Prevalence of Ehrlichia canis (Rickettsiales: Anaplasmataceae) in dogs and Rhipicephalus sanguineus (Acari: Ixodidae) ticks from Brazil.

Abstract

The current study evaluated the prevalence of Ehrlichia canis Donatien and Lestoquard in domestic dogs, Canis familiaris L., and Rhipicephalus sanguineus (Latreille) (Acari: Ixodidae) ticks from different areas of Brazil. In Monte Negro County (state of Rondônia, Brazilian western Amazon), the indirect immunofluorescence assay detected E. canis-reactive antibodies (titer > or = 40) in 58/153 (37.9%) and 40/161 (24.8%) dogs from the urban and rural areas, respectively. These values were significantly different between the two areas. Ticks from a household in the urban area of Monte Negro, and from households in three other localities (162-165 adult ticks per household) in the state of São Paulo (SP) were tested by polymerase chain reaction (PCR) targeting an erlichial dsb gene fragment. The prevalence of infected ticks (given as minimal infection rate) was 2.3, 6.2, and 3.7% for populations 1 (Monte Negro), 2 (Jundiaí, SP), and 3 (São Paulo I, SP), respectively, which were statistically similar. In contrast, no infected tick was detected in population 4 (São Paulo II, SP). DNA sequences were determined for some of the PCR products generated from ticks and dogs from populations 1-3, being all identical to each other and to available sequences of E. canis in GenBank. These results reinforce previous records of E. canis-infecting dogs in Brazil. Natural infection of R. sanguineus ticks by E. canis is reported for the first time in Brazil, where this tick is the commonest species infesting dogs.

Full text

Page 1

VECTOR-BORNE DISEASES, SURVEILLANCE, PREVENTION
Prevalence of Ehrlichia canis (Rickettsiales: Anaplasmataceae) in Dogs
and Rhipicephalus sanguineus (Acari: Ixodidae) Ticks from Brazil
DANIEL M. AGUIAR,1,2GUACYARA T. CAVALCANTE,2ADRIANO PINTER,2,3
SOLANGE M. GENNARI,2LUIS MARCELO A. CAMARGO,4,5AND MARCELO B. LABRUNA2,6
J. Med. Entomol. 44(1): 126Ð132 (2007)
ThecurrentstudyevaluatedtheprevalenceofEhrlichiacanisDonatienandLestoquardin
domestic dogs, Canis familiaris L., and Rhipicephalus sanguineus (Latreille) (Acari: Ixodidae) ticks from
different areas of Brazil. In Monte Negro County (state of Rondo ˆnia, Brazilian western Amazon), the
indirectimmunoßuorescenceassaydetectedE.canis-reactiveantibodies(titer?40)in58/153(37.9%)and
40/161(24.8%)dogsfromtheurbanandruralareas,respectively.ThesevaluesweresigniÞcantlydifferent
between the two areas. Ticks from a household in the urban area of Monte Negro, and from households
inthreeotherlocalities(162Ð165adultticksperhousehold)inthestateofSa ˜oPaulo(SP)weretestedby
polymerasechainreaction(PCR)targetinganerlichialdsbgenefragment.Theprevalenceofinfectedticks
(given as minimal infection rate) was 2.3, 6.2, and 3.7% for populations 1 (Monte Negro), 2 (Jundiaõ ´, SP),
and 3 (Sa ˜o Paulo I, SP), respectively, which were statistically similar. In contrast, no infected tick was
detectedinpopulation4(Sa ˜oPauloII,SP).DNAsequencesweredeterminedforsomeofthePCRproducts
generated from ticks and dogs from populations 1Ð3, being all identical to each other and to available
