CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Sept. 2002, p. 1079–1084 Vol. 9, No.

5
1071-412X/02/$04.00⫹0 DOI: 10.1128/CDLI.9.5.1079–1084.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Cytokine Gene Expression by Peripheral Blood Leukocytes in Horses

Experimentally Infected with Anaplasma phagocytophila
Hyung-Yong Kim, Jason Mott, Ning Zhi, Tomoko Tajima, and Yasuko Rikihisa*
Department of Veterinary Biosciences, The Ohio State University, Columbus, Ohio 43210-1093
Received 19 November 2001/Returned for modification 4 May 2002/Accepted 21 June 2002

Human granulocytic ehrlichiosis (HGE), a tick-borne zoonosis, is caused by an obligatory intragranulocytic

bacterium, the HGE agent, a strain of Anaplasma phagocytophila. The equine model of HGE is considered
valuable in understanding pathogenic and immune mechanisms of HGE. In the present study, cytokine mRNA
expression by peripheral blood leukocytes (PBLs) in horses was examined during the course of infection by
intravenous inoculation of A. phagocytophila or by allowing feeding by infected ticks. The p44 genes encoding
the major outer membrane protein P44s of A. phagocytophila were detected by PCR in PBLs of all four horses
from 4 to 20 days postexposure. During the 20-day infection period, interleukin-1␤ (IL-1␤) and tumor necrosis
factor alpha (TNF-␣) mRNA expression was upregulated in PBLs of all four horses, and IL-8 mRNA expres-
sion was upregulated in three horses. Gamma interferon, IL-10, and IL-12 p35 mRNAs were weakly expressed
in only one horse each. IL-2, IL-4, IL-6, and IL-12 p40 mRNA expression , however, could not be detected in
the PBLs of any of the four horses. These results suggest that IL-1␤, TNF-␣, and IL-8 generation during A.
phagocytophila infection has a primary role in HGE pathogenesis and immunomodulation.

Human granulocytic ehrlichiosis (HGE) is characterized by
fever, chills, headache, myalgia, and laboratory findings includ-
ing leukopenia, anemia, and thrombocytopenia, as well as el-
evated liver enzyme activities (3, 8). Delayed treatment, mis-
diagnosis, and/or the presence of immunosuppression may
lead to a severe or fatal outcome (13, 21). HGE has been
increasingly recognized in the United States (27) and various
parts of Europe (20, 23, 30). HGE is caused by a strain of
Anaplasma phagocytophila (the HGE agent), a gram-negative
obligatory intragranulocytic bacterium. The white-footed
mouse (Peromyscus leucopus) is considered to be a major wild-
animal reservoir of A. phagocytophila in the northeastern
United States, and Ixodes spp. ticks, the vector of Borrelia
burgdorferi, the agent of Lyme disease, are considered to be the
vector (12, 14, 29, 34). Strains of A. phagocytophila have also
been known to cause tick-borne fever in ruminants in Europe
and to cause equine ehrlichiosis in the United States (it has
hence formerly been called Ehrlichia phagocytophila and Ehr-
lichia equi).
Pathogenesis and cellular immune responses of HGE are
not well defined. The presence of low levels of A. phagocyto-
phila in the blood of patients in the acute stage and autopsied
patients’ tissues (3, 21) and the nature of clinical signs and
laboratory findings suggest the involvement of proinflamma-
tory cytokines and chemokines in the pathogenesis. In vitro, we
found that A. phagocytophila strain HZ, isolated from an acute-
stage HGE patient, induced rapid and strong proinflammatory
cytokine (interleukin-1␤ [IL-1␤], tumor necrosis factor ␣
[TNF-␣], and IL-6) mRNA expression by human peripheral
blood leukocytes (PBLs) and monocytes within 2 h and protein
secretion within 24 h (16, 17). However, only IL-1␤ is upregu-
* Corresponding author. Mailing address: Department of Veteri-
nary Biosciences, College of Veterinary Medicine, The Ohio State
University, 1925 Coffey Rd., Columbus, OH 43210-1093. Phone: (614)
292-5661. Fax: (614) 292-6473. E-mail: rikihisa.1@osu.edu.
lated in neutrophils, and the mRNA of these three cytokines is
not upregulated in HL-60 cells. Within 2 h of incubation with
A. phagocytophila, IL-8, IL-10, gamma interferon (IFN-␥),
IL-2, and transforming growth factor ␤ mRNAs are not con-
sistently upregulated in human PBLs, suggesting that for these
cytokines, gene expression either is not induced or is induced
at later time points in vitro. Akkoyunlu et al. (2) reported IL-8
production by retinoic acid-treated HL-60 cells (human pro-
myelocytic leukemia cell line) after 24 h (negative at 12 h for
both mRNA and protein) and by human neutrophils 7 h
postincubation with A. phagocytophila in vitro. That report also
indicated that no IL-1␣, IL-1␤, or TNF-␣ is detected in the
culture supernatant of retinoic acid-treated and nontreated
HL-60 cells at 6 days after the addition of A. phagocytophila.
