Genomic Data Mining Reveals Abundant Uncharacterized Transporters in Coccidioides immitis and Coccidioides posadasii

Coccidioides immitis and Coccidioides posadasii are causative agents of coccidioidomycosis, commonly known as Valley Fever. The increasing Valley Fever cases in the past decades, the expansion of endemic regions, and the rising azole drug-resistant strains have underscored an urgent need for a better understanding of Coccidioides biology and new antifungal strategies. Transporters play essential roles in pathogen survival, growth, infection, and adaptation, and are considered as potential drug targets. However, the composition and roles of transport machinery in Coccidioides remain largely unknown. In this study, genomic data mining revealed an abundant, uncharacterized repertoire of transporters in Coccidioides genomes. The catalog included 1288 and 1235 transporter homologs in C. immitis and C. posadasii, respectively. They were further annotated to class, subclass, family, subfamily and range of substrates based on the Transport Classification (TC) system. They may play diverse roles in nutrient uptake, metabolite secretion, ion homeostasis, drug efflux, or signaling. This study represents an initial effort for a systems-level characterization of the transport machinery in these understudied fungal pathogens.


Introduction
Soil-dwelling dimorphic fungi, Coccidioides immitis and Coccidioides posadasii, are causative agents of coccidioidomycosis, commonly known as Valley Fever [1]. While about 60% of infected people show no or minimum symptoms, the remainder develop clinical symptoms, ranging from pneumonia to life threatening, disseminated coccidioidomycosis. There is currently no clinically available vaccine against coccidioidomycosis and treatments are based on standard antifungal therapies.
The development of new antifungal drugs targeting Coccidioides is urgently needed because (1) The rise of incidence of coccidioidomycosis [2]. Incidence was reported to increase by almost 800% from 2000 to 2018 in California, and in 2018 over 15,000 cases were reported to the Centers for Disease Control and Prevention of the United States [3,4]; (2) The increase in the area of endemicity. Historically endemic in the arid and semiarid areas of the southwestern USA, Mexico, Central America, and South America [5], Coccidioides were recently reported to expand to Utah, Oregon, and Washington state [6][7][8][9][10]; (3) The emergence of clinical isolates conferring resistance to antifungals [11][12][13].
The search for new targets against Coccidioides requires a better understanding of Coccidioides biology. Despite recent advances in epidemiology, ecology and population biology [1,14,15], the molecular mechanisms underlying fungal growth, adaptation to the host environment, pathogenesis, and virulence, remain elusive. Difficulties in Coccidioides Figure 1 depicts the pipeline of our genomic analysis. The completed reference genome sequences of C. immitis RS strain (BioProject PRJNA12883) and C. posadasii SOWgp strain (BioProject PRJNA9616) were downloaded from the NCBI Genbank database (https: //www.ncbi.nlm.nih.gov/data-hub/genome/?taxon=5500, accessed on 20 July 2021) (Table 1) [16,17]. All the amino acid sequences and annotated features were retrieved.
The transporter classification database TCDB, https://www.tcdb.org/ (accessed on 20 July 2021), was used as the knowledge database for the identification and classification of transporters in Coccidioides [27,28]. TCDB includes curated transporter sequences, classification, structural, evolutionary, mechanistic, medical and biotechnological information about transport systems from a variety of organisms. A total of 21,373 amino acid sequences of the transporters collected in the TCDB was downloaded.

Identification, Classification, and Characterization of Putative Transporters in Coccidioides
The BLASTP query of all the proteins in C. immitis RS strain and C. posadasii SOWgp strain against the TCDB database was conducted to identify Coccidioides proteins that were homologs to known or predicted transporters [29]. The cutoff for homologous genes were set as: BLASTP E-value < 10 −20 and greatest positive percentage >50. The annotation of Coccidioides transporters was based on the hits in the TCDB with the lowest E-value, and the highest similarity score.
The predicted Coccidioides transporters were further classified into families and subfamilies based on the 5-letter Transport Classification (TC) system [13]. Similar to the Enzyme Commission (EC) system for classification of enzymes, the TC system is formally adopted by the International Union of Biochemistry and Molecular Biology for transporter classification and nomenclature. The 5-letter TC number, in the form of VWXYZ corresponds transporter class, subclass, family, subfamily and the substrate/substrates transported [27].
Conserved domains/motifs in predicted Coccidioides transporter sequences were identified by searching against the Pfam 35.0 database [30], which is a collection of protein families based on hidden Markov models implemented in HMMER [31]. The TMHMM program was used to analyze the transmembrane structures and predict transmembrane segments (TMSs) [32]. The substrates for predicted Coccidioides transporters were predicted based on the Chemical Entities of Biological Interest (ChEBI), an ontology and dictionary focused on small chemical compounds [33].

