American Journal of Epidemiology
Copyright O 1999 by The Johns Hopkins University School of Hygiene and Public Health
All rights reserved
Vol.149, No. 8
Printed In USA.
Antibody Levels to RecombinantTick Calreticulin Increase in Humans after
Exposure to Ixodes scapularis (Say) and Are Correlated with Tick
Engorgement Indices
Martin L. Sanders,1 Gregory E. Glass,2 Robert B. Nadelman,3 Gary P. Wormser,3 Alan L. Scott,2 Syamal Raha,6
Bruce C. Ritchie,6 Deborah C. Jaworski,4 and Brian S. Schwartz5
The antibody responses of subjects who presented with a definite Ixodes scapularis (Say) tick bite were
measured to determine the utility of the antibody response against recombinant tick calreticulin (rTC) as a
biologic marker of tick exposure. Subjects bitten by /. scapularis evidenced an increase in antj-rTC antibody
levels between visit 1 and visit 2 from 24.3 to 27.1 ng/uJ serum (n = 88, p = 0.003), and levels remained elevated
at visit 3 (p = 0.005). These anti-rTC antibody levels during visits 2 and 3 were significantly higher than those in
four non-exposed controls. Tick engorgement indices, measured on the biting ticks, were found to be correlated
with anti-rTC antibody levels (e.g., for visit 3: Pearson's r = 0.357, p = 0.001). Tick engorgement index (TEI), ratio
of body length to scutal width, was identified to be the only independent predictor of anti-rTC antibody levels in
linear regression models. Logistic regression revealed that a bite from an /. scapularis tick that became engorged
(TEI >3.4) was a risk factor for anti-rTC antibody seropositivity (adjusted odds ratio for age and bite location =
7.4 (95% confidence interval 2.1-26.4)). The anti-rTC antibody test had a sensitivity of 0.50 and a specificity of
0.86 for a bite from /. scapularis that became engorged. Immunoblotting revealed that subjects made a specific
anti-rTC antibody response. Am J Epidemiol 1999;149:777-84,
antibody; biologic marker; Ixodes; Lyme disease; tick
In epidemiologic research, host antibody responses
to pathogens are commonly used to assess exposure
and disease risk. Studies show humans and animals
produce antibodies to salivary proteins from various
blood feeding arthropods, including mosquitoes (1),
biting flies (2), and ticks (3). In ticks, salivary glands
are the largest glands and perform a variety of impor-
tant functions that assist feeding and inhibit the host
immune response (4-7). Many tick saliva proteins
have been shown to be immunogenic to mammalian
hosts (4).
Previous studies suggest antibodies to tick salivary
gland proteins are biologic markers of exposure to
ticks (8-12). For example, levels of anti-tick saliva
antibody (ATSA), using sonicated whole Ixodes
scapularis Say or Amblyomma americanum L. salivary
glands as antigen, are associated with self-reported
tick exposure, Lyme disease seroprevalence, and an
index of tick engorgement (9). ATSA levels are
inversely associated with self-reported personal protective behaviors and have been found to have a sensitivity of 0.81 and a specificity of 0.56 for a bite by /.
scapularis that became engorged (9).
It was hypothesized that the sensitivity and specificity of ATSA as a biomarker of tick exposure may be
improved by measuring the antibody response to a limited number of specific salivary proteins rather than
the antibody response to whole sonicated glands. In
the present study, the samples from Westchester
County, New York, were re-assayed with a newly identified recombinant tick saliva protein (recombinant
tick calreticulin, rTC). Calreticulin is a major calciumbinding protein of the endoplasmic reticulum, and it
Received for publication January 23,1998, and accepted for publication August 25, 1998.
Abbreviations: ATSA, anti-tick saliva antibody; Cl, confidence
interval; ELISA, enzyme-linked immunosorbent assay; OR, odds
ratio; rTC, recombinant tick calreticulin; TEI, tick engorgement
index.
1
Maryland Department of Health and Mental Hygiene, Division of
Outbreak Investigation, Baltimore, MD.
2
Department of Molecular Microbiology and Immunology, Johns
Hopkins University, School of Hygiene and Public Health, Baltimore,
MD.
3
Division of Infectious Diseases, New York Medical College,
Westchester County Medical Center, Valhalla, NY.
4
Department of Molecular Biology and Biochemistry, University of
California at Irvine, Irvine, CA.
