Background

Bartonella spp. are Gram-negative bacteria from the family Bartonellaceae with more than 23 defined species that infect domestic and wild mammals and humans [1,2,3]. Bartonella henselae, B. clarridgeiae, B. quintana, and B. bacilliformis are the most common species associated with human diseases [1, 2]. Among these, B. henselae and B. clarridgeiae cause cat-scratch disease while B. quintana causes trench fever disease. Both diseases are called bartonellosis and manifest with symptoms such as fever, bacteremia, bacillary angiomatosis and endocarditis [2, 4]. Bartonella koehlerae, B. elizabethae and B. alsatica also have been associated with sporadic cases of endocarditis in humans [5, 6].

The principal reservoir hosts for B. henselae, B. clarridgeiae and B. koehlerae are domestic cats [3, 7, 8]. Other Bartonella species, such as B. rochalimae, B. elizabethae, B. quintana and B. grahamii, have been detected in cats [9]. Cats can become infected with many Bartonella species, but they usually show no symptoms. However, uveitis and endocarditis have been associated with B. henselae infection in cats [10], and lymphadenopathy, fever and neurological signs have been reported in experimentally infected cats [11].

Various diagnostic methods are currently in use to diagnose bartonellosis and/or applied during epidemiological surveys, such as culture, PCR assay, histopathology and serology. Among these methods, PCR assays targeting Bartonella-specific gene sequences have become a very important tool for the diagnosis of Bartonella species, which are very difficult to isolate from blood or tissue samples [11, 12]. The 16S ribosomal RNA (rRNA) gene was initially used during the molecular diagnosis of Bartonella species, but was subsequently shown that it could not provide sufficient distinction in phylogenetic analysis at the species level [13]. More reliable phylogenetic results and species distinctions are obtained by analyzing the 16S–23S rDNA intergenic spacer (ITS) and gltA genes [13,14,15].

The prevalence of Bartonella in cats has been reported to vary from 4% to 70% using blood culture methods [16], and the seroprevalence of antibodies against Bartonella in cats also varies, ranging from 0 to 80%. An increased prevalence has been especially reported in warmer regions; for example, in a study conducted in California, the seroprevalence in cats was 80% compared to 0% in a study conducted in Norway [17, 18]. In studies conducted in the Middle East, including Saudi Arabia and Iraq, seroprevalence rates in cats were found to be 15% for B. henselae and 12.6% for B. clarridgeiae whereas Bartonella DNA positivity was 9.25% [19, 20]. In Iran, Bartonella DNA positivity was reported to vary from 14% to 74.2% in dogs [21, 22]. In the same region, Bartonella DNA positivity was reported to be 7.14% and 1.42% in nail and saliva samples collected from cats [23]. In Turkey, the prevalence of Bartonella was found to be 9.4% in domestic cats by blood culture methods but seroprevalence reached up to 40% [7, 8].

Since the zoonotic transmission of Bartonella occurs by a cat scratch or through the bite of a vector, the prevalence of Bartonella in stray cats that are in close contact with humans is frequently being screened in many countries [8, 9, 24,25,26]. Although the weather is very hot in Izmir, Turkey, especially during the summer, and the city is home to many stray cats, the prevalence of Bartonella and species of Bartonella have not been investigated. Therefore, the aim of this study was to investigate the molecular prevalence of Bartonella in a large number of blood samples collected from stray cats and to sequence the positive samples for species identification. In addition, genetic diversity within each detected species was investigated by haplotype analysis.

Methods

Blood samples

Blood samples (n = 1012) were collected from stray cats in Izmir city that had been brought to veterinary clinics located in the districts of Balçova (n = 110), Bayraklı (n = 43), Bornova (n = 48), Buca (n = 54), Çiğli (n = 6), Gaziemir (n = 4), Güzelbahçe (n = 10), Karabağlar (n = 115), Karşıyaka (n = 4), Konak (n = 614) and Narlıdere (n = 4). These stray cats were captured by persons in an animal friendly manner, without any harm being inflicted on the animal, and brought to the clinics for sterilization. The probability sampling method was used for sampling.

