Assessing cross-species transmission of hemoplasmas at the wild-domestic felid interface in Chile using genetic and landscape variables analysis

The co-occurrence of domestic cats (Felis silvestris catus) and wild felids in rural landscapes can facilitate pathogen transmission. However, in the relatively-isolated regions of southern South America there have been no comprehensive studies to assess disease transmission risks between domestic cats and forest-dwelling wild felids such as guigna (Leopardus guigna). We evaluated hemoplasma infection and the possibility of transmission between domestic cats and guignas by comparing spatial and phylogenetic patterns of pathogen prevalence. Blood/spleen samples were collected from 102 wild guignas and 262 co-occurring rural domestic cats across the entire distribution range of guigna in Chile. Hemoplasma infection was assessed by direct sequencing of the 16S RNA gene. Infection with hemoplasmas was common and geographically widespread across different bioclimatic areas for both species. The most common feline Mycoplasma species in guigna and domestic cats were Candidatus M. haemominutum (CMhm) (15.7% guigna; 10.3% domestic cat) and Mycoplasma haemofelis (Mhf) (9.8% guigna, 6.1% domestic cat). A previously undescribed Mycoplasma sp. sequence was found in two guignas and one cat. Continuous forest-landscapes were associated with higher hemoplasma-prevalence in guignas. Shared hemoplasma nucleotide sequence types between guigna and domestic cats were rare, suggesting that cross-species transmission between guignas and domestic cats may occur, but is probably uncommon. Ectoparasites, which have been linked with hemoplasma transmission, were not found on guignas and were infrequent on domestic cats. Our results suggest that transmission pathways vary among hemoplasma species and, contrary to our predictions, domestic cats did not appear to be the main driver of hemoplasma infection in guignas in these human-dominated landscapes.

animals and wildlife in human-dominated landscapes 2 . Phylogenetic relatedness, interspecific interactions and geographic overlap among host and vector species are main predictors of pathogen spillover 3 . Domestic cats and wild felids, which share a common ancestor a relatively recent 10 million-years ago 4 , often share genetic, physiological and ecological traits and thus share similar susceptibility to many infectious agents 2 . These shared biological features and the ubiquitous presence of domestic cats around human settlements make co-occurring domestic and wild felids an important study system to investigate frequency and impact of pathogen spillover at the wildlife-human interface 5 . Transmission of pathogens from domestic to wild felids has been linked with population declines among wild species, such as the case in bobcats (Lynx rufus) with the outbreak of feline panleukopenia virus in 1988 6 , and the outbreaks of feline leukemia virus (FeLV) in Iberian lynx (Lynx pardinus) in 2006 7 and Florida panther (Puma concolor coryi) in 2001-2004 8 and 2010-2016 9 . Therefore, gaining a better understanding of the mechanisms, ubiquity, and patterns of pathogen transmission among domestic and wild felids has epidemiological, ecological and conservation implications.
Mhf is also the most pathogenic and can cause hemolytic anemia in immunocompetent cats. Hemoplasma acute infection can induce hemolytic anemia, anorexia, lethargy, dehydration, weight loss and sudden death 17 . The chronic form has been associated with the absence of clinical signs. More studies to elucidate the long-term implications of hemoplasma infection are necessary 11 .
The epidemiology and transmission patterns of hemoplasmas are poorly understood 12 . Geographical aggregation of infection in some studies relates transmission with an arthropod vector 29 , and the cat flea, Ctenocephalides felis, has been proposed to be a potential vector in Mhf 30 . However, other experimental studies have found no evidence of hemoplasma transmission by fleas 31 . Alternatively, cat fights could also be a route of hemoplasma transmission since subcutaneous inoculation of CMt infected blood in laboratory settings resulted in transmission, whereas saliva did not 32 . If this were common, then transmission by aggressive interaction would be more likely than transmission by social contact (saliva via mutual grooming) 32 . However, for CMhm (but not for Mhf), there is evidence of horizontal transmission by direct contact between cats but no observed signs of aggressive interaction or vectors 29 . Additionally, the discovery of the feline hemoplasma species, CMt, which is closely related to rodent hemoplasmas, supports the hypothesis of predator-prey transmission of these agents between mice and cats 33 . Finally, it has been suggested that blood transfusions might be a potential route of transmission 34 .
