Journal of Cell Science, Supplement 19,13-19 (1995)Printed in Great Britain © The Company of Biologists Limited 1995

13

The influence of 5' and 3' end structures on pre-mRNA metabolism

Joe D. Lewis, Samuel I. Gunderson and lain W. MattajGene Expression Programme, European Molecular Biology Laboratory, Meyerhofstrassel, 69117 Heidelberg, Germany

SUMMARY

The 5' cap structure of RNA polymerase II transcripts and the poly(A) tail found at the 3' end of most mRNAs have been demonstrated to play multiple roles in gene expression and its regulation. In the first part of this review we will concentrate on the role played by the cap in pre- mRNA splicing and how it may contribute to efficient and specific substrate recognition. In the second half, we will

discuss the roles that polyadenylation has been demon- stated to play in RNA metabolism and will concentrate in particular on an elegant mechanism where regulation of polyadenylation is used to control gene expression.

Key words: pre-messenger RNA processing, splicing, nuclear cap binding complex, polyadenylation

INTRODUCTION

The 5 ' cap structure of RNA polymerase II (pol II) transcripts, (m7G(5/)ppp(5')N), and the poly(A) tail found at the 3' end of most mRNAs have been shown to play multiple roles in gene expression and its regulation. The cap structure is added co- transcriptionally (Salditt-Georgieff et al., 1980) and is in general not further modified. One of the exceptions to this are the pol II transcribed spliceosomal U snRNAs. After export from the nucleus, the cap of these RNAs is modified in the cytoplasm from a m7G to a m2,2’7G cap structure in a reaction which is dependent on an intact Sm binding site, the site through which these RNAs bind to a group of common (Sm) proteins. In the case of the U snRNAs this modification can facilitate nuclear import (reviewed by Izaurralde and Mattaj, 1992a).

The poly(A) tail found at the 3' end of most mRNAs is, in contrast to the cap, a highly dynamic modification. The length of the poly(A) tail has been shown to vary at different stages during the gene expression pathway. It is thought to play a role in RNA stability (Decker and Parker, 1994), and also to be important in the developmental regulation of mRNA translation (Wickens, 1990). Polyadenylation of the 3' end of prokaryotic transcripts has also been shown to play a role in RNA stability (reviewed by Cohen, 1995) indicating perhaps the ancient nature of this mechanism for modulating gene expression.

The remainder of this review is in two sections. In the first part, recent developments in elucidating the role played by the cap in pre-mRNA splicing are discussed. In the second part, the factors involved in 3' end processing and polyadenylation, and how the U1A protein interacts with them to autoregulate expression of its own mRNA, are reviewed.

THE CAP STRUCTURE

The cap structure, (m7G(5')ppp(5')N), found at the 5' end of

all RNA pol II transcripts, has been long known to play multiple roles in the gene expression pathway. Capping of nascent transcripts occurs co-transcriptionally (Salditt- Georgieff et al., 1980). In nuclear run-on experiments it has been demonstrated, in the case of the Drosophila heat shock genes, that capping takes place during a window when approximately 20-30 nucleotides of the nascent RNA have been synthesized (Rasmussen and Lis, 1993). Thus it is likely that capping is the first postranscriptional RNA modification to take place. Characterisation of mutants in the RNA guanylyltrans- ferase (capping enzyme) have also demonstrated the importance of this modification for viability in both Saccharomyces cerevisiae and Schizosaccharomyces pombe (Shibagaki et al., 1992; Shuman et al., 1994).

Roles of the cap structureThe ability of the cap to protect mRNA from 5' exoribonu- cleases was one of its first described functions (Furuichi et al., 1977; Shimotohno et al., 1977). Other processes where the cap has been shown to be of importance are shown dia- gramatically in Fig. 1. The cap has been shown to play a role in pre-mRNA splicing both in vitro (Konarska et al., 1984; Krainer et al., 1984; Edery and Sonnenberg, 1985; Patzelt et al., 1987, Izaurralde et al., 1994) and in vivo using microinjected Xenopus oocytes (Inoue et al., 1989). Nuclear export of mRNAs (Jaramolowski et al., 1994) and especially of U snRNAs has also been shown to be facilitated by the cap in vivo (Hamm and Mattaj, 1990, Izaurralde et al., 1992). Finally, efficient mRNA translation also requires the cap stucture. In translation, the cap is recognised by the cap binding protein eIF-4E which is part of a multi-protein complex (eIF-4F) required for translational initiation (Sonnenberg, 1988; Rhoads, 1988).

