In eukaryotic cells, three distinct RNA polymerases—Pol I, II, and III—carry out DNA-dependent synthesis of RNA. Although the properties of these polymerases resemble those of bacterial RNA polymerase in many ways, the eukaryotic polymerases require many additional proteins, called transcription factors, to begin efficient transcription at promoter sequences. These factors help assemble transcription complexes on chromatin, the compacted form of DNA that makes up eukaryotic genomes. Like bacterial sigma factors, each eukaryotic transcription factor binds to a specific promoter sequence and to a particular RNA polymerase, bridging the two to initiate transcription. Using a variety of general transcription factors, eukaryotic cells promote the transcription of many sets of genes under varying conditions (Figure 15-18). Specific transcription factors bind DNA at a long distance upstream from the promoter, at sequences known as enhancers, and can stimulate or repress transcription in various ways. Transcription factors, both general and specific, have important roles in gene regulation and cell development. Indeed, recent studies show that differentiated cells, once believed to be committed to a particular cell type, can be converted to another cell type simply by manipulating the expression of transcription factors (Highlight 15-2). Transcription regulation proteins, including transcription factors, and the DNA sites to which they bind, including enhancers, are discussed in detail in Chapters 19–22.

Because transcription is a fundamental process in all cells, it is not surprising that some eukaryotic RNA polymerase subunits are homologous to those of bacterial polymerase. Furthermore, some subunits are common to all three of the eukaryotic polymerases (see Table 15-1). Relative to bacteria, eukaryotes require additional factors to help RNA polymerases find and access promoters in the cell nucleus. This is because eukaryotic DNA is packaged into chromatin through the formation of nucleosomes (see Chapter 10). In addition, the sheer size of eukaryotic genomes, and the large number of promoters to be sorted through, probably requires additional transcription machinery.

We begin with a brief discussion of all three eukaryotic polymerases and their promoters, and then focus on Pol II transcription. As the polymerase responsible for transcribing the genes that encode proteins, Pol II is the most extensively studied of the three eukaryotic polymerases.

Each of the three types of RNA polymerase that make up the eukaryotic transcription machinery transcribes only certain classes of genes, and thus each type binds to specific and distinct promoter sequences. Pol I binds to a single type of promoter that controls the expression of the pre-ribosomal RNA (pre-rRNA) transcript, from which rRNAs are derived. Pol II, which synthesizes mRNAs, microRNAs, and some other noncoding RNAs, can recognize thousands of promoters that vary greatly in sequence. Pol III recognizes well-characterized promoter sequences for tRNAs, the 5S rRNA, and some other small regulatory RNAs, sequences that in many cases are located within the transcribed region itself rather than in more conventional locations upstream from the RNA start site.

Although each polymerase works with its own unique set of transcription factors, all three types use a factor called the TATA-binding protein (TBP). This protein, so-named because of its binding to a 5′-TATAAA sequence (known as the TATA box) near position −30, plays a major role in transcription initiation. Genomic sequencing studies have shown that only about a quarter of human genes include a TATA box in the core promoter, the region responsible for recruiting the essential transcription machinery. Nonetheless, TBP is used for transcription initiation of all genes, and in most of those that lack a TATA box, TBP is recruited to the gene through proteins called TBP-associated factors (TAFs) that recognize other promoter sequences. A summary of the eukaryotic polymerases and the types of RNA produced by each, along with their promoter elements, is given in Table 15-3.

538

RNA Polymerase I Promoters Synthesis of pre-rRNA accounts for 80% of all the transcription in eukaryotic cells (as measured for the yeast S. cerevisiae). The precursor rRNA transcript produced by mammalian Pol I is processed into the mature 5.8S, 18S, and 28S rRNAs, which, together with the 5S rRNA transcribed by Pol III, are the major catalytic and architectural components of the ribosome (discussed in Chapter 18). Transcription of rRNA genes, which occurs in the nucleolus, begins with the recruitment and assembly of Pol I and transcription factors into a multiprotein complex at the rRNA gene promoter. The promoter includes a core sequence, essential for accurate transcription initiation, and an upstream control element (UCE), located 100 to 150 bp upstream from the transcription start site (Figure 15-19).