sequences of E. canis in GenBank. These results reinforce previous records of E. canis-infecting dogs in
Brazil.NaturalinfectionofR.sanguineusticksbyE.canisisreportedfortheÞrsttimeinBrazil,wherethis
tick is the commonest species infesting dogs.
ABSTRACT
RESUMO
emca ˜esdome ´sticos,CanisfamiliarisL.,ecarrapatosRhipicephalussanguineus(Latreille)(Acari:Ixodidae)
em diferentes a ´reas do Brasil. Pela Reac ¸a ˜o de Imunoßuoresce ˆncia Indireta (RIFI) (tõ ´tulo ?40), foram
detectados anticorpos reativos para E. canis em 37.9% de 153 ca ˜es da a ´rea urbana e 24.8% de 161 ca ˜es da
a ´rea rural do municõ ´pio de Monte Negro, estado de Rondo ˆnia, Amazo ˆnia brasileira. Carrapatos R. san-
guineus (162 a 165 carrapatos adultos por populac ¸a ˜o) de uma reside ˆncia da a ´rea urbana de Monte Negro
e de outras tre ˆs localidades do estado de Sa ˜o Paulo foram testados por PCR para um fragmento do gene
dsb de Ehrlichia. A prevale ˆncia de carrapatos infectados [avaliada como Prevale ˆncia Mõ ´nima de Infecc ¸a ˜o
(PMI)] foi de 2.3, 6.2, e 3.7% para as populac ¸o ˜es 1 (Monte Negro), 2 (Jundiaõ ´, SP), e 3 (Sa ˜o Paulo I, SP),
respectivamente,sendoestesresultadossimilares.Nenhumcarrapatoinfectadofoidetectadonapopulac ¸a ˜o
4 (Sa ˜o Paulo II, SP). Os produtos da PCR de alguns carrapatos e ca ˜es das populac ¸o ˜es 1, 2 e 3 foram
sequ ¨enciados, sendo as sequ ¨e ˆncias obtidas ide ˆnticas entre si e a ` sequ ¨e ˆncia de E. canis disponõ ´vel no
GenBank. Estes resultados reforc ¸am estudos anteriores que relataram a infecc ¸a ˜o por E. canis em ca ˜es do
Brasil,contudorelatapelaprimeiraveznoBrasilainfecc ¸a ˜onaturalporE.canisemcarrapatosR.sanguineus,
tido como o principal carrapato de ca ˜es no paõ ´s.
Opresenteestudoavaliouaprevale ˆnciadainfecc ¸a ˜oporEhrlichiacanisDonatieneLestoquard
KEY WORDS
Ehrlichia canis, Rhipicephalus sanguineus, dogs, Brazil, prevalence
Ehrlichia canis Donatien and Lestoquard is a cosmo-
politan tick-transmitted intracellular bacterium, the
etiological agent of canine monocytic ehrlichiosis
(CME). E. canis is transmitted by nymphs and adults
of the brown dog tick, Rhipicephalus sanguineus (La-
treille)(Acari:Ixodidae).Larvaearenotimportantfor
transmission because transovarial transmission of E.
canisintickshasnotbeendemonstrated(Grovesetal.
1975, Bremer et al. 2005). The organism multiplies in
the midgut and salivary glands of R. sanguineus ticks,
whichtransmittheagentwhilefeedingondogs,Canis
familiaris L. (Groves et al. 1975). In dogs, E. canis
infects blood mononuclear cells, causing persistent
infection lasting possibly for life (Waner et al. 2001).
1Age ˆnciaPaulistadeTecnologiadoAgronego ´cio,Po ´loRegionalda
Alta Sorocabana, Presidente Prudente SP, Brazil.
2Departamento de Medicina Veterina ´ria Preventiva e Sau ´de An-
imal, Faculdade de Medicina Veterina ´ria e Zootecnia, Universidade
de Sa ˜o Paulo, Sa ˜o Paulo, SP, Brazil.
3Superintende ˆnciadeControledeEndemias(SUCEN),Sa ˜oPaulo,
SP, Brazil.
4Departamento de Parasitologia, Instituto de Cie ˆncias Biome ´dicas
V, Universidade de Sa ˜o Paulo, Monte Negro, RO, Brazil.
5Faculdade Sa ˜o Lucas, Porto Velho, RO, Brazil.
6Corresponding author, e-mail: labruna@usp.br.
0022-2585/07/0126Ð0132$04.00/0 ? 2007 Entomological Society of America