Here we assess the prevalence and cross-species pathogen transmission among small felids in rural landscapes of central and southern Chile. In these areas non-feral free-roaming domestic cats are abundant and can roam up to 2 km from their human households during incursions into native forest habitat (unpublished GPS tracking devices and camera trap data). This roaming behavior increases the probability of contact with forest-dwelling wildlife species and other domestic animals and thus increases the chances of coming in contact with a range of pathogens. Moreover, these domestic cats are rarely subjected to any prophylactic programs or veterinary care, increasing their possibility of being infected with pathogens 35 . To date, there have been no systematic studies or records documenting the existence of completely feral cat populations in Chile.
The forest-dwelling guigna (Leopardus guigna) is a small wild felid inhabiting central and southern Chile (30°-48°S) and a narrow strip of land in southwestern Argentina (39°-46°S west of 70°W) 34 . It is one of the most threatened wild felids in the Americas and is classified by the IUCN as Vulnerable with declining populations 36 . Guignas are solitary and closely associated with native forests, as they depend on vegetation cover [37][38][39] and the small mammals and birds prey species. However, they also inhabit small forest patches surrounded by a matrix of human-modified habitats (i.e., livestock and agriculture) where domestic cats are often present 40 , providing opportunities for contact that may facilitate pathogen spillover. Moreover, guignas in these human-dominated landscapes have disperse greater distances and occupy larger ranges, rising the chances of contact with other species. These populations also have reduced genetic diversity, making them potentially more susceptible to infectious diseases 41 . Possible cross-species transmission from domestic cats to guignas of FeLV and feline immunodeficiency virus (FIV) has been recorded in human-perturbed landscapes in southern Chile 42 . However, there is no information about other pathogens that could be infecting guignas.
In Chile, hemoplasma has been detected in domestic cats in the city of Valdivia (southern Chile) with infection prevalence rates between 1-8% 43 depending on the bacterial species. Similarly the Darwin's fox (Lycalopex fulvipes), a small forest-dwelling canid endemic to southern Chile, has a high infection prevalence of Mycoplasma spp. (57% of 30 individuals) 44 . Although they are found throughout the current distribution of Darwin's foxes and coexist with rural domestic cats in human-perturbed landscapes, hemoplasma infection has never been described in guignas.
Here, we used epidemiology, phylogenetics and landscape ecology to understand patterns of hemoplasma variation at the wildlife-domestic interface. We carried out an extensive molecular survey for hemoplasma infection in guigna across its entire distribution range in Chile and in co-occurring rural domestic cats. We evaluated hemoplasma infection and prevalence and modelled possible transmission pathways and landscape drivers associated with infection. We conducted genetic analyses based on sequences of the bacterial 16S rRNA gene to identify hemoplasma species infecting several known host species and evaluate the existence of shared genotypes as evidence of cross-species transmission. To understand the possible impact of hemoplasma infection on guigna health, we also assessed the clinical status of hemoplasma-infected guignas.
We hypothesized that human-dominated landscapes provided increased opportunities for domestic cat-guigna encounters and cross-species pathogen transmission. Therefore, we expected to find i) higher hemoplasma prevalence in guignas inhabiting human-dominated landscapes compared with pristine continuous forest habitats, and ii) high genetic similarity (and shared genotypes) among domestic cat and guigna hemoplasma sequences.

Materials and Methods
Study area. The study was conducted across the entire current distribution of guignas in Chile (33°S-46°S) 36 , encompassing four different bioclimatic regions (Fig. 1). The study sites included a gradient of different landscape types, ranging from continuous pristine native forest with no human presence to human-perturbed landscapes with high human densities and where fragments of remnant forest are surrounded by a matrix of agricultural, livestock along with their domestic pets (cats and dogs) and other human activities.
