Presence of Leptospira spp. and absence of Bartonella spp. in urban rodents of Buenos Aires province, Argentina

ABSTRACT

Big cities of Argentina are characterized by a strong social and economic fragmentation. This context enables the presence of urban rodents in close contact to the human population, mostly in the peripheral areas of the cities. Urban rodents can harbor a large variety of zoonotic pathogens. The aim of this study was to molecularly characterize Leptospira spp. and Bartonella spp. in urban rodents from the area of Gran La Plata, Buenos Aires province, Argentina. The species of urban rodents captured and tested were Rattus norvegicus, Rattus rattus, and Mus musculus. Leptospira interrogans and L. borgpetersenii were detected in R. norvegicus and M. musculus respectively. Bartonella spp. DNA was not detected in any of the kidney samples tested. No significant differences were observed between the prevalence of bacteria and rodent and environmental variables such as host sex, presence of stream and season by Generalized Linear Model analysis. These results confirm the role of urban rodents as infection sources of Leptospira spp., suggesting the need to implement public health measures to prevent the transmission of Leptospira spp. and other zoonotic pathogens from rodents to humans. Bartonella was not detected in this set of samples.

Graphical abstract

1. Introduction

Cities have significantly higher land surface temperature than the natural rural areas surrounding them creating the so called ‘urban heat islands’ [1].

Many environmental changes caused by urbanization can alter prevalence of zoonotic pathogens [2], and these environments may also contribute in converting certain mammalian species into ‘synanthropic’ (species that can benefit from human-modified habitats) [2]. Zoonotic infections hosted by synanthropic animals occur 15 times more commonly in comparison to the wild animals [3]. Rodents are one of the mammals adapted to life in the cities and they have an important role because of their close association with human activities [4]. In addition, the risk of acquiring rodent transmitted zoonoses is higher in developing countries due to the large number of people living in poor sanitary conditions [5,6].

Argentina is characterized by a strong social and economic fragmentation [7,8]. Most of the population lives in densely populated cities, which has generated the establishment of peri-urban neighborhoods and slums, where the precarious living conditions enable the presence of rodents, and consequently the risk of rodent-associated zoonosis emerges [9, 10,11].

Rodents play an important role as reservoirs in the transmission of pathogens to humans [12–15]. Urban rodents include the house mouse (Mus musculus Linnaeus, 1758), the Norway rat (Rattus norvegicus Berkenhout, 1769), and the black rat (Rattus rattus Linnaeus, 1758). Rodent-associated zoonoses can be bacterial, parasitic, or viral [13,16].

Leptospira spp. and Bartonella spp. are important zoonotic agents with worldwide distribution [17–20, 2,21–23]. Numerous studies worldwide reported the association of these bacteria and urban rodents [2,21,23–25]. Leptospirosis is caused by spirochetal bacteria from the genus Leptospira (family Leptospiraceae), with 64 species classified into two clades: pathogenic, or P clade, which includes isolates that cause human or animal infections, and saprophytic, or S clade, which includes species isolated from the environment that do not cause infections [26]. Species clustering in the pathogenic clade are considered more relevant due to their ability to cause a wide range of clinical signs, including multiple organ failure and even death [5,27–31]. Urban rodents are the main source of infection for humans and other vulnerable vertebrate hosts for leptospirosis. Humans may become infected with Leptospira spp. by direct contact with an infected animal or by indirect contact with soil or water contaminated with urine from an infected animal [18,30,31].

Bartonellosis is caused by bacteria of the genus Bartonella (Bartonellaceae). These organisms are vector-borne, blood-borne, intracellular, Gram-negative bacteria that can induce prolonged infection in the host. These persistent infections make domestic and wild animals’ important reservoirs of Bartonella spp. in nature and can serve as a source for inadvertent human infection [2,32–37]. Urban and wild rodents are considered the primary source of infection for ten species of Bartonella, at least five of them, including – Bartonella elizabethae, Bartonella tribocorum, Bartonella grahamii, Bartonella vinsonii subsp. arupensis, and Bartonella washoensis have been implicated as the cause of human infections [38–40].

In Argentina, several studies molecularly characterized Leptospira spp. isolates from rodents [41,42,43,44,45–47]. In contrast, only a few studies regarding Bartonella spp. were performed in Argentina and, to the best of our knowledge, this is the first study focused on the detection of Bartonella spp. in urban rodents [48–50].

