Vet. Res. 38 (2007) 697–710
c ? INRA, EDP Sciences, 2007
Available online at:
Canine bartonellosis: serological and molecular
prevalence in Brazil and evidence of co-infection
with Bartonella henselae and Bartonella vinsonii
Pedro Paulo Vissotto De Paiva Da, Ricardo Guillermo Ma,
Denise Saretta Sb, María Belén Ca, Julie Meredith
Ba, Barbara Ha, Edward Bealmear Ba*
aIntracellular Pathogens Research Laboratory, Center for Comparative Medicine and Translational
Research, Department of Clinical Sciences, College of Veterinary Medicine,
North Carolina State University, 4700 Hillsborough Street, Raleigh, NC, 27606, USA
bSchool of Veterinary Medicine and Animal Science, São Paulo State University (FMVZ – UNESP),
Botucatu, São Paulo, Brazil
(Received 17 November 2006; accepted 4 April 2007)
Abstract – The purpose of this study was to determine the serological and molecular prevalence of
Bartonella spp. infection in a sick dog population from Brazil. At the São Paulo State University
Veterinary Teaching Hospital in Botucatu, 198 consecutive dogs with clinicopathological abnor-
malities consistent with tick-borne infections were sampled. Antibodies to Bartonella henselae and
Bartonella vinsonii subsp. berkhoffii were detected in 2.0% (4/197) and 1.5% (3/197) of the dogs,
respectively. Using 16S-23S rRNA intergenic transcribed spacer (ITS) primers, Bartonella DNA
was amplified from only 1/198 blood samples. Bartonella seroreactive and/or PCR positive blood
samples (n = 8) were inoculated into a liquid pre-enrichment growth medium (BAPGM) and sub-
sequently sub-inoculated onto BAPGM/blood-agar plates. PCR targeting the ITS region, pap31 and
rpoB genes amplified B. henselae from the blood and/or isolates of the PCR positive dog (ITS:
DQ346666; pap31 gene: DQ351240; rpoB: EF196806). B. henselae and B. vinsonii subsp. berkhof-
fii (pap31: DQ906160; rpoB: EF196805) co-infection was found in one of the B. vinsonii subsp.
berkhoffii seroreactive dogs. We conclude that dogs in this study population were infrequently ex-
posed to or infected with a Bartonella species. The B. henselae and B. vinsonii subsp. berkhoffii
strains identified in this study are genetically similar to strains isolated from septicemic cats, dogs,
coyotes and human beings from other parts of the world. To our knowledge, these isolates provide
the firstBrazilianDNA sequences from theseBartonella species and thefirst evidence of Bartonella
co-infection in dogs.
dogs / Bartonella infections / heart disease / culture / Brazil
Bartonella are fastidious hemotropic
bacteria, which cause long-lasting bac-
teremia in mammals and are transmit-
ted by animal bites and scratches or by
many different vectors including sand-
flies, body lice, fleas and potentially
ticks. The genus Bartonella contains nu-
merous recently described species, many
Article available at http://www.vetres.org or http://dx.doi.org/10.1051/vetres:2007023 Article available at http://www.vetres.org or http://dx.doi.org/10.1051/vetres:2007023
698P.P.V.P. Diniz et al.
of which are new and emerging human
In dogs, infection with five Bartonella
species has been described:B. clarridgeiae
and B. vinsonii subsp. berkhoffii (here-
after B. v. berkhoffii) DNA was detected
in aortic valves of six dogs with vegeta-
tive endocarditis, based upon independent
studies [9,14,32];B. elizabethae DNA was
detected in a dog with weight loss and
sudden death ; B. henselae DNA was
sequencedfrom a dog with peliosis hepatis
and from a dog with granulomatous hep-
atitis ; and B. washoensis was isolated
from a dog with mitral valve endocardi-
tis . A potentially new species closely
related to B. clarridgeiae (proposed name
B. rochalimaea) has been isolated from
three dogs and from 22 gray foxes in
To date, no studies have reported sero-
logical or molecular evidence of Bar-
tonella infection in dogs from Brazil or
other South American countries. With the
notable exception of B. bacilliformis, there
are only a limited number of reports
describing human infection with other
Bartonella species in South America. In
Brazil, B. quintana was detected by PCR
in an infant  and two fatal cases of
endocarditis were associated with high B.
henselae antibody titers . In contrast to
humans, B. henselae seroreactivity was re-
ported in 46% of 102 cats2suggesting that
exposure to this organism in Brazil may be
1Henn J.B., Koehler J.E., Gabriel M., Kasten
R.W., Brown R.N., Papageorgiou S., Chomel
B.B., Zoonotic Bartonella spp. in domestic
dogs and gray foxes from California, in: Proc.
20th Meeting of the American Society for Rick-
ettsiology and 5th International Conference
on Bartonella as Emerging Pathogens, Pacific
Grove, 2006, pp. 70.
D’Auria S.R.N., Soroprevalência de Bartonella
henselae em gatos do municipio de São Paulo,
Rev. Soc. Bras. Med. Trop. (1996) 29:104
more prevalent than was previously recog-
The objective of this study was to deter-
mine the serological and molecular preva-
lence of Bartonella infection in a subset of
the sick dog population examined at a Vet-
erinary Teaching Hospital in southeastern
Brazil. Successful culture and molecular
characterization of B. henselae and B. v.
berkhoffii from two clinically ill dogs is
also reported, including the first evidence
of simultaneous infection with these two
organisms in the same host.