The available evidence suggests that the nuclear functions of the cap are mediated through protein factors which recognise this modification specifically. Several such activities have been described in nuclear extracts although most have not

14 J. D. Lewis, S. I. Gunderson and I. W. Mattaj

Fig. 1. The cap is involved in multiple aspects of RNA metabolism. Shortly after the initiation of transcription RNA polymerase II, the RNAs are capped. Splicing of the pre-mRNA, nuclear export and translation all show varying degrees of dependence on the cap structure. This is indicated by an arrow connecting the cap and the process affected.

been fully characterised (e.g. Patzelt et al., 1983; Rozen and Sonnenberg, 1987).

W e have recently described the purification o f a nuclear cap binding com plex (CBC) which specifically recognises the mono-m ethyl guanosine cap structure. CBC has two com ponents, CBP80 (Ohno et al., 1990; Izaurralde et al., 1992) and a second smaller subunit CBP20 (Izaurralde et al., 1994). We dem onstrated that im m uno-depletion of CBC from nuclear splicing extracts could efficiently inhibit splicing of an Adenovirus pre-mRNA. Analysis o f the splicing defect caused by CBC depletion showed that an early step in spliceosome assembly was inhibited. This inhibition of splicing could be relieved by adding back CBC purified from HeLa cell extracts.

The spliceosome cycle and pre-mRNA splicingSplicing out of intronic sequences from pre-m RNAs takes place in a large dedicated ribonucleoprotein complex called the spliceosome. W ithin the spliceosome two sequential trans- esterification reactions take place to produce the mature mRNA and the intron lariat (Fig. 2).

Many o f the com ponents o f the spliceosome have been described, although the functions of most remain poorly characterised. The spliceosome consists o f the five spliceosomal small nuclear RNAs (snRNAs) (U l, U2, U4/U6 and U5) and a large num ber of associated proteins. The snRNAs are found as small nuclear ribonucleoprotein particles (snRNPs) com posed o f one RNA (with the exception of the U4/U6

Fig. 2. Simplified schematic representation of the spliceosome assembly/disassembly cycle. The pre-mRNA branch point is marked by a filled circle. The various stages of spliceosome assembly which can be indentified in vitro, together with their snRNP composition, are labelled. A summary of the assembly steps is given in the text. For reasons of clarity the numerous non-snRNP protein splicing factors, with the exception of U2AF and CBC, are omitted. The evidence that CBC stays associated with the RNA throughout the cycle is discussed in the text.

particle), a set of Sm core proteins, which are common to all spliceosomal snRNPs, and a num ber o f proteins specific for each particle (for example the U l snRNP specific A protein; reviewed by Luhrm ann et al., 1990). In addition to the snRNPs there are a large number of non-snRNP splicing factors whose role in splicing is currently under active investigation (reviewed by Lam m and Lam ond, 1994).

The basic steps o f spliceosome assembly which can be separated in vitro are shown in Fig. 2. All the spliceosomal snRNPs have been shown, but for simplicity most o f the non- snRNP factors have been omitted. Prior to its sequestration to the splicing pathway, the pre-m RNA is com plexed with the heterogenous nuclear RNP proteins. This packaging may be im portant for the following steps. The first pre-splicing com plex which can be detected is the E (for early) complex which forms on the pre-mRNA in the absence o f ATP (M ichaud and Reed, 1991). Analysis o f this com plex has shown that U l snRNP is bound to the 5 ' splice site and the

