Leishmaniases of men and animals caused by L. (Mundinia) species are emerging all over the world. The diseases in humans are characterized by symptoms varying from self-healing skin lesions [2, 12] to visceral forms. The latter prevail in HIV-positive patients [8, 16] but was also observed in immunocompetent humans [12, 17]. Very little is known about the life-cycle of these ancient and neglected species and an appropriate animal model is necessary for closer understanding of their biology.
Guinea pigs were chosen for the experimental model in our study as they are the only known non-human mammalian hosts of L. (Mundinia) species, except for kangaroos, cows and horses [6, 7, 9, 10, 18], which are not practicable for most laboratory investigations. Leishmania enriettii was repeatedly isolated from domestic guinea pigs from various localities in Brazil [6, 19]. Interestingly, individual cases were separated by long time periods, which does not agree with the fact that guinea pigs are popular pets, and according to several studies, they are very susceptible to infection [5, 20, 21]. We suggest that this rare incidence may have two different explanations. First, the prevalence of infection is actually much higher, but the owners of infected guinea pigs do not take them for veterinary checks, therefore, parasites are not isolated. Alternatively, guinea pigs are only incidental hosts and the primary reservoir hosts (and primary insect vectors) are not present in close vicinity to households. In this case, secondary vectors and/or reservoirs may be temporarily involved in transmission to domestic localities and domestic guinea pigs.
Our experiments confirmed the susceptibility of guinea pigs to L. enriettii. All infected animals showed development of typical ear lesions and the animals were infectious to sand flies. The numbers of positive sand flies were significantly higher at week 4 pi than at the later time interval, week 8 pi. This decrease of infectivity was also observed previously by Seblova et al. . At week 12 pi, the animals did not show any more external signs of infection and Leishmania DNA was not detected in any of the examined tissue samples. Spontaneous healing of lesions was observed also by Paranaiba et al. . In their experiments initiated by intradermal inoculation of 105 promastigotes, a different virulence between the two strains used was observed. The Cobaia strain did not develop any lesions, while strain L88 developed lesions that were growing by weeks 4–6 pi, and then diminution of lesions was observed until the end of the experiment. The authors also described development of larger lesions in groups where sand fly salivary glands were added to the inoculum.
However, the virulence of the L. enriettii parasite strain and the presence of sand fly salivary glands are not the sole factors influencing the degree of pathogenicity for guinea pigs. The outcome of infections is also dependent on the method of their initiation, i.e. on parasite numbers and stages (amastigotes vs promastigotes) used as an inoculum. Thomaz-Soccol et al.  described the development of serious symptoms of disease, such as the dissemination of parasites and subsequent death of all tested animals, when the inoculum consisted of amastigotes of a strain identical to L88 (according to isoenzyme analyses). Wide dissemination of L. enriettii in animals was also observed by Paraense et al. , who infected guinea pigs with amastigotes from lesion homogenates, and by Seblova et al.  who infected animals with 107 culture derived promastigotes.
Development of L. enriettii has also been tested in hamsters (Mesocricetus auratus), where infections were characterized by the development of temporary lesions at the site of inoculation and their subsequent healing [5, 24]. In experiments with wild guinea pigs (Cavia aperea), rhesus macaques and dogs , no animals showed any signs of infection, so domestic guinea pigs remain the best laboratory model for L. enriettii at present.
Here, we compared the susceptibility of guinea pigs to four other L. (Mundinia) species. The infections were lost in animals infected with L. macropodum, Leishmania sp. from Ghana and L. martiniquensis strain CU1. Animals infected with a second L. martiniquensis strain, MAR1, and with L. orientalis developed only temporary changes on the ears and the animals were not infectious to sand flies. However, PCR analysis showed no presence of leishmanial DNA by week 12 pi in any of the tested samples. We suggest that L. martiniquensis and L. orientalis are capable of temporary survival at the site of inoculation, but they cannot disseminate to other tissues of guinea pigs.
We suggest that for a better understanding of L. (Mundinia) biology it is necessary to focus on other model host organisms. The first choice could be BALB/c mice or hamsters as the most common animal models for research used with many L. (Leishmania) and L. (Viannia) species. Infections of these standard laboratory animals with their controlled genetic background may bring valuable information. Alternatively, genetically polymorphic models like wild rodents mimicking natural hosts could be used. These less common models can allow a better understanding of the dynamics of infection and host-parasite relationships related more closely to the situation in the wild . On the other hand, when infected with L. enriettii and L. orientalis, guinea pigs could serve as a potential model for spontaneous healing, which could be informative for the design of vaccines.