2. MATERIALS AND METHODS
2.1. Study population
One hundred and ninety-eight EDTA-
anticoagulated blood and serum samples
from dogs with historical, clinical and/or
laboratory findings consistent with tick-
borne diseases were prospectively se-
lected at a Veterinary Teaching Hospi-
tal at the College of Veterinary Medicine
and Animal Husbandry (FMVZ – Un-
esp) in Botucatu, southeast Brazil (latitude
22◦53?09??S, longitude 48◦26?42??W)
between October 2002 and November
2003. Dogs included in this study had
at least three of the following clini-
cal or laboratory criteria: presence of
tick infestation at the time of examina-
tion, bleeding disorders (epistaxis, me-
lena, ecchymosis, bleeding gums), neu-
rological signs, inflammatory ocular dis-
ease (uveitis, chorioretinitis, hyphema),
fever (rectal temperature > 39.4◦C), ane-
mia (PCV < 35%), leukopenia (WBC
< 6 000 cells/µL), thrombocytopenia
(platelets < 150000 cells/µL) and hyper-
proteinemia (total protein > 7.8 g/dL).
Dogs treated with tetracycline or imi-
docarb dipropionate during the previous
30 days were excludedfrom the study pop-
ulation. This study was approved by the
Medical Ethical and Animal Care Com-
mittee of São Paulo State University. In
Bartonella species in Brazilian dogs699
addition, each owner authorized blood col-
lection and enrollment of the dog into the
From the selected dogs, two had
Bartonella species detected by culture
(techniques described below). Dog #1
was a 6.3 kg 2-year-old male Poodle
that was examined because of anorexia,
bleeding gums, cutaneous petechiae, and
ecchymoses.The dogwas febrile(39.7◦C)
and infested with ticks. Hematological
(PCV = 14%; normal ? 39%), leukope-
nia (2 692/µL; normal ? 6 000/µL),
? 190 000/µL), and hypoproteinemia
(4.6 g/dL; normal ? 6.1 g/dL). Cardiovas-
cular abnormalities included hypotension
(systolic blood pressure = 75 mmHg;
normal ? 90 mmHg), myocardial dysfunc-
tion (fractional shortening = 24%; normal
? 30%) and increased serum cTnI concen-
Electrocardiographic abnormalities were
not detected. B. henselae was isolated
from the EDTA-anticoagulatedblood sam-
ple. By PCR testing, this dog was found
to be co-infected with Ehrlichia canis,
but was PCR negative for Anaplasma
Ehrlichia chaffeensis, Ehrlichia ewingii
and Rickettsia spp. DNA. Antibodies to
B. burgdorferi, L. chagasi, L. interrogans,
R. rickettsii and D. immitis antigens were
Dog #2, a 34 kg 6-year-old male
German Shepherd, was determined to
be co-infected with B. henselae, B. v.
berkhoffii and E. canis. This dog was ex-
amined due to bilateral epistaxis and had
history of tick attachment. The dog was
febrile (42◦C) at the time of examination.
mild anemia (PCV = 34%), monocytosis
(1936/µL, normal ? 1100/µL), thrombocy-
topenia (50500/µL) and hyperproteinemia
(11.6 g/dL, normal ? 7.8 g/dL). ECG
blood pressure was normal; however,
echocardiography identified an increased
left ventricle wall (14.5 mm) thickness
resulting in reduction of left ventricular
chamber size (33 mm) during diastole. All
other infectious disease diagnostic tests
described above for Dog #1 were also
negative for Dog #2.
2.2. Serology-based assays
(IFA) for the detection of anti-B. v.
#51672) and anti-B. henselae (strain
Houston-1, ATCC #49882) antibodies
in canine sera was performed as de-
scribed . The starting dilution was
at 1:16 and the cut-off for seroreactivity
was defined at 1:64. Endpoint titers were
determined as the last dilution at which
brightly stained bacteria could be detected
on a fluorescence microscope. Exposure
to Leishmania chagasi and Rickettsia
rickettsii was evaluated by IFA as de-
scribed [42, 45]. The cut-off dilution for
seroreactivity was defined at 1:64 for R.
rickettsii, and 1:80 for L. chagasi. An
ELISA-based test kit (SNAP 4Dx, IDEXX
Laboratories, Inc., Westbrook, ME, USA)
was used according to the manufacturer’s
instructions to detect Borrelia burgdor-
feri antibodies, as well as antigens of
Dirofilaria immitis from dog serum. A
microagglutination test for Leptospira
interrogans with a cut-off dilution at 1:100
was performed as described . One
dog was not tested because of insufficient
2.3. DNA extraction
DNA was extracted from 300 µL of
each dog’s frozen EDTA-blood pellet us-
ing a commercially available GFX Ge-
nomic Blood DNA Purification Kit (Amer-
sham Biosciences, Piscataway, NJ, USA).
700P.P.V.P. Diniz et al.
The final eluted volume was 200 µL per
sample. The absence of PCR inhibitors
was demonstrated by the amplification of
a fragment of the constitutive gene for
the glyceraldehyde-3-phosphate dehydro-
genase (GAPDH) protein, an enzyme re-
lated to the glycolysis pathway that is ex-
pressed in all mammal cells [5,7].
2.4. DNA amplification-based assays
2.4.1. Bartonella 16S-23S ribosomal
RNA intergenic spacer (ITS) PCR
Bartonella genus screening was per-
formed as described previously . Am-
plification was performed in a 25 µL
final volume reaction containing 2mM
MgCl2, 0.625 units of Taq DNA poly-
merase (Takara Ex Taq, Takara Bio Inc,
Shiga, Japan), 12.5 pmol of each primer
(325s 5’-CTT CAG ATG ATG ATC CCA
AGC CTT YTG GCG -3?and 1100as 5?-
GAA CCG ACG ACC CCC TGC TTG
CAA AGC A-3?), 200 µM (each) dATP,
dTTP, dCTP, and dGTP (Fisher Scientific,
Pittsburgh, PA, USA) and 1 to 5 µL of
DNA template, according to DNA concen-
tration determined for each sample (5 to
200 ng/reaction). B. v. berkhoffii identity
was established using specific ITS primers
(46s: 5?- CCT CAT TCT TTA AAA AAA
GAG GGC TTT TTA AG-3?and 590as:
GAA AGC GCT AAC CCC TAA ACC
GAT T-3?) and the same PCR conditions.
