Skip to main content
NCBI home page
Applied and Environmental Microbiology logoLink to Applied and Environmental Microbiology
. 2015 Jul 21;81(16):5613–5621. doi: 10.1128/AEM.01370-15

Relationship between the Presence of Bartonella Species and Bacterial Loads in Cats and Cat Fleas (Ctenocephalides felis) under Natural Conditions

Ricardo Gutiérrez 1, Yaarit Nachum-Biala 1, Shimon Harrus 1,✉
Editor: H Goodrich-Blair
PMCID: PMC4510192  PMID: 26070666

Abstract

Cats are considered the main reservoir of three zoonotic Bartonella species: Bartonella henselae, Bartonella clarridgeiae, and Bartonella koehlerae. Cat fleas (Ctenocephalides felis) have been experimentally demonstrated to be a competent vector of B. henselae and have been proposed as the potential vector of the two other Bartonella species. Previous studies have reported a lack of association between the Bartonella species infection status (infected or uninfected) and/or bacteremia levels of cats and the infection status of the fleas they host. Nevertheless, to date, no study has compared the quantitative distributions of these bacteria in both cats and their fleas under natural conditions. Thus, the present study explored these relationships by identifying and quantifying the different Bartonella species in both cats and their fleas. Therefore, EDTA-blood samples and fleas collected from stray cats were screened for Bartonella bacteria. Bacterial loads were quantified by high-resolution melt real-time quantitative PCR assays. The results indicated a moderate correlation between the Bartonella bacterial loads in the cats and their fleas when both were infected with the same Bartonella species. Moreover, a positive effect of the host infection status on the Bartonella bacterial loads of the fleas was observed. Conversely, the cat bacterial loads were not affected by the infection status of their fleas. Our results suggest that the Bartonella bacterial loads of fleas are positively affected by the presence of the bacteria in their feline host, probably by multiple acquisitions/accumulation and/or multiplication events.

INTRODUCTION

Bartonellae are vector-borne hemotropic bacteria of numerous mammalian hosts, in which they typically establish persistent and subclinical infections (1). Several Bartonella species are considered pathogens of many incidental hosts, including humans and domesticated animals (2). Cats (Felis catus) are the reservoirs of three zoonotic Bartonella species: Bartonella henselae, Bartonella clarridgeiae, and Bartonella koehlerae (2). Of these, B. henselae is considered the main causative agent of cat scratch disease (CSD) in humans (3), while B. clarridgeiae has been implicated as a cause of a CSD-like disease in several cases (4–6), and B. koehlerae was reported to cause endocarditis in humans and dogs (7, 8). Commonly, infected cats are subclinical persistent carriers of these Bartonella species (9–11). Moreover, other Bartonella species, including Bartonella quintana (12, 13), Bartonella bovis (14), and Bartonella vinsonii subsp. berkhoffii (15), were occasionally isolated from cats. Cat fleas (Ctenocephalides felis) are considered the major vector of feline bartonellae. They have been experimentally proved to be a competent vector of B. henselae (11) and have been proposed as the potential vector of B. clarridgeiae and B. koehlerae, since DNA sequences from these Bartonella species have been commonly detected in C. felis collected from cats worldwide (16).

The distribution of Bartonella species in cat populations varies notably across geographic regions. Infection rates of 0 to 62% in cats have been reported worldwide (17–19) but can reach even higher percentages in isolated cat populations (11). Although the natural transmission of Bartonella species among cats requires the presence of fleas (11, 20) and no vertical transmission of bartonellae has been proven in either cats (maternal transmission; 21, 22) or cat fleas (23), previous studies reported a lack of association between Bartonella infection in cats and their fleas (11, 18). The presence and/or level of infection (bacterial loads) of the Bartonella species in the cats tested did not mirror the infection status of the fleas they hosted (11, 18). It has also been observed that infected cats can host negative fleas and vice versa or can be infected with different Bartonella species (18). The latter studies have based the infection status and quantification of bacterial loads on bacterial isolation (11, 18). However, it has been shown that isolation-based methods underestimate the rates of Bartonella species infection in cats in comparison to molecular-analysis-based assays (24). Thus, investigation of the distribution of Bartonella species and their infection loads in both hosts and vectors through a molecular-analysis-based quantitative approach was warranted. This study aimed to further investigate the relationships between Bartonella bacterial loads in cats and their fleas under natural conditions by quantifying their infection loads by high-resolution melt (HRM) real-time quantitative PCR (qPCR) assays and evaluating ecological factors that could influence the infection status and/or the level of Bartonella bacteremia.

MATERIALS AND METHODS

Animal sample collection.

Thirty-six stray cats (21 females and 15 males) were caught in the eastern suburbs of Rishon-LeZion, Israel, during August to November 2012 as part of a neutering campaign by the Municipal Veterinary Services of the city. Nineteen of the 36 cats, were kittens (<1 year old), and 17 were adults (1 to 4 years old). The animals were anesthetized with a combination of ketamine at 10 mg/kg and xylazine at 1 mg/kg injected intramuscularly. Blood samples were drawn from the cephalic or jugular vein of each cat into EDTA tubes after disinfection of the skin with 70% ethanol. All of the fleas found on each cat (one to six per cat), were collected and kept in 70% ethanol until further analyzed. After collection, EDTA-blood and flea samples were kept in a cool box (4°C), transported to the laboratory, and kept frozen at −20°C and room temperature, respectively. Prior to molecular analysis, the fleas were taxonomically identified and sexed by morphological characteristics by microscopy. The study was approved by the Hebrew University Institutional Animal Care and Use Committee (MD-12-13461-2).

Screening for feline retroviruses FeLV and FIV.

Cat EDTA-blood samples were screened for feline immunodeficiency virus (FIV) antibodies and feline leukemia virus (FeLV) antigen with the commercial combined SNAP Combo Plus test (IDEXX Inc., Westbrook, ME).

DNA extraction. (i) Blood samples.

Genomic DNA was extracted from 50 μl of EDTA-blood from each cat with a DNA extraction kit (BiOstic Bacteremia DNA isolation kit; MO BIO Laboratories, Inc., Carlsbad, CA). DNA was obtained in 50 μl of elution buffer. For quality assurance, a sample with all of the reagents except blood was processed in parallel with the blood samples and used as a negative control.