Influence of 5' and 3' end structures on pre-mRNA metabolism 15

non-snRNP splicing factor U2 auxiliary factor (U2AF) is bound to the poly-pyrimidine tract and may bring the 5' and 3' splice sites into close proximity (Michaud and Reed, 1993). Additional proteins are also found in E complex although their roles in complex formation are not clear. It has been postulated that formation of E complex is the step which commits the pre- mRNA to the splicing pathway (Michaud and Reed, 1991) although definitive proof is still required. A U1 containing ‘commitment complex’ that is the first detectable precursor to the spliceosome has been previously described in yeast (Rosbash and Séraphin, 1991), and this may be the equivalent to the E complex. In the next step, U2 snRNP associates with the branch point sequence to form the A complex in a process which requires ATP and U2AF (Fig. 2). The mature spliceosome is then formed when the U4/6.U5 tri-snRNP joins the A complex to form B complex, the mature spliceosome. This undergoes the first step of catalysis (complex C) and the second catalytic step takes place to produce the mature messenger RNA and the intron lariat. The mature mRNA is then released and exported to the cytoplasm while the intron lariat is debranched and degraded. In vivo the spliceosome is then disassembled and the factors take part in further rounds of splicing. This process of assembly, splicing, disassembly and release of the spliceosome components is referred to as the splicing cycle.

CBC is required for early steps in spliceosome assemblyThe complexity of the spliceosome may in part reflect the need to accurately recognise and process pre-mRNAs, and evidence is accumulating that multiple proof-reading events take place during spliceosome assembly to ensure that only bona fide introns are recognised. However, although the role of the snRNAs as determinants of splice site selection and their role in accuracy of cleavage has been extensively characterised (reviewed by Newman, 1994) very little is known about how an intron is recognised by protein factors in the steps preceding spliceosome assembly. Recent evidence from our laboratory suggests that the cap of the pre-mRNA may be involved in this process.

When splicing is inhibited, either in vitro or in vivo by cap analogues, no accumulation of splicing intermediates is observed (Konarska et al., 1984; Krainer et al., 1984; Edery and Sonnenberg, 1985; Patzelt et al., 1987; Inoue et al., 1989; Izaurralde et al., 1994). Native gel analysis of spliceosome formation in CBC depleted extracts showed that levels of A complex formation were markedly reduced compared to the control extracts (Izaurralde et al., 1994). As a consequence, the levels of B and C complex formation were also reduced (see Fig. 2). Together, these observations raise the exciting possibility that CBC may be involved in the steps required to recognise the pre-mRNA. In support of this idea, recent immuno-precipitation experiments have shown that CBC is found to be associated with the pre-mRNA as well as the splicing intermediates and mature message (Joe Lewis, unpublished observations; also summarised in Fig. 2). One attractive model is that CBC bound to the cap defines an RNA as a potential splicing substrate and can directly or indirectly facilitate binding of the U1 snRNP to the 5' splice site of the pre- mRNA. One prediction of this model would be that CBC should be required for E complex formation. In order to test

this model, work is underway to determine at exactly which step CBC is required in pre-spliceosome complex formation.

Another related, and by no means mutually exclusive, model is that CBC interacts in concert with members of the SR splicing factor family to promote recognition of the pre- mRNA. These proteins are characterised by the presence of arginine-serine dipeptide repeats and an RNA recognition motif (Zahler et al., 1992; reviewed by Lamm and Lamond, 1994). Evidence is accumulating that members of this family play a role in early steps of spliceosome assembly (Horowitz and Krainer, 1994). Perhaps the best example in mammalian systems is the alternative splicing of the bovine growth hormone pre-mRNA. Splicing of intron D is dependent on the presence of an exonic splicing enhancer (ESE) which binds specifically the SR protein SF2/ASF (Sun et al., 1993). Increasing the amount of SF2/ASF in extracts stimulates splicing of intron D through the ESE showing that it plays a positive role in intron recognition. Recent evidence has shown that physical depletion of U1 snRNP from, or inactivation of U1 snRNP in, splicing extracts and the consequent inhibition of splicing could be overcome by the addition of a fraction enriched in SR proteins (Crispino et al., 1994; Tarn and Steitz, 1994, respectively). However, splicing inhibition caused by inactivation of the U2 snRNP could not be overcome by addition of exogenous SR proteins or the splicing factor SC35 (Tarn and Steitz, 1994) suggesting that they are required only for stages preceeding A complex formation. This is supported by data which show that SR proteins can enhance E complex formation in splicing extracts (Staknis and Reed, 1994). What is clear is that recognition of intron-containing RNAs will involve many different factors which may not be identical between different RNAs, although cap recognition by CBC may be common to all. Nevertheless it is hoped that some general principles regarding pre-mRNA recognition should emerge from continuing studies in this area of research.