PCR was performed in a Mastercycler
EP (Eppendorf, Hamburg, Germany) un-
der the following conditions: a single hot-
start cycle at 95◦C for 2 min followed
by 55 cycles of denaturing at 94◦C for
15 s, annealing at 66◦C for 15 s, and ex-
tension at 72◦C for 15 s. Amplification
was completed by an additional cycle at
72◦C for 1 min, and products were an-
alyzed by 2% agarose gel electrophoresis
under UV exposure. DNA from a healthy,
specific pathogen-free dog was used as a
PCR negative control. Plasmid clones of
partial sequences of 16-23S ITS region of
B. henselae Houston-1 (L35101) and B. v.
berkhoffii type I (AF167988) were used as
positive controls. The detection limit ob-
served in 100% of 10 replicate reactions
was 2.5 DNA copies of B. henselae and
5 copies of B. v. berkhoffii per reaction.
2.4.2. Bartonella heme-binding protein
gene (pap31) PCR amplification
ITS region PCR positive dogs were
subsequently tested for the pap31 bacte-
riophage associated gene . Oligonu-
cleotides for Bartonella genus 1s 5?-ACT
TCT GTT ATC GCT TTGATT TCR RCT-
3?and 688as 5?-CAC CAC CAG CAA
AAT AAG GCA TMA Y-3?were used
at the same concentration and amplifica-
tion conditions as described above, where
the annealing temperature was 58◦C. B.
henselae identity was established using
specific pap31 forward primer (pap31 Bh-
s: 5?-TAA GGT TGA AAT AAC TGA
TCC GAA T-3?) in conjunction with the
688as reverse primer. Plasmid clones of
partial sequences of pap31 gene of B.
henselae Houston-1 (BX897699) and B. v.
berkhoffii type I (AY663045) was used as
a positive control. The detection limit ob-
served in 100% of 10 replicate reactions
was 5 copies per reaction for both organ-
2.4.3. Bartonella RNA polymerase
beta subunit (rpoB) gene PCR
ITS regionPCR positive dogs were sub-
sequently tested for the rpoB gene. The
following oligonucleotides were manually
designed for Bartonella genus amplifica-
tion: rpoB 1615s 5?-ATY ACY CAT AAR
CGY CGT CTT TCT GCT CTT GG-
3?and rpoB 2267as 5?-GGA TCT AAA
Bartonella species in Brazilian dogs701
TCT TCY GTY GCA CGR ATA CG-
3?. The PCR conditions were similar as
described above, with annealing tempera-
ture at 62◦C. As positive controls, DNA
extracted from isolates of B. henselae
Houston-1 (BX897699) and B. v. berkhof-
fii type I (ATCC strain 93-CO-1) were
used as genomic equivalent (GE) serially
diluted in DNA obtained from a specific
pathogen-free dog. The detection limit for
both organisms was 2.5 GE per reaction.
2.4.4. Other tick-borne pathogen DNA
Because of the risk of polymicrobial in-
fection, other tick-borne pathogens were
evaluated. Anaplasma and Ehrlichia 16S
rRNA gene was targeted by conventional
PCR [10,30] with a detection limit of 10
copies per reaction in 100% of 10 repli-
cates. Rickettsia spotted-feverspecies were
screened by real-time PCR targeting the
ompA gene  with a detection limit of
5 copies per reaction in 96% of 22 repli-
2.5. Bartonella species culture
Two hundred microliters of blood
from Bartonella ITS region-PCR posi-
tive dogs and/or IFA seroreactive dogs
were inoculated into liquid Bartonella
(BAPGM)  and incubated at 35◦C in
5% CO2, water-saturated atmosphere. Af-
ter a 14-day incubation period, 1 mL from
liquid culture was sub-inoculated onto
BAPGM/blood-agar plates and incubated
for 14 days. For maximization of Bar-
tonella DNA recovery, multiple bacterial
colonies were collectively swabbed from
the surface of each BAPGM/blood-agar
plate, resuspended in sucrose-phosphate-
glutamate (SPG) buffer and stored at
−80◦C until processed for DNA extrac-
tion. DNA from 200 mL of BAPGM liquid
culture and from frozen SPG bacterial
colonies was extracted using a commer-
cially available MagAttract DNA blood kit
(BioRobot M48, Qiagen, Chatsworth, CA,
2.6. Cloning and sequencing of ITS
region and pap31 gene amplicons
PCR products obtained from EDTA-
anti-coagulated blood, BAPGM liquid cul-
ture and/or the plate colonies were cloned
into plasmid pGEM-T Easy Vector System
(Promega, Madison, WI, USA) and the re-
combinants selected based on white/blue
screening, EcoR I digestion and 2%
agarose gel electrophoresis. After plas-
mid purification using commercial kit
(QIAprep Spin Miniprep, Qiagen, Valen-
cia, CA, USA), clones were sequenced
by Davis Sequencing (Davis, CA, USA).
Chromatogram evaluation and sequence
alignment were performed using Contig-
Suite 10.1, Invitrogen Corp., Carlsbad,
CA, USA). Bacteria species and strain
were defined by comparing similarities
with other sequences deposited in the Gen-
Bank database prior to April 2007 using
the Basic Local Alignment Search Tool
(Blast version 2.0) .
2.7. Cardiac evaluation
Bartonella infected and/or seroreac-
tive dogs were evaluated for cardiac dis-
turbances. Non-invasive blood pressure
was established using a Doppler ultra-
sonographic method (DV-10, Microem,
Brazil) . An electrocardiogram (ECG)
was recorded with a 12-lead computer-
based system (ECG-PC, TEB, Brazil) .
Echocardiography was performed with a
5 MHz electronic ultrasound probe (EUB-
405, Hitachi, Norcross, GA, USA) .
702P.P.V.P. Diniz et al.
Table I. Serology and PCR results from 8 seroreactive or PCR positive sick dogs enrolled in this
study. Bartonella titers ? 64 were considered seroreactive.
IFA reciprocal titer
Bartonella genus ITS region PCR
Blood Liquid culture
B. v. berkhoffii
* By DNA sequencing, Dog #1 was infected with B. henselae and Dog #2 was co-infected with B. hense-
lae and B. vinsonii subsp. berkhoffii.
Serum cardiac troponin I (cTnI) was mea-
sured using human cTnI kits (Dimen-
sionRxL clinical chemistryanalyzer,Dade
Behring, Deerfield, IL, USA), previously
validated for testing dog serum .
3.1. Serological and molecular
B. henselae and B. v. berkhoffii anti-
bodies were detected in 2.0% (4/197) and
1.5% (3/197) of the 198 dogs in this study,
respectively (Tab. I). ITS region amplifi-
cation for Bartonella genus was not ob-
tained from the blood of any of the 7
Bartonella seroreactive dogs. However, B.
henselae DNA was detected from 1/191
non-seroreactive dog blood samples (Dog
#1, Tab. I). After blood culture (described
in detail below), one additional dog (#2)
was detected as co-infected with B. hense-
lae and B. v. berkhoffii. This dog had a
reciprocal titer of 64 for B. v. berkhoffii but
no Bartonella DNA was amplified from
the blood sample. Based upon PCR testing
of blood and blood culture of seroreactive
dogs, the B. henselae and B. v. berkhof-
fii molecular prevalence in this population
was 1% (2/198) and 0.5% (1/198), respec-
Six of seven Bartonella seroreactive
dogs detected in this study were co-
infected with E. canis. Dog #1, Bartonella
PCR positive but seronegative, also was
infected with E. canis. The only E. ca-
nis PCR negative dog (#5, Tab. I) was
co-infected with A. platys, and had a re-
ciprocal B. henselae titer of 256. DNA
of A. phagocytophilum, E. chaffeensis, E.
ewingii, Rickettsia spp. or exposure to B.
burgdorferi, L. chagasi, L. interrogans, R.
rickettsii orantigensof D. immitis werenot
detected in any Bartonella seroreactive or
PCR positive dogs.
3.2. Culture and molecular
characterization of B. henselae
and B. vinsonii subsp. berkhoffii
From the only PCR-positive blood sam-
ple (Dog #1), several colonies were ob-
tained from the BAPGM/blood-agar cul-
ture, from which DNA was collectively
extracted. The ITS region and pap31 gene
amplicons from these colonies were of
Bartonella species in Brazilian dogs703
Figure 1. PCR results of 16S-23S rRNA intergenic spacer region for Bartonella genus from Dog
#1. Lane 2, blood sample; lane 3, BAPGM liquid pre-enrichment medium; lane 4, BAPGM/blood-
agar plate colonies; lane 5, negative dog DNA; lane 6, B. henselae positive control; lanes 1 and 7,
1-Kb DNA ladder.
the expected size for B. henselae (Fig. 1).
DNA sequences from blood, BAPGM liq-
uid culture, and the plate isolates were
100% homologous. Sequences were de-
posited into the GenBank database under
accession numbers DQ346666 (ITS) and
DQ351240 (pap31 gene).
seroreactive dogs was also cultured, and
from a second patient (Dog #2) several
colonies were obtained. After extracting
DNA of these colonies conjointly, ITS
region PCR amplicons were 100% ho-
mologous (604/604 bp) with B. henselae
strain Houston-1 (BX897699). However,
cloned pap31 sequences derived from the
pooled colonies were 99.5% homologous
(545/548 bp) with B. v. berkhoffii type
I (Tab. II). This sequence was deposited
under accession number DQ906160. The
co-infection was confirmed using species-
specific primers targeting the B. henselae
pap31 gene and the B. v. berkhoffii ITS
region. Consensus of multiple sequences
of pap31 gene was 100% homologous
(398/398 bp) to B. henselae Houston-1
(Tab. III) and of consensus of ITS region
was 99.8% homologous(553/552bp) to B.
v. berkhoffii type I (AF167988). Dog #2
was B. v. berkhoffii seroreactive(reciprocal
titer 64), but was not B. henselae seroreac-
PCR for rpoB gene was positive only
from blood from Dog #1 and agar plate
colonies from Dog #2. Alignment of mul-
tiple sequences confirmedpreviousresults:
Dog #1 was infected with B. henselae
Houston-1 (655/656 bp – 99.8% homol-
ogous to BX897699) and Dog #2 was
infected with B. v. berkhoffii (656/656 bp
– 100% homologous to AF165989). B.
henselae and B. v. berkhoffii rpoB se-
quences were deposited under accession
numbers EF196806 and EF196805, re-
3.3. Cardiac evaluation
In addition to the two cases reported
in detail above, four other Bartonella spp.
seroreactive dogs had evidence of car-
diac disease. Dogs #3 and #4 (Tab. I)
were seroreactive to B. henselae. Dog #3
704P.P.V.P. Diniz et al.
Table II. Bartonella vinsonii subsp. berkhoffii heme-binding protein gene (pap31) nucleotide differ-
ences of the sequence from this study (Iso-756) compared to GenBank database sequences available
on April 2007.