Klein et al. reported that in infected dimethyl sulfoxide-treated
HL-60 cells or human bone marrow cells, IL-8 and other che-
mokine levels, but not that of IL-1, IL-6, or TNF-␣ in the
culture medium, were significantly elevated at 48 h postinfec-
tion (18).
IL-8 levels are significantly increased in the blood of HGE
patients (2). Dumler et al. reported that IFN-␥ and IL-10 levels
are elevated in acute-phase (median, 4 days after onset) sera
compared with convalescent (median, 31 days after onset) sera
from HGE patients or normal controls whereas serum IL-1␤,
TNF-␣, and IL-4 levels are not elevated (9). Using C3H mice,
which do not show clinical signs and spontaneously clear the
infection by 15 days after intraperitoneal inoculation of A.
phagocytophila, splenic IL-12 and IFN-␥ mRNA and serum
IFN-␥ levels are significantly elevated starting at 2 days postin-
fection, although the clearance of A. phagocytophila is not
delayed in IFN-␥⫺/⫺ mice (1).
Considering apparently conflicting data in vitro using cell
lines, bone marrow cells, or human PBLs which were assayed
at varied time points of incubation or in vivo using a transient
mouse infection model or patients’ sera, an equine model of
HGE may be useful for clarifying the importance of these

1079
1080 KIM ET AL.
cytokines in HGE. Several reports have suggested that the
horse serves as a more useful infection and disease model of
HGE than the mouse, since A. phagocytophila isolated from
patients induces clinical signs in horses similar to those of
HGE and equine ehrlichiosis (7, 24, 25, 31, 32). However,
cytokine responses in horses infected with A. phagocytophila
have not been described so far. In the present study, we inves-
tigated cytokine mRNA expression by PBLs in horses follow-
ing transmission of A. phagocytophila by intravenous inocula-
tion or tick bite.
MATERIALS AND METHODS
A. phagocytophila. HZ strain A. phagocytophila, isolated from an HGE patient
(33), was propagated in HL-60 cells as previously described (15). The percentage
of infected cells was determined by Diff-Quik (Baxter Scientific Products, Obetz,
Ohio) staining as previously described (33).
Horse infection. Four horses purchased from Ohio horse farms were kept in
vector-proof stalls in the Equine Isolation Unit (College of Veterinary Medicine,
The Ohio State University). All horses were mixed breed. Horse 1 was a 2-year-
old male (380-kg body weight [BW]), horse 2 was a 5-year-old male (200-kg BW),
horse 3 was a 5-year-old male (180-kg BW), and horse 4 was a 3-year-old female
(170-kg BW). Prior to A. phagocytophila infection, these animals were confirmed
as seronegative and PCR negative for A. phagocytophila. Horses 1 and 2 were
intravenously (i.v.) inoculated with A. phagocytophila of high passage (HP; i.e.,
more than 50 passages in HL-60 cells) and low passage (LP; i.e., fewer than 10
passages), respectively, at 107 infected HL-60 cells (approximately 80% infected
cells) in 5 ml of RPMI 1640 medium. A total of 100 and a total of 47 infected
Ixodes scapularis adults were allowed to feed on horses 3 and 4, respectively. As
nymphs, these ticks were acquisition fed on experimentally infected DBA/2 mice
(37). Uninfected nymphs were kindly provided by S. R. Telford (Harvard School
of Public Health, Boston, Mass.). All horses were euthanized at 24 days postin-
oculation (PI) or postattachment (PA). The use of horses for this study has been
approved by the Ohio State University Institutional Animal Care and Use Com-
mittee.
Clinical evaluation, hematology, PBL preparation, morula examination, and
indirect fluorescent antibody (IFA) testing. Fever, depression, anorexia, diar-
rhea, leukopenia, laminitis, and other clinical signs were monitored daily for 20
days. For complete blood cell counts, blood samples were taken from the jugular
veins of horses prior to injection or tick attachment (day 0) and every 4 days until
20 days PI or PA. To prepare the PBL fraction, the blood samples, which were
collected in acid citrate dextrose anticoagulant tubes, were centrifuged at 500 ⫻
g for 5 min. Erythrocytes in the pellet were lysed in sterile 0.83% NH4Cl solution
for 3 min at room temperature. Cells were washed twice by centrifugation (500
⫻ g for 5 min) in phosphate-buffered saline (137 mM NaCl, 10 mM Na2HPO4,
2.7 mM KCl, 1.8 mM KH2PO4 [pH 7.2]). Viabilities of PBLs were ⬎98%, as
assessed by the trypan blue dye exclusion test. Intracytoplasmic microcolonies
(morulae) of A. phagocytophila were examined in PBLs by using Diff-Quik
staining and IFA using monoclonal antibody (MAb) 5C11 (15). The IFA test was
performed as previously described (33). The serum antibody titer was expressed
as the reciprocal of the highest dilution of serum that showed a positive reaction.
