Revisiting epidemiology of leishmaniasis in central Asia: lessons learnt

Abstract Abstract In this work we reviewed historical and recent data on Leishmania spp. infection combining data collected in Turkmenistan, Uzbekistan, Kazakhstan, Kyrgyzstan, Iran, China and Mongolia. We specifically focused on a complex of co-existing species (Leishmania major, Leishmania turanica and Leishmania gerbilli) sharing the same animal reservoirs and vectors. In addition, we analysed the presence of dsRNA viruses in these species and discussed future research directions to identify species-specific traits, which may determine susceptibility of different Leishmania spp. to viral infection.


Introduction
Leishmaniasis is one of the neglected vector-borne tropical diseases endemic in almost 100 countries worldwide caused by Leishmania spp. (Euglenozoa: Trypanosomatidae) (Bruschi and Gradoni, 2018;Kostygov et al., 2021b). Between 10 and 15 million people in the world are infected, and the annual rate of new infections is over 2 million cases (WHO, 2022). The mortality per year from leishmaniasis is second only to malaria among all parasitic diseases (Pace, 2014). Its clinical manifestations range from cutaneous ulcers to systemic multiorgan diseases in the cases of cutaneous leishmaniasis (CL) and visceral leishmaniasis (VL), respectively (Bruschi and Gradoni, 2018).
In the Old World, the main areas of CL circulation are northern Africa, central Asia (hereafter, Kazakhstan, Kyrgyzstan, Mongolia, Turkmenistan and Uzbekistan) and the Middle East (Alvar et al., 2012;Torres-Guerrero et al., 2017). The most common Leishmania spp. documented in the Old World are Leishmania aethiopica, Leishmania major, Leishmania tropica and species of the Leishmania donovani complex (L. donovani and Leishmania infantum) (Lukeš et al., 2007;Bruschi and Gradoni, 2018). The VL in humans is mainly caused by members of the L. donovani complex and may manifest in damages to the liver, spleen, lymph nodes and bone marrow often resulting in death of a patient, if not diagnosed and treated in a timely manner (Strelkova et al., 2015;Mann et al., 2021). Infection with these species may also present skin manifestations in the cases of post-kala-azar dermal leishmaniasis or atypical leishmaniases (Guan et al., 2013;Zhang et al., 2014;Ben-Shimol et al., 2016;Zijlstra, 2016). Leishmania aethiopica, L. major and L. tropica mostly cause CL, although some isolates of L. major and L. tropica were occasionally identified from patients with VL (Alborzi et al., 2008;Bruschi and Gradoni, 2018;Charyyeva et al., 2021). The CL can be further subdivided into anthroponotic (ACL) and zoonotic (ZCL) forms, which are predominantly caused by L. tropica and L. major, respectively (Akilov et al., 2007;Ghatee et al., 2020). Great gerbils (Rhombomys opimus) and fat sand rats (Psammomys obesus) serve as the main animal reservoirs for ZCL in central Asia and the Middle East, correspondingly (Elfari et al., 2005;Akhavan et al., 2010c), although other animal speciesfor example, Libyan jird (Meriones libycus), Shaw's jird (Meriones shawi), Indian gerbil (Tatera indica) or Indian desert gerbil (Meriones hurrianae)may play this role in particular geographic regions (Yaghoobi-Ershadi et al., 1996;Rassi et al., 2001;Mohebali et al., 2004;Ghawar et al., 2011;Akhoundi et al., 2013).
The great gerbils may simultaneously host several species of Leishmania. In addition to pathogenic to humans L. major, they may also be infected by gerbil-restricted Leishmania turanica and Leishmania gerbilli (Strelkova et al., 1990b(Strelkova et al., , 2001Akhavan et al., 2010b). Here, we reviewed the literature on the mixed infections of L. major, L. turanica and L. gerbilli in central Asia and neighbouring countries with a focus on their natural animal reservoirs. As a second aim, we wanted to highlight some important papers on this topic published in Russian, and, as such, not well-known to the researchers in other countries.

