Pheromone expression reveals putative mechanism of unisexuality in a saprobic ascomycete fungus

Homothallism (self-fertility) describes a wide variety of sexual strategies that enable a fungus to reproduce in the absence of a mating partner. Unisexual reproduction, a form of homothallism, is a process whereby a fungus can progress through sexual reproduction in the absence of mating genes previously considered essential for self-fertility. In this study, we consider the molecular mechanisms that allow for this unique sexual behaviour in the saprotrophic ascomycete; Huntiella moniliformis. These molecular mechanisms are also compared to the underlying mechanisms that control sex in Huntiella omanensis, a closely related, but self-sterile, species. The main finding was that H. omanensis displayed mating-type dependent expression of the a- and α-pheromones. This was in contrast to H. moniliformis where both pheromones were co-expressed during vegetative growth and sexual development. Furthermore, H. moniliformis also expressed the receptors of both pheromones. Consequently, this fungus is likely able to recognize and respond to the endogenously produced pheromones, allowing for self-fertility in the absence of other key mating genes. Overall, these results are concomitant with those reported for other unisexual species, but represent the first detailed study considering the unisexual behaviour of a filamentous fungus.


Introduction
Sexual reproduction in fungi commonly requires an interaction between strains that carry different genetic information at the mating type locus [1]. This region, known as the MAT1 locus in ascomycetes, harbours genes that represent either the MAT1-1 or the MAT1-2 idiomorph [2]. The MAT1-1-1 gene defines the MAT1-1 idiomorph, but can co-exist at this locus with other genes such as MAT1-1-2 and MAT1- [1][2][3]. Similarly, the MAT1-2 idiomorph is defined by the MAT1-2-1 gene, though other genes can also be present [2]. These genes typically encode proteins with DNA binding domains and are known to regulate the global gene expression patterns associated with sexual reproduction [3]. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Foundation (NRF) Centre of Excellence in Tree Health Biotechnology (CTHB). This work is based on the research supported in part by a number of grants from the National Research Foundation of South Africa (including Grant specific unique reference number, UID 83924). This research was also funded in part by Prof Brenda Wingfield's SARCHi grant. The grant holders acknowledge that opinions, findings, conclusions and/or recommendations expressed in any publication generated by the NRF supported research are that of the author(s), and that the NRF accepts no liability whatsoever in this regard. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
were used. One of these was sexually-competent, producing abundant ascomata and ascospores, while the other maintained a strictly vegetative growth form. For H. omanensis, three different isolates were used. Isolate CMW11056 (CBS 118113) is a mixed mating-type culture (MAT1xMAT2) representing the sexually reproducing isolate type. The vegetative isolates (CMW44450, CBS 143823 and CMW44442, CBS 143822), represent MAT1 and MAT2 isolates, respectively. The two vegetative cultures were isolated as single ascospore progeny from the mixed mating-type culture CMW11056 [13].
Cultures were grown and maintained on 2% malt extract agar plates (20 g.L -1 malt extract (Biolab, Merck, South Africa) and 20 g.L -1 agar (Biolab, Merck, South Africa) supplemented with thiamine hydrochloride (100 mg.L -1 , SIGMA, Steinheim, Germany) and streptomycin sulphate salt (150 mg.L -1 , SIGMA, Steinheim, Germany). These are forthwith referred to as MEA-ST plates. The cultures were maintained at room temperature for the duration of the study. All cultures used in this study are preserved in the culture collection (CMW) of the Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, South Africa and at the Westerdijk Fungal Biodiversity Institute, The Netherlands.

