Leishmania enriettii visceralises in the trachea, lungs, and spleen of Cavia porcellus

BACKGROUND Leishmania (Mundinia) enriettii is a species commonly found in the guinea pig, Cavia porcellus. Although it is a dermotropic species, there is still an uncertainty regarding its ability to visceralise during Leishmania life cycle. OBJECTIVE Here, we investigated the ability of L. enriettii (strain L88) to visceralise in lungs, trachea, spleen, and liver of C. porcellus, its natural vertebrate host. METHODS Animals were infected sub-cutaneously in the nose and followed for 12 weeks using histological (hematoxilin-eosin) and molecular tools (polymerase chain reaction-restriction fragment length polymorphism - PCR-RFLP). To isolate parasite from C. porcellus, animals were experimentally infected for viscera removal and PCR typing targeting hsp70 gene. FINDINGS Histological analysis revealed intense and diffuse inflammation with the presence of amastigotes in the trachea, lung, and spleen up to 12 weeks post-infection (PI). Molecular analysis of paraffin-embedded tissues detected parasite DNA in the trachea and spleen between the 4th and 8th weeks PI. At the 12th PI, no parasite DNA was detected in any of the organs. To confirm that the spleen could serve as a temporary site for L. enriettii, we performed additional in vivo experiments. During 6th week PI, the parasite was isolated from the spleen confirming previous histopathological and PCR observations. MAIN CONCLUSION Leishmania enriettii (strain L88) was able to visceralise in the trachea, lung, and spleen of C. porcellus.

several L. enriettii strains in the State of Paraná, Brazil. It did not detect the parasite in the wild reservoirs reinforcing the role of C. porcellus as a the main, if not the only, domestic host. (14) Several research groups have been studied L. enriettii to understand its biology and some aspects of hostparasite interaction. (15,16) A distinguished feature of this species (unlike other Mundinia members) is its ability to specifically infect guinea pigs. (17) In C. porcellus, L. enriettii causes a severe ulcerated cutaneous lesion (CL) that heals over time. For this reason, C. porcellus was used as a model for CL for many years. (18,19) Experimental infection of guinea pigs by L. enriettii did not use salivary glands extracts (SGE) until recent biological and immunopathological studies. (20,21) Lutzomyia longipalpis SGEs modulated the infection by preventing the attraction of monocytes and CD163 macrophages throughout the course of infection. After 12 weeks postinfection (PI) regardless the use of SGEs, the CL lesions healed spontaneously. Early and recent reports (14,18,21,22,23) already demonstrated the ability of L. enriettii to visceralise in guinea pigs after several inoculation routes. It was, however, unknown if this phenomenon is permanent or transient requiring fast transmission to vector(s).
Over the 75 years since the discovery of L. enriettii, there are several aspects of infection that still need to be better elucidated. In our previous study, we reported the proinflammatory features of L. enriettii infection in the skin. (20) After the removal of the skin for immunopatho-logical studies, we also collected trachea, lungs, liver, and spleen in the same PI time intervals. Here, as part of a wider study of L. enriettii, we provide a more detailed study on the immunopathological aspects in different organs using molecular and histological approaches in C. porcellus viscera.
In vivo experiments -For visceralisation confirmation, six experimental animals were divided in two groups including: (1) two negative controls inoculated with (PBS+SGE) and (2) four animals infected with L. enriettii. Guinea pigs were infected intradermally with 10 5 L. enriettii promastigotes in PBS + SGE from L. longipalpis as previously reported. (20) After six weeks PI, tissue fragments were recovered for parasite isolation.
Parasite recovery and typing -Tissue fragments from trachea, lungs, liver, and spleen were seeded in NNN-Schneider's medium and incubated at 26ºC. Cultures were followed until appearance of promastigote forms. Positive cultures were expanded for DNA extraction and polymerase chain reaction (PCR) typing using hsp70 gene as above. The ~1,300-bp fragment was subjected to digestion with HaeIII (New England) and restriction profiles were analysed in a 4% agarose gel. (25) Ethical considerations -This work was approved by the Ethics Committee for Animal Use (CEUA), Oswaldo Cruz Foundation (FIOCRUZ) (LW-24/19). Young male C. porcellus were bred according to the International Standards for the Breeding and Use of Laboratory Animals, (26) obtained from the Institute of Science and Technology in Biomodels (ICTB/Fiocruz).