Page 2

Three clinical stages have been proposed for CME:
acute, subclinical, and chronic (Harrus et al. 1997).
DogsalsoareconsideredthemainreservoirofE.canis
in nature, because the agent does not persist in ticks
for more than one generation (Groves et al. 1975,
Smith et al. 1976).
The tick R. sanguineus is native from the Old World
and was introduced in the New World, possibly with
domestic dogs among human migrations before or
after the Columbian period. As a typical nidicolous
tick, R. sanguineus has colonized households in urban
areas of the entire Brazilian land, where it is active
throughout the year (Labruna and Pereira 2001, La-
bruna 2004). Although not very common, some R.
sanguineus populations are also established in rural
areas, especially when dogs are reared conÞned to
smallareas,offeringsuitableconditionsforthetickÕs
nidicolous behavior (Labruna and Pereira 2001, La-
bruna et al. 2001). Although indigenous tick species
(Amblyomma spp) in rural areas usually quest on
vegetation inside forest or bush areas, R. sanguineus
in not Þnding questing on vegetation; rather, they
quest inside homes or dog shelters, where they col-
onize cracks and crevices surrounding dog resting
places.
Becauseofthewidespreaddistributionofitsvector
(R. sanguineus), CME is one of the most important
infectiousdiseasesofdogsinBrazil.Localdatashowed
that 20Ð30% of dogs admitted to veterinary hospitals
presented antibodies reacting with E. canis antigens
and/or were polymerase chain reaction (PCR) posi-
tive for the E. canis 16S rRNA gene (Trapp et al. 2002,
Dagnone et al. 2003, Labarthe et al. 2003, Bulla et al.
2004). In addition, infestation by R. sanguineus has
beendemonstratedtobeasigniÞcantriskfactorforE.
canis seropositivity in Brazil (Trapp et al. 2002). Al-
though the role of R. sanguineus as vector for E. canis
hasbeenwellestablished,theprevalenceofE.canisin
R.sanguineuspopulationshasneverbeeninvestigated.
In addition, previous reports of prevalence of E. canis
among dog populations in Brazil have been biased to
dogs attended to veterinary hospitals.
Monte Negro County is located in the state of Ron-
do ˆnia, in the western Amazonian region of Brazil. A
previous study (Labruna et al. 2005) demonstrated
that most of the domestic dogs living in the rural area
of Monte Negro were infested by the indigenous tick
species, Amblyomma ovale Koch and Amblyomma ob-
longoguttatum Koch, whereas dogs living in the urban
area of Monte Negro were infested mostly by R. san-
guineus ticks. In addition, Labruna et al. (2007) re-
ported E. canis infection in four dogs from the urban
area of Monte Negro.
The current study evaluated the prevalence of E.
canis-reactive antibodies in dogs living in the rural or
urban areas of Monte Negro. In addition, ehrlichial
infection rate was determined by PCR in four R. san-
guineuspopulations(onefromMonteNegroandthree
from the state of Sa ˜o Paulo).
Materials and Methods
Blood Samples. Dog serum samples were obtained
in Monte Negro County, State of Rondo ˆnia, in the
western Amazonian region of Brazil (10? 18? S; 63? 14?
W).Ithasapopulationof?13,000inhabitants,mostof
them rural dwellers in small, family-employed farms.
The region has a high rainfall averaging 2,000 mm
annually; there is a moderate drought period during
AprilÐOctober. Temperature ranges from 25 to 29?C
and the relative humidity is 70Ð80% throughout the
year (Camargo et al. 2002).
Blood samples were collected from a jugular or
brachial vein of dogs living in the urban (during July
2001) and rural (from May to October 2002) areas. In
thelaboratory,serawereseparatelybycentrifugation,
and stored at ?20?C until analyses. At the time of the
blood collection, the urban area of Monte Negro had
671dogsdistributedin85ofthe94blocks.Atleastone
dog from each block was sampled (range, 1Ð4 dogs),
totaling153dogsfromtheurbanarea.Wefollowedthe
protocol that agrees with Ethical Principles in Animal
Research adopted by the Brazilian College of Animal
Experimentation and that was approved by the
FaculdadedeMedicinaVeterina ´riaeZootecnia/USP-
Ethical Committee for Animal Research. Permits and
approvals are on Þle in the ofÞce of D.M.A.
TheruralareaofMonteNegrowascomposedof722
cattle farms, with the majority having at least one dog
(Aguiaretal.2006).Thenumberofsampledfarmswas
determined with an estimated prevalence of 50%, ab-
solute precision desired of 10%, and conÞdence inter-
val of 95%. As a result, 86 farms were randomly se-
lected and all dogs that were present in these farms
were sampled, totaling 161 dogs (range of one to six
dogs per farm). Sampling analyses were performed
using the program EpiInfo 6.04 (Center for Disease
Control and Prevention, Atlanta, GA).
During blood collection in both urban and rural
areas, a questionnaire was given to each dog owner to
obtain information related to dog health status, age,
sex, breed, conÞnement (domiciled or free-roaming),
and hunting practices.
Serological Testing. Individual canine serum sam-
ples were tested by the indirect immunoßuorescence
assay (IFA) by using E. canis-infected DH82 cells as
antigen, performed with the Jaboticabal strain of E.
canis from Brazil, isolated from infected dog blood
(Machado 2004, Aguiar et al. 2007). Reactions were
performed with ßuorescein-conjugated anti-dog IgG
(Sigma-Aldrich, St. Louis, MO) previous titrated to
the best working dilution (1:1600), as described pre-
viously (Ristic et al. 1972, Aguiar et al. 2007). Serum
was considered to contain antibodies reactive to E.
canis if it displayed a reaction at the 1:40 dilution
(McBrideetal.2001).Ineachslide,aserumpreviously
shown to be nonreactive (negative control) and a
knownreactiveserum(positivecontrol;endpointtiter
of1,280)weretestedatthe1:40dilution.Samplesthat
reacted at the screening dilution (1:40) were then
titrated using serial two-fold dilutions to determine
endpoint titers.
January 2007AGUIAR ET AL.: Ehrlichia canis IN DOGS AND R. sanguineus TICKS
127