Historical notes
Parasites of the genus Leishmania were first formally described by Leishman and Donovan in 1903 in patients infected with kala-azar in India (Donovan, 1903;Leishman, 1903).
The parasite causing tropical ulcer was described as Helcosoma tropicum by Wright in the same year (Wright, 1903) and renamed as Leishmania tropica in 1906 by Lühe (1906). Yet, the first scientist documented the presence of a parasite now known as L. tropica was Cunningham in 1885 (Cunningham, 1885). Its protistan nature ('class of protozoa') was discovered by Borovsky (1898), but remained unrecognized until much later (Hoare, 1938). In 1914, Yakimov identified 2 variants of Leishmania sp., based on the size of amastigotes in the macrophages of patients and named them L. tropica minor and L. tropica major (Yakimov and Schokhor, 1914). Later studies revealed that L. t. minor causes dry ulcers usually lasting for over a year and it is more commonly spread in the cities. In contrast, L. t. major manifests in wet ulcers, the course of the disease is shorter and it is more commonly spread in rural areas (Latyshev and Krukova, 1941;Kozevnikov, 1963;Schnur, 1987). In 1973, summarizing the accumulated data, Bray proposed to reclassify parasites as L. tropica for the causative agent of ACL and L. major for the causative agent of ZCL (Bray et al., 1973).
It is important to note that for about 50 years all isolates of Leishmania coming from animals and people in central Asia and neighbouring countries with a characteristic clinical picture were classified as L. major, with the only exception being a description of another Leishmania sp., L. gerbilli, from R. opimus in 1964 in China (Wang et al., 1964). In line with that, in vitro experiments in animals demonstrated that different isolates have different levels of virulence. The highly virulent (HV) strains were invariably isolated from humans. They caused a progressive disease with obligatory ulceration in golden hamsters and domestic mice. Conversely, the strains with virulence ranging from low (low virulent strains, LV) to high could be isolated from gerbils. The LV strains caused a slow course of the disease that was limited to infiltrates and never led to ulceration. The strains with intermediate virulence caused a prolonged disease manifesting in small abortive ulcers in the later stages (Kellina, 1965;Lavrova et al., 1973;Kellina et al., 1981). Experimental infection of different animal species (in which Leishmania presence was documented in nature) with clonal cultures or strains of Leishmania with different virulence (HV and LV) revealed that HV parasites infected all the tested animalsgreat gerbils, Libyan jirds, Severtzov's jerboa (Allactaga severtzovi), long-eared hedgehogs (Hemiechinus auritus), domestic mice (Mus musculus) and golden hamsters (Mesocricetus auratus). In contrast, only the great gerbils, some Libyan jirds and golden hamsters could be infected by the LV clones or strains Strelkova et al., 1980). The mystery of strains with different virulence was solved only with an advent of molecular techniques in the late 1980s. The isoenzyme analysis revealed that the strains previously identified as L. major include 3 independent species -L. major sensu stricto, L. turanica and L. gerbilli (Strelkova, 1990;Strelkova et al., 1990b). These experiments also confirmed that HV and LV strains belonged to L. major and L. turanica/L. gerbilli, respectively. Notably, all strains isolated from humans were L. major, implying that L. turanica and L. gerbilli are restricted to gerbils. To sum up, all 3 abovementioned species can infect great gerbils, golden hamsters and Libyan jirds, but only L. major (neither L. turanica nor L. gerbilli) can infect Severtzov's jerboa, long-eared hedgehogs and domestic outbred mice.
Communal epidemiology and ecology of L. gerbilli, L. major and L. turanica

Ecology
In the natural foci of the ZCL on the territory of Turkmenistan, Uzbekistan, Kazakhstan, Kyrgyzstan and Mongolia great gerbils are the main natural hosts of Leishmania spp. discussed above (Eliseev and Kellina, 1963;Dubrovsky, 1978;Eliseev and Neronov, 1997;Bruschi and Gradoni, 2018) (Fig. 1, Table 1). Parasites' life cycle not involving R. opimus was shown in some rare cases (e.g. in the lower reaches of the Surkhandarya river it involves M. libycus) . Great gerbils form topical and, as a result, trophic relationships with sand flies ( , 1978)] may cohabitate with the greater gerbils for some time usually occupying the outer parts of the colony. This significantly reduces the likelihood and intensity of sand flies' feeding on them.

Hosts and vectors
In central Asian countries, the spectra of documented hosts and vectors vary for different Leishmania spp.: L. major has been isolated from humans, great gerbils, Libyan jirds, Phlebotomus papatasi and Phlebotomus andrejevi, while L. turanica and L. gerbilli have been found in great gerbils, P. papatasi, P. andrejevi, Phlebotomus caucasicus, Phlebotomus mongolensis, Phlebotomus alexandri, Sergentomyia clydei, and great gerbils and P. mongolensis, respectively (Strelkova, 1996). Out of all the identified sand fly species, only P. papatasi and P. mongolensis are anthropophilic (Killick-Kendrick, 1990;Guan et al., 1995). Notably, experimental coinfections of L. major and L. turanica in P. papatasi revealed that 2 parasite species do not outcompete each other and develop in parallel (Chajbullinova et al., 2012). In addition to (albeit, loose) vector specificity and host specificity, different Leishmania spp. inhabit somewhat different geographic areas. While L. turanica was found infecting R. opimus throughout its distribution areas (see Fig. 1 in Strelkova, 1996), L. major parasitizes great gerbils and Libyan jirds in more southern parts of their distribution areas predominantly in river valleys, oases, areas adjacent to the oases and foothill plains (Neronov et al., 1987;Strelkova et al., 1990a). Conversely, in the southern and southwestern areas of Iran not populated by great gerbils, the circulation of CL is unstable confirming the role of these animals as the main reservoirs (Neronov and Farang-Azad, 1973) (Table 1, see below). Leishmania gerbilli is not as widespread as the other 2 species discussed above and epizootics caused by this parasite are limited to the local populations of R. opimus in several regions of China, Iran, Kazakhstan, Turkmenistan and Uzbekistan (Strelkova, 1996;Strelkova et al., 2003).

Infection prevalence and seasonal dynamics
The prevalence of L. turanica infection in R. opimus is high (over 50%, sometimes achieving 100% in certain populations) and fairly stable over the years. The transmission season lasts from late April to mid-October and from late May to early September for southern and northern parts of the area, respectively; it usually peaks around late June-early July and remains stable until the end of transmission period. The prevalence of L. major in R. opimus is lower than that of L. turanica; it peaks in July-August achieving 10-50% in different years and places. The prevalence of this parasite in M. libycus (usually, within the R. opimus distribution range) does not exceed 3-4%. The prevalence of L. gerbilli infection in R. opimus is even lower than that of L. major (Strelkova, 1996).

Coinfections, sympatric and allopatric populations
A remarkably high proportion of great gerbils in central Asia is infected by more than 1 Leishmania spp.; occasionally, all 3 species have been documented in the same animal (Strelkova, 1990;Strelkova et al., 1990a) (Fig. 1, Table 1). Populations of L. gerbilli, L. major and L. turanica are sympatric in some areas of R. opimus distribution (Strelkova et al., 2003). The human ZCL caused by L. major in these areas always occurs on the background of L. turanica or, more rarely, L. gerbilli infection. It has been demonstrated that coinfection with L. turanica facilitates L. major perseverance in rodents during the 6-10 months break in the sand fliesmediated transmission of these parasites in central Asia (Strelkova et al., 2001). In contrast, the populations of L. turanica in northern Kazakhstan and Mongolia are allopatric, as the Iran-Turanian part of the ZCL range is geographically isolated from the central Asian one (Shurkhal et al., 1985;Neronov et al., 1987). The dynamics of parasite presence in sympatric populations was investigated in the cases of predominant L. turanica and subsidiary L. major in great gerbils in central Asia (Eliseev et al., 1991; Strelkova et al., 1993Strelkova et al., , 2003. Three main types of natural foci are described in R. opimus (Lysenko and Beljaev, 1987). The hypoendemic foci are characterized by the very low-density circulation of L. major; here L. turanica predominates throughout the year, often for several years in a row. As a consequence, human ZCL in these foci is rare. In the hyperendemic foci, the proportion of L. major goes up during the transmission period (May-September), in some cases achieving 50% by the end of the season. This results in a substantially higher incidence of human ZCL. Leishmania major in the mesoendemic foci may accumulate up to 50% prevalence in R. opimus for several years, while being virtually absent in certain years in-between (Eliseev et al., 1991;Strelkova, 1996). Coinfections with L. turanica and L. major in great gerbils appear to be evolutionarily beneficial over the single-species infections. Indeed, while individual experimental infections with L. major, L. turanica and L. gerbilli lasted for 7, 15 and 18 months, respectively, the coinfection of L. major and L. turanica lasted for over 25 months (Strelkova, 1991), indirectly confirming an early observation that duration of Leishmania spp. infection may be comparable with the life time of a host, R. opimus (Shishliaeva-Matova et al., 1966). While infections with L. major alone were mostly self-healing, those with L. gerbilli, L. turanica, or coinfections with L. major and L. turanica led to chronic diseases in vast majority of cases (Strelkova, 1991).
The situation in neighbouring countries (particularly, Iran and China) is somewhat different and deserves special discussion.

Hosts and vectors
The animal hosts of L. major are R. opimus in the north-eastern and central Iran, M. libycus in the south-western and central regions, T. indica in the south-west, west and south of the country and M. hurrianae in the south-east (Mohebali et al., 2004;Akhoundi et al., 2013). The presence of L. turanica was documented in great gerbils from the central and north-eastern Iran (Mohebali et al., 2004;Mirzaei et al., 2011), M. libycus from Fars and Esfahan provinces (Akhoundi et al., 2013;Asl et al., 2022), Rattus norvegicus and T. indica in Bushehr province (Yaghoobi-Ershadi et al., 2013), and short-tailed bandicoot rats (Nesokia indica) in the western province of Kermanshah (Hajjaran et al., 2009) (Fig. 1, Table 1).

Infection prevalence, seasonal dynamics and coinfections
The reported prevalence of single infections and coinfections with L. major and L. turanica in great gerbils varied greatly between different studies (Akhavan et al., 2010b(Akhavan et al., , 2010cAkhoundi et al., 2013;Hajjaran et al., 2013;Mirzaei et al., 2014;Asl et al., 2022), indicating that more studies are needed to address this question. Leishmania gerbilli was previously detected only in the Esfahan and Bushehr provinces in single infections and coinfections with L. major, L. turanica or both, but usually in a small fraction of animals (Mirzaei et al., 2011;Yaghoobi-Ershadi et al., 2013). A recent study reported a higher frequency of L. gerbilli in the central Iran: out of 162 R. opimus and M. libycus analysed, 28 and 43 were infected by L. gerbilli alone and L. gerbilli with L. turanica, respectively (Asl et al., 2022). The seasonal dynamics of R. opimus infection in Iran also differs from that described for the former USSR: while single L. major infections were observed in autumn and winter, and coinfections with L. major and L. turanica in all seasons except for summer, the single L. turanica infections were present throughout the year (Akhavan et al., 2010a(Akhavan et al., , 2010c. The proven vectors of L. major in this country include P. papatasi, Phlebotomus salehi, P. caucasicus and (suspected) Phlebotomus ansari (Azizi et al., 2012; Yaghoobi-Ershadi, 2012; Maroli et al., 2013;Rafizadeh et al., 2016). The presence of L. major, L. turanica, and L. gerbilli in Iranian sand flies (P. papatasi, P. caucasicus, P. mongolensis, P. salehi, Phlebotomus sergenti, P. ansari, P. alexandri and Sergentomyia sintoni) was reported mostly from north-eastern and central regions Bakhshi et al., 2013;Bordbar and Parvizi, 2014a;Sharbatkhori et al., 2014;Rafizadeh et al., 2016). Coinfections with L. major and L. turanica were rare and documented only in P. papatasi Darvishi et al., 2015;Rafizadeh et al., 2016). Coinfections with L. turanica and L. gerbilli were seen in P. papatasi and P. caucasicus (Bakhshi et al., 2013;Darvishi et al., 2015). Overall, P. papatasi is the species, in which L. turanica has been found most frequently in Iran, both in single and coinfections with L. major or L. gerbilli. As its competence to support the development of L. turanica has also been demonstrated in the laboratory (Chajbullinova et al., 2012), P. papatasi is probably the main vector of L. turanica in Iran. Its vector competence for L. gerbilli has yet to be established.

China
In China, both VL and CL leishmaniasis are endemic in the western and north-western areas, with a predominance of VL and rather rare cases of CL (Wang et al., 2010). Leishmania major in these regions has not been documented in rodents or humans, and human CL is caused by L. infantum transmitted by Phlebotomus major wui (Guan et al., 2013). Leishmania gerbilli has been described from desert areas inhabited by R. opimus since the 1960s (Wang et al., 1964), and L. turanica was first identified in R. opimus in mid-1990s in Xinjiang (Guan et al., 1995). Phlebotomus mongolensis and P. andrejevi are considered the main vectors; both are ecologically connected to R. opimus (Table 1). While P. andrejevi is scarcely captured in residential quarters, P. mongolensis is anthropophilic and dominates in human baits. In theory, reptilian reservoirs could also be involved in the circulation of L. turanica in the arid desert areas of northwestern China, as this parasite has been found in 4 and 2 species of the genera Phrynocephalus and Eremias, respectively (Zhang et al., 2019).
In summary, the division of previously recognized as a singlespecies L. major into 3 Leishmania spp., of which only 1 is pathogenic for humans, facilitated revision of the earlier concepts about the epidemiology and endemicity of L. major in the territory that was considered well-studied.

Pathogenicity of L. turanica infections for humans
In Iran, asymptomatic Leishmania infections prevail in the natural rodent hosts: approximately 90% of the infected R. opimus showed no cutaneous lesions on their earlobes (Akhavan et al., 2010b(Akhavan et al., , 2010cAkhoundi et al., 2013). Notably, among symptomatic R. opimus, both L. major and L. turanica were detected but not scrutinized further (Hajjaran et al., 2013). In line with these, single infections with L. turanica can cause lesions in humans too, as was demonstrated in a single study in northern Iran (Bordbar and Parvizi, 2014b). Subcutaneous inoculation of human volunteers with 2 strains of L. turanica (1 from a Mongolian great gerbil and 1 from a sand fly from Uzbekistan) resulted in mild cutaneous manifestations. The Mongolian isolate infection manifested in a small nodule, which persisted for 2 weeks and resolved with no ulceration. Infection with a strain originated from a sand fly led to a prolonged incubation period and ulceration. The lesions persisted for about 2.5 months and healed spontaneously leaving small scars (Strelkova et al., 1990b). In China, 2 healthy volunteers injected with L. turanica also had mild cutaneous manifestations (Guan et al., 1995).
Collectively, L. turanica, a species of Leishmania might be categorized better as being low in pathogenicity, rather than being non-pathogenic in humans. Further studies are warranted to clarify this issue, as the number of volunteers involved in experimental infections was too small to generalize.

Genomics
Genomes of all Leishmania spp. discussed above have been sequenced. The iconic one, L. major MHOM/IL/81/Friedlin had been sequenced earlier in 2005  and since then was used as a reference in numerous comparative studies (El-Sayed et al., 2005;Peacock et al., 2007;Lukeš et al., 2018;Zakharova et al., 2022). The genomes of L. turanica MRHO/ SU/65/VL (LEM423) and L. gerbilli MRHO/CN/60/GERBILLI (LEM452) are also available, but they have not been scrupulously analysed yet (Warren et al., 2021). It goes without saying that more strains of these Leishmania spp. need to be sequenced and analysed in order to shed light on molecular mechanisms driving speciation in allopatric and sympatric populations of Leishmania spp., as well as the details defining hypoendemic, mesoendemic and hyperendemic foci of this parasite in central Asia.

Presence of leishmaniaviruses
Another interesting and very important topic of investigation is interaction between Leishmania spp. and their endosymbiotic viruses. The fact that viruses can infect Leishmania was established half-a-century ago (Molyneux, 1974), but the real breakthrough came only after about 40 years of active investigation (Tarr et al., 1988;Stuart et al., 1992;Widmer and Dooley, 1995;Grybchuk et al., 2018a), when it was demonstrated that the presence of Leishmania RNA virus 1 (Leishmaniavirus 1, LRV-1) in Leishmania guyanensis controls the severity of mucocutaneous leishmaniasis (Ives et al., 2011). The LRV-1 and LRV-2 are restricted to the New and Old World Leishmania spp., respectively and generally co-evolve with their hosts (Scheffter et al., 1995;Widmer and Dooley, 1995;Cantanhêde et al., 2021;Kostygov et al., 2021a). Interestingly, their elimination from L. major and L. guyanensis in vitro prompts different cellular responses . Our previous analyses of the available central Asian strains of L. gerbilli (n = 2), L. major (n = 14) and L. turanica . The prevalence of L. major infection by LRV-2 in these studies varied between 2 and 68% and the presence of LRV-2 was confirmed in 87 out of 221 isolates of L. major analysed (only 1 isolate of L. turanica was included and reported as negative and no isolates of L. gerbilli were analysed). These findings also pose an intriguing questionwhy L. turanica and L. gerbilli (note that conclusion about L. gerbilli is based on analysis of a very few isolates) are not infected with LRV-2? This is unlikely to be a virus-related feature, as LRVs were shown to be able to cross species barriers and even infect trypanosomatids of other genera (Grybchuk et al., 2018b). This is also unlikely to be determined by the vector, because all 3 Leishmania spp. are transmitted by the same or closely related species of sand flies (Strelkova, 1996). The most parsimonious explanation is that L. gerbilli and L. turanica are naturally resistant against LRV-2. If this is really the case, analysis of additional strains of L. gerbilli and L. turanica will help us to discover genetic elements making them resilient to viral infections possibly providing new tools for treatment of leishmaniases. It is also important to consider how LRVs are transmitted. While to the best of our knowledge no experimental studies were performed on LRV-2s, in the cases of LRV-1 and New World Leishmania spp. it has been shown that viruses are transmitted horizontally via extracellular vesicles (Atayde et al., 2019;Olivier and Zamboni, 2020).
In conclusion, we are now on the verge of new and exciting findings fuelled by recent development in sequencing technologies that may dramatically expand our knowledge about parasites causing leishmaniasis and change the way this disease is diagnosed and treated.

Author's contributions
Writing and revisionall authors. Conflict of interest. None.