Kaitlin K. Rainwater, MPH,1,4 Jacob IJdo,2 MD, PhD, Ana Capuano, MPS,3 Mary J.R. Gilchrist, PhD,1 James S. Gill, MD, PhD1

1
Kaitlin K. Rainwater, MPH,1,4 Jacob IJdo,2 MD,
PhD, Ana Capuano, MPS,3 Mary J.R. Gilchrist,
PhD,1 James S. Gill, MD, PhD1 1University of Iowa
Hygienic Laboratory, Iowa City, IA 2University
of Iowa Carver College of Medicine, Iowa City,
IA 3University of Iowa Center for Emerging
Infectious Diseases, Iowa City, IA 4Current
Affiliation Plum Island Animal Disease Center,
Agricultural Research Service, USDA, Greenport, NY

• Anaplasma phagocytophilum is primarily maintained
  in nature by a rodent-tick cycle with the
  white-footed mouse (Peromyscus leucopus) serving
  as the rodent reservoir and Ixodes scapularis,
  the deer tick, as the principal tick vector
  (Magnarelli et al. 1995, Magnarelli et al. 1997,
  Telford et al. 1996). The preferred host of adult
  I. scapularis is white-tailed deer (WTD
  Odocoileus virginianus) (Lane et al. 1991),
  which have been shown to serve as competent
  sentinel animals for the surveillance of A.
  phagocytophilum (Belongia et al. 1997, Magnarelli
  et al. 1999, Magnarelli et al. 2004, Tate et al.
  2005). While I. scapularis populations are known
  to exist in the state of Iowa, they have
  historically been concentrated in the
  northeastern and eastern areas of the state
  (Bartholomew et al. 1995). The assumption that A.
  phagocytophilum is present in these regions of
  Iowa is consistent with the fact that the state
  shares its northeastern and northern borders with
  Wisconsin and Minnesota, respectively, where the
  initial human cases of HGA were reported in 1994
  and continue to be reported at consistently
  elevated levels (Chen et al. 1994, Demma et al.
  2005).
• Recent evidence suggests that I. scapularis is
  increasing in numbers and is extending its range
  into Iowa (Lingren et al. 2005). Although there
  are no studies specifically investigating the
  presence of A. phagocytophilum in Iowa, five
  cases of HGA have been reported in Iowa since the
  year 2000 (University of Iowa Hygienic
  Laboratory, unpublished data). While this may be
  an indication of increased awareness and
  recognition of HGA by physicians, it may also
  reflect a geographic expansion beyond the two
  endemic foci of the Northeast and the upper
  Midwest and a true increase in incidence rates of
  the disease.
• Main objectives
• Investigate if A. phagocytophilum is present in
  Iowa by conducting a serosurvey of WTD across the
  state in 2004
• Compare seropositivity rates from a single,
  common location in 1999 and 2004
• Assess the effects of age and gender on
  seropositivity rates

To our knowledge this study is the first to
report that WTD in the state of Iowa have been
exposed to A. phagocytophilum, thus expanding the
previously identified geographic range of this
organism beyond the endemic adjacent area of
Minnesota and Wisconsin (Demma et al. 2005).
Eleven of the 13 sites in this serosurvey
revealed at least one positive serum sample by
ELISA or WB. Eight of the 13 sites showed
positive sera by both methods. Interestingly,
comparison of seropositivity rates at F.W. Kent
Park (Site 3) and the Iowa City area (Site 4)
were drastically different despite the close
proximity of the two sites (13 miles). Seropreval
ence of WTD in F.W. Kent Park was 76.9 (95 CI
54 to 99.8) by ELISA, the highest of any site
surveyed. However, seroprevalence at the Iowa
City site in the same year was only 1.5 (95 CI
0 to 3.7) using the same testing method. This
difference is significant despite the relatively
small F.W. Kent Park sample (n13). This may
suggest that some sites are relatively isolated
and differences in the surrounding environment
may affect transmission of A. phagocytophilum.
While the area in which Iowa City samples were
collected is more urban with heavy traffic,
housing, and businesses, F.W. Kent Park is a
densely wooded, rural area that may have
significant populations of I. scapularis and P.
leucopus. We intend to perform future studies in
this area that will include dragging and flagging
for ticks and small mammal capture in order to
determine the significance of these
populations. Although the data do not suggest a
significant adjusted odds ratio when comparing
the eastern and western regions of Iowa, part of
the variation between regions was due to age
variation (Table 3). The analysis indicates a
higher seroprevalence rate among adult WTD in
Iowa. In fact, among eastern Iowa adult WTD, we
observed increased odds compared to adults of
other regions. We expect the seroprevalence rate
in central Iowa adult WTD to rise over the next
several years in conjunction with the recent
identification of I. scapularis populations in
this region of the state (Lingren et al.
2005). Importantly, A. phagocytophilum is now
the second ehrlichial agent to be found in Iowa.
Previously, WTD were found to have antibodies to
Ehrlichia chaffeensis (Mueller-Anneling et al.
2000). Furthermore, Borrelia burgdorferi, the
causative agent of Lyme disease, has been
identified in I. scapularis ticks in Iowa
(Lingren et al. 2005) and several
laboratory-confirmed cases of Lyme disease are
reported in Iowa each year (University of Iowa
Hygienic Laboratory, unpublished data). These
observations have important public health
implications of which health care providers in
the state of Iowa should be aware.
Figure 1. Map of Iowa indicating sites from which
WTD serum samples were collected in 2004. Samples
were collected from Site 4 in 1999.

• Sera collection
• 628 WTD at thirteen controlled deer harvest sites
  across Iowa (Figure 1, Table 1) during September
  2004 through February 2005.
• 282 previously untested serum samples collected
  in a similar manner from Iowa City (Site 4,
  Figure 1) during the fall of 1999.
• Age group (fawn, yearling, or adult) and sex
  information were determined.
• Enzyme-linked ImmunoSorbent Assay (ELISA)
• Performed on all 910 samples as previously
  described with changes (IJdo et al. 1999).
  Briefly, the p44 antigen was generated as a p44
  fusion protein with maltose binding protein (MBP)
  produced in Escherichia coli and purified by
  affinity column chromatography. Ninety-six-well
  polystyrene plates were coated with p44-MBP,
  blocked with 5 nonfat milk solution, and washed
  with PBS-Tween. Sera diluted in PBS were added to
  duplicate wells, plates incubated and washed as
  described above. HRP-conjugated rabbit anti-deer
  IgG was added, plates incubated and washed as
  described above. TMB and 1N H2SO4 was added and
  absorbance measured at 405nm. Absorbance values
  of ?0.300 were recorded as positive. Serum
  samples at four sites in southwestern Iowa where
  I. scapularis populations have not been reported
  (Bartholomew et al. 1995, Lingren et al. 2005)
  were used to calculate cutoff values based on
  three standard deviations (SD) above the mean.
• Western Immunoblotting (WB)
• Performed on all ELISA-positive serum samples
  (n62) and on a random selection of 36
  ELISA-negative serum samples as previously
  described with minor variations (IJdo et al.
  1999). Briefly, cell lysates of HL-60 cells
  infected with the NCH-1 strain of A.
  phagocytophilum (Telford et al. 1996) were placed
  in sample buffer, heated, and separated on a 12
  SDS-PAGE gel. Proteins transferred to
  nitrocellulose and unbound sites were blocked
  using 3 bovine serum albumin (BSA). Blots were
  cut into strips and probed with deer sera diluted
  in PBS. Strips were washed three times with
  PBS-Tween, placed in HRP-conjugated rabbit
  anti-deer immunoglobulin G diluted in PBS-Tween,
  washed as described above, placed in SuperSignal
  West Femto Maximum Sensitivity Substrate, and
  exposed to Xray film. The presence of the p44
  band was required as evidence of a positive
  immunoblot result (Figure 2).
• Statistical Analysis
• A two-sided Fishers exact test was used in the
  initial analysis of age, gender, location and
  year of collection. Binomial intervals were
  computed for risk factors for the crude
  prevalence of A. phagocytophilum infection in
  WTD. For data collected in 2004, unadjusted and
  adjusted odds ratios were computed for each risk
  factor using logistic regression. A final
  multivariable model was defined using a saturated
  model and manual backward elimination.
  Ninety-five percent confidence intervals (95 CI)
  were calculated at each step. Analyses were
  performed using SAS software version 9.1.

Figure 2. Representative immunoblots of deer sera
collected in Iowa showing reactivities to the p44
antigen of A. phagocytophilum. Molecular mass
standards are shown in kilodaltons along the
left. Lanes marked with (-) or () are negative
and positive controls, respectively. Lanes 2 and
3 show WTD serum samples with antibodies to the
p44 antigen, and lanes 1 and 4 are negative serum
samples.
We thank Thomas Gahan of the University of Iowa
Hygienic Laboratory for providing relevant data
and technical assistance Adam Carlson of the
University of Iowa Carver College of Medicine for
laboratory assistance and Dr. Louis Magnarelli
and Tia Mastrone of the Connecticut Agricultural
Experiment Station for providing WTD serum
controls. K.K.R. was supported by the
Association of Public Health Laboratories and
Centers for Disease Control and Prevention
through an appointment to the Emerging Infectious
Diseases Fellowship Program.