Transcription Regulation in Archaea

Abstract

The known diversity of metabolic strategies and physiological adaptations of archaeal species to extreme environments is extraordinary. Accurate and responsive mechanisms to ensure that gene expression patterns match the needs of the cell necessitate regulatory strategies that control the activities and output of the archaeal transcription apparatus. Archaea are reliant on a single RNA polymerase for all transcription, and many of the known regulatory mechanisms employed for archaeal transcription mimic strategies also employed for eukaryotic and bacterial species. Novel mechanisms of transcription regulation have become apparent by increasingly sophisticated in vivo and in vitro investigations of archaeal species. This review emphasizes recent progress in understanding archaeal transcription regulatory mechanisms and highlights insights gained from studies of the influence of archaeal chromatin on transcription.

INTRODUCTION

RNA polymerase (RNAP) is a well-conserved, multisubunit essential enzyme that transcribes DNA to generate RNA in all cells. Although RNA synthesis is carried out by RNAP, the activities of RNAP during each phase of transcription are subject to basal and regulatory transcription factors. Substantial differences in transcription regulatory strategies exist in the three domains (Bacteria, Archaea, and Eukarya). Only a single transcription factor (NusG or Spt5) is universally conserved (1, 2), and the roles of many archaeon-encoded factors have not been evaluated using in vivo and in vitro techniques. Archaea are reliant on a transcription apparatus that is homologous to the eukaryotic transcription machinery; similarities include additional RNAP subunits that form a discrete subdomain of RNAP (3, 4) as well as basal transcription factors that direct transcription initiation and elongation (5–8). The shared homology of archaeal-eukaryotic transcription components aligns with the shared ancestry of Archaea and Eukarya, and this homology often is exclusive of Bacteria. Archaea are prokaryotic, but the transcription apparatus of Bacteria differs significantly from that of Archaea and Eukarya.

The archaeal transcription apparatus is most commonly summarized as a simplified version of the eukaryotic machinery. In some respects this is true, as homologs of only a few eukaryotic transcription factors are encoded in archaeal genomes, and archaeal transcription in vitro can be supported by just a few transcription factors. However, much regulatory activity in eukaryotes is devoted to posttranslational modifications of chromatin, RNAP, and transcription factors, and this complexity seemingly does not transfer to the Archaea, where few posttranslational modifications or chromatin-imposed regulation events are currently known. The ostensible simplicity of archaeal transcription is under constant revision, as more detailed examinations of archaeon-encoded factors become possible through increasingly sophisticated in vivo and in vitro techniques. This review will highlight the current understanding of archaeal transcription, emphasizing the roles of factors that regulate archaeal RNAP throughout each stage of the transcription cycle and also highlighting outstanding issues in the field.

THE ARCHAEAL TRANSCRIPTION CYCLE

Transcription is highly regulated, and the transcription cycle is typically demarcated into three phases: initiation, elongation, and termination (9–13) (Fig. 1). An abbreviated and overall introduction to this cycle is presented first, with sections below detailing the activities of RNAP and associated factors during each stage of transcription. Briefly, archaeal transcription initiation requires that RNAP be directed to promoter sequences defined by the binding of TATA binding protein (TBP) and transcription factor B (TFB). TBP, TFB, and RNAP are sufficient to generate a single-stranded section of DNA (the transcription bubble) and feed the template strand into the bipartite active center of RNAP (7, 14). RNAP can initiate transcript synthesis de novo, and continued synthesis then competes with favorable promoter and initiation factor contacts until promoter escape can be achieved. Release of RNAP from the initiating factors classically defines the end of initiation, although in reality no clear boundary separates the last stages of initiation from the early stages of elongation. Although TFB and TBP are necessary and sufficient to permit promoter-directed transcription initiation, a third conserved factor, transcription factor E (TFE), can also assist in transcription initiation and leaves the promoter with RNAP during the early stages of transcript elongation (15–18). Transition to a stable, long-lived elongation complex is believed to involve internal rearrangements of RNAP. This transition involves the exchange of initiation factors for stably bound elongation factors that monitor RNA synthesis for accuracy, respond to regulatory DNA sequences, react to regulatory inputs of more transiently associated transcription factors, and influence processivity of RNAP. Elongation is, in general, very stable, but specific sequences can lower the overall energy of the transcription elongation complex, permitting either spontaneous intrinsic or factor-assisted termination (19, 20). Transcription termination results in release of both the transcript and RNAP from the DNA template.