Sample collection. From 2008 to 2018, 102 free-ranging guignas were sampled after being captured in tomahawk-like live traps baited with chicken and olfactory attractor (Hawbakers wildcat lure; Fort Loudon, PA, USA) (n = 52) or after being received injured in wildlife rescue and rehabilitation centers (WRRC; n = 8). Whole blood samples were collected from these 60 animals by cephalic or jugular venipuncture in EDTA tubes. In addition, 42 spleen samples were collected during complete necropsies of "road-killed" cats found dead or following euthanasia at WRRC. Guigna captures and tissue collection were conducted using standard protocols 41 that conformed with handling and supervision protocols within bioethical and animal welfare frameworks 45  The anesthesia protocol (5 mg/kg ketamine and 0.05 mg/kg dexmedetomidine) was adapted from methods described in other wild felid species from South America as L. colocolo 46 . For each guigna we recorded the GPS location of collection/capture, sex, age class (estimated from dentition patterns), physical condition, season and presence of ectoparasites. Any clinical signs of disease were assessed and recorded by a veterinarian. Of 38 sampled females and 64 males, 63 were adults and 16 juveniles (no age data was available for 23 of the animals).
We sampled 262 free-roaming domestic cats from rural communities across the entire distribution of guigna in Chile. Whole blood samples (n = 254) were collected by cephalic or jugular venipuncture using manual restraint after the owner's consent. Eight spleen samples were collected during complete necropsies of domestic cats found as road-kills or euthanized at veterinary clinics. The sex, age class, season and "capture" location of www.nature.com/scientificreports www.nature.com/scientificreports/ each cat was recorded. The presence of ectoparasites was assessed in 125 domestic cats (48%). In total, 129 females and 133 males, 226 adults and 36 juveniles were sampled.
All procedures were conducted with full consideration of animal welfare and ethical protocols and with the approval of Animal Ethics Committee of the Institute of Ecology and Biodiversity in Universidad de Chile (resolution of 20 November 2015). Samples were stored frozen at −20 °C until needed for molecular analyses.
Genetic analysis of hemoplasma sequences. Total DNA from 100 µl of blood and 10 mg of spleen was extracted using a DNeasy ® Blood & Tissue Kit (Qiagen, Germany) following the manufacturer's instructions. To monitor for cross-contamination during the extraction process, negative controls consisting of 100 µl phosphate-saline buffer were concurrently prepared with each batch of 15 samples. We used universal primers 44 to amplify 391 bp (base pairs) of the 16S rRNA gene from a broad range of hemoplasma species. PCR conditions consisted in 20 µl volume containing 2.5 µL 10x PCR Buffer, 2.0 µL MgCl2 50 mM, 2.5 µL dNTP, 0.3 U of Platinum Taq DNA polymerase, and 1.0 µl of each forward and reverse primers 10 pmol/µl). Thermocycling parameters consisted on an initial denaturation at 95 °C for 5 min, 40 cycles of 94 °C for 30 seconds, 58 °C for 30 seconds, 72 °C for 30 seconds, and a final extension of 72 °C for 7 min. All samples that were determined to be positive were molecularly characterized by amplifying the complete 16S rRNA gene using a semi nested PCR that amplified 492 bp and 1107 bp of the 16S rRNA gene, with an overlapping fragment of 171 bp to facilitate sequence assembly. The first round conditions were the same for the two PCRs, using Hemo-F1 and Hemo-R2 published primers and PCR conditions 47 , and were subsequently amplified through two different reactions using Hemo-F1/Hemo-R1 primers, and Hemo F2/Hemo-R2 primers 47 (Table 1).
To assess the possibility of coinfection by different hemoplasma species, we screened positive samples using species-specific PCR primers that targeted the three recognized species of feline hemoplasma. To detect Mhf-and CMhm-derived DNA, one-step PCR amplification was performed following published PCR conditions 21,48 and using two forward and one reverse oligonucleotide, including a OH-OK Mhf-specific forward primer, a CA-B2 CMhm-specific forward primer, and a 00CR-r1 common reverse primer 48 (Table 1). The CMt PCR amplification used primers CMt-F and CMt-R 21 .
For all PCR assays, ultrapure water was used as a negative control. PCR products were visualized in 2% agarose electrophoresis gel stained with GelRed fluorescent dye. After PCR, all products were purified and sequenced in both directions at Macrogen (South Korea) with primers described above. All sequences were aligned using CLUSTALW algorithm (Geneious ® ) and compared to those of the GenBank database to assess their uniqueness.