Page 3

Tick Populations. Because R. sanguineus is a typical
nidicolous tick species, we considered a tick popula-
tion to be composed by ticks inhabiting the backyard
and indoor areas of a single household, where dogs
were permanently present. Thus, R. sanguineus ticks
were collected in four different populations (four
households)fromdifferentareas:population1,Monte
NegroCounty(10?18?S;63?14?W),stateofRondo ˆnia,
collected during October 2004; population 2, Jundiaõ ´
County (23? 11? S; 46? 52? W), state of Sa ˜o Paulo,
collected during December 2004; population 3, Sa ˜o
PauloCounty(23?32?S;46?38?W),stateofSa ˜oPaulo,
collected during May 2005; population 4, Sa ˜o Paulo
County, collected during January 2005 (this popula-
tion was geographically separated from population 3
by ?5 km). Populations 1Ð3 were selected for the
study because they were sustained by at least one dog
showing clinical signs compatible with CME (e.g.,
apathy, pale mucous membrane), thus enhancing our
chances to detect E. canis-infected ticks. However,
population 4 was selected because there was no his-
tory of CME; thus, this population was selected to be
a “negative control” of the tick study. Only adult ticks
were sampled, because very few nymphs were ob-
served during Þeldwork.
Ticks from population 1 were collected while at-
tached on dogs; most of them were partially engorged
females. Ticks from populations two included both
unfed adult ticks and full engorged females (recently
detached from dogs), which were all collected on the
walls of dog resting places. All ticks from populations
3 and 4 were unfed adult ticks collected on the walls
of dog resting places. Ticks from the environment
wererandomlycollectedatanyareaofthehousehold,
and they were detected climbing the walls or hidden
in cracks and crevices. No tick was collected outside
thehouseholdforeachpopulation.Theinconsistency
of tick sampling (unfed, feeding, or engorged ticks)
wasbecauseoftheavailabilityofticksinthecollection
dates.
For each tick population, the sample size was de-
termined considering a 10% prevalence of infected
ticks in an inÞnite population [based on maximum
prevalence values reported for Ehrlichia chaffeensis
Anderson, Dawson & Wilson in Amblyomma ameri-
canum (L.) ticks; Burket et al. 1998, Whitlock et al.,
2000], because data for E. canis in R. sanguineus are
lacking],absoluteprecisionof5%,and95%conÞdence
interval. Thus, the minimal number of ticks to be
tested in each population was determined to be 138.
Collected ticks were preserved in absolute isopropyl
alcohol until DNA extraction described below.
Tickpopulations1,2,3,and4weresustainedbyÞve,
three, one, and one dog, respectively. Blood samples
werecollectedinEDTAfromallthesedogs,exceptfor
the Þve dogs in population 1, which had previously
beensampledinanotherstudythatfoundfourofthem
tobeinfectedbyE.canisbypolymerasechainreaction
(PCR) and DNA sequencing (Labruna et al. 2007).
Bloodsampleswerekeptrefrigeratedandtakentothe
laboratory to be subjected to DNA extraction.
DNAExtractionfromTicksandCanineBloodSam-
ples. Tick DNA was extracted in pools of every three
individual ticks. For this purpose, each tick pool was
triturated in a sterile microtube by using a sterile
disposable micropestle. For engorged females from
population 2, ticks were previously dissected for sep-
arating their salivary glands, which were used to form
pools of salivary glands from every three ticks. There-
after, DNA was extracted using the DNeasy tissue kit
(QIAGEN, Valencia, CA) following the manufactur-
erÕs protocol for isolation of DNA from insects. Five
microliters of template containing 100Ð400 ng DNA
from each tick or salivary gland pool was used in the
PCR protocol described below. DNA extraction from
canine blood samples were individually performed
using the DNeasy tissue kit following the manufac-
turerÕs protocol for isolation of DNA from blood.
PCR Amplification. Tick pool samples were pro-
cessed by a PCR assay with primers dsb-330 (5?-GAT-
GATGTCTGAAGATATGAAACAAAT-3?)anddsb-728
(5?-CTGCTCGTCTATTTTACTTCTTAAAGT-3?) de-
signed to amplify a 409-bp fragment of the dsb gene of
Ehrlichia spp. (Doyle et al. 2005, Labruna et al. 2007).
EachampliÞcationreactionwascarriedoutwith0.4?M
each primer, 1.25 U of Taq Polymerase (Platinum
TaqDNA polymerase, Invitrogen, Carlsbad, CA), PCR
buffer(50mMKCland20mMTris-HCl),3mMMgCl2,
dNTPmixture(0.2mMeach),5?lofDNAtemplate,and
watertoaÞnalvolumeof50?l.PCRcyclingconditions
were one cycle at 95?C for 2 min followed by 50 cycles
of15sat95?C,30sat58?C,and30sat72?C,followedby
72?C for 5 min. PCR products were analyzed by 1.5%
agarose gel electrophoresis stained with ethidium bro-
mide and examined by UV transillumination. To mini-
mize the potential risks for contaminations, DNA ex-
traction, PCR, and agarose gel electrophoresis were
performed in physically separate rooms. Positive
(DNA extracted from tissue cultured E. chaffeensis)
and negative controls (water) were included in all
PCR assays.
Sequence Analysis. PCR products ampliÞed from
each dog blood and from at least one tick pool from
each tick population were puriÞed using ExoSAP-IT
(USB, Cleveland, OH), and the nucleotide sequence
was determined by direct sequencing of the amplicon
with both forward and reverse primers (dsb-330 and
dsb-728) in an automated ABI Prism 310 genetic an-
alyzer (Applied Biosystems, Foster City, CA). The
BLAST program (National Center for Biotechnology
Information, Bethesda, MD) was used to compare
appropriate similarities of the dsb partial sequences
generated in the current study. Because we used E.
chaffeensis DNA as positive control in the PCR assays,
all resulting DNA sequence that matched with the E.
canisdsbpartialsequenceinGenBankwasconsidered
valid,withminimalpossibilityoflaboratoryDNAcon-
tamination.
Statistical Analyses. Prevalence values of seroposi-
tivity between urban and rural dogs were compared
by the chi-square test as well as the prevalence values
with canine sex, age, and rearing methods (e.g., free
roaming or domiciled) and hunting practices. Preva-
128JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 44, no. 1