3' END PROCESSING AND POLYADENYLATION

The 3' end of almost all eukaryotic mRNAs comprises a homopolymer of adenosine residues ranging from 20 to 250 nucleotides in length. The poly(A) tail is post-transcriptionally added to pre-mRNA in the nucleus by a two step process called cleavage and polyadenylation (Fig. 3) which is catalyzed by a complex of proteins. The exact location on the pre-mRNA to which the poly(A) tail is added is called the polyadenylation site. Since it was first detected, the polyadenosine, or poly(A), tail has been proposed to function in the regulation of nearly every aspect in the biosynthesis and activity of mRNA. Of these functions, the ones for which experimental evidence exists include: (1) regulation of translational efficiency of mRNA during oocyte maturation (Wickens, 1990); (2) regulation of the mRNA degradation pathway (Bernstein and Ross, 1989; Decker and Parker, 1994); (3) being required for translation of yeast mRNAs (Sachs, 1993; Jackson and Standart, 1990), which would help to direct ribosomes to translate only intact mRNA; and possibly (4) being required for mRNA export from the nucleus. Besides affecting the biological activity of mRNA, the choice of the polyadenylation site can have dramatic effects on both the spatial and temporal expression of mRNA-encoded genes. The majority of tran

16 J. D. Lewis, S. I. Gunderson and I. W. Mattaj

Fig. 3. Schematic representation of the cleavage and polyadenylation reaction. Within the nucleus the factors necessary for cleavage are assembled onto the AAUAAA sequence found in the capped pre- mRNA transcript. DSE, downstream sequence element. The abbreviations for the proteins can be found in the text. The scissors indicates that the complex is poised to endonucleolytically cleave the substrate pre-mRNA in two. The downstream half is degraded in the nucleus while the upstream half undergoes polyadenylation. When about 250 adenosine (A) residues are added the reaction stops and the processed mRNA is exported to the cytoplasm.

scription units have only one polyadenylation site. However, a num ber of genes are known to have alternative polyadenylation sites which are used to regulate the expression of their protein products (Leff et al., 1986).

Cleavage and polyadenylation o f pre-mRNAs has been best characterized in vertebrate systems. It is only recently that attention has becom e focused on understanding these reactions in yeast, therefore this review will focus primarily on what has been learned from vertebrates.

RNA sequencesThe AAUAAA motif, which is invariably found 10-30 nucleotides upstream o f the polyadenylation site in higher eukaryotes, is the best understood and most well-conserved ds-acting RNA element that is required for cleavage and polyadenylation to occur. Saturation mutagenesis of this m otif along with analysis of the frequencies o f naturally occuring variants leads to the conclusion that more than 98% of these motifs are o f the AAUAAA or AUUAAA type (Sheets et al.,

1990). Because AAUAAA sequences are found at other locations scattered throughout mRNAs it was self-evident that at least one other RNA elem ent must be necessary to define the polyadenylation site. One such additional element, the DSE, downstream sequence element, is required for cleavage and polyadenylation and is found downstream o f the cleavage site (reviewed by W ahle, 1995). DSEs are found either alone or in several copies and contain a rather degenerate consensus sequence. They tend to be U- or GU rich, and have accordingly also been called Uor G-U rich elem ents. DSEs are thought to function in selecting where cleavage and poladeny- lation will take place (Chou et al., 1994; M acDonald et al., 1994). In addition to the DSE there are other RNA elements which modulate the efficiency o f cleavage and polyadenylation. These other elements are found in close proximity, upstream or downstream, to the AAUAAA sequence. They have been studied in viruses, where life-cycle dependent changes in polyadenylation site usage are common.