539

The number of rRNA genes varies among organisms. Pol I promoter sequences also vary, but within a species, all Pol I promoters are the same. Low levels of transcription can be observed in the presence of a preinitiation complex comprising Pol I and selectivity factor 1 (SL1), which is a complex of TBP and three TAFs. Higher levels of transcription require, in addition to Pol I and SL1, an upstream binding factor (UBF). UBF binds to both the UCE and the core promoter and to SL1, stabilizing the complex with Pol I and helping recruit the polymerase to the promoter.

RNA Polymerase II Promoters Many Pol II promoters share certain sequence features, including a TATA box near −30 and an initiator sequence (Inr) near the RNA start site at +1 (Figure 15-20). Pol II promoters also sometimes include a sequence upstream from the TATA box, called a TFIIB recognition element (BRE), and a sequence downstream from the initiator, the downstream promoter element (DPE). These sequences comprise the core promoter. Other sequences are also needed for efficient Pol II recognition and transcription in the cell, such as upstream promoter elements and enhancers. These regulatory sequences, which can be located many thousands of base pairs away from the promoter they influence, bind a variety of specific transcription factors that either activate or repress transcription, depending on various stimuli. The proteins that bind to the elements in the promoter are discussed shortly.

540

RNA Polymerase III Promoters Pol III is the largest RNA polymerase with the greatest number of subunits. All of its transcription products are short, untranslated RNAs, most less than 300 nucleotides long. In addition to 5S rRNA, they include tRNAs; 7SL RNA, which is required for introducing proteins into membranes as part of the signal recognition particle (see Chapter 18); and several RNAs involved in mRNA, tRNA, and rRNA processing. Perhaps reflecting their varied gene products, Pol III promoters differ in sequence and in components. The promoters of tRNA genes include two segments, Box A and Box B, located a short distance apart within the tRNA-coding sequence (Figure 15-21a). The 5S rRNA gene promoter includes Box A and Box C (Figure 15-21b). Other promoters contain the TATA box to which TBP can bind directly, just as for Pol II promoters.

Like the other eukaryotic polymerases, Pol III requires transcription factors. The tRNA genes require TFIIIB and TFIIIC, whereas the 5S rRNA gene requires TFIIIB, TFIIIC, and TFIIIA. Transcription of tRNA genes begins when TFIIIC binds to the promoter boxes within the gene, and then recruits TFIIIB. TFIIIB includes TBP and recognizes the DNA just upstream from the transcription start site. Together, these factors recruit Pol III to the transcription start site; TFIIIC is transiently displaced as the polymerase transcribes through its binding site in the DNA. In 5S rRNA transcription, TFIIIA binds to the DNA within the transcribed region and helps recruit TFIIIC.

Pol II–catalyzed transcription is responsible for producing all mRNAs in the eukaryotic cell, as well as transcripts, such as microRNAs, that can base-pair with mRNAs and help regulate their expression (see Chapter 22). Consisting of 12 subunits, Pol II is strikingly more complex than its bacterial counterpart, yet it has remarkable similarities in structure, function, and mechanism (Figure 15-22). The largest subunit (RPB1) exhibits a high degree of homology to the β′ subunit of bacterial RNA polymerase. Another subunit (RPB2) is structurally similar to the bacterial β subunit, and two others (RPB3 and RPB11) show some structural homology to the two bacterial α subunits (see Table 15-1). Pol II must function with genomes that have multiple chromosomes and with DNA molecules more elaborately packaged than those in bacteria. The need for protein-protein interactions with the numerous other protein factors required to navigate this labyrinth largely accounts for the added complexity of Pol II and the other eukaryotic polymerases.

