Multiplex SYBR® green-real time PCR (qPCR) assay for the detection and differentiation of Bartonella henselae and Bartonella clarridgeiae in cats.

A novel SYBR® green-real time polymerase chain reaction (qPCR) was developed to detect two Bartonella species, B. henselae and B. clarridgeiae, directly from blood samples. The test was used in blood samples obtained from cats living in animal shelters in Southern Brazil. Results were compared with those obtained by conventional PCR targeting Bartonella spp. Among the 47 samples analyzed, eight were positive using the conventional PCR and 12 were positive using qPCR. Importantly, the new qPCR detected the presence of both B. henselae and B. clarridgeiae in two samples. The results show that the qPCR described here may be a reliable tool for the screening and differentiation of two important Bartonella species.

INTRODUCTION

Bartonella spp. are emerging infectious zoonotic agents[13,23] endemic in some South American countries[19]. Bartonella can infect humans as well as domestic and wild mammals[2]. There are more than 23 described species within this genus[6,8], among which, 13 are related to human diseases[15]. The most commonly reported species are B. bacilliformis, B. quintana, B. henselae [1] and B. clarridgeiae [18].

The main vectors associated with human infections are fleas (Ctenocephalides felis), lice (Pediculus humanus), sandflies (Lutzomyia verrucarum, Lutzomyia peruensis)[15] and ticks (Ixodes pacificus)[10]. The main animal reservoirs are cats, dogs and rodents[15]. The infection is generally more severe in immunocompromised hosts, but may also occur in healthy subjects[15]. Bartonella bacilliformis is an important cause of severe illness and death among immunocompetent adults and children[19].

B. henselae and B. clarridgeiae are known as the causative agents of cat-scratch disease (CSD), which is characterized by chronic lymphadenopathy[11]. Furthermore, in some cases, these bacteria may result in Parinaud oculoglandular syndrome, encephalopathy, convulsions, endocarditis, hepatosplenomegaly, glomerulonephritis, pleurisy, mediastinal adenopathy, nodules in the head of the pancreas, bacillary angiomatosis, osteomyelitis, atypical pneumonia, mammary tumours, haemolyticanaemia and eosinophilic purpura[7].

More than 4,000 cases of CSD are registered each year in the United States, resulting in over 2,000 hospitalizations in the same period, with a rate of 0.77 to 0.88 / 100,000 hospitalizations, with 55% of cases occurring in patients younger than 18 years old[14]. In the Netherlands, an estimated 2,000 cases occur per year with a rate of 12.5 cases/100 thousand inhabitants[3]. In Asian countries, reports show that the seroprevalence of B. henselae among the cat population ranges from 9.1% to 15.1% in Japan, 68% in the Philippines, 48% in Singapore and 54% in Indonesia[7].

Bartonella spp are difficult-to-culture bacteria, hampering laboratory diagnosis of these infections. Other diagnostic alternatives, such as Giemsa-stained blood smears[4] and serology[7], are available with limited sensitivity, which may further delay proper diagnosis.

Giemsa stained blood smears can be useful in screening blood samples, however, the presence of artifacts might be a confusing factor[24]. Serology has only a retrospective value, suggesting that the patient might have been infected during some period of their life, and hence is also of limited value. To date, no gold standard method for the detection of Bartonella species is available[9].

Molecular techniques are powerful tools that may be used to screen for the presence of these pathogens in clinical samples[16]. Conventional polymerase chain reaction (PCR) has been used for the detection of several Bartonella species, and most of those assays use gene targets located on the 16S rRNA gene[5], ribCgene[17], rpo B gene[22], 16S-23S intergenic spacer region (ITS)[20] and gltA gene[21]. However, real-time PCR is substituting conventional PCR in many laboratories as it is not only quicker but also more sensitive and specific than its predecessor[12].

The aim of this study was to evaluate a SYBR® green qPCR assay to detect and identify the presence of two Bartonella species directly from feline blood samples.

METHODS

Blood samples from 47 cats were collected during 2009 in two municipal animal shelters (Novo Hamburgo and São Leopoldo, south of Brazil), and have been described previously[24]. All procedures were performed under veterinary supervision and approved by the Animal Ethics Committee from Universidade Feevale, under the protocol number 2.12.03.09.1391.

