Generally, DNA marker is now the most widely used method in identification and classification of Leishmania, since it is both effective and efficient. While different evolutionary rate of diverse gene markers may lead to different classification result. Thus, to help us understand interspecific relationship more comprehensively, the usage of more dissimilar identification method would be necessary. RAPD has been widely used in genetic maps construction, breeding lines identification and gene markers screening, as has been in the realm of genetic evolution of parasites such as trypanosome, schistosome and trichinella spiralis [30–32]. As we all know, RAPD has its innate drawbacks in stability and repeatability of bands due to the highly random hybrid sites with template DNA. Therefore, to gain stable bands and repeatable results of RAPD, we followed the PCR amplified condition whose stability and repeatability has been already reported [27, 28]and had validated the results before formal experiments. Meanwhile, the usage of commercial Taq DNA Mix from the same batch replaced the addition of dNTP, Mg2+ and Taq DNA polymerase one by one also improved the reaction stability in this study. Theoretically, the larger the number of polymorphic sites of RAPD, the more reliable the genetic relationship can be concluded. In this study, 10 of 20 random primers has produced 121 polymorphic bands, with the proportion of polymorphic bands was more than 99%, which could reflect genetic diversity among these isolates credibly.
As is shown in Fig. 2 and Table 1, these 17 strains were clustered into two major Clades. Three isolates SC10H2, SD and GL, which identified as L. (SauroLeishmania) sp. based on gene marker cyt b and HSP70 previously [11], had a lower genetic similarity with other strains and firstly clustered as Clade I. This result from the view of genome adds to evidence that in China there exist an undescribed Leishmania species and it is a distinct branch which has low homology with Chinese L. donovani strains [9, 33, 34]. Particularly, this RAPD result also demonstrated the discrimination and differentiation in the relationship among subspecies of L. donovani complex. Three isolates Cy, WenChuan and 801, which isolated from Gansu, Sichuan and Xinjiang respectively, were clustered together and separated from other L. donovani isolates. Meanwhile, they had the closest genetic relatives with L. donovani reference strain DD8. This result confirmed that the genetic differentiation really existed in Chinese L. donovani, which was in accord with the phylogenetic analysis on HSP70 [11]. However, the dendrogram of UPGMA showed that L. donovani reference strain DD8 did not cluster with KXG-918, KXG-927, KXG-XU, KXG-LIU, 9044, KXG-65 and SC6 that identified as L. donovani previously. Instead, it had the closest genetic relatives with Cy, WenChuan and 801. This result indicated there were differences between these Chinese L. donovani strains and L. donovani reference strain from India on genomic level, which was inconsistent with the phylogenetic analysis results of gene markers [8, 34]. For this, as a gene marker only contains partial information of genome and the selective pressure is various among different genes, intraspecific genetic differentiation probably could not be reflected fully. On the other hand, great genetic variation is generally generated between species or genera in RAPD amplification, so that using an individual represents a species may cause deviation of phylogenic results. Consequently, the divergence in this study need to be further verified by enlarging the sample size or combining with other methods. In addition, the cluster dendrogram showed that the clade B divided into two small branches: one is Sichuan isolate SC6 from hill foci, another included Shandong isolate 9044 from plain foci and other five Xinjiang isolates from desert foci. It revealed that there were still differences among VL isolates from hill, plain and desert in China, which supported the previous report [35]. Although RAPD technology has been waned gradually, while it is very sensitive to identify slight intraspecific differences, which is suitable for differentiation of sibling species.
By means of RAPD, the population differentiation of Leishmania could be analyzed, and the random amplified segments reflected abundant polymorphic information of genome. Furthermore, the bioinformatics research of species-specific segments is benefit for further genetic information mining and the develop of specific genetic markers, which could help to species identification or disease diagnosis. In this study, we developed three L. donovani complex species-specific DNA markers in available 17 Leishmania strains. Although the markers have been verified and showed stably repeatable results in differentiating L. donovani complex from other Leishmania species, according to their distribution of Blast hits in NCBI, the primers of marker 1-AD17 has a greater specificity for amplification of L. donovani complex since it had no matching primer binding sites with other Leishmania species sequences. We considered that the differences of primer binding sites or annealing sites of amplification was the cause for generation of differential DNA fragments of diverse species in RAPD, which was speculated in previous report [25]. Thus, the SCAR marker 1-AD17 have the potentiality to be developed into a rapid diagnosis marker of kala-azar. Admittedly, the L. donovani complex specific DNA marker in this study still have certain limits because of diverse species of Leishmania and less genomic information is available. Therefore, more parasite samples and patient specimens would be needed to test the specificity.
Through bioinformatic analysis, the three markers were all located in large chromosomes instead of kinetoplast, which was similar with some other reports [28, 36, 37]. It may be related to that multicopy genes are found preferentially on disomic chromosomes [38], which would increase the probabilities of random primers binding to them. As genome sequences of different Leishmania species are highly conserved [38], the amplification loci of RAPD frequently located in variable regions. In this study, although there were 4 to 7 ORFs in the three markers, in which only 3-O13 has two potential promotors and its ORF-4 has the potential of encoding proteins. The following protein prediction analysis showed that the hypothetical protein had higher antigenic index and surface probability. Nevertheless, all these need further experiments to be verified.