An overview of the Pol II transcription complex is shown in Figure 15-23. Playing a role much like that of sigma factors in helping bacterial RNA polymerase recognize and bind promoter sequences, general transcription factors associate with promoter DNA and recruit Pol II to form a preinitiation complex (Figure 15-23a, step 1). The preinitiation complex is converted to an initiation complex by unwinding the DNA (step 2). During initiation (step 3), the C-terminal domain (CTD) of Pol II is phosphorylated and some transcription factors are released (Figure 15-23b). Elongation (step 4) proceeds as in bacteria. Transcription is terminated (step 5) and the Pol II CTD is dephosphorylated. Each step is associated with characteristic proteins.

The transcription initiation mechanism has been most extensively studied for Pol II. Recruitment begins with binding of the TATA box by TFIID. Like many transcription factors, TFIID is a multiprotein complex, which includes TBP and TAFs. The TAFs fine-tune TFIID by changing the affinity of TBP for DNA, helping the transcription factor bind certain promoters. Once the TBP-DNA interaction is stabilized, other transcription factors, and Pol II itself, can stably associate with the promoter to form the preinitiation complex.

541

542

The discovery of TBP and its importance as a general transcription factor required for all transcription by Pol II raised questions about how and why it binds so specifically to the TATA element. This mystery of the transcription initiation process was solved when the research groups of Paul Sigler and Stephen Burley independently determined the molecular structure of TBP bound to DNA, using x-ray crystallography. The structure revealed that TBP sits on the DNA double helix much like a saddle, with an extended β sheet and loop “stirrups” in contact with the minor groove of the TATA box sequence (Figure 15-24). This unconventional mode of DNA recognition—most DNA-binding proteins recognize DNA by inserting α helices into the major groove—bends the DNA by positioning two pairs of Phe residue side chains between base pairs at each end of the recognition sequence. The bending opens and widens the minor groove, enabling hydrogen bonding between protein side chains and the minor-groove edges of the DNA bases. The observed helical bending explains why A=T base pairs are favored in the recognition sequence: they are more easily distorted to allow opening of the minor groove. Because TBP is used by all three classes of eukaryotic polymerases, a similar mechanism may account for promoter recognition in all cases.

In addition to TBP, Pol II requires an array of transcription factors to form an active transcription complex. The general transcription factors required at every Pol II promoter are highly conserved in all eukaryotes. Using cell-free systems pioneered by Robert Roeder at Rockefeller University, in which purified proteins were added back to the reaction mix to reconstitute active transcription complexes, it was possible to determine the identity and order of proteins needed for transcription initiation. When TBP, as part of TFIID, binds to the TATA box, it is bound in turn by the transcription factor TFIIB, which binds a larger site on the DNA than TBP alone. A third transcription factor, TFIIA, is not always essential in experiments using purified proteins to monitor transcription. However, mutant cells that lack TFIIA are not viable, showing that this factor is essential in vivo. TFIIA, when it binds to TFIID, unmasks TBP and enables it to bind efficiently to the TATA box. Pol II is bound to the complex through a mutual interaction with TFIIF. Finally, TFIIE and TFIIH bind to create the closed complex, analogous to the closed complex described for bacterial RNA polymerase (see Section 15.2). TFIIH has DNA helicase activity that promotes unwinding of the DNA near the transcription start site. This unwinding creates an open complex that is competent to begin transcription. Counting all the subunits of the various essential factors (excluding TFIIA), this minimal active assemblage has more than 30 polypeptides!

543

Once assembled on a promoter, Pol II typically produces a few abortive transcripts before entering the elongation phase of transcription—behavior similar to that of bacterial RNA polymerase. In contrast to bacterial polymerase, however, Pol II must be chemically modified by the addition of phosphate groups to its CTD to disengage from the promoter and begin elongating a transcript.

The CTD, part of the largest polymerase subunit, consists of multiple repeats of the seven amino acid sequence Tyr-Ser-Pro-Thr-Ser-Pro-Ser. In yeast and animals, the CTD mainly functions as a docking platform to recruit transcription and processing factors during the transcription cycle. CTD-associated factors have a variety of functions, including to catalyze mRNA 5′ capping and 3′-end processing, pre-mRNA splicing, and histone modification. To recruit all these different proteins, the CTD uses distinct chemical codes. Reversible phosphorylations of the serines in the second and fifth positions (Ser² and Ser⁵) of the CTD repeat sequence are the primary CTD codes and are crucial for regulating transcription and the binding of mRNA processing factors. The protein enzymes responsible for these phosphorylations, the CTD kinases, are conserved from yeast to metazoans. To enable recruitment of other kinds of proteins, the CTD adopts additional modifications, including phosphorylation of Tyr¹, Ser⁷, and Thr⁴, as well as cis-trans isomerization of Pro³ and Pro⁶, in the CTD repeat sequence.

TFIIE and TFIIH are released during synthesis of the initial 60 to 70 nucleotides of RNA, as Pol II enters the elongation phase of transcription. Notably, phosphorylation of Pol II CTD also influences downstream processing of the RNA transcript, providing a mechanism for coupling transcription to RNA splicing and intracellular transport (as discussed in Chapter 16).

In the elongation phase, polymerase activity is greatly enhanced by elongation factors. They suppress pausing during transcription, enhance polymerase editing of misincorporated bases by hydrolysis (as for bacterial RNA polymerase), and recruit protein complexes involved in posttranscriptional processing of the mRNA. Once the RNA transcript is completed, transcription is terminated. In eukaryotes, termination is often triggered by endonucleases that recognize and cleave specific sequences in the newly synthesized RNA, leading to disassembly and dissociation of the transcription complex. Pol II is then dephosphorylated and recycled, readying it to initiate another transcript (see Figure 15-23).

As we have seen, the initiation and control of eukaryotic mRNA synthesis requires a large set of evolutionarily conserved general transcription factors that function at most, if not all, genes. These include initiation factors TFIIB, TFIID (which includes the TATA-binding protein, TBP), TFIIE, TFIIF, and TFIIH—which comprise the minimal set of helper proteins necessary and sufficient for in vitro selective binding and accurate transcription initiation by Pol II from core promoters. In vivo, in yeast cells, the multiprotein Mediator complex is also required for the regulated transcription of nearly all Pol II–dependent genes. Its presence also in humans implies a similar central role in Pol II–catalyzed transcription.

Mediator functions as an intermediary between specific transcription factors bound at upstream promoter elements or enhancers and the Pol II complex and general initiation factors bound at the core promoter (Figure 15-25a). First discovered and purified from yeast by Roger Kornberg and his colleagues, Mediator was found to be required for transcriptional activation by specific activators in vitro, using a reconstituted enzyme system containing purified Pol II and general initiation factors. Yeast Mediator has 20 subunits in three distinct subdomains, referred to as the head, middle, and tail modules (Figure 15-25b). An additional module, which includes a kinase enzyme complex, is associated with a subset of yeast Mediator complexes. The presence of the kinase corresponds to repression of a subset of genes, suggesting a role for Mediator in transcriptional down-regulation as well as in activation.

Human Mediator contains a set of consensus subunits similar to those in yeast. As in yeast, multiple forms of Mediator seem to function differently in the transcriptional control of different sets of genes. In particular, the kinase module can exert a repressive effect when associated with the mammalian Mediator, whereas other auxiliary proteins are associated with an activating form of Mediator.

The mechanisms by which Mediator complexes control mRNA synthesis involve direct interactions with DNA-binding transcription activators bound at upstream promoter elements and enhancers, interactions with Pol II, and interactions with one or more of the general initiation factors bound at the core promoter. Mediator supports transcriptional activation, at least in part, by increasing the rate and/or efficiency of assembly of the Pol II preinitiation complex. Mammalian Mediator complexes influence several steps during this assembly, including the recruitment of TFIID (or TBP), Pol II, and the other general initiation factors to the core promoter.

544

The three eukaryotic RNA polymerases use different strategies for terminating transcription, although these mechanisms have some aspects in common. The Pol III and Pol I termination pathways seem to be simpler than the Pol II pathway. Pol III terminates transcription at T-rich sequences in the DNA template located a short distance from the 3′ end of the mature RNA, assisted by just a few protein factors. Pol I terminates at a terminator site located downstream from the pre-rRNA sequence and requires terminator recognition by specific protein factors.

In contrast, Pol II termination does not occur at a conserved site or at a constant distance from the 3′ end of mature RNAs. In mammals, it takes place anywhere from a few base pairs to several kilobase pairs downstream from the 3′ end of the mature transcript. The 3′ end of the mature mRNA includes a stretch of A nucleotides, called a poly(A) tail, that is essential for translation into protein (see Chapter 18). A polyadenylation signal sequence (typically AAUAAA) is present in the primary transcript (and directly encoded by the DNA). Factors responsible for cleavage of the primary transcript bind to the AAUAAA sequence, resulting in cleavage somewhat downstream from that position. Only after this cleavage is the poly(A) tail added. Pol II termination is coupled to 3′-end processing of precursor mRNA transcripts, and the intact polyadenylation signal is necessary for termination of transcription of protein-coding genes in human and yeast cells.

Two different models have been proposed to explain how 3′-end processing contributes to Pol II transcription termination. The first, known as the allosteric or antiterminator model, proposes that transcription through the poly(A) site triggers conformational changes in the Pol II elongation complex caused by the dissociation of elongation factors and/or association of termination factors. This is analogous to the hairpin model of termination in bacteria (see Section 15.2). According to the allosteric model, these conformational changes in Pol II cause it to fall off the DNA template. The second model, the torpedo model, suggests that after mRNA synthesis is complete, Pol II remains associated with the DNA template and continues the transcription reaction to extend the 3′ end of the mRNA. Protein complexes cleave the mRNA at the polyadenylation site, producing a new 3′ end that can be recognized and extended by the enzyme poly(A) polymerase. The new 5′ end of the downstream, or residual, mRNA strand becomes a substrate for an enzyme called Xrn2, a 5′→3′ exonuclease (an exoribonuclease) that attaches to the CTD of Pol II (Figure 15-26). Xrn2 proceeds to degrade the uncapped residual RNA in the 5′→3′ direction until it reaches Pol II. Similar to the ρ factor in ρ-dependent termination in bacteria, Xrn2 triggers dissociation of Pol II by either pushing the polymerase off the DNA template or pulling the template out of the RNA polymerase.

545

In eukaryotes, transcription is coupled to other activities, including the repair of damaged DNA and various kinds of RNA processing and transport events. Researchers noticed that DNA damage repair and mRNA processing and transport are more efficient for genes that are actively being transcribed. Furthermore, DNA lesions in the template strand are repaired somewhat more efficiently than lesions in the coding (nontemplate) strand. For DNA repair, these remarkable observations are explained by the alternative functions of the TFIIH subunits. Not only does TFIIH participate in forming the closed complex during assembly of a transcription complex, but some of its subunits are also essential components of the separate nucleotide excision repair complex (see Chapter 12). When Pol II transcription stalls at the site of a DNA lesion, TFIIH reassociates with the DNA and transcription machinery. TFIIH can then interact with the lesion and recruit the entire nucleotide excision repair complex. Mutations causing deletion of certain TFIIH subunits produce human diseases. Two examples are xeroderma pigmentosum, with its associated photosensitivity and tumor susceptibility, and Cockayne syndrome, which is characterized by arrested growth, photosensitivity, and neurological disorders.

Eukaryotic mRNA is processed in a variety of ways before it is shipped across the nuclear membrane to the cytoplasm for translation. We discuss the mechanisms of these processing events in Chapter 16, but it is important to note here that like DNA repair, mRNA processing is naturally linked to transcription. This is possible because some of the same proteins required for elongating RNA transcripts are also required for 5′-end processing (5′ capping) of the RNA. Because these activities are coupled, transcripts can be processed as they are synthesized.

547

548

549