Dog - Bvb Type I
Dog - Bvb Type II
Fox - Bvb Type III
Dog - Bvb Type IV A
Nucleotide difference at positiona
32 55-56 70 97 173 245 251 268 286 316 385 417 435 501 536
G CGA GAAGA
aThe number represents the nucleotide position of B. vinsonii subsp. berkhoffii Iso-756; • same base
compared with Brazilian isolate.
had diastolic left ventricular enlargement
and increased cTnI of 0.28 ng/mL (nor-
mal ? 0.11 ng/mL). Dog #4 had increased
cTnI (0.51 ng/mL). Dogs #7 and #8 were
B. v. berkhoffii seroreactive (Tab. I) and,
despite cTnI levels within normal range,
Dog #7 had evidence of myocardial dys-
function (fractional shortening = 22.3%;
normal ? 30%) and Dog #8 had left ventri-
cle enlargement. No dog had echocardio-
graphic findings consistent with infective
endocarditis or was hypertensive. In total,
6of8 IFAand/orPCR Bartonellaspp.pos-
itive dogs had cardiac abnormalities.
To our knowledge, this study provides
the first data relative to B. henselae and
B. v. berkhoffii serological and molecu-
lar prevalence in dogs from Brazil. In
addition, this is the first report of Bar-
tonella species isolation from dogs from
South America and the first documenta-
tion of co-infection with two Bartonella
species in a dog. In 2004, B. henselae
DNA was reported for the first time in
a blood culture isolate obtained from a
dog from Gabon ; however, no in-
formation about amplicon size, strain and
genetic sequence was provided. The Bar-
tonella seroprevalences detected in this
study were lower than expected for a
country with tropical and sub-tropical cli-
mates. B. henselae seroreactivity was de-
tected from 3.0% (3/100) of dogs in the
United Kingdom , 7.7% (4/52) of dogs
in Japan  and 27.2% (82/301) of sick
dogs in North Carolina . Bartonella
vinsonii subsp. berkhoffii antibodies were
detected in 10% (4/40) of dogs from Is-
rael  and in 38% (19/49) of dogs from
Thailand . Bartonellosis may predis-
pose the host to insidious nonspecific man-
ifestations; therefore, we can not exclude
a selection bias, based upon narrowly de-
fined entry criteria of this study.
Cats are the main reservoir of Bar-
tonella spp. infection for humans .
human cases of bartonellosis have also
implicated dogs as a source of B. hense-
lae infection [28, 52]. B. v. berkhoffii has
been detected by PCR in dogs, coyotes
and foxes [29,36]; however, the extent to
which dogs contribute to the transmission
of Bartonella to humans is unknown .
In Brazil, Rhipicephalus sanguineus, the
most prevalent tick that infests dogs 
also parasitizes humans [17,21]. Epidemi-
ological evidences suggest that R. san-
guineus may be involved in the transmis-
sion of Bartonella species in dogs .
Bartonella species in Brazilian dogs705
Table III. Bartonella henselae heme-binding protein gene (pap31) nucleotide differences of the strains Brazil-1 and Brazil-2 compared to GenBank
database sequences available on April 2007.
Nucleotide difference at positiona
16 81 244-245 325 329 334 337 350 353-354 359 375 386 390 394 398 407-408 412 418-419 421-422 425-426 429
San Ant 2
aThe number represents the nucleotide position of B. henselae Brazil-1; • same base as the Brazilian isolate; / sequence not available; S, degenerate oligonu-
cleotide (C or G).
* Sequence not deposited at GenBank database because it is 100% homologous to other sequences.
706P.P.V.P. Diniz et al.
Table IV. Bartonella henselae 16S-23S ribosomal RNA intergenic spacer (ITS) nucleotide differ-
ences for the strains Brazil-1 and Brazil-2 compared to GenBank database sequences available on
Nucleotide difference at positiona
San Ant 2
aThe number represents the nucleotide position of B. henselae Brazil-1; • same base as the Brazilian
isolate; –, deletion; W, degenerate oligonucleotide (A or T); insertion consisting of: ATTGCTTC-
* Sequence not deposited in GenBank because homology was 100% with B. henselae Houston-1.
Thus, this tick potentially could pose a risk
for human infection.
Dog #1 in this study, co-infected with
B. henselae and E. canis, had a clini-
cal diagnosis compatible with myocardi-
tis, based upon cardiac dysfunction and
highcTnIlevel, a cardiacbiomarkerhighly
specific and sensitive for myocardial in-
jury . Moreover, 6 of 8 (75%) Bar-
tonella seroreactive or PCR positive dogs
had evidence of cardiac disease. Myocar-
dial injury has been reported in infective
endocarditis in dogs [11, 32] and in hu-
man bartonellosis [25, 47]. Surprisingly,
there was not echocardiographic evidence
of endocarditis in any dog in this study.
Seven of eight Bartonella seropositive or
PCR positive dogs were infected with E.
canis, a tick-transmitted, obligatory intra-
cellular bacterium that is also associated
with myocardialinjury3. Hence,abnormal-
ities detected in this study could have been
3Diniz P.P.V.P., De Morais H.S.A., Schwartz
D.S., Cardiac troponin I in dogs naturally in-
fected by Ehrlichia canis, J. Vet. Intern. Med.
