Bartonella spp. and Yersinia pestis Reservoirs, Cusco, Peru

To the Editor: Bartonella spp. are gram-negative alphaproteobacteria that are transmitted between the reservoir and mammal host by hematophagous insects (1). The genus Yersinia comprises 11 species, of which Y. pestis is the causative agent of plague, a deadly rodent-associated, fleaborne zoonosis (2). Despite the large number of plague cases reported in humans and the large amount of data about human-infecting Bartonella spp. in Peru (3), no data have been published about which rodent species are reservoirs of these pathogens in this country. 
 
La Convencion Province, where cases of bartonellosis occurred during 1998 (4), is located in the northeastern part of Cusco, Peru. Although, to our knowledge, no human cases of plague have been reported in this province, a plague outbreak was recently detected in Junin Province (M.A. Quispe-Riclade, pers. comm.), which is located northwest of La Convencion Province. 
 
A total of 28 rodents were captured during 2010–2011 in 3 villages (Alto Ivochote, Aguas Calientes, and Yomentoni) in the Echarate District, La Convencion Province. Traps had been set in intradomiciliary, peridomiciliary, and extradomiciliary settings. Spleens of animals were obtained, and DNA was isolated by using the Illustra Tissue and Cells Genomic Prep Mini Spin Kit (GE Healthcare, Little Chalfont, UK). 
 
Rodents were examined for Bartonella spp. DNA by using a PCR and primers CS443f and CS1210r specific for a 767-bp fragment of the citrate synthase gene (5). Screening for plague was performed by using PCR primers Yp1 and Yp2 specific for a plasminogen activator protein (pla) encoded by the Y. pestis–specific pPLA plasmid (6). 
 
New and published Bartonella spp. and Y. pestis sequences were obtained from GenBank and compared by using the nucleotide-nucleotide basic local sequence alignment tool (BLAST) (blastn) program (www.ncbi.nlm.nih.gov/Class/MLACourse/Modules/BLAST/nucleotide_blast.html). The χ2 test was used to determine statistical differences in the prevalence of both pathogens among host species and villages. 
 
Overall prevalences for Y. pestis and Bartonella spp. were 17.9% and 21.4%, respectively (Table). Co-infections with both bacteria were found in 3 (10.7%) rodents: 2 Hylaeamys perenensis rodents and 1 Oecomys spp. rodent. Bartonella prevalence was higher in H. perenensis rats than in Rattus rattus rodents (p<0.001). Rodents positive for Bartonella spp. were found in the 3 study villages, and prevalence for Aguas Calientes was higher than that for Alto Ivochote (p<0.001). One of 8 rodents trapped inside houses and 1 of 2 rodents trapped at peridomestic sites were positive for Bartonella spp. 
 
 
 
Table 
 
Prevalence of Bartonella spp. and Yersinia pestis in rodents from Echarate District, Cusco, Peru 
 
 
 
Sequence analysis identified 3 citrate synthase gene sequences (GenBank accession nos. {"type":"entrez-nucleotide-range","attrs":{"text":"KF021602-KF021604","start_term":"KF021602","end_term":"KF021604","start_term_id":"512134467","end_term_id":"512134471"}}KF021602-KF021604) that had 98% and 99% sequence similarity to genotypic variant A3 of the undescribed Bartonella genogroup A, which was obtained from Oryzomis palustris rats in the southeastern United States (7). One genotype (isolate B259) was identified in H. perenensis rats, and 2 other genotypes were identified in 1 H. perenensis rodent and 1 Oecomys spp. rodent (isolates B273 and B280, respectively).We propose that the genotype of isolates B273 and B280 is variant A6 and the genotype of isolate B259 is variant A7. A previous study reported that the A, B, and C genogroups contain independent species (8). 
 
The pla amplicons (GenBank accession nos. {"type":"entrez-nucleotide-range","attrs":{"text":"KF214264-KF214266","start_term":"KF214264","end_term":"KF214266","start_term_id":"631512920","end_term_id":"631512924"}}KF214264-KF214266) had 98% sequence identity with Y. pestis reference sequences. Plague prevalence was higher in H. perenensis rats than in R. rattus rats (p<0.05). Infected rodents were found in all villages studied except Yomentoni, and prevalence in Aguas Calientes was higher than in Alto Ivochote (p<0.01). Two (25%) of 8 rodents trapped inside houses were infected with Y. pestis. 
 