(ii) Flea samples.

DNA was extracted individually from 90 fleas (1 to 3 per cat; 35 female fleas, 19 male fleas, and 36 fleas whose gender was undetermined) as follows. Each flea was washed once in 1 ml of 70% ethanol for 5 min and three times in 1 ml of sterile phosphate-buffered saline (PBS) for 5 min. Thereafter, the flea was homogenized in 50 μl of sterile PBS with a sterile pestle until a clear solution was obtained. Finally, the DNA was extracted with a DNA extraction kit (Illustra Tissue & Cells GenomicPrep Mini Spin kit; GE Healthcare, Buckinghamshire, United Kingdom) with an incubation step in lysis buffer and proteinase K for 2.5 h. DNA was obtained in 100 μl of elution buffer. For quality assurance, a sample with all of the reagents but no flea components was processed in parallel with the samples and used as a negative control.

Molecular quantification of Bartonella species bacterial loads in cats and fleas by HRM qPCR for Bartonella DNA.

An intercalating-dye-based qPCR assay was developed to quantify the Bartonella bacterial loads in the cat and flea samples. Accordingly, a 190-bp fragment of the internal transcribed spacer (ITS) was targeted with primers 321s (AGATGATGATCCCAAGCCTTCTGG) and H493as (TGAACCTCCGACCTCACGCTTATC) (25). All reactions were performed in the Life Technologies StepOnePlus real-time PCR system (Thermo Fisher Scientific, Waltham, MA) in 96-well plates. The amplification protocol used was 4 min at 95°C, followed by 50 cycles of 5 s at 95°C, 30 s at 60°C (data collection on HRM reporter), and 2 s at 72°C. The HRM stage was performed at the end of the cycling stage as follows: 15 s at 95°C, followed by a temperature increase from 70 to 95°C (data collection set in 0.3%, HRM reporter). Genomic DNA samples obtained from Bartonella-free cat blood and Bartonella-free fleas were used as negative controls, and ultrapure water was used as a nontemplate control (NTC). The expected HRM profiles of Bartonella species ranged from 80 to 85°C (24); thus, melt profiles out of this range were considered nonspecific products.

The qPCR tests were carried out with a 20-μl final volume containing 0.5 μl of a 10 μM solution of each primer, 0.6 μl of a 50 μM Syto9 solution (Invitrogen, Carlsbad, CA), 10 μl of Maxima Hot-Start PCR Master Mix (2X) (Thermo Scientific, Surrey, United Kingdom), 3.9 μl of ultrapure water (Sigma-Aldrich, St. Louis, MO), 0.5 μl of a 25 μM solution of MgCl2, and 4 μl of each genomic DNA. Cat blood DNA was diluted 1:2 in ultrapure water for the analyses. Each qPCR run included duplicates of each (blood or flea) DNA sample, B. henselae reference points (described below), negative-control samples, and the NTC. Any amplification with a quantification cycle (Cq) value of <40 and with an HRM pattern within the expected range was sequenced and confirmed as a positive result.

All samples were retested with an additional Bartonella real-time qPCR assay in order to evaluate the performance of the ITS assay (see the supplemental material). Accordingly, a 380-bp fragment of the citrate synthase gene (gltA) was targeted with primers 443F (GCTATGTCTGCATTCTATCA) (26) and 781R (CCACCATGAGCTGGTCCCC) (based on Bhcs.781F from Norman et al. [27]). All reactions were run under amplification conditions equal to those used for the ITS HRM qPCR assay. The PCR tests were carried out with a 20-μl final volume containing 0.5 μl of a 10 μM solution of each primer, 0.6 μl of a 50 μM Syto9 solution (Invitrogen, Carlsbad, CA), 10 μl of Maxima Hot-Start PCR Master Mix (2X) (Thermo Scientific, Surrey, United Kingdom), 4.4 μl of ultrapure water (Sigma-Aldrich, St. Louis, MO), and 4 μl of each genomic DNA.

Sequencing.

All positive PCR products obtained in the ITS and gltA assays were purified with a PCR purification kit (Exo-SAP; New England BioLabs, Inc., Ipswich, MA) and sequenced by BigDye Terminator cycle sequencing chemistry with the Applied Biosystems ABI 3700 DNA Analyzer and the ABI Data Collection and Sequence Analysis software (ABI, Carlsbad, CA). Further analyses of sequences were done with MEGA alignment software, version 5.05 (The Biodesign Institute, Tempe, AZ).

B. henselae reference points for the standard curve.

A B. henselae isolate (cultured from blood of a stray cat from Israel) was used as the reference for quantification of Bartonella infection loads. Accordingly, a highly concentrated inoculum of fresh B. henselae colonies was homogenized in 200 μl of sterile PBS, and 10-fold dilutions ranging from 10−1 to 10−8 were prepared in sterile PBS (final volume of 400 μl). DNA was then extracted from 50 μl of each of the 10-fold bacterial dilutions and used as the reference points for the standard curve of the qPCR assay. DNA was extracted with the Illustra Tissue & Cells GenomicPrep Mini Spin DNA extraction kit as described above. The concentration of the original bacterial solution in CFU per microliter was determined by seeding 100 μl of each of the bacterial dilutions (10−6 to 10−8) in chocolate agar plates (in duplicate). The plates were incubated at 37°C in a 5% CO2 atmosphere for 7 days.

Standard curve for the ITS HRM qPCR assay.

The standard curve for the ITS HRM-qPCR assay was obtained by running triplicates of the B. henselae reference points and plotting their Cq values against the log of the number of Bartonella CFU/μl. The linear equation of the curve and the correlation coefficients (r2 values) were obtained with Microsoft Excel 2010 software (Microsoft Corporation, Redmond, WA). The amplification efficiency (E) of the qPCR was calculated with the formula E = (10−1/slope – 1) × 100. Threshold values were adjusted manually according to the geometric phase of the amplification plot of the reference points. Baselines were fixed manually (start 6, end 12 cycles) for all runs.

Real-time PCR for C. felis DNA.