RNA extraction and sequencing
Total RNA was extracted from 5-7 day-old cultures. This time period ensured that cultures did not reach the stationery phase, but gave enough time for sexual reproduction to take place. Nine MEA-ST plates were used per culture, allowing for three biological and three technical replicates. Mycelium was harvested from the vegetatively growing H. moniliformis isolate as well as the MAT1 and MAT2 H. omanensis isolates. A mixture of mycelium, ascomata and ascospores was harvested from the sexually reproducing H. moniliformis (unisexual) and H. omanensis (heterosexual) cultures.
The harvested tissue was flash frozen in liquid nitrogen and ground to a fine powder using a mortar and pestle. RNA extractions were performed using the RNeasy 1 Mini Kit (Qiagen, Limburg, The Netherlands) following the manufacturer's protocols with the following modifications: the RLC buffer was used as the extraction buffer and the optional on-column DNase 1 digestion was included. The integrity of the total RNA was assessed by 2% (w/v) agarose gel electrophoresis at 120V for 25 min and the concentration was estimated using an ND 1000 Spectrophotometer (ThermoScientific, Waltham, USA). The RNA was further subjected to quality testing using the Experion™ automated electrophoresis system (BioRad Laboratories, California, USA) at the ACGT Microarray Facility (University of Pretoria, South Africa). Dynabead 1 -based mRNA enrichment (ThermoFisher Scientific, Carlsbad, USA), cDNA synthesis and library preparation were performed at the Central Analytical Facilities (CAF) at the University of Stellenbosch (South Africa). The RNA sequencing was completed at the same facility using the Ion Proton Platform and PI™ Chip system.

RNA-seq analysis
CLC Genomics Workbench V7.5 (CLC bio, Aarhus, Denmark) was used to filter out raw reads with a Phred quality score of below 20 (Q 0.01) and a read length of more than 300 bp. Up to two terminal ambiguous nucleotides were trimmed from the remaining reads. To remove any reads representing the rRNA fraction of the total RNA, the filtered and/or trimmed reads were mapped to a species-specific contig harbouring only the rRNA sequences using the RNA Seq (Legacy) tool in CLC Genomics Workbench. To ensure optimal mapping, a minimum length fraction of 0.5 and a minimum similarity fraction of 0.8 were used. Reads that had not mapped to the rRNA contigs were then used in further analyses.
Gene prediction and annotation was conducted on the draft genomes assemblies of H. moniliformis (accession JMSH01000000, [29]) and H. omanensis (accession JSUI00000000, [30]) using Web AUGUSTUS with the Fusarium graminearum gene models [31,32]. These gene models were used because F. graminearum is the Huntiella species' closest relative for which gene models are available in the AUGUSTUS library. To detect differential expression of these predicted genes, the non-rRNA reads from each isolate type were mapped onto the species-appropriate gene annotations. Read mapping was conducted using the same length and similarity fractions as for the rRNA mapping. Intra-specific gene expression level comparisons were conducted using CLC Genomics Workbench following the method proposed by Mortazavi et al. [33].
For the accurate comparisons of gene expression across samples, a two-step normalization of the expression values was conducted. Firstly, RPKM values were used to represent relative expression. RPKM (Reads Per Kilobase of exon model per Million mapped reads) corrects for differences in expression levels between genes of different lengths and allows for accurate intra-and inter-sample comparisons [33]. Quantile normalization (CQN) was subsequently performed on the RPKM values in order to account for the expected technical variation specifically found in RNA-Seq data [34]. Baggerley's Z test [35] was used to test for expression differences across comparable samples, using the normalized RPKM values. To further minimize measurement errors, a corrected p-value was used to test for statistical significance. In this case, the p-value was corrected using the Benjamini-Hochberg false discovery rate (FDR) method to adjust for multiple test correction [36].
Genes displaying a 2-fold change in expression at an FDR-corrected p-value of 0.05 and lower were considered as significantly DE. When detecting expression level changes for the MAT genes in H. omanensis, a different approach was employed because the sexually reproducing culture is likely made up of approximately half MAT1 material and half MAT2 material. For example, if the expression of the MAT1-1-1 gene has the same RPKM value in both the vegetative and sexually reproducing cultures, there would have been a 2-fold expression increase in the sexually reproducing isolates. This is because the MAT1 individual expressing the gene would represent only~50% of the culture used for RNA extraction. The same was also true for the MAT1-1-2 gene and for the MAT1-2-1 and MAT1-2-7 genes in MAT2 individuals. Consequently, we considered genes displaying equal RPKM values as being significantly DE. BLAST2GO [37] was used to assign Gene Ontology (GO) terms, InterPro identities and KEGG enzyme codes to all the genes predicted using Web AUGUSTUS. BLAST2GO also assesses GO term enrichment of one set of genes with respect to another and was used to determine whether any GO terms were enriched in the highly expressed and DE gene lists with respect to the complete list of genes from the genome. This enrichment was confirmed statistically using the Fishers Exact Test (FET) in the same program, with a p-value of 0.05 or less indicating significant enrichment. REViGO was then used to condense the enriched GO terms by removing redundant descriptions using the "small, 0.5 allowed similarity" setting. The REViGO output was then visualized using Tableau V10.1.