Protein factors and the reaction pathwayThe developm ent o f an in vitro cleavage and polyadenylation system about 10 years ago (M oore and Sharp, 1985) has led to the identification, cloning and characterizing o f most o f the protein factors involved in cleavage and polyadenylation. In vivo the cleavage and polyadenylation reactions are tightly coupled (the cleavage interm ediate is not detectable) making it difficult to study these two reactions separately. However, methods were developed that permitted cleavage and polyadenylation to occur as independent reactions in vitro, enabling each o f the protein factors to be assigned to a particular reaction or to both reactions. The results o f a large body of work are summ arized in Fig. 3 which gives the current state of knowledge about the factors that mediate the cleavage and polyadenylation reaction (for reviews see W ahle, 1995; and W ahle and Keller, 1992). W hile the details o f what is shown in Fig. 3 may change as w ork progresses it is likely that the basic outline will be correct.

CleavageUpon synthesis o f the pre-mRNA, a com plex of proteins assembles at or around the AAUAAA sequence (see Fig. 3; Gilmartin and Nevins, 1989). Probably one o f the first factors to bind is called CPSF for cleavage and polyadenylation specificity factor (Bienroth et al., 1991; Murthy and Manley, 1992). Highly purified CPSF can bind specifically to RNAs containing the AAUAAA sequence (Keller et al., 1991). The CPSF- AAUAAA com plex serves as a nucleation point through which the rem aining cleavage factors are recruited. CStF, cleavage and stimulatory factor, has been cloned and binds to the DSE which is found downstream of the cleavage site (Takagaki et al., 1990; Gilmartin and Nevins, 1991). Subsequent to binding the RNA, CStF stabilizes the CPSF-AAU AAA complex. Cleavage factors (CFs) are also necessary for the cleavage reaction to occur, but they remain poorly characterized. The protein factor(s) which carries out the endonucleolytic cleavage of the RNA backbone has not been identified but it is believed that it will be one o f these CFs. The final factor necessary for efficient cleavage in vitro is, curiously, poly(A) polym erase (PAP), the enzym e which catalyzes the addition of the poly(A) tail (Christofori and Keller, 1989), although this may not be the case for all polyadenylation signals (Takagaki

Influence of 5' and 3' end structures on pre-mRNA metabolism 17

et al., 1989). O nce the com plex is assembled the phosphodi- ester backbone o f the RNA is cleaved, resulting in two RNA fragments. The upstream RNA fragment has a 3'-OH and subsequently undergoes polyadenylation (see below) whereas the downstream RNA fragm ent is rapidly degraded in the nucleus. A fter cleavage, CStF and the CFs may exit the cleavage and polyadenylation com plex leaving only CPSF and PAP still bound to the AAUAAA sequence. The evidence for this comes from in vitro experim ents where it is possible to reconstitute specific polyadenylation (see below) in the absence of CStF and the CFs.

PolyadenylationIn the polyadenylation reaction a poly(A) tail is added to the 3'-OH end of the upstream RNA fragment by PAP. Purified PAP from either HeLa cells or recom binant PAP are capable o f inefficiently polyadenylating any RNA (including tRNAs, rRNA or snRNAs) that has a free 3'-OH (W ahle et al., 1991; Raabe et al., 1991). Such polyadenylation does not occur in vivo and is therefore called non-specific polyadenylation in order to differentiate it from specific polyadenylation in which only RNAs having an AAUAAA sequence are polyadenylated. Specific polyadenylation also requires the presence o f CPSF. Thus, in vitro, these two factors are sufficient to reconstitute AA UAAA-dependent polyadenylation. Reconstituted polyadenylation occurs in 3 distinct phases (Sheets and W ickens, 1989). In the first phase poly(A) addition is slow and distributive and depends on the AAUAAA sequence and CPSF. In the second phase, when the growing poly(A) tail reaches a length of 10 residues, poly(A) addition becomes rapid and processive and depends on the poly(A) tail and a third factor which enters the reaction called PABII, poly(A) binding protein II (W ahle, 1991). In the third phase, after the addition of about 200 adenosine residues, the rate o f polyadenylation slows considerably, eventually stopping at around 250 adenosine residues, which is the length of poly(A) tails added to mRNAs in the nucleus in vivo. PABII also functions in controlling the final length of the poly (A) tail.