DNA was extracted from blood samples with the QIAmpDNA blood mini Kit (Qiagen). Two new primer pairs were especially designed from consensus genome regions obtained from the DNA Data Bank of Japan (DDBJ). Primers targeting fragments of the Citrate synthase gene were: one generic forward primer (BART-LC-GEN-F: 5′ – ATGGGTTTTGGTCATCGAGT – 3′); one specie-specific reverse B. henselae primer (BART-LC-HEN-R: 5′ –AA ATCGACATTAGGGTAAAGTTTTT – 3′); and one specie-specific reverse B. clarridgeiae primer (BART-LC-CLA-R: 5′-CAAGAAGTGGATCATCTTGG – 3′). Specificity of the assay was assessed by testing known positive samples (B. henselae, B. clarridgeiae and B. bacilliformis, confirmed by DNA sequencing[24]), and several other clinical and ATCC-derived bacterial species. No false-positive reaction was noted with these tests. Apart from that, the specificity of the assay was also confirmed by in silico analysis of the primers, which demonstrated no cross-reaction with other Bartonella species. All positive results were confirmed by DNA sequencing reaction to make sure that these are true positive results and no false-positive samples were present. The results were analyzed by the DNA Data Bank of Japan (DDBJ Blast).

The reaction mix contained 2 µL of extracted DNA, 10 pM each primer, 1µL of BSA, 10 µL of SYBR green qPCR Supermix (Invitrogen) for a final volume of 20 µL. The multiplex qPCR was performed using the LightCycler®1.5 Real-Time PCR System (Roche Diagnostics) under the following conditions: 96 °C for two min (DNA denaturation), followed by 40 cycles of 96 °C for two sec, 60 °C for five sec and 72 °C for eight sec. A melting curve analysis was performed at the end of the amplification cycles. Positive reactions were recognized by typical melting temperatures (Tm ∼ 82 °C). Since both species showed the same Tm, identification of the species among positive samples was performed by a second qPCR using species-specific primers (Singleplex) under the same conditions.

RESULTS

The multiplex SYBR® green qPCR allowed the detection of Bartonella DNA in 25.5% (12/47) of the blood samples. In a previous study, conventional PCR analysis targeting Bartonella spp in the same blood specimens resulted in a detection rate of 17.02% (8/47)[24]. DNA sequencing analysis of the PCR products obtained by conventional PCR revealed the presence of only two Bartonella species among those samples, which consisted of B. henselae, present in 10.63% (5/47) of the samples, and B. clarridgeiae in 6.38% (3/47)[24]. When using the multiplex qPCR, additional positive samples were observed with B. henselae and B. clarridgeiae detected in 17.02% (8/47) and 12.76% (6/47) of the samples, respectively (Table 1). Moreover, this assay allowed us to observe these species co-infecting two samples, this was not previously detected using conventional PCR.

Table 1

Results by qPCR and Conventional PCR for the diagnosis of Bartonella infection

Samples	qPCR	PCR
1	B. c.	B. c.
2	B. h.	B. h.
3	B. c.	Neg.
4	B. h.	Neg.
5	B. h.	Neg.
6	B. h. / B. c.	Neg.
7	B. h.	B. h.
8	B. c.	B. c.
9	B. c.	B. c.
10	B. h.	B. h.
11	B. h. / B. c.	B. h.
12	B. h.	B. h.
13-47	Neg. by both methods

B.h. (Bartonella henselae), B.c. (Bartonella clarridgeiae), Neg. (Negative).

DISCUSSION

Using multiplex qPCR, an additional four positive samples were found, which may represent an increase in sensitivity in comparison with the conventional PCR method. Moreover, the whole technique, including DNA extraction, could be performed within 2-3 hours, without any need of subsequent DNA sequencing for species identification. Conventional PCR, targeting consensus regions of the Citrate synthase gene may not amplify all Bartonella species with the same efficiency, thus explaining the differences in sensitivities obtained by our qPCR assay. Moreover, the identification of the species in positive samples depends on subsequent DNA sequencing. The advantage of the conventional PCR followed by DNA sequencing analysis would be the opportunity to identify the other or newer species that might be present in the blood samples. In this case, in a direct comparison, using the new specific primers for the detection of the two species identified by DNA sequencing, the qPCR assay was able to detect 100% of these species.

An advantage for the qPCR is the possibility of identifying the most prevalent Bartonella species in cat's blood samples without the need of subsequent DNA sequencing, which is laborious and, in our case, was not able to detect the presence of co-infection in two samples.

In conclusion, our results suggest that the multiplex SYBR® green qPCR may be a useful technique to detect and differentiate the two most common species of Bartonella directly from blood samples. With the lack of a defined gold standard method to detect the presence of different Bartonella species, there is a need for improved, clinically useful methods[9] for diagnosis of these infections, and the method developed here may be a useful diagnostic tool. Furthermore, compared with conventional PCR, the multiplex SYBR® green qPCR seems to be more sensitive; however, this hypothesis needs to be further investigated using a more robust number of specimens. The real time PCR described here is faster than conventional PCR and can be adapted to detect othe Bartonella species if necessary. The species found here are those associated with CSD, which, therefore, may pose some risk to public health, mainly for HIV-positive individuals.