The aim of this study was to molecularly characterize Leptospira spp. and Bartonella spp. in urban rodents from the area of Gran La Plata, Buenos Aires province, Argentina.

2. Materials and methods

2.1. Study area and sample collection

The area of Gran La Plata is located on the coast of Río de la Plata in the southern area of the Metropolitan territory of Buenos Aires, 60 km away from the City of Buenos Aires. This area is divided into three departments: La Plata, Berisso and Ensenada, with an extension of 1162 km² and a population of 787.294 inhabitants [51].

Rodents from seven neighborhoods were trapped; five of them were peripheral to La Plata. The neighborhoods include, Malvinas Argentinas (34°56´43”S, 58°00´36” W), La Isla (34°53′28′′S, 57°59′25′′W), El Retiro (34° 57′51′′S, 58°00′17′′W), La Latita (34°58′31′′S, 57°58′30′′W), and Abasto (34°58′05′′S, 58° 01′47′′W); one neighborhood peripheral to Berisso: El Carmen (34°55′33′′S, 57°53′09′′W); and one neighborhood of the inner city of La Plata (34°55′16′′S, 57° 57′16′′W). The neighborhoods included in the study have different levels of urbanization; most of them have poor sanitary conditions, such as inadequate garbage removal, lack of sanitation networks, lack of potable water, and pavement. These neighbors also have areas susceptible to flooding and people have domestic animals that do not have veterinary care.

Rodents were trapped seasonally from September 2014 to September 2015 and captured using live traps (15×16×31 cm) and then euthanized. After necropsy, renal tissue was stored at -20° C until further testing. During seasonal samplings, between 30 and 50 traps were set daily for three consecutive nights, inside or in the backyard of houses of each neighborhood.

Rodents were identified by mammalogist specialists at the Centro de Estudios Parasitológicos y de Vectores (CEPAVE) of La Plata and Museo Argentino de Ciencias Naturales Bernardino Rivadavia (MACN) of Buenos Aires city (see Acknowledgments), while DNA samples were analyzed at the Centers for Disease Control and Prevention’s Division of Vector-Borne Diseases in Fort Collins, Colorado.

2.2. Ethical statement

This study was supported by the Dirección de Flora y Fauna, Ministerio de Asuntos Agrarios de la Provincia de Buenos Aires (File no. 22,500–7981/10) and was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The specimens were studied and sacrificed following the procedures and protocols approved by the national laws and the Ethics Committee for Research on Laboratory Animals, Farm and Obtained from Nature of the National Council of Scientific and Technical Research (CONICET). No endangered species were involved.

2.3. Molecular detection of Bartonella spp. and Leptospira spp. by real-time PCR and conventional PCR

Kidneys were washed extensively in 0.9% saline solution and stored in 70% ethanol until used for DNA extraction. Genomic DNA from kidney was extracted using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA) according to the manufacturer´s protocol and the samples were kept at -20°C until they were analyzed at the CDC. Quality of extractions was assessed using 0.8% agarose gel electrophoresis and ethidium bromide staining.

The real-time PCR was performed on kidney DNA, targeting the 32-kDa lipoprotein (lipL32) gene [52,53] for Leptospira spp., and the transfer-messenger RNA (ssrA) gene for Bartonella spp [35]. The 25 ul reaction mixture contained the following components: 12.5 ul of 2× PerfeCta MultiPlex qPCR SuperMix (Quanta Biosciences, Gaithersburg, MD), forward and reverse primers at a final concentration of 500 nM, probe at a final concentration of 100 nM, and 5 ul of DNA template. The real-time PCR was performed on an CFX96 Touch™ Real-Time PCR Detection System (BioRad, Hercules, CA, USA) with the following parameters: 1 cycle of 50°C for 2 min, 1 cycle of 95°C for 10 min followed by 44 cycles of 95°C for 15 sec and 60°C for 60 sec for Leptospira spp., and 1 cycle of 95°C for 2 min followed by 45 cycles of 95°C for 15 sec and 60°C for 60 sec for Bartonella spp.

Positive samples were further analyzed by conventional PCR. For Leptospira spp., targeting the lipL32 gene [24], the PCR was performed using the primers, lipL32/270 F and lipL32/692 R (Table 1). The PCR conditions used were as follows: 1 cycle at 95°C for 5 min followed by 35 cycles of denaturing at 94°C for 1 min, annealing at 55°C for 1 min, and extension at 72°C for 5 min. Amplification was completed by a final extension at 72°C for 7 min.