The best model of evolution among sequences was determined with the program jModelTest2 (version 2.1.6) 49 , under Akaike Information Criterion (AIC) 50 (both for each hemoplasma species separately and altogether) to identify the level of complexity of the model of nucleotide substitution that best fit the combined dataset. Phylogenetic trees were constructed based on Neighbor-Joining, Maximum likelihood and Bayesian methods using GTR + I + G model. Maximum-likelihood analysis was conducted using RaXML software version 1.5 51 . The data set was resampled 1,000 times to generate bootstrap values. We used software iTOL for visualization 52 . Bayesian inference analysis was performed with Mr. Bayes 3.1.2 53 . Four Markov chain Monte Carlo (MCMC) simulations were run for 10 9 generations with a sampling frequency of every 1000 generations and a burn-in of 25%. Mixing and convergence of the MCMC was assessed through the average standard deviation of the split frequencies (ASDSF), being the optimum closest to zero. Estimated sample size (ESS) values were assess for stabilized variance, targeting values greater than 100. Genetic divergence between clades was calculated as genetic distance (Ks) with DnaSP.5 software 54 .
A nucleotide sequence type (ntST) network was generated using the median joining approach method implemented in popart software (Population Analysis with Reticulate Trees) 55 . Phylogeographic structure of CMhm and Mhf in guignas and domestic cats was assessed comparing frequency-based G ST with the pairwise difference-based N ST indices 56 implemented in Permut 57,58 . The level of significance was assessed with 1000 permutations. Genetic structure of CMhm and Mhf in guigna and domestic cat host species was estimated using Within each buffer area, we described and quantified seven landscape variables of land use and human occupation, biologically relevant to assess our questions: (1) percent vegetation cover 63 , (2) presence of houses, 3) number of houses, (4) distance of guigna capture location to the nearest house, (5) land use (fragmented landscape or continuous forest), (6) local administrative region and (7) bioclimatic region (Mediterranean region, rainy temperate region, Chiloe Island: rainy temperate to oceanic cold temperate transition region and oceanic cold temperate region) 64 . GIS layers were obtained from the Ministerio de Bienes Nacionales website 65 and QGIS 2.14 ® software was used to extract the landscape attributes.
We tested the collinearity among predictors by running bivariate Pearson correlations among pairs of variables. No statistically significant correlations were found.
To address spatial autocorrelation in our data, we conducted a Global Moran's I test using ArcGIS Pro. We obtained non-significant results (Moran's index = 0.38, z-score = 0.46, p-value = 0.64) suggesting there is no pattern of data spatial clustering.
Statistical analyses. Spatial (landscape use) and biological (host age and sex) independent variables and hemoplasma infection (binary response variable) were assessed with multivariate logistic regression analysis (function glm), calculating crude and adjusted odds ratios (ORs) (function confint) within 95% confidence intervals (CIs). Goodness of fit models were assessed using the Hosmer Lemeshow test (function hoslem.test) and an analysis of residuals.
Differences in hemoplasma spp. infection prevalence between domestic cat and guigna, as well as between seasons, were analyzed by Mann-Whitney U tests (function wilcox.test). Hematological and biochemical parameters of hemoplasma-infected and non-infected guignas were compared by Kruskal-Wallis tests (function kurskall. test). Comparisons between flea-infected and hemoplasma-infected domestic cats were conducted using Fisher exact tests. All statistical analyses were performed in R software 66 with a significance level of p < 0.05.
Results of species-specific PCR revealed coinfection of CMhm and Mhf in three guignas (2.9%; 95% CI = 0.3-6.2%) and five domestic cats (1.9%, 95% CI = 0. www.nature.com/scientificreports www.nature.com/scientificreports/ D) that corresponded with the three identified hemoplasma species and an unidentified Mycoplasma sp. group or OTU (operational taxonomic unit) (Figs 2 and S1). The ntSTs from our sampled guignas and domestic cats from Chile were positioned within clades shared with domestic cats, wild felids and other carnivore species from around the world. However, a separate group consisting of two guigna sequences (A5, A8) were observed within the CMhm clade. Genetic distance between clades C and D is lower (0.037) than between clades A and C (0.067), and clades A and B (0.086).