U1A protein autoregulation by inhibition of polyadenylationThe U1 snRNP particle is involved in splicing of pre-mRNA by binding to the 5 ' splice site o f introns (see first part o f review). The U1 snRNP particle com prises U1 snRNA com plexed with the Sm core proteins and three U1 specific proteins. One o f these is the U1A protein which binds with high affinity to the second stem loop of U 1 snRNA (Scherly et al., 1989). W ithin the loop part o f this hairpin is the sequence AUUGCAC that has been shown to be required for specific binding o f U1A protein. M ore recently it was discovered that the U1A protein also specifically binds to a RNA sequence found adjacent to the polyadenylation signal o f the U1A pre- mRNA (Boelens et al., 1993). As shown in Fig. 4 this RNA sequence contains 2 short motifs which closely or perfectly match the U 1A binding site in stem-loop 2 of U 1 snRNA. Also shown is the secondary structure for this RNA sequence, which was determ ined by a com bination o f com puter prediction, phylogenetic analysis, enzymatic and chemical structure probing, and mutagenesis (van Gelder et al., 1993). Note that the 2 sequences (AUUGC/UAC) within this structure are in single stranded loops (as they are found in stem-loop 2 of U1 snRNA)

Fig. 4. Mechanism for how U1A protein autoregulates its own production by inhibition of polyadenylation. Vertebrate CPSF and PAP, the factors needed for the polyadenylation step of the cleavage and polyadenylation reaction, are shown bound to the U1A pre- mRNA. The cleavage reaction has already occurred. The shaded circle is U1A protein which, when in excess, binds as two molecules to the U1A pre-mRNA, shown in its characteristic secondary structure. Once bound the UlA-pre-mRNA complex bypasses RNA- bound CPSF and specifically interacts with and inhibits the activity o f vertebrate PAP. The shaded area in the PAP oval represents a domain necessary for PAP to bind to and be inhibited by RNA- bound U1A protein. Likewise, the stippled area within the U1A shaded circle represent a small domain which is needed to interact with and inhibit PAP.

and that one molecule of U1A protein binds each sequence. W hen both sites are occupied by U1A protein, polyadenylation o f the U1A pre-mRNA is inhibited both in vitro and in vivo (Boelens et al., 1993). M utagenesis studies dem onstrated that one molecule of RNA-bound U1A protein is not sufficient to efficiently inhibit polyadenylation (van Gelder et al., 1993). U1A protein inhibition of its own polyadenylation results in a down-regulation of the U1A mRNA levels in vivo. Fig. 4 shows how U1A autoregulation is thought to work. W hen there is excess U1A protein in the cell, it binds its own pre-mRNA, inhibiting the cleavage and polyadenylation reaction. Unpolyadenylated U1A pre-mRNA is thought to remain in the nucleus (where it is probably rapidly degraded). This reduces the effective concentration of U1A mRNA resulting in a reduction of U1A protein synthesis.

In vitro studies have elucidated the m echanism o f U1A protein inhibition o f polyadenylation (Gunderson et al., 1994). Surprisingly, U1A protein com plexed with U1A pre-mRNA has no detectable effect on the efficiency of cleavage. It is, instead, the polyadenylation rection which is inhibited. This may result in a more effective form o f regulation because any cryptic downstream polyadenylation sites which could be potentially used will be removed from the RNA.

Since specific polyadenylation requires only CPSF and PAP, U1A protein must be blocking one or both of these factors. It was shown that RNA-bound U1A protein had no effect on CPSF binding to the AAUAAA recognition sequence,

18 J. D. Lewis, S. I. Gunderson and I. W. Mattaj

however, it efficiently blocks the non-specific polyadenylation activity of mammalian PAP (Gunderson et al., 1994). This regulatory mechanism can be reconstituted in vitro using only recombinant PAP, U1A protein and substrate RNA. More unexpectedly, the RNA-bound U1A protein does not inhibit the polyadenylation activity of yeast PAP indicating that RNA- bound U1A protein is not sterically blocking access of PAP to the 3'-OH of the substrate RNA. In fact, RNA bound U1A protein specifically interacts directly with mammalian, but not yeast, PAP. More recently it has been shown that small domains in both U1A and PAP are needed for this interaction (S. I. Gunderson, unpublished observations).