The variability in flea size between the samples, which could potentially affect the DNA extraction yield, was corrected by targeting an internal reference gene of C. felis in a real-time PCR assay (see data analysis section below). Accordingly, a 120-bp fragment of the cytochrome oxidase subunit 2 gene (cox-2) of C. felis was targeted with primers cox-2F (CTGCTACCGATGTTCTTCATTCA) and cox-2R (TGTCCAAAATATAATCCTGGTCGAT) (this study). The qPCRs were carried out with a 20-μl final volume containing 0.15 μl of a 10 μM solution of each primer, 0.6 μl of a 50 μM Syto9 solution (Invitrogen, Carlsbad, CA), 5.1 μl of ultrapure water (Sigma-Aldrich, St. Louis, MO), 10 μl of Maxima Hot-Start PCR Master Mix (2X) (Thermo Scientific, Surrey, United Kingdom), and 4 μl of each flea DNA sample. The amplification protocol was the same as for the ITS HRM qPCR assay. The specificity of this assay for C. felis only was corroborated with DNA extracted from Ctenocephalides canis, Xenopsylla cheopis, Pulex irritans, and Bartonella species. All flea samples were run in duplicate. Threshold values were adjusted manually and baselines were fixed automatically according the StepOnePlus software version 2.2.2.

Real-time PCR for F. catus DNA.

An internal reference cat gene was used for a cat blood meal size estimation of all flea samples (see data analysis section below). Thus, a 286-bp amplicon of the cytochrome b fragment gene (cytB) from F. catus was screened with primers F3 (ATCTCAGCCTTAGCAGGAGTACAC) and R2 (TGGATCGGAGAATTGCGTATGCGA) (28). The PCR tests were carried out with a 20-μl final volume containing 0.3 μl of a 10 μM solution of each primer, 0.6 μl of a 50 μM Syto9 solution (Invitrogen, Carlsbad, CA), 4.8 μl of ultrapure water (Sigma-Aldrich, St. Louis, MO), 10 μl of Maxima Hot-Start PCR Master Mix (2X) (Thermo Scientific, Surrey, United Kingdom), and 4 μl of each flea DNA sample. The amplification protocol was identical to that of the other PCR assays (described above). The specificity of this assay for cat blood DNA only was corroborated with DNA extracted from dog blood, cow blood, Bartonella species, and C. felis fleas collected from dogs. All flea samples were run in duplicate. Threshold values were adjusted manually and baselines were fixed automatically according the StepOnePlus software version 2.2.2.

Data analysis.

The Bartonella ITS Cq values of positive samples were used to quantify the Bartonella bacterial load (number of CFU per flea or milliliter of blood) in each sample. First, the variability between runs was corrected by calculating the difference between the Cq of the reference point of each run and the Cq of the standard curve point. The latter was subtracted from the mean Cq of each positive sample. Additionally, the Bartonella ITS Cq values from fleas were corrected according to flea size, with the internal Cq values of the internal reference gene for C. felis (cox-2) as described elsewhere (23, 29). Finally, the Bartonella bacterial load (number of CFU per μl) in each sample was calculated according to the equation of the linear regression of the standard curve. For a blood meal size analysis, the difference between the mean Cq value of the cytB cat gene in each flea and the cox-2 flea gene Cq value was calculated (ΔCq) and then divided by the ratio of the Bartonella bacterial load of the host's blood and the median bacterial load of the cat population tested in this study. The result of the latter equation was used as a relative measurement of the blood meal size of each flea.

Statistical analyses.

The Bartonella bacterial loads obtained by ITS quantification were compared with those obtained by gltA quantification by Pearson's correlation test. The bacterial loads in the cat blood samples were compared according to the cat gender, categorical age (kitten versus adult), Bartonella species detected, FeLV infection status, and the Bartonella infection status of the fleas by the nonparametric Mann-Whitney U and Kruskal-Wallis tests for two or more than two independent groups, respectively. The relationship between the categorical variables of Bartonella and FeLV infection statuses was evaluated by Fisher's exact test. Bartonella bacterial loads in fleas were quantified according to flea sex by paired analysis with the Wilcoxon signed-rank test. The effects of the cat host infection status and the Bartonella species detected in the fleas were evaluated by the Kruskal-Wallis and Mann-Whitney U tests. Spearman's rho correlation coefficient test was used to evaluate the Bartonella loads in cat blood and the average bacterial load of the fleas infected with the same Bartonella species and to evaluate if there is a correlation between the bacterial load in the fleas and their relative blood meal size (both analyses were performed only for fleas infected with the same Bartonella species as their hosts). All statistical analyses were performed in IBM SPSS Statistics software version 20 (IBM Corp., Armonk, NY). Median values and interquartile ranges (IQRs) of the groups were calculated. Statistical significance was defined as P < 0.05. Significance levels were adjusted for multiple tests by Bonferroni correction.

RESULTS

ITS HRM qPCR assay performance.

The concentration of the original B. henselae bacterial solution was 9.7 × 106 CFU/μl. The standard curve for the ITS HRM qPCR assay showed an efficiency (E) of 90.2%, a y intercept of 34.204, a slope of −3.583, and a linear correlation (R2) of 0.9941. The assay showed a detection limit of 0.32 CFU/4 μl of extracted DNA. However, the assay maintained linearity until a concentration of 1.5 CFU/4 μl of extracted DNA; thus, the latter was considered the quantifiable detection limit (QDL), corresponding to a Cq of 33.6. Consequently, the Bartonella bacterial loads in positive samples (confirmed by sequencing) with Cq values higher than 33.6 (QDL) were defined as 1.5 CFU/4 μl. The Bartonella bacterial loads estimated by the ITS qPCR assay correlated significantly with those obtained by the confirmatory gltA qPCR assay (Pearson's correlation, r = 0.907, P < 0.001; see Fig. S1 in the supplemental material).

Detection of Bartonella DNA in cats and fleas.

The detection of Bartonella DNA in cats and their fleas is shown in Table 1. Sixty-four percent (23/36) of the cats were positive for Bartonella DNA. Of these, B. clarridgeiae DNA was identified in 36.1% (13/36), B. henselae DNA was identified in 22.2% (8/36), and B. koehlerae DNA was identified in 5.6% (2/36). The Bartonella species was confirmed in 91.3% (21/23) of the cases by the gltA assay (see Table S1 in the supplemental material). No additional coinfecting Bartonella species was detected in the positive cat blood samples, either when a different locus was targeted or by detection of a double HRM pattern. Seventy-eight percent (18/23) of the positive cats harbored at least one flea infected with the same Bartonella species, 47.8% (11/23) of the bacteremic cats harbored at least one flea infected with a different Bartonella species, and 26.1% (6/23) of the bacteremic cats hosted at least one negative flea. Moreover, 84.6% (11/13) of the nonbacteremic cats harbored at least one Bartonella-positive flea (Table 1).