Genetic analysis of the most prevalent bacterial species (CMhm and Mhf) described here along with world-wide hemoplasma sequences obtained from GenBank revealed no geographic structure (all non-significant www.nature.com/scientificreports www.nature.com/scientificreports/ paired Phi st tests) or genetic structure among domestic and wild carnivore host species (CMhm: S nn = 0.55, p = 0.25; Phi st = 0.07, p = 0.15; Mhf: S nn = 0.5, p = 0.2; Phi st = 0.006, p = 0.45) (Fig. 3). Gst-Nst test could not be conducted due to low number of sequences per location.
In contrast, genetic structure based on host species for CMhm was visually apparent, as all ntST were species-specific, either unique to domestic cats or unique to guignas (Fig. 4). Statistically significant genetic structure of CMhm sequences between guignas and domestic cats was found, based on contrasting ntST frequencies (Phi st = 0.65, p = 0.00) and nucleotide sequence-based statistics (S nn = 0.95, p = 0.00). On the other hand, two cases of shared ntSTs between both host species were observed (1 for Mhf, 1 for Mycoplasma sp.) and no genetic structure of Mhf sequences between guignas and domestic cats was found (Phi st = 0.19, p = 0.39; S nn = 0.58, p = 0.43). exposure risk factor analysis. Although hemoplasma-infected guignas were found in all four bioclimatic areas sampled in Chile (Fig. 1, Table 4), multivariate logistic regression analysis revealed a significantly higher prevalence of total hemoplasma species and for CMhm specifically (Tables 4 and S1, Fig. S2) in the rainy temperate area. We also found a significantly higher prevalence of total hemoplasma species and specifically of CMhm and Mhf (Table S1, Fig. S2) in guignas from continuous forest compared with fragmented landscapes. However, there were no differences in infection prevalence between sampling seasons, males and females or among different age classes in guignas. In domestic cats, higher prevalence for total hemoplasma species was observed in males compared to females (p = 0.05) but without significant differences. There were no differences between sampling seasons, age classes or bioclimatic areas (Tables S1 and S2).
Ectoparasites were found in seven of 125 evaluated domestic cats (5.6%; 95% CI = 1.5-9.6). All ectoparasites were fleas and no ticks were found. Two hemoplasma-positive domestic cats had fleas (1 CMhm, 1 Mhf). There was no statistical difference in hemoplasma infection prevalence between flea-positive and negative animals (Fisher exact test; p = 0.3). The seven flea-positive domestic cats were found in both cold and warm seasons. We were unable to compare ectoparasite prevalence between bioclimatic zones due to the small number of flea-positive animals. No ectoparasites were found on the guignas.  (Table S3, Fig. S3). Hematological parameters were obtained from 11 hemoplasma-infected and 8 hemoplasma-negative guignas, while biochemical parameters were obtained from 10 hemoplasma-infected and 8 hemoplasma-negative guignas (Tables S3 and S4). Comparisons revealed no statistically significant differences in hematological and biochemical values between hemoplasma infected and non-infected guignas, although sample numbers for these comparisons were low. Comparison of hematological and biochemical values of hemoplasma-infected guignas with the baseline data from the closely-related Leopardus geoffroyi 67 suggested that two juvenile male guignas had nonregenerative anemia. Both of these individuals were infected by Mhf and one was coinfected with CMhm. Both also presented abnormal ALT, albumin and AST values, but neither presented inclusions in red blood cells observed in blood smears. Asymptomatic animals infected by CMhm, Mhf, or Mycoplasma sp. and coinfected by CMhm and Mhf and Mycoplasma sp. and Mhf were also found.

Discussion
To our knowledge, this study represents the first large-scale molecular survey of hemoplasma for any co-occurring domestic/wild carnivores. Our remarkably large sample size of a threatened, elusive wild felid such as the guigna, including samples from its entire distribution in Chile, was matched with a large sample of rural domestic cats from the same areas. Genetic analyses revealed that infection with genetically diverse hemoplasmas is common and geographically widespread in both species.
