Molecular survey and phylogenetic analysis of Bartonella sp., Coxiella sp., and hemoplamas in pudu (Pudu puda) from Chile: first report of Bartonella henselae in a wild ungulate species

Introduction Recent evidence shows a high diversity of infectious agents in wildlife that represent a threat to human, domestic, and wild animal health. In Chile, wild populations of the most common cervid species, pudu (Pudu puda), have been reported as hosts for novel pathogens such as Mycoplasma ovis-like and a novel ecotype of Anaplasma phagocytophilum. A better understanding of the epidemiology of this group and other intracellular bacteria that might have cervids as hosts would enlighten their population relevance. This study aimed to determine the occurrence and genetic diversity of Bartonella spp., hemotropic mycoplasmas, and Coxiella burnetii in pudus from Chile. Methods The DNA was extracted from the blood samples of 69 wild free-ranging and 30 captive pudus from Chile. A combination of real-time (nouG gene for Bartonella and IS1111 element for C. burnetii) and conventional PCR (16S rRNA for hemotropic Mycoplasma spp. and rpoB, gltA, and ITS for Bartonella spp.) was used for pathogen screening and molecular characterization. Results DNA of Bartonella spp. was detected in 10.1% [95% CI (5.2–18.2%)] samples, hemotropic Mycoplasma spp. in 1.7% [95% CI (0.08–10.1%)], and C. burnetii in 1.0% [95% CI (0.05–6.3%)] samples. Two sequenced samples were identified as Mycoplasma ovis-like, and one free-ranging pudu was positive for C. burnetii. While one captive and two free-ranging pudus were positive for Bartonella henselae, one wild pudu was co-positive for B. henselae and Bartonella sp., similar to Bartonellae identified in ruminants. Discussion To the best of our knowledge, this is the first report of B. henselae in wild ungulate species, and C. burnetii and Bartonella spp. in wild ungulate species in South America. Further research will be necessary to evaluate the potential role of pudu as reservoirs of infection and identify the sources for disease transmission among humans and wild and domestic animals.


Introduction
The recognition of the role of wildlife as reservoirs of pathogens that threaten the health of humans and/or livestock species has increased in the past several decades; accordingly, the relevance of infectious agents in the wildlife conservation field has also amplified (1,2).As expected, there are differences between regions of the world.In South America, for example, there is a lack of scientific publications on infectious diseases in wildlife when compared with more developed countries (3)(4)(5)(6).
Bartonella spp., Coxiella burnetii, and hemotropic Mycoplasma spp.are intracellular bacteria that infect a wide range of animals (7)(8)(9) and humans.Hemotropic mycoplasmas (hemoplasmas) are obligate epi-erythrocytic, cell wall-deficient bacteria that usually generate hemolytic anemia in numerous animal species.Routes of transmission are not fully elucidated, but aggressive interactions and possibly fleas and ticks might be involved.The pathogenic potential of hemotropic mycoplasmas, as a cause of human disease, has not been clearly defined; the public health implications derived from these emerging zoonotic pathogens are underestimated (10).Bartonella is composed of gram-negative fastidious, facultative intracellular microorganisms transmitted by fleas and other vectors that provoke a long-lasting bacteremia in the mammal host.The zoonotic potential of these bacteria is well described, and the term bartonellosis has been implemented to refer to human diseases (11).Coxiella burnetii is a zoonotic, strictly intracellular gram-negative bacterium that infects a wide range of animals.In its sylvatic cycle, it can be transmitted by ticks.In humans, it is considered the causal agent of query fever (Q-fever), and the Centers for Disease Control and Prevention (CDC) has classified this microorganism as a potential bioterrorism agent (12).
There is increased evidence that wildlife species are also susceptible (13)(14)(15) and have the potential to be zoonotic (7,16,17).In Chile, several domestic and wildlife species have been identified as potential hosts for several hemoplasmas (18)(19)(20)(21)(22) and Bartonella spp.(23)(24)(25)(26)(27)(28).Information on Coxiella burnetii is much more limited, with only one report with molecular evidence in bats from Chile (26).Despite being commonly reported in domestic and wild ruminants in Europe and North America (13,29,30), there are no studies for the detection of Bartonella spp.and C. burnetii in these taxa in Chile, and only until recently has it been possible to identify hemoplasmas in domestic camelids, llamas (Lama glama), and alpacas (Vicugna pacos) (31).Finally, the native pudu (Pudu puda) has been identified as the potential host species of several hemoplasmas (10), including Mycoplasma ovis-like, in the Chilean template forest.Pudu is the most common cervid in Argentina and Chile and is considered threatened in both countries (32, 33), as shown in CITES Appendix I.In Chile, pudus inhabit temperate forests heavily affected by anthropic factors such as deforestation, housing construction, free-ranging dogs, and livestock (34).Additionally, a high diversity of infectious agents that could be a threat to their health status has recently been identified (10,(35)(36)(37)(38). Recently, pudus were identified as potential reservoir hosts for the bovine viral diarrhea virus, which is a cause of major disease in cattle (39).This study aimed to determine the occurrence and genetic diversity of Bartonella spp., hemotropic mycoplasmas, and C. burnetii in free-ranging and captive pudus from Chile.