Table 1.

PCR results by pathogen and by target gene

Agent	Target gene	Positive/tested (%)
Bartonella	Real-time ssrA	37/126 (29.36%)
	Conventional N-gltA	0/37 (0%)
	Conventional ITS	0/37 (0%)
Leptospira	Real-time lipL32	18/126 (14.28%)
	Conventional lipL32	13/18 (72.22%)

For Bartonella spp., two separate PCRs were performed: conventional PCR targeting the 16S-23S intergenic spacer region (ITS) and nested PCR targeting a specific region of the citrate synthase gene (gltA) (Table 1). ITS amplification was performed in a 25 ul reaction volume containing: 12.5 ul of 2× Green GoTaq® Reaction Buffer (Promega, Madison, WI, USA), 0.4 uM of each primer, 200 uM each dNTP, 1 U Taq DNA polymerase (Promega), and 5 ul of template DNA. The PCR was performed in a C1000 Touch™ Thermal Cycler (BioRad, Hercules, CA, USA) under the following conditions: 1 cycle at 95°C for 2 min followed by 55 cycles of denaturing at 94°C for 15 sec, annealing at 66°C for 15 sec, and extension at 72°C for 15 sec. Amplification was completed by an additional cycle at 72°C for 1 min [36]. Nested-PCR for the gltA gene was performed using the external primers, gltA-443 and gltA-1210 [32] and internal primers BhCS-781 F and BhCS-1137 R [37]. Both PCRs were performed in the same manner as the one described above. The PCR conditions used for the first amplification were as follows: 1 cycle at 95°C for 2 min followed by 40 cycles of denaturing at 95°C for 30 sec, annealing at 48°C for 30 sec, extension at 72°C for 2 min, and final extension at 72°C for 7 min. The PCR conditions for the second amplification were: 1 cycle at 95°C for 3 min followed by 40 cycles of denaturing at 95°C for 30 sec, annealing at 55°C for 30 sec, extension at 72°C for 2 min and a final extension at 72°C for 7 min.

Samples positive for Leptospira spp. were further identified by sequencing. The PCR amplicons were purified using a QIAquick PCR Purification Kit (Qiagen, Valencia, CA) according to manufacturer’s instructions, and then sequenced in both directions with the Applied Biosystems 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA).

Forward and reverse sequences were assembled using the SeqMan Pro program in Lasergene v.14 (DNASTAR, Madison, WI). A phylogenetic tree was generated by comparing assembled sequences between themselves and with reference sequences available in GenBank. The sequences were aligned by MUSCLE and exported into MEGA X for the selection of the most appropriate nucleotide substitution model. The phylogenetic analysis was performed with Neighbor-Joining algorithm using MEGA X and bootstrap proportions were assessed by the analysis of 1000 replicates.

2.4. Data analyses

The Difference of Proportions Test and the Fisher Test were used to evaluate prevalence differences between host sexes, groups of neighborhoods (G1 vs. G2), and seasons [Quantitative Parasitology 3.0 software, 54]. Due to the low number of cases in some variables, neighborhoods were classified in two groups according to the presence/absence of highly contaminated streams flowing through them. Group 1 (G1) comprises El Retiro, El Carmen, and La Isla neighbors and these neighbors have stream flowing through them and Group 2 (G2) includes Malvinas Argentina, La Latita, Abasto, and the inner city of La Plata neighbors and there is no stream flowing through them.

Seasons were grouped in two-year periods: ‘warm period’ (spring–summer, with mean monthly temperatures equal or above 16°C and monthly precipitation averaging 100 mm), and ‘cold period’ (autumn–winter, with monthly temperatures below 16°C and monthly precipitation below 100 mm).

Associations between the probability of Leptospira infection of rodents and explanatory variables [host species, host sexes, groups of neighborhoods (G1 vs. G2), and seasons], were evaluated using a generalized linear model (GLM) with logit function. Support for competing models was evaluated using the Akaike information. The analyses were performed with software R version 3.6.1 [55].

3. Results

A total of 126 rodent kidneys (one kidney per animal) were tested during the study; 80, 25, and 21 samples were collected from M. musculus, R. norvegicus, and R. rattus species respectively. Though 37 samples screened positive for Bartonella spp. by the real-time PCR, none were confirmed by both conventional PCRs. Eighteen samples were positive for Leptospira spp by using the real-time PCR and 13 out of the 18 positives were confirmed positive by conventional PCR (Table 1).