In domestic cats, higher hemoplasma prevalence have been found in free-roaming domestic cats (e.g. 37% in Japan 20 ) than in confined cats in urban areas (2-8% Valdivia city, Chile 43 ; 4% different USA cities 18 ), probably because greater exposure with bloodsucking arthropods and a higher incidence of wounds increase their exposure to hemoplasmas 18 . In our study, the proportion of infected cats ranged from 11% to 42%, depending on the area, and was not significantly associated with tick or fly infestation. We found no differences in hemoplasma spp. infection rates between sex and age classes, although other authors have reported higher hemoplasma prevalence in older domestic cat males 12 . This may be because the free-ranging behavior of all the sampled domestic cats, regardless of sex or age, may increase their exposure to a variety of bloodsucking vectors (e.g., mosquitoes, horseflies) and lead to increased social contact, including fighting. In addition, the possibility of vertical transmission and/or onward horizontal transmission could explain the lack of significant differences on infection among age classes. www.nature.com/scientificreports www.nature.com/scientificreports/ In our survey, the most prevalent hemoplasma species, both in guigna and domestic cats was CMhm, followed by Mhf. This is similar to results from previous studies of other felid populations 12,20,21,23,68 . CMhm may infect guignas and cats more persistently and efficiently than other hemoplasma species, or perhaps is less virulent, allowing the coexistence between host and bacteria. Alternatively, these differences could be associated to different transmission methods among hemoplasma species.
Molecular characterization of 16S rRNA gene (1289 bp) was obtained for 11 of the 24 hemoplasma-positive guignas and 22 of 40 hemoplasma-positive domestic cats. The relatively low success rates of the assays was due to the quality of field samples, which were sometimes suboptimal due to degradation in field conditions (e.g, roadkills). Additionally, fluctuating pathogen loads on blood samples, where are correlated with the state of infection of the host, may also have affected the efficient amplification of long gene fragments (more than 1000 bp).
Phylogenetic analysis revealed well-supported clades that corresponded with the three identified hemoplasma species. NtSTs from sampled guignas and domestic cats from Chile were positioned within clades shared by domestic cats and wild felid and carnivore host species from around the world, suggesting a worldwide distribution of hemoplasmas and no phylogeographic differentiation among continents within each species-specific clade. Worldwide interspecific transmission of these pathogens between wild and domestic carnivore host species has been described by other authors 23,25,27 .
The ntST networks and genetic relationships among guigna and domestic cat hemoplasma ntSTs, Mhf and Mycoplasma sp. showed some evidence of cross-species transmission, as there was one shared ntST in each of these hemoplasma species. For Mhf, we also observed private ntSTs for domestic cats and for guignas (including that with the higher frequency in this species), and no genetic structure between hosts.
However, CMhm ntST networks of sampled guignas and domestic cats from Chile showed a clear pattern of genetic clustering by host species, where all ntSTs were either unique to domestic cats or to guignas, and shared ntSTs were not found. The most prevalent ntSTs in guigna were all unique to this species. Therefore, a statistically significant high-level of genetic structure between guigna and domestic cat hemoplasma ntSTs was found. These results suggest that cross-species transmission events between both hosts may occur for some hemoplasma species but are most-likely relatively rare. As other authors have suggested 23,27 , domestic cats may be the source of the global distribution of multiple strains of hemoplasmas in wild felids, and these may persist through onward transmission following spillover. Overall, we found evidence for higher levels of intraspecific, versus interspecific, transmission in both guignas and domestic cats. This pattern is also supported by the analysis of landscape features, which found significant associations between higher hemoplasma prevalence of infection in guignas with pristine, continuous native forest devoid of domestic cats. Therefore, contrary to our hypothesis and the conceptual framework for pathogen emergence in wildlife, domestic cats in human-dominated landscapes appear not to be reservoirs or main drivers of hemoplasma infection in guignas.
Other authors have also described higher prevalence of hemoplasma infection in wild species inhabiting pristine areas compared to urban areas, including raccoons in USA 68 , where higher hemoplasma prevalence in pristine areas was associated with higher tick infestation rates in host species. In some cases, the clustered geographical distribution of hemoplasma, mainly in regions with warmer climates [69][70][71][72] , has been related to the role of an arthropod vector (mostly ticks) as one of the main transmission modes 29 . In our study, the widespread distribution of hemoplasma-positive animals in different seasons (cold and warm) and across study areas, including regions with marked differences in bioclimatic characteristics, and the higher prevalence of hemoplasma infected guignas in the rainy temperate bioclimatic area where ticks potentially associated to hemoplasma transmission are absent 73,74 , refute the idea that ticks are the main hemoplasma vectors. In this study, higher hemoplasma prevalence in pristine areas was not associated with higher tick or fly infestation rates in host species from these areas. Fleas were found in only 5.6% of evaluated domestic cats and no other ectoparasites were found infesting domestic cats or guignas. However, this does not rule out the possible role of other bloodsucking vectors, such as mosquitoes or horseflies.