RESULTS
Histopathological evaluation -Histopathological changes in the trachea, lung, spleen, and liver of infected animals were analysed at the 4th, 8th, and 12th weeks PI using HE. Negative controls with uninfected animals did not show any histological alterations ( Fig. 1A-D). Based on our previous work, the cutaneous lesions peak at 6th-8th weeks PI. (21) For this reason, we chose representative images at 8th and 12th weeks PI (Figs 2-5). In the mucosa and submucosal region, trachea showed intense and diffuse mononuclear inflammatory infiltrate with amastigotes inside macrophages ( Fig. 2A-B). In the lungs, the main lesion observed was a diffuse and intense chronic inflammatory reaction with granuloma formation char-acterising an interstitial pneumonitis (Fig. 2C). In fact, the alveolar wall was thickened because of the presence of the chronic exudates of plasma cells, lymphocytes, and several parasitised macrophages with numerous amastigotes of L. enriettii (Fig. 2D).
In the spleen, the histological alterations included thickening and inflammation of the capsule and, hyperplasia/hypertrophy of the white and red pulp (Fig. 3A). The hyperplasia and hypertrophy were characterised by numerous enlarged macrophages usually parasitised with L. enriettii amastigotes. In parallel, the sinusoid vessels of the red pulp were always enlarged and congested (Fig.  3B). Histopathological alterations and amastigote forms were not documented in the liver ( Fig. 3C-D).
At 12 weeks PI, most of the histopathological features in the organs persisted. Trachea, lungs, and spleen still had inflammation with productive proinflammatory infiltrate and parasite presence, whereas liver did not change (Figs 4-5). However, an interesting feature was shown in the lungs (Fig. 4B). Although pneumonitis was still evident, a decrease in the granulomas was noticed. Leishmania enrietti DNA detection in C. porcellus viscera -The expected ~1,300-bp fragment was detected in the positive control (PC) (L88 strain of L. enriettii), trachea (Tra) (4th week PI) and spleen (4th and 8th weeks PI). At 12 weeks PI, no DNA was detected in the organs (Fig. 6).
Experimental infection of C. porcellus with L. enriettii -Confirming our previous findings, (20,21) L88 strain successfully infected C. porcellus showing cutaneous lesion development and swelling of the nasal plane at 4th and 6th weeks PI (Fig. 7A-B). Macroscopically, the cutaneous lesion showed an expected nodular protuberance with alopecia and ulceration common to L. enriettii. As expected, control animals did not develop any cutaneous lesions (Fig. 7C).
Parasitological and molecular confirmation of L. enriettii visceralisation in C. porcellus -After 18 days, spleen-derived cultures revealed presence of promastigotes from one animal (Fig. 8). As expected, the ~1,300 bp fragment was amplified from both samples (Fig.  9A). After treatment with HaeIII, their profiles were the same (Fig. 9B).