TABLE 1.

Detection of Bartonella DNA in cats and their fleas, arranged according to the Bartonella species in the host cata

No. Host ID Bartonella species identified in:
Host cat Flea 1 Flea 2 Flea 3
1 3 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. clarridgeiae
2 29 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. clarridgeiae
3 30 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. clarridgeiae
4 32 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. clarridgeiae
5 15 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. elizabethae-like organism
6 38 B. clarridgeiae B. clarridgeiae B. clarridgeiae B. clarridgeiae, B. henselaeb
7 7 B. clarridgeiae B. henselae B. henselae Negative
8 26 B. clarridgeiae B. clarridgeiae B. clarridgeiae Negative
9 20 B. clarridgeiae B. clarridgeiae B. clarridgeiae
10 18 B. clarridgeiae B. clarridgeiae B. henselae
11 36 B. clarridgeiae B. henselae B. henselae
12 1 B. clarridgeiae Negative Negative
13 21 B. clarridgeiae B. clarridgeiae
14 8 B. henselae B. henselae B. henselae B. henselae
15 10 B. henselae B. henselae B. henselae B. henselae
16 11 B. henselae B. henselae B. henselae B. henselae
17 14 B. henselae B. henselae B. henselae B. henselae, B. koehleraeb
18 35 B. henselae B. henselae B. henselae B. elizabethae-like organism
19 37 B. henselae B. henselae B. elizabethae-like organism Negative
20 9 B. henselae B. henselae B. clarridgeiae Negative
21 19 B. henselae B. elizabethae-like organism Negative Negative
22 12 B. koehlerae B. clarridgeiae B. clarridgeiae
23 16 B. koehlerae B. koehlerae
24 22 Negative B. clarridgeiae B. clarridgeiae B. clarridgeiae
25 5 Negative B. clarridgeiae B. clarridgeiae Negative
26 2 Negative B. clarridgeiae Negative Negative
27 33 Negative B. henselae Negative Negative
28 28 Negative B. elizabethae-like organism Negative Negative
29 27 Negative Negative Negative Negative
30 4 Negative B. clarridgeiae B. clarridgeiae
31 24 Negative B. henselae B. clarridgeiae
32 6 Negative B. elizabethae-like organism Negative
33 13 Negative B. clarridgeiae Negative
34 23 Negative B. henselae
35 34 Negative B. henselae
36 17 Negative Negative Negative
Open in a new tab
a

Sequence identification was performed according to the GenBank database (identities of 97 to 100%).

b

Coinfection case identified while sequencing different Bartonella loci (ITS and gltA).

Bartonella DNA was detected in 75.6% (68/90) of the C. felis fleas. DNA of four different Bartonella species was identified in these fleas. Single infection with B. clarridgeiae was detected in 38.9% (35/90), B. henselae in 26.7% (24/90), Bartonella elizabethae-like bacteria in 6.7% (6/90), and B. koehlerae in 1.1% (1/90) of the fleas. In 79.4% (54/68) of the ITS-positive samples, the gltA fragment was also amplified (see Table S1 in the supplemental material). In 96% (52/54) of these cases, the Bartonella species was confirmed. Fleas infected with B. elizabethae-like organisms harbored ITS sequences 100% identical to that of B. elizabethae (GenBank accession number L35103.1) and gltA sequences 98% identical to that of Bartonella tribocorum (GenBank accession number HG969192.1). It is noteworthy that both ITS and gltA sequences were found to be 100% identical to Bartonella sp. strain Tel Aviv Rr, which was previously isolated from commensal rats in Israel (30). In two cases, the ITS and gltA sequences from the same flea corresponded to different feline Bartonella species and were defined as coinfections (Table 1; see Table S1 in the supplemental material). Moreover, one sample was found to be positive for B. clarridgeiae DNA only by gltA real-time PCR assay (see Table S1). Of the fleas collected from bacteremic cats, 66.7% (40/60) were positive for the same Bartonella species as their host, while 20.0% (12/60) harbored a different Bartonella species and 13.3% (8/60) were negative for Bartonella DNA.

FeLV and FIV infection status of the cats.

Eighty-one percent (29/36) of the cats were positive for the FeLV antigen, while only 5.6% (2 cats) carried antibodies against FIV. None of the cats was positive for both FeLV antigen and FIV antibodies.

Quantification of Bartonella bacterial loads in cats and fleas.

The Bartonella bacterial loads determined in the positive cats ranged from 7.5 × 102 to 5.7 × 105 CFU/ml of blood, with a median bacterial load of 2.7 × 103 CFU/ml of blood (IQR, 1.1 × 104). The bacterial loads were not significantly different between female and male positive cats (Mann-Whitney U test, P = 0.727) or between kittens and adult cats that were positive (Mann-Whitney U test, P = 0.309). No differences in bacterial loads were observed between the cats either according to the Bartonella species identified (Kruskal-Wallis test, P = 0.133) or according to the FeLV infection status (infected versus uninfected; Mann-Whitney U test, P = 0.754). Moreover, no relationship was found between Bartonella infection status and FeLV infection status (Fisher's exact test, P = 0.501).

The bacterial loads in single Bartonella-infected fleas ranged from 3.8 × 101 to 2.4 × 106 CFU/flea. The median bacterial loads of female and male fleas were 1.2 × 103 (IQR, 2.0 × 105) and 1.2 × 102 (IQR, 6.1 × 103) CFU/flea, respectively. No statistically significant difference was found between female and male flea pairs collected from the same host and infected with the same Bartonella species (paired analysis, Wilcoxon signed-rank test, P = 0.237). The Bartonella bacterial loads varied significantly according to the Bartonella species identified in the fleas (Kruskal-Wallis test, P = 0.009). The greater bacterial loads were determined in fleas infected with the three feline Bartonellae species (B. henselae, B. clarridgeiae, and B. koehlerae), while fleas infected with the B. elizabethae-like bacteria had bacterial loads below the QDL (37.5 CFU/flea). No significant difference was observed between the bacterial loads of fleas infected with the most prevalent feline Bartonella species, B. clarridgeiae and B. henselae (Mann-Whitney U test, P = 0.415). Moreover, positive fleas with the same Bartonella species as their hosts showed no correlation between their relative blood meal size and their Bartonella bacterial load (Spearman's rho, rs = −0.322, P > 0.05).