Functional annotation
Genes known to be involved in sexual reproduction in other ascomycete fungi were investigated for their presence and possible DE in the two Huntiella species (Table 1). These included the MAT genes already known to be present at the MAT loci of the two species [13], as well as other sex-related genes from Saccharomyces cerevisiae [38,39] and Aspergillus species [40]. The a-and α-factor pheromone genes identified based on homology to previously described Ascomycete pheromone proteins [41] were also included. In order to identify these genes in the genomes of the two Huntiella species, gene sequences were downloaded from the NCBI Gene Database (www.ncbi.nlm.nih.gov/gene) and used as queries in local BLASTn and tBLASTx searches against the draft genome assemblies using CLC Genomics Workbench.

Saccharomyces cerevisiae
Endoprotease alpha-factor processor

RNA-seq statistics
High quality RNA (RQI > 9) was extracted from each of the cultures used in this study (S1 Table). Sequencing of the individual samples produced an average of~19 million reads per library for the H. moniliformis samples and an average of~21 million reads per library for the H. omanensis samples (S2 Table). The raw RNA-seq reads used for this study are available on the NCBI SRA database under the following accession number: SRP108437. Quality filtering and trimming of the sequence reads retained 99.9% and 99.7% of the H. moniliformis and H. omanensis reads, respectively. Once reads mapping to the rRNA contigs of the two species had been discarded, an average of more than 14 million reads per library remained for use in the subsequent analyses (S2 Table).
Of the remaining non-rRNA reads, an average of 71% and 62% mapped to the respective H. moniliformis and H. omanensis genomes (S2 Table). Unmapped reads were most likely the result of sequencing errors, reads from repetitive or unassembled genome regions [42] and/or the presence of poly(A) sequences [43]. The percentage of reads mapped to the two Huntiella draft genomes was similar to reports from other eukaryotic RNA mapping projects: in S. cerevisiae, approximately 56% reads mapped to the assembled genome [44], 69% in Heterobasidion species [45], about 50% in human [46], and 60% in cattle [43].

Detection of gene expression
Of the predicted 6 864 genes in the 25 Mb H. moniliformis genome, 6 636 (97%) were expressed across both isolate types. In the 31 Mb genome of H. omanensis, 7 923 (94%) of the total 8 394 predicted genes were expressed across the three isolate types. A gene was considered expressed if three or more unique gene reads mapped to its annotation [47] and a normalized RPKM value of at least 0.1 was observed [48]. The majority of the expressed genes in both species were expressed by all isolates (Fig 1).
The sexually reproducing isolates of both species exhibited the highest number of uniquely expressed genes (Fig 1). Of the genes expressed in the vegetative H. moniliformis isolate, 156 were highly expressed (RPKM > 1 000, S3 Table) with 117 being unique to this isolate. The highly expressed genes were enriched for biological processes such as primary metabolism, macromolecule metabolism and biosynthesis (Fig 2, S4 Table). In the sexually reproducing isolate, 85 genes were highly expressed (S3 Table), with only 46 being unique to this isolate type. The highly expressed genes were enriched for processes such as cell wall organization, protein kinase C signalling, generation of energy and developmental processes (Fig 2, S4 Table). In the two vegetative isolates of H. omanensis, 107 genes were highly expressed, including 60 genes that were unique to the vegetative isolate type (S5 Table). These highly expressed genes were enriched for processes including glucose and carbohydrate metabolism, translational elongation and developmental processes (Fig 2, S6 Table). This is in contrast to the 74 highly expressed genes (S5 Table) in the sexually reproducing culture that included only 24 unique genes. These genes were enriched for carbohydrate catabolism, response to stress/stimulus and sexual reproduction (Fig 2, S6 Table).

Differentially expressed genes
Of the total gene complement of H. moniliformis, 41% of the genes were differentially expressed between vegetative and sexually reproducing isolates (Fig 3A, S7 Table). In contrast, only 24% of the total H. omanensis gene complement was differentially expressed in the same comparison (Fig 3B, S8 Table). In H. moniliformis, approximately half of the DE genes were up-regulated in the vegetative isolate and were enriched for housekeeping functions such as transcription and translation as well as RNA processing and protein transport (S9 Table). The remaining genes were up-regulated in the sexually reproducing isolates and were enriched for cell communication, signalling and signal transduction as well as carbohydrate metabolism (S9 Table). In contrast, the majority of DE genes across the H. omanensis comparisons were up-regulated in the vegetative isolates. These genes included those involved in general housekeeping processes such as carbohydrate transport and metabolism, hyphal growth and cell wall biogenesis. The remaining genes were up-regulated in the sexually reproducing isolate and included those involved in the termination of G-protein coupled receptor signalling pathways, carbon utilization and microtubule organization.

