On the genetic basis of the effect of Spiroplasma on the male reproductive fitness of Glossina fuscipes fuscipes.

Son and colleagues recently reported [1] the phenotypic effects of the infection of the endosymbiont Spiroplasma in the tsetse fly Glossina fuscipes fuscipes, a major disease vector in humans. The authors examined the impact of Spiroplasma on the reproductive biology of G. f. fuscipes, showing that infection by the pathogen modifies sex-biased expression in reproductive tissues, ultimately affecting female fecundity as well as sperm motility and competitiveness among other phenotypes. The effect on male reproductive biology is presented in connection to the Vector Base gene ID GFUI025244, i.e., a presumed ortholog to the Drosophila melanogaster coding gene Sperm-specific dynein intermediate chain (Sdic). In fact, the interpretation of the results is heavily reliant on the functional and phenotypic attributes of Sdic in D. melanogaster.

In D. melanogaster, Sdic is a defective duplicate of the parental gene short wing (sw) [2], which encodes a cytoplasmic dynein intermediate chain [3] that acts as a subunit in protein complexes called dyneins–a class of molecular motors responsible for microtubule-based motility in the cell [4]. Expression profiling of Sdic showed that it is primarily expressed in male testes [2] but also in ovaries and other somatic tissues of both sexes [5]. Knockout experiments revealed that Sdic impacts sperm performance in competitive settings [6], likely through enhancing sperm displacement or retention ability in the sperm storage organs of the female [7]. Sdic has undergone multiple tandem duplications [8], forming a cluster on the X chromosome, which is flanked by the genes sw and obst-A upstream, and by AnxB10 and CG9581 downstream (Fig 1A). The Sdic proteins encoded by different paralogs are highly conserved except for a stretch at their carboxyl termini [5,8]. Crucially, Sdic has been found in all strains of D. melanogaster examined so far, but not in any other Drosophila or insect species [2,5,9]. Therefore, Sdic is a D. melanogaster-specific gene, conflicting with its assumed presence in G. f. fuscipes [1].

Fig 1

The ortholog to the G. f. fuscipes gene GFUI025244 in D. melanogaster is sw and not Sdic.

(A) Gene organization surrounding Sdic and sw in D. melanogaster according to the annotation of the Berlin genome assembly for the reference strain ISO-1 [5] compared to that around the gene GFUI025244 in the supercontig KK351816 of the G. f. fuscipes genome assembly 3.0.2 (GenBank accession: GCA_000671735.1). For G. f. fuscipes, the name of the presumed ortholog in D. melanogaster is also provided. Gene size and spacing are not to scale. Based on the current assembly of the region, only one of the two genes, Sdic or sw, is present in G. f. fuscipes. (B) Sequence alignment of the D. melanogaster proteins Sdic and sw, and the protein predicted by VectorBase to correspond to Sdic in G. f. fuscipes and three close relatives. Only the first 200 amino acids from the N-terminus are shown. As a Sdic protein of D. melanogaster, only the product encoded by the paralog Sdic1 in the reference strain ISO-1 is included in the alignment; all encoded proteins by D. melanogaster Sdic paralogs have an identical 5’ termini [5]. The amino acid sequence encoded by different exons of the genes sw and Sdic in D. melanogaster and G. f. fuscipes are color coded. The gene ID in each Glossina species is indicated before the species name. The presumed Sdic protein in a fourth species, G. austeni, was omitted as it is fragmented into three partial ORFs. As shown in the alignment, the encoded protein by GFUI025244 exhibits high sequence identity with the entire sw protein of D. melanogaster, which is not the case in relation to the Sdic1 protein as it is defective for the residues coded by the first two exons of sw.

To clarify this disparity, we retrieved the ~16kb of genomic sequence centered around the presumed ortholog of Sdic in G. f. fuscipes (GFUI025244), which resides in the supercontig KK351816 (Fig 1A). As confirmed by BLASTP homology searches against D. melanogaster, this genomic fragment includes the orthologs to the coding genes obst-A (GFUI025254; E-value = 10−106) and AnxB10 (GFUIO25246; E-value = 10−143) as outermost genes. According to the annotation in Vector Base, there is one additional coding gene immediately upstream (GFUI025251) and another one downstream (GFUIO25261) of GFUI025244. Reciprocal best match searches using BLASTP revealed that GFUI25251 is a likely ortholog of CG14384 (E-value = 0), a gene located on the 3L of D. melanogaster, while GFUI025261 does not have any significant match, an observation that holds when other Diptera are considered. Surprisingly, there is no annotation in Vector Base that corresponds to the ortholog of the gene sw, an essential gene in D. melanogaster [3]. In order to discard that the sw ortholog could be located elsewhere in the genome of G. f. fuscipes, we performed BLASTP searches against G. f. fuscipes. We found one main significant blast hit involving GFUI025244 (E-value = 10−143), i.e., the gene model annotated as the ortholog to Sdic, and two secondary hits associated with additional dynein intermediate chain encoding genes in D. melanogaster (GFUI019652, E-value = 10−18, the putative ortholog to CG1571; GFUI019660, E-value = 10−10, the putative ortholog to Dnai2). To discard that this region of the G. f. fuscipes genome was misassembled, thus providing a distorted reconstruction of the region where the putative Sdic ortholog could reside, we replicated the same analyses in the closely related species G. austeni, G. morsitans, G. palpalis, and G. pallidipes, finding identical results. This outcome directly challenges that the ortholog to GFUI025244 in D. melanogaster is Sdic (Fig 1A).

We then aligned the proteins encoded by the presumed Sdic in G. f. fuscipes and closely related species, as well as Sdic1 and sw of D. melanogaster (Fig 1B). Except for G. pallidipes, species that exhibits a 180 amino acid residue deletion at the 5’ terminus, the protein sequence encoded by the presumed ortholog to Sdic is virtually identical across Glossina species. For alignable intervals, the encoded proteins in the Glossina species exhibit sequence identity values of 86.3–86.7% along their entire length with the protein sw of D. melanogaster. In contrast, the sequence identity of these proteins with that of Sdic of D. melanogaster, although also high (79.3–83.4%), is essentially absent across the 123 amino acid residues at the 5’ terminus. In D. melanogaster, the Sdic protein also lacks the 100 residues encoded by the first two exons of the sw protein. Notably, the amino acid residues encoded by the first exon of Sdic, which is conserved across all Sdic paralogs and unique to Sdic–therefore absent from sw–[5], are also lacking in the presumed Sdic proteins from the Glossina species. Collectively, these results strongly suggest that Son and colleagues [1] misidentified the ortholog to sw as the ortholog to Sdic, following the existing annotation in Vector Base.

Spiroplasma infection preferentially affects the expression of sex-biased genes in reproductive tissues of both males and females of G. f. fuscipes [1]. In relation to male reproductive fitness, the transcript of the putative ortholog to Sdic, as assumed by the authors [1]–here shown to be actually sw–, is particularly abundant in the spermathecae of females that mated with males free of Spiroplasma compared to equivalent matings involving infected males. In stark contrast, no differences in expression levels for the putative ortholog to Sdic were found in the male reproductive tract of infected versus noninfected males. Subsequent phenotypic tests determined that although the number of stored sperm in females did not differ when mated with infected versus noninfected males, the fill with sperm of the spermatheca decreased in infected females compared to noninfected females, and the beating frequency of the sperm from infected males was significantly lower compared to that of the sperm from noninfected males. Although conspicuous, these observations are merely correlative. In D. melanogaster, neither sw nor Sdic are exclusively expressed in sperm cells as expression has been documented in female and somatic tissues [5,10]. Additionally, it is not clear from the data presented whether the other genes expressed in the sperm are also downregulated in the spermatheca of females that mated with infected males. Therefore, any impaired sperm motility could be the result of other affected biological processes and not of one necessarily dependent on the sw ortholog.

As part of its general role in cytoplasmic functions in D. melanogaster, the protein sw interacts with the p150 subunit of the dynactin protein complex through the 5’ terminus and with the dynein light and heavy chains of the cytoplasmic dynein complex through the 3’ terminus [3]. The role of this protein in the sperm remains uncharacterized [11]. Therefore, even if the sw ortholog is the most impacted gene by Spiroplasma infection in G. f. fuscipes, it is unclear how this could affect sperm motility. Further, and in relation to the impact of Spiroplasma on reducing male fertility in G. f. fuscipes, which could have valuable consequences for vector control, several experiments should be performed: (i) to characterize the precise role of the ortholog to sw compared to other dynein subunit-encoding genes; (ii) to quantify any male fertility reduction in competition settings [12]; and (iii) to determine whether the ortholog to sw is essential, as it is in D. melanogaster, or at least relevant for overall organismal fitness. Particularly this latter point is crucial as it is unclear what is the true probability of success of a potential intervention that relies on the use of Spiroplasma infected males as this type of male might have a substantially lower fitness compared to uninfected males due to an impaired functionality of the sw ortholog in G. f. fuscipes.