The infection status of the cat host and/or the flea ectoparasites had contrasting effects on the Bartonella bacterial loads in the cats and/or the fleas. The Bartonella bacterial loads in fleas differed significantly according to the infection status of their host (Kruskal-Wallis test, P < 0.0005; Fig. 1). Fleas infected with the same Bartonella species as their host cat harbored greater bacterial loads than fleas collected from nonbacteremic cats (Mann-Whitney U test, P < 0.0005) or fleas infected with a Bartonella species different from that of their cat hosts (Mann-Whitney U test, P = 0.012). On the contrary, the cat bacterial loads did not vary significantly according to their fleas' infection status (Fig. 2). The Bartonella bacterial loads of cats that harbored at least one flea infected with the same Bartonella species did not show any significant difference from the Bartonella bacterial loads of cats that did not harbor any flea with the same infection status (Mann-Whitney U test, P = 0.259). Nevertheless, in cases where the flea and the cat hosted the same Bartonella species, a moderate positive correlation was observed between the average Bartonella loads of the positive fleas and the bacterial loads of their host cats (Spearman's rho, rs = 0.553, P = 0.017; Fig. 3). This correlation was also confirmed with the gltA qPCR data (Spearman's rho, rs = 0.742, P = 0.002; see Fig. S2 in the supplemental material).

FIG 1.

FIG 1

Open in a new tab

Effect of the Bartonella infection status of host cats on the Bartonella bacterial loads of their fleas. Box plots illustrate the distribution of the samples according to the infection status of the host cat. Boxes represent IQRs, and horizontal black thick lines represent median values. Vertical lines (whiskers) represent the distribution of maximum and minimum values. The symbols ° and * represent outliers. The values on the y axis are on a log scale. Kruskal-Wallis test, P < 0.0005.

FIG 2.

FIG 2

Open in a new tab

Effect of the Bartonella infection status of fleas on the Bartonella loads of their host cats. Box plots illustrate the distribution of the samples according to the infection status of the fleas. Boxes represent IQRs, and horizontal black thick lines representing median values. Vertical lines (whiskers) represent the distribution of maximum and minimum values. The symbol ° represents an outlier. The values on the y axis are on a log scale. Mann-Whitney U test, P = 0.259.

FIG 3.

FIG 3

Open in a new tab

Correlation of Bartonella bacterial loads of host cats and their fleas infected with the same Bartonella species. Each flea bacterial load point represents the average of the fleas collected from a host cat in which the same Bartonella species was detected. The values on the axes are on log scales. Spearman's rho correlation coefficient, rs = 0.553, P = 0.017.

DISCUSSION

In this study, Bartonella bacterial loads were measured and compared between cats and their fleas in a confined stray cat community. Our results showed a quantitative relationship between the presence of Bartonella bacteria in host cats and their fleas, corresponding to the dominant Bartonella species involved. Accordingly, a moderate positive correlation was found between the infection loads of the host cats and their fleas infected with the same Bartonella species. Moreover, fleas infected with the same Bartonella species as their host cats harbored larger numbers of bacteria than did fleas that carried Bartonella species other than those of their hosts or fleas collected from nonbacteremic cats. On the contrary, no significant differences were noted between the bacterial loads of cats that carried positive fleas for the same Bartonella species and those that carried fleas infected with other Bartonella species. Thus, the presence of a particular Bartonella species in the cat host seemed to have a positive effect on the bacterial loads of its fleas, but not the opposite. This relationship could be explained by continuous acquisitions/accumulations of bartonellae and/or multiplication events that might have occurred in the fleas, as opposed to the more complex infection dynamics in the host cats. It has been reported that B. henselae can persist and multiply in the gut of the flea (i.e., multiplication event) (23, 31), and since a flea can consume multiple meals from its host cat (32), new acquisitions of Bartonella organisms from the host may increase the number of bacteria in the flea (i.e., accumulation event). In addition, no correlation was observed between the relative amount of blood consumed by the flea and its Bartonella bacterial loads, suggesting that the Bartonella loads estimated are not necessarily a result of a recent blood meal. On the other hand, cyclic bacteremia (9, 22, 33) due to apparent cyclic release of bartonellae from a primary niche (or from secondary niches yet to be elucidated) into the bloodstream may explain the apparent independence of cat bacterial loads from the infection status of the fleas they host (infected or not infected with the same Bartonella species as the cat). Thus, it seems reasonable that, at a particular moment, the bacterial load of a cat is less affected by the presence of the Bartonella bacteria in its fleas, while bacterial loads in the fleas seem to be associated with new bacterial acquisition or input from the infected host.

The detection of a Bartonella species in a particular cat and its fleas has shown some inconsistencies in previous studies (11, 18, 34, 35). It has been observed that Bartonella-positive cats can harbor negative fleas or fleas infected with other Bartonella species and that nonbacteremic cats can harbor Bartonella-positive fleas. Our results also demonstrated these phenomena (Table 1). However, it seems that these events are less likely to occur than in cases in which the cat and its fleas host the same Bartonella species. In bacteremic cats, a larger percentage (66.7%) of the fleas collected contained the same Bartonella species as their hosts than that of fleas that harbored a different Bartonella species (20.0%) or did not harbor bartonellae (13.3%). The nonmatched cases could be explained by a previous encounter of those fleas with another positive host or by a hidden coinfection that was below the assay's detection limit in either the cat or the flea. Coinfection was detected in only two fleas, which were harboring two Bartonella species, including the Bartonella species that was present in their hosts. Nevertheless, we are aware that the methodology applied in this study is biased toward the detection of the dominant Bartonella species; thus, other coinfections could have been overlooked. The negative fleas hosted by bacteremic cats could be the result of a recent feline acquisition of noninfected fleas or newly emerged fleas that had not become infected at the time of molecular screening or had not reached detectable levels of bartonellae. On the other hand, almost all of the nonbacteremic cats (11/13) were carrying Bartonella-positive fleas. Interestingly, these fleas presented the lower bacterial loads in this study, further strengthening our above-mentioned presumptions about the Bartonella acquisition/accumulation phenomenon, which could not take place in these fleas. The apparent dissimilar scenarios of the Bartonella infection of hosts compared to their vectors can be further explained by unexplored events such as the recent clearance of a previous Bartonella species from the cats and/or fleas, nonbacteremic periods of infected cats due to the cyclic pattern of Bartonella bacteremia, and potential competition between different Bartonella species in the cats and/or fleas.