. Animal sampling
Blood samples from frozen banks in rescue centers and zoos/breeding centers were used.The frozen bank samples were opportunistically collected from 69 free-ranging pudus between 2016 and 2022 on admission day from two wildlife rehabilitation centers in the template forest ecosystem of southern Chile in Los Lagos District, one (USS: Cerefas, Universidad San Sebastian) located in the continental area and the other (Ch.S: Chiloe Silvestre NGO) in Chiloe island.Additionally, blood samples from 30 captive pudus were collected between 2017 and 2021 during preventive medicine procedures in two facilities, one located in the Mediterranean ecosystem of Central Chile in the Metropolitan District and the other in Los Lagos District, and do not have contact between centers.Blood samples were obtained by venipuncture of the jugular vein using an evacuated tube system (Vacutainer, Beckon, Dickson and Company, Franklin Lakes, New Jersey, USA) and stored at −20 • C within 6 h of collection.For extensive sampling details, refer to the study mentioned in the reference (10).

. . Conventional (c) PCR for Bartonella spp. characterization
All positive Bartonella spp.nuoG-qPCR positive samples were subjected to cPCR amplification of a fragment of three loci [gltA (43), rpoB (44), and ITS (45)] by cPCR with the aim to molecularly characterize Bartonella spp.cPCR amplification reactions were performed in a T100 Bio-Rad thermocycler (Bio-Rad©, Hercules, CA, U.S.A.), and the details of the amplification conditions are presented in Table 1.Bartonella henselae genomic DNA from a cat tested in a previous study was used as a positive control (42).

. . Quantitative real-time PCR for Coxiella burnetii screening
The screening real-time qPCR targeted a 295-bp fragment of the multicopy insertion element IS1111 and is used for sensitive detection of C. burnetii in biological samples (46) (Table 1).Primers Cox-F (GTC TTA AGG TGG GCT GCG TG) and Cox-R (CCC CGA ATC TCA TTG ATC AGC) and hydrolysis probe Cox-TM (FAM-AGC GAA CCA TTG GTA TCG GAC GTT-TAMRA-TAT GG) were used.Standard curves were constructed using 10-fold serial dilutions (2.0 × 10 7 to 2.0 × 10 0 ) of a gBlock R (Integrated DNA Technologies, Coralville, IA, USA), encoding a 295-bp fragment of the IS1111 element of C. burnetii (Integrated DNA Technologies, Coralville, IA, USA).Amplification efficiency (E) was calculated from the standard curve slope in each run using the following formula (E = 10-1/slope).Coxiella burnetii genomic DNA from a cow was used as a positive control.All PCR runs were performed with nuclease-free water (Thermo Scientific©, Waltham, MA, USA) as a negative control.Replicates showing a Cq difference higher than 0.5 were retested.