(2004) 18:454-455 (Abstract).
due to infection with either E. canis or
E. canis, B. v. berkhoffii and B. henselae
co-infection, and/or due to other undeter-
Three genes were targeted in this study
to confirm Bartonella co-infection in Dog
#2 and to better characterize the B. hense-
lae isolate from Dog #1. Partial ITS re-
100% homologous to B. henselae CAL-1
(546 bp/546 bp) (Tab. IV), which was iso-
lated from a septicemic human being in
the USA . However, partial pap31 se-
quences were 99.8% homologous to strain
Fizz (514 bp/514 bp) (Tab. III), which
was isolated from a bacteremic cat in
Switzerland . These two strains (CAL-
1 and Fizz) are 98% homologous within
the ITS region (971/983 bp) and 100%
homologous within pap31 gene. This sug-
gests that B. henselae strains CAL-1 and
Fizz may belong to the same genetic vari-
ant, as suggested previously . The rpoB
gene sequences from Dog #1 were 99.8%
homologous (655/656 bp) to B. hense-
lae Houston-1.Potentially, the discrepancy
Bartonella species in Brazilian dogs707
among sequenced genes from Dog #1 can
be explained by the absence of CAL-1
and Fizz rpoB sequences in the GenBank
database and by the high homology of
rpoB gene among Bartonella species. Se-
quences of longer length and targeting ad-
ditional genes would be necessary for a
more complete phylogenetic classification
of this B. henselae isolate .
In this study, unanticipated co-infection
with B. henselae and B. v. berkhoffii was
successfully documented in a dog for the
first time (Dog #2). By independently tar-
geting the ITS region and pap31 gene
with species-specific primers, Dog #2 was
found to be co-infected with B. henselae
Houston-1 and B. v. berkhoffii type I. The
B. v. berkhoffii pap31 sequence from Dog
#2 differed in 1 bp among 548 bp from
a B. v. berkhoffii type I isolate (Y-12) ob-
tained from a coyote  and 3 bp from B.
v. berkhoffii type I (strain 93-CO-1, ATCC
51672) isolated from a dog with endo-
carditis .Targetinga third gene(rpoB),
infection with B. v. berkhoffii was con-
strain 93-CO-1). Because colonies con-
taining both organisms were unknowingly
pooled and extracted together, molecular
confirmation of co-infection by targeting
multiple genes was required. This obser-
vation indicates that dogs, as is well rec-
ognized for cats, can be co-infected with
more than one Bartonella species.
Bartonella PCR and IFA test results
were discrepant among many of the dogs
in this study (Tab. I). Different antigenic
expression among Bartonella strains could
result in false negative serology results, as
previously shown in humans  and doc-
umented in our laboratory . Gene ex-
pression and pathogenicity of B. henselae
strains are based upon incompletely char-
acterized mechanisms and do not directly
correlate with genetic differences .
Furthermore, since clinically ill dogs were
selected for this study, acute infection
cases might have been preferentially in-
cluded, resulting in PCR positivity prior to
development of a detectable humoral im-
mune response, as found in Dog #1.
plifiable Bartonella DNA in conjunction
with low B. henselae and B. v. berkhoffii
seroreactivity detected in this study indi-
transmitted to dogs with a high risk of
tick infestation in southeastern Brazil. The
two species of Bartonella detected from
sick dogs in Brazil were homologous to
other species previously isolated from sep-
ticemic human beings, coyotes or dogs
from the USA. Bartonella co-infection
with E. canis or A. platys might increase
disease severity and contribute to myocar-
This study was supported in part by the
state of North Carolina and grants provided by
Fundação de Amparo à Pesquisa do Estado de
São Paulo – Brazil (FAPESP, grants number
02/11117-1, 03/02627-9 and 04/09313-2), Co-
ordenação de Aperfeiçoamento de Pessoal de
Ensino Superior – Brazil (CAPES), Waltham
foundation (grant number 2006-0255), and
 Altschul S.F., Gish W., Miller W., Myers
E.W., Lipman D.J., Basic local alignment
search tool, J. Mol. Biol. (1990) 215:403–
 ArvandM., Schubert
Emergenceof distinct genetic variants
in the population of primary Bartonella
henselae isolates, Microbes Infect. (2006)
 Azevedo Z.M., Higa L.Y., Boechat P.R.,
Boechat M.B., Klaplauch F., Cat-scratch dis-
ease caused by Bartonella quintana in an
infant: an unusual presentation, Rev. Soc.
Bras. Med. Trop. (2000) 33:313–317 (in
708P.P.V.P. Diniz et al.
 Baneth G., Breitschwerdt E.B., Hegarty
B.C., Pappalardo B., Ryan J., A survey of
tick-borne bacteria and protozoa in naturally
exposed dogs from Israel, Vet. Parasitol.
 Barber R.D., Harmer D.W., Coleman R.A.,
Clark B.J., GAPDH as a housekeeping
gene: analysis of GAPDH mRNA expres-
sion in a panel of 72 human tissues, Physiol.
Genomics (2005) 21:389–395.
 Barnes A., Bell S.C., Isherwood D.R.,
Bennett M., Carter S.D., Evidence of
Bartonella henselae infection in cats and
dogs in the United Kingdom, Vet. Rec.
 Birkenheuer A.J., Levy M.G., Breitschwerdt
E.B., Development and evaluation of a sem-
inested PCR for detection and differentiation
of Babesia gibsoni (Asian genotype) and B.
canis DNA in canine blood samples, J. Clin.
Microbiol. (2003) 41:4172–4177.
 Boulouis H.J., Chang C.C., Henn J.B.,
Kasten R.W., Chomel B.B., Factors associ-
ated with the rapid emergence of zoonotic
Bartonella infections, Vet. Res. (2005)
 Breitschwerdt E.B., Kordick D.L., Malarkey
D.E., Keene B., Hadfield T.L., Wilson K.,
Endocarditis in a dog due to infection
with a novel Bartonella subspecies, J. Clin.
Microbiol. (1995) 33:154–160.
 Breitschwerdt E.B., Hegarty B.C., Hancock
S.I., Sequential evaluation of dogs naturally
infected with Ehrlichia canis, Ehrlichia
ewingii, or Bartonella vinsonii, J. Clin.
Microbiol. (1998) 36:2645–2651.
 Breitschwerdt E.B.,AtkinsC.E.,Brown T.T.,
Kordick D.L., Snyder P.S., Bartonella vin-
sonii subsp. berkhoffii and related members
of thealpha subdivision ofthe Proteobacteria
in dogs with cardiac arrhythmias, endocardi-
tis, or myocarditis, J. Clin. Microbiol. (1999)
Bartonella infection in animals: carrier-
ship, reservoir potential, pathogenicity, and
zoonotic potential for human infection, Clin.
Microbiol. Rev. (2000) 13:428–438.
 Brown D.J., Rush J.E., MacGregor J., Ross
J.N. Jr., Brewer B., Rand W.M., M-mode
echocardiographic ratio indices in normal
dogs, cats, and horses: a novel quantitative
method, J. Vet. Intern. Med. (2003) 17:653–
 Chomel B.B., Mac Donald K.A., Kasten
R.W., Chang C.C., Wey A.C., Foley J.E.,
Thomas W.P., Kittleson M.D., Aortic valve
endocarditis in a dog due to Bartonella clar-
ridgeiae, J.Clin.Microbiol.(2001) 39:3548–
 Chomel B.B., Wey A.C., Kasten R.W.,
Isolation of Bartonella washoensis from a
dog with mitral valve endocarditis, J. Clin.
Microbiol. (2003) 41:5327–5332.
 Chomel B.B., Boulouis H.J., Maruyama S.,
Breitschwerdt E.B., Bartonella spp. in pets
and effect on human health, Emerg. Infect.
Dis. (2006) 12:389–394.
 Dantas-Torres F., Figueredo L.A., Brandao-
Filho S.P., Rhipicephalus sanguineus (Acari:
Ixodidae), the brown dog tick, parasitizing
humans inBrazil, Rev. Soc. Bras. Med. Trop.
 De Lemos E.R., Machado R.D., Coura J.R.,
Guimaraes M.A., Freire N.M., Amorim M.,
Gazeta G.S., Epidemiological aspects of the
Brazilian spotted fever: seasonal activity of
ticks collected in an endemic area in Sao
Paulo, Brazil, Rev. Soc. Bras. Med. Trop.
 Drancourt M., Moal V., Brunet P., Dussol
B., Berland Y., Raoult D., Bartonella
seronegative hemodialyzed patient, J. Clin.
Microbiol. (1996) 34:1158–1160.
 Duncan A.W., Maggi R.G., Breitschwerdt
E.B., A combined approach for the enhanced
detection and isolation of Bartonella species
in dog blood samples: pre-enrichment liquid
culture followed by PCR and subculture onto
agar plates, J. Microbiol. Methods (2007)
 Figueiredo L.T., Badra S.J., Pereira L.E.,
Szabo M.P., Report on ticks collected in the
Southeast and Mid-West regions of Brazil:
analyzing the potential transmission of tick-
borne pathogens to man, Rev. Soc. Bras.
Med. Trop. (1999) 32:613–619.
 Gillespie T.N., Washabau R.J., Goldschmidt
Breitschwerdt E.B., Detection of Bartonella
henselae and Bartonella clarridgeiae DNA
in hepatic specimens from two dogs with
hepatic disease, J. Am. Vet. Med. Assoc.
(2003) 222:47–51, 35.
 Gundi V.A., Bourry O., Davous B., Raoult
D., La Scola B., Bartonella clarridgeiae and
B. henselae in dogs, Gabon, Emerg. Infect.
Dis. (2004) 10:2261–2262.
 Higgins R., Emerging or re-emerging bac-
terial zoonotic diseases: bartonellosis, lep-
tospirosis, Lyme borreliosis, plague, Rev.
Sci. Tech. (2004) 23:569–581.
infection in a
Bartonella species in Brazilian dogs709
 Holmes A.H., Greenough T.C., Balady G.J.,
Regnery R.L., Anderson B.E., O’Keane
J.C., Fonger J.D., McCrone E.L., Bartonella
henselae endocarditis in an immunocompe-
tent adult, Clin. Infect. Dis. (1995) 21:1004–
 Houpikian P., Raoult D., 16S/23S rRNA
intergenic spacer regions for phylogenetic
analysis, identification, and subtyping of
J. Clin. Microbiol.
 Joblet C., Roux V., Drancourt M., Gouvernet
J., Raoult D., Identification of Bartonella
(Rochalimaea) species among fastidious
gram-negative bacteria on the basis of the
partial sequence of the citrate-synthase gene,
J. Clin. Microbiol. (1995) 33:1879–1883.
 Keret D., Giladi M., Kletter Y., Wientroub
S., Cat-scratch disease osteomyelitis from a
dog scratch, J. Bone Joint Surg. Br. (1998)
 Kordick D.L.,Swaminathan B., Greene C.E.,
Wilson K.H., Whitney A.M., O’Connor
S., Hollis D.G., Matar G.M., Steigerwalt
A.G., Malcolm G.B., Hayes P.S., Hadfield
T.L., Breitschwerdt E.B., Brenner D.J.,
Bartonella vinsonii subsp. berkhoffii subsp.
nov., isolated from dogs; Bartonella vinsonii
subsp. vinsonii; and emended description of
Bartonella vinsonii, Int. J. Syst. Bacteriol.
 Kordick S.K., Breitschwerdt E.B., Hegarty
Hancock S.I., Bradley J.M., Rumbough
R., McPherson J.T., MacCormack J.N.,
pathogens in a Walker Hound kennel in
North Carolina, J. Clin. Microbiol. (1999)
 Li W., Chomel B.B., Maruyama S., Guptil
L., Sander A., Raoult D., Fournier P.E.,
Multispacer typing to study the genotypic
distribution of Bartonella henselae popula-
tions, J. Clin. Microbiol. (2006) 44:2499–
 MacDonald K.A., Chomel B.B., Kittleson
M.D., Kasten R.W., Thomas W.P., Pesavento
P., A prospective study of canine infective
endocarditis in northern California (1999–
2001): emergence of Bartonella as a preva-
lent etiologic agent, J. Vet. Intern. Med.
 Maggi R.G., Breitschwerdt E.B., Isolation
of bacteriophages from Bartonella vinsonii
subsp. berkhoffii and the characterization of
Pap31 gene sequences from bacterial and
phage DNA, J. Mol. Microbiol. Biotechnol.
 Maggi R.G., Breitschwerdt E.B., Potential
limitations of the 16S-23S rRNA intergenic
region for molecular detection of Bartonella
species, J. Clin. Microbiol. (2005) 43:1171–
 Maggi R.G., Duncan A.W., Breitschwerdt
E.B., Novel chemically modified liquid
medium that will support the growth of
seven Bartonella species, J. Clin. Microbiol.
 Maggi R.G., Chomel B., Hegarty B.C., Henn
J., Breitschwerdt E.B., A Bartonella vinsonii
berkhoffii typing scheme based upon 16S-
23S ITS and Pap31 sequences from dog,
coyote, gray fox, and human isolates, Mol.
Cell. Probes (2006) 20:128–134.
 Mexas A.M., Hancock S.I., Breitschwerdt
E.B., Bartonella henselae and Bartonella
elizabethae as potential canine pathogens, J.
Clin. Microbiol. (2002) 40:4670–4674.
 O’Brien P.J., Smith D.E., Knechtel T.J.,
Marchak M.A., Pruimboom-Brees I., Brees
D.J.,Spratt D.P., Archer F.J., Butler P., Potter
A.N., Provost J.P., Richard J., Snyder P.A.,
Reagan W.J., Cardiac troponin I is a sen-
sitive, specific biomarker of cardiac injury
in laboratory animals, Lab. Anim. (2006)
 Pappalardo B.L.,
Epidemiologic evaluation of the risk factors
associated with exposure and seroreactivity
to Bartonella vinsonii in dogs, Am. J. Vet.
Res. (1997) 58:467–471.
 Riess T., Raddatz G., Linke D., Schafer A.,
Kempf V.A., Analysis of Bartonella adhesin
A expression reveals differences between
various B. henselae strains, Infect. Immun.
 Siciliano R.F., Strabelli T.M., Zeigler R.,
Rodrigues C., Castelli J.B., Grinberg M.,
Colombo S., da Silva L.J., Mendes do
Nascimento E.M., Pereira dos Santos F.C.,
Uip D.E., Infective endocarditis due to
Bartonella spp. and Coxiella burnetii: ex-
perience at a cardiology hospital in Sao
Paulo, Brazil, Ann. NY Acad. Sci. (2006)
 Silva E.S., Gontijo C.M., Pacheco R.S.,
Fiuza V.O., Brazil R.P., Visceral leishma-
niasis in the Metropolitan Region of Belo
Horizonte, State of Minas Gerais, Brazil,
Mem. Inst. Oswaldo Cruz (2001) 96:285–
710P.P.V.P. Diniz et al.
 Solano-Gallego L., Bradley J., Hegarty B.,
Sigmon B., Breitschwerdt E., Bartonella
henselae IgG antibodies are prevalent in
dogs from southeastern USA, Vet. Res.
 Solano-Gallego L., Kidd L., Trotta M.,
Di Marco M., Caldin M., Furlanello T.,
Breitschwerdt E., Febrile illness associated
with Rickettsia conorii infection in dogs
from Sicily, Emerg. Infect. Dis. (2006)
 Solano-Gallego L., Llull J., Osso M.,
Hegarty B., Breitschwerdt E., A serolog-
ical study of exposure to arthropod-borne
pathogens in dogs from northeastern Spain,
Vet. Res. (2006) 37:231–244.
 Stepien R.L., Rapoport G.S., Henik R.A.,
Wenholz L., Thomas C.B., Comparative di-
agnostic test characteristics of oscillometric
and Doppler ultrasonographic methods inthe
detection of systolic hypertension in dogs, J.
Vet. Intern. Med. (2003) 17:65–72.
 Stollberger C., Molzer G., Finsterer J.,
Seroprevalence of antibodies to microorgan-
isms known to cause arterial and myocardial
damage in patients with or without coronary
stenosis, Clin. Diagn. Lab. Immunol. (2001)
 Suksawat J., Xuejie Y., Hancock S.I.,
Hegarty B.C., Nilkumhang P., Breitschwerdt
E.B., Serologic and molecular evidence
of coinfection with multiple vector-borne
pathogens in dogs from Thailand, J. Vet.
Intern. Med. (2001) 15:453–462.
 Tilley L.P., Essentials of canine and feline
electrocardiography: interpretation and treat-
ment, Lea & Febiger, 3rd ed., 1992, p. 470.
 Tsukahara M., Tsuneoka H., Iino H., Ohno
K., Murano I., Bartonella henselae infection
from a dog, Lancet (1998) 352:1682.
 WHO, Human leptospirosis: guidance for
diagnosis, surveillance and control, World
Health Organization, 2003, p. 122.
 Yamanouchi H., Izumikawa K., Hisamatsu
T., Yoshinaga M., Sasaki E., Izumikawa
K., Hayakawa T., Hara K., Maruyama
S., Ohtani H., Shimokawa I., A case of
Bartonella henselae infection from a dog,
Kansenshogaku Zasshi (2004) 78:270–273.