This study suggests that infections of rodents with Bartonella spp. and Y. pestis are common and widespread throughout the Echarate District. It also shows the role of H. perenensis and Oecomys spp. rodents as reservoirs of both pathogens. This role was confirmed by amplifying the chromosomal ferric iron uptake regulation gene by using PCR primers Ypfur1 and Ypfur2, as described by Hinnebusch et al. (9). The epidemiologic role of rodent-borne Bartonella spp. as a cause of disease in humans is emerging in the Americas. This role has been suggested by identification of a novel rodent-associated Bartonella strain causing febrile illness in the rural southwestern United States (10) and a strain of pathogenic B. elizabethae, a bacteria that can cause human endocarditis, in the Huayllacallan Valley in Peru (3). 
 
Because most identified Bartonella spp. have been reported as infectious agents for humans, our results should prompt public health concern. However, our findings require further investigation about the pathogenicity of these Bartonella genotypes. The detection of both pathogens in intradomestic and peridomestic areas where humans are in close contact with rodents could indicate that the incidence of both diseases in humans from Echarate District might be underestimated.

Rodents were examined for Bartonella spp.DNA by using a PCR and primers CS443f and CS1210r specific for a 767-bp fragment of the citrate synthase gene (5).Screening for plague was performed by using PCR primers Yp1 and Yp2 specific for a plasminogen activator protein (pla) encoded by the Y. pestis-specific pPLA plasmid (6).
New and published Bartonella spp.and Y. pestis sequences were obtained from GenBank and compared by using the nucleotide-nucleotide basic local sequence alignment tool (BLAST) (blastn) program (www.ncbi.nlm.nih.gov/Class/MLACourse/Modules/BLAST/nucleotide_blast.html).The χ 2 test was used to determine statistical differences in the prevalence of both pathogens among host species and villages.
Overall prevalences for Y. pestis and Bartonella spp.were 17.9% and 21.4%, respectively (Table ).Co-infections with both bacteria were found in 3 (10.7%)rodents: 2 Hylaeamys perenensis rodents and 1 Oecomys spp.rodent.Bartonella prevalence was higher in H. perenensis rats than in Rattus rattus rodents (p<0.001).Rodents positive for Bartonella spp.were found in the 3 study villages, and prevalence for Aguas Calientes was higher than that for Alto Ivochote (p<0.001).One of 8 rodents trapped inside houses and 1 of 2 rodents trapped at peridomestic sites were positive for Bartonella spp.
Sequence analysis identified 3 citrate synthase gene sequences (GenBank accession nos.KF021602-KF021604) that had 98% and 99% sequence similarity to genotypic variant A3 of the undescribed Bartonella genogroup A, which was obtained from Oryzomis palustris rats in the southeastern United States (7).One genotype (isolate B259) was identified in H. perenensis rats, and 2 other genotypes were identified in 1 H. perenensis rodent and 1 Oecomys spp.rodent (isolates B273 and B280, respectively).We propose that the genotype of isolates B273 and B280 is variant A6 and the genotype of isolate B259 is variant A7.A previous study reported that the A, B, and C genogroups contain independent species (8).
The pla amplicons (GenBank accession nos.KF214264-KF214266) had 98% sequence identity with Y. pestis reference sequences.Plague prevalence was higher in H. perenensis rats than in R. rattus rats (p<0.05).Infected rodents were found in all villages studied except Yomentoni, and prevalence in Aguas Calientes was higher than in Alto Ivochote (p<0.01).Two (25%) of 8 rodents trapped inside houses were infected with Y. pestis.
This study suggests that infections of rodents with Bartonella spp.and Y. pestis are common and widespread throughout the Echarate District.It also shows the role of H. perenensis and Oecomys spp.rodents as reservoirs of both pathogens.This role was confirmed by amplifying the chromosomal ferric iron uptake regulation gene by using PCR primers Ypfur1 and Ypfur2, as described by Hinnebusch et al. (9).The epidemiologic role of rodent-borne Bartonella spp. as a cause of disease in humans is emerging in the Americas.This role has been suggested by identification of a novel rodent-associated Bartonella strain causing febrile illness in the

Table .
Prevalence of Bartonella spp.and Yersinia pestis in rodents from Echarate District, Cusco, Peru Figure, panel A).During 2007-2012, a total of 573 clinical cases of Buruli ulcer disease were reported to WHO by the National Leprosy, Buruli Ulcer and Yaws Control Program in Congo.We report 108 cases (19% of all cases reported) that 5 (17.9)Emerging Infectious Diseases • www.cdc.gov/eid• Vol. 20, No. 6, June 2014