Reverse transcription (RT)-PCR for cytokine genes. Total RNA was extracted
from equine PBLs (107) by TRIzol reagent (GIBCO-BRL) and resuspended in
90 ␮l of diethyl pyrocarbonate-treated sterile water. As a positive control, PBLs
(107) of uninfected healthy horses were stimulated with either concanavalin A
(ConA; Sigma, St. Louis, Mo.) (10 ␮g/ml) or Escherichia coli lipopolysaccharide
(LPS) (10 ␮g/ml) in RPMI 1640 medium at 37°C for 4 h. Total RNA (2 ␮g) was
reverse transcribed in a 30-␮l reaction mixture containing 1⫻ reaction buffer (50
mM Tris-HCl [pH 8.3], 75 mM KCl, 3 mM MgCl2), 0.5 mM each deoxynucleo-
side triphosphate (dNTP), 1 U of RNase inhibitor (GIBCO-BRL), 1.5 ␮M
oligo(dT) primer, and 10 U of Moloney murine leukemia virus reverse transcrip-
tase (GIBCO-BRL) at 42°C for 1 h. The reaction was terminated by heating at
94°C for 5 min, and the cDNA was used in the PCR. The cDNA (2 ␮l) was
amplified in a 50-␮l reaction mixture containing 1⫻ PCR buffer (10 mM Tris-
HCl [pH 8.3] and 50 mM KCl), 1.5 mM MgCl2, 0.2 mM each dNTP, and 0.4 ␮M
(each) 3⬘ and 5⬘ primers (0.5 ␮M each primer for equine IL-8 mRNA). The
primers used for equine IL-1␤, IL-2, IL-4, IL-6, IL-10, IL-12 p35/40, TNF-␣,
IFN-␥, and ␤-actin mRNA detection were as previously described (10, 11).
Primer sequences for equine IL-8 mRNA expression, designed based on the
equine cDNA sequence (GenBank accession number AF062377), were kindly
provided by S. Giguere (Department of Large Animal Clinical Sciences, College
CLIN. DIAGN. LAB. IMMUNOL.
of Veterinary Medicine, University of Florida). The sequences were as follows:
sense, 5⬘-GACTTCCAAGCTGGCTGTTGC-3⬘, and antisense, 5⬘-GTCCTCTT
TAGAAACGCCTGC-3⬘. To reduce nonspecific priming, all PCRs were per-
formed by the hot-start method: Taq DNA polymerase (2 U/reaction; GIBCO-
BRL) was added after incubation of the mixture at 94°C for 5 min. PCR was 25
cycles (27 cycles for equine IL-8), consisting of denaturation at 94°C for 45 s,
annealing at 60°C (62°C for IL-8) for 45 s, and extension at 72°C for 2 min. The
final extension was for 7 min. These conditions were within the linear range for
PCR, as determined previously (16). PCR products (10 ␮l each) were electro-
phoresed in 1.5% agarose gel containing ethidium bromide (final concentration,
0.5 ␮g/ml) at 95 V for 1 h and photographed under UV illumination with a gel
video system (Gel Print 2000i; Biophotonics Corporation, Ann Arbor, Mich.).
DNA size markers (HaeIII fragments of ␾X174 replicative-form [RF] DNA;
GIBCO-BRL) providing bands from 1,353 to 72 bp were run in parallel.
PCR of p44 genes of A. phagocytophila. To detect A. phagocytophila DNA in
equine PBLs, nested PCR, based on the multigene family p44 genes encoding
immunodominant outer membrane protein P44s (36), was performed. The two
primer pairs p3726 (5⬘-GCTAAGGAATTAGCTTATGA-3⬘)-p4257 (5⬘-AGAA
GATCATAACAAGCATTG-3⬘) and p3761 (5⬘-CTGCTCKGCCAARACCTC-
3⬘)-p4183 (5⬘-CAATAGTYTTAGCTAGTAACC-3⬘) (K ⫽ T⫹G; R ⫽ A⫹G; Y
⫽ C⫹T [mixed bases]) (22) that are located in conserved regions of p44 genes
(Q. Lin, N. Zhi, H.-Y. Kim, H. Horowitz, G. Wormser, and Y. Rikihisa, 101st
Gen. Meet. Am. Soc. Microbiol., abstr. D-204, p. 319, 2001) were used. The first
and second PCR product sizes were 531 and 422 bp, respectively. The first PCR
was performed with 3 min of denaturation at 94°C followed by 27 cycles con-
sisting of 1 min each of denaturation at 94°C, annealing at 55°C, and extension
at 72°C. In the first cycle, each 50-␮l PCR mixture contained 2 ␮g of RNA, 5 ␮l
of 10⫻ reaction buffer, 0.2 mM each dNTP, 1.5 mM MgCl2, 1.25 U of Taq
polymerase, and 25 pmol of each primer of the primer pair p3726-p4257. The
second PCR was performed the same as the first PCR, except that the annealing
temperature was 52°C, 1 ␮l of the first PCR product was used as template, and
10 pmol of each primer of the primer pair p3761-p4183 was used.
