Multilocus microsatellite typing (MLMT) of strains from Turkey and Cyprus reveals a novel monophyletic L. donovani sensu lato group

Abstract

Background: New foci of human CL caused by strains of the Leishmania donovani (L. donovani) complex have been recently described in Cyprus and the Çukurova region in Turkey (L. infantum) situated 150 km north of Cyprus. Cypriot strains were typed by Multilocus Enzyme Electrophoresis (MLEE) using the Montpellier (MON) system as L. donovani zymodeme MON-37. However, multilocus microsatellite typing (MLMT) has shown that this zymodeme is paraphyletic; composed of distantly related genetic subgroups of different geographical origin. Consequently the origin of the Cypriot strains remained enigmatic.

Methodology/principal findings: The Cypriot strains were compared with a set of Turkish isolates obtained from a CL patient and sand fly vectors in south-east Turkey (Çukurova region; CUK strains) and from a VL patient in the south-west (Kuşadasi; EP59 strain). These Turkish strains were initially analyzed using the K26-PCR assay that discriminates MON-1 strains by their amplicon size. In line with previous DNA-based data, the strains were inferred to the L. donovani complex and characterized as non MON-1. For these strains MLEE typing revealed two novel zymodemes; L. donovani MON-309 (CUK strains) and MON-308 (EP59). A population genetic analysis of the Turkish isolates was performed using 14 hyper-variable microsatellite loci. The genotypic profiles of 68 previously analyzed L. donovani complex strains from major endemic regions were included for comparison. Population structures were inferred by combination of bayesian model-based and distance-based approaches. MLMT placed the Turkish and Cypriot strains in a subclade of a newly discovered, genetically distinct L. infantum monophyletic group, suggesting that the Cypriot strains may originate from Turkey.

Conclusion: The discovery of a genetically distinct L. infantum monophyletic group in the south-eastern Mediterranean stresses the importance of species genetic characterization towards better understanding, monitoring and controlling the spread of leishmaniasis in this region.

Figure 1. K26-PCR confirms that CUK1, CUK2 and EP59 strains belong to the L. donovani complex.

Lanes: M, 100 bp DNA ladder; CH35 (MON-37, 700 bp); PM1 (MON-1, 626 bp); SC23 (MON-38, 515 bp); CUK1 (MON-309, 480 bp); CUK2 (MON-309, 480 bp); HUSSEN (MON-31, 430 bp); EP59 (MON-308, 385 bp); BUCK (MON-78, 385 bp). Strains CUK3, CUK4, CUK7 and CUK 10 gave identical results to CUK1 and CUK2 (not shown).

Figure 2. Estimated population structure of the 76 L. donovani s.l. strains as inferred by STRUCTURE.

MLMT profiles are based on variations in 14 microsatellite markers. Each strain is represented by a single vertical line divided into K colours, where K is the number of populations assumed. Each colour represents one population and the length of the coloured segments shows the strain's estimated proportion of membership in that population. Strains with mixed memberships to the different populations are represented by different coloured segments in the vertical bar, which are proportional to the membership coefficient. A. When the likelihood of population number is calculated according to Evanno et al. , the derived graph for ΔK shows a peak at K = 2 indicating the existence of two main populations in the studied strain set. However, eight populations are observed based on a log-likelihood plot, which plateaus at K = 8. At K = 8, all Turkish non MON-1 strains group with MON-37 strains from Cyprus, forming the TR/CY non MON-1 population. Bar plots for K = 3 and K = 4 are also shown to help determine ancestral populations. B. When the L. infantum non MON-1 & TR/CY non MON-1 strains are re-analysed separately the log-likelihood plot plateaus at K = 5, while the ΔΚ graph shows a major peak at K = 2 and a minor one at K = 5. At K = 5 a split of the TR/CY non MON-1 population was observed.

Figure 3. Midpoint rooted Neighbor-joining tree for the 76 L. donovani complex strains analysed.

This midpoint rooted tree was inferred from Dps-distances calculated for the MLMT data (14 microsatellite markers) of our sample set. The tree leaves are coloured according to the populations defined by structure analysis (Fig. 2). Bootstrap values (1000 re-samplings) above 50% are indicated at key nodes. The EP59 strain was excluded from this analysis due to its hybrid profile.

Figure 4. Factorial correspondence analysis (FCA) of the 76 L. donovani complex strains studied.

CY, Cyprus; ET, Ethiopia; GR, Greece; IN, India; KE, Kenya; LK, Sri Lanka; PT, Portugal; SD, Sudan; SP, Spain; TR, Turkey; Populations designated as MON-1 (red squares) and non MON-1 (orange and brown squares) include respectively L. infantum MON-1 strains from Greece, Turkey, Spain and Portugal and L. infantum non MON-1 strains from Spain, Portugal, France and Italy, corresponding to previous analyses , ; populations designated as IN1 (India 1), IN3 (India 3) and LK (Sri Lanka) (yellow squares); SD/ET (Sudan/Ethiopia) (pink squares); KE/IN2 (Kenya/India 2) (blue squares) compose the L. donovani genetic group, as detected previously , , , ; L. infantum non MON-1 strains from Turkey and Cyprus are designated as TR (dark purple squares) and CY (light purple squares), respectively. The MON-37 clone of CD44 strain from Cyprus groups with the other CY non MON-1 strains. The TR/CY strains are placed between the L. infantum MON-1 and non MON-1 populations and the hybrid strain EP59 from Turkey between the TR non MON-1 and MON-1 populations. The BUCK strain is placed very close to the TR non MON-1 strains, as previously described , .