Page 4

lence of infected ticks was calculated as the minimum
infectionrate(MIR).Forthispurpose,thenumberof
tick pools yielding expected PCR products was di-
videdbythetotalnumberofexaminedticksfromeach
population (Burket et al. 1998). Prevalence values of
infected ticks were compared within (salivary glands
versus whole tick body in case of population 2) and
between tick populations by chi-square or Fisher ex-
act tests. Analyses were performed using the program
Epiinfo 6.04. Variables were considered signiÞcantly
different if P ? 0.05.
Results
Dogs from Monte Negro County. In total, 314 dogs
(153fromtheurbanarea,and161fromtheruralarea)
wereevaluatedinthisstudy.Theywereallapparently
health, ranging from 3 months to 13 yr old, and en-
compassing different breeds (although mostly they
werecrossbreeds);127(40.5%)dogswerefemalesand
187(59.5%)weremales.Intheurbanarea,105(68.6%)
dogs were free roaming and 48 (31.4%) were com-
pletely domiciled. All rural dogs were free roaming,
from which 93 (57.7%) used to go hunting in the
jungle.
The IFA test detected antibodies reactive with E.
canis (titer ?40) in 58 (37.9%) and 40 (24.8%) dogs
fromtheurbanandruralareas,respectively(Table1).
Endpoint titers ranged from 40 to 40,960, with a me-
dian value of 160 for either urban or rural dogs. The
overall prevalence of dogs positive for E. canis-reac-
tive antibodies was signiÞcantly higher (P ? 0.05) for
urban than for rural dogs. The frequency of E. canis-
reactivedogswascomparedbetweenthreeagegroups
(0Ð12, 12Ð24, and ?24 mo old) with no statistical
difference(P?0.05)withineitherurbanorruraldogs
(Table1).Overall,therewasnosigniÞcantdifference
(P ? 0.05) for the E. canis seroprevalence between
male (63/187; 33.7%) and female (35/127; 27.6%)
dogs. In the urban area, the seroprevalence was sta-
tistically similar (P ? 0.05) for free-roaming (41/105;
39.0%) and domiciled (17/48; 35.4%) dogs (P ? 0.05).
In the rural area, the seroprevalence in hunting dogs
(25/93; 26.9%) was similar (P ? 0.05) to dogs that did
not hunt (15/68; 22.1%).
Tick Populations. Four R. sanguineus populations
were examined for ehrlichial infection by PCR tar-
geting an ehrlichial dsb gene fragment. The preva-
lenceofinfectedticks(givenasMIR)was2.3,6.2,and
3.7% for populations 1 (Monte Negro), 2 (Jundiaõ ´),
and 3 (Sa ˜o Paulo I), respectively, which were statis-
tically similar (P ? 0.05). In contrast, the prevalence
in population 4 (Sa ˜o Paulo II) was signiÞcantly dif-
ferent (P ? 0.05), with no infected tick detected
(Table 2).
In population 2, PCR was performed with DNA
extracted from 24 pools composed by salivary glands
dissected from engorged females, and 30 pools com-
posed by whole tick bodies of unfed females. These
two different DNA sources resulted in statistically
similar (P ? 0.05) MIR values, being 6.9% (5/72) for
salivaryglands,and5.5%(5/90)forwholetickbodies.
Then, these data were pooled to represent a single
MIR value (6.2%) for tick population 2 (Table 2).
Table 1.
County, state of Rondônia, Brazilian Western Amazon
Prevalence of dogs positive for Ehrlichia canis-reactive antibodies by age group in urban and rural areas of Monte Negro
Dog age (mo)
Urban areaRural area
No. of dogs
tested
No. of positive
dogs
% prevalencea
No. of dogs
tested
No. of positive
dogs
% prevalencea
0Ð12
?12Ð24
?24
No determined
Totalb
19
19
112
4
8
21.0A
42.1A
40.1A
33.3
37.9a
31
37
80
13
161
5
7
16.1A
19.0A
32.5A
15.4
24.9b
45
1
58
26
2
40
3
153
aSame capital letters on the same column indicates values signiÞcantly similar (P ? 0.05).
bDifferent lowercase letters on this line indicates values signiÞcantly different (P ? 0.05).
Table 2.
from four populations from Brazil, and in DNA extracted from dogs sustaining the tick populations
Summary of PCR amplifications for the dsb gene fragment of Ehrlichia in DNA extracted from adult R. sanguineus ticks
Tick population
(locality)a
TicksDogs
No. of ticks
tested
Tick developmental stage
No.
pools
No. pools
positive
MIRb
95% CI
No. of dogs sustaining
tick pop/no. of
E. canis-infected dogs
1 (Monte Negro)
2 (Jundiaõ ´)
165
162
Feeding adults
Full engorged females and
unfed adults
Unfed adults
Unfed adults
55
54
42.4a
6.2a
0.7Ð5.7
3.0Ð10.7
5/4c
3/310
3 (Sa ˜o Paulo I)
4 (Sa ˜o Paulo II)
162
162
54
54
6
0
3.7a
0b
1.5Ð7.5 1/1
1/0
MIR, minimum infection rate, assuming that each PCR-positive pool contained only one infected tick.
aEach tick population was composed by ticks inhabiting a single household, where dogs were permanently present.
bDifferent letters on the same column indicates values signiÞcantly different (P ? 0.05).
cReported by Labruna et al. (2006).
January 2007AGUIAR ET AL.: Ehrlichia canis IN DOGS AND R. sanguineus TICKS
129

