The protozoan parasiteLeishmania sp. is responsible for significant worldwide human morbidity and mortality, and the clinical features of diseases vary depending on the causative species. The genusLeishmaniabelongs to the family of trypanosomatids with, among others,Trypanosoma bruceiandT. cruzi, responsible for sleeping sickness and Chagas' disease, respectively. Taxonomically, it is separated into two sub-genera: L. (Leishmania) andL. (Viannia). The differentiation between both appears very ancient, as it is estimated to be contemporary to the formation of the Gondwana, i.e. around 60 million years . However, comparative genomics showed the general structure of the genome to be remarkably stable within this genus. Chromosomal synteny groups are entirely conserved for all Old World species (subgenusL. (Leishmania)) where 36 heterologous chromosomes have been identified . As regards the New World, only two and one large chromosomal rearrangements, according to the subgenus, have been shown as compared with Old World species. This leads to a chromosome number of 34 and 35 forL. (Leishmania) andL. (Viannia) species, respectively, but with all Old World species linkage groups remaining conserved . Similarly, chromosomal restriction maps showed a complete conservation of collinearity of markers between species . Finally, the sequencing of the 'TriTryp' genomes also demonstrated a high degree of synteny among those three protozoa (L. major, T. bruceiandT. cruzi) .
Trypanosomatids exhibit a number of highly original molecular and cellular biological features. Among those, one may cite systematic trans-splicing, consisting in the addition of a 39 nt-spliced leader RNA at the 5'end of all mRNAs, and the near absence of promoters for polymerase II, implying an absence of regulation of gene expression at the transcriptional level. One of the most extraordinary features revealed by these genome projects was the gene organisation into large collinear clusters present on a single strand and comparable to prokaryotic polycistronic units, except that the genes present have no common nor akin function [6–8]. These large directional gene clusters (DGCs) are separated by short sequences of a few kbs termed coding strand switches or strand-switch regions (SSRs), where the transcription sense converges or diverges. This remarkable position led several authors to express different hypotheses upon the putative function of these regions inLeishmania.
The comparativein silicoanalysis of several SSRs inL. majoronly revealed low homologies at the structural as well as nucleotide sequence level [5,8,9] that render difficult a common interpretation as for their putative role. AT distribution and hairpin content analysis failed revealing common features to different SSRs but allowed separating them into two groups with respect to the adjacent gene clusters transcription orientation (divergent or convergent) . The same authors showed that the five SSRs analysed presented a very high intrinsic DNA curvature, the latter being classically associated with transcription as well as replication and centromere functions (reviewed in ).
However, experimental data showed that the deletion of the sole SSR of chromosome 1 did not affect mitotic stability, hence it was not necessary for chromosomal replication nor segregation . This goes against the hypotheses of this SSR being a single replication origin  or a centromere . It is noteworthy that, whereas inT. cruzi, a 16-kb SSR, made for a large part of retroelements, has been identified as a centromere , inLeishmania, the centromeric function on chromosome 1 could be attributed to a subtelomeric 20-kb satellite repeat cluster . On the other hand, run-on experiments showed that this same SSR onL. major chromosome 1 contained a single site of bi-directional initiation of transcription for both gene clusters . This might explain why the deletion of this SSR could not be realised on the three copies of chromosome 1 in the reference strainL. major'Friedlin' . On the other hand, the expression of a reporter gene inserted into one of both gene clusters was not affected by the deletion of the SSR , which shows that transcription is possible without the presence of the SSR. Myler et al. explained the latter fact by a minor 'residual' level of expression on the chromosome that would not initiate at the SSR itself : still, this level of transcription is sufficiently high to allow the expression of the reporter protein.
Here, we bring new elements in the knowledge of the structure and conservation of these regions using an infrageneric compared genomics approach. We have sequenced the SSR of chromosome 1 in 15 highly divergent species ofLeishmaniaandSauroleishmania(of which the inclusion withinLeishmaniaremains uncertain)  and show the presence of conserved structural elements and motifs that were overlooked during sequence annotation or using inter-genera comparative genomics.