Results from the Leptospira-positive samples revealed that two sequences were 100% identical to L. interrogans isolated from a human sample in Brazil (GeneBank accession number AY461907) and 11 were identical to L. borgpetersenii isolated from a human sample in Sri Lanka (GeneBank accession number AY461898; Table 2). The phylogenetic analysis placed these sequences in separate and strongly bootstrap supported clades (99% for each clade; Figure 1).

Table 2.

Leptospira spp. PCR results by rodent species and by Leptospira species

Rodent spp.	Total	Positive (%)	Leptospira spp.
M. musculus	80	11 (8.73)	L. borgpetersenii
R. norvegicus	25	2 (1.59)	L. interrogans
R. rattus	21	0	–
TOTAL	126	13 (10.32)

Figure 1.

Phylogenetic relationships of Leptospira spp. inferred by the neighbor joining analysis of the – lipL32 gene nucleotide sequences. The acronyms corresponding to Leptospira spp. from four neighborhoods: LAT (La Latita), IS (La Isla), MATU (El Carmen), RETI (El Retiro), and others obtained of Gen Bank.

The prevalence of Leptospira spp. in relation to host species, host sex, groups of neighborhoods (G1 and G2), and season is showed in the (Table 3). No positive samples were detected in Rattus rattus. No significant differences were observed concerning host sexes, groups of neighborhoods, and seasons (p > 0.05) by GLM.

Table 3.

Prevalence of L. interrogans and L. borgpetersenii and confidence intervals (CI) based on host species, host sex, neighborhoods, and seasons

	L. borgpetersenii		L. interrogans
Host species (N)	Positive (P)	95% CI	Positive (P)	IC 95%
R. norvegicus (25)	0	0.00 ± 16.12	2 (8%)	1.24 ± 26.34
R. rattus (21)	0	0.00 ± 18.63	0	0.00 ± 18.63
M. musculus (80)	11 (13.75%)	7.74 ± 23.21	0	0.00 ± 5.64
Host sex
Males	5 (13.89 %)	5.72 ± 29.28	1 (5.88 %)	0.00 ± 29.25
Females	6 (13.64 %)	6.12 ± 27.21	1 (12.50 %)	0.50 ± 49.50
Neighborhoods
G1	7 (22.58 %)	11.23 ± 40.19	2 (10.53 %)	1.90 ± 32.88
G2	4 (6.78 %)	2.28 ± 16.77	-
Year period
Warm	8 (11.94 %)	5.99 ± 22.18	2 (25 %)	6.66 ± 60.01
Cold	3 (23.08 %)	7.75 ± 51.07	-

None of the variables considered in the global GLM was associated with infection of L. borgpetersenii in rodents. For the initial model obtained for all the variables an AIC of 56.4 (log likelihood with residual deviance and degrees of freedom (DF) = 46.40) was recorded. No significant associations were identified in the first top model between L. borgpetersenii infection and each variable (host sexes, groups of neighborhoods and seasons). Leptospira interrogans was not analyzed since only two positive samples were detected.

4. Discussion

Leptospirosis has a worldwide distribution and Latin America, the Caribbean, Southeast Asia, Oceania, and the Indian subcontinent are considered endemic areas for this infection [56]. In these regions, including Argentina, poor levels of sanitation, limited access to health care, and high populations of reservoir hosts create perfect conditions for maintaining and disseminating the disease [56,57].

Almost all mammals can serve as hosts of leptospirosis, carrying the bacteria in their proximal renal tubules and shedding them through the urine [56,58]. However, rodents serve as the major carriers, and it has been shown that the Norway rat is the main reservoir for L. interrogans, and the house mouse is the main reservoir of L. borgpetersenii. [57,59–61]. This is consistent with our results, where two R. norvegicus (Norway rat) were positive for L. interrogans and 11 M. musculus (house mouse) were positive for L. borgpetersenii.

No significant differences were observed between the pathogens studied and host sex, neighborhoods (G1 and G2) as previously reported [24]. However, the prevalence of Leptospira spp. trended higher in the rodents from neighbor G1 (streams flowing through) than in the ones from neighbor G2. L. borgpetersenii cannot survive long in the environment without a host. L. interrogans is also transient in the environment but can survive without a host. It is likely that the drier environment is related to the populations of carrier mammals, which for L. borgpetersenii are more common in mice, mongoose, and cattle. L. interrogans is most common in rats but can also be found in other mammals. Leptospira spp. are more prevalent in areas susceptible to flooding, however, wet soil may be more important than flowing water for transmission of leptospirosis.