Page 5

PCR products were ampliÞed from the blood of
three and one dogs reared in the localities of tick
populations2and3,respectively.NoPCRproductwas
ampliÞedfromthesingledogrearedintheareaoftick
population 4 (Table 2).
DNASequencing.DNAsequencingwasperformed
on the PCR products from one tick pool from popu-
lation 1 (Monte Negro), four tick pools and three dog
blood samples from population 2 (Jundiaõ ´), and Þve
tick pools and one dog blood sample from population
3 (Sa ˜o Paulo I). All sequences (355 nucleotides with-
out the primer sequences) were identical to each
other and to the corresponding sequence of the dsb
gene fragment of a North American strain of E. canis
(AF403710) available in GenBank. In addition, DNA
sequencing previously performed on PCR products
from the four PCR-positive dogs from population 1
were also identical to each other and to the corre-
sponding sequence of E. canis (Labruna et al. 2007).
Discussion
The current study reports the presence of E. canis
in dogs and R. sanguineus ticks from two different
states of Brazil (Rondo ˆnia and Sa ˜o Paulo). Recently,
Labrunaetal.(2007)reportedthepresenceofE.canis
DNAinthebloodoffourdogslivingintheurbanarea
of Monte Negro, Rondo ˆnia. The state of Rondo ˆnia is
located within the western Amazonian region of Bra-
zil, a recently colonized area where most of the coun-
ties have been founded during the last 20 yr. A recent
study (Labruna et al. 2005) demonstrated that R.
sanguineus ticks were widespread in the urban area
of Monte Negro, where very few other tick species
(Amblyommaspp.)werefoundinfestingdogs.Incon-
trast,dogsfromtheruralareawereinfestedchießyby
the ticks A. ovale and A. oblongoguttatum, and much
less frequently by R. sanguineus. Although R. san-
guineus is the predominant vector of E. canis world-
wide(Dumleretal.2001),therehasbeennoEhrlichia
species associated with Amblyomma ticks in Brazil.
Recently, Labruna et al. (2007) tested a total of 279
Amblyomma spp. by PCR (including 121 A. ovale and
30 A. oblongoguttatum) collected in the rural areas of
Monte Negro and in a neighboring County (Gover-
nador Jorge Teixeira). No PCR products were gener-
atedfromtheseticks.Thus,itseemsthatdogslivingin
the urban area of Monte Negro are at greater risks of
acquiring erlichial infection than dogs from the rural
area. This statement is in agreement with our sero-
logical results, which demonstrated a signiÞcantly
higher prevalence of E. canis-reactive antibodies in
urban dogs than in rural dogs from Monte Negro.
Serological testing such as IFA cannot consistently
distinguish between infections with E. canis and other
Ehrlichiaspecies(Waneretal.2001).However,ourIFA
results for the urban dogs of Monte Negro (some dogs
showingextremelyhightiters,e.g.,10,240Ð40,960)asso-
ciated with epidemiological data (R. sanguineus was the
major tick infesting dogs in the area; E. canis DNA was
sequenced from ticks and dogs from the area) strongly
indicate that E. canis is the main or the only Ehrlichia
species infecting dogs in that urban area.
However, the 24.8% prevalence of E. canis-reactive
antibodies for the rural dogs of Monte Negro (Table
1)contrastswiththefewreportsofR.sanguineusticks
infesting dogs in that area (Labruna et al. 2005). The
presence of Ehrlichia species other than E. canis in-
fecting those rural dogs is not supported by previous
studies, which found no Ehrlichia species infecting
Amblyomma ticks or febrile humans in this same area
(Labrunaetal.2007).However,atleasttworuraldogs
presented high antibody titers (20,480) to E. canis
(data not shown), suggesting homologous serological
response.Itisnoteworthythattheinterchangeofdogs
from the urban to the rural area (and vice versa) of
Monte Negro was frequently observed (Aguiar et al.
2006)forseveralreasons(e.g.,manyfarmownersused
to live in the urban area or at least used to visit it
periodically for business or personal purposes). Thus,
dogmovementmighthaveaccountedsigniÞcantlyfor
the prevalence of E. canis-reactive dogs in the rural
area, because E. canis causes persistent infection ap-
parently lifelong, which means that some of the sero-
logicalresponsecouldbearesultofapastinfectionin
a much earlier moment of the dogÕs life.
In the current study, ehrlichial DNA was detected
in 2.4Ð6.2% of the R. sanguineus ticks from three pop-
ulationssustainedbyatleastoneE.canis-infecteddog.
BecausetheE.canisinfectionindogsispersistent,the
availability of E. canis-infected dog for ticks of these
three populations was constant (none of these dogs
hadhistoryofantibiotictherapy;datanotshown).For
example,thetickpopulation1wassustainedbyatotal
of Þve dogs, from which four were shown to be in-
fected by E. canis; tick population 2 was sustained by
threedogs,allinfectedbyE.canis;andtickpopulation
3wassustainedbyasingleE.canis-infecteddog(Table
2). Because the MIR values reported for these tick
populations were relatively low (?7%), it means that
not every bloodmeal that a R. sanguineus tick takes on
an E. canis-infected dog results in the infection of the
tick. It has been demonstrated that when dogs were
experimentally infected with E. canis and further in-
fested with R. sanguineus ticks during the acute phase
of the disease, the proportion of the ticks that suc-
cessfully acquired the infection varied from 50 to 70%
(Bremer et al. 2005). In addition, attempts to transmit
E.caniswithadultticksthatfedtorepletionasnymphs
on dogs during the subclinical or chronic phases
(asymptomatic) of infection were unsuccessful (Lewis
et al. 1977). Similar observations were reported when
the acquisition of E. canis by Dermacentor variabilis
Say ticks was compared between ticks fed on dogs
during the acute and subclinical phases of the infec-
tion (Johnson et al. 1998).
Because our PCR protocol was based on Ehrlichia
genus-speciÞcprimers,EhrlichiaspeciesotherthanE.
caniscouldhavebeendetected.DNAsequencingwas
performed on 18 (eight dogs, 10 tick pools) of 28
PCR-positive samples. All sequences were identiÞed
as being from E. canis. Thus, we assumed that the
remaining positive, not sequenced samples (derived
130 JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 44, no. 1