RESULTS
Clinical signs and hematology. Horse 1, which had been
inoculated i.v. with HP A. phagocytophila, developed fever
(body temperature ⬎ 38.9°C; normal range, 37.5 to 38.6°C
[19]) during days 5 to 8 and 15 to 18 PI and thrombocytopenia
(⬍1.1 ⫻ 1011 platelets/liter; normal value for total equine
platelets, [2.2 ⫾ 0.93] ⫻ 1011/liter [6]) during days 4 to 8 PI.
Horse 2, inoculated i.v. with LP A. phagocytophila, developed
fever only at day 8 and thrombocytopenia (1.1 ⫻ 1011 platelets/
liter) and slight neutropenia (1.5 ⫻ 109 leukocytes/liter; normal
value, [4.4 ⫾ 2.0] ⫻ 109/liter [6]) on day 16. Horses 3 and 4 did
not show any significant clinical signs. Total red blood cell
counts (5.9 ⫻ 1012 to 13.9 ⫻ 1012 cells/liter; normal value, [7.5
⫾ 2.5] ⫻ 1012/liter [6]) of all four horses were within the
normal range during the 20 days PI. However, at 8 days PI,
variable sizes (anisocytosis) and abnormal shapes (poikilocy-
tosis) were detected in less than 1% of the red blood cells in all
4 horses. Leukocyte levels (5.1 ⫻ 109 to 16.7 ⫻ 109/liter; nor-
mal value, [7.6 ⫾ 3.0] ⫻ 109/liter [6]) and differential cell count
levels (1.0 ⫻ 109 to 7.9 ⫻ 109 lymphocytes/liter [normal value,
(3.0 ⫾ 1.9) ⫻ 109/liter] and 0.1 ⫻ 109 to 0.8 ⫻ 109 monocytes/
liter [normal value, (0.25 ⫾ 0.25) ⫻ 109/liter] [6]) were within
normal ranges in all four horses. The reactive lymphocytes
(less than 1%) were observed for horse 1 (days 8, 12, 16, and
20), for horse 2 (day 16), for horse 3 (days 16 and 20), and for
horse 4 (every day examined). Slight eosinophilia was observed
during days 12 to 16 in horse 3 (0.4 ⫻ 109 to 0.5 ⫻ 109
eosinophils/liter; normal value, [0.15 ⫾ 0.15] ⫻ 109/liter [6]).
In horses 1 and 2, which were inoculated i.v., examination of
PBLs by using Diff-Quik staining or IFA staining with MAb
5C11 revealed the presence of A. phagocytophila in neutrophils
during days 4 to 12 PI (horse 1, 1.0% of neutrophils infected on
VOL. 9, 2002 CYTOKINE GENE EXPRESSION IN EQUINE HGE MODEL 1081

FIG. 1. Microcolonies (morulae) of A. phagocytophila (black ar-

rowheads) found in peripheral blood neutrophils from HP A. phago-
cytophila-infected horse 1 by using Diff-Quik staining and IFA with
MAb 5C11 (insert, white arrowhead). Magnification, ⫻625.

day 4, 4.9% infected on day 8, and 2.2% infected on day 12;

horse 2, 1% of neutrophils infected on day 4 and 3% infected
during days 8 to 16) (Fig. 1). Morulae were not detected in the
granulocytes of horses 3 and 4 throughout 20 days. FIG. 2. p44 gene detection of A. phagocytophila in the PBLs. The
Detection of p44 genes of A. phagocytophila. The presence of levels of A. phagocytophila in PBLs in horses infected by i.v. inoculation
immunodominant major outer membrane protein P44s has and/or by attachment of infected ticks were examined by PCR using
p44 gene-specific primers. Input DNA was normalized by ␤-actin PCR.
been demonstrated in all strains of A. phagocytophila so far The PCR products were resolved on agarose gels containing ethidium
examined (22, 36). Since P44 proteins are encoded by a mul- bromide. M, DNA size markers (HaeIII fragments of ␾X174 RF
tigene family, we have developed a highly sensitive p44-based DNA); N, no template; P, positive-control DNA from cell-cultured A.
nested-PCR method for HGE diagnosis (Lin et al., abstr. 101st phagocytophila HZ.
Gen. Meet. Am. Soc. Microbiol., 2001). ␤-Actin DNA served
as a control for the amount of horse PBL DNA across the
samples. In all four horses, p44 genes were detected during TNF-␣, IL-6, IFN-␥, IL-8, IL-2, IL-4, IL-12 p35, IL-12 p40, and
days 4 to 20 PI (Fig. 2), indicating that ehrlichemia was estab- IL-10 mRNAs were clearly detectable in uninfected equine
lished in all four horses. Although the nested PCR used was PBLs, using the primers and our RT-PCR conditions (Fig. 4).
not quantitative, lack of progressive increase in band intensi- For horse 1 only, the results for all 10 cytokine and ␤-actin
ties following the day of infection in all four horses suggests an mRNAs examined are shown, and for the remaining horses,
immune mechanism to contain the infection. the results for only the cytokines whose mRNAs were ex-
IFA titers. In the two horses inoculated i.v. with HP and LP pressed and IL-8 and ␤-actin mRNAs are shown (Fig. 4).