The U1A autoregulatory circuit is the best understood example at the molecular level of regulation of RNA processing, and is notable because U1A protein and PAP are involved in distinct separable steps (splicing versus cleavage and polyadenylation) in the pre-mRNA processing pathway. Recent reports have blurred the lines separating these two steps. For example it has now been shown in vivo that splicing of the final intron is coupled to the cleavage and polyadenylation reaction (Niwa and Berget, 1991). In support of this, anti-Ul snRNP antibodies, and not other anti-RNP antibodies, specifically block cleavage and polyadenylation in HeLa nuclear extracts (Moore and Sharp, 1985; Hashimoto and Steitz, 1986). There are also a number of reports which have shown that the U1 snRNP particle is in some way associated with PAP (Raju and Jacob, 1988; Wassarmann and Steitz, 1993). Perhaps the U1A protein within the particle is contributing to this interaction with PAP. What the biological function of this interaction is remains an open question, however, integration of the multiple processing events that must occur to any pre-mRNA is clearly a desirable goal. Additionally, the expression of a growing number of genes seems to be controlled by a competition between utilization of adjacent splicing and cleavage/ polyadenylation signals. All of these examples point towards the idea that these two processing reactions are intimately connected in vivo. Thus, we can expect in the near future to learn more about how and why the splicing and polyadenylation machineries communicate with each other.

REFERENCES

Bernstein, P. and Ross, J . (1989). Poly(A) binding protein and the regulation of mRNA stability. Trends Biochem. Sci. 14, 373-377.

Bienroth, S., Wahle, E., Suter-Crazzolara, C. and Keller, W. (1991). Purification of the cleavage and polyadenylation factor involved in 3'- processing of messenger RNA precursors. J. Biol. Chem. 266, 19768-19776.

Boelens, W. C., Jansen, E. J . R., van Venrooij, W. J., Stripecke, R., M attaj, I. W. and Gunderson, S. I. (1993). The human U1 snRNP-specific U1A protein inhibits polyadenylation of its own pre-mRNA. Cell 72 , 881-892.

Chou, Z.-F., Chen, F. and Wilusz, J. (1994). Sequence and position requirements for uridylate-rich downstrean elements of polyadenylation signals. Nucl. Acids Res. 22 , 2525-2531.

Christofori, G. and Keller, W. (1989). Poly (A) polymerase purified from HeLa cell nuclear extract is required for both cleavage and polyadenylation of pre-mRNA in vitro. Mol. Cell. Biol. 9,193-203.

Cohen, S. (1995). Surprises at the 3 'end of prokaryotic RNA. Cell 80 ,819-832.Crispino, J., Blencowe, B. and Sharp, P. (1994). Complementation by SR

proteins of pre-mRNA splicing reactions depleted of U 1 snRNP. Science 265, 1866-1869.

Decker, C. J. and Parker, R. (1994). Mechanism of mRNA degradation in eukaryotes. Trends Biochem. Sci. 19, 336-340.

Edery, I. and Sonenberg, N. (1985). Cap-dependent RNA splicing in a HeLa nuclear extract. Proc. Nat. Acad. Sci. USA 82,7590-7594.

Furuichi, Y., LaFiandra, A. and Shatkin, A. J . (1977). 5'-Terminal structure and mRNA stability. Nature 266, 235-239.

Gilmartin, G. M. and Nevins, J . R. (1989). An ordered pathway of assembly of components required for polyadenylation site recognition and processing. Genes Dev. 3,2180-2189.

Gilmartin, G. M. and Nevins, J . R. (1991). Molecular analyses of 2 poly(A) site-processing factors that determine the recognition and efficiency of cleavage of pre-mRNA. Mol. Cell. Biol. 11, 2432-2438.

Gunderson, S. I., Beyer, K., M artin, G., Keller, W., Boelens, W. C. and M attaj, I. W. (1994). The human U1A snRNP protein regulates polyadenylation via a direct interaction with poly(A) polymerase. Cell 76 , 531-541.

Hamm, J. and M attaj I. W. (1990). Monomethylated Cap structures facilitate RNA export from the nucleus. Cell 63, 109-118.