Conventional PCR

DNA was isolated from the blood samples collected from the stray cats using a commercial kit (Qiagen DNA Extraction Kit; Qiagen, Hilden, Germany) in accordance with the manufacturer’s instructions. The 16S-23S rRNA ITS region in the extracted DNA samples was targeted for the diagnosis of Bartonella species [27]. During PCR analysis, a 489-bp fragment was amplified using the primer pairs 325s (5-CTTCAGATGATGATCCCAAGCCTTCTGGCG-3) and 1100as (5-GAACCGACGACCCCCTGCTTGCAAAGCA-3)(Eurofins Genomics Germany GmbH, Ebersberg, Germany)[27]. The amplification reaction mixture (30 µl) consisted of 5 µl template DNA, 1 µl of each primer (10 µM), 12.5 µl 2× PCR master mix (GeneMark, Taichung, Taiwan) and 10.5 µl nuclease-free water. PCR cycling program consisted of an initial denaturation of 2 min at 95 °C, followed by 35 cycles at 94 °C for 15 s, 66 °C for 15 s, and 72 °C for 15 s, with a final elongation at 72 °C for 1 min.

Species identification

For species identification of Bartonella PCR-positive samples, sequences obtained by Sanger sequencing (Eurofins Genomics Germany GmbH) were aligned with MEGA 7.0 software and subject to BLAST analysis against the GenBank database. In addition, the obtained results also were confirmed by phylogenetic analysis performed by maximum likelihood method using the Kimura 2-parameter gamma distribution (K2 + G) model with 1000 bootstrap replications [28]. Anaplasma phagocytophilum was used as an outgroup. For sequences with identical nucleotides (100% identity), only one was used for phylogenetic analysis. The reference 16S-23S rRNA ITS sequences used in this study are given in Additional file 1: Table S1.

Haplotype analysis

Haplotype analysis was performed using the DNASP program [29] using Bartonella isolates detected in this study and reference B. henselae, B. clarridgeiae and B. koehlerae strains isolated from cats in different countries. A haplotype network was generated in PopArt using the TCS network [30, 31]. For Bartonella species detected in the present study, the number of variable sites (VS), C + G content (GC%), number of haplotypes (h), haplotype diversity (Hd), nucleotide diversity (π), number of nucleotide differences (K) and standard deviation (SD) were calculated using the DNASP program. Sequences belonging to B. henselae, B. clarridgeiae and B. koehlerae from cats were retrieved in GenBank and used in the haplotype analysis. These included 24 B. henselae sequences from 10 countries (Spain, Malta, Brazil, Paraguay, Taiwan, Oklahoma, Guatemala, Korea, Australia and Malaysia), 40 B. clarridgeiae sequences from 13 countries (Spain, Malta, Portugal, Philippines, Brazil, Paraguay, China, USA, Taiwan, Indonesia, Japan, Greece and Iran), and five B. koehlerae sequences from two countries (Brazil and Malta).

Statistical analysis

Bartonella positivity values detected in stray cats in Izmir were computed with the exact binomial confidence intervals of 95% (95% CI), and comparison of the proportions was performed by the Chi-square test using PASW Statistics version 18 software. Statistically significant differences were determined at P < 0.05.

Results

Molecular prevalence of Bartonella spp.

DNA extracted from 1012 blood samples collected from stray cats was screened by PCR; of these 1012 DNA samples, 12.1% (122/1012; 95% CI 10.1–14.2%) were positive for Bartonella DNA. The highest prevalence was detected in samples collected from cats in Güzelbahçe (20%), followed by Bayraklı (14%) (Fig. 1). There was no statistically significant difference in detected Bartonella positivity values between the districts of Izmir sampled (Chi-square test, χ2 = 0.003, df = 1, P = 0.955).