While L. interrogans remains viable for long periods in humid environments, L. borgpetersenii was more commonly detected at sites without natural bodies of water [62]. Though our findings do not support results reported by [62], we have to admit that we tested a relatively small number of samples.

Rattus rattus can be associated with the same Leptospira spp. along with R. norvegicus [62,63]. However, in this study all R. rattus samples tested negative for Leptospira DNA. This might be explained by behavior difference observed in the rat hosts. For example, R. rattus live on higher grounds and avoid swimming, while R. norvegicus prefer lower, flooded areas as this species is an excellent swimmer that generates a higher exposure to Leptospira spp. infection [41]. Another factor could be related to the difference in existing competitive hierarchy between these two species, according to which R. norvegicus dominates R. rattus and leads to its displacement. These factors are supported by studies carried out in different parts of the world [41,64,65].

Regarding M. musculus, the relationship of this rodent with L. borgpetersenii has been recorded in other countries [66–68]. Moreover, the increase in human leptospirosis associated with L. borgpetersenii serogroup Ballum has been reported worldwide [66,69,70].

Though real-time PCR detected a number of positive rats for Bartonella DNA, no confirmation by conventional PCR and sequencing of the positive samples was achieved. The markers that we used in the study (ssrA as a screening test and gltA and ITS as confirmatory tests) are among the most commonly used genetic loci [71]. The real time PCR was used as a screening test since it is a Genus-Specific Assay that can amplify over 30 Bartonella species, subspecies and strains [35]. This assay was developed in the same laboratory where the rat samples were tested and it was effective for many other investigators [35]. However, identification of Bartonella DNA based on sequence information is the most important step in confirming occurrence of this infection in animals considering the high diversity of this group of bacteria and unknown level of specificity of used genetic markers.

The failure to amplify Bartonella spp. DNA from the rodent samples can relate to the type of sample tested in the study (kidney), but likely represent the real absence of Bartonella spp. DNA in these samples. The most commonly used sample to detect Bartonella spp. is blood, although other tissue samples tested are the spleen, liver, heart, and kidney [71]. There is limited information regarding the variation of detection of Bartonella between tissues [71]. Since Bartonella colonize the vascular endothelial cells and every five days are released into the blood stream, infecting the erythrocytes where they multiply, the majority of the studies used blood samples [71]. Five studies used kidney samples for detection of Bartonella DNA [71]. For an example, [72],detected Bartonella DNA in kidneys of 13.2% of Bandicota bengalensis, 13.3% of Rattus rattus brunneusculus, and 53.3% of Suncus murinus from several areas of Nepal.

In many areas of the world, Bartonella spp. can reach a very high prevalence in rodents, while in some places the prevalence was very low or the bacteria were absent [2]. For instance, rats from coastal cities on both the Pacific and Atlantic coasts in the United States and Canada carried Bartonella spp. while other rat populations from noncoastal cities were free of Bartonella spp [73]. The prevalence of Bartonella spp. in rodents in Japan varied considerably, while rodents from Kanagawa, Nagano, and Ehime prefectures reported significantly high prevalence rates (>80%), rodents captured from urban areas (Yokohama and Shimoda) were all negative (0/345) for Bartonella species [74]. Another study in Cotonou, Benin, reported that both R. rattus and R. norvegicus were Bartonella-free [75].

5. Conclusions

Our findings of two species of Leptospira (L. interrogans in R. norvegicus and L. borgpetersenii in M. musculus) support previous reports, for example studies carried out in Brazil, Canada and New Caledonia were these host-pathogen associations were found [12, 76]. Nevertheless, it is necessary to generate new studies to evaluate factors that affect the distribution of the bacteria in these two host species and in R. rattus, where no positive samples were found.

Regarding Bartonella, further studies testing larger numbers of urban rodents could yield a more comprehensive understanding of the prevalence of Bartonella spp. and other rodent zoonotic infections in the area. Our results underscore the importance of sanitation and rodent control for the protection of public health.

Funding Statement

This study was funded by PIO (Proyecto de Investigación Orientado) CONICET-UNLP.

Disclosure statement

No potential conflict of interest was reported by the author(s).