Page 6

from 10 tick pools) were infected by E. canis because
they were from the same tick populations and house-
hold were E. canis partial sequences were generated
fromsamplesofdogbloodandotherticks.Inaddition,
no other Ehrlichia species is known to infect the cos-
mopolitan tick R. sanguineus. Finally, we are aware
thatticksamplingmethodswereinconsistentbetween
the populations (e.g., unfed versus engorged ticks),
what could invalidate comparisons of prevalence val-
ues between the tick populations. Regardless of the
samplingmethods,ehrlichialinfectionwasalwayslow,
with no signiÞcant difference of the prevalence be-
tween the three tick populations infected by E. canis
(Table 2).
The results of the current study suggest that the
epidemiological scenario of CME-endemic areas is
characterized by a relatively high proportion of in-
fecteddogssustainingaR.sanguineuspopulationwith
a relatively low proportion of infected ticks. A similar
scenario has been observed with E. chaffeensis, A.
americanum, and white-tailed deer [Odocoileus vir-
ginianus (Zimmermann)] in the United States (i.e.,
whereastheprevalenceofE.chaffeensisindeerishigh
[45Ð90%], its prevalence in the tick vector (A. ameri-
canum)isgenerallylowerthan10%;Burketetal.1998,
Steiner et al. 1999, Whitlock et al., 2000, Arens et al.
2003). These unexpected low prevalence values of
Ehrlichia spp. on their vector population are possibly
linked to the fact that vertebrate hosts might be a
suitable source of infection to ticks chießy during the
acuteinfectionphaseorduringarecrudescencephase
in chronically infected hosts, when blood ehrlichial
loads tend to be higher.
The current study reinforces previous studies that
reported E. canis infecting dogs in Brazil (Dagnone et
al.2003,Bullaetal.2004,Machado2004,Labrunaetal.
2006).Inaddition,wereportfortheÞrsttimeinBrazil
the natural infection of E. canis in its vector, R. san-
guineus, which is the commonest tick infesting dog in
this country (Labruna 2004). In addition, we deter-
mined the prevalence of E. canis infection in four R.
sanguineus populations. By our knowledge, the prev-
alence of E. canis in natural populations of R. san-
guineus has never been reported. Previous reports of
E. canis in R. sanguineus dealt with mixed samples of
ticks from different populations and/or with sample
numbers not previously calculated to represent a tick
population (Murphy et al. 1998, Unver et al. 2001).
Acknowledgments
This work was supported by the Fundac ¸a ˜o de Amparo a `
Pesquisa do Estado de Sa ˜o PauloÐFAPESP (grants to D.M.A.
and M.B.L.) and Conselho Nacional de Desenvolvimento
Cientõ ´Þco e Tecnolo ´gicoÐCNPq (grants to S.M.G. and
M.B.L.).
References Cited
Aguiar, D. M., G. T. Cavalcante, A. A. R. Rodrigues, M. B.
Labruna, L. M. A. Camargo, E. P. Camargo, and S. M.
Gennari. 2006. Prevalenceofanti-Neosporacaninuman-
tibodiesincattleanddogsfromWesternAmazon,Brazil,
in association with some possible risk factors. Vet. Para-
sitol. 30: 71Ð77.
Aguiar, D. M., T. B. Saito, M. K. Hagiwara, R. Z. Machado,
and M. B. Labruna. 2007. Diagno ´stico sorolo ´gico de
erliquiose canina com antõ ´geno brasileiro de Ehrlichia
canis. Cien. Rural. (in press).
Arens, M. Q., A. M. Liddell, G. Buening, M. Gaudreault-
Keener, J. W. Sumner, J. A. Comer, R. S. Buller, and G.
Storch. 2003. Detection of Ehrlichia spp. in the blood of
wildWhite-TailedDeerinMissouribyPCRandserologic
analysis. J. Clin. Microbiol. 41: 1263Ð1265.
Bremer, W. G., J. J. Schaefer, E. R. Wagner, S. A. Ewing, Y.
Rikihisa, G. R. Needham, S. Jittapalapong, D. L. Moore,
and R. W. Stich. 2005. Transstadial and intrastadial
experimental transmission of Ehrlichia canis by male
Rhipicephalus sanguineus. Vet. Parasitol. 131: 95Ð105.
Bulla, C., R. K. Takahira, J. P. Araujo Jr., L. A. Trinca, R. S.
Lopes,andC.E.Weidmeyer. 2004. Therelationshipbe-
tween the degree of thrombocytopenia and infection
with Ehrlichia canis in an endemic area. Vet. Res. 35:
141Ð146.
Burket,C.T.,C.N.Vann,R.R.Pinger,C.L.Chatot,andF.E.
Steiner. 1998. Minimum infection rate of Amblyomma
americanum (Acari: Ixodidae) by Ehrlichia chaffeensis
(Rickettsiales: Ehrlichieae) in southern Indiana. J. Med.
Entomol. 35: 653Ð659.
Camargo,L.M.,M.M.Moura,V.Engracia,R.C.Pagotto,S.A.
Basano,L.H.daSilva,E.P.Camargo,B.Beiguelman,and
H.Krieger. 2002. AruralcommunityinaBrazilianWest-
ern Amazonian Region: some demographic and epide-
miological patterns. Mem. Inst. Oswaldo Cruz 97: 193Ð
195.
Dagnone, A. S., H. S. M. Autran, M. C. Vidotto, F. S. Jojima,
and O. Vidotto. 2003. Ehrlichiosis in anemic, thrombo-
cytopenic, or tick-infested dogs from a hospital popula-
tion in south Brazil. Vet. Parasitol. 117: 285Ð290.
Doyle, C. K., M. B. Labruna, E. B. Breitschwerdt, Y. Tang,
R. E. Corstvet, B. C. Hegarty, K. C. Bloch, P. Li, D. H.
Walker, and J. W. Mcbride. 2005. Detection of medi-
cally important Ehrlichia by quantitative multicolor Taq-
Man Real-time PCR of the dsb gene. J. Mol. Diagn. 7:
504Ð510.
Dumler, J. S., A. F. Barbet, C. P. Bekker, G. A. Dasch, G. H.
Palmer, S. C. Ray, Y. Rikihisa, and F. R. Rurangirwa.
2001. Reorganization of genera in the families Rickett-
siaceae and Anaplasmataceae in the order Rickettsiales:
uniÞcation of some species of Ehrlichia with Anaplasma,
Cowdria with Ehrlichia and Ehrlichia with Neorickettsia,
descriptions of six new species combinations and desig-
nation of Ehrlichia equi and HGE agentÕ as subjective
synonyms of Ehrlichia phagocytophila. Int. J. Syst. Evol.
Microbiol. 51: 2145Ð2165.
Groves, M. G., G. L. Dennis, H. L. Amyx, and D. L. Huxsoll.
1975. Transmission of Ehrlichia canis to dogs by ticks
(Rhipicephalus sanguineus). Am. J. Vet. Res. 36: 937Ð940.
Harrus, S., T. Waner, and T. Bark. 1997. Canine monocytic
ehrlichiosisÐan update. Comp. Contrib. Educ. Pract. Vet.
19: 431Ð444.
Johnson, E. M., S. A. Ewing, R. W. Barker, J. C. Fox, D. W.
Crow, and K. M. Kocan. 1998. Experimental transmis-
sionofEhrlichiacanis(Rickettsiales:Ehrlichieae)byDer-
macentor variabilis (Acari: Ixodidae). Vet. Parasitol. 74:
277Ð288.
Labarthe, N., M. Campos Pereira, O. Balbarini, W. Mckee,
C. A. Coimbra, and J. Hoskins. 2003. Serologic preva-
lence of Dirofilaria immintis, Ehrlichia canis and Borrelia
burgdorferi infection in Brazil. Vet. Ther. 4: 67Ð75.
January 2007AGUIAR ET AL.: Ehrlichia canis IN DOGS AND R. sanguineus TICKS
131