Four distinct Bartonella species were detected in the fleas sampled in this study, including the three acknowledged feline-associated Bartonella species (B. henselae, B. clarridgeiae, and B. koehlerae) and one rodent-associated Bartonella species (a B. elizabethae-like isolate), while only the three feline-associated Bartonella species were detected in the cats. The detection of a rodent-associated Bartonella strain closely related to Bartonella sp. strain Tel Aviv Rr, B. elizabethae, and B. tribocorum in the C. felis fleas studied supports the notion that these arthropods are in close contact with other mammal hosts and not restricted to one host species only, even though the mobility of the fleas between hosts has been estimated to be low (36). Previous experimental studies have proven that C. felis can acquire and maintain persistent infection with non-feline-associated bartonellae, including B. tribocorum (23, 37), and evidence of C. felis (collected from dogs) carrying B. elizabethae DNA has been recently reported (38). Interestingly, the transmission of this rodent-associated Bartonella strain from fleas to cats and the potential establishment of infection in the cats seem to be limited in the population studied, as no cat was found to be infected with this strain. In a recent experimental-infection study, Chomel et al. (33) have shown that the ability to produce bacteremia in cats varies among non-feline-associated Bartonella species. In a previous study performed by our group, a different rodent Bartonella genotype was detected in a cat (24), suggesting that cats may be infected with other rodent-associated strains. Thus, the biological role of C. felis fleas in the transmission of nonfeline bartonellae to cats and other mammalian hosts (including humans) needs to be further investigated.

Cat populations can represent substantial reservoirs of bartonellae under natural conditions (39). In a cat population with close interactions, Bartonella transmission can reach high percentages, as observed in an isolated cattery with infection rates that varied from 67 to 100% over a year (11). Accordingly, the cat population studied appeared to be a focus of Bartonella infection in Israel, since the infection rate of the cats sampled (64%) was more than twice the general Bartonella prevalence in Israeli stray cats determined in a previous study (30.7%) (24). Moreover, the distribution of the Bartonella species differed from the general Israeli prevalence, since the percentage of infected cats with B. clarridgeiae (36.1%) surpassed that of cats infected with B. henselae (22.2%). Similarly, B. koehlerae was the least prevalent species detected (5.6% in both studies). The distribution of the Bartonella species in fleas mirrored this prevalence order. Interestingly, in other regions, such as the Palestinian Authority and France, Bartonella species in fleas showed a similar distribution, in which B. clarridgeiae was the most prevalent, more than B. henselae and B. koehlerae (40, 41). Furthermore, no differences were observed in the bacterial loads of cats according to gender (female versus male) or categorical age (kittens versus adults), reflecting a very homogeneous distribution of Bartonella infection within the cat population. In addition, most of the cats were infected by the immunosuppressing retrovirus FeLV (81%) or FIV (5.6%). The high percentage of retroviral infection of the cats in this study suggests that their immunological status was compromised. This might explain the higher Bartonella prevalence than in the Israeli stray cat population (42). However, no difference or association was observed between the FeLV infection status and the Bartonella loads and Bartonella infection status of the cats, respectively. These findings might have resulted from the low number of FeLV-negative cats included in this study. Thus, the cat population from the eastern suburbs of Rishon-LeZion, Israel, enabled us to add valuable information relating to the distribution of bartonellae in cats and fleas under natural conditions.

In conclusion, a quantitative relationship between the bacterial loads in cats and their fleas was found, especially when both harbored the same Bartonella species. The positive host effect on the bacterial loads of the fleas, but not the opposite, supports the notion that cats act as the source of bartonellae while fleas act as accumulators/proliferators and vehicles for dispersal. Moreover, the detection of rodent-associated bartonellae in the fleas evidences the movement of fleas between different host species. The present study suggests that the exploration and analysis of the distribution of Bartonella species within hosts and vectors need to be accomplished by quantitative and qualitative approaches. Accordingly, disregarding one approach might result in erroneous epidemiological conclusions.

Supplementary Material

Supplemental material
supp_81_16_5613__index.html (1.4KB, html)

ACKNOWLEDGMENTS

We thank Yehonathan Even-Zur, Tomer Nessimyan, and Tomer Iakooby of the Veterinary Clinic of Rishon-LeZion for their assistance in collecting blood samples and fleas and Ofer Peleg for his professional assistance.

This research was supported by a stipend from the Ministerio de Ciencia y Tecnología and the Consejo Nacional para Investigaciones Científicas y Tecnológicas, Costa Rica, to Ricardo Gutiérrez.

Footnotes

Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.01370-15.