Fig. 1
figure 1

Map showing the prevalence of Bartonella PCR-positivity detected in sampled cats according to district of Izmir

Full size image

Species identification

Among the Bartonella-positive samples, 100 (100/122; 81.96%) were successfully sequenced. BLAST and phylogenetic analyses revealed the presence of B. henselae, B. clarridgeiae and B. koehlerae in the positive samples. Bartonella henselae was the most common species detected in the stray cats (45%; 45/100) detected, followed by B. clarridgeiae (29%; 29/100) and then by B. koehlerae (26%; 26/100) (Fig. 2).

Fig. 2
figure 2

Map showing the distribution of Bartonella species detected in sampled cats according to district of Izmir

Full size image

Phylogenetic analysis and haplotype diversity

All Bartonella species detected in this study clustered with reference sequences, forming well-defined groups separated by moderate and high bootstrap values (Fig. 3).

Fig. 3
figure 3

Phylogenetic tree shows the relationship of Bartonella species detected with reference Bartonella isolates. Phylogenetic analysis was performed by maximum likelihood method using the Kimura 2-parameter gamma distribution (K2 + G) model with 1000 bootstrap replications. Anaplasma phagocytophilum was used as an outgroup. Only bootstrap values > 50 are shown. Reference 16S-23S ribosomal RNA internal transcribed spacer sequences used in this study are given in Additional file 1: Table S1

Full size image

Bartonella clarridgeiae isolates (n = 69) belonged to 19 haplotypes (H-1 to H-19). Among these haplotypes, the most prevalent haplotype was H-1, which contained 47 B. clarridgeiae isolates from 14 countries, including Turkey (Fig. 4). The B. clarridgeiae sequences generated in this study belonged to different haplotypes only from Turkey (Fig. 4). Similarly, some B. clarridgeiae sequences from Paraguay and Spain also belonged to different haplotypes (Fig. 4). Haplotype analysis performed for B. henselae sequences (n = 59) belonged to 14 haplotypes (H-1 to H-14). Among these haplotypes, the most prevalent haplotype was H-1, which included 40 B. henselae sequences from 10 countries, including Turkey (Fig. 5). In addition to H-1, there were two haplotypes (H-2 and H-3) containing sequences from different countries (Fig. 5). Also, B. henselae detected in this study belonged to different haplotypes only from Turkey (Fig. 5). All B. koehlerae sequences (n = 31) belonged to eight haplotypes (H-1 to H-8). Among these haplotypes, the most prevalent haplotype was H-1, which contained 21 B. koehlerae sequences from three countries, including Turkey (Fig. 6). The B. koehlerae sequences detected in this study belonged to different haplotypes containing only isolates from Turkey (Fig. 6). The VS, GC%, h, Hd, π, K and SD for each Bartonella species detected in this study are presented in Table 1.

Fig. 4
figure 4

Haplotype analysis conducted for Bartonella clarridgeiae isolates. The haplotype network was generated in PopArt using the TCS network. According to this analysis, haplotype I containing 47 B. clarridgeiae isolated from 14 different countries, including Spain, Malta, Portugal, Philippines, Brazil, Paraguay, China, USA, Taiwan, Indonesia, Japan, Greece, Iran and Turkey, is the most prevalent haplotype. Haplotype I also represents the similar Bartonella isolates that are frequently detected in these countries. Each remaining haplotype represents unique Bartonella isolates to any country. Each color represents a country, as shown in Fig. 4

Full size image
Fig. 5
figure 5

Haplotype analysis conducted for Bartonella henselae isolates. The haplotype network was generated in PopArt using the TCS network. According to this analysis, haplotype I containing 40 B. henselae isolated from 10 different countries, including Spain, Malta, Brazil, Paraguay, Taiwan, Guatemala, Korea, Australia, Malaysia and Turkey, is the most prevalent haplotype. Haplotype II and III containing more than one Bartonella isolate from different countries are among the prevalent haplotypes. Each remaining haplotype represents unique Bartonella isolates to any country. Each color represents a country, as shown in Fig. 5