5
Johns Hopkins School of Hygiene and Public Health, Division of
Occupational and Environmental Health, Baltimore, MD.
6
Dept. of Medicine, University of Alberta, Edmonton, Alberta,
Canada.
Reprint requests to Dr. Brian Schwartz, Johns Hopkins School of
Hygiene and Public Health, Division of Occupational and
Environmental Health, Room 7041, 615 North Wolfe Street,
Baltimore, MD 21205.
777
778
Sanders et al.
appears to be secreted in the saliva of A. americanum
and Dermacentor variabilis Say (13). The presence of
calreticulin has been demonstrated in the salivary
glands of /. scapularis (D. Jaworski, UC Irvine, personal communication, 1998), but, as of this writing, it
has not been isolated and expressed for use in biomarker assays. Calreticulin is not detectable in the
salivary glands of unfed ticks, but is observable by the
third day of feeding (13).
Anti-rTC antibodies are produced by animals experimentally fed upon by D. variabilis ticks (12). In contrast, gerbils exposed to Aedes aegypti did not develop
detectable levels of anti-rTC antibodies. Anti-rTC antibody seropositivity also was associated with selfreported protective behaviors and fort of origin in military personnel on maneuvers in areas at Fort Chaffee,
Arkansas, that were infested with A. americanum (12).
These observations all suggest that anti-rTC antibody
is a biologic marker of tick exposure and that this
recombinant tick salivary gland protein can be used in
the place of whole tick sab'vary glands in the measurement of antibodies to tick salivary gland proteins.
This study examines the kinetics of anti-rTC antibody levels in subjects from Westchester County, New
York, with a recent and confirmed tick bite and normal
controls. Subjects were part of a randomized trial of
empiric antibiotic treatment to prevent Lyme disease
after a definite /. scapularis bite in Westchester
County, New York. Tick bite subjects retained their
ticks and submitted them for acarologic assessment,
allowing analysis of anti-rTC antibody response relative to a number of acarologic variables.
MATERIALS AND METHODS
Study population
The study population has been previously described
(9). Briefly, subjects were recruited from Westchester
County, New York, an area hyperendemic for Lyme
disease transmission (14). In this study, three subsets
of study subjects were examined: 1) persons with a
recent (removed <72 hours before blood specimen
obtained) /. scapularis bite (n = 95), 88 of whom provided a tick that was suitable for determination of the
tick engorgement index, 2) persons with erythema
migrans (n = 7), and 3) normal controls without history of tick bite (n = 4). Subjects had three blood samples collected over approximately 6 weeks.
Antigen preparation
Recombinant tick calreticulin (rTC) was prepared as
described previously (Raha et al., unpublished manuscript). Briefly, a truncated A. americanum calreticulin
cDNA clone (missing 90 base pairs of the 5' end) was
subcloned into pRSETB (Invitrogen, San Diego,
California) and used to transform TOPP 5 cells
(Stratagene, La Jolla, California). Expression and
purification of the fusion protein was accomplished by
the protocols provided by Invitrogen. A single colony
was grown in LB medium at 30°C. Cells in the exponential growth phase were heat induced at 42°C for 30
minutes, then grown at 37°C for 2 hours. Cells were
harvested by centrifugation and lysed in guanidinium
buffer. The guanidinium was removed using a
Centriprep 3 (Amicon, Beverly, Massachusetts) and
the protein was purified over a nickel column (immobilized anion chromatography).
Study variables
Study variables included demographic characteristics of study subjects, time since tick removal, estimated time from initial attachment of tick to acquisition of
blood specimens, study group (/. scapularis bite, erythema migrans, nonexposed control), and several
study variables such as tick type, stage, tick engorgement index (TEI), i.e., ratio of body length to scutum
width (15), history of prior tick bite, history of clinical
Lyme disease, self-reported second tick bite during the
study period, and attachment site.
Human anti-rTC antibody ELISA
All serologic assays were conducted without knowledge of the status of the study subject, and all samples
for a given study subject were assayed on the same plate.
Enzyme-linked immunosorbent assay (ELISA) plates
were coated with 0.2 mg rTC (in 100 (ll phosphatebuffered saline (PBS)) per well overnight at 4°C. The
plates were blocked with blotto (2 percent nonfat milk in
PBS) for 90 minutes at 37°C. Sera were diluted 1:200 in
blotto and 100 microliters per well was incubated
overnight at 4°C. After washing, a 1:1000 dilution of
goat anti-human immunoglobulin G (IgG) conjugated to
horseradish peroxidase (The Binding Site, San Diego,
California) in blotto was added to the plates and incubated for 75 minutes at 37°C. The plates were developed
with ABTS (Kirkegaard & Perry Laboratories, Inc.,
Gaithersburg, MD) and the optical density (OD) measured at 405 nm. The serum concentration of human
anti-rTC antibody in ng/(0.1 (nanograms per microliter)
was determined based on a kinetic measure of anti-rTC
antibody ELISA OD using a VAX kinetic plate reader
(Molecular Devices, Sunnyvale, California). Values
obtained from the kinetic measure of ELISA OD were
adjusted for background and converted to nanograms of
human IgG per microliter human serum using a standard
curve of known human IgG concentrations.
Am J Epidemiol
Vol. 149, No. 8, 1999
Anti-Tick Calreticulin Antibody: A Biomarker of Tick Bites
Statistical analysis
Analyses were performed with BMDP statistical
software programs (16). Descriptive analyses were
performed on all study variables. Analysis of variance
was used to compare the mean anti-rTC antibody levels in the different subject groups. Pearson's correlation (r) was used to examine the association between
pairs of continuous variables. Paired Mests were used
to evaluate changes in anti-rTC antibody levels over
time.
Multiple linear regression was used to model antirTC antibody response as a function of the human- and
tick-derived study variables. Analyses involving the
engorgement index were performed only on the subjects with an /. scapularis bite when an engorgement
index was obtained (n = 88).
Anti-rTC antibody levels were dichotomized at the
75th percentile of the anti-rTC antibody distribution,
dividing tick bite subjects into "positive" and "negative" anti-rTC antibody serologic results. The 75th
percentile of the distribution corresponded to the mean
anti-rTC antibody level plus 4.9 standard deviations of
the four control subjects. Subsequently, stratified
analysis and logistic regression were used to identify
risk factors for anti-rTC antibody seropositivity.
Continuous variables were dichotomized so that odds
ratios and 95 percent confidence intervals could be
determined.
The sensitivity and specificity of anti-rTC antibody
serologic status were calculated for individuals with a
tick engorgement index of ^3.4 (75th percentile of the
TEI distribution) compared with those with TEI below
this cutoff. Experiments in animals suggest that a TEI
of 3.4 corresponds to an experimental feeding duration
of 69 hours for nymphal /. scapularis and 50 hours for
adult /. scapularis (17).
779
domly selected, and sera from each of their three visits
were used. The NC strips were washed with PBST and
incubated for one hour with rocking at room temperature with horseradish peroxidase labeled goat antihuman IgG (gamma chain specific) (the Binding Site,
San Diego, California), diluted 1:1,000 in blotto. The
NC strips were washed with PBST and PBS, and then
developed with 4-chloro-l-naphthol (4CN Peroxidase
Substrate, Kirkegaard & Perry Laboratories, Inc.,
Gaithersburg, MD).
RESULTS
Of the 95 subjects bitten by /. scapularis, five developed symptoms compatible with clinical Lyme disease, nine reported a prior diagnosis of Lyme disease,
27 reported a prior tick bite, and 17 reported receiving
a second tick bite during the study period (table 1). Of
the /. scapularis ticks with a measured TEI, approximately 57 percent were nymphs and 43 percent were
adult females.
Anti-rTC antibody levels increased over time
Mean anti-rTC antibody levels were higher in tick
bite subjects than in controls for visits 2 and 3; there
was no significant difference between tick bite subjects and controls for visit 1 (table 2). Subjects with
erythema migrans had significantly higher anti-rTC
antibody levels for visits 2 and 3 compared with controls, but not for visit 1. In subjects who were bitten by
a tick with a measurable TEI, mean anti-rTC antibody
levels were significantly higher at visits 2 and 3 compared with visit 1; there was no significant difference
between visits 2 and 3 (table 2). Erythema migrans
subjects demonstrated a similar result, while control
subjects evidenced no change over time.
Immunoblotting
Approximately 40 |ig of rTC was electrophoresed by
SDS-PAGE on a pre-cast, 4-20 percent polyacrylamide
gradient denaturing mini-gel (Jules Biotechnology,
New Haven, Connecticut). Proteins were transferred to
a nitrocellulose (NC) membrane (Trans-Blot, Bio-Rad
Laboratories, Hercules, California) in a Mini-Protean II
Electrophoresis/Transfer Unit (Bio-Rad Laboratories)
containing 192 mM glycine, 89 mM Tris, and 20 percent (v/v) methanol in distilled water, at 4°C, for 3.5
hours at 170 mA.
Following transfer, the NC membrane was cut into 4
mm wide strips and blocked with blotto. The NC strips
were incubated with human sera (diluted 1:100 in blotto) overnight at 4°C with rocking. Three tick bite subjects with a tick on which TEI was measured were ranAm J Epidemiol
Vol. 149, No. 8, 1999
TABLE 1. Characteristics of 95 study subjects reporting an
Ixodea scapularis bite, Westchester County, New York, 1990
Characteristic
No.
Age (years), mean (SD*)
Sex (% male)
History of Lyme disease (%)
History of prior tick bite (%)
Second bite during study (%)
Duration (days), mean (SD) until
Visit 1
Visit 2
Visit 3
Self-reported duration of bite
(hours), mean (SD)
1
SD, standard deviation.
Tick bite subjects
95
46.0 (14.9)
47.6
10.5
32.1
17.9
0.80 (0.85)
22.90 (3.60)
46.20 (6.01)
19.2(20.5)
780
Sanders et al.
TABLE 2. Antibody response against recombinant tick calretlculin (antl-rTC, antibody levels in nanograms per microllter (ng/u.1)
serum) determined by ELISAt in subjects with an Ixodes scapularls bite and tick engorgement index (n = 88), subjects presenting with erythema mlgrans, and normal controls, Westchester County, New York, 1990
VlsH 1*
Study group
/. scapularis bite with TElt
(n = 88)§
Erythema migrans (n = 7)
Controls (n = 4)
Mean
ng/(J
serum
24.3
24.5
14.7
VlsH 2*
95% Clf
Mean
ng/uJ
serum
19.1-29.5
12.5-36.5
6.2-23.2
27.1
28.3
11.8
visit 3*
p values}:
95% Cl
Mean
ng/jil
serum
95% Cl
Visit
1 vs. 2
Vlsjt
1 vs. 3
Visit
2vs 3
20.5-33.6*
15.1-41.5*
6.2-17.4
25.9
26.9
10.1
20.3-31.6*
14.7-39.1*
5.9-14.3
0.003
0.001
0.151
0.005
0.003
0.240
0.296
0.100
0.460
* Mean duration of time between visits for tick bite subjects is shown in table 1. For erythema migrans subjects, visit 1 occurred at presentation of erythema migrans, visit 2 at approximately 3 weeks, and visit 3 at approximately 7 weeks.
t ELISA, enzyme-linked immunosorbent assay; Cl, confidence interval; TEI, tick engorgement index, ratio of body length to scutal width,
i p value for paired Mest comparing mean anti-rTC antibody levels from visit 1 and 2, visit 1 and 3, and visit 2 and 3.
§ Visit 3: n = 78, 10 individuals had no sera collected for visit 3.
# These anti-rTC antibody levels were statistically significantly different from levels in the controls from the same study visit.
TEI Is correlated with anti-rTC antibody levels
Risk factors for anti-rTC seroposltivlty
TEI was found to be correlated with anti-rTC antibody levels for all three visits (table 3 and figure 1). A
previous study found that engorgement index was correlated with ATSA levels (9), and those data are shown
for comparison in table 3. ATSA levels were poorly
correlated with the corresponding anti-rTC antibody
levels for all visits (table 2).
TEI was divided into quartiles (<2.45, 2.46-2.71,
2.72-3.4, and >3.4) and associations between TEI and
anti-rTC antibody levels were examined (table 4).
Subjects who were bitten by a tick with a TEI between
2.72 and 3.4 had approximately a threefold increased
risk of anti-rTC antibody seropositivity (odds ratio
(OR) = 2.94, 95 percent confidence interval (Cl)
0.42-25.5), while subjects bitten by ticks with a TEI of
>3.4 demonstrated a tenfold increased risk (OR = 10.0,
95 percent Cl 1.6-80.1).
Logistic regression next was used to model the association between anti-rTC antibody seropositivity and
TEI, controlling for age (<46 vs. >46 years; median
age = 46 years) and other confounding variables (table
5). For these analyses, TEI was dichotomized at 3.4,
the 75th percentile in its distribution. Only location of
tick bite was associated with anti-rTC antibody
seropositivity. Using the four most common bite locations (thigh, forearm, arm, and groin) compared with
all other locations, only the thigh was found to be significantly associated with seropositivity for anti-rTC
antibody. When tick bite location was added to the
model, the odds ratio for the association of TEI with
anti-rTC antibody seropositivity increased to 7.4 (95
percent Cl 2.1-26.4). Individuals who were bitten on
the thigh had a 12-fold increased risk of seropositivity
for anti-rTC antibody (OR = 12.4, 95 percent Cl
2.54-61.0).
Predictors of anti-rTC antibody levels
The anti-rTC antibody levels at visit 3, the change in
the antibody levels between visits 1 and 2 and the
change in the antibody levels between visits 1 and 3
were modeled using linear regression. TEI was found to
predict anti-rTC antibody levels during all three visits.
No other variables were found to be associated with the
anti-rTC antibody levels when added to the linear
regression models. No study variables were found to
predict the change in anti-rTC antibody levels over time.
TABLE 3. Correlations between tick engorgement index
(TEI), antibody response against recombinant tick calretlculum (antl-rTC), and anti-tick saliva antibody (ATSA) levels* in
88 subjects bitten by Ixodes scapularls with measurable TEI,
Westchester County, New York, 1990
Variable 1
TEI
anti-rTC
Variable 2
Visit
no.
Pearson's
rvalue
P
value
antl-rTCr
anti-rTC
anti-rTC
ATSA
ATSA
ATSA
ATSA
ATSA
ATSA
T
2
3
1
2
3
1
2
3
0.35
0.33
0.36
0.35
0.36
0.38
-0.017
-0.045
-0.043
0.001
0.003
0.001
0.001
<0.001
<0.001
0.87
0.65
0.67
• ATSA values from Schwartz et al. (10).
Sensitivity and specificity of anti-rTC antibody
test
Anti-rTC antibody seropositivity as a test to identify
a subject bitten by a tick that became engorged (TEI
>3.4) had a sensitivity of 50 percent and a specificity
of 86 percent.
Am J Epidemiol Vol. 149, No. 8, 1999
Anti-Tick Calreticulin Antibody: A Biomarker of Tick Bites 781
160 -
o
100 -
i
vel (
O)
Anti-rTC = 9.5388 (El) - 3.6568.
140 - N = 78
r = 0.357
120 - p = 0.001
80 •o
C antil
O
n
60 40 20 -
1.J-!
0 -
*->
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
Engorgement index
FIGURE 1 . Relation between tick engorgement index and anti-rTC antibody level at visit 3 (mean (standard deviation) days from tick bite to
visit 3 was 46.20 (6.01) days), in 78 subjects in Westchester County, New York, 1990.
TABLE 4. Association of seroposltlvlty of antibody response
against recomblnant tick calreticulin (anti-rTC) with tick
engorgement index In 88 subjects bitten by Ixodes scapularis,
Westchester County, New York, 1990
Posittvef
Negative
Total
Crude
odds
ratio
<2.45
2.46-2.71
2.72-3.4
>3.4
2
2
5
11
20
20
17
11
22
22
22
22
1.0*
1.0
2.94
10.0
Total§
20
68
88
Anti-rTC antibody status
TEI*
95% Cl*
0.09-11.4
0.42-25.5
1.6-80.1
* TEI, tick engorgement index, ratio of body length to scutal
width; Cl, confidence interval; SD, standard deviation.
t Defined as 75th percentile in distribution of anti-rTC antibody
levels, 4.9 SD* above mean of four subjects without a history of tick
bites.
t Reference group.
§ Chi-square, p = 0.003; chi-square test for linear trend, p =
0.0006.
Immunoblots
Immunoblots using rTC as antigen were performed
in order to determine the analytical specificity of the
antibody response detected by ELISA (figure 2). All of
the tested human serum samples had IgG that recognized a band of approximately 64 kDa for all three visits. Little qualitative difference in band intensity
between time points was observed in any of the three
subjects; no quantitative measurements of band intenAm J Epidemiol
Vol. 149, No. 8, 1999
TABLE 5. Risk factors for seroposrtjvtty* for antibody
response against recomblnant tick calretlculum (anti-rTC)
antibody after tick bite In 88 subjects bitten by Ixodes
scapularis, Westchester County, New York, 1990
Variable
Bite by a tick that
became engorged
vs. unengorged
tick§ (TElf £3.4
vs. <3.4)
Age (years) (£46 vs.
<46 years)
Crude
odds
ratio
95%Clf
Adjusted);
odds
ratio
95% Cl
5.3
1.8-16.1
7.4
2.1-26.4
2.1
0.4-6.2
1.7
0.5-5.7
• Defined as 75th percentile in distribution of anti-rTC antibody
levels, 4.9 SDt above mean of four subjects without a history of tick
bites.
t Cl, confidence interval; TEI, tick engorgement index, ratio of
body length to scutal width; SD, standard deviation.
% Adjusted for age and bite location.
§ A TEI £3.4 was the 75th percentile in the TEI distribution.
sity were made. The results indicate that the antibody
response made by humans, as measured by ELISA,
appears to include specific anti-rTC antibody.
DISCUSSION
The present study describes associations between
anti-rTC antibody seropositivity and several acarologic
variables. While several similarities exist between antirTC antibody and ATSA, a previously described bio-
782
Sanders et al.
MWM
1
2
3
4
5
5
Z
8
9
208
144
87
44.1
32.7
•
17.7
7.1
FIGURE 2. Immunoblots of sera of three human subjects bitten by /. scapularis using rTC as antigen. All subjects provided ticks which had
measured tick engorgement index (TEI); only subject 1 was bitten by a tick which became engorged (TEI >3.4). Lanes 1, 2, and 3: subject 1,
visits 1, 2, and 3, respectively; lanes 4, 5, and 6: subject 2, visits 1, 2, and 3, respectively; lanes 7, 8, and 9: subject 3, visits 1, 2, and 3, respectively.
logic marker of tick exposure, the present study demonstrates that anti-rTC antibody levels increase after a single tick bite. Anti-rTC antibody seropositivity also has
lower sensitivity and higher specificity, compared with
ATSA, to identify subjects bitten by a tick that became
engorged. These results demonstrate that there are differences between the host antibody response directed
against rTC and that directed against ATSA.
Anti-rTC antibody levels increased during the time
from visit 1 to visit 2 (approximately 3 weeks) in study
subjects bitten by a tick with a determined TEI. This
group was also shown to have significantly higher levels of anti-rTC antibody than nonexposed controls during visits 2 and 3. Although the tick bite subjects had
higher anti-rTC antibody levels than nonexposed controls at visit 1, this difference was not statistically significant. The increase in levels of specific anti-rTC
IgG antibody over time demonstrates that rTC is recognized by the immune system of the study subjects
after a bite from a feeding tick, probably representing
secondary exposure to ticks in this hyperendemic area.
Subjects who reported erythema migrans (« = 7)
were part of a second study and did not have ticks
available for measurement of TEI. This group demonstrated an increase in anti-rTC antibody levels similar
to that of the larger group who were bitten by a tick on
which TEI was measured (n = 88). Erythema migrans
would be indicative of a recent, longer duration tick
bite, and thus these individuals would be expected to
develop anti-rTC antibody levels similar to those subjects with observed recent tick bite.
Immunoblots of individuals reveal that humans
make a specific anti-rTC antibody response. This
result is similar to previous anti-rTC antibody
immunoblots done with human sera (12), which found
that study subjects from Fort Chaffee, Arkansas, visualized a band at approximately 64 kDa, the molecular
weight of rTC.
TEI, measured from ticks removed from study subjects, was shown to be correlated with anti-rTC antibody levels at all three visits. When four subjects with
high TEI but low anti-rTC antibody levels were
removed from the scatterplot, the Pearson's r increased
to 0.493 (p = 0.001). It is possible that inter-species
differences in calreticulin could explain those results.
The recombinant tick calreticulin used in this study
was cloned from the salivary glands of A. americanum
(13), and is, to our knowledge, the only recombinant
tick saliva protein available. As stated above, subjects
in the current study were bitten by /. scapularis.
Individuals with long duration tick bites from /. scapularis (high TEI) may not develop high levels of antirTC antibodies to A. americanum derived calreticulin,
depending on the extent of their cross-reactivity. In
addition, antibodies produced by subjects in the current study were produced in response to a bite from /.
Am J Epidemiol Vol. 149, No. 8, 1999
Anti-Tick Calreticulin Antibody: A Biomarker of Tick Bites
scapularis, but the rTC used was obtained from A.
americanum. It is possible that anti-/. scapularis antibodies only recognize a subset of cross-reactive epitopes on the A. americanum derived calreticulin, thereby affecting the ability to detect intensity differences
based on immunoblotting.
In a previous study of a biologic marker of tick
exposure in the same study population (9), TEI was
shown to be a predictor of ATS A levels. In that study,
ATSA levels did not increase significantly over time,
in contrast to anti-rTC antibody levels. In the current
study, ATSA and anti-rTC antibody levels were poorly
correlated, although both were correlated with TEI.
The lack of correlation between the two biologic markers (r = -0.017 to -0.045 for the three visits) was found
to be influenced by six individuals with high levels of
anti-rTC antibody and low ATSA levels. When these
six subjects were dropped from the analysis, the
Pearson's r increased to 0.22 (p = 0.071). In five out of
these six subjects (83.8 percent), the biting tick was a
nymph; this value could be compared with the 55.6
percent of the remaining tick bite subjects who were
bitten by a nymph. It is possible that there are differences in calreticulin kinetics in the saliva of adult ticks
in contrast to nymphal ticks, accounting for this observation. A dose-response effect may also account for the
antibody response of these subjects; nymphal ticks
may secrete smaller amounts of saliva overall, resulting in a lower ATSA response, but the proportion of
calreticulin in the saliva may be higher, resulting in an
increased anti-rTC antibody response. This is speculative and further work is needed to characterize the
dynamics of calreticulin production in the salivary
glands of immature and adult ticks.
The sensitivity and specificity of the anti-rTC antibody serologic test can be compared with the sensitivity and specificity of ATSA determined in a previous
study. ATSA was found to have a sensitivity of 0.81
and a specificity of 0.56 to identify individuals bitten
by an /. scapularis that became engorged (TEI >3.4).
The current study revealed that anti-rTC antibody
seropositivity had a sensitivity of 0.50 and a specificity
of 0.86 to identify individuals after the same tick bite.
As opposed to ATSA, anti-rTC antibodies are raised
against a specific recombinant tick protein. Because
the rTC has no carbohydrate epitopes, and is a single
isolated tick protein, the amount of potential crossreactivity is greatly reduced. This would reduce the
number of false positive results due to cross-reactivity,
thus increasing the specificity of the assay. Tick calreticulin has also been shown to be undetected in tick
salivary glands until after approximately 3 days of
attachment (13). Anti-rTC antibody seropositivity
would therefore be expected in a smaller subset of tick
Am J Epidemiol
Vol. 149, No. 8, 1999
783
bite subjects, those with longer duration bites, and
would have far less potential for cross-reactivity
resulting in decreased sensitivity and increased specificity compared with ATSA seropositivity.
In terms of practical application, the two tests could
be used together in a parallel testing technique, producing results that would be over 80 percent sensitive
and specific for an /. scapularis bite which became
engorged. Anti-vector antibody tests would be particularly useful in situations where the anti-vector borne
pathogen test has low sensitivity. For example, serologic tests for anti-fi. burgdorferi antibody have sensitivities <50 percent in the early stages of Lyme disease. An assay (or combination of assays) with high
sensitivity and specificity for tick bites conferring
Lyme disease risk (duration >48 hours) may assist in
assessing risk of early Lyme disease. It is probable that
anti-arthropod saliva antibody assays are primarily
useful as research tools for the study of arthropod
exposure. Additional research is necessary to determine if these assays will have clinical utility. In a
patient presenting with symptoms compatible with a
tick-borne disease, a diagnostic test that confirms a
recent tick bite may assist the clinician. However, there
are several host-, vector-, and pathogen-related factors
that must first be elucidated. For example, it has been
demonstrated (19) that nymphal /. scapularis ticks can
transmit Lyme disease to mice in less than 48 hours,
and the dynamics of calreticulin in the saliva of
nymphal ticks has yet to be described.
Currently, the study of vector borne disease epidemiology lacks reliable biologic markers of arthropod exposure, and therefore relies heavily on selfreporting as a measure of exposure to arthropod
vectors, especially ticks. Self-reporting of tick bites
has been shown to be a poor indicator of tick bites
(18). A biologic marker of tick bites would allow for a
more accurate measure of tick exposure, both at the
population level and at the individual level. This would
improve the study of tick exposure in epidemiologic
studies, as well as providing a possible mechanism for
the estimation of individual risk of tickborne disease
after tick bites.
