Abstract
High prevalence of Rickettsia felis in patients with fever of unknown origin was revealed in the north-central Bangladesh from 2012 to 2013. Subsequently, in this study, prevalence of R. felis in cats and cat fleas (Ctenocephalides felis), together with febrile patients, was studied by PCR detection of 17 kDa antigen gene and DNA sequencing. R. felis was detected in 28% (28/100) and 21% (14/68) of cat blood and cat flea samples, respectively, whereas 42% (21/50) of patients were positive for R. felis. R. felis-positive cat fleas were detected at significantly higher rate on R. felis-positive cats. The results suggested a potential role of cats and cat fleas for transmission of R. felis to humans in Bangladesh.
Introduction
As a first report in Bangladesh, we previously identified R. felis in blood samples from sporadic patients with fever of unknown origin at high rate (46%) by PCR followed by sequencing of 17 kDa antigen gene, gltA, 16S rRNA gene, ompA, and ompB genes (Ferdouse et al. 2015). The high prevalence among patients suggested endemic distribution of R. felis in Bangladesh, which was presumed to be associated with large number of stray cats and dogs carrying cat fleas. However, there have been no report for vectors and reservoirs for R. felis in Bangladesh. In this study, we investigated the prevalence of R. felis in cat fleas and cats to understand their role in rickettsial transmission to humans in this country.
Materials and Methods
This cross-sectional study was conducted from November 2013 to June 2014, and was approved by the ethical review committee of Mymensingh Medical College. Blood samples as well as cat fleas were collected from stray cats in Mymensingh city and neighboring rural areas located in central-northern Bangladesh. Venous blood samples were also collected from febrile patients with suspected rickettsial fever (fever of unknown origin), which showed positive result of Weil–Felix test in the Mymensingh Medical College Hospital as described previously (Ferdouse et al. 2015). In brief, patients with fever (axillary temperature >37.5°C) persisting for more than 15 days, not responding to commonly used antibiotics, associated with any additional clinical features (headache, rash, myalgia, etc.) were taken into account for this study. For detection of Rickettsia, nested PCR targeting 17 kDa antigen gene was performed as described previously (Schriefer et al. 1994) with utmost attention to avoid contamination. DNA extraction and PCR were done in separate laboratories and on different benches on different days depending on specimens from human, cat, or cat fleas. Double-distilled water was used as a negative control during any PCR amplification. Nucleotide sequence of PCR product was directly determined by the Sanger method with the use of automated sequencer, and their sequence identity by BLAST search for identification of rickettsial species was analyzed. Five representative sequences were deposited to GenBank under accession nos. KT390150–KT390154.
Results and Discussion
Blood samples were collected from a total of 100 stray cats, from which 68 flea pools were obtained. Human blood samples were collected from 50 suspect cases of rickettsial illness, who had prolonged unknown fever associated with positive results of Weil–Felix test, in a period from January to June 2014. By PCR, Rickettsia was detected in 28% (28/100) of cat blood samples, 20.6% (14/68) of cat fleas, and 42% (21/50) in human blood samples. Sequences of the partial 17 kDa antigen genes from cat, cat flea, and human samples showed 98–100% identity to that of R. felis URRWXCa12 (GenBank acc. no. CP000053). Cat fleas were obtained from Rickettsia-positive cats at higher rate (85.7%) than Rickettsia-negative cats (61.1%), and the rate of Rickettsia-positive fleas was significantly higher in Rickettsia-positive cats (37.5%) than in Rickettsia-negative cats (11.4%) (Table 1).
p = 0.033.
p = 0.026.
In this study, R. felis was detected in human cases at almost the same rate as our previous study during 2012–2013 (46%). The high detection rate of R. felis in this and previous studies in Bangladesh might have arisen from appropriate selection of suspect cases with rickettsial illness, and putative high endemicity of R. felis in the study area. The detection rate in cat fleas was similar to those in recent reports from Southeast Asia (e.g., 21.4% in Taiwan and 32.2% in Malaysia) (Tsai et al. 2011, Tay et al. 2014), although it varies considerably depending on reports. Seropositive rates of R. felis in cats were described as 5.2–7.5% in Spain (Segura et al. 2014) and 21% in China (Zhang et al. 2014). However, detection of R. felis DNA in cat blood by PCR has not yet been reported to date, except for experimentally infected cats showing positive rate of 31% (5/16) (Wedincamp and Foil 2000). Detection of rickettsial gene in cats at a rate of 28% in this study is considered to be achieved by a sensitive method, that is, nested PCR targeting at 17 kDa antigen gene. It is also conceivable that this Rickettsia is highly prevalent among cats in Bangladesh, which would be confirmed by a serological study.
Partial 17 kDa antigen genes determined in this study were discriminated into two sequences with five nucleotide difference, four of which cause amino acid difference (Fig. 1). One sequence was identical to that of R. felis URRWXCa12 as well as MMC-Rick7 (GenBank acc. no. KP318088) detected in Bangladesh in our previous study (Ferdouse et al. 2015), the other sequence was identical to that described as Rickettsia sp. Rf31 (GenBank acc. no. AY953285) (Parola et al. 2003). Although only URRWXCa12 sequence was detected from human blood samples, both sequences were found in cat blood samples and cat flea pools, with each sample showing either sequence. In only a cat blood sample BGCa-Rf28, at four among the five positions of the nucleotide difference, both nucleotides from the two sequences were detected (Fig. 1), suggesting mixed infection of R. felis and Rickettsia sp. Rf31. The Rickettsia sp. Rf31 is R. felis-like but unnamed rickettsial species, and has been reported in cat fleas from Thailand, Malaysia, and the United States (South Carolina) (Parola et al. 2003, Reeves et al. 2005, Tay et al. 2014). This study first revealed the presence of Rickettsia sp. Rf31 in Bangladesh and also in cat blood, suggesting its wide distribution and potential to cause infection in animals. Further genetic and serological characterizations on Rickettsia sp. Rf31 may be required to understand actual state of its distribution and to distinguish it from R. felis.

Sequence alignment of partial 17 kDa antigen genes of Rickettsia from human, cat, and flea. Dots indicate nucleotides identical to those of Rickettsia felis strain URRWXCa21. Codons of URRWXCal2 in which different amino acids are coded in some other Rickettsia are underlined with amino acids shown in parentheses below alignment (amino acids of URRWXCa12 are underlined). Nucleotide numbers of URRWXCa21 genome (GenBank accession no. CP000053) are indicated for the first and final nucleotides. GenBank accession numbers of Rf31 and MMC-Rick7 are AY953285 and KP318088, respectively. R, A, or G; Y, C, or T; M, A, or C.
In conclusion, prevalence of R. felis in cats and cat fleas was first demonstrated in Bangladesh, suggesting a potential role of cats and cat fleas for transmission of R. felis to humans. Further studies in cats, other mammals such as dogs, and various arthropods are necessary to clarify ecological nature and transmission route of R. felis.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