A. phagocytophila, respectively, similar IFA immunoglobulin G Regardless of method of inoculation (i.e., intravenous inocu-
(IgG) titers were observed (Fig. 3). In horses 1 and 2, sero- lation or tick transmission), expression of IL-1␤ and TNF-␣
conversions (⬎1:20) were detected from day 8 and day 4 PI, mRNAs was induced in all four horses by A. phagocytophila
respectively. The maximum IgG titers of the 2 horses were infection (Fig. 4).
1:320 on days 12 and 16 PI, respectively. The IgG titers in
horse 3, which had been infected by attaching infected ticks,
were less than 1:20 throughout 20 days, and horse 4 showed a
peak IgG titer of 1:160 on day 12 PA (Fig. 3).
Equine cytokine mRNA expression. Expression of several
equine mRNAs (IL-1␤, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12
p35/40, TNF-␣, IFN-␥, and ␤-actin) by PBLs in the four horses
infected with A. phagocytophila by i.v. inoculation or by attach-
ing infected ticks was measured following the time course of
infection by RT-PCR at a linear range of 25 or 27 PCR cycles.
Contamination of RNA preparations with the genomic DNA
was negligible, because PCR products from “minus-reverse
transcriptase” negative controls were not detected in all spec-
FIG. 3. IFA titers, showing development of IgG antibodies against
imens. Constitutively expressed ␤-actin mRNA served as a A. phagocytophila in the four horses infected by intravenous inocula-
control for the amount of input RNA across the samples. As tion of HP or LP A. phagocytophila or by attachment of HP A. phago-
positive controls, ConA- and/or E. coli LPS-induced IL-1␤, cytophila-infected ticks.
1082 KIM ET AL.
FIG. 4. Induction of cytokine mRNAs in PBLs from four horses in-
fected with A. phagocytophila. Total RNA was extracted from equine
PBLs and subjected to RT-PCR. The amount of cDNAs was normalized
against ␤-actin mRNA in corresponding samples. M, DNA size markers
(HaeIII fragments of ␾X174 RF DNA); N, no template. Uninfected horse
PBLs were stimulated with ConA (C) or E. coli LPS (L) for 4 h in vitro.
The results shown are representative of more than three assays.
IL-8 mRNA expression was detected in horses 1, 2, and 4 but
not in horse 3. IFN-␥ mRNA expression was weakly induced
only in horse 2. A weak IL-10 mRNA expression level was also
detected in horse 1. In horse 4, IL-12 p35 mRNA was weakly
CLIN. DIAGN. LAB. IMMUNOL.
detected (Fig. 4). Expression of IL-2, IL-4, IL-6, and IL-12 p40
mRNAs was not detectable in the PBLs of any infected horses.
DISCUSSION
In our equine model of HGE, only IL-1␤ and TNF-␣
mRNAs were consistently upregulated in all four horses, re-
gardless of A. phagocytophila cell culture passage number or
route of infection. This suggests that A. phagocytophila-induced
mild fever, neutropenia, and thrombocytopenia are mediated
in horses through the induction of these two major proinflam-
matory cytokines, which are known to cause these changes
(28). In our previous study, A. phagocytophila or rP44 induced
IL-1␤, TNF-␣, and IL-6 mRNAs within 2 h PI and induced
protein within 24 h PI in human PBLs in vitro in a dose-
dependent manner (16, 17). However, IL-6 mRNA expression
was not detectable in our equine model. The lack of IL-6
mRNA induction by PBLs in infected horses may be due to low
numbers of A. phagocytophila in the blood of horses, since IL-6
mRNA induction in vitro requires 10-fold higher levels of A.
phagocytophila than TNF-␣ or IL-1␤ mRNA induction does
(16). The lack of IL-6 mRNA induction may be also related to
the relatively mild nature of clinical signs of our horses. Coin-
fection of C3H mice with B. burgdorferi and A. phagocytophila
was reported to increase the levels of A. phagocytophila and
elevate IL-6 levels in the sera at 2 weeks postinfection beyond
those of infection with B. burgdorferi alone (35). One report
observed that in sera of acute-phase patients, the levels of
IL-1␤, TNF-␣, and IL-6 are not elevated (9). Since, unlike
those of experimentally infected animals, it is difficult to pre-
serve these rapid-turnover cytokines in the specimens of hu-
man patients (5), it is premature to dismiss the role these
cytokines play in HGE pathogenesis.
Acute-phase HGE patient sera contain significantly higher
levels of IFN-␥ than convalescent-phase sera (9). A. phagocy-
tophila can transiently establish infection in C3H mice without
any clinical signs, and these mice develop significant IFN-␥
levels in serum during days 2 to 8 postintraperitoneal inocula-
tion with A. phagocytophila-containing blood from infected
SCID mice (1). Although in IFN-␥⫺/⫺ mice, initial infection
levels with A. phagocytophila are greater than in control mice,
the speed of clearance of A. phagocytophila does not differ (1).
In our study, IFN-␥ mRNA expression was seen only in one
horse. This finding may be related to the lack of spontaneous
clearance of A. phagocytophila from the blood of all horses at
even 20 days PI or PA. The role of IFN-␥ in clearance of A.
phagocytophila remains to be studied.
In the present study, IL-8 mRNA expression was upregu-
lated in 3 horses; however, horse 3 had the weakest cytokine
response overall and did not show IL-8 mRNA expression.
Recently, Akkoyunlu et al. (2) proposed that IL-8 generated by
neutrophils in response to A. phagocytophila facilitates the
infection. Klein et al. reported that in dimethyl sulfoxide-
treated infected HL-60 cells or human bone marrow cells, IL-8
and other chemokine levels, but not that of IL-1, IL-6, or
TNF-␣ in the culture medium, were significantly elevated at
48 h postinfection (18). Since findings, including those of our
present horse study, indicate that IL-8 mRNA is not upregu-
lated at early stages of infection in vitro and in vivo, IL-8 does
not seem to be critical for the initial establishment of infection.
VOL. 9, 2002 CYTOKINE GENE EXPRESSION IN EQUINE HGE MODEL
Overall clinical signs for our horses, which were inoculated
with a strain from an HGE patient in the state of New York,
were milder than those reported with studies using the BDS or
Webster strain from Wisconsin (4, 25, 31) but rather similar to
those using inoculation with a New York strain as reported by
Chang et al. (7). In the study of Pusterla et al. (32), at day 16
after i.v. inoculation and at day 18 after inoculation by tick
attachment, A. phagocytophila Webster and BDS strains be-
came undetectable by real-time quantitative PCR assays, based
on the presence of the 16S rRNA gene in the blood of horses.
However, A. phagocytophila HZ strain p44 genes were detected
at day 20 of infection in four horses in the present study,
suggesting that A. phagocytophila establishes subclinical persis-
tent infection in horses, as suggested by Chang et al. (7). A
pioneering phylogenic study, comparing different A. phagocy-
tophila strains from patients and animals in the United States
and Europe by using the ank gene as a marker, showed that
Wisconsin strains are closer to Ehrlichia equi (A. phagocyto-
phila, horse isolate, 99.18 to 99.24% amino acid identity) than
New York strains (97.90 to 97.92% amino acid identity) (26).
Thus, it is possible that Wisconsin strains are more pathogenic
to horses but cleared more rapidly than New York strains.
Disease severity in horses inoculated i.v. with A. phagocyto-
phila strain Webster was reported to be dependent on A.
phagocytophila passage numbers in culture (31). This finding
appears to be among those of studies that are not evident
unless larger number of animals are tested, since we did not
observe clear differences in clinical manifestations between
horses infected with HP and LP A. phagocytophila HZ strains.
Using a real-time quantitative PCR assay, Pusterla et al. re-
ported 55-fold-lower initial ehrlichial load in PBLs from tick-
infected horses than in those from horses inoculated i.v. (32).
In the same study, however, higher ehrlichial levels in the
blood of horses infected by tick attachment were observed at 7
days PI. According to the findings based on the use of PCR
with p44 genes, our study appears to support their observation
with respect to the lower rate of increase of ehrlichial growth
with tick transmission than with intravenous inoculation. Pus-
terla et al. (32) did not see significant difference in antibody
titer at day 30 PI between intravenous and tick transmission. In
the present study, antibody responses were lower by tick at-
tachment than by intravenous inoculation. More studies are
needed to learn whether tick inoculation modulates host im-
mune response to facilitate the infection.
ACKNOWLEDGMENTS
This research was supported by grants AI30010 and AI40934 from
the National Institutes of Health.
We thank Roger W. Stich and Debra Grove at the Ohio State
University for technical assistance with tick attachment and S. R.
Telford at Harvard School of Public Health, Boston, Mass., for unin-
fected nymphs. We thank Quan Lin at the Ohio State University for
helping in extraction of DNA from blood samples.
REFERENCES
1. Akkoyunlu, M., and E. Fikrig. 2000. Gamma interferon dominates the mu-
rine cytokine response to the agent of human granulocytic ehrlichiosis and
helps to control the degree of early rickettsemia. Infect. Immun. 68:1827–
1833.
2. Akkoyunlu, M., S. E. Malawista, J. Anguita, and E. Fikrig. 2001. Exploita-
tion of interleukin-8-induced neutrophil chemotaxis by the agent of human
granulocytic ehrlichiosis. Infect. Immun. 69:5577–5588.
3. Bakken, J. S., J. Krueth, C. Wilson-Nordskog, R. L. Tilden, K. Asanovich,
1083
and J. S. Dumler. 1996. Clinical and laboratory characteristics of human
granulocytic ehrlichiosis. JAMA 275:199–205.