Expression of the MAT genes
The MAT genes of the Huntiella species (Table 1) were identified in a previous study [13]. In H. omanensis, the primary MAT genes (MAT1-1-1 and MAT1-2-1) were up-regulated in the sexually reproducing culture, while MAT1-1-2 showed no evidence of differential expression. MAT1-2-7 also showed a general pattern of up-regulation during sexual reproduction, but high intra-sample variation rendered this result insignificant. This gene was discovered using bioinformatics analyses in a previous study [13] and has no similarity to any other known proteins. Given that six other genes had already been described at the MAT1-2 idiomorphs in a variety of other ascomycetes, this gene was named MAT1-2-7. This gene has recently been reported in Knoxdaviesia species [49]. Its association with the MAT locus in H. omanensis and the increase in its expression during sexual reproduction suggests that it has some unknown

Expression of the pheromone response pathway
Many of the genes involved in the pheromone cascade were identified in the genomes of the two Huntiella species (Table 1). These included the a-and α-factor pheromones, the two αpheromone processing factors, both pheromone receptors as well as several genes that play a role in the G-protein-coupled signal transduction pathway. The pheromone genes in H. omanensis were expressed in a mating-type dependent manner, with MAT1 individuals expressing the α-factor pheromone and MAT2 individuals expressing the a-factor pheromone. This was in contrast to the expression patterns of the pheromones in H. moniliformis (Fig 4), where no mating-type dependent expression was observed. The MAT2 individual expressed both the αfactor and a-factor pheromones at appreciable levels. The remaining genes were all expressed in H. omanensis and H. moniliformis in all isolate types.
There was some evidence to suggest that the pheromone receptor genes in H. omanensis exhibited mating-type dependent expression. The expression data indicated that the α-factor pheromone receptor was expressed solely by the MAT2 individual. The a-factor pheromone receptor was expressed almost exclusively by the MAT1 isolate, but was still expressed by the MAT2 isolate, albeit at much lower levels. This expression pattern did not extend to H. moniliformis. Rather, both receptor genes were expressed at a low, constitutive level in both the vegetative and sexually reproducing cultures.
The only genes that were differentially expressed in this pathway in H. omanensis were the α-factor pheromone (up-regulated) and its receptor (down-regulated). Again, H. moniliformis exhibited a different pattern and while the a-factor pheromone showed a decrease in expression in the sexually reproducing culture, the expression of the α-factor pheromone increased by almost 20-fold (p<0.001). Additionally, the mpkB, ste20 and ste23 genes (all involved in the signal transduction section of the response pheromone pathway) were also upregulated in the sexually reproducing H. moniliformis culture. Both pheromone factor genes showed differential expression, with the a-factor showing a significant decrease in expression in the sexually reproducing isolate while the α-factor showed a significant increase. The pheromone receptors showed no evidence of differential gene expression and seem to be expressed constitutively throughout the lifecycle. A logarithmic scale is used for the Y-axis. Ã indicates that there is a significant difference between two bars (2-fold change at FDR-corrected p-value < 0.05). https://doi.org/10.1371/journal.pone.0192517.g004 Pheromone expression reveals putative mechanism of unisexuality