Anti-rTC antibody may be a more specific biomarker
of tick bites than ATSA, although ATSA has been
shown to have greater sensitivity. Used together, the
two assays may have the potential for improving the
detection of long-duration (>48 hours) tick bites in
human subjects through an adequate balance of sensitivity and specificity. The development of recombinant
tick salivary gland proteins for other tick species may
lead to even more specific biomarkers of tick bites and
an increase in the utility of anti-tick salivary protein
antibodies in the diagnosis of tickborne disease.
784
Sanders et al.
ACKNOWLEDGMENTS
This research was funded, in part, by Research Grant
AI31608 from MAID (to Dr. Schwartz), NIH Training
Grant 5-T32-ESO-7141 (to Dr. Sanders), and scholarship
funds received from Achievement Rewards for College
Scientists (ARCS) Inc. and the Delta Omega Honor Society
(both to Dr. Sanders).
The authors thank the Acarology Laboratory at Oklahoma
State University for supplying the ticks used in the study.
REFERENCES
1. Al-Ahdal MN, Al-Hussain K, Thorogood RJ, et al. Protein
constituents of mosquito saliva: studies on Culex molestus. J
Trop Med Hyg 1990;93:98-105.
2. Lello E, Peracoli MTS. Cell-mediated and humoral immune
responses in immunized and/or Dermatobia hominis infested
rabbits. Vet Parasitol 1993;47:129-38.
3. Willadsen P. Immunity to ticks. Adv Parasitol 1980;18:
293-313.
4. Sonenshine DE. Biology of ticks, vol. 1. New York: Oxford
University Press, 1991.
5. Ribeiro JMC. Role of saliva in blood feeding by arthropods.
Ann Rev Entomol 1987;32:463-78.
6. Nuttall PA, Jones LD, Labuda M, et al. Adaptations of
arboviruses to ticks. J Med Entomol 1994;31:1-9.
7. Binnington KC, Kemp DH. Role of tick salivary glands in
feeding and disease transmission. Adv Parasitol 1980; 18:
315-39.
8. Schwartz BS, Ford DP, Childs JE, et al. Anti-tick saliva antibody: biologic marker of tick exposure that is a risk factor for
Lyme disease seropositivity. Am J Epidemiol 1991;134:86-95.
9. Schwartz BS, Nadelman RB, Fish D, et al. Entomologic and
demographic correlates of anti-tick saliva antibody in a
prospective study of tick bite subjects in Westchester County,
New York. Am J Trop Med Hyg 1993;48:50-7.
10. Schwartz BS, Goldstein MD, Childs JE. A statewide survey of
antibodies to Borrelia burgdorferi and tick salivary gland proteins in New Jersey outdoor workers. Am J Public Health
1993;83:1746-8.
11. Schwartz BS, Sanchez J, Sanders ML, et al. Tick avoidance
behaviors associated with a decreased risk of anti-ticksalivary-gland-protein antibody seropositivity in military personnel exposed to Amblyomma americanum in Arkansas,
USA. Am J Trop Med Hyg 1996;55:410-16.
12. Sanders ML, Jaworski DC, Sanchez J, et al. Antibody to a
cDNA-derived calreticulin protein from Amblyomma americanum as a biomarker of tick exposure in humans. Am J Trop
Med Hyg 1998;59:279-85.
13. Jaworski DC, Simmen FA, Lamoreaux W, et al. A secreted calreticulin protein in ixodid tick (Amblyomma americanum) saliva. J Insect Physiol 1995 ;41:369-75.
14. White DJ, Chang HG, Benach JL, et al. The geographic spread
and temporal increase of the Lyme disease epidemic. JAMA
1991;261:1230-6.
15. Piesman J, Spielman A. Human babesiosis on Nantucket
Island: prevalence of Babesia microti in ticks. Am J Trop Med
Hyg 1980;29:742-6.
16. Dixon WJ. BMDP statistical software manual, vols 1 and 2.
Berkeley, CA: University of California Press, 1990.
17. Falco RC, Fish D, Piesman J. Duration of tick bites in a Lyme
disease endemic area. Am J Epidemiol 1996; 143:187-92.
18. Schwartz BS, Ribeiro JMC, Goldstein MD. Anti-tick antibodies: an epidemiologic tool in Lyme disease research. Am J
Epidemiol 1990;132:58-66.
19. Piesman J. Dynamics of Borrelia burgdorferi transmission
by nymphal Ixodes dammini ticks. J Infect Dis 1993;
167:1082-5.
Am J Epidemiol Vol. 149, No. 8, 1999