. . Conventional (c) PCR for hemotropic Mycoplasma spp.
All positive samples in the irbp cPCR were subjected to a cPCR protocol targeting the 16S rRNA hemotropic Mycoplasma spp.gene (620 bp), using HemMycop16S-322s and HemMycop16S-938as primers (Table 1), according to a previously described protocol (7).All cPCR runs were performed with nuclease-free water (Thermo Scientific©) as a negative control, and a cat sample known to be infected by M. haemofelis was used as a positive control.This protocol was used for screening and later sequencing for molecular characterization of detected hemoplasmas.

. . Electrophoresis
Conventional PCR products were separated by 1.5% agarose gel electrophoresis (LE Agarose Seakem R , Lonza) and stained with SYBR© safe DNA gel stain (Thermo Scientific©).The DNA products with the expected size were purified and sequenced.

. . Purification and sequencing
Only positive samples presenting strong band intensity (Bartonella spp.and hemotropic Mycoplasma spp.) were purified by enzymatic reaction using ExoSAP-IT TM PCR Product Cleanup Reagent (Thermo Scientific©, Carlsbad, CA, U.S.A.), following the manufacturer's instructions.Purified DNA was sent to MACROGEN (Seoul, Korea) for sequencing by the Sanger method in an automatic sequencer (A.B.I Prism 310 genetic analyzer; Applied Biosystem ©/PerkinElmer) for species identification.Forward and reverse sequences were analyzed in Geneious 7.1 (https://www.geneious.com),to obtain consensus sequences.Identity percentages were obtained using BLASTn (47).

. Phylogenetic analysis
Before constructing the phylogenetic inference, sequences belonging to different samples, but representing the same bacterial species, were aligned with Geneious 7.1 (https://www.geneious.com)using the MAFFT alignment  method ( 48) and subsequently analyzed for detection of polymorphism and haplotype identification using DnaSP v5 software (49).
The sequences of the present study were aligned with other sequences from the database (GenBank) through the MAFFT program (Multiple Alignment by Fast Fourier Transform) (48) incorporated in Geneious 7.1 software (https://www.geneious.com).Then, multiple alignments were analyzed using BMGE (Block Mapping and Gathering with Entropy) software to remove ambiguously aligned regions (50).
Representative sequences of Bartonella spp.gltA, ITS, and rpoB genes were deposited in GenBank (55) under the accession numbers OQ162290, OQ137267, and OQ162291.Within sequences that represented the same haplotype, only one representative sequence (with a higher size) was deposited in GenBank and used for phylogenetic analysis.

. . Bartonella spp. phylogenetic analysis
The rpoB sequences of the present study were allocated in the same taxa, sharing a clade with B. henselae Houston 1 (AF171070), B. henselae from a cat from Paraguay (MW514660), and B. henselae from Urva auropunctata from St. Kitts and Nevis (MW728257) (Figure 1).The rpoB diversity analyses are represented on Table 2.
The gltA phylogenetic reconstruction evidenced that the sequence of the present study was allocated to the same clade with Bartonella sp. from a Cervus from Japan (CP019781), Bartonella sp. from a cattle tail louse from Israel (KJ522487), and B. capreoli, B. schoenbuchensis, and B. chomeli (Figure 2).