REFERENCES

  • 1.Dehio C. 2005. Bartonella-host-cell interactions and vascular tumour formation. Nat Rev Microbiol 3:621–631. doi: 10.1038/nrmicro1209. [DOI] [PubMed] [Google Scholar]
  • 2.Chomel BB, Kasten RW. 2010. Bartonellosis, an increasingly recognized zoonosis. J Appl Microbiol 109:743–750. doi: 10.1111/j.1365-2672.2010.04679.x. [DOI] [PubMed] [Google Scholar]
  • 3.Mainardi JL, Figliolini C, Goldstein FW, Blanche P, Baret-Rigoulet M, Galezowski N, Fournier PE, Raoult D. 1998. Cat scratch disease due to Bartonella henselae serotype Marseille (Swiss cat) in a seronegative patient. J Clin Microbiol 36:2800. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kordick DL, Hilyard EJ, Hadfield TL, Wilson KH, Steigerwalt AG, Brenner DJ, Breitschwerdt EB. 1997. Bartonella clarridgeiae, a newly recognized zoonotic pathogen causing inoculation papules, fever, and lymphadenopathy (cat scratch disease). J Clin Microbiol 35:1813–1818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Margileth AM, Baehren DF. 1998. Chest-wall abscess due to cat-scratch disease (CSD) in an adult with antibodies to Bartonella clarridgeiae: case report and review of the thoracopulmonary manifestations of CSD. Clin Infect Dis 27:353–357. doi: 10.1086/514671. [DOI] [PubMed] [Google Scholar]
  • 6.Sander A, Zagrosek A, Bredt W, Schiltz E, Piemont Y, Lanz C, Dehio C. 2000. Characterization of Bartonella clarridgeiae flagellin (FlaA) and detection of antiflagellin antibodies in patients with lymphadenopathy. J Clin Microbiol 38:2943–2948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Avidor B, Graidy M, Efrat G, Leibowitz C, Shapira G, Schattner A, Zimhony O, Giladi M. 2004. Bartonella koehlerae, a new cat-associated agent of culture-negative human endocarditis. J Clin Microbiol 42:3462–3468. doi: 10.1128/JCM.42.8.3462-3468.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ohad DG, Morick D, Avidor B, Harrus S. 2010. Molecular detection of Bartonella henselae and Bartonella koehlerae from aortic valves of Boxer dogs with infective endocarditis. Vet Microbiol 141:182–185. doi: 10.1016/j.vetmic.2009.08.005. [DOI] [PubMed] [Google Scholar]
  • 9.Kabeya H, Maruyama S, Irei M, Takahashi R, Yamashita M, Mikami T. 2002. Genomic variations among Bartonella henselae isolates derived from naturally infected cats. Vet Microbiol 89:211–221. doi: 10.1016/S0378-1135(02)00175-X. [DOI] [PubMed] [Google Scholar]
  • 10.Yamamoto K, Chomel BB, Kasten RW, Hew CM, Weber DK, Lee WI, Droz S, Koehler JE. 2002. Experimental infection of domestic cats with Bartonella koehlerae and comparison of protein and DNA profiles with those of other Bartonella species infecting felines. J Clin Microbiol 40:466–474. doi: 10.1128/JCM.40.2.466-474.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Chomel BB, Kasten RW, Floyd-Hawkins K, Chi B, Yamamoto K, Roberts-Wilson J, Gurfield AN, Abbott RC, Pedersen NC, Koehler JE. 1996. Experimental transmission of Bartonella henselae by the cat flea. J Clin Microbiol 34:1952–1956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.La VD, Tran-Hung L, Aboudharam G, Raoult D, Drancourt M. 2005. Bartonella quintana in domestic cat. Emerg Infect Dis 11:1287–1289. doi: 10.3201/eid1108.050101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Breitschwerdt EB, Maggi RG, Sigmon B, Nicholson WL. 2007. Isolation of Bartonella quintana from a woman and a cat following putative bite transmission. J Clin Microbiol 45:270–272. doi: 10.1128/JCM.01451-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Chomel BB, Boulouis HJ, Breitschwerdt EB. 2004. Cat scratch disease and other zoonotic Bartonella infections. J Am Vet Med Assoc 224:1270–1279. doi: 10.2460/javma.2004.224.1270. [DOI] [PubMed] [Google Scholar]
  • 15.Varanat M, Travis A, Lee W, Maggi RG, Bissett SA, Linder KE, Breitschwerdt EB. 2009. Recurrent osteomyelitis in a cat due to infection with Bartonella vinsonii subsp. berkhoffii genotype II. J Vet Intern Med 23:1273–1277. doi: 10.1111/j.1939-1676.2009.0372.x. [DOI] [PubMed] [Google Scholar]
  • 16.Tsai YL, Chang CC, Chuang ST, Chomel BB. 2011. Bartonella species and their ectoparasites: selective host adaptation or strain selection between the vector and the mammalian host? Comp Immunol Microbiol Infect Dis 34:299–314. doi: 10.1016/j.cimid.2011.04.005. [DOI] [PubMed] [Google Scholar]
  • 17.Bergh K, Bevanger L, Hanssen I, Loseth K. 2002. Low prevalence of Bartonella henselae infections in Norwegian domestic and feral cats. APMIS 110:309–314. doi: 10.1034/j.1600-0463.2002.100405.x. [DOI] [PubMed] [Google Scholar]
  • 18.La Scola B, Davoust B, Boni M, Raoult D. 2002. Lack of correlation between Bartonella DNA detection within fleas, serological results, and results of blood culture in a Bartonella-infected stray cat population. Clin Microbiol Infect 8:345–351. doi: 10.1046/j.1469-0691.2002.00434.x. [DOI] [PubMed] [Google Scholar]
  • 19.Chomel BB, Carlos ET, Kasten RW, Yamamoto K, Chang CC, Carlos RS, Abenes MV, Pajares CM. 1999. Bartonella henselae and Bartonella clarridgeiae infection in domestic cats from The Philippines. Am J Trop Med Hyg 60:593–597. [DOI] [PubMed] [Google Scholar]
  • 20.Boulouis HJ, Chang CC, Henn JB, Kasten RW, Chomel BB. 2005. Factors associated with the rapid emergence of zoonotic Bartonella infections. Vet Res 36:383–410. doi: 10.1051/vetres:2005009. [DOI] [PubMed] [Google Scholar]
  • 21.Guptill L, Slater LN, Wu CC, Lin TL, Glickman LT, Welch DF, Tobolski J, HogenEsch H. 1998. Evidence of reproductive failure and lack of perinatal transmission of Bartonella henselae in experimentally infected cats. Vet Immunol Immunopathol 65:177–189. doi: 10.1016/S0165-2427(98)00153-6. [DOI] [PubMed] [Google Scholar]