Discussion
Unisexual reproduction enables the production of sexual spores in fungi that usually require an opposite mating partner. To date, unisexuality has been observed in only five species; four ascomycetes and one basidiomycete. Although three of these species are filamentous fungi, the underlying molecular mechanisms enabling this form of self-fertility have only been characterized in the two yeast species. We compared the gene expression changes associated with unisexual reproduction in the filamentous ascomycete H. moniliformis to similar changes associated with heterothallic mating in a closely related species. The results showed that there are significant differences in the pheromone production in these two species.
A comparative transcriptomics approach revealed that changes in the expression of the pheromone precursor genes in the heterothallic H. omanensis were very similar to the general expression profiles in self-sterile model fungi such as N. crassa. Although individual isolates of N. crassa possess genes for both the αand a-factor pheromones, they are expressed in a mating-type specific manner [50]. Thus, mat A (= MAT1) individuals express only the α-factor pheromone, while mat a (= MAT2) individuals express only the a-factor pheromone. Sexual reproduction is then initiated only in the presence of a mating partner of the opposite matingtype, despite the fact that both partners express both pheromone receptor genes [51]. Our results showed that pheromone expression in H. omanensis also follows this pattern. Interestingly, there was also some evidence of mating-type dependent receptor expression, a profile not observed in N. crassa or other heterothallic species [51].
The expression pattern of the pheromones in H. moniliformis provides a putative molecular mechanism that enables the species to reproduce unisexually. These genes exhibit neither the same mating-type dependent expression nor the differential expression seen in H. omanensis. However, they do display expression profiles similar to other homothallic species such as S. macrospora [17] and thus, both pheromones are expressed by a single isolate. It is important to recognize that species exhibiting primary homothallism, such as S. macrospora, possess both MAT1-1 and MAT1-2 genes, and thus, despite these genes being under mating-type control, both pheromones can be expressed. In contrast, H. moniliformis is able to achieve this expression profile, while harbouring only the MAT1-2-1 gene.
Pheromone expression is only part of the sexual activation pathway. This is because pheromones can only elicit their intended response if recognized by the appropriate pheromone receptor. In our study, both receptors were expressed at appreciable levels by both fungal species studied. This indicates that the pheromone response pathway is being utilized. In H. omanensis, the α-factor receptor showed down-regulation in the sexually reproducing isolate, while the H. omanensis a-factor receptor as well as both receptors in H. moniliformis showed no evidence of differential expression. This result indicates that differential expression of the receptors is not important to sexual reproduction in these species and that constitutive expression is sufficient. This could also indicate that differential expression of the pheromones themselves controls the activation of sexual reproduction, irrespective of whether it is a heterothallic or homothallic interaction.
The results of this study have also provided the first experimental validation for the expression of the MAT1-2 idiomorph-associated gene, MAT1-2-7. This gene was previously detected using de novo gene prediction software [13] and is thought to produce a full-length protein of 155 aa in H. omanensis. Its position upstream of the archetypical MAT1-2-1 gene in MAT2 isolates coupled with its absence in the MAT1 isolates highlighted the role that it might play in mating. In the present study, MAT1-2-7 had no detectable expression level in vegetatively growing isolates, but was shown to be expressed in sexually reproducing cultures. This expression pattern, in addition to its idiomorphic position, strongly supports a role for the MAT1-2-7 gene in sexual reproduction. Because it is truncated in H. moniliformis, we believe it is likely that this gene plays a role in the inhibition of self-fertility.
H. omanensis was included in this study because it represents a typical heterothallic species. It was consequently able to provide a basis for comparison of gene expression changes that occur during a sexual interaction in Huntiella species. This was especially important due to the unavailability of MAT1 H. moniliformis isolates. Thus, following the model proposed for H. omanensis, expression of the a-factor pheromone in H. moniliformis is either directly or indirectly influenced by genes present at the MAT1-2 idiomorph. This same model would suggest that the α-factor pheromone should be controlled by genes at the MAT1-1 idiomorph. However, H. moniliformis MAT2 individuals also express this gene. Taken collectively, these results suggest that MAT2 individuals of H. moniliformis have been able to overcome the mating-type dependent expression of the pheromones that would otherwise prevent homothallic behaviour in typically heterothallic species (Fig 5).
A modification of the typical heterothallic pheromone expression pattern represents the most likely molecular mechanism that underlies unisexual reproduction in H. moniliformis. Although, the genetic elements that govern this system remain unknown, they would likely have enabled a transition from typical heterothallism to unisexual reproduction via a change in the control of pheromone expression. Comparisons between the expression patterns of the pheromone genes as well as the genic content of the MAT1-2 loci in the two Huntiella species provides a basis for some speculation regarding the origin of mating-type independent pheromone expression observed in H. moniliformis. In this case, it is plausible that the pheromone expression patterns of the unisexual species could be attributed to the truncation of MAT1-2-7 These pheromones are secreted (2) and recognized by the pheromone receptors of a suitable mating partner (3). This recognition activates the pheromone transduction pathway, a MAP kinase cascade (4). Assuming other sex-favouring environmental conditions are met, this cascade alters gene expression patterns within the cells (5) and allows for the conversion of vegetative mycelia into sexually-competent tissue (6). B: The proposed pheromone system in H. moniliformis. The two species systems likely work in a very similar manner, except that H. moniliformis is able to express, secrete and recognize both the αand a-factor pheromones (1, 2 & 3). Once the MAP kinase cascade has been activated (4), the genetic and physiological changes are likely very similar to those in H. omanensis (5 & 6). https://doi.org/10.1371/journal.pone.0192517.g005 Pheromone expression reveals putative mechanism of unisexuality gene. This hypothesis rests on the assumption that a fully functional MAT1-2-7 protein acts as a repressor of the α-factor pheromone in MAT2 individuals, with its absence in MAT1 individuals allowing for expression of this pheromone. The genes of the MAT locus have been shown to be essential for pheromone expression in N. crassa [50], S. macrospora [52] and C. neoformans [25] and provides a precedent for the direct involvement of the MAT proteins in the pheromone system. In fact, a recent study has shown that the inactivation of the Mat2 gene in C. neoformans disables the pheromone pathway and preferentially enables unisexual behaviour over bisexual mating [24].
The hypothesis that the truncated MAT1-2-7 gene is responsible for the unisexual behaviour of H. moniliformis is further supported by mutation studies performed on the heterothallic P. anserina [53]. Here, disruptions in a number of the MAT genes resulted in self-fertility. Importantly, it has been shown that the wild type mat+ genes (= MAT1-2) are not only required for the activation of mat+ functions, but also for the repression of mat-(= MAT1-1) fertilization functions. Thus, when there is a mutation in the mat+ gene, FPR1, the repression of fertilization is lifted and the mutant is rendered self-fertile. We suggest that this could be similar to what occurs in H. omanensis, where a full length MAT1-2-7 gene represses the α-factor pheromone, rendering the MAT2 isolates self-sterile. However, in H. moniliformis, the MAT1-2-7 gene is truncated, the repression of the α-factor pheromone does not occur and mating-type independent expression is possible. This in turn leads to self-activation, self-fertility and culminates in unisexual reproduction.
The results of this study suggest that the mating-type independent expression of the a-and α-factor pheromones is responsible for the unique ability of the MAT2 individuals of H. moniliformis to complete a sexual cycle in the absence of a MAT1 mating partner. As presented above, the role of pheromone expression in unisexuality has been previously investigated in both Ca. albicans and Cr. neoformans [9,20,[23][24][25][26]. The results presented here are most similar to observations for Ca. albicans, and we hypothesize that endogenous production of both the a-and α-factor pheromones by a single MAT2 individual is sufficient to initiate the sexual cycle in H. moniliformis. We believe that this is the most probable explanation for the data presented here but acknowledge that other plausible hypotheses may also exist.
The unisexual pathway exhibited by H. moniliformis was originally assumed to be a sexual process. This was partly due to the similarity of the reproductive structures and spores found in H. moniliformis and the heterothallic H. omanensis [13]. The apparent importance of pheromone expression in the unisexual pathway of H. moniliformis further substantiates this assumption. However, it is worth noting that another spore-producing process, known as haploid fruiting, also enables spore production in single mating type isolates. Although similar to unisexuality in that way, this process is not necessarily linked to sexual reproduction or the formation of a diploid state [54][55][56]. It may even require different environmental conditions to those necessary for sexual reproduction [57]. Haploid fruiting can be induced in Schizophyllum commune when the mycelium is physically damaged [55] or when a fruiting-inducing chemical is present in the environment [54]. While this raises an alternative hypothesis for the process through which H. moniliformis produces progeny, the currently available data supports a sexual process.
This study utilized a transcriptomics approach that has some inherent limitations. The most important is that conclusions drawn from the data assume that gene expression is directly correlated with the production of a functional protein product. In addition, pheromones are often expressed as pre-pheromone proteins that require post-translational modification. Therefore, future studies will include the functional characterization of the two pheromone genes as well as the MAT1-2-7 gene in order to enhance our understanding of the role these genes play in unisexuality. Furthermore, studies utilizing synthetic pheromones will be conducted in an attempt to initiate unisexual behaviour in H. omanensis.
Supporting information S1