FIG 1.

The archaeal transcription cycle. (A) The euryarchaeal RNA polymerase crystal structure from Thermococcus kodakarensis (PDB ID no. 4QIW) is shown in a surface representation. The clamp and stalk domains are highlighted. A simplified cartoon structure of RNA polymerase is shown below this in light green; the bipartite active site and RNA exit channel are highlighted in dark green. (B) Steps in the transcription cycle. (i) RNAP is recruited to the promoter by transcription factors TFB, TFE, and TBP during transcription initiation. (ii) RNAP escapes the promoter, and early elongation begins with TFE bound to RNAP. (iii) TFE is replaced by elongation factor Spt5 during elongation. (iv) Factor-dependent termination is predicted to occur in archaea by an unknown factor. (v) Intrinsic termination sequences are characterized by a run of T's on the nontemplate strand. (vi) The transcript is released, and RNAP is recycled for another round of transcription.

REGULATED TRANSCRIPTION INITIATION

Transcription initiation is tightly regulated by both transcription factors and DNA elements. The minimal, necessary proteins and DNA elements for archaeal transcription initiation are now well defined and characterized (21–28). A recent excellent review (29) summarizes the actions of repressors and activators that function during initiation in archaeal species. We focus here on the roles of new DNA elements and newly discovered strategies of basal initiation factors.

BASAL TRANSCRIPTION FACTORS

TBP and TFB are the only transcription factors required for in vitro transcription under optimized conditions, and TFE has been shown to assist promoter opening when conditions are suboptimal (16). In vivo studies have shown that Archaea must retain at least one gene encoding TBP and one gene encoding TFB, although many archaeal species encode multiple TBP and TFB isoforms (6, 21, 30–35). Some differences in promoter sequence preferences and protein pairing have been noted in TBP-TFB isoform pairs (36–41), but these minor differences are not on par with the clear but not always radical promoter sequence differences noted for alternative σ factors in bacterial transcription (39, 42). TFE also appears essential, and it is currently unclear if this essentiality is due to necessary activities during transcription initiation or some other role in the transcription cycle (26, 43, 44).

All three of the aforementioned transcription factors have close eukaryotic homologs: archaeal TBPs are nearly identical to eukaryotic TBPs (45); archaeal TFB proteins are homologous to eukaryotic transcription factor IIB (TFIIB) proteins (46), with homology also seen with the Pol III initiation factor BRF1 (47) and Pol I initiation factor Rrn7/TAF1B (48); and archaeal TFE proteins are homologous to the N-terminal half of the eukaryotic alpha subunit of TFIIE, or TFIIEα, and very recent evidence identified a separate homolog in some lineages to the eukaryotic beta subunit, TFIIEβ (17). TBP is needed to recognize the TATA box, bend the DNA, and recruit TFB (46); its role had therefore been deemed equivalent to the role of eukaryotic TBPs. Recent, sophisticated total internal-reflection fluorescence–fluorescence resonance energy transfer measurements now detail differences in the activities of archaeal and eukaryotic TBPs, despite the nearly identical three-dimensional folds of these factors (6). In some cases, archaeal TBPs require the cobinding of TFB to stably bind and bend the promoter DNA (6, 22, 49, 50). It is tempting to speculate that different promoter sequences may be regulated by different TFB-TBP pairs based on the interdependence, or lack thereof, of cooperative DNA bending for establishing a stable platform for RNAP recruitment. Recent studies suggest that select isoforms of TFB and TBP can result in differences in transcription output, but further studies will be needed to determine if these effects on such preliminary steps of transcription initiation are a direct mode of regulation resulting in phenotypic differences (37, 51).