Discussion
This is the first study to document the presence of DNA of B. henselae in a wild ungulate species and Bartonella spp.and C. burnetii in wild ungulate species in South America.The circulation of Mycoplasma ovis-like in free-ranging pudu in Chile is also confirmed (10).The presence of these intracellular bacteria in free-living pudu could suggest an increase in the interaction between domestic species and their ectoparasites and these native species in their natural habitats.Unlike studies in wildlife in other regions (56)(57)(58), no co-infection with the three evaluated pathogens was found in pudus.However, one pudu was possibly co-infected with more than one Bartonella species, since B. henselae and Bartonella sp., similar to Bartonellae identified in ruminants, were detected.The co-occurrence of different Bartonella species in the bloodstream of reservoir animals such as pudus was earlier described in cats and rodents (59-61), and it illustrates the outstanding tolerance of these hosts to harbor mixed Bartonella infections.This could be mediated by an arthropod vector via multiplication and interaction of different Bartonella genetic variants in their digestive tract, with subsequent simultaneous transmission to the mammal host (62, 63).Culture and further molecular characterization of the isolates (64) should be attempted with these samples in future to confirm the co-positivity with multiple species of Bartonella.
Coxiella burnetii is an important bacterial zoonotic pathogen that can cause Q fever in humans.The bacterium has the potential to cause large-scale outbreaks due to its low infectious dose, environmental resistance, and ability to spread airborne through aerosolization of the pathogen, and is a potential biological threat classified as a "Select Agent" in the USA.Coxiella burnetii has a worldwide geographical distribution, apart from Antarctica and New Zealand, and has a wide and diverse host range.The pathogen primarily affects sheep, goats, and cattle, which are considered their primary reservoirs and sources for human outbreaks (65).The livestock species can be infected with C. burnetii and appear healthy, and people often become exposed by breathing in dust contaminated with animal feces, urine, and birth products.Wild ungulate species have been reported commonly exposed to C. burnetii infection in Europe and North America (66), including eight cervid species, but this report in pudu represents the first in deer from the Southern Hemisphere (65).In   Maximum likelihood phylogenies for a subset of Bartonella spp.inferred using an alignment ( , bp) of the gene encoding citrate synthase (gltA).Calculated substitution model was K P+G (partition ), TIM +F+G (partition ), and TNe+G (partition ).Best models were chosen using the Bayesian information criterion (BIC).
the Basque region in Spain, the prevalence has been categorized as stable throughout time.Therefore, the roe deer (Capreolus capreolus) plays a role in the sylvatic cycle of Q fever (67).In South America, there is no evidence of C. burnetii DNA in blood samples of wild boar (Sus scrofa), marsh deer (Blastocerus dichotomus), brown brocket deer (Mazama gouazoubira), small red brocket deer (Mazama bororo), red brocket deer (Mazama americana), and pampas deer (Ozotocerus bezoarticus) (12, 68).