4. Barlough, J. E., J. E. Madigan, E. Derock, J. S. Dumber, and J. S. Bakken.
1995. Protection against Ehrlichia equi is conferred by prior infection with
the human granulocytic ehrlichia (HGE agent). J. Clin. Microbiol. 33:3333–
3334.
5. Beutler, B. A., I. W. Milsark, and A. Cerami. 1985. Cachectin/tumor necrosis
factor: production, distribution, and metabolic fate in vivo. J. Immunol.
135:3972–3977.
6. Beutler, E., M. A. Lichtman, B. S. Coller, and T. J. Kipps. 1995. Williams
hematology: examination of the blood, p. 8–15; comparative hematology, p.
77–85. McGraw-Hill, Inc., New York, N.Y.
7. Chang, Y.-F., V. Novosel, E. Dubovi, S. J. Wong, F. K. Chu, C. F. Chang, F.
Del Piero, S. Shin, and D. H. Lein. 1998. Experimental infection of the
human granulocytic ehrlichiosis agent in horses. Vet. Parasitol. 78:137–145.
8. Chen, S.-M., J. S. Dumler, J. S. Bakken, and D. H. Walker. 1994. Identifi-
cation of a granulocytotropic Ehrlichia species as the etiologic agent of
human disease. J. Clin. Microbiol. 32:589–595.
9. Dumler, J. S., E. R. Trigiani, J. S. Bakken, M. E. Aguero-Rosenfeld, and
G. P. Wormser. 2000. Serum cytokine responses during acute human gran-
ulocytic ehrlichiosis. Clin. Diagn. Lab. Immunol. 7:6–8.
10. Franchini, M., U. Gilli, M. K. Akens, R. V. Fellenberg, and V. Bracher. 1998.
The role of neutrophil chemotactic cytokines in the pathogenesis of equine
chronic obstructive pulmonary disease (COPD). Vet. Immunol. Immuno-
pathol. 66:52–65.
11. Giguere, S., and J. F. Prescott. 1999. Quantitation of equine cytokine mRNA
expression by reverse transcription-competitive polymerase chain reaction.
Vet. Immunol. Immunopathol. 67:1–15.
12. Hodzic, E., D. Fish, C. M. Maretzki, A. M. De Silva, S. Feng, and S. W.
Barthold. 1998. Acquisition and transmission of the agent of human granu-
locytic ehrlichiosis by Ixodes scapularis ticks. J. Clin. Microbiol. 36:3574–
3578.
13. Jahangir, A., C. Kolbert, W. Edwards, P. Mitchell, J. S. Dumler, and D. H.
Persing. 1998. Fatal pancarditis associated with human granulocytic ehrli-
chiosis in a 44-year-old man. Clin. Infect. Dis. 27:1424–1427.
14. Katavolos, P., P. M. Armstrong, J. E. Dawson, and S. R. Telford III. 1998.
Duration of tick attachment required for transmission of granulocytic ehr-
lichiosis. J. Infect. Dis. 177:1422–1425.
15. Kim, H.-Y., and Y. Rikihisa. 1998. Characterization of monoclonal antibod-
ies to the 44-kilodalton major outer membrane protein of the human gran-
ulocytic ehrlichiosis agent. J. Clin. Microbiol. 36:3278–3284.
16. Kim, H.-Y., and Y. Rikihisa. 2000. Expression of interleukin-1␤, tumor
necrosis factor-␣, and interleukin-6 in human peripheral blood leukocytes
exposed to human granulocytic ehrlichiosis agent or recombinant major
surface protein P44. Infect. Immun. 68:3394–3402.
17. Kim, H.-Y., and Y. Rikihisa. 2002. Roles of p38 mitogen-activated protein
kinase, NF-␬B, and protein kinase C in proinflammatory cytokine mRNA
expression by human peripheral blood leukocytes in response to Anaplasma
phagocytophila. Infect. Immun. 70:4132–4141.
18. Klein, M. B., S. Hu, C. C. Chao, and J. L. Goodman. 2000. The agent of
human granulocytic ehrlichiosis induces the production of myelosuppressing
chemokines without induction of proinflammatory cytokines. J. Infect. Dis.
182:200–205.
19. Kobluk, C. N., T. R. Ames, and R. J. Geor. 1995. The horse disease and
clinical management: equine gastrointestinal surgery, p. 329–361. W. B.
Saunders Company, Philadelphia, Pa.
20. Lebech, A. M., K. Hansen, P. Pancholi, L. M. Sloan, J. M. Magera, and D. H.
Persing. 1998. Immunoserologic evidence of human granulocytic ehrlichiosis
in Danish patients with Lyme neuroborreliosis. Scand. J. Infect. Dis. 30:173–
176.
21. Lepidi, H., J. E. Bunnell, M. E. Martin, J. E. Madigan, S. Stuen, and J. S.
Dumler. 2000. Comparative pathology and immunohistology associated with
clinical illness after Ehrlichia phagocytophila-group infections. Am. J. Trop.
Med. Hyg. 62:29–37.
22. Lin, Q., N. Zhi, N. Ohashi, H. W. Horowitz, M. E. Aguero-Rosenfeld, J.
Raffalli, G. P. Wormser, and Y. Rikihisa. 2002. Analysis of sequences and
loci of p44 homologs expressed by Anaplasma phagocytophila in acutely
infected patients. J. Clin. Microbiol. 40:2981–2988.
23. Lotric-Furlan, S., T. Avsic-Zupanc, M. Petrovec, W. L. Nicholson, J. W.
Sumner, J. E. Childs, and F. Strle. 2001. Clinical and serological follow-up
of patients with human granulocytic ehrlichiosis in Slovenia. Clin. Diagn.
Lab. Immunol. 8:899–903.
24. Madigan, J. E., J. E. Barlough, J. S. Dumler, N. S. Schankman, and E.
DeRock. 1996. Equine granulocytic ehrlichiosis in Connecticut caused by an
agent resembling the human granulocytotropic ehrlichia. J. Clin. Microbiol.
34:434–435.
25. Madigan, J. E., P. J. Richter, R. B. Kimsey, J. E. Barlough, J. S. Bakken, and
J. S. Dumler. 1995. Transmission and passage in horses of the agent of
human granulocytic ehrlichiosis. J. Infect. Dis. 172:1141–1144.
26. Massung, R. F., J. H. Owens, D. Ross, K. D. Reed, M. Petrovec, A. Bjoers-
dorff, R. T. Coughlin, G. A. Beltz, and C. Murphy. 2000. Sequence analysis
of the ank gene of granulocytic ehrlichiae. J. Clin. Microbiol. 38:2917–2922.
1084 KIM ET AL.
27. McQuiston, J. H., C. D. Paddock, R. C. Holman, and J. E. Childs. 1999. The
human ehrlichioses in the United States. Emerg. Infect. Dis. 5:635–642.
28. Morris, D. D. 1991. Endotoxemia in horses: a review of cellular and humoral
mediators involved in its pathogenesis. J. Vet. Intern. Med. 5:167–181.
29. Pancholi, P., C. P. Kolbert, P. D. Mitchell, K. D. Reed, Jr., J. S. Dumler, J. S.
Bakken, S. R. Telford III, and D. H. Persing. 1995. Ixodes dammini as a
potential vector of human granulocytic ehrlichiosis. J. Infect. Dis. 172:1007–
1012.
30. Petrovec, M., S. Lotric-Furlan, T. Avsic-Zupanc, F. Strle, P. Brouqui, V.
Roux, and J. S. Dumler. 1997. Human disease in Europe caused by a
granulocytic Ehrlichia species. J. Clin. Microbiol. 35:1556–1559.
31. Pusterla, N., J. E. Madigan, K. M. Asanovich, J. S. Chae, E. Derock, C. M.
Leutenegger, J. B. Pusterla, H. Lutz, and J. S. Dumler. 2000. Experimental
inoculation with human granulocytic ehrlichia agent derived from high- and
low-passage cell culture in horses. J. Clin. Microbiol. 38:1276–1278.
32. Pusterla, N., C. M. Leutenegger, J.-S. Chae, H. Lutz, R. B. Kimsey, J. S.
Dumler, and J. E. Madigan. 1999. Quantitative evaluation of ehrlichial
burden in horses after experimental transmission of human granulocytic
ehrlichia agent by intravenous inoculation with infected leukocytes and by
infected ticks. J. Clin. Microbiol. 37:4042–4044.
CLIN. DIAGN. LAB. IMMUNOL.
33. Rikihisa, Y., N. Zhi, G. P. Wormser, B. Wen, H. W. Horowitz, and K. E.
Hechemy. 1997. Ultrastructural and antigenic characterization of a granulo-
cytic ehrlichiosis agent directly isolated and stably cultivated from a patient
in New York State. J. Infect. Dis. 175:210–213.
34. Telford, S. R., III, J. E. Dawson, P. Katavolos, C. K. Warner, C. P. Kolbert,
and D. H. Persing. 1996. Perpetuation of the agent of human granulocytic
ehrlichiosis in a deer tick-rodent cycle. Proc. Natl. Acad. Sci. USA 93:6209–
6214.
35. Thomas, V., J. Anguita, S. W. Barthold, and E. Fikrig. 2001. Coinfection with
Borrelia burgdorferi and the agent of human granulocytic ehrlichiosis alters
murine immune responses, pathogen burden, and severity of Lyme arthritis.
Infect. Immun. 69:3359–3371.
36. Zhi, N., N. Ohashi, and Y. Rikihisa. 1999. Multiple p44 genes encoding
major outer membrane proteins are expressed in the human granulocytic
ehrlichiosis agent. J. Biol. Chem. 274:17828–17836.
37. Zhi, N., N. Ohashi, T. Tajima, J. Mott, R. W. Stich, D. Grover, S. R. Telford
III, Q. Lin, and Y. Rikihisa. 2002. Transcript heterogeneity of p44 multigene
family in human granulocytic ehrlichiosis agent transmitted by ticks. Infect.
Immun. 70:1175–1184.