Transcription machinery of the minimalist: comparative genomic analysis provides insights into the (de)regulated transcription mechanism of microsporidia — fungal-relative parasites

ABSTRACT

Microsporidia are eukaryotic obligate intracellular parasites closely related to fungi. Co-evolving with infected hosts, microsporidia have highly reduced their genomes and lacked several biological components. As it is beneficial for intracellular parasites like microsporidia to reduce their genome size, it is therefore reasonable to assume that genes encoding multifactorial complex machinery of transcription could be a potential target to be excluded from microsporidian genomes during the reductive evolution. In such a case, an evolutionary dilemma occurs because microsporidia cannot remove all transcription-machinery-encoding genes, products of which are essential for initialthe initial steps of gene expression. Here, I propose that while genes encoding core machinery are conserved, several genes known to function in fine-tune regulation of transcription are absent. This genome compaction strategy may come at the cost of loosely regulated or less controllable transcription. Alternatively, analogous to microsporidian polar tube, the parasites may have specialized factors to regulate their RNA synthesis.

1. Introduction

Microsporidia are opportunistic intracellular parasites whichparasites, which infect many hosts within a broad taxonomic range, including humans and economically important animals [1–5]. Although they are phylogenetically related to fungi, the lengthy periods of parasitism make them develop their own unique characteristics [4,5]. For example, microsporidia possess a specialized invasion structure, so-called polar tube, which rapidly everts out of the microsporidian spore when it encounters appropriate environmental conditions [2,6–8]. Besides, certain organelles (e.g., mitochondria), proteins involved in vesicular transport (e.g., Target of Rapamycin (TOR) and clathrin), and many metabolic enzymes have been lost in microsporidian species [1,2,7,9]. As these small fungal relatives have undergone drastic genome reduction, they largely depend on host metabolisms for their survival as well as reproduction and are considered as true obligate parasites [2,5,9]. Compared with other fungi, microsporidia are highly reduced in every level: from cell size and morphology, metabolic enzymesenzymes, and pathways, to the level of genes and genome.

To decode the genome, the process of gene expression starts with transcription, the synthesis of RNA molecule. Transcription of eukaryotic protein-coding messenger RNA (mRNA) is catalyzed by the twelve-subunit DNA-dependent RNA polymerase II (RNAPII) with help from a number of basal transcription factors [10–12]. For instance, multi-subunit protein complexes of general transcription factors (GTFs) participate in the recognition of core promoter and the formation of a pre-initiation complex (PIC). Evolutionarily conserved multi-subunit protein complex of Mediator serves as a bridge between specific transcription factors and basal transcriptional machinery [13–15]. Transcription initiation and elongation factors modulate RNAPII activity and facilitate elongation of the RNA transcript on chromatin templates [16–21]. Moreover, several transcription factors are also implicated in the interconnection between transcription and transcription-coupled mRNA processing [22–25]. Many proteins play important roles in the regulation of transcription in a gene-specific manner. To date, more than hundreds of transcription factors have been identified in budding yeast Saccharomyces cerevisiae and more are expected to be present in higher eukaryotes.

Given that it is beneficial for intracellular parasites like microsporidia to reduce their genome size in order to decrease doubling times, to overcome restrictive nutrient limitations, and to save space within their host cells [7,26], it is therefore reasonable to assume that genes encoding multi-player transcription machinery – RNA polymerase(s), GTFs, the Mediator, transcription initiation and elongation factors, transcription termination and 3’-end processing factors – could be potential targets to be excluded from microsporidian genomes during the reductive evolution. In such a case, an evolutionary dilemma may occur because microsporidia cannot remove all transcription-machinery-encoding genes, products of which are essential for the initial step of gene expression. It is therefore intriguing to determine how much the transcription machinery of these parasites have (d)evolved.

Focusing on the mRNA transcription machinery in microsporidia, I here bioinformatically demonstrate that while several core subunits of RNAPII are highly conserved, some are shortened, especially the C-terminal domain (CTD) of Rpb1, the largest subunit of RNAPII. The tandem repeated CTD heptad, a hallmark and regulatory region of Rpb1, is strikingly much reduced or completely lost. In many microsporidia, certain peripheral subunits of the polymerasepolymerase, such as Rpb4 and Rpb8 seem to be absent, strikingly resembling the subunit-losing pattern of RNA polymerase complex of a giant Vaccinia virus of the poxvirus family. While many GTFs are conserved, the Mediator complex and a vast majority of transcription elongation factors, especially Prp19 and transcription-export (TREX) complexes, are not. Since many factors involving regulatory controls of RNA synthesis seem to be lost, RNAPII transcription of microsporidia might be less controllable; otherwise, they may gain unidentified set of transcription factors, analogous to microsporidian polar tube, to regulate their RNA synthesis.

2. Methods

Identification of transcription machinery in the microsporidian genomes

To identify protein components of the transcription machinery in microsporidian genomes, information of each protein was obtained from the UniProt database [27]. Encephalitozoon romaleae (NCBI: txid571949), Encephalitozoon hellem (NCBI: txid27973), Encephalitozoon intestinalis (NCBI: txid58839), Encephalitozoon cuniculi (NCBI: txid6035), Ordospora colligata (NCBI: txid174685), Nosema ceranae (NCBI: txid40302), Enterocytozoon bieneusi (NCBI: txid31281), Enterospora canceri (NCBI: txid1081671), Enterocytozoon hepatopenaei (NCBI: txid646526), Hepatospora eriocheir (NCBI: txid1081669), Vittaforma corneae (NCBI: txid42399), Trachipleistophora hominis (NCBI: txid72359), Pseudoloma neurophilia (NCBI: txid146866), Edhazardia aedis (NCBI: txid70536), Anncaliia algerae (NCBI: txid723287), and Nematocida displodere (NCBI: txid1805483) were selected as canonical microsporidia, Amphiamblys sp. WSBS2006 (NCBI: txid1866961) as a Metchnikovellid, and Mitosporidium daphnia (NCBI: txid1485682) and Paramicrosporidium saccamoebae (NCBI: txid1246581) as early diverging microsporidia, and Rozella allomycis (NCBI: txid281847) as a closely related obligate endoparasitic fungus [2]. Unless stated otherwise, Saccharomyces cerevisiae yeast proteins involving transcription machinery were employed as queries in batch in a protein Basic Local Alignment and Search Tool (BLASTP) against the non-redundant (nr) database for microsporidian protein sequences on the National Center for Biotechnology Information (NCBI) website with an Expect (E)-value cutoff of 10⁻⁵ [28,29]. For proteins that no ortholog was detected by BLASTP search, the translated nucleotide BLAST (TBLASTN) operation mode was employed against whole-genome shotgun contigs (wgs) of microsporidian genomes with an E-value cutoff of 10⁻⁵ [28,29].

3. Evolutionary conservation of transcription machinery in microsporidia

In order to obtain the first glimpse into landscape of transcription machinery in microsporidia, I first obtained information of each protein from the UniProt database and used its proteome gene identifier (ID) as query in the protein Basic Local Alignment and Search Tool (BLASTP) against the non-redundant protein sequences (nr) database with an Expect (E)-value cutoff of 10⁻⁵ [27,28]. As assembly of certain microsporidian species may yet to be completed and/or certain sequences may be missing in the proteome database of annotated proteins, I also employed the translated nucleotide BLAST (TBLASTN) operation mode against the whole-genome shotgun contigs (wgs) with the same E-value cutoff if the initial BLASTP failed to identify a significant hit [28].

RNA polymerase II First, amino acid sequences of all twelve subunits, Rpb1 to Rpb12, of S. cerevisiae RNA polymerase II (RNAPII) were used as queries to identify microsporidian homologs. As expected, all subunits of RNAPII were likely present in the closest relatives of fungi: R. allomycis, P. saccamoebae, M. daphniae, and Amphiamblys sp. (Figure 1 and Table S1). Intriguingly, however, Rpb4 and Rpb8 are strikingly absent in almost all canonical microsporidia, while Rpb5 was undetected in certain species, especially the whole genus of Encephalitozoon (Figure 1 and Table S1). The lack of both Rpb4 and Rpb8 is noticeably somewhat similar to that of reduced RNA polymerases of certain viruses [30,31], suggesting the stripped-down version of RNA polymerase of this kind.

Figure 1.

Conservation of proteins in the RNA polymerase II complex in microsporidia and closely related fungal species. Each row is a species, and each column is a protein. Phylogeny of sequenced genomes is based on previously reported calculation [2]. Parasitic fungus is highlighted in brown, three early diverging microsporidia in medium brown, and canonical microsporidia in light brown. Numbers in parentheses show the size of genome of each organism in mega base pairs (Mbp). Ortholog assignments were determined and confirmed using BLAST with a reciprocal E-value cutoff of 10⁻⁵; orthologs available shown in blue, unavailable in light yellow. Twelve known subunits of eukaryotic RNA polymerase II (left panel), C-terminal domain (CTD) of the largest subunit Rpb1, and CTD phosphatases (right panel) are shown. Numbers in the box show the number of CTD heptad repeats.

ThoughAlthough highly conserved, the largest subunit of the polymerase Rpb1 in microsporidia may differ from those of other eukaryotes. The protein is known to possess a unique C-terminal domain (CTD) consisting of tandem heptad repeats with a consensus sequence of YSPTSPS ^11,21,32,33]. The number of repeats varies and reflects the complexity of the organism, e.g., 26 repeats in S. cerevisiae and 52 repeats in human. The CTD is crucial in connecting transcription with many processes of RNA metabolism and is tightly regulated by reversible phosphorylation [11,21,32]. After closely analyzing the C-terminal tail of microsporidian Rpb1, I found that the CTD region in the parasites is relatively shorter than that of yeast (Table 1). The number of tandem repeats has also clearly reduced in all microsporidia, some of which even showed no clear consensus sequence at all (Figure 1 and Table 1). It is interesting to note that if the Tyr residue of the heptad repeat was placed to the seventh position (i.e., S¹P²T³S⁴P⁵S⁶Y⁷) instead of the first position (i.e., Y¹S²P³T⁴S⁵P⁶S⁷), the residue probability of Tyr was higher than that of conventional assignment, particularly distinguishable in E. romaleae, E. intestinalis, E. cuniculi, O. colligate, and E. aedis (Table 1). Anyhow, although homologs of CTD kinases Ctk1, Bur1 and phosphatase Fcp1 were identified in all microsporidia, it is unclear whether they are true functional counterparts because the protein domains are often used in other biological pathways, too. Intriguingly, however, the presence of CTD heptad consensus repeat(s) correlates well with the presence of putative homologs of CTD phosphatase Ssu72 (Figure 1). These findings infer that microsporidian CTD of Rpb1 might also be post-translationally modified and involved in modulating functions of RNAPII. In summary, the CTD-mediated transcriptional regulation may occur in some of these fungal-relatives, but likely to a lesser extent than in other eukaryotes.

Table 1.

Structural characteristics of the C-terminal domains (CTD) of Rpb1 genes from microsporidian species. Putative heptad repeats within CTD are printed in bold (underlines were used merely for making a clearer distinction between adjacent heptad). Sequence conservation of amino acids (sequence logos) of the heptad repeats were generated using by WebLogo (https://weblogo.threeplusone.com/).

Microsporidian species	Length of Rpb1 (aa)	Length of the CTE (aa)	Sequence of the CTD	Total number of repeats	Consensus repeat if starts with SPTSPSY	Consensus repeat if starts with YSPTSPS
Encephalitozoon romaleae	1515	106	VKKLDKAIPLSNPVFKPNEPVTPVISTPSSDSFSISSGNWSPTHLEMAYSRDVDGRLSPTSPSYSPTSPSYSPTSPSYSISTSGFSNKSKSKEQDGDKKRRNDNNF	3
Encephalitozoon hellem	1478	69	VKKLDKAIPLSNPVFKPNDPVTPMISTPSSDSLSINSGNWSPTHLEMAYPRDLSGRLSPTSPSYSPTSP	1
Encephalitozoon intestinalis	1543	134	VKKLDKAIPLSNPVFKPNEPVTPVISTPSSDSFSISSGNWSPTHLEMAYPRDLSGRLSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSVLTPNFSNKNKSKDQDGNKKRRNEEPF	7
Encephalitozoon cuniculi	1487	78	MKKLDKAIPLSNPVFKPNEPATPVISTPSSDSFSISSGNWSPTHLEMAYSRDLGERLSPTSPSYSPTSPSYSPTSPSY	3
Ordospora colligata	1578	169	MNKLDKAIPLANPVFKANDPDTPVINTPMSDSFSISSGAWSPMHIDMGYNRDIVGRISPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSITSPSFAGKNKNKDQDGDRKRKNSHPF	12
Nosema ceranae	1605	219	VSKLEFAIPLSKPDYNYEDVDTPFIYSPVSESQSISSGNWSPAYLAEESKYSPKLSLFSVGSPSYGSASPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTGYKKDKDKKNERKRRHEN	18
Enterocytozoon bieneusi	1541	188	IKKLDQIVNKAFINTVDKNLIMLNSPAFTEALLIEQENCILSEDIGLFSPEHEELSDFNNIKIPEYNDSPNSEPNTYNSLNTPISSEKVNNSVMYTPNSFYFNQQPESIYFPGTNGSYFAASGEYFQSNNFHSNTILNSGGYQEYMANSNDYDNLSQQYKPIDSPKYNYDNITIYNYSSEEEKKEEKE	-	-	-
Enterospora canceri	1503	192	TERLDQMLFSAMNEAKSRATGGLLLNSPAFSEIAAVASNNLRLAEEAAPFSPEHENAESVTFSVPEYFSGMSPASLYAVSPASNAYMSPGSNFSGSAIYSPNSSYVNPGSTFYQSPGSNYYQPASNQNSAYFTPGSSNYFGGSGSYQPGSGLKYNAYSAGSNQNIYSIGSNNSRSKQSEYKLEEENKSKDGE	-	-	-
Enterocytozoon hepatopenaei	1542	185	KEKLATLNVTKTTTIANSVVLNSPAYTSDSDVVENEIIGKIQEDRGEFSPEYEHAEVVASRVAEFNMTPGSGVLYSSPGSGAMFSSPGSNSYAPGSTFGGSTFFQGAGSQGYYTQNTTAYQPGSGLGSYMGSGQAGSASMKPGNAYYNAYSVGSGQYSEKHDAYAVSNYENESTEYKEESSSEKK	-	-	-
Hepatospora eriocheir	1555	207	FDKLKEFNKDEVDDDSFEGALDQINAPDIDFISNSSAFYSEFTDMNTMKGFSPDYSANAPYSMTLQSTTAPYNSATFRADSSTFNPTSNNFNMVGSSRAAYYNPSIRSNLSPYQSRSVGDIMYGSGNRSTYENRSSYGSSPAAKFSGLSSNRYRPSSGSPGSSPYANNSSTAYSPGYLGYNPDIDYNNSYSPGHYSPGSYKKDSEDK	-	-	-
Vittaforma corneae	1569	193	IKKIEHIIPRYVSPSKEGMSAPLYSPSSDSVVSWSSGNYSPIGTGGFSPMSPLSAGFSPDYSGYNSIQSQTYSPTSAGYNIRSPVYNATSQVYNPRTPNFAPVSPTYNPMTSMYAPTSFDTSLSSDHHKPTSPSYTPLSPSNSPASPRYYGSYEPASSSYSPATPGFKLHSPTYSPSSPSFNEKWKDDGKGKK	-	-	-
Trachipleistophora hominis	1748	161	VHALGEKAFEFEHAPRQNVLWSPESGTCTDGVLSPAARYSPRYTPASPTYSRISSPSY SPTSPAYTPTSPKYVPMNAKYVLGSSRYSPTSPSYSPASPRYNPGRSAGSPQYYTRASPSYSPTSVRYYSPTSTRYVKGGVSPLYGQDDGEDKEEKDEKADKD	3
Pseudoloma neurophilia	925	95	TSSTTTVSSYKRESTPLISSKDCNKDDNKGDNKDKSKDNSKDGNGNKNVSHENILSYDKNKYDELLQILSSKLQNIPSYSYNNIHMMVSTGSKGS	-	-	-
Edhazardia aedis	2270	359	IQKLKDAVEICDKDFELEVRSPNSGYSPIIGLWSPIPEDSPLGKAYKIESSPSHSTQGGYSPISAGYSPISPGYNTSPNATYSPVSPGYGTIGGFSPSSPSYIPASPSYTPSGFSPSGMPISGTNAYAPSSPTYAPSSPAYNSPFTPGVQSSAQNTVTSPSYTPSVSSFASNRPTNSSAFTPVSPAYSSSRQPGSILNRPMNLGYTPTSPSYNPSSTFGRSSNVGYTPVSPSYTGSFQGSAYRPSNLSYAPTSPSYTSINSMMPPKTPKPVTTPVTKFVPQSPTNVYESMPQIPNVEHERSSSVGAFVTSSDKRNQQVENVNSKNDISEFDLERSGKNDEYVHASNKEKNDQHKDTKKE	13
Anncaliia algerae	1616	170	MEILQSVIPLARKYEFLREINTPLILSPDTITTPNRISMGYSPATLPGYSPTTPIYNPTSPVYSPTTPVYSPTSPEYSPTSPVYSPTTPIYSPTSPIYSPTTPVYSPTSPVYIPQSPKNISKTSQYNPKSPVYKPTSPAYRVTSPKNKTSDKRKLESEEECKRIEKDKDK	11
Nematocida displodere	1606	192	KDSLEMGIMAMEKYEFE DEDSGDKTPPDFASPSGSGYSEMSPYASQGGSPGFIWSPQVDSGYSPDASGGSGMFGYSPTSPGPASPSPYGGANAYTPMASYSFGYPGFSPSSPQNRPGSSSPIPTYAPLSPSRGYVEPYTPGRAFSPSTATLYMPAPPTSRREHKDSPKRLWKDDEKSESSDESFQLDKKQKY	-	-	-
Amphiamblys sp. WSBS2006	1552	155	DGEIKNIFGPVGEGDLDAQVEGLEMSPRMEASPLYSPTSIGSSGMAFSPLYQPETVRNAQRKDTQGDGGSSGYSPARQTYSPMSPAYSPTSPAYSPASPAYSPTSAAYSPASPAYSPMSPAYSPMSPAYSPMSPAYSPASPAYSPMSPQAPHTRR	9
Mitosporidium daphniae	1809	295	EDRLAMGALVPALSATAYSMVAAQATNMMSTVSGSATPGSLGFRSPWVQTGAGSSPQPYLKGSRSPTFRSSGDGYRTPGDIGFSPFSTSSGFSSGSSPSWWGGRSPAALVGRFSPTSPGYSPTSPGYSPTSPGYSPTSPGYSPTSPGYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPSYSPTSPKYSPSSPRPGAFSQTAFSPKSPGYSPSLSSSGQRHKK	22
Paramicrosporidium saccamoebae	1672	227	ESALNHALSYSQNFAAADWQNNGARTGDYTSYPDGQRSPAYSMYSGAPMSPASGSFSPAWSDAAAGGKWSPFRQSGSASPRYSPASPAYYSPASPAYEAGGASPRYSPASPAYSPASPAYSPASPAYSPASPAYSPASPAYSPASPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSPTSPAYSQSTTQGKQ	18
Rozella allomycis	1868	394	ESQVEQAIPIYYHPLLSSLNNGSQTPFLDSKTPYWNGSQTPSGAYTPHDSSFSPFNDSGSFSPGLSPTSPAYAPVSPGYSPKTPGFNISPGSGYSPGFKSTASPSYTPGSPVYNPSGQTQYSPGQTQYSPGQTQSYSPDQGGYSPSPYSPNQGAYSPSPYSPTSPLASPNMSYSPTSPSHRQTSPNYSPTSPLYSPASPHYSPTSPQYSPTSPQYSPTSPKYSPGSGGYSPTSPKYSPTSPQYSPTSPQYSPTSPKYSPGSGGYSPTSPKYSPTSPQYSPTSPKYSPTSPQYSPTSPKYTPGSGGYSPTSPQYSPTSPQYSPTSPQYSPGSPKYSPQGYSPDQGYSPKYSATGKYTSTPKYSPDDEQNYSPDTPKYSPDAQYSPDGSYSPAEKK	22

General transcription factors In order to correctly initiate transcription, general transcription factors (GTFs) are required to help the RNAPII position correctly on a promoter, to unwind the two strands of DNA, and to subsequently prepare RNAPII into elongation mode [10,34–36]. GTFs are generally required in all gene promoters transcribed by RNAPII, thus called TFII for transcription factor for RNA polymerase II: TFIIA, TFIIB, TFIID, TFIIE, TFIIF, and TFIIH. They are sequentially recruited and function differently during transcription initiation steps.

Since the group of GTFs is composed of more than thirty proteins in budding yeast, I next asked, to what extent, the GTF genes have been depleted in microsporidian reducedmicrosporidian-reduced genomes. TATA-binding protein (TBP), which recognizes and directly binds to the TATA-box sequence within a promoter, seemed to be highly conserved in the parasites as it may be important for initial recognition of the promoter (Figure 2 and Table S2). TFIIA and TFIIB, the two GTF components that form a complex with TBP and the TATA-box, could also be identified in most microsporidian species, suggesting their important role in the assembly (Figure 2). As parts of TFIID, TBP-associated factors (TAFs) are most likely reduced in microsporidia (Figure 2). Intriguingly, however, DNA-binding subunits of TFIID – TAF1 and TAF2 whichTAF2, which bind to the initiator element; TAF6 and TAF9 whichTAF9, which bind to the downstream promoter element – are all conserved in the parasites (Figure 2 and Table S2) [36]. By contrast, non-DNA-binding TAF subunits, which are thought to help integrate signals from transcription activator and/or Mediator complex, were strikingly undetected, inferring that sequence/gene-specific transcriptional regulation may be limited in microsporidia (see also below) [13,15,36].

Figure 2.

Conservation of general transcription factors (GTFs) in microsporidia and closely related fungal species. Similar to Figure 1, but GTFs were analyzed. TBP: TATA-binding protein; IIA: TFIIA; IIB: TFIIB; IID: TFIID; IIE: TFIIE; IIF: TFIIF; IIH: TFIIH.

TFIIH, a multifunctional complex with established roles in RNAPII transcription as well as transcription-coupled DNA repair, and TFIIE involving in TFIIH recruitment seemed to be partially conserved (Figure 2) [37–39]. Ssl2 and Rad3, the 3’-5’ and 5’-3’ ATP-dependent helicases, respectively, as well as their key regulators, Tfb2 and Ssl1, are present in all microsporidian species, while the Tfb5, which is a tiny ATPase-stimulating co-factor of Ssl2, seems to be absent (Figure 2). Two major components of CDK-Activating Kinase (CAK) were also identified. Interestingly, Tfb1, the component that interacts with many DNA repair factors, and Tfb4 that bridges Tfb1 and other TFIIH components, are both absent. The loss of these two DNA repair-related subunits correlates well with the absence of Rad4, the factor that recruits TFIIH to the nucleotide excision repair (NER) complex (Table S2). These observations imply that microsporidian TFIIH may function as general transcription factor but likely have lost the activities whichactivities, which are necessary for transcription-coupled DNA repair.

Particularly interesting is the presence/absence of TFIIF, a GTF whichGTF that plays an important role in the formation of stable pre-initiation complex, in particular for retention and positioning of TFIIB [10]. My BLASTP/TBLASTN analyses showed no apparent homolog of any subunit of yeast TFIIF (Figure 2 and Table S2). As mentioned above, TFIIB homologs is likely present in all microsporidia and the role of TFIIF in the core initiation complex is highly conserved in eukaryotes, even other polymerase systems – RNA polymerases I and II – both contain a TFIIF-like complex [10]. It is therefore reasonable to assume that TFIIF-like proteins might also exist in microsporidia. Recently, TFIIF-like complex has been identified in Trypanosoma brucei, an obligate intracellular parasite [40]. Although functional attributes suggest the complex represents TFIIF in the protist, the amino acid sequences are so divergent from those of most eukaryotes [40]. Perhaps, through a gradual evolutionary process similar to T. brucei, the TFIIF of microsporidia may be remotely diverged and could not be identified by simple BLAST analyses.

It is particularly interesting to compare the presence of the basal transcription machinery of microsporidia with previously reported those of several parasites: Entamoeba histolytica, Toxoplasma gondii, Plasmodium falciparum, Kinetoplasta, and Giardia lamblia [41]. First, though the twelve subunits of RNA polymerase II of T. gondii and P. falciparum are conserved, E. histolytica, Kinetoplasta, and G. lamblia seem to lack Rpb4 similarly to microsporidia. Moreover, while Rpb8 is likely conserved in the three species, Rpb12 seems to be absent. It is intriguing to note that both Rbp8 and Rpb12 are closely located at the position adjacent to Rpb11. Besides RNAPII, general transcription factors of those parasites seem to be slightly distinct to microsporidia. While TBP and TFIIB are highly conserved, apparent homologs of TFIIA are unfound. TFIIE, TFIIF, and TFIIH components are similarly conserved in all five parasites. The major discrepancy is the TFIID complex: E. histolytica is the only species possessing TFIID subunits similarly to microsporidia; other parasites, on the other hand, seem to completely lack the complex. The contribution of the difference in RNAPII and GTFs to the unique structure and regulation of the polymerase, if any, shall be investigated.

Mediator complex Mediator is an evolutionarily conserved multi-subunit complex comprising 25 and 30 subunits in S. cerevisiae and in human, respectively [13–15]. The complex is generally required for transcription by bridging the gap between RNAPII, GTFs, and sequence-specific DNA-binding transcription factors, which bind to specific DNA sequences in the regulatory regions and serve as repressors or activators [13–15]. Mediator comprises four distinct parts, which are head, middle, tail, and Cdk8 kinase modules [15]. The head and middle modules constitute the essential core of the complex and are necessary for transcription regulation. The tail and Cdk8 modules, on the other hand, serve regulatory functions [14].

In microsporidia, the Mediator complex is extensively reduced and/or diverged. BLASTP/TBLASTN analyses showed that all Mediator components except Med6 of the head, Med31 of the middle, and Cdk8 of the kinase modules are most likely absent (Figure 3 and Table S3). In yeast and human, all three proteins do not interact with one another [14] and therefore most likely are not able to form a stable complex in the parasites. Despite the negative data, the possibility that distant homologs existing in microsporidia cannot be absolutely excluded because it has been shown that the primary sequences of Mediator subunits are divergent in different organisms with limited identity and/or similarity [42]. To clarify the uncertainties, biochemical and genetic investigations are required.

Figure 3.

(Non)conservation of Mediator complex in microsporidia and closely related fungal species. Similar to Figure 1, but the Mediator complex was analyzed.

Capping enzyme and cap-binding complex Messenger RNA transcribed by RNAPII undergoes several modifications, including the addition of a reversed orientation of 7-methyl-guanosine cap to the 5’-end of the transcript [43]. The reaction is catalyzed by three enzymes – RNA triphosphatase, RNA guanylyltransferase, and RNA (guanine-7)-methyltransferase – and occurs soon after the initiation of transcription [43]. As a hallmark for RNAPII-transcribed RNA, the cap structure, which is recognized and bound by the nuclear cap-binding complex (CBC; Cbp20-Cbp80) and cytoplasmic cap-binding protein eIF4E, stimulates downstream processes such as pre-mRNA splicing, transcription elongation, mRNA nuclear export, protein synthesis by ribosome, and so on [43].

Microsporidia most likely have functional cap structure and relevant capping enzymes. Initial BLAST analysis using S. cerevisiae RNA triphosphatase Cet1 as query unfortunately failed to detect any microsporidian homolog (Figure 4 and Table S4). However, when previously characterized Cet1 of Schizosaccharomyces pombe and Encephalitozoon cuniculi were used, homologs of the enzymes were identified in many, but not all, microsporidia (Figure 4 and Table S4) [44,45]. On a side note, there were no detectable homologs of RNA triphosphatases of Caenorhabditis elegans, Arabidopsis thaliana, and Homo sapiens, either (data not shown). By contrast, the other two capping enzymes, RNA guanylyltransferase and RNA (guanine-7)-methyltransferase, were found by a simple BLAST search in almost all species across the group (Figure 4). This discrepancy could be explained in part by the notion that across evolution only RNA triphosphatase differs with regard to structure and catalytic mechanisms and is largely categorized into two families: (1) metal-dependent RNA triphosphatase mostly found in fungi and protists, and (2) metal-independent RNA triphosphatase found in nematodes, plants, and mammals [43,46]. Microsporidian RNA triphosphatases are probably of the first category.

Figure 4.

Conservation of transcription elongation factors in microsporidia and closely related fungal species. Similar to Figure 1, but 5’-cap-related proteins, transcription elongation factors were analyzed. CE: capping enzyme; CBC: cap-binding complex; FACT: FAcilitates Chromatin Transcription complex; Paf1C: Paf1 complex; Prp19C: Prp19 complex; TREX: transcription and export complex.

Microsporidia seem to have both nuclear and cytoplasmic cap-binding factors. While apparent homologs of Cbp80 were unfound in canonical microsporidia, Cbp20 and eIF4E counterparts were detected (Figure 4 and Table S4). Cbp80 is thought to help stabilize the mRNA with Cbp20, which is the subunit in direct contact with the 5’-cap structure, and recruit splicing and transcription elongation factors to the transcription machinery [47,48]. In microsporidia, Cbp20 and 5’-Cap might have evolved to function without the assistance of Cbp80. Besides, highly reduced spliceosome and a lack of many transcription elongation factors in microsporidia (see also below) may be in part the reasons for the loss of Cbp80. Taken together, I conjecture from the above observations that microsporidian mRNA contains a cap structure, which could be recognized and bound by cap-binding proteins.

Transcription elongation factors Following transcription initiation, transcription elongation is a process during which a new chain of RNA is synthesized as RNAPII keeps moving on the DNA template strand. Over the past decades, a number of transcription factors have been reported for their functions in controlling the processivity of RNAPII during the elongation stage [16–20]. One of the first characterized transcription elongation factors is the Spt4-Spt5 complex, also called DRB sensitivity inducing factor (DSIF) in higher eukaryotes [49–51]. Besides, TFIIS interacts directly with RNAPII and stimulates the intrinsic endonucleolytic activity of the pausing RNAPII to cleave the 3’-stretch of the nascent RNA, thereby allowing the restart of transcription [52]. As mis-incorporate the backtrackings favor backtracking of the RNAPII, TFIIS-induced cleavage also promotes transcription fidelity. Activities of Spt4-Spt5 and TFIIS are probably so important for processive RNA synthesis that functional homologs are also present in prokaryotes [51–54]. In fact, most microsporidia also have homologs of these proteins, though Spt5 is slightly less conserved than others (Figure 4 and Table S5). ByIn contrast, another elongation factor Spt6, which binds to phosphorylatedthe phosphorylated CTD of Rpb1 and facilitates co-transcriptional events such as pre-mRNA splicing and histone modification, seems to be absent in almost all canonical microsporidia (Figure 4 and Table S5) [55]. In the context of chromatin, yeast and mammal RNAPII requires FACT (for FAcilitates Chromatin Transcription), a dimeric histone chaperone of Spt16-Pob3 whichSpt16-Pob3 that facilitates transcription through chromatin environment [55–57]. Homologs of both subunits seemed to be present in many microsporidian species, suggesting that transcription regulation at the chromatin level may still exist in the parasites (Figure 4 and Table S5). Altogether, these observations suggest that RNAPII of canonical microsporidia may still require the activities of Spt4-Spt5, FACT, and TFIIS, but perhaps not Spt6, during transcription elongation.

Though several monomeric- and dimeric-subunit transcription elongation factors seemed to be present, multi-subunit complexes may be extensively lost from microsporidian genomes. First, the Paf1 complex, comprising five subunits, i.e., Paf1, Ctr9, Cdc73, Leo1, and Rtf1, were found only in early emerging microsporidia but completely unfound in all canonical species, even though the complex has been shown to interact genetically and physically with Spt4-Spt5 and the FACT complexes in budding yeast S. cerevisiae (Figure 4 and Table S5) [58,59]. Next, multifunctional Prp19 complex also seemed to be absent (Figure 4 and Table S5) [17,60]. Furthermore, two transcription and export (TREX-1 and TREX-2) complexes, which play major roles in coupling transcription elongation to mRNA nuclear export, and the transport receptor Mex67-Mtr2 (or TAP-p15 in humans) were largely undetected (Figure 4 and Table S5) [61–63]. Nevertheless, the nuclear export factor Xpo1 involving in export of biomolecules, including mRNA, proteins, and ribosomal subunits, from the nucleus to cytoplasm seemed to be present in all parasites (Figure 4 and Table S5). Taken together, I conclude that while the microsporidian transcription machinery is still essentially conserved for fundamental RNA synthesis, but it may lack many regulatory proteins and transcription elongation factors, particularly the components whichcomponents that fine-tune functionsthe functions of RNAPII and facilitate co-transcriptional events.

4. Conclusion and outstanding questions

Although microsporidia are closely related to fungi, their characteristics are distinct to their relatives. Due to reductive evolution, microsporidia have lost many genes encoding many cellular pathways and organelles; meanwhile, they have acquired an invasion apparatus with a highly specialized structure of polar tube for their unique parasitic lifestyles [4,7,9]. Gene losses and gains have shaped the structure and function of microsporidian molecular machines, including RNAPII transcription machinery.

Here, I propose that microsporidian RNA synthesis and the control thereof are most likely distinct to their fungal relatives. Though highly conserved, RNAPII of the parasites lacks certain regulatory subunits, i.e.that is, Rpb4 and Rpb8, and Rpb1 CTD, coordinating recruitment of many transcription factors to the transcription complex, has partly or entirely become diminished. While many GTFs seem to be present to help RNAPII recognize gene promoter and form transcriptional pre-initiation complex, regulatory multifactorial transcription factors suchfactors, such as Mediator, Prp19, TREX1/2 complexes seem to be massively absent. Since many genes of transcription machinery missing in microsporidian genomes are known to play roles in the fine-tune regulation of gene expression at transcription level, it is most likely that the genome compaction strategy of the parasites may come at the cost of loosely regulated or less controllable gene expression. Alternatively, if not completely lost, the parasite may theoretically have a set of specialized factors, analogous to the polar tube invasion machinery, to regulate their RNA synthesis.

For many reasons, the knowledge landscape of microsporidian transcription is still limited. Below I highlight a few, among many, important questions and relevant points for future investigations.

• The lack of Rpb4 and Rpb8 of RNAPII is strikingly similar to that of the vaccinia virus. But due to the protein sequence divergence, would the structure and function of the polymerases be different? Also, structures and functions of other reduced complexes could also be studied.

• Deteriorated C-terminal domain of microsporidian Rpb1 may directly and indirectly affect the coupling between transcription initiation, elongation, RNA processing, export, DNA damage repair, and so on. To what extent are microsporidian co-transcriptional processes impaired?

• In this study, certain crucial transcription factors such as TFIIF and Mediator complex, amino acid sequences of which could be divergent from their counterparts in other eukaryotes, are unidentified in several microsporidia [40]. If they indeed exist, how would they look like and do their activities the same or different to their counterparts?

• Since microsporidia lack many regulatory proteins, subunits, and/or most likely certain protein domains for transcription regulations, how would it affect the quality and quantity of gene expression at the RNA level of the parasites?

• It has been shown that a potential CCC- or GGG-like transcriptional initiation signal is present in 5’ upstream of all transcripts [64], how would the preinitiation complex be formed? Is it mediated by the conserved transcription factors? Or do microsporidia have any unknown specialized factors for the activity? Do they have a unique regulation?

• As certain chromatin-specific transcription factors seem to be present in the parasites, it is interesting to see how genes are transcribed in the chromatin context with less transcription factors. Also, do histone modification and epigenetic control still exist?

• Though this work only focuses on RNAPII transcription and closely-relatedclosely related machinery, other compartments involving RNA metabolism might also be reduced. For example, RNAPI and RNAPIII, the spliceosome, nuclear pore complex, RNA degradation and surveillance. To what extent have they been reduced?

• To my current knowledge, there are yet no efficient genetic tools for studying microsporidian genetics. Development thereof would facilitate a better understanding of microsporidian transcription from biochemistry and molecular biology aspects.

In conclusion, transcription machinery of microsporidia most likely has been stripped down to the core; while essential proteins and domains are conserved, fine-tuning regulators are not. In other words, microsporidia may lack major controls for their RNA syntheses. It would be intriguing to understand how the parasites use the minimal apparatus to shape their transcriptomes, thereby modifying the host cell biology and determining infection and host specificity.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Funding Statement

S.C. lab is supported by Mahidol University, the National Science, Research and Innovation Fund (NSRF) via the Program Management Unit for Human Resources & Institutional Development, Research and Innovation (PMUB; B05F640138), and the Research Grant for New Scholar from the Ministry of High Education, Science, Research and Innovation (RGNS63-178).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21541264.2023.2174765