On the other hand, guigna population densities are higher in pristine forest areas compared to fragmented landscapes 41 , suggesting hemoplasma transmission could be density-dependent. Therefore, direct transmission via infected blood or saliva 31,32 should also be taken into account as a possible transmission mode explaining the widespread occurrence of hemoplasmas in wild guigna populations, especially considering possible transmission networks that include other forest-dwelling native carnivores (e.g., Darwin's fox). No statistical differences on infection prevalence between adults and juveniles in guigna reveal the possibility of transplacental or vertical transmission, as well as horizontal onward transmission. The absence of statistical differences in infection prevalence among seasons suggests that there might be different transmission pathways, perhaps including one or more species of bloodsucking vectors (e.g. mosquitoes, horseflies), in addition to direct (vertical and horizontal) transmission.  www.nature.com/scientificreports www.nature.com/scientificreports/ Phylogenetic analysis also revealed the presence of a well-supported clade, not related to any previously described mycoplasma species worldwide. Two guignas and one domestic cat from Chile were positioned within this clade (shared ntST), which is also shared with a hemoplasma sequences found in Darwin's fox from Chile and a Brazilian rodent. This previously undescribed Mycoplasma sp. group may correspond to a new species of hemoplasma, although this hypothesis must be further validated. Whether fox-guigna transmission occurs, or both individuals got infected through predation on rodents, is unknown. Interestingly, the Darwin's fox is a www.nature.com/scientificreports www.nature.com/scientificreports/ forest-specialist carnivore, co-occurring with guignas in most of their distribution ranges, supporting the hypothesis that hemoplasma infection in guigna is associated to pristine, continuous native forest landscapes.
The pathogenicity of hemoplasma infection in wild felids is uncertain 75 , but most wild felids have been asymptomatic 22,25,26,76 . However, clinical signs (pale mucous membranes and nonregenerative anemia) associated with hemoplasma infection have been observed in some species, including the the Iberian lynx 23 . In our study, two hemoplasma infected guignas (juvenile males), of a total of 11 positive guignas for which hematological analysis were conducted, presented nonregenerative anemia. Both were infected by Mhf and one was coinfected with CMhm. In domestic cats, Mhf has been described as the most pathogenic hemoplasma species [11][12][13]28 , which may explain our findings in guignas. However, Mhf-infected guignas not showing clinical signs were also found. The lack of differences in hemoplasma spp. infection prevalence between adult and juvenile guignas and the presence  www.nature.com/scientificreports www.nature.com/scientificreports/ of infected animals in all seasons makes it difficult to define if these are endemic pathogens in guigna populations, or how long they persist in infected individuals. These infections may be chronic in some cases, as has been suggested by other authors in domestic cats 10 , showing mild or no clinical implications. Hence more health studies correlated with hematological analysis are necessary to clarify the pathogenic importance of hemoplasma infection in guignas.
In a related study, we observed high genetic similarity between domestic cat and guigna FIV and FeLV sequences and widespread shared ntSTs, suggesting that domestic cats were the source of infection 40 (unpublished data). In contrast, our results here indicate that guigna infection patterns with mycoplasma species were not associated with human-modified landscapes, but instead, was significantly linked with pristine continuous forests. This indicates that modes of transmission of hemoplasma species and FeLV/FIV are different. Further effort is needed to document the ecological interactions among guigna and other forest-dwelling wild felids and carnivores (e.g. Darwin's fox), which are currently unknown in southern Chile. Understanding multi-host pathogens, including hemoplasma species, which can be transmitted among numerous host species and via diverse transmission pathways, has critical implications for host species management strategies. Future studies should therefore focus on clarifying hemoplasma transmission patterns, characterizing the role of native carnivore species have in transmission networks, and documenting the pathogenic significance of feline hemoplasma infections in wild felid populations.

Data availability
Derived data supporting the findings of this study are available from the corresponding author upon request.