Page 7

Labruna, M. B. 2004. Biologica-ecologia de Rhipicephalus
sanguineus (Acari: Ixodidae). Rev. Bras. Parasitol. Vet. 13
(Suppl): 123Ð124.
Labruna,M.B.,andM.C.Pereira. 2001. Carrapatosemca ˜es
no Brasil. Clin. Vet. 30: 24Ð32.
Labruna, M. B., S. L. P. Souza, J. S. Guimara ˜es, Jr., R. C.
Pacheco,A.Pinter,andS.M.Gennari. 2001. Prevale ˆncia
de carrapatos em ca ˜es de a ´reas rurais da regia ˜o norte do
Estado do Parana ´. Arq. Bras. Med. Vet. Zootec. 53: 553Ð
556.
Labruna, M. B., L. M. A. Camargo, F. A. Terrassini, F. Fer-
reira, T. T. Schumaker, and E. P. Camargo. 2005. Ticks
(Acari: Ixodidae) from the state of Rondo ˆnia, western
Amazon, Brazil. Syst. Appl. Acarol. 10: 17Ð32.
Labruna, M. B., J. W. McBride, L. M. A. Camargo, D. M.
Aguiar, M. J. Yabsley, W. R. Davidson, E. Y. Stromdahl,
P. C. Williamson, R. W. Stich, S. W. Long, et al. 2007.
Investigation of Ehrlichia species in ticks, humans, dogs,
and capybaras from Brazil. Vet. Parasitol. (in press).
Lewis, G. E., M. Ristic, R. D. Smith, T. Lincoln, and E. H.
Stephenson. 1977. The brown dog tick Rhipicephalus
sanguineusandthedogasexperimentalhostsofEhrlichia
canis. Am. J. Vet. Res. 38: 1953Ð1955.
Machado,R.Z. 2004. Ehrlichiosecanina.Rev.Bras.Parasitol.
Vet. 13 (Suppl.): 53Ð57.
McBride,J.W.,R.E.Corstvet,E.B.Breitschwerdt,andD.H.
Walker. 2001. ImmunodiagnosisofEhrlichiacanisinfec-
tion with recombinant proteins. J. Clin. Microbiol. 39:
315Ð322.
Murphy, G. L., S. A. Ewing, L. C. Whitworth, J. C. Fox, and
A. A. Kocan. 1998. A molecular and serologic survey of
Ehrlichia canis, E. chaffeensis, and E. ewingii in dogs and
ticks from Oklahoma. Vet. Parasitol. 79: 325Ð339.
Ristic, M., D. L. Huxsoll, R. M. Weisiger, P. K. Hildebrandt,
and B. A. Nyindo. 1972. Serological diagnosis of tropical
canine pancytopenia by indirect immunoßuorescence.
Infect. Immunol. 6: 226Ð231.
Smith, R. D., D. M. Sells, E. H. Stephenson, M. Ristic, and
D. L. Huxsoll. 1976. Development of Ehrlichia canis,
causative agent of canine ehrlichiosis, in the tick Rhipi-
cephalus sanguineus and its differentiation from a sym-
biotic Rickettsia. Am. J. Vet. Res. 37: 119Ð126.
Steiner, F. E., R. R. Pinger, and C. N. Vann. 1999. Infection
rates of Amblyomma americanum (Acari: Ixodidae) by
Ehrlichiachaffeensis(Rickettsiales:Ehrlichiae)andprev-
alenceofE.chaffeensis-reactiveantibodiesinwhite-tailed
deer in southern Indiana, 1997. J. Med. Entomol. 36:
715Ð719.
Trapp, S. M., A. S. Dagnone, O. Vidotto, R. L. Freire, and
H. S. A.Morais. 2002. Seroepidemiologyofcaninebabe-
siosis and ehrlichiosis in a hospital population in south
Brazil. J. Vet. Int. Med. 16: 365.
Unver, A., M. Perez, N. Orellana, H. Huang, and Y. Rikihisa.
2001. Molecular and antigenic comparison of Ehrlichia
canisisolatesfromdogs,ticks,andahumaninVenezuela.
J. Clin. Microbiol. 39: 2788Ð2793.
Waner, T., S. Harrus, F. Jogenjan, H. Bark, A. Keysary, and
A.W.Cornelissen. 2001. SigniÞcanceofserologicaltest-
ingforehrlichialdiseasesindogswithspecialemphasison
the diagnosis of canine monocytic ehrlichiosis caused by
Ehrlichia canis. Vet. Parasitol. 95: 1Ð15.
Whitlock, J. E., Q. Q. Fang, L. A. Durden, J. H. Oliver, Jr.
2000. Prevalence of Ehrlichia chaffeensis (Rickettsiales:
Rickettsiaceae) in Amblyomma americanum (Acari: Ixo-
didae)fromtheGeorgiaCoastandBarrierIslands.J.Med.
Entomol. 37: 276Ð280.
Received 29 May 2006; accepted 23 September 2006.
132JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 44, no. 1