In contrast to eukaryotic transcription, archaeal promoter opening is not an energy-dependent process (7). Therefore, TBP and TFB alone are capable of assisting RNAP in the formation of the transcription bubble. In all Archaea, TFB is responsible for stabilizing the TBP-bound DNA complex and, together, this bipartite protein platform recruits RNAP (52), but how these molecular interactions melt the DNA is still unresolved. Reconstructions and analyses of open complexes using archaeal components reveal an overall architecture of the open promoter complex and provide the first placement of the nontemplate strand within the complex (52). TBP and TFB are located closer to RNAP than would be the case for eukaryotic promoters, and this proximity may provide more intimate contacts that collectively provide the energy to open the promoter DNA. The tight network of interactions in the archaeal open complex may torsionally strain the DNA, and melting is likely to relieve this strain and result in open complex formation.

Several new insights into TFE activity and evolution have been recently described. The archaeal TFE had previously been characterized as a monomer and as a homologue of the alpha subunit of eukaryotic TFIIE, termed TFIIEα (16, 18, 53). Eukaryotic TFIIE is a heterodimeric complex of TFIIEα and TFIIEβ, but archaeal genomes had previously only been shown to encode a homologue of only the alpha subunit (54, 55). Eukaryotic RNAPs differ in their requirements for initiation, with RNAP III incorporating homologues of several RNAP II initiation factors as core components of RNAP III (56–58). Comparisons of the RNAP III subunit hRCP39 revealed a well-conserved archaeal homolog (termed TFEβ) that directly and extensively interacts with TFE (now named TFEα) (17). Although TFEβ is not conserved in all Archaea, TFEβ is essential for some Crenarchaea; when employed in vitro, TFEα-TFEβ complexes are effective in binding RNAP, stabilizing open complex formation, and stimulating total transcriptional output (17).

The mechanism of TFE recruitment to the initiation complex and its activities during initiation has been partially resolved. TFEα simultaneously binds TBP, RNAP, and downstream DNA and has been shown to stimulate transcription at noncanonical promoter sequences and at reduced temperatures in vitro (16, 18, 59). Several studies have identified critical interactions between TFE and the preinitiation complex that have furthered our understanding of TFE function during initiation (2, 15, 26, 53, 59). TFEα consists of two domains: a winged helix (WH) domain and a zinc ribbon domain (60, 61); TFEβ contains a conserved WH domain and an FeS domain (17). The WH domain of TFEα contacts the upstream, nontemplate strand of DNA and helps form the open promoter complex through an unknown mechanism (15, 52). Several studies have shown that the presence of the RNAP stalk domain—unique to archaeoeukaryotic RNAPs and comprised of two subunits, RpoE and RpoF in archaea and Rpo4 and Rpo7 in eukaryotes—is essential for the full activity of TFEα (59, 62, 63). The predicted interaction between TFEα and the stalk domain was bolstered by copurification of TFEα with intact RNAP and the loss of TFEα from RNAP preparations wherein the stalk domain was missing (44). A recent structure-function study identified critical interactions between TFEα and RpoE of the stalk domain (26). TFE may have an essential role in modulating intramolecular movements of RNAP during the transcription cycle, most notably movements of the clamp domain. Interaction of TFEα with both the stalk and clamp domains of RNAP during transcription initiation may retain the clamp domain in an open conformation necessary for initiation and early elongation. Replacement of TFE by Spt4/5 during early elongation may alter clamp positioning and further stabilize the elongation complex (2).

DNA ELEMENTS

Transcription initiation is regulated by DNA elements that are recognized by basal transcription factors and that influence subsequent steps in promoter opening. There are four DNA elements currently known to regulate archaeal transcription initiation: (i) the TATA box located approximately 25 bp upstream of the site of transcription initiation (64–66), (ii) the TFB recognition element (BRE) located immediately upstream of the TATA box (6), (iii) the initiator element (INR) located within the initially transcribed region, and (iv) the promoter proximal element (PPE) located between the TATA box and the site of transcription initiation (67–69). Of these four, only the TATA box and the BRE are required for transcription initiation, although alterations to all four elements can influence the total output of a promoter.