Full size image
Fig. 6
figure 6

Haplotype analysis conducted for Bartonella koehlerae isolates. The haplotype network was generated in PopArt using the TCS network. According to this analysis, haplotype I containing 21 B. koehlerae isolated from three different countries, including Brazil, Malta and Turkey, is the most prevalent haplotype. This haplotype also represents the similar Bartonella isolates that are frequently detected in these countries. Each remaining haplotype represents unique Bartonella isolates to any country. Each color represents a country, as shown in Fig. 6

Full size image
Table 1 Genetic diversity among Bartonella species detected in this study
Full size table

Discussion

In the present study we investigated the prevalence of Bartonella spp. in stray cats and identified the species of Bartonella present in the DNA collected from Bartonella-positive samples by sequencing. A haplotype analysis was also performed to reveal the genetic diversity of each Bartonella species detected. Bartonella DNA was detected in 12.1% of the samples collected from the stray cats. This prevalence is comparable with that reported in previous studies conducted in Turkey. In a study analyzing 256 samples from cats in Ankara, Turkey, Bartonella was detected in 9.4% of samples by blood culture [7] while the seroprevalence of B. henselae in the cats was 18.6%. Higher Bartonella prevalence values in cats also were reported in different studies using molecular or serological methods. Accordingly, the seroprevalence of B. henselae in cats was determined to be 41.3, 33.9, 27.5, 32.3, 17.9 and 12.5% in Bursa, Adana, Aydın, Burdur, Kayseri and Istanbul, respectively [8]. A study conducted in Tekirdağ reported a prevalence of 40.1% for B. henselae based on an analysis of samples collected from 167 client-owned symptomatic cats using PCR [32]. All of these results, obtained by blood culture, molecular or serological methods, indicate that Bartonella is prevalent in cats living in different locations of Turkey.

Bartonella henselae, B. clarridgeiae and B. koehlerae were the species detected in stray cats in this study. Among the Bartonella-positive samples, B. henselae was found to be the predominant species (prevalence: 45%) together with B. clarridgeiae and B. koehlerae. While less frequent than B. henselae, B. clarridgeiae is accepted as a causative agent for cat scratch disease [9] and B. koehlerae has been linked to endocarditis in humans [5].

The highest haplotype diversity was detected among B. clarridgeiae sequences. Bartonella clarridgeiae H-1 has been detected in Spain [33], Brazil [34], USA [35], Malta [36], Greece [37], Portugal [38], Philippines [39], Paraguay [40], Japan [41], Taiwan [39] and Indonesia [39]. Also, B. henselae H-1 has been reported in Spain [33], Malta [36], Brazil [42], Paraguay [40], Taiwan [33], Guatemala [43], Korea [44] and Australia [45]. Bartonella henselae H-2 was detected in Spain [33] and Brazil [42] whereas B. henselae H-3 has been reported in Spain [33]. Finally, B. koehlerae H-1 has been detected in Brazil [42] and Malta [36]. Within each Bartonella species, there were haplotypes that are apparently unique to Turkey in addition to haplotypes from different countries including Turkey (Figs. 4–6). Nonetheless, most of the Bartonella sequences obtained in this study belong to haplotypes that have also detected in cats in different countries.

Since previous studies carried out in Turkey reported anti-B. henselae antibodies in different human groups such as adult and pediatric patients [46], healthy blood donors [47], cattle breeders and veterinarians [48] and kidney transplant patients [49], stray cats could be an important source for transmission of Bartonella infection to humans in this country.

Conclusion

In conclusion, we detected a 12.1% prevalence of Bartonella spp. infection in stray cats in Turkey, with B. henselae, B. clarridgeiae and B. koehlerae being the species detected.