  • 22.Abbott RC, Chomel BB, Kasten RW, Floyd-Hawkins KA, Kikuchi Y, Koehler JE, Pedersen NC. 1997. Experimental and natural infection with Bartonella henselae in domestic cats. Comp Immunol Microbiol Infect Dis 20:41–51. doi: 10.1016/S0147-9571(96)00025-2. [DOI] [PubMed] [Google Scholar]
  • 23.Bouhsira E, Ferrandez Y, Liu MF, Franc M, Boulouis HJ, Biville F. 2013. Ctenocephalides felis an in vitro potential vector for five Bartonella species. Comp Immunol Microbiol Infect Dis 36:105–111. doi: 10.1016/j.cimid.2012.10.004. [DOI] [PubMed] [Google Scholar]
  • 24.Gutiérrez R, Morick D, Gross I, Winkler R, Abdeen Z, Harrus S. 2013. Bartonellae in domestic and stray cats from Israel: comparison of bacterial cultures and high-resolution melt real-time PCR as diagnostic methods. Vector Borne Zoonotic Dis 13:857–864. doi: 10.1089/vbz.2013.1308. [DOI] [PubMed] [Google Scholar]
  • 25.Maggi RG, Breitschwerdt EB. 2005. Potential limitations of the 16S-23S rRNA intergenic region for molecular detection of Bartonella species. J Clin Microbiol 43:1171–1176. doi: 10.1128/JCM.43.3.1171-1176.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Birtles RJ, Raoult D. 1996. Comparison of partial citrate synthase gene (gltA) sequences for phylogenetic analysis of Bartonella species. Int J Syst Bacteriol 46:891–897. doi: 10.1099/00207713-46-4-891. [DOI] [PubMed] [Google Scholar]
  • 27.Norman AF, Regnery R, Jameson P, Greene C, Krause DC. 1995. Differentiation of Bartonella-like isolates at the species level by PCR-restriction fragment length polymorphism in the citrate synthase gene. J Clin Microbiol 33:1797–1803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Natonek-Wisniewska M. 2009. Species Identification of feline DNA based on analysis of cytochrome b. Ann Anim Sci 9:379–383. http://www.izoo.krakow.pl/czasopisma/annals/2009/4/6.pdf. [Google Scholar]
  • 29.Robinson MT, Morgan ER, Woods D, Shaw SE. 2010. Real-time and multiplex real-time polymerase chain reactions for the detection of Bartonella henselae within cat flea, Ctenocephalides felis, samples. Med Vet Entomol 24:449–455. doi: 10.1111/j.1365-2915.2010.00901.x. [DOI] [PubMed] [Google Scholar]
  • 30.Harrus S, Bar-Gal GK, Golan A, Elazari-Volcani R, Kosoy MY, Morick D, Avidor B, Baneth G. 2009. Isolation and genetic characterization of a Bartonella strain closely related to Bartonella tribocorum and Bartonella elizabethae in Israeli commensal rats. Am J Trop Med Hyg 81:55–58. [PubMed] [Google Scholar]
  • 31.Higgins JA, Radulovic S, Jaworski DC, Azad AF. 1996. Acquisition of the cat scratch disease agent Bartonella henselae by cat fleas (Siphonaptera:Pulicidae). J Med Entomol 33:490–495. doi: 10.1093/jmedent/33.3.490. [DOI] [PubMed] [Google Scholar]
  • 32.Rust MK, Dryden MW. 1997. The biology, ecology, and management of the cat flea. Annu Rev Entomol 42:451–473. doi: 10.1146/annurev.ento.42.1.451. [DOI] [PubMed] [Google Scholar]
  • 33.Chomel BB, Kasten RW, Stuckey MJ, Breitschwerdt EB, Maggi RG, Henn JB, Koehler JE, Chang CC. 2014. Experimental infection of cats with Afipia felis and various Bartonella species or subspecies. Vet Microbiol 172:505–510. doi: 10.1016/j.vetmic.2014.05.033. [DOI] [PubMed] [Google Scholar]
  • 34.Bergmans AM, de Jong CM, van Amerongen G, Schot CS, Schouls LM. 1997. Prevalence of Bartonella species in domestic cats in The Netherlands. J Clin Microbiol 35:2256–2261. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Lappin MR, Hawley J. 2009. Presence of Bartonella species and Rickettsia species DNA in the blood, oral cavity, skin and claw beds of cats in the United States. Vet Dermatol 20:509–514. doi: 10.1111/j.1365-3164.2009.00800.x. [DOI] [PubMed] [Google Scholar]
  • 36.Rust MK. 1994. Interhost movement of adult cat fleas (Siphonaptera: Pulicidae). J Med Entomol 31:486–489. doi: 10.1093/jmedent/31.3.486. [DOI] [PubMed] [Google Scholar]
  • 37.Kernif T, Leulmi H, Socolovschi C, Berenger JM, Lepidi H, Bitam I, Rolain JM, Raoult D, Parola P. 2014. Acquisition and excretion of Bartonella quintana by the cat flea, Ctenocephalides felis felis. Mol Ecol 23:1204–1212. doi: 10.1111/mec.12663. [DOI] [PubMed] [Google Scholar]
  • 38.Sofer S, Gutiérrez R, Morick D, Mumcuoglu KY, Harrus S. 11 April 2015. Molecular detection of zoonotic bartonellae (B. henselae, B. elizabethae and B. rochalimae) in fleas collected from dogs in Israel. Med Vet Entomol. doi: 10.1111/mve.12116. [DOI] [PubMed] [Google Scholar]
  • 39.Breitschwerdt EB, Kordick DL. 2000. Bartonella infection in animals: carriership, reservoir potential, pathogenicity, and zoonotic potential for human infection. Clin Microbiol Rev 13:428–438. doi: 10.1128/CMR.13.3.428-438.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Nasereddin A, Risheq A, Harrus S, Azmi K, Ereqat S, Baneth G, Salant H, Mumcuoglu KY, Abdeen Z. 2014. Bartonella species in fleas from Palestinian territories: prevalence and genetic diversity. J Vector Ecol 39:261–270. doi: 10.1111/jvec.12100. [DOI] [PubMed] [Google Scholar]
  • 41.Rolain JM, Franc M, Davoust B, Raoult D. 2003. Molecular detection of Bartonella quintana, B. koehlerae, B. henselae, B. clarridgeiae, Rickettsia felis, and Wolbachia pipientis in cat fleas, France. Emerg Infect Dis 9:338–342. doi: 10.3201/eid0903.020278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Harrus S, Klement E, Aroch I, Stein T, Bark H, Lavy E, Mazaki-Tovi M, Baneth G. 2002. Retrospective study of 46 cases of feline haemobartonellosis in Israel and their relationships with FeLV and FIV infections. Vet Rec 151:82–85. doi: 10.1136/vr.151.3.82. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental material
supp_81_16_5613__index.html (1.4KB, html)
AEM.01370-15_zam999116486so1.pdf (284.6KB, pdf)

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES