The multifaceted roles of Myb domain–containing proteins in apicomplexan parasites

Abstract

Apicomplexan parasites are a large and diverse clade of protists responsible for significant diseases of humans and animals. Central to the ability of these parasites to colonize their host and evade immune responses is an expanded repertoire of gene-expression programs that requires the coordinated action of complex transcriptional networks. DNA-binding proteins and chromatin regulators are essential orchestrators of apicomplexan gene expression that often act in concert. Although apicomplexan genomes encode various families of putative DNA-binding proteins, most remain functionally and mechanistically unexplored. This review highlights the versatile role of myeloblastosis (Myb) domain–containing proteins in apicomplexan parasites as transcription factors and chromatin regulators. We explore the diversity of Myb domain structure and use phylogenetic analysis to identify common features across the phylum. This provides a framework to discuss functional heterogeneity and regulation of Myb domain–containing proteins particularly emphasizing their role in parasite differentiation.

I. Introduction

The phylum of apicomplexan parasites is a large and diverse clade of protists that is responsible for significant diseases of humans and animals [1]. Comparative studies with the closely related Chromera velia have shown that apicomplexans evolved from free-living phototrophic algae, experiencing gene reduction and adaptation to engage in parasitism [2,3]. As single-celled obligate parasites, apicomplexans have developed multiple morphologically distinct forms adapted to asexual replication or sexual recombination in one or more host species. Early branching apicomplexans such as gregarines (e.g. Gregarina niphandrodes) or Cryptosporidia (e.g. Cryptosporidium parvum) undergo both sexual and asexual replication in a single host. By contrast, sexual and asexual replication of other apicomplexans, like Haemosporida (e.g. Plasmodium spp.) and Piroplasmida (e.g. Babesia spp. and Theileria spp.), require alternating infections of vertebrate and invertebrate hosts, where asexual and sexual replication take place, respectively [4]. Other groups show greater flexibility. For example, the completion of the coccidian life cycle can be achieved in a single host (e.g. Eimeria tenella infections in poultry) or involve multiple host species (e.g. Toxoplasma gondii specifically exhibits sexual replication in the intestinal epithelium of felids) [5]. In order to maintain this intricate life cycle and adapt to constantly changing surroundings, apicomplexan parasites rely on distinct gene-expression programs that tune responses to varying nutrient availability and other environmental cues. For example, Plasmodium falciparum induces a protective heat-shock response during the febrile episodes of their hosts by changing the transcript levels of over 300 genes [6]. In addition, species such as Plasmodium vivax or T. gondii form dormant cell types that mediate persistence for months or years before reactivation [7,8]. Therefore, the ability of apicomplexan parasites to colonize their hosts, evade immune responses, and transmit to new hosts requires the coordinated action of numerous transcriptional regulators [9–11].

Transcriptional regulation in apicomplexan parasites

Transcriptional regulation plays a crucial role in orchestrating gene expression changes across apicomplexan life cycles [12–16]. Several features of the apicomplexan transcriptional machinery are shared with other eukaryotes, including factors associated with the RNA-polymerase, such as the TATA-binding protein (TBP) or TBP-associated factors, and basal transcription factors (TFs) that form the preinitiation complex at core promoter elements [17]. The accessibility of promoter elements is determined by chromatin structure, which can be regulated by chromatin-remodeling complexes and histone-modifying enzymes. In apicomplexan parasites, both conserved and unique histone modifications have been detected [18]. Combinations of histone marks and histone variants have been associated with transcriptional activation, repression, and elongation. Consequently, histone-modifications are under the strict control of several histone modifiers [9]. In T. gondii, the histone deacetylase HDAC3 interacts with a unique version of the MORC chromatin remodeler and other factors to repress sexual-stage transcripts. The MORC-HDAC3 complex and other chromatin remodelers are often guided to specific DNA motifs, directly or indirectly, through proteins harboring DNA-binding domains [19,20]. This makes DNA-binding proteins essential for orchestrating complex gene expression networks, ensuring proper cellular development, differentiation, and responses to environmental cues.

Apicomplexan parasites have lost many conventional eukaryotic DNA-binding proteins and their distinctive binding sites such as Forkhead box family TFs or basic helix-loop-helix TFs [21]. However, the ratio of TFs to proteome size is comparable between apicomplexans and other eukaryotes, due to the lineage-specific expansion of a single class of DNA-binding proteins: the plant-like Apetala2/ethylene response factor (AP2) TFs [22]. Many TFs exhibit lineage-specific expansions, suggesting considerable variability in transcriptional regulation and gene expression even among closely related eukaryotic lineages [21,23]. Apicomplexan AP2 (ApiAP2) domain–containing proteins (previously reviewed [24]) are important drivers of differentiation, parasite virulence, and cell-cycle regulation [25–27]. For example, AP2-G is the master regulator of sexual differentiation in P. falciparum while AP2IX-5 is a vital regulator of asexual cell division in T. gondii [25,26]. In addition to ApiAP2 domain–containing proteins, apicomplexan genomes encode various smaller families of DNA-binding proteins. These include Zn-finger, myeloblastosis (Myb), forkhead-associated, and high mobility group (HMG) DNA-binding domain–containing protein families [21]. Although some of these putative TFs have crucial functions throughout the parasite life cycle, most remain functionally and mechanistically unexplored. This review comprehensively explores the role of Myb domain–containing proteins (Mybs) in apicomplexan parasites to examine their structural features, functional diversity, and regulatory mechanisms. In so doing, we hope to highlight their emerging roles in apicomplexan gene expression.

II. General features of Myb domain–containing proteins

TFs are grouped into families based on sequence and structural similarity of their DNA-binding domains [28]. The Myb domain family was first recognized in the avian myeloblastosis virus, where v-Myb regulates host gene expression by binding to specific DNA sequences [29]. Since then, proteins encoding various numbers of Myb domain repeats have been identified across the spectrum of eukaryotic organisms, becoming one of the largest and most diverse family of transcriptional regulators. The prototypical member, the mammalian c-Myb, contains three Myb domains that are referred to as R1, R2, and R3 [30]. Based on the number of adjacent Myb domains, and their similarity to the repeats of c-Myb, four subfamilies have been defined: 1R-Mybs (Myb-related proteins), 2R-Mybs (R2R3-Mybs), 3R-Mybs (R1R2R3-Mybs), and 4R-Mybs (atypical Mybs) [31,32].

Diversity of Myb domain–containing proteins

The origins of Mybs can be traced back to the last eukaryotic common ancestor, together with components of the basal TF machinery, HMG domain–containing proteins, and Forkhead domain–containing proteins. Mybs can also be found in over 250 bacterial species, where they also affect gene expression. Multicellular plants often contain over 200 distinct Mybs—a majority are 2R-Mybs followed by 1R-Mybs—representing the principal family of TFs. By comparison, 3R-Mybs, which are common in animals, are rare in multicellular plants [33]. 3R-Mybs are involved in cell cycle regulation and differentiation in plants and animals, while 2R-Mybs are involved in various processes including metabolism, differentiation, and stress responses [34–36]. 1R-Mybs are highly heterogeneous in function and can be of type R3, which includes regulators of cellular morphogenesis and secondary metabolism, or of type R1/R2, which includes telomere-binding proteins or components of the circadian oscillator [32,37]. Additionally, several 1R-Mybs have recently been described as important modulators of plant adaptation to abiotic stresses [38]. 4R-Mybs are rare but conserved in a broad range of eukaryotes. One 4R-Myb was identified in plants acting as a subunit of the small nuclear RNA-activating protein complex [39].

Myb domain structure and function

Myb domains are typically N-terminal, followed by intrinsically disordered regions that vary considerably in sequence and length [40]. Intrinsically disordered regions are often found in eukaryotic TFs and mediate dynamic protein-protein interactions and promoter selection [41]. Myb domains have also been found in the C terminus of proteins including the 1R-Myb telomeric repeat–binding factors TRF1 and TRF2, which bind DNA with high affinity as homodimers [42]. Inter alia, characterization of LUX ARRHYTHMO (LUX), a member of the plant circadian clock evening complex, has shown that homodimerization is not required for efficient DNA binding of all 1R-Mybs. By contrast, two Myb domains are necessary for DNA binding of either 2R-Mybs or 3R-Mybs [43,44] (Fig.1A, B).

Figure 1. Classification of Myb and SANT domains.

A-B. Myb domain structures of Arabidopsis thaliana 1R-Myb transcriptional repressor LUX (A; PDB: 6QEC [43]) and 2R-Myb transcriptional activator Myb66 (B; PDB: 6KKS [46]). Myb domains are shown in green and DNA in white. Amino acid residues interacting with the major groove of the DNA are highlighted in red. The electrostatic-surface diagrams of the substrate interaction surface show mostly basic residues corresponding to a high pI. C. SANT domain structure of the Saccharomyces cerevisiae 1R SANT domain–containing Myb Ada2 highlighted in blue in complex with GCN5 in white (PDB: 6CW3 [99]). Amino acid residues that have been found to be important for interacting with GCN5 are highlighted in red. The electrostatic-surface diagram of the substrate-interaction surface shows mostly acidic and neutral residues corresponding to a low pI. Structures were generated using Chimerax-1.6.1. In the electrostatic-surface diagrams, residues are colored from red to white to blue, ranging from most acidic to most basic. D. Alignment of representative T. gondii and C. parvum Myb domains for each clade identified in Figure 2 colored according to their side-chain chemistry. Negatively charged residues are represented in red, positively charged residues in blue, polar residues in green, aromatic in purple and nonpolar residues are colored in gray. Consensus sequences for Myb and SANT domains are indicated above and below, respectively. Asterisks highlight conserved residues. Secondary structures are annotated according to [46,48]. Alignment generated with MEGA11 (version 11.0.13) using the MUSCLE implementation and the ggmsa package (version 1.6.0) in RStudio (version 2023.06.0+421) and R (version 4.3.0).

Generally, each Myb domain consists of approximately 50 amino acids that encode three alpha-helices. Most alpha-helices contain a conserved tryptophan residue important for establishing the hydrophobic core required for DNA binding [30]. In plants, the first tryptophan residue of R3 is generally replaced by another aromatic or hydrophobic residue such as phenylalanine or isoleucine [45]. The second and third helix of each Myb domain form a helix-turn-helix secondary structure, but only amino acids from the R2 and R3 helix-turn-helix bind DNA directly in a sequence-specific manner [30]. The amino acids that dictate sequence specificity vary between Mybs but often involve lysines and asparagines in R3 that form H-bonds with the nucleobases of the classic 5’-AMCNR-3’ DNA recognition motif (Fig. 1A, B). Other residues, such as leucine, play a role in sensing the DNA methylation status of the recognition motif. By contrast, R2 amino acids involved in DNA binding are variable which contributes to the range of target DNA sequence associated with 2R and 3R-Mybs, generally found in close proximity to the transcriptional start site (TSS) of target genes [46]. 1R-Mybs also exhibit a wide range of DNA target sequences [30]. Some 1R-Mybs bind DNA targets similar to those bound by 2R-Mybs and may therefore function as competitors [47].

The recognition of specific DNA sequence motifs is the principal function attributed to Mybs. However, a subgroup of 1R- and 2R-Mybs, collectively referred to as SANT domain proteins, facilitates protein-protein interactions. Several key residues that interact with DNA are not conserved in the SANT domain. In contrast to DNA binding Myb domains, the SANT domain has a relatively acidic isoelectric point (pI) and a negative electrostatic surface. These features make the SANT domain incapable of DNA binding [48]. SANT domains are commonly found in chromatin-remodeling and modifying complex subunits including Ada2 of the SAGA histone acetylation complex (Fig. 1C) and SMRT of the histone deacetylase complex. Deletion of the Ada2 SANT domain results in reduced histone-H3 binding and SAGA complex activity, while single amino acid changes in the first SANT domain of SMRT disrupt the binding to HDAC3 [49,50]. In this analysis, we aim to differentiate between putative DNA-binding Myb domains and SANT domains based on their pI. The pI values will be used together with conserved features, such as the glutamic acid in helix 3, to categorize the domains into putative DNA-binding Myb domains or putative SANT domains in apicomplexan parasites (Fig. 1D).

III. Myb domain–containing proteins in apicomplexan parasite biology

Myb domain–containing proteins are broadly distributed across apicomplexans. A systematic analysis of eight diverse apicomplexan genomes and a related free-living species (C. velia) identified 123 Mybs, including 96 1R-Mybs, 25 2R-Mybs, one 3R-Myb (TGME49_264120), and one 4R-Myb (cgd3_1120). Of the identified Mybs, 80 (72 1R-Mybs and eight 2R-Mybs) exhibit features of SANT domains including low theoretical pI values and conserved acidic amino acid residues [48]. In general, the number of identified Mybs scales with genome size. Accordingly, the 14-Mb genome of G. niphandrodes encodes 7 Mybs, the 23-Mb genome of P. falciparum encodes 10 Mybs, the 53-Mb genome of E. tenella encodes 14 Mybs, the 66-Mb genome of T. gondii encodes 15 Mybs, and the 194-Mb genome of C. velia encodes 23 Mybs. However, piroplasms like Babesia microti (6.4 Mb, 14 Mybs) and Theileria annulata (8.4 Mb, 14 Mybs), and Cryptosporidia like C. parvum (9.1 Mb, 16 Mybs) represent outliers with comparatively small genomes but numbers of Mybs mirroring those of the largest apicomplexan genomes (Fig. 2A, B). By contrast, ApiAP2 TFs are reduced in piroplasms, Cryptosporidia and gregarines, when taking into account their genome sizes [24,51,52]—although these findings are limited by the accuracy of the available genome annotation for each species. Even though Mybs generally occur at lower frequency compared to AP2s in apicomplexan genomes, they nevertheless represent a substantial proportion of the proteins expected to participate in gene regulation.

Figure 2. Comparison of apicomplexan Mybs.

A. Phylogenetic tree drawn with Myb domains (SM00717) identified through a homology-based search of a broadly representative group of apicomplexan species including T. gondii, E. tenella, P. falciparum, P. berghei, T. annulata, B. microti, C. parvum and G. niphandrodes and the related species C. velia with SMART 9.0 (e-value < 0.05) and filtering the results according to prior knowledge of Myb domain length (30–70 amino acids). The initial set of 110 Mybs was further enhanced by 13 additional Mybs using BLASTP against all analyzed species. Reciprocal BLASTP using the complete set of Myb domains derived from all 123 Mybs did not result in additional hits. A maximum likelihood (ML) phylogenetic tree of all Mybs was generated by concatenating all Myb domains for each Myb and aligning them using the MUSCLE implementation of MEGA11. The best model to construct the ML tree was calculated using ModelTest-NG version 0.2.0. The best scoring model (LG+2G) was used to construct the ML tree using MEGA11 with 1000 bootstrap replicates. Bootstrap values are represented by circles. The human c-Myb was used as an outgroup. Dotted lines represent shortened branches. Clades containing > 2 proteins with similar domain architecture as identified by SMART are indicated. Group 1 represents a dispersed class of spread out 2R-Mybs while group 2 refers to all Mybs which could not be assigned to any clade or group 1. CHY and SWIRM domains have not been identified by SMART and have been annotated according to HHPred analysis. Myb domains were differentiated from SANT domains by calculating the theoretical pI of the cumulative Myb domains per Myb using the R package seqinr version 4.2.30. Myb domains with a theoretical pI < 8.5 were considered SANT domains. This initial set of 53 putative SANT domain-containing proteins was further enriched by 27 proteins considering SANT-specific features including missing hydrophobic residues in helix 3 or a conserved glutamic acid in helix 3 identified by sequence alignment of the initial set of putative SANT domains. Putative DNA-binding Myb–containing clades are indicated in green while putative SANT domain–containing Myb clades are indicated in blue. Clade J contains SANT and Myb domains and is highlighted in light gray. Representative domain structures for each clade are shown in circles with domains not present in all proteins in the respective clade shown in brackets. B. Heat map indicating the number of Mybs identified in a clade for each species. Species ordered according to their evolutionary distance from T. gondii.

Of the 123 apicomplexan Mybs analyzed, 107 could be grouped into 13 clades representing specific domain architectures. The remaining Mybs can be assigned to two dispersed groups. Group 1 represents a class of 2R-Mybs containing Myb or SANT domains that do not occur as tandem repeats but are separated by more than 210 amino acids. Group 2 contains all Mybs that could not be assigned to any other group or clade (Fig. 2A, B). Five clades were fully conserved across all apicomplexan species analyzed (clades C, F, J, K, and L; Fig. 2B). Four of the five conserved clades contain putative SANT domains, indicating that protein-protein interacting SANT domains exhibit a higher degree of conservation compared with putative DNA-binding Myb domains.

SANT domains in chromatin regulator complexes

Many conserved SANT domain–containing Mybs are involved in the regulation of chromatin modifications and accessibility. Clade L Mybs share domain homology with the histone acetyltransferase (HAT) complex member Ada2, which typically contains an additional ZZ-type zinc finger (ZZ) domain (Fig. 2A). Metazoans encode two HAT complexes: the ATAC HAT complex that includes Ada2a and the SAGA HAT complex that includes Ada2b [53]. Both complexes target the lysine acetylase GCN5 to specific chromatin regions. Previous studies identified a single Ada2 ortholog in P. falciparum (PF3D7_1014600) and Plasmodium berghei (PBANKA_1213000) and two Ada2s in T. gondii (Ada2a: TGME49_217050 and Ada2b: TGME49_262420) [54,55]. TgAda2a interacts with TgGCN5b, while TgAda2b is able to interact with either TgGCN5a or TgGCN5b by yeast two-hybrid [55]. Interestingly, TgAda2b and TgGCN5a are dispensable under normal growth conditions; however, TgGCN5a knockout parasites have an impaired response to stress [56]. By contrast, TgAda2a and TgGCN5b are essential for parasite viability under normal growth conditions, as is the case with PbAda2 [57–59]. Pulldown experiments of TgGCN5b could not identify any other conserved ATAC or SAGA components besides TgAda2a. Instead, the TgGCN5b-Ada2a core complex associates with multiple ApiAP2 TFs, some of which have been found to be specific to the parasite’s stress response [57]. B. microti, T. annulata, and E. tenella also encode two Ada2 orthologs from clade L. Even though the putative ZZ domain of the Ada2b ortholog ETH2_1202600 was of low confidence (e-value: > 0.05) and therefore disregarded, the existence of two Ada2 paralogs in E. tenella is supported by the presence of two GCN5 histone acetyltransferases identified by homology to T. gondii GCN5s. By contrast, C. velia appears to have no SANT domain–containing Ada2 orthologs despite two proteins containing Ada2 like ZZ domains (Cvel_18768 and Cvel_16886), perhaps reflecting issues with the current gene models.

Clades E and K contain homologs of subunits belonging to chromatin remodeling complexes that typically utilize ATP hydrolysis to alter the packing state of chromatin by moving, ejecting, or restructuring nucleosomes [60]. Clade K is of type SWI3 while clade E is of type ISW1 and lacks orthologs in G. niphandrodes, P. falciparum, and P. berghei (Fig. 2A). SWI3 represents a noncatalytic subunit that dimerizes and forms the backbone of the SWI/SNF chromatin remodeling complex together with SWI1. During this process, the SANT domain of SWI3 interacts with various components of the complex while the SWIRM domain contacts the nucleosome-binding module [61]. SWI3 (TGME49_286920) was previously identified in T. gondii during an insertional mutagenesis screen for mutants impaired in chronic differentiation. When parasites were exposed to alkaline stress, the TgSWI3 mutant displayed reduced transcript levels for some bradyzoite-induced genes while others remained unaffected [62]. In contrast to clade K, members of the ISW1 type clade E contain DEXDc+HELICc domains which are well conserved in eukaryotes and characteristic of the ATPase subunits of chromatin remodeling complexes (Figure 2a). These can be further classified based on the presence of additional domains whereby the imitation switch (ISWI) subfamily ATPases ISW1 and ISW2 are the only ATPases containing an additional SANT domain—involved in SNF2 binding and stabilization—but no chromodomain like the chromodomain helicase DNA-binding (CHD) subfamily [63]. ISWI complexes exist in various variants that are often directed by DNA-binding proteins and involved in multiple aspects of cell physiology [64]. In T. gondii, TGME49_321440 was identified as a member of the ISWI subfamily but has not been further characterized [65]. Despite lacking the characteristic C-terminal SANT domain, PF3D7_0624600 has been described as the P. falciparum ISWI, based on the homology of its DEXDc+HELICc domains to other ISWI subfamily members [66]. However, the presence of N-terminal plant homeodomain (PHD) finger and chromo domains (identified by HHPred [67]) suggests that PF3D7_0624600 instead belongs to the CHD family of chromatin remodeling complexes. SANT domains play a central role in the regulation of chromosomal DNA accessibility by mediating the interaction of different chromatin regulator complex components. However, some SANT domain–containing proteins contribute to transcriptional regulation through non-chromatin regulatory mechanisms.

Non-chromatin regulatory Mybs with SANT domains

Mybs with SANT domains that have not been found to participate in chromatin regulation include predominantly 1R-Mybs in apicomplexan parasites, many of which lack co-domains and have yet to be characterized. There are two well conserved clades of SANT/Myb domain–containing 1R-Mybs that include either a CHY zinc finger domain (clade J) or a HMG domain for some Mybs (clade C) (Fig. 2A). CHY zinc finger domains can be found in most eukaryotic species but remain functionally uncharacterized [68]. Notably, several apicomplexan CHY domain–containing Mybs, including TGME49_306040 and PBANKA_0831100 of clade J, have been found to be dispensable in asexual stages [58,59]. By contrast, clade C Mybs, some of which contain a HMG DNA-binding domain (e.g., TGME49_203950 and PF3D7_1205800) are essential in asexual stages [58,59]. These domains are widespread in nonhistone chromosomal proteins involved in DNA and nucleosome binding [69]. PF3D7_1205800 is annotated as PfHMGB3 based on homology to PfHMGB1 [70]. In a study that performed DNA pull-down experiments using DNA motifs enriched in chromatin regions showing stage-specific accessibility, PfHMGB3 was enriched at the motif 5’-GCACTTTTATTGCA-3’ together with TF AP2-I and chromatin-related factors such as BDP1-3 and SWIB. Several of these factors are conserved from yeast to humans and have been found to form a complex in P. falciparum where they control the expression of genes related to invasion [71–73]. In P. berghei the 1R-Myb PBANKA_0604500 is annotated as PbHMGB3; however, both HMG domains are predicted with low confidence by SMART analysis (e-value: > 0.05), and have therefore been disregarded in our analysis. We did not identify any other low-confidence HMG domains in other Mybs of the analyzed species including clades B, D and H which do not contain any co-domains. In apicomplexan parasites, all SANT domain–containing clades are single-repeat Mybs with the exception of clade I.

Clade I 2R SANT domain configurations are specific to DNAJC in apicomplexan parasites and C. velia (Fig. 2A, B). All analyzed species with the exception of G. niphandrodes contain exactly one DnaJ domain–containing Myb. The conserved N-terminal DnaJ domain is essential for interacting with the ATPase domain of DnaK, stimulating its activity and target affinity. The DnaJ-DnaK system is involved in folding and translocation of nascent polypeptides, stress responses, and protein degradation. DnaK usually has more than one DnaJ interacting partner which can be of class A, B or C depending on other structural features. A DnaJ domain in combination with SANT domains has mostly been found in DNAJC1 and DNAJC2, which are typically localized in the nuclear or ER membrane, or the nucleus and cytosol, respectively [74]. Both DNAJC1 and DNAJC2 have been shown to act as molecular chaperones for folding nascent polypeptides. In addition, DNAJC2 is also involved in transcriptional activation by displacing polycomb-repressive complex 1 at the beginning of cellular differentiation through its binding of ubiquitinated H2A with its R2 SANT domain [74–76]. Furthermore, the R2 SANT domain of DNAJC1 has been shown to bind α₁-antichymotrypsin demonstrating that SANT domains can also act as molecular sensors [76]. Thus, despite the similarity between SANT and DNA-binding Myb domains, SANT domain–containing Mybs have undergone functional diversification distinct from DNA binding. In apicomplexan parasites, SANT domains can be found in various types of proteins which highlights their multifaceted roles in a wide range of biological processes.

DNA-binding Mybs have a heterogenous domain architecture

Four clades of Mybs (A, F, G, and M) exhibit the amino acid composition and theoretical pI values consistent with DNA binding (Fig. 2A). Clades A and F contain 1R- and 2R-Mybs, respectively. We could not identify other common features of clade A Mybs; however, a recent study identified a missense mutation outside of the Myb domain of ETH2_1257900 in a monensin-resistant strain of E. tenella, demonstrating the existence of functional sites outside the Myb domain [77]. Another clade A Myb, TGME49_306320, was found to be regulated in a cell-cycle dependent manner with elevated transcript levels after prolonged passage of T. gondii in cell culture [78,79].

Based on homology, cluster F represents orthologs of CDC5L, which is highly conserved in eukaryotes and participates in diverse processes [80] (Fig. 2B). For example, the plant CDC5L is a positive regulator of cell cycle G2/M progression while also regulating the transcription of miRNAs by interacting with RNA-pol II [81]. CDC5L is also an essential member of the Prp19 non-small nuclear ribonucleoprotein (snRNP) splicing complex, supporting specific spliceosomal conformations. Interestingly, the Myb domains of CDC5L simultaneously bind DNA and the spliceosomal RNA U6, which stabilizes the catalytic U2/U6 RNA structure of the major U2-type spliceosome [82]. While the minor U12-type spliceosome and several snRNA-trafficking proteins are absent in apicomplexan parasites, the U2-type spliceosome has acquired apicomplexan-specific features including the unusually long 3’ poly(A) extension of spliceosomal RNAs and a different spliceosomal core complex organization [83–85].

In contrast to clades A and F, clades G and M are heterogeneous in their Myb domain architecture and possess either 2–4R-Mybs or 1–3R-Mybs, respectively (Fig. 2A). The presence of Mybs with different numbers of repeats in the same clade suggests that these Mybs have diverged from one another through domain loss or gain from an ancestral Myb [39]. These Mybs have likely evolved species-specific functions and include several that have been identified as key regulators for apicomplexan differentiation.

IV. Myb domain–containing proteins as regulators of apicomplexan differentiation

Differentiating between cell fates requires the orchestration of networks of transcriptional regulators. Within this regulatory landscape, a single node often assumes the role of a master regulator, exerting dominant control over the fate and identity of cells, which makes it necessary and sufficient for differentiation [86,87]. Several Myb domain–containing proteins have been identified as key transcriptional regulators or master regulators in apicomplexan parasites.

In P. falciparum, Myb1 (PF3D7_1315800) has been characterized as an essential TF during intraerythrocytic development [88]. Knockdown of PfMyb1 results in a 40 % reduction of parasite growth and subsequent mortality during the differentiation from trophozoite to schizont stages. PfMyb1 target genes are also involved in cell cycle regulation such as calcium dependent kinases [89]. PfMyb1 has no orthologs outside the Plasmodium genus and was therefore assigned to group 2 (Fig. 2A). Interestingly, both PfMyb1 and its P. berghei ortholog (PBANKA_1414300) contain one and two low-confidence Myb domains (e-value: > 0.05) in addition to their single high confidence Myb domain, respectively. While PfMyb1 regulates asexual development in Plasmodium, Myb-M (cgd6_2250) of clade G (Fig. 2A) has recently been shown to be a key regulator of the male gamete transcriptional program in C. parvum. CpMyb-M is specifically expressed in early male gamonts. Conditional overexpression of CpMyb-M overrides the time-dependent developmental program of C. parvum and induces the formation of non-proliferative male gametes. Knockout attempts have not been successful, suggesting that CpMyb-M may not only be sufficient but also required for male gamete formation (Katelyn Walzer and Boris Striepen, personal communication). The characterization of both PfMyb1 and CpMyb-M highlights the important roles putative DNA-binding Mybs play in orchestrating apicomplexan differentiation.

To date, T. gondii BFD1 (TGME49_200385) represents the only Myb in apicomplexan parasites characterized as a bona fide master regulator, being both necessary and sufficient for the reprogramming of cell fate (Fig. 2A). Identified via a Cas9-mediated screen, TgBFD1 was found to be necessary for the differentiation of proliferative stage tachyzoites to dormant stage bradyzoites. Parasites lacking TgBFD1 fail to establish chronic infections in mice but grow normally under standard conditions. In addition, the overexpression of TgBFD1 is sufficient to drive parasite differentiation into bradyzoites [14]. In accordance with its function as a TF, TgBFD1 targets the 5’-CACTGG-3’ motif closely resembling the classic DNA recognition motif of other 2R-Mybs [14,46]. Like most cis-regulatory sequences, the TgBFD1-binding motif is enriched in the nucleosome-depleted region of genes upregulated in bradyzoites with the highest motif frequency ~98 bp upstream of the TSS [90]. Therefore, TgBFD1 is expected to directly activate transcription of bradyzoite-specific genes, including the commonly used bradyzoite marker BAG1 and metabolic enzymes such as LDH2 or ENO1. In accordance with its function as a master regulator, TgBFD1 also binds upstream of its own promoter as well as upstream of other known bradyzoite specific TFs such as AP2IX-9 [14]. Although the molecular mechanism of TgBFD1-dependent transcriptional activation remains undefined, its comprehensive characterization represents an important milestone for the role of Mybs in apicomplexan differentiation.

V. Regulation of Myb domain–containing proteins

Myb domain–containing proteins are heterogeneous in their function and domain composition and are therefore highly regulated. Many putatively DNA-binding Mybs are differentially expressed between apicomplexan life cycle stages, including CpMyb-M which is specifically expressed in early gamonts and PfMyb1 whose expression level peaks in P. falciparum trophozoite stages [88] (Katelyn Walzer and Boris Striepen, personal communication). However, other processes must influence the activity of master regulators since they are, by definition, at the top of the transcriptional hierarchy. Therefore, master regulators require translational or post-translational regulation [86]. TgBFD1 is transcribed throughout the T. gondii asexual cycle but can only be detected at the protein level in stressed or differentiating parasites [14]. Two independent studies identified the CCCH-type zinc-finger RNA binding protein TgBFD2 (also known as TgROCY1), as a regulator of TgBFD1 translation. TgBFD2 interacts with the TgBFD1 messenger RNA specifically under stress conditions to facilitate its translation. Reciprocally, TgBFD1 has been found to increase TgBFD2 transcription, leading to a positive feedback loop that enforces the bradyzoite-specific gene-expression program [91,92]. Similar positive-feedback loops involving master regulators are often found at the center of cell fate decisions [93]. Interestingly, TgBFD2 is a target of protein phosphatase 2A (PP2A) which also targets TGME49_237520, a 1R SANT domain–containing Myb that is part of the uncharacterized clade D (Fig. 2A). Disruption of TgPP2A leads to increased starch accumulation and blocks T. gondii differentiation to bradyzoites implying that TGME49_237520 plays a role in parsite metabolism and differentiation [94].

Post-translational modifications (PTMs) such as phosphorylation, sumoylation and O-GlcNAcetylation often orchestrate the activity of TFs by changing their DNA-binding specificity, subcellular localization, degradation, or ability to interact with other proteins [95]. In Mybs, PTMs mostly occur in the more variable intrinsically disordered regions where they can positively or negatively influence TF activity [96]. A well-characterized example includes c-Myb, which becomes inactive due to conformational changes when phosphorylated at the C terminus. Consequently, the introduction of a phospho-null mutation increases the transcription of a subset of its target genes [97]. This demonstrates that mere access to DNA does not determine sequence specificity of Myb domains. An additional layer of regulation occurs through interactions with co-factors and chromatin remodelers. The P. falciparum SAGA complex interacts with several TFs in a stage-specific manner to activate transcription. Analogously, the T. gondii MORC-HDAC3 complex has been shown to interact with AP2XII-1 and AP2XI-2 to repress the transcription of merozoite transcripts [19,20,98]. Although many Myb domain–containing proteins bear homology to specific chromatin regulator domains or TFs, their roles in apicomplexan transcriptional regulation remains largely unexplored.

VI. Conclusion

This overview of Myb domain–containing proteins in apicomplexan parasites has focused on their structural features, functional diversity and regulation. Our phylogenetic analysis provides a framework for understanding the evolutionary relationships of Myb domain–containing proteins. We have identified three clades of Mybs that seem to be involved in chromatin regulation. The SAGA complex, which contains one of the Mybs described, has been previously examined while SWI/SNF and ISWI chromatin remodeling complexes remain poorly understood in apicomplexan parasites. Five additional clades contain Mybs featuring SANT domains including Mybs with CHY, HMG and DnaJ co-domains. In P. falciparum, several HMG domain–containing Mybs are enriched in chromatin regions displaying stage-specific accessibility, while DnaJ domain–containing proteins are well conserved and have been characterized in eukaryotes. Four clades of Mybs, putatively act as DNA-binding proteins, which include orthologs of CDC5L but also apicomplexan specific TFs that play important roles in various parasitic stages; among these, the master regulator for bradyzoite differentiation TgBFD1, but also important regulators of P. falciparum blood-stage development and C. parvum male gamete formation. Despite the important roles of many Mybs during the apicomplexan life cycle and the identification of many as critical for parasite fitness, this family of putative transcriptional regulators remains poorly understood. Further research on Myb domain–containing proteins holds the potential for advancing our understanding of transcriptional regulation in apicomplexan parasites and developing therapeutic interventions targeting parasitic diseases.

Supplementary Material

1

Table S1. SMART and BLASTP search parameters and domain structure characteristics in identified Mybs, related to Figure 2. SMART and BLASTP search parameters and results for Myb domain (SM00717) containing proteins of the apicomplexan parasites T. gondii, E. tenella, P. falciparum, P. berghei, T. annulata, B. microti, C. parvum and G. niphandrodes and the related species C. velia.