 533
diagnosis of Lyme disease in clinically affected canine patients
from areas in which it is endemic. The canine Lyme multiplex
assay can distinguish between these antibodies and has a broad
quantification range, providing the increased accuracy of anti-
body detection during infection. The assay allows the simultane-
ous analysis of antibodies to multiple infection markers that are
to 5 months; OspC?/OspF?/C6?), or late infection (?5 months;
OspC?/OspF?/C6?). These advantages of the canine Lyme mul-
tiplex assay allow an interpretation of the stage of infection that
cannot be provided by other assays and is beneficial for the diag-
nosis, prognosis, and treatment of Lyme disease in dogs.
The multiplex assay development and the analysis of serum samples were
funded by the Method Development Funds of the Animal Health Diag-
nostic Center at Cornell University. The experimental infection study,
We acknowledge the expert contributions to this study of Nicole
Animal Health. We also thank T. N. Mather (University of Rhode Island)
for collecting the ticks and for determining tick infection levels.
financial interests in the context of this article. B.W. has submitted a
patent application entitled “Methods for diagnosing Lyme disease” that
uses technology described in the manuscript. This study has been per-
formed as a collaborative research project between P.M. at Pfizer Animal
Health and B.W. at Cornell University. There were no fees or financial
Pfizer Animal Health, which does not have a diagnostic division.
B. Wagner developed the multiplex evaluation approach, performed
some of the data analysis, interpreted the data, and drafted the manu-
was involved in the statistical analysis and revising the manuscript. C.
Earnhardt performed the B. burgdorferi DNA extraction and the FlaB
Meeus had oversight of the experimental infection study design and con-
tributed to critically revising the manuscript.
1. Akin E, McHugh GL, Flavell RA, Fikrig E, Steere AC. 1999. The
immunoglobulin (IgG) antibody response to OspA and OspB correlates
with severe and prolonged Lyme arthritis and the IgG response to P35
correlates with mild and brief arthritis. Infect. Immun. 67:173–181.
2. Appel MJG, et al. 1993. Experimental Lyme disease in dogs produces
arthritis and persistent infection. J. Infect. Dis. 167:651–664.
and disease caused by Borrelia burgdorferi. Infect. Immun. 63:3543–3549.
4. Conlon JA, Mather TN, Tanner P, Gallo G, Jacobson RH. 2000. Efficacy
after challenge by ticks naturally infected with Borrelia burgdorferi. Vet.
5. Duncan AW, Correa MT, Levine JF, Breitschwerdt EB. 2005. The dog as
a sentinel for human infection: prevalence of B. burgdorferi C6 antibodies
6. Fikrig E, Barthold SW, Kantor FS, Flavell RA. 1990. Protection of mice
against the Lyme disease agent by immunizing with recombinant OspA.
7. Grimm D, et al. 2004. Outer-surface protein C of the Lyme disease
spirochete: a protein induced in ticks for infection of mammals. Proc.
Natl. Acad. Sci. U. S. A. 101:3142–3147.
8. Guerra MA, Walker ED, Kitron U. 2000. Quantitative approach for the
serodiagnosis of canine Lyme disease by the immunoblot procedure. J.
Clin. Microbiol. 38:2628–2632.
9. Hovius JWR, van Dam AP, Fikrig E. 2007. Tick-host-pathogen interac-
tion in Lyme borreliosis. Trends Parasitol. 23:434–438.
10. Hutton TA, et al. 2008. Search for Borrelia burgdorferi in kidneys of dogs
with suspected “Lyme nephritis.” J. Vet. Intern. Med. 22:860–865.
11. Jacobson RH, Chang YF, Shin SJ. 1996. Lyme disease: laboratory diag-
nosis of infected and vaccinated symptomatic dogs. Semin. Vet. Med.
Surg. (Small Anim.) 11:172–182.
12. Levy SA, Magnarelli LA. 1992. Relationship between development of
of limb/joint borreliosis. J. Am. Vet. Med. Assoc. 200:344–347.
13. Levy SA, et al. 2008. Quantitative measurement of C6 antibody following
antibiotic treatment of Borrelia burgdorferi antibody-positive nonclinical
dogs. Clin. Vaccine Immunol. 15:115–119.
14. Liang FT, et al. 1999. An immunodominant conserved region within the
variable domain of VlsE, the variable surface antigen of Borrelia burgdor-
feri. J. Immunol. 163:5566–5573.
15. Liang FT, Philipp MT. 1999. Analysis of antibody response to invariable
16. Liang FT, Jacobson RH, Straubinger RK, Grooters A, Phillipp MT.
ful in canine Lyme disease serodiagnosis by enzyme-linked immunosor-
bent assay. J. Clin. Microbiol. 38:4160–4166.
17. Lindenmayer J, Weber M, Bryant J, Marquez E, Onderdonk A. 1990.
Comparison of indirect immunofluorescent-antibody assay, enzyme-
of Lyme disease in dogs. J. Clin. Microbiol. 28:92–96.
18. Magnarelli LA, Flavell RA, Padula SJ, Anderson JF, Fikrig E. 1997.
Serologic diagnosis of canine and equine borreliosis: use of recombinant
antigens in enzyme-linked immunosorbent assays. J. Clin. Microbiol. 35:
19. Magnarelli LA, et al. 2001. Reactivity of dog sera to whole-cell or recom-
J. Med. Microbiol. 50:889–895.
20. Mather TN, SR Telford III, Moore SI, Spielman A. 1990. Borrelia
to vector ticks (Ixodes dammini). Exp. Parasitol. 70:55–61.
21. McDowell JV, Sung SY, Price G, Marconi RT. 2001. Demonstration of
the genetic stability and temporal expression of select members of the
Lyme disease spirochete OspF protein family during infection in mice.
Infect. Immun. 69:4831–4838.
22. Morgan E, et al. 2004. Cytometric bead array: a multiplexed assay plat-
form with applications in various areas of biology. Clin. Immunol. 110:
laris mediated by outer surface protein A. J. Clin. Investig. 106:561–569.
24. Pal U, Fikrig E. 2003. Adaptation of Borrelia burgdorferi in the vector and
vertebrate host. Microb. Infect. 5:659–666.
25. Pal U, et al. 2004. OspC facilitates Borrelia burgdorferi invasion of Ixodes
scapularis salivary glands. J. Clin. Investig. 113:220–230.
26. Philipp MT, et al. 2001. Antibody response to IR6, a conserved immu-
nodominant region of the VlsE lipoprotein, wanes rapidly after antibiotic
humans. J. Infect. Dis. 184:870–878.
27. Prabhakar U, Eirikis E, Miller BE, Davis HM. 2005. Multiplexed cyto-
kine sandwich immunoassays: clinical applications. Methods Mol. Med.
28. Ramamoorthi N, et al. 2005. The Lyme disease agent exploits a tick
protein to infect the mammalian host. Nature 436:573–577.
29. Schaible UE, et al. 1990. Monoclonal antibodies specific for outer surface
protein A (OspA) of Borrelia burgdorferi prevent Lyme borreliosis in se-
vere combined immunodeficiency (SCID) mice. Proc. Natl. Acad. Sci.
U. S. A. 87:3768–3772.
30. Schwan TG, Piesman J, Golde WT, Dolan MC, Rosa PA. 1995. Induc-
tion of outer surface protein on Borrelia burgdorferi during tick feeding.
Proc. Natl. Acad. Sci. U. S. A. 92:2909–2913.
31. Shin SJ, et al. 1993. Cross-reactivity between B. burgdorferi and other
Lyme disease agent in dogs. Vet. Microbiol. 36:161–174.
32. Steere AC. 2001. Lyme disease. N. Engl. J. Med. 345:115–125.
33. Stone EG, Lacombe EH, Rand PW. 2005. Antibody testing and Lyme
disease risk. Emerging Infect. Dis. 11:722–724.
34. Straubinger RK, Straubinger AF, Summers BA, Jacobson RH. 2000.
Wagner et al.
cvi.asm.orgClinical and Vaccine Immunology
Status of Borrelia burgdorferi infection after antibiotic treatment and the Download full-text
effects of corticosteroids: an experimental study. J. Infect. Dis. 181:1069–
disease in the dog. J. Comp. Pathol. 133:1–13.
mission through the tick. J. Exp. Med. 199:603–606.
37. Töpfer KH, Straubinger RK. 2007. Characterization of the humoral
A study with five commercial vaccines using two different vaccination
schedules. Vaccine 25:314–326.
for simultaneous quantification of cytokines in horses. Vet. Immunol.
39. Wagner B, Freer H, Rollins A, Erb HN. 2011. A fluorescent bead-based
multiplex assay for the simultaneous detection of antibodies to B. burg-
dorferi outer surface proteins in canine serum. Vet. Immunol. Immuno-
40. Wagner B, et al. 2011. Development of a multiplex assay for detection of
antibodies to Borrelia burgdorferi in horses and its validation using Bayes-
ian and conventional statistical methods. Vet. Immunol. Immunopathol.
41. Wieneke CA, et al. 2000. Evaluation of whole-cell and OspC enzyme-
linked immunosorbent assays for discrimination of early lyme borreliosis
from OspA vaccination. J. Clin. Microbiol. 38:313–317.
42. Wikle RE, Fretwell B, Jarecki M, Jarecki-Black JC. 2006. Canine Lyme
ovalent Lyme vaccine. Intern. J. Appl. Res. Vet. Med. 4:23–28.
43. Wittenbrink MM, Failing K, Krauss H. 1996. Enzyme-linked immu-
nosorbent assay and immunoblot analysis for detection of antibodies
to Borrelia burgdorferi in dogs. The impact of serum absorption
with homologous and heterologous bacteria. Vet. Microbiol. 48:257–
44. Yang XF, Pal U, Alani SM, Fikrig E, Norgard MV. 2004. Essential role
for OspA/B in the life cycle of the Lyme disease spirochete. J. Exp. Med.
B. burgdorferi Early and Late Infection Markers in Dog
April 2012 Volume 19 Number 4 cvi.asm.org 535