FIGURE
Maximum likelihood phylogenies for a subset of Bartonella spp.inferred using an alignment ( bp) of the internal transcribed spacer (ITS).Calculated substitution model was TPM u+F+G .Best models were chosen using the Akaike information criterion (AIC).
A recent study (12) found that 5.32% of the sampled deer was seropositive for C. burnetii by an indirect immunofluorescence assay (IFA) for IgG antibodies (anti-phase I); to date, it is the only evidence of exposure to this pathogen in deer in the region.
In Chile, DNA findings of C. burnetii have been reported in samples of animal origin only in bats and bulk tank milk from cows (26,69).The last human Q fever outbreak in Chile was declared in 2017 in the Los Lagos District, the same region where molecular evidence was found in pudu in our study (70).This district is a part of the southern macrozone where seropositivity for humans (6%) was significantly higher than in other regions of the country (70).It is likely that the source of infection for pudu is of anthropogenic origin (livestock), or from exotic deer species, red deer, and/or fallow deer, which have been reported in the area (71) and have been commonly reported infected by C. burnetii in Europe (60,65), or from rodent species previously found to be a source of livestock coxiellosis (72).Other serological or molecular studies in dogs and Darwin fox (Lycalopex fulvipes) in the southern macrozone found no evidence of C. burnetii infection (18,73,74).The finding of only one pudu being positive for the bacterium and the low prevalence of C. burnetii in Chile make serological and molecular screening necessary for a much larger number of pudu samples from the Los Lagos region, to evaluate their potential role as a host of infection for transmission to animals and humans.Additionally, the reports of infectious abortions in captive pudus in Chile (39) added to the evidence that C. burnetii has been involved in reproductive loss in captive exotic ungulates, mainly in bovid species (65), making it necessary to include in the differential diagnosis of possible causes of abortion in pudu.Moreover, future studies should attempt to molecularly characterize via sequencing the C. bunetii found in pudus from Chile.
In this study, B. henselae, an emerging zoonotic pathogen that causes scratch disease in humans and whose transmission mainly involves domestic cats as the main reservoir and cat fleas (Ctenocephalides felis) as the main vector (75), is described for the first time in wild ungulate species.Otherwise, to a lesser extent, it is reported in other mammals, bovines (76, 77) and rodents among them (78-81), suggesting that they have a permissive cycle in nature, being detected in several ecological niches (hosts and vectors) (77).Thus, this finding could indicate the circulation of B. henselae in an infected vector, favoring B. henselae transmission among domestic and wild mammals.Nonetheless, further epidemiological and genotyping studies are necessary to confirm this hypothesis.The DNA of Bartonella bacteria has been widely described in cervid species from Europe (30, 58, 82-84), North America (29, 85-87), and Asia (88, 89), usually with a higher prevalence (between 4.9 and 77.7%) than reported in our study.In South America, there are reports of Bartonella spp. in vector species of wild ungulates (68, 90) but not in their blood samples (68).In Chile, during the last decade, there have been reports of the presence of Bartonella spp. in cats, dogs, minks, and bats (23-26, 28, 91, 92).Bartonella henselae in pudu was similar to B. henselae reported in small Indian mongooses in the Caribbean (93) and cats from Brazil (94) and Paraguay (27).Bartonella sp., related to those reported infecting ruminants, such as B. schoenbuchensis [CP019789, HG977197 (95)], B. chomeli (KM215718) (96), B. melophagi (JF834886) (97), B. bovis, and B. capreoli, was also detected in a pudu in this study.More screening will be necessary to confirm the role of pudu in the epidemiology of this infectious agent and its impact on animal health.
Hemoplasma bacteria have been extensively studied in wild and domestic carnivores in Chile during the last decade.Darwin foxes (Lycalopex fulvipes) present a high prevalence of M. haemocanis causing enzootic and asymptomatic infections (18,19) that could be a source of infection for pudu since both share the same habitat within the Los Lagos region.Hemotropic Mycoplasma spp.have been recently reported in llamas (12.8%) and alpacas (6.3%) (Candidatus Mycoplasma haemolamae) (31) and Mycoplasma ovislike in free-living pudu in southern Chile (14%) (10).Molecular screening of hemotropic Mycoplasma spp. in sheep, livestock, and native (huemul) and exotic ungulates (wild boars, red deer) in the Los Lagos District is recommended to understand the epidemiology of these infectious agents and the possible role of pudu as a host.It is also recommended to evaluate the pathogenicity of Mycoplasma ovis-like in pudu.
For the first time in pudu, the finding of B. henselae and C. burnetii, both zoonotic pathogens, could be relevant to public health.Both B. henselae and C. burnetii are pathogens related to occupational diseases, with evidence of health risks for those working with infected species through occupational exposure in rehabilitation centers, breeding centers, and zoological parks (98,99), representing an important factor to consider in medical and management practices with this animal species.

Conclusion
This study expands the knowledge of bacteria with zoonotic potential carried by pudu.Mycoplasma ovis-like was confirmed in pudus, while Bartonella spp., Bartonella henselae, and C. burnetii were described for the first time in South American ungulates.The results of this study suggest an anthropic impact on wildlife species with domestic species interacting epidemiologically with pudus in their natural habitats.Further research will be necessary to evaluate the potential role of pudu as a host and reservoir of infection, and identify the sources for disease transmission among humans and wild and domestic animals.and writing-reviewing and editing.All authors contributed to the article and approved the submitted version.

FIGURE
FIGUREMaximum likelihood phylogenies for a subset of Bartonella spp.inferred using an alignment ( , bp) of the gene encoding the β subunit of RNA polymerase (rpoB).Calculated substitution model was TPM u+F+G (partition ), TNe+G (partition ), and TIM e+G (partition ).Best models were chosen using the Bayesian information criterion (BIC).

FIGURE
FIGURE

FIGURE
FIGUREMaximum likelihood phylogenies for a subset of Mycoplasma spp.inferred using an alignment ( bp) of the S rRNA gene.Calculated substitution model was GTR+F+G .Best models were chosen using the Akaike information criterion (AIC).
TABLE Summary information of the conventional and Real time PCR primer sets, amplification conditions and their amplicon sizes used in the present study.
TABLE Polymorphism and genetic diversity of rpoB Bartonella species sequences identified in pudu from Chile.