DISCUSSION
Histopathological and molecular studies involving species from the subgenus Mundinia have so far been restricted to the dermis region. (3,17,20,21) However, an unknown aspect of the infection outcome is its progression to the parasitological cure or latency. In this work we carried out a more detailed analyses in the viscera of C. porcellus infected with L. enriettii to corroborate findings of its visceralisation. In the previous papers, the ability of L. enriettii to infect several organs, including spleen, liver, adrenal glands, bone marrow, lymphatic ganglia and even blood, was reported. (14,18,19,23) Here, we reported for the first time the infection in trachea and lungs. By the time abovementioned papers were published, molecular tools were not available, and the techniques were mainly restricted to parasitological and microscopical methods. Consistent with previous observations, we detected amastigote forms of L. enriettii in trachea, lung, and spleen. Our histopathological findings showed intense inflammatory processes in those organs, except for liver that did not show any changes. Lungs were the most affected organs, where the inflammatory profile was more exacerbated with the presence of numerous granulomas and interstitial pneumonitis. However, they disappeared at 12th weeks PI, suggesting a healing process. This atypical alteration in this organ (granulomes) was similar to those caused by typical viscerotropic species in hamsters' liver. (27,28) This is different from dogs naturally infected with L. infantum, where only chronic interstitial pneumonitis with notable thickening of the interalveolar septa were noted. (29) Although there are evident pneumonia associated with granuloma and pyogranulomas, the findings of amastigotes in dogs lungs are very scarce. This is different from guinea pigs, where it was possible to visualise amastigotes in several areas. An interesting feature of this work is the apparent absence of amastigotes and/or pro-inflammatory processes in the liver. On the other hand, those were detected in the spleen during the histological assessment with noticeable intracellular amastigote forms in the parenchyma, especially inside macrophages. Altogether, our data indicate that L. enriettii can visceralise and cause immunopathological events in trachea, lungs, and spleen up to 12 weeks PI. While our previous paper (21) demonstrated that the infection in the skin was completely cured at 12th PI, it remained active in the viscera.
To confirm this, paraffin-embedded tissues were assayed for the presence of parasite DNA using molecular tools. Recently, hsp70 gene has been successfully used as a tool for discriminating L. martiniquensis and L. orientalis. (30) Here, we amplified the specific fragment of this target (~1,300-bp) from DNA extracted from paraffin-embedded tissues. (25) This has allowed us to trace movement of L. enriettii from trachea to spleen in the first 8 weeks PI. After two months, it disappears from the trachea and increases its presence in the spleen. No parasite DNA was found at 12 weeks PI suggesting a possible clearance of the parasite in the organs. However, this result is conflicting with the histopathological findings that showed amastigote forms in the tissues  during this period. This also does not correlate with what occurs in the skin of the same animals, whose lesions reached their peak at 6 weeks PI and are completely healed at 12 weeks. (20,21) Confirming our previous his-tological findings, no DNA was found in the liver in all time points analysed. However, it was also not detected in the lungs by PCR, where amastigotes and a severe pro-inflammatory milieu were seen, implying sensi- tivity limits of PCR detection. Interestingly, our histopathological analysis showed the presence of tracheal, lung and spleen amastigotes in all analysed weeks, including the 12th week PI. Although those techniques may be complementary, inconsistencies are expected since DNA extraction from paraffin-embedded tissues is laborious and may affect DNA quality, especially at 12th weeks PI, where fewer parasites were seen. This is different from previous studies, where detection of parasite DNA in paraffin blocks with virtual absence of amastigotes was achieved. (29,31) This reinforces the need of using different methods for description of parasitological events in the hosts.
Based on our data, it is likely that L. enriettii can reach the viscera, cause severe histopathological alterations that decrease, but not completely disappear, for example, in the skin 8-12-weeks PI. It is interesting to notice that under our controlled experimental conditions the animals were apparently active and did not have any clinical signs. (20,21) However, this is different from previous studies with infected field-collected guinea pigs, where the severity of the lesions seemed to have an impact in those animals. (14,22) We suggest that in nature a higher severity of the lesions may provide more infective blood meal to the vector. However, a proven vector for L. enriettii is yet to be determined. (15)    Leishmaniasis comprises a wide spectrum of clinical manifestations including diseases caused by dermotropic and viscerotropic species. Depending on the region, intraspecies tropism variations may occur. For example, Leishmania infantum (viscerotropic) can cause cutaneous lesions in Honduras, Central America. (32) Leishmania donovani, another viscerotropic species, can cause unusual dermal lesions after therapeutic failure known as Post-kala-azar dermal Leishmaniasis (PKDL). (33) The opposite is also true for Leishmania amazonensis (dermotropic). It was isolated from lymph nodes from dogs showing symptoms of canine visceral leishmaniasis, similar to that caused by L. infantum, (34) and from rodents' viscera. (35) Finally, L. martiniquensis, another member of the Mundinia subgenus, causes not only LV, but also CL in HIV-positive patients. (36) Consistent with those observations, L. enriettii showed a dual profile being able to cause dermotropic and vis-cerotropic lesions under our experimental conditions in its main host C. porcellus.
To confirm our previous histological and molecular findings, next we performed an in vivo infection targeting the spleen as a possible source for parasite isolation and observed the same pattern of lesion development as reported elsewhere at four-eight weeks PI. (20,21) After euthanasia, we failed to isolate parasites from trachea, lungs, and liver fragments after six weeks PI. After 18 days, we succeed in detecting promastigotes in only one spleen fragment from one animal. After performing polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP), we confirmed that cultured promastigotes used for animal infection had the same profile as those, recovered from the spleen. This finding confirms that live parasites can be isolated from spleen.
Leishmania enriettii (strain L88) visceralised in C. porcellus causing severe lesions in lungs, trachea, and spleen. Despite the clinical cure of the skin lesion, the parasite was not eliminated from the tissues during the 12 weeks PI as judged by the histological studies. In contrast to the infected field-collected guinea pigs, L. enriettii under laboratory conditions did not cause any clinical sign to them. This finding is of epidemiological importance, since those animals may be potential sources for vector infection facilitating parasite maintenance and spreading in the environment. This phenomenon warrants further investigation to address the role of apparently cutaneous healed animals in L. enriettii transmission.

AUTHORS' CONTRIBUTION
RPS and WLT conceived and designed the study protocol; EVAS, LJP, LFP, ICF and PMP performed experiments; NFG provided SGE from Lutzomyia longipalpis for in vivo experiments; RPS, LJP, WT and MDL performed data analysis and interpretation; RPS and LJP drafted the manuscript; MDL contributed to critical revision of the manuscript. All authors approved the final version of the manuscript.