Hashimoto, C. and Steitz, J . A. (1986). A small nuclear ribonucleoprotein associates with the AAUAAA polyadenylation signal in vitro. Cell 45 , 581591.

Horowitz, D. and Krainer, A. (1994). Mechanisms for selecting 5' splice sites in mammalian pre-mRNA splicing. Trends Genet. 10, 100-106

Inoue, K. 1., Ohno, M., Sakamoto, H. and Shimura, Y. (1989). Effect of the cap structure on pre-mRNA splicing in Xenopus oocyte nuclei. Genes Dev. 3, 1472-1479.

Izaurralde E. and M attaj, I. W. (1992). Transport of RNA between nucleus and cytoplasm Semin. Cell Biol. 3,279-288.

Izaurralde, E., Stepinski, J., Darzynkiewicz, E. and M attaj, I. W. (1992). A cap binding protein that may mediate nuclear export of RNA polymerase II- transcribed RNAs. J. Cell Biol. 118, 1287-1295.

Izaurralde, E., Lewis, J., McGuigan, C., Jankowska, M., Darzynciewicz, E. and M attaj, I. W. (1994). A nuclear cap binding complex involved in pre- mRNA splicing. Cell 78, 657-668.

Jackson, R. J . and Standart, N. (1990). Do the poly(A) tail and 3' untranslated region control mRNA translation? Cell 62, 15-24.

Jarmolowski, A., Boelens, W., Izaurralde, E. and M attaj, I. W. (1994). Nuclear export of different classes of RNA is mediated by specific factors. J. Cell Biol. 124, 627-635.

Keller, W., Bienroth, S., Lang, K. and Christofori, G. (1991). Cleavage and polyadenylation factor CPF specifically interacts with the pre-mRNA 3' processing signal AAUAAA. EMBO J. 10, 4241-4249.

Konarska, M. M., Padgett, R. A. and Sharp, P. A. (1984). Recognition of cap structure in splicing in vitro of mRNA precursors. Cell 38, 731736.

K rainer, A. R., M aniatis, T., Ruskin, B. and Green, M. R. (1984). Normal and mutant human p-globin pre-mRNAs are faithfully and efficiently spliced in vitro. Cell 36 , 993-1005.

Lamm, G. and Lamond, A. I. (1994). Non-snRNP splicing factors. Biochim. Biophys. Acta 1173, 247-265.

Leff, S. E., Rosenfeld, M. G. and Evans, R. M. (1986). Complex transcriptional units, diversity in gene expression by alternative RNA processing. Annu. Rev. Biochem. 55, 1091-1117.

Liihrmann, R., Kastner, B. and Bach, M. (1990). Structure of spliceosomal snRNPs and their role in pre-mRNA splicing. Biochim. Biophys. Acta 1087, 265-292.

MacDonald, C. C., Wilusz, J . and Shenk, T. (1994). The 64-kiloDalton subunit of the CstF polyadenylation factor binds to pre-mRNAs downstream of the cleavage site and influences cleavage site location. Mol. Cell. Biol. 14, 6647-6654.

Michaud, S. and Reed, R. (1991). An ATP-independent complex commits pre-mRNA to the mammalian spliceosome assembly pathway. Genes Dev. 5 , 2534-2546.

Michaud, S. and Reed, R. (1993). A functional association between the 5' and 3' splice sites is established in the earliest prespliceosome complex (E) in mammals. Genes Dev. 7 , 1008-1020.

Moore, C. L. and Sharp, P. A. ( 1985). Accurate cleavage and polyadenylation of exogenous RNA substrate. Cell 41, 845-855.

M urthy, K. G. K. and Manley, J . L. (1992). Characterization of the multisubunit cleavage-polyadenylation specificity factor from calf thymus. J. Biol. Chem. 267, 14804-14811.

Newman, A. (1994). RNA splicing. Activity in the spliceosome. Curr. Opin. Cell Biol. 6, 360-367.

Niwa, M. and Berget, S. M. (1991). Mutation of the AAUAAA polyadenylation signal depresses in vitro splicing of proximal but not distal introns. Genes and Dev. 5,2086-2095.

Ohno, M., Kataoka, N. and Shimura, Y. (1990). A nuclear cap binding protