by Ann Yang Harvard College ‘22
ABSTRACT: Eukaryotic transcription is a complex process that requires the precise and orchestrated recruitment of many initiation, elongation, and termination factors. Many of these transcription factors are not structurally or functionally well understood. In particular, the elongation factor Spt6 has broad and diverse roles in regulating transcription which have been studied to varying degrees of detail. While many of the functions of Spt6 have been elucidated, many of the finer mechanisms of these functions remain unknown, making Spt6 a particularly key protein of interest for transcription studies. This review summarizes recent studies of the many roles of Spt6 in regulating transcription with an emphasis on the control of chromatin structure as a mechanism of transcriptional regulation. Additionally, interactions between Spt6 and several key proteins are discussed in the context of transcriptional regulation. The relationship between Spt6 and the relatively unexplored elongation factor Elf1 is highlighted and structurally analyzed for possible direct interaction. Structural analysis approximately reconstitutes the in vivo spatial relationship between Spt6 and Elf1, thus providing insight on the likelihood of a physical interaction between these proteins. As part of examining the Spt6-Elf1 relationship, this review also covers the characterization of Elf1 and potential roles for Elf1 independent of Spt6. Our current knowledge of Spt6 is extensive, but future experiments are necessary to establish the connections between Spt6 and less-studied proteins such as Elf1, particularly for analyzing indirect interactions which may occur through multiprotein pathways.
Introduction
The budding yeast Saccharomyces cerevisiae is the leading model organism for studying gene expression. S. cerevisiae is a single-celled eukaryotic organism, able to divide rapidly under simple growth conditions. However, the true power of yeast as a model system is in the ease with which its genes and gene expression can be mutated and controlled either by plasmids or the yeast chromosomes (Duina et al., 2014). As a result, yeast is a powerful and easily manipulatable model system for investigating transcription. Furthermore, given the high degree of conservation among eukaryotes, yeast is an excellent system for understanding transcription in higher eukaryotes as well. Regulation of transcription at the genetic level is crucial for proper gene expression and cell maintenance. Transcription in eukaryotic cells is a complex process that requires the recruitment of many transcription factors. RNA polymerase II (RNAPII), aided by these factors, transcribes DNA into RNA. This process can be divided into three distinct phases: initiation, elongation, and termination (Venkatesh and Workman, 2015). During initiation, RNAPII recognizes and binds to promoter sequences with the help of general transcription factors (GTFs). These GTFs unwind double-stranded DNA at the transcription start, thus forming a transcription bubble and initiating RNAPII activity. After initiation, these factors are replaced with elongation factors which promote RNA synthesis during the elongation phase. As the RNAPII elongation complex moves along the template strand, DNA is unwound to allow for transcription of the pre-mRNA. However, as the complex passes, the pre-transcription chromatin structure is re-established (Jonkers and Lis, 2015). Elongation is followed by termination when RNA is released from RNAPII and RNAPII is released from the DNA template (Venkatesh and Workman, 2015). Regulation of this process is necessary for ensuring proper cell function and viability and may be accomplished by a variety of methods including activity of regulatory factors, DNA modifications, and control of chromatin structure (Jonkers and Lis, 2015; Venkatesh and Workman, 2015). The yeast genome is packaged as chromatin. In eukaryotic cells, DNA is wound ~1.7 times around an octamer of histone proteins to form nucleosomes. The core structure consists of two copies of each histone: H3, H4, H2A and H2B. Nucleosomes in turn coil and stack to form chromatin fibers which loop and fold to form chromosomes (Luger et al., 1997; Venkatesh and Workman, 2015). Typical haploid S. cerevisiae cells contain ~12,000 kb of genomic DNA organized this way into 16 chromosomes (Duina et al., 2014). Chromatin is a dynamic structure. In addition to efficiently packaging the genome into the nucleus, chromatin structure plays key roles in controlling multiple cellular processes such as transcription, recombination, and DNA replication by regulating the accessibility of DNA to the machinery involved in each of these processes. Both biochemical and genetic studies have advanced our understanding of how chromatin structure regulates transcription. For example, in vitro experiments revealed the barrier that nucleosomes pose to transcription initiation and elongation. In vivo, genetic methods identified roles for nucleosomes and histones in regulating transcription. Yeast studies employing both approaches have been instrumental in advancing our knowledge of chromatin, transcription, and their relationship (Rando and Winston, 2012). Proper regulation of transcription elongation involves the recruitment of many regulatory factors. During elongation, RNAPII forms a complex with multiple regulatory factors such as the PAF complex, Dst1, Spt4/5, Spt6, Spn1, Elf1, and FACT (Schier and Taatjes, 2020). Several of these protein factors have roles in processing nascent RNA, activating RNAPII, modifying chromatin, and controlling chromatin structure (Buratowski, 2009; Joo et al., 2019). Advancements in structural biology along with mutagenesis and functional assays have increased the extent to which the RNAPII elongation complex has been characterized. However, many of the dynamics of elongation and the factors involved in regulating this process still remain unclear. The class of factors that regulate transcription through control of chromatin structure operate through a diverse set of pathways. For example, the chromatin remodeler Swi/Snf promotes transcription by altering histone-DNA interactions to remove nucleosome barriers to transcription. The chromatin modifying complex SAGA is involved in histone H3 acetylation and histone H2B ubiquitylation during elongation. Histone chaperones FACT and Spt6 interact directly with nucleosomes to alter chromatin structure. FACT is a complex composed of two proteins, Spt16 and Pob3, that functions in nucleosome reassembly and disassembly. Spt6 is a large protein with many domains and interactions that exhibits broad and varied effects on chromatin structure (Rando and Winston, 2012). This paper reviews studies covering major findings on the roles of Spt6 and a related factor, Elf1, in regulating transcription with an emphasis on these factors’ control of chromatin structure. In particular, the role of Spt6 in regulating transcriptional elongation through nucleosome dynamics is highlighted. While Spt6 is known to regulate nucleosome dynamics, many of the mechanisms of this regulation are not yet known. It is the goal of this review to organize the current mechanistic models of Spt6 regulation of nucleosome dynamics. The majority of the studies reviewed use yeast as a model system. While studies employing other model systems have been crucial to informing our current understanding of Spt6 and Elf1, many of the key discoveries discussed in this paper come from yeast studies. However, given the high degree of conservation among eukaryotes, studies of larger eukaryotes are included when appropriate.
The Many Roles of Spt6
Spt6 is a multifunctional transcriptional regulator. Spt6-mediated chromatin effects have consequences for transcription. One possible way that Spt6 regulates transcription is by controlling chromatin structure, thereby affecting accessibility of DNA to elongating RNAPII complexes and subsequently increasing or decreasing levels of transcription. Both in vitro and in vivo experiments have shown Spt6 to affect nucleosome occupancy and nucleosome dynamics. Furthermore, Spt6 interactions with histones, nucleosomes, RNAPII, the elongation factor Spn1, and casein kinase II (CKII) likely mediate these changes in chromatin structure. However, many of the mechanisms operating behind Spt6-dependent changes in chromatin structure remain to be elucidated. Additionally, while not the focus of this review, Spt6 may also regulate transcription directly. It should be noted that Spt6 regulation of chromatin structure and transcription are not mutually exclusive. Spt6 is a key regulator of elongation with central roles in mediating both chromatin structure and transcription.
Interaction with RNAPII
Spt6 may directly regulate transcription through interactions with RNAPII. While the role of this interaction and its significance for transcription is unclear, Spt6 is well-established as an elongation factor that travels with elongating RNAPII. Studies on the mammalian and Drosophila orthologues of Spt6 characterize it specifically as a positive regulator of transcription elongation (Andrulis et al., 2000; Ardehali et al., 2009; Endoh et al., 2004; Kaplan et al., 2000). Furthermore, these studies also identified direct interactions between Spt6 and RNAPII. The interaction between Spt6 and RNAPII is mediated by multiple domains, modifications, and factors. C-terminal tandem Src homology 2 (tSH2) domains in yeast Spt6 are crucial for this interaction (Close et al., 2011; Diebold et al., 2010; Yoh et al., 2007). Additionally, phosphorylation of key residues in the C-terminal domain (CTD) of RNAPII promotes recruitment of Spt6 to transcribed genes (Brázda et al., 2020; Burugula et al., 2014; Yoh et al., 2007). Furthermore, the Spt6-RNAPII interaction may be partially mediated by the histone deacetylases Rpd3 and Hos2 (Burugula et al., 2014). Recently the linker region of the Rpb1 subunit of RNAPII was found to bind the tSH2 domains of Spt6 while the C-terminal residues were dispensable for binding Spt6. Phosphorylation of residues S1493, T1471, and Y1473 in this linker region promotes Spt6 binding to Rpb1 (Sdano et al., 2017). However, these findings are not necessarily in contradiction to the previous studies establishing the interaction between Spt6 and the CTD of RNAPII. While the linker region may serve as the primary binding site, the tSH2 domains of Spt6 also bind the RNAPII C-terminal domain in vitro. This Spt6-CTD interaction may be important for stabilizing recruitment of Spt6 or retaining Spt6 at transcribed genes, which is consistent with the Rpb1 linker recruiting Spt6 (Brázda et al., 2020; Sdano et al., 2017). The different interactions between Spt6 and RNAPII may regulate different functions necessary for recruitment of Spt6 to RNAPII. The Spt6-RNAPII interaction plays roles in regulating chromatin structure. While Spt6 may regulate elongation directly through interacting with RNAPII, this interaction also has important consequences for chromatin structure. Recruitment of Spt6 by Ser2 in the CTD of RNAPII has been proposed to regulate histone occupancy at the 5’ end of coding regions, and facilitation of this recruitment by histone deacetylases is consistent with a model in which Spt6 reassembles histones displaced by RNAPII elongation (Burugula et al., 2014). Interaction between Spt6 and the Rpb1 linker also affects chromatin structure. Mutation of Rpb1 S1493 is associated with a phenotype indicative of an active promoter that is normally repressed by the local chromatin structure. This defect in chromatin structure suggests a role for the Spt6-Rpb1 interaction in maintaining repressive chromatin structures during transcription (Sdano et al., 2017). These findings are consistent with a role for Spt6 in reassembling nucleosomes after passage of an elongating RNAPII molecule. Whether this interaction regulates chromatin structure directly or indirectly is uncertain. However, there is evidence to suggest that the Spt6-RNAPII interaction facilitates recruitment of hIws1, the human analog of the factor Spn1, a conserved transcription factor implicated in recruitment of chromatin-remodeling factors (Yoh et al., 2007; Zhang et al., 2008). Thus, there are multiple mechanisms by which the Spt6-RNAPII interaction may control chromatin structure, either directly in the wake of elongation or through recruitment of an additional factor.
Effects of Spt6 on Nucleosome Occupancy and Dynamics
Spt6 affects nucleosome occupancy. Loss of Spt6 in S. cerevisiae results in decreased nucleosome occupancy over coding regions, preferentially over highly transcribed genes (Doris et al., 2018; Ivanovska et al., 2011). Furthermore, experiments done using an spt6 mutant allele in the closely related fission yeast, Schizosaccharomyces pombe, also found disrupted nucleosome positioning and occupancy over highly transcribed regions, including a general loss of the +1 nucleosome. However, nucleosome loss was not found to correlate with changes in levels of mRNA transcripts of these genes, suggesting a non-regulatory function for Spt6 over transcribed regions (DeGennaro et al., 2013). Conversely, analysis of the regulatory region of CHA1, a gene repressed by Spt6 activity, shows a shift in nucleosome positioning and de-repression of the gene under rich growth conditions, suggesting a role for Spt6 in remodeling the +1 nucleosomes at regulatory regions of coding sequences (Ivanovska et al., 2011). In addition, genome-wide mapping indicates a role for Spt6 in maintaining 5’ and 3’ nucleosome-depleted regions (NDRs) and maintaining nucleosomes in the regions flanking these NDRs (Perales et al., 2013). These effects on nucleosome occupancy indicate a transcription-dependent remodeling of nucleosomes by Spt6. One possible explanation for decreased nucleosome occupancy over transcribed genes in spt6 mutants is a role for Spt6 in mediating nucleosome reassembly and disassembly. Current evidence supports a model in which Spt6 functions in nucleosome reassembly in the wake of transcription. In this model, over regions of low or moderate transcription, nucleosomes should have sufficient time to reassemble before passage of the next RNAPII molecule, thereby maintaining nucleosome levels. However, at highly transcribed genes, nucleosomes do not have sufficient time to reassemble nucleosomes before disassembly upon passage of the next RNAPII molecule, resulting in reduced nucleosome occupancy at these genes. Reduced nucleosome occupancy seen in spt6 mutants and slowed nucleosome reassembly rates are consistent with this model (Adkins and Tyler, 2006; Kaplan et al., 2003). Furthermore, Spt6-mediated nucleosome dynamics have consequences for transcriptional regulation. Spt6 plays roles in maintaining chromatin structure during elongation, thus repressing initiation of cryptic transcription from within coding regions (DeGennaro et al., 2013; Kaplan et al., 2003). One way Spt6 may repress cryptic transcription is by preventing accumulation of histone variant H2A.Z from gene bodies, as this accumulation can contribute to cryptic transcription (Jeronimo et al., 2015). Nucleosome reassembly at promoters was also found to require Spt6 and function in repression of transcription of the corresponding genes (Ivanovska et al., 2011; Perales et al., 2013). Thus Spt6 likely reassembles nucleosomes, either directly or indirectly, in the wake of transcription elongation.
Spt6 Interaction with Histones
Spt6 may mediate nucleosome assembly through interaction with histones. Spt6 is classically characterized as an H3-H4 histone chaperone. An spt6 mutation was found to confer the same alterations to chromatin structure as a deletion of one pair of the genes encoding histones H2A-H2B, suggesting the possibility of an interaction between Spt6 and histones. In vitro, Spt6 binds to H3-H4 and H2A-H2B, indicating that Spt6 can interact directly with each histone. Furthermore, overexpression of H3-H4 and H3 alone, but not overexpression of H4 alone, suppresses spt6 phenotypes. These results are consistent with the strong direct interaction between Spt6 and H3 detected in vitro; the weaker in vitro interaction between Spt6 and H4 is suggestive of a weak in vivo interaction (Bortvin and Winston, 1996). Mutations in histones H2A and H2B are also able to suppress phenotypes conferred by an spt6-F249K mutation (McCullough et al., 2015). Thus, Spt6 interacts with H3-H4 and H2A-H2B in vivo. Given the necessity of histones for nucleosome formation, Spt6 interaction with histones is likely to function in assembling nucleosomes. The histone-Spt6 interaction may function in nucleosome assembly. Spt6 is able to assemble nucleosomes in vitro in the presence of histones, suggesting a role for the histone-Spt6 interaction for nucleosome assembly in vivo (Bortvin and Winston, 1996). Mutations of residues in the globular domain of histone H3 that interact with Spt6 are associated with reduction of nucleosome occupancy at highly transcribed regions (Hainer and Martens, 2011, 2016), and slowed nucleosome dynamics (Hainer and Martens, 2016). Additionally, H3 mutants fail to suppress mutant spt6 phenotypes, and co-immunoprecipitation experiments indicate a reduced interaction between Spt6 and H3 in H3 mutants (Bortvin and Winston, 1996; Hainer and Martens, 2016). Thus, this reduced interaction is associated with defects in nucleosome occupancy, reassembly, and disassembly, which supports a role for the Spt6-H3 interaction in regulating these processes in wild-type cells. One mechanism that has been proposed is Spt6 interacting with H3 to alter histone-DNA interactions, thus regulating chromatin structure. However, alternatively, the reduced Spt6-H3 interaction may be the result of changes in nucleosome dynamics in the H3 mutants (Hainer and Martens, 2016). Future experiments will be necessary to distinguish the reduced Spt6-H3 interaction as the cause or effect of altered nucleosome dynamics in H3 mutants. Furthermore, while H2A-H2B interacts with Spt6, the effects of this interaction on nucleosome dynamics have not been elucidated (McCullough et al., 2015). Thus, whether Spt6 mediates nucleosome assembly through an interaction with H2A-H2B is unknown. Altogether, these findings suggest a role for the interaction between histones and Spt6 in nucleosome assembly. The interaction between Spt6 and H3 is associated with defects in nucleosome occupancy and dynamics. However, the mechanistic role of this interaction in promoting nucleosome assembly is unclear. Furthermore, an interaction between Spt6 and H2A-H2B may present another pathway through which Spt6 mediates nucleosome assembly, although this interaction has not been extensively studied in the context of chromatin structure. Ultimately, the defects in nucleosome occupancy and assembly seen in spt6 and histone mutants is consistent with a role for histone-Spt6 interactions in maintaining normal chromatin structure. Future experiments aimed at determining the mechanistic basis for this assembly of nucleosomes will be necessary to fully understand how histones affect the ability of Spt6 to assemble nucleosomes.
Spt6 Interaction with Spn1
The Spt6-Spn1 interaction is implicated in regulating chromatin structure. Structural analysis suggests that nucleosomes and Spn1 compete for binding to Spt6 in vitro. Therefore, the Spt6- Spn1 interaction may act to regulate the interaction of Spt6 with nucleosomes, suggesting important consequences for maintaining chromatin structure (McDonald et al., 2010). Interestingly, H2A and H2B mutations at the dimer interface that destabilize the nucleosome are able to suppress phenotypes seen in an spt6 mutant that has reduced binding to Spn1. These results support a model in which the destabilized nucleosome activates another system of chromatin maintenance that bypasses the need for the Spt6- Spn1 interaction (McCullough et al., 2015). Furthermore, Spn1 regulates recruitment of Spt6 during transcription. Analysis of Spt6-Spn1 dynamics at the inducible CYC1 gene show defects in Spt6 recruitment when Spn1 occupancy is decreased. Spt6 is only recruited upon induction of high transcription levels, and Spt6 is recruited before the chromatin-remodeling complex Swi/Snf. In the absence of Spn1, Swi/Snf is constitutively recruited. Thus, Spn1 may inhibit recruitment of Swi/Snf, an inhibition that is itself abolished by binding of Spt6 to Spn1 (Zhang et al., 2008). Altogether, the interaction between Spt6 and Spn1 likely has roles in regulating chromatin structure, either directly, through activation of alternative pathways, or through regulating recruitment of other factors implicated in controlling chromatin structure.
Spt6 Interaction with CKII
Spt6 interacts with casein kinase II (CKII), an essential protein kinase involved in many cellular processes. CKII has roles in modulating chromatin structure through regulation of histone H3 and key transcription factors, such as Spt2, the PAF complex, and the FACT complex in addition to Spt6. For example, phosphorylation of Spt2 by CKII disrupts the interaction between Spt2 and Spt6, thereby promoting the association of Spt2 with coding regions and repressing spurious transcription (Gouot et al., 2018). CKII interacts with and phosphorylates the N terminus of Spt6 both in vivo and in vitro. This interaction has consequences for chromatin structure and transcription (Dronamraju et al., 2018; Gouot et al., 2018) The Spt6-CKII interaction has roles for regulating chromatin structure and transcription. Strains expressing spt6 mutations defective for phosphorylation by CKII show defects in maintaining nucleosome occupancy (Dronamraju et al., 2018). Furthermore, analyses using the reporter system pGAL1- FLO8-HIS3 (Gouot et al., 2018) and the well-characterized SRG1-SER3 system (Dronamraju et al., 2018) show defects in chromatin structure in cells expressing spt6 mutations. These cells also show defects in Spt6 protein stability and cellular Spt6 protein levels, indicating the importance of CKII-mediated phosphorylation of Spt6 for proper Spt6 function (Dronamraju et al., 2018; Gouot et al., 2018). However, the exact nature of this phosphorylation’s role in maintaining proper Spt6 function and proper chromatin structure is not clear. Consistent with effects on chromatin structure, phosphorylation of Spt6 by CKII also has consequences for transcription. RNA-seq experiments using the same spt6 mutants showed increased levels of antisense and cryptic transcripts compared to wild-type strains, suggesting a role for this phosphorylation in repressing antisense and cryptic transcription (Dronamraju et al., 2018; Gouot et al., 2018). CKII-dependent phosphorylation of Spt6 may have a role in regulating Spt6’s interactions with other proteins. Depletion of CKII was associated with a decrease in phosphorylation of the RNAPII CTD residue Serine 2 (Gouot et al., 2018), a residue known to interact with the tSH2 domains of Spt6 (Burugula et al., 2014; Yoh et al., 2007). However, the consequences of this CKII depletion on the interaction between Spt6 and RNAPII were not elucidated in this study. Future experiments will be necessary to determine the nature of these effects, if any. Furthermore, phosphorylation of Spt6 by CKII presents a mechanism by which the Spt6-Spn1 interaction may be regulated. In an spt6 mutant without CKII phosphorylation sites, Spn1 association with Spt6 was significantly decreased compared to wild-type, suggesting a role for CKII-mediated phosphorylation of Spt6 in maintaining the Spn1-Spt6 interaction (Dronamraju et al., 2018). Thus, CKII phosphorylation of Spt6 may be able to promote normal chromatin structure by promoting normal interaction of Spt6 with other proteins including Spn1. Overall multifunctionality of Spt6 The histone chaperone Spt6, a highly conserved transcription factor, is a multifunctional protein able to regulate transcription through many pathways. One model for how Spt6 regulates transcription is by controlling chromatin structure. Multiple studies have established that changes in chromatin structure such as nucleosome occupancy and nucleosome assembly occur when Spt6 is mutated. Alternatively, Spt6 may regulate transcription directly by interacting with RNAPII. Furthermore, Spt6’s interactions with histones, nucleosomes, Spn1, and CKII are crucial to its ability to regulate chromatin structure and transcription.
Connections between Spt6 and the Mysterious Protein Elf1
In addition to its interactions with RNAPII, Spn1, and CKII, Spt6 likely also interacts with unidentified or less well-characterized factors. One factor that genetically interacts with Spt6 is Elf1 (elongation factor 1). While Elf1 has been characterized as a transcription elongation factor and shown to make several genetic interactions, its role in transcription and the role of these genetic interactions is largely unknown. Identification of an interaction between these two proteins could indicate another pathway Spt6 is involved in and serve to better characterize Elf1 as a transcription factor.
Characterization of Elf1
While its exact roles are unclear, Elf1 has been established as an elongation factor. Elf1 was initially identified through a genetic screen for mutations in genes that cause lethality in combination with mutations in the genes encoding the known elongation factors TFIIS (dst1Δ) and Spt6 (spt6-14). Sensitivity to mycophenolic acid (MPA) and 6-azauracil (6AU), chemicals known to inhibit transcription elongation, in elf1 mutation strains suggested Elf1 as an elongation factor. Furthermore, elf1Δ genetically interacts with mutations in genes encoding elongation factors. When combined with a deletion mutation in the gene encoding the elongation factor TFIIS (dst1Δ), elf1Δdst1Δ cells show increased sensitivity to 6AU and MPA compared to either single mutation as well as inability to grow on galactose, indicating a genetic interaction between elf1Δand dst1Δ. In addition to dst1Δ, tetrad and synthetic genetic array analyses showed elf1Δ to genetically interact with spt4Δ, several spt5 and spt6 missense mutations, and deletion mutations in genes encoding subunits of the Paf1 complex. These interactions with mutations in genes encoding elongation factors suggest a role for Elf1 in transcription elongation (Prather et al., 2005).
Relationship between Spt6 and Elf1
Elf1 preferentially localizes to actively transcribed regions in a manner partially dependent on Spt4 and Spt6. ChIP experiments show high levels of Elf1 localization at the coding regions of transcribed genes and reduced localization over the promoter and activator sequences of these genes. However, the recruitment of Elf1 to transcribed regions was reduced in certain spt4 and spt6 mutant strains, suggesting a role for Spt4 and Spt6 in recruiting Elf1 to coding regions (Prather et al., 2005). The role of Elf1 in transcription elongation may be related to that of Spt6. Genetic interactions between elf1Δ and spt6 mutations and reduced recruitment of Elf1 to transcribed regions in an spt6 mutant suggest a role for Elf1 related to Spt6 function. Given the role for Spt6 in controlling chromatin structure over transcribed regions, it is possible that Elf1 also affects chromatin structure. In an assay for initiation of cryptic transcription, elf1Δ cells initiated cryptic transcription to a greater extent than wild-type cells under high transcription conditions. Increased levels of cryptic initiation indicate a defect in maintaining the nucleosomal barrier to transcription in elf1Δ cells and suggest a role for Elf1 in maintaining proper chromatin structure during transcription (Prather et al., 2005). These results parallel Spt6’s function in maintaining chromatin structure during elongation. However, whether or not Elf1’s role is dependent on Spt6 function is unknown.
Interaction of Elf1 with Other Proteins
Understanding the interaction of Elf1 with other cellular proteins is crucial to understanding the functions of Elf1 and whether there are consequences for the Spt6-Elf1 relationship. One such interaction is between Elf1 and CKII. Copurification experiments show that Elf1 associates with both the catalytic and regulatory subunits of CKII. This physical association suggests that CKII-mediated phosphorylation of Elf1 is important for Elf1 function (Prather et al., 2005). Consistent with this idea, Elf1 is phosphorylated by monomeric and tetrameric forms of CKII in vitro. Furthermore, Elf1 contains several putative CKII sites with the most likely in vitro phosphorylation sites identified by mass spectrometry as Ser95 and Ser117 (Kubinski et al., 2006; Prather et al., 2005). However, the significance and role of this interaction in vivo remains unclear. Elf1 also copurifies with RNAPII and Spn1. While proteomics approaches (Gavin et al., 2002; Krogan et al., 2002) and early copurification experiments with Elf1 did not identify interactions between Elf1 and other proteins, a genome-wide profiling of RNAPII transcription factors showed similar profiles for Elf1 and Spn1 (Mayer et al., 2010). Spn1 copurifies with RNAPII and Elf1 but does not interact directly with Elf1. This result is consistent with an indirect interaction between Elf1 and Spn1 within the RNAPII complex (Mayer et al., 2010). Furthermore, Elf1 does not copurify with other elongation factors. One explanation for this lack of direct interactions may be a requirement for Elf1 association with chromatin in order to physically associate with these proteins, which standard purification methods cannot identify (Prather et al., 2005).
Elf1 Structure in the RNAPII Elongation Complex
Structural studies have also provided a way to investigate the role of Elf1 and its spatial connections to Spt6. Within the elongation complex, Elf1 bridges the RNAPII central cleft. Recent structural studies have employed cryo-EM and X-ray crystallography approaches to solve the structure of the RNAPII elongation complex bound to Elf1 and the elongation factor Spt4/5 from the yeast Komagataella pastoris (Ehara and Sekine, 2018; Ehara et al., 2017, 2019). In addition to locating Elf1 between the lobe and clamp domains of the RNAPII elongation complex, these studies also showed Elf1 to be oriented with its C-terminal α helix interacting with the Rpb2 lobe and the opposite side contacting the Rpb1 clamp (Ehara et al., 2017, 2019). In this position, Elf1 fills the gap over the RNAPII central cleft, thus closing the “DNA entry tunnel” for passage of downstream DNA during elongation (Figure 1). Furthermore, the disordered N-terminal tail of Elf1 may interact with downstream DNA. This interaction is consistent with a role for Elf1 in preventing DNA dissociation from the EC, thus stabilizing the RNAPII elongation complex (Ehara et al., 2017). In this case, the interaction between Elf1 and RNAPII may not be detected in the absence of chromatin and thus cannot be identified by standard purification approaches. However, in vitro experiments suggest that Elf1 slows the rate of elongation, an effect that is diminished when its RNAPII-interacting residues or DNA-interacting N-terminal tail are mutated, supporting a role for Elf1 in preventing DNA dissociation (Ehara and Sekine, 2018; Ehara et al., 2017).
Elf1 has been proposed to prevent DNA dissociation by altering histone-DNA interactions. Recently, cryo-EM structures of transcribing RNAPII elongation complexes show that Elf1 cooperates with Spt4/5 to separate downstream DNA from the nucleosome, thus preventing DNA reassociation to histones and DNA dissociation from the elongation complex (Ehara et al., 2019). Additionally, the N-terminal tail of Elf1 may compete with histones H3-H4 for interactions with DNA, thus promoting dissociation of DNA from histones. Thus, Elf1 may cooperate with Spt4/5 to lower the nucleosomal barrier to transcription elongation (Ehara and Sekine, 2018). Moreover, in the absence of Elf1 and Spt4/5, the nucleosome is trapped in the central cleft between the Rpb2 lobe and Rpb1 clamp that Elf1 normally occupies during elongation (Figure 2). By occupying this position, Elf1 reduces the RNAPIInucleosome interaction and prevents the elongation complex from becoming trapped (Ehara et al., 2019). The structural relationship(s) between Elf1 and other elongation factors remains an area of potential investigation. Copurification experiments suggest that Elf1 and Spn1 interact indirectly within an RNAPII complex (Mayer et al., 2010). Given the interaction between Spn1 and Spt6, Elf1 may also affect Spt6 through the elongation complex. Alternatively, Elf1 may indirectly interact with Spt6 via Spn1. However, based on previous copurification experiments and the putative structural position of Spt6 relative to Elf1, a direct interaction is unlikely (Figure 3; Prather et al., 2005). Interestingly, both Elf1 and Spt6 contact domains of the highly conserved core transcription factor Spt5 in the RNAPII elongation complex. Thus, Spt5 may also mediate an indirect interaction between Elf1 and Spt6.
Conservation of Elf1 and implications
Orthologues of Elf1 are highly conserved and multifunctional. Multiple orthologues of Elf1 have been identified by sequence similarity (Daniels et al., 2009; Prather et al., 2005). The initial characterization of Elf1 in yeast has implications for understanding the potential roles of its orthologues. In Archaea, only some of the genomes containing an Elf1 orthologue also coded for histones, suggesting a histone-independent function for this orthologue of Elf1. Additionally, both Archaea and eukaryotes are able to regulate transcription at the chromatin level; the conservation of Elf1 suggests that this mechanism of regulating transcription is related rather than independently-evolved (Daniels et al., 2009). Mammalian Elf1, Elof1, is also implicated in embryonic development and post-transcriptional regulation of gene expression. Elof1 mutants exhibit clear morphological defects and fail to initiate gastrulation. This observation, in combination with results from an alternative splicing system, suggests that Elof1 functions in regulating alternative splicing (Tellier et al., 2019). While these morphological defects cannot be observed in single-celled yeast, Elf1 may be implicated in regulating alternative splicing. Although not a transcriptional role, future experiments investigating this topic may determine if such a function is dependent on Spt6 and thus advance our understanding of both proteins beyond their involvement in eukaryotic transcription.
Conclusion
Yeast studies have been crucial for advancing our understanding of transcription and key transcription factors. The relationship between chromatin structure and transcription has become increasingly apparent and significant. Proper regulation of transcription requires recruitment of many regulatory factors, many of which have roles in controlling chromatin structure. The elongation factor Spt6 falls into this class of proteins. Spt6- mediated chromatin effects have extensive and diverse consequences for transcription. Genetic studies have shown that defects in Spt6 are associated with changes in nucleosome occupancy and nucleosome dynamics. Furthermore, the control Spt6 exerts over chromatin structure and transcription may be influenced by its interactions with other proteins such as histones, RNAPII, Spn1, and CKII. Genetic interactions between Spt6 and the comparatively less-investigated elongation factor Elf1 have been important for elucidating the role of Elf1 in regulating transcription. Recent structural insights suggest that Elf1 prevents DNA dissociation from the RNAPII elongation complex. However, this insight remains to be confirmed in vivo.
Future directions
The implication of Elf1 in transcription elongation, coupled with the relatively small amount of information known about it, makes Elf1 an especially exciting and interesting target for future investigations. Particularly, the connection to Spt6 may be key to understanding the role of Elf1 in regulating chromatin structure and elongation. One approach that future studies may take to investigate this topic is identifying elf1 mutations that suppress defects in chromatin structure caused by an spt6 allele, analogous to the approach taken by McCullough et al. (2015). Structural insight on a yeast RNAPII elongation complex bound to Elf1 and Spt6 may also further elucidate the relationship between the two factors as well as the role of Elf1 in transcription. Finally, computational approaches aimed at uncovering the pathways Elf1 is implicated in have the potential to reveal any connections between Spt6 and Elf1.