Multiple Alignment and Phylogenetic Analysis
Multiple sequence alignments were obtained using the MUSCLE program [34,35]. Phylogenetic trees were inferred by the neighbor-joining methods [36] and the maximum likelihood [37] using MEGA11 [38]. Bootstrap resampling with 1000 pseudo replicates was used to assess statistical support for each branch [39].

C. immitis and C. posadasii Possess a Rich Repertoire of Transporters
To gain insight into the transport machinery of Coccidioides, the protein sequences in the C. immitis RS strain and C. posadasii SOWgp strain were subjected to an exhaustive search against the TCDB database, which has a catalog and a structure-, mechanic-, and phylogeny-based classification of transporters. Stringent threshold of E-value < 10 −20 and positive percentage >50 were adopted to ensure the high coverage with low false-positives. A total of 1288 and 1235 transporter homologs were identified in C. immitis and C. posadasii, respectively, which account for 13.0% and 17.1% of their respective proteome ( Table 1). The transporter composition in the two organisms was highly similar (Supplementary Table S1): Mutual BLASTP analysis between C. immitis and C. posadasii showed that only 23 transporters were specific to C. immitis, and six transporters were specific to C. posadasii. Our new catalog of transporters included 197 and 103 transporters that were previously annotated by genome-sequencing projects in C. immitis and C. posadasii, respectively [17].
Combining domain specification and TC nomenclature, we further divided the 1288 and 1235 transporter homologs into seven classes, 25 subclasses, and 269 families in C. immitis and C. posadasii ( Figure 2 and Table 2). The majority of these predicted transporters were not previously characterized. The most abundant class of transporters is electrochemical potential-driven transporters (Class 2) (Supplementary Table S1). These 358 and 343 secondary carriers account for approximately over 27% of all the transporters in C. immitis and C. posadasii. The majority of the Class 2 transporters belong to porters (TC Subclass 2.A), consisting of uniporters, symporters, and antiporters [40,41]. The porters in fungal species catalyze the uptake of nutrients such as sugars, amino acids, and efflux of toxic compounds and drugs. Among these porters, the major facilitator superfamily (MFS) transporters have been implicated in multidrug resistance in Candida albicans [42][43][44][45], C. glabrata [46,47], C. tropicalis [48], and Aspergillus fumigatus [49]. We identified 140 and 135 MFS transporters in C. immitis and C. posadasii, respectively. See Section 3.2.2 for detailed analysis. The second most abundant class is primary active transporters (Class 3) (Supplementary Table S1). They represent about 22% of the transporter repertoire in C. immitis and C. posadasii. This class of transporters uses a primary source of energy such as ATP to transport solutes across a membrane against their electrochemical gradient. The members of this class are major players in the uptake and excursion of diverse solutes. Most notably, a superfamily of ATP-binding cassette (ABC) transporters are known to confer efflux-mediated antifungal resistance in pathogenic fungi [21]. Our analysis identified 44 and 38 putative ABC transporters in C. immitis and C. posadasii, respectively. The genomic and phylogenetic analyses of these ABC transporters are seen in Section 3.2.1. In addition, P-type ATPases were found to be present in fungi species such as A. fumigatus, A. nidulans, A. oryzae, C. neoformans, Neurospora crassa, Saccharomyces cerevisiae, and Schizosaccharomyces pombe [50]. These active pumps play important roles in ion homeostasis for fungal cell physiology [51]. C. immitis and C. posadasii genomes contain 20 and 18 P-type ATPases, respectively, including ATPases that translocate calcium, magnesium, copper, phospholipid, and potassium-sodium (Supplementary Table S1).
Class 1 channel/pore proteins are ubiquitous in all living organisms. Over 19% of the transporter homologs in C. immitis and C. posadasii fall within this class. These transporters facilitate diffusion of solutes in an energy-independent mode. The two largest subclasses in Class 1 present in Coccidioides are α-type channels (TC Subclass 1.A) [52], and membranebounded channels that form pore complexes (TC Subclass 1.I) [53]. Coccidioides possess various cation channels, including calcium, potassium, and transient receptor potential channels with potential roles in cellular signaling and homeostasis. Homologs of these cation channels are present in S. cerevisiae [54,55], C. albicans [56], Aspergillus spp. [57], and C. neoformans [58]. Although the detailed physiological function of fungal channels is largely unknown, they are considered as potential drug targets [59]. In addition to channels, Coccidioides also possess Nuclear Pore Complex (NPC) transporters (TC family 1.I.1) that bidirectionally transfer macromolecules between the cytoplasm and the nucleus. Besides facilitating nucleocytoplasmic trafficking, the homologs of these porins were shown to be involved in chromatin organization, and gene expression regulation in fungi [53,60,61].
Taking a systems perspective, the TC system also groups various accessory factors that facilitate the transport, but do not directly transport solutes into Class 8. Over 140 predicted homologs fall into Class 8, representing 11% of the total transporter homologs in C. immitis and C. posadasii. The physiological roles of these putative transporters are largely unknown.

Potential Functionally Important Transporters in C. immitis and C. posadasii
Among the 269 families of predicted transporters, undoubtedly some of the uncharacterized transporters perform important functions in the Coccidioides life cycle, for example, the ABC transporter superfamily (TC 3.A.1) and the major facilitator superfamily (MFS) (TC 2.A.1). Examples of potentially important transporters in Coccidioides and their homologs in other pathogenic fungi are shown in Table 3.
Our genomic analyses identified 44 and 38 putative ABC transporters in C. immitis and C. posadasii, respectively. Phylogenetic analysis revealed that these ABC transporters could be divided into five distinct efflux groups ( Figure 3).  The evolutionary history was inferred using the neighbor-joining method [36]. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches [39]. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method [97] and are in the units of the number of amino acid substitutions per site. Evolutionary analyses were conducted in MEGA11 [38]. Accession numbers: XP_0012XXXXX (C. immitis sequences), XP_004445935.1 (C. immitis), and XP_0030XXXXX (C. posadasii sequences). Each colored diamond corresponds to the branching point leading to a specific ABC group. ABCB, the largest group, includes three families. The Multidrug Resistance Exporter Family (TC 3.A.1.201) has been widely studied in other fungi species such as S. cerevisiae, S. pombe, C. albicans, A. fumigatus, and C. neoformans [21]. Four homologs in C. immitis and C. posadasii are closely related to the MDR1 gene in A. fumigatus, which confers resistance to Cilofungin [87] (Supplementary Table S1). The second member of the ABCB group is the Heavy Metal Transporter (HMT) Family (TC 3.A.1.210). Eight C. immitis and three C. posadasii HMT homologs are closely related to the vacuolar HMT1 gene in S. pombe [98], which is capable of enhancing heavy metal tolerance in a high calcium content. A mitochondrial ATM1 gene in C. posadasii is homologous to mitochondrial iron transporter ATM1 in S. cerevisiae, which is essential for biogenesis of cytosolic iron/sulfur proteins [99][100][101][102]. The third member of the ABCB group is the Mitochondrial Peptide Exporter (MPE) Family (TC 3.A.1.212). ATP-dependent permease MDL1 in C. immitis and its homolog in C. posadasii are evolutionarily related to the MDR2 gene in Trichophyton tonsurans (scalp ringworm fungus), which plays an important role in susceptibility to multiple antifungal drugs [91].
The ABCC group includes the Drug Conjugate Transporter (DCT) Family (TC 3.A.1.208). The Coccidioides sequences in this family are homologous to various drug transporters, for instance, the YOR1 gene in S. cerevisiae conferring resistance to oligomycin, rhodamine B, tetracycline, verapamil, eosin Y and ethidium bromide [103], the YCF1 gene in S. cerevisiae involved in vacuolar metal resistance and drug detoxification [104], the bile acid transporter BAT1 gene in S. cerevisiae, and the EcdL gene in A. fumigatus conferring resistance to antifungal Echnocandin B [90].
The ABCD group includes the Peroxysomal Fatty Acyl CoA Transporter (P-FAT) Family (TC 3.A.1.203). C. immitis and C. posadasii each possesses two paralogs of P-FAT genes, which may be involved in the fatty acid transport across the peroxisomal membrane [105].
Cholesterol/Phospholipid/Retinal (CPR) Flippase Family (TC 3.A.1.211) is the single member of the ABCA group in C. immitis and C. posadasii, which may mediate the efflux of cellular cholesterol and phospholipids [106].

Major Facilitator Superfamily (MFS)
MFS (TC 2.A.1) constitutes a large and diverse superfamily of secondary active transporters. Widespread across all three domains of living organisms, MFS transporters move a broad spectrum of small molecules across membranes to maintain important physiological function of cells [107][108][109]. Similar to the ABC superfamily, MFS has been widely recognized in various pharmacological processes by active excursion of cytotoxic compounds [110]. Mounting evidence suggests that MFS transporters are key mediators of antifungal resistance [20,[111][112][113][114][115][116]. While an MFS transporter was shown to have undergone fast evolution in the Coccidioides lineage, indicating potential significance of MFS in Coccidioides adaptation [17], our knowledge of the MFS system in Coccidioides remains minimum.
We identified 140 and 135 MFS transporters in C. immitis and C. posadasii, respectively, representing nearly 11% of the predicted transporters in their genomes. They belonged to 18 families. Of particular interest, abundant members were found in the Drug:H+ Antiporter-1 (DHA1) Family and the DHA2 Family. DHA1 and DHA2 transporters have demonstrated roles in antifungal resistance in fungal species in the genus of Saccharomyces, Candida, Cryptococcus and Aspergillus [20,114].
Phylogenetic analysis revealed that DHA1 and DHA2 homologs in Coccidioides are distributed into two clusters, which is consistent with their distinct structural properties (Figure 4). The sequences in the DHA1 family are marked in green, and the sequences in the DHA2 family are marked in red. The evolutionary history was inferred using the neighbor-joining method [36]. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches [39]. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method [97] and are in the units of the number of amino acid substitutions per site. Evolutionary analyses were conducted in MEGA11 [38]. Accession numbers: XP_0012XXXXX (C. immitis sequences), XP_004445935.1 (C. immitis), and XP_0030XXXXX (C. posadasii sequences).
The DHA1 cluster (TC 2.A.1.2) includes 29 and 28 members in C. immitis and C. posadasii, respectively. These genes display sequence homology to DHA1 genes characterized in other fungi (Supplementary Table S1). For example, both Coccidioides genomes contain paralogs to the HOL1 gene in S. cerevisiae, which is capable of nonselective uptake of histidinol and other cations [117]. Notably, C. immitis and C. posadasii each possesses two paralogs with high sequence similarity to the QDR2 genes in Candida species. QDR2 is known to confer prevalent resistance to a broad spectrum of antifungals including miconazole, clotrimazole, tioconazole, and ketoconazole and quinidine [47,92,118]. One of these two paralogs, XP_001240201.2, appeared to have undergone rapid evolution in C. immitis [17].
The DHA2 cluster (TC 2.A.1.3) includes 24 and 21 members in C. immitis and C. posadasii, respectively. Two paralogs of aflT genes are in the Coccidioides genomes; AflT is an efflux pump in the aflatoxin pathway in filamentous fungus Aspergillus parasiticus [93]. Four paralogs are present in each of the Coccidioides genomes, closely related to the YOR378W gene in S. cerevisiae, which is involved in boron stress tolerance [119]. Abundant copies (nine and seven) of genes are present in C. immitis and C. posadasii, respectively, with high homology to MFS1 genes, which is related to the antifungal resistance in the wheat fungal pathogen Zymoseptoria tritici [94].

Other Novel Transporters
Our catalog of Coccidioides transporters also includes novel transporters with broad implications in fungal physiology. For example, we found members in the Nucleobase: Cation Symporter-1 (NCS1) Family (TC 2.A.39) and in the Nucleobase/Ascorbate Transporter (NAT) or Nucleobase: Cation Symporter-2 (NCS2) Family (TC 2.A.40) [95,120,121], which may be important components of salvage pathways for purine, pyrimidine, and related metabolites [96]. Two paralogs of TRK genes (TC 2.A.38) were identified in Coccidioides. TRK1 and TRK2 were found to participate in potassium uptake and response to internal and external signals in S. cerevisiae and C. albicans [122,123].

Structural and Biochemical Features of Transporters in C. immitis and C. posadasii
Transmembrane segments (TMSs) are important structural components for transporters to translocate solutes across membranes. The topology of a transporter is specified by the number of the TMSs, and its overall orientation in the membrane [124]. The number of TMSs is a characteristic feature of fungal transporters; for example, different subfamilies of fungal ABC proteins were shown to possess distinct number of TMSs and nucleotide-binding domains, suggesting structure-function correlations [21,125]. To reveal transmembrane topology of predicted transporters in Coccidioides, we performed TMHMM analysis [126]. Similar distributions of TMS topology were shown in two Coccidioides genomes. The number of TMSs range from 0 to 22 (Table 4 and Figure 5). While the structural-functional significance of the TMS topology in Coccidioides is yet to be elucidated, we found that the majority of membrane transporters are channels/pores, electrochemical potential-driven transporters, group translocators, or electron carriers. Table 4. Distribution of topological types of transporters in Coccidioides genomes.

Discussion
Transporters are essential components for the survival of living organisms. The roles of transporters have been demonstrated in a variety of fungi. Our knowledge of the transport machinery in Coccidioides, however, remains limited. The genome annotation laid the groundwork for transporter characterization in C. immitis and C. posadasii [16,17]. To date, only a small number of transporters in Coccidioides have been reported or studied, including MDR1 (a multidrug resistance protein, accession number XP_003069119.1) and PSP1 (a hypothetical lipid transporter, accession number XP_003069236.1) in C. posadasii [127], a transmembrane amino acid transporter CIMG_11858 (accession number XP_012214138.1) and a major facilitator superfamily transporter CIMG_09822 (accession number XP_001240201.2) that showed fast evolution in the Coccidioides lineage [17], an ABC multidrug transporter CIMG_09753 (accession number XP_001240132.1), and a copper transporter CIMG_10037 (accession number XP_001239015.1) in C. immitis [128].
To fill in the critical knowledge gaps in Coccidioides biology, here, for the first time, we present a catalog of 1288 and 1235 putative transporters in C. immitis and C. posadasii, based on exhaustive homology search and comparative genomic analysis. These transporters fall into seven classes, 25 subclasses, and 269 families, with diverse transmembrane topologies and a wide array of substrates. Our hypothesis that Coccidioides fungi possess a rich and powerful transport machinery is justified.
It was estimated by the TCDB that transporters constitute about 10% of all cellular proteins [27]. The high content of transporters found in Coccidioides is likely an attribute of their adaptation to the complex soil ecosystem and the alien mammalian host system. Soil represents one of most challenging natural environments in which microorganisms scavenge nutrients, produce toxins to competing organisms and resist the effects of such cytotoxic substances. A similarly high content of transporters was observed in soil microbials, especially in the genus of Streptomyces [129,130]. However, unlike most Ascomycetes, which are plant pathogens or plant associated, clear evidence suggested that Coccidioides have undergone extensive genomic evolution to adapt to the animal host niche [17]. Such adaptation includes at least two major challenges: first, surviving from a plant-associated to an animal-associated nutritional environment, in a desert or a semi-dessert setting, and second, surviving from host immune detection and defense. The distribution of substrates in the Coccidioides transporters shows that a large number of transporters use proteins, amino acids and derivatives as substrates (Supplementary Tables S2 and S3), indicative of the need for an animal-associated nutritional niche. Moreover, the observed lineagespecific expansion of transporter families including ABC and MFS may be an outcome of positive natural selection in response to the external stress within a human host and to antifungal treatments, thereby contributing to the development of infectious phenotypes in Coccidioides [16,17].
We are bearing in mind that the catalog presented in this study are in silico predictions that await experimental validation. Historically, Coccidioides are understudied, partly due to the required biosafety level 3 (BSL3) laboratory containment and the special expertise needed to manage the aerosol risk posed by the large amounts of spores, and the severity of coccidioidomycosis. Experimental assays standard for other organisms can often be time-consuming or cost-prohibitive. Thanks to the advent of the genomic era, the in silico approach is becoming a cost-effective and efficient approach to identify and prioritize genes for wetlab characterization, especially suitable for non-model organisms such as Coccidioides.
While the roles of Coccidioides transporters in physiology, pathogenesis, and stress response are yet to be fully investigated, this study represents an initial attempt to a systemslevel understanding of the mechanisms underlying Coccidioides survival and infection. Transporters on the catalog are likely members of the cellular networks associated with nutrient uptake, ion balance, drug excursion, signaling, and regulation. With the availability of high throughput assays, it is possible to integrate various types of omic data to interrogate the expression profiles and associations of transporter genes and their upstream regulators and downstream substrates/effectors in a network perspective [128,[131][132][133]. By combining in silico omics-based discovery with wetlab characterization, there is an increased likelihood of identifying new therapeutic targets for these neglected fungal pathogens in the genus of Coccidioides.
In addition to providing a catalog of Coccidioides transporters for characterization, this genomic study also raises an important unanswered question: what are the content and diversity of transporters among other pathogenic fungal species? Published work to date mostly focuses on specific transporters or transporter families. For example, Costa et al. [20] surveyed MFS multidrug transporters in pathogenic fungi such as C. albicans, C. tropicalis, C. parapsilosis, C. guilliermondii, C. lusitaniae, C. glabrata, A. fumigatus, and C. neoformans. Kovalchuk and Driessen [21] performed phylogenetic analysis of fungal ABC transporters. Our future direction will be to predict and classify transporters in other pathogenic fungi and conduct comprehensive comparative genomic analyses of the transporters on their diversity and lineage-specific features.  Acknowledgments: The authors are grateful to Chiung-Yu Hung for helpful discussion. We thank Milton Saier and scientists at the TCDB for making the transporter resources publicly available.

Conflicts of Interest:
The authors declare no conflict of interest.