Intergenomic synteny and whole-genome duplication
All seven corresponding chromosomes between the FCDK and FCSH genomes exhibit conserved syntenic relationships (Figure 2a). MCScanX identified 212 syntenic blocks containing 27,559 (59.55%) PCGs from the two genomes. Both chromosomes 1 and 4 of FCDK showed perfect matches with the corresponding chromosomes of FCSH; however, some obvious large-scale chromosomal structural variations were present between other corresponding chromosomes of the two strains. Large-scale inversions occurred at least one, three, two, and four times on corresponding chromosomes 2, 3, 5 and 7, respectively, of FCDK and FCSH. Meanwhile, many translocation events occurred on corresponding chromosomes 6 and 7, respectively of the two strains.
Several large distal regions of FCDK chromosomes 1 (~10.51 Mb), 3 (~3.76 Mb), 4 (~7.71 Mb) and 7 (~9.13 Mb) did not share any syntenic blocks with FCSH (Figure 2a). The top enrichment GO terms of ~2,300 PCGs located in these FCDK-specific regions were mainly related to nutrient responses, immunity, stress tolerance and protein conformation, etc. (Figure 2b), which may associate with the better adaptation of this strain to the environment relative to FCSH. Most of other enriched GO terms (Figure 2b) and most enriched KEGG pathways (Figure 2c) were associated with viruses, bacteria and fungi as well as interactions between these organisms and FCDK. A large number of genes may have been generated from ancient horizontal gene transfers (HGTs), as revealed by Faddeeva-Vakhrusheva et al. (2017) in the FCBL strain.
We tested the WGD hypotheses using within-genome collinear patterns, paranome Ks distributions, and the estimated number of HOX gene clusters. First, 14 syntenic blocks (198 genes) were identified within the FCDK genome, and nine (131 genes) were found within the FCSH genome (Figure 1i), among which only nine FCDK blocks were distributed on different chromosomes. Most of the blocks occurred in genomic regions with a high density of transposons. Second, KDEs fit the distributions very well without obvious peaks in the Ks distributions (Figure 2d). When BGMM mixture components were fit to Ks distributions, none of the four default hypotheses (1~4 components/WGDs) was found to be predominant (Figure S2). Third, only a single, complete HOX gene cluster was identified on chromosome 5 (Figure 1i), with a length of approximately 5 Mb, in both the FCDK and FCSH genomes. The Hox gene cluster showed the same order of Scr , Ftz , Antp ,Ubx , lab , pb , Hox3 , Dfd , Abd-Aand Abd-B reported in a previous study (Faddeeva-Vakhrusheva et al., 2007). Therefore, three lines of evidence (i.e., few within-genome syntenic blocks, an absence of obvious peaks in Ks distribution plots, and single HOX clusters in the genome) absolutely supported an absence of WGD events in F. candida .