The INR is not a required DNA element for transcription initiation; however, it is a regulatory element that can increase the strength of the promoter in a TATA- and BRE-dependent manner. The INR is a core promoter element located in the 5′ untranslated region, and it has sequence similarity to the TATA box. The INR has been shown to be targeted by some transcriptional activators, and its high AT content may facilitate promoter opening in some instances. Many archaeal transcripts are leaderless, so the INR is not consistently identifiable, and the regulatory influence of INR sequences does not appear to extend to RNA half-life or alter the translational capacity (70). PPEs, centered approximately 10 bps upstream of the site of initiation, have been shown to increase transcription output through recruitment of TFB (67, 68). Additionally, permanganate footprinting data of the preinitiation complex demonstrated that the border of the transcription bubble is at the PPE and that this region is important for the activity of TFEα-TFEβ (17).

REGULATION OF ELONGATION

As transcription transitions from initiation to elongation, RNAP undergoes a conformational change accompanied by the replacement of initiation factors with elongation factors (2, 12, 71–74). It is plausible that the emerging nascent transcript stimulates the swap of regulatory factors and initiates the intramolecular movements that result in stable elongation complex formation (62, 75). Very few transcription elongation factors have been bioinformatically identified within archaeal genomes, and it is probable that archaeon-specific factors await discovery. It is worth noting what is seemingly not encoded in archaeal genomes, given that so much of archaeal and eukaryotic transcription machinery is shared. Archaeal genomes do not appear to encode any coactivator complexes or megacomplexes for chromatin modification or rearrangements. There does not appear to be machinery for regulated posttranslational modifications of the archaeal transcription apparatus nor of chromatin, with the exception of acetylation/deacetylation of the small chromatin-associated protein Alba (76–79). Furthermore, archaeal transcripts are not capped, do not require nuclear export, and, with the exception of self-splicing introns, are intronless; thus, factors responsible for these activities are similarly lacking from archaeal genomes (80–82).

Transcription elongation factors have various roles, including increasing processivity and fidelity of RNAP and/or increasing genome stability. Only two archaeal elongation factors have been experimentally studied: the aforementioned universally conserved elongation factor Spt5, often with a conserved binding partner Spt4 (Spt4/5) (2, 83, 84), and transcription factor S (TFS) (85, 86). Several recent studies have shed light on the roles of Spt5 during elongation (1, 72, 87, 88). TFS, with homology to the C-terminal domain of eukaryotic TFIIS and functionally analogous to GreA/GreB in Bacteria (8, 89–91), can stimulate endonucleolytic cleavage of the RNA from backtracked RNAP complexes (85, 91–93). The finding of multiple TFS homologues in some archaeal lineages offers the possibility of unique regulatory roles of specific isoforms.

TRANSCRIPTION FACTOR Spt5

Archaeal Spt5, homologous to bacterially encoded NusG, consists of two domains: the NusG N-terminal (NGN) domain and a single C-terminal Kyrpides-Ouzounis-Woese (KOW) domain with affinity for single-stranded RNA (83, 84, 87); eukaryotic Spt5 typically contain three to six repeats of the C-terminal KOW domain (94–96). Critical, direct molecular interactions between Spt5 and RNAP have been identified in both Bacteria and Archaea (83, 84, 87, 88, 95, 97–99), and the conservation of RNAP and Spt5 infers that these same interactions are used in Eukarya. Briefly, a hydrophobic depression on the NGN domain interacts with the mobile clamp domain of RNAP, with additional interactions between the NGN domain and RNAP jaw domain likely fixing the location of the clamp domain in a closed configuration (11, 98). Spt5 interaction with RNAP is not necessary for productive and processive elongation in vitro, but the interaction does increase the total output of transcription systems (1). It is plausible that Spt5 increases elongation rates and processivity, as NusG in Escherichia coli does, and it is further possible that the increased efficiency of transcription results from the stabilization of the clamp domain that in turn stabilizes the DNA-RNA hybrid in place during transcription elongation (87, 100–102). The NGN domain also contacts the upstream strands of DNA, offering protection from backtracking, and, by inference, may reduce pausing of the transcription elongation complex (87, 88, 103, 104). It is of importance to note that NusG/Spt5 can have a positive and/or negative effect on elongation rates and pause events of RNAP. In Thermus thermophilus, NusG slows down RNA elongation rather than increases elongation rates (105). In Bacillus subtilis, sequence-specific interactions of the NGN and nontemplate DNA strand within the paused transcription bubble stabilize the pause event in the trp operon (103, 106). Furthermore, evidence has shown that Spt4/5 induces pauses during early elongation of Pol I but promotes elongation downstream (107). Although NusG can elicit opposite roles on transcription elongation, the NusG-RNAP binding sites remain well conserved across various species. Archaeal and eukaryotic genomes often encode an additional elongation factor, Spt4 (annotated as RpoE″/RpoE2 in Archaea), that forms a complex with Spt5 and stabilizes the Spt5-RNAP interaction (1, 84, 95). Spt4 does not appear to be essential; however, the affinity of Spt5 for RNAP decreases in the absence of Spt4 in vitro (1).

The primary interacting partners (e.g., RNAP and Spt4) of the Spt5-NGN domain have been established in molecular detail; however, no specific interacting partners of the KOW domain have been identified in archaea. It is possible that the affinity of the KOW domain for RNA leads to nonspecific interactions with the emerging transcript; however, it is tempting to speculate about greater involvement of the KOW domain based on the known activities of the C terminus of bacterial NusG (108). Bacterial NusG can facilitate elongation or termination depending on its binding partner (99–101, 109–111). The bacterial NusG KOW domain can interact with the S10 ribosomal subunit (NusE) during elongation, thereby linking the leading ribosome with the transcription apparatus (110, 111). When not bound to a trailing ribosome, the bacterial NusG-KOW domain can be bound by and stimulate the activity of the transcription termination factor Rho (109, 112, 113). Archaeal transcription and translation are similarly coupled (114, 115), and it is reasonable to venture that archaeal Spt5 can also link the archaeal transcription and translation apparatuses and also potentially interact with termination factors.

INTRAMOLECULAR REARRANGEMENTS OF RNAP MAY INCREASE PROCESSIVITY

The archaeal and three eukaryotic RNAPs can be reduced in complexity to three large domains: the core, the mobile clamp, and the stalk (4, 73, 116). The archaeoeukaryotic stalk, absent from bacterial RNAP, is used by a host of archaeal and eukaryotic transcription factors to bind and regulate the activities of RNAP. Increasing evidence from biochemical, biophysical, and in vivo approaches indicate that transcription factor binding often stimulates intramolecular movements of RNAP that appear necessary for transitions between phases of the transcription cycle (2, 4, 26, 88, 97, 117).

Hinge-like movement of the mobile clamp domain has been demonstrated for the bacterial RNAP (71). The movements of the mobile clamp are sufficiently large enough to open the main channel of RNAP, such that double-stranded DNA can easily enter and exit when the clamp is open, whereas double-stranded DNA—or the RNA-DNA hybrid—would be trapped inside RNAP when the clamp is closed. The bacterial RNAP clamp is open during initiation but remains closed during processive elongation (71), leading to a simple model of encapsulation of the nucleic acids to explain the dramatic stability of the elongation complex. It is logical to propose mechanistic actions of transcription factors that may modulate the clamp positioning with respect to the core and stalk domains of RNAP and thus alter the stability and transitions of RNAP throughout the transcription cycle. TFE is predicted to make contacts with both the clamp and stalk domain of RNAP, thereby fixing the clamp into the open conformation critical for initiation (26, 59, 117–119). As transcription transitions into the elongation phase, RNA emerges from the enzyme and interacts with the stalk domain (62, 75), where a predicted steric clash occurs between the RNA and the TFE, likely driving TFE to disengage from RNAP. The disengagement of TFE allows for Spt5 to bind to the clamp and core domains of RNAP and lock the clamp in the closed position, thus ensuring processivity during elongation (87).

RNAP clamp movement is predicted to be universal; however, both the archaeal and the eukaryotic RNAP contain additional subunits, including the stalk domain (2, 73, 116, 118, 119), and previous structural data predicted that the stalk domain would sterically limit or abolish major movements of the clamp domain. Recent crystallographic evidence of the complete euryarchaeal RNAP demonstrated that the clamp is able to open without a steric clash with the stalk domain through a coordinated swing and rotation movement of both the clamp and stalk domains (73). This evidence supports the bacterial mechanism of the clamp opening and closing during initiation/termination or elongation, respectively, thus supporting a universal model of clamp movement.

TERMINATION

Transcription termination occurs when the transcription elongation complex becomes sufficiently unstable and fails to maintain contact between RNAP and the encapsulated nucleic acids. The stability of the transcription elongation complex is derived from (i) contacts between RNAP and the RNA-DNA hybrid, (ii) contacts between RNAP and single-stranded RNA in the exit channel, (iii) contacts between RNAP and the downstream DNA, and (iv) the base pairing of the RNA-DNA hybrid (116, 120–126). The first and last of these contacts are most likely to be altered during the termination process. Transcription through specific DNA sequences can result in stronger or weaker base pairing within the RNA-DNA hybrid, and contacts between RNAP and the nucleic acids are most easily modified by movements of the clamp domain that relieve movements of the hybrid with respect to the core of RNAP (127–129). Release of the nascent RNA may be possible through continued translocation in the absence of synthesis, or the RNA-DNA hybrid could be released in bulk if the clamp domain transitions from a closed to an open position. The gene-dense nature of many archaeal genomes necessitates timely termination of transcription to prevent aberrant transcription of neighboring genes. It is predicted that there are two mechanisms of termination across all domains: intrinsic termination and factor-dependent termination (Fig. 1B).

INTRINSIC TERMINATION

Intrinsic transcription termination is driven primarily by weak base pairing within the RNA-DNA hybrid and occurs independent of the activity of transcription factors (130, 131). Intrinsic transcription termination has been established in all three domains (19, 20, 132, 133), with some differences in sequence and structural requirements (130, 132, 134–136). The archaeal RNAP, like eukaryotic RNAP III, is sensitive to intrinsic termination (19, 133, 137, 138). Eukaryotic RNAP I and RNAP II do respond to DNA sequence context in the form of pauses and arrests but rarely release the transcript at such positions (139–141). Archaeal intrinsic termination is characterized by a run of 5 to 10 thymidine residues in the nontemplate strand, encoding a poly(U) run at the 3′ end of the nascent RNA (19, 20). The weak rU:dA RNA-DNA hybrid at or near the positions of termination is seemingly insufficiently energy rich to maintain the stability of the elongation complex; RNAP III similarly spontaneously dissociates upon transcription of poly(T) nontemplate tracts.

IDENTIFICATION OF FACTOR-DEPENDENT TERMINATION

Transcription factors involved in initiation and elongation have been characterized in all domains, while a transcription termination factor(s) has been characterized only in Bacteria and Eukarya (142–145). By inference, from known termination factors that are employed in bacterial and eukaryotic systems, it is easily argued that protein factors are encoded in archaeal genomes that have the capacity to direct transcription termination in vivo. Bioinformatic analyses reveal some potential targets that remain to be more fully evaluated, but there are no easily identified homologues of known eukaryotic or bacterial termination factors. Two well-studied transcription bacterial termination factors, Rho and Mfd (13, 146–150), lack clear homologues in archaeal genomes, but there are hints that analogous activities may be present in archaeal species. Rho is a homohexamer helicase that represses phage transcription and mediates polar repression of downstream genes when transcription and translation become uncoupled (142, 151–153). Archaea demonstrate polar repression of downstream genes in the absence of continued translation, and it is likely that a factor or factors mediate polarity in archaea (115). It is tempting to use the bacterial model of NusG-Rho interactions to conjure a similar picture for Spt5-KOW interactions with an archaeal transcription termination factor; Rho is capable of terminating a stalled archaeal RNAP in vitro (19). The bacterial Mfd protein can remove RNAP from sites of DNA damage and initiate transcription-coupled DNA repair (146, 148, 150, 154). Recent evidence that the archaeal RNAP halts synthesis and forms long-lived complexes at the site of lesions in DNA in vitro predicts that mechanisms exist to remove RNAP from the site of damage (T. J. Santangelo, unpublished results).

CHROMATIN ARCHITECTURE AFFECTS THE TRANSCRIPTION CYCLE

Archaea employ two seemingly distinct mechanisms to compact, wrap, and condense their genomes to fit within the cell (Fig. 2) (155). Most euryarchaeal species are polyploid (156–160) and encode histone proteins that dominate chromatin architecture (156–160); archaeal histones mimic the core eukaryotic histone fold (161). In contrast, most crenarchaeal species are diploid and are reliant on small, basic nucleoid proteins to organize their genomes (162, 163). Condensation demands organization of the genome and offers regulatory opportunities by controlling the accessibility of promoter sequences, the introduction of local superhelicities that may promote or inhibit promoter opening, and the potential for the introduction of chromatin-based obstacles to transcription elongation. The overall role of genome architecture with respect to archaeal transcription is an emerging area, with several recent studies highlighting the breadth of influences that genome architecture can have on transcription output at the organismal level.

FIG 2.

Transcription in the context of archaeal chromatin. (A) The structure of histone A from Methothermus fervidus (PDB ID no. 1B67) is overlaid by a cartoon representation of each histone dimer with ∼60 bp of DNA wrapping the complex. (B) The crystal structure of an Alba dimer from Sulfolobus solfataricus (PDB ID no. 1H0X) bound to DNA is overlaid by a cartoon representation. (C) Transcription elongation continues in a chromatin environment. Accessibility of the TATA box and BRE is altered by localized chromatin structure.

Archaeal histone-based chromatin is composed of nucleosome particles that wrap and condense the genome. The best-described complexes are homo- or hetero-histone tetramers, homologous to the H3/H4 tetramer in eukaryotes, that associate with ∼60 bp of double-stranded DNA. Archaeal histones share similar biases with eukaryotic nucleosomes for flexible DNA sequences and are, in general, absent from the core promoters of archaeal genes (164, 165). Archaeal histone proteins share the same core fold as eukaryotic histones but lack the extensions from this fold (i.e., tails) that are highly modified and essential for proper nucleosome dynamics in eukaryotes (166). Higher-order structure has been demonstrated in Thermococcus kodakarensis in the form of dynamic histone polymers that have the ability to wrap up to 180 bp (167). Archaeal nucleosomes present a surmountable barrier to the progression of the transcription elongation complex, although traversion does slow the elongation complex (168). The lack of known modifications to archaeal histones, and the lack of known machinery for the repositioning or movement of archaeal nucleosomes, suggests that transcription elongation complexes simply traverse the nucleosomes and that chromatin organization spontaneously reforms when the histones gain access to preferred binding positions following the departure of RNAP. This mechanism of elongation through the histones is similar to the mechanism of Pol III in eukaryotes (168–170).

The activities or stimulatory effects of archaeal elongation factors on transcription through archaeal histone-based chromatin remain to be explored; the substantial pausing and delayed progress of RNAP on chromatinized templates suggest that elongation factors will accelerate progress of the transcription elongation complex. Any role of chromatin architecture in transcription termination is similarly unexplored. The topology of naked DNA templates does influence the positions and efficiencies of intrinsic terminators, suggesting that chromatin templates may also influence termination patterns. Nucleosomes are depleted not only from promoter regions but also from predicted termination regions, suggesting a potential regulatory role for chromatin architecture on termination of transcription (164).

HISTONE-BASED REGULATION OF TRANSCRIPTION

Several genetic studies have addressed the role of archaeal histone-based chromatin on gene expression at the organismal level, with surprisingly different results. In some halophilic species, singular histone-encoding genes are nonessential, and histone proteins appear to function more akin to site-specific transcription factors, moderately influencing the expression of only a few genes (171). These studies contrast the view of histone proteins as general organizational factors with the global influence on gene expression and minimally suggest that the archaeal chromatin of some species is dependent on the activities of many nucleoid-associated proteins. When histone-encoding genes have been deleted, or have been attempted to be deleted from other species, more global disruption of gene expression has been noted (161, 164, 165, 167, 171–176). Some species are reliant on at least one histone protein, and it is unclear at this point whether the noted global changes in gene expression seen in deletion strains stem from reorganization or disorganization of the archaeal genomes or the primary, secondary, and tertiary effects of localized disruptions that lead to additional differences in regulation at remote sites.

NUCLEOSOME OCCUPANCY AT THE PROMOTER

Chromatin architecture at a promoter could influence or prevent transcription initiation by occluding transcription factor binding or inhibiting DNA melting (164, 167, 168, 177). Crenarchaeon-encoded nucleoid-associated proteins have been shown to influence transcription output through the acetylation/deaceytlation of Alba in vitro (76), although Alba has not yet been shown to influence transcription in vivo. It is possible that Alba regulates transcription, given that Alba proteins can loop, condense, bridge, and even saturate DNA in vitro, but the in vivo dynamics remain unknown (178–182). In the euryarchaeal organism Methanococcus voltae the deletion of the gene encoding Alba resulted in the upregulation of only a small number of genes, implying that Alba-based regulation may be limited in scope (173). Additional research may reveal a clearer picture of transcriptional regulation through the binding of Alba.

The binding preferences and genomic locations of stable euryarchaeal histone proteins interactions have been mapped, and it has been shown that regions directly upstream from the start codon are nucleosome depleted on a global scale (164, 165). The presence of histones bound at the promoter has been correlated with a decrease in total transcription in vitro (177), and it was suggested that both steric and torsional effects limited binding of basal transcription factors to the DNA (177). Although most data support the lack of nucleosomes at the promoter, specific promoters can be regulated by nucleosome occupancy. This appears to be a general mechanism of histone-based regulation in some halophiles and a more specialized mechanism of regulation in other species. The transcriptional activator Ptr2 from Methanocaldococcus jannaschii must outcompete histones for binding to the promoter to activate transcription of select genes (183).

CONCLUSIONS AND OUTSTANDING ISSUES

Exploration of archaeal transcription and regulation continues to yield a bounty of evolutionary, biophysical, and mechanistic details of transcription mechanisms that are often applicable to all extant life. The ability to reconstitute the complete archaeal transcription apparatus permits biophysical studies not possible with eukaryotic components, and the simplicity and explicit homology of many factors provide meaningful insights into the mechanistic roles of individual factors and even of specific domains and residues of archaeal transcription components. The development and recent advances in genetic techniques for more archaeal species are now offering complementary in vivo studies to probe regulatory strategies and rationally manipulate protein interfaces and activities in the cell. Although discussion of transcriptome mapping of archaeal organisms is outside the scope of this review, the mapping is becoming more frequent (36, 184–189) and offers invaluable insight into noncoding RNA, transcription start site selection and redundancy, and expression levels under various growth conditions (36, 171, 190–193).

There is still much to be learned regarding archaeal transcription regulation and mechanisms. The identification and characterization of additional archaeal elongation and potential termination factors offer the opportunity to examine archaeon-specific mechanisms of regulation. Factors that regulate the organization and dynamics of archaeal chromatin are likely to be identified and should offer contrasting regulatory potential with the network of regulatory strategies employed for eukaryotic chromatin. Continued insightful biophysical probing of shared archaeoeukaryotic factors will surely reveal conserved regulatory strategies for promoter recognition, DNA melting, transcription factor swapping, and elongation through chromatinized templates. Advances in genetic techniques will quickly move studies of archaeal transcription inside the cell, and the application of omics approaches to gene expression in modified strains should answer outstanding question surrounding archaeal responses to external signals and ever-changing environments. Given the extremophilic nature of many experimentally utilized Archaea, the evolutionary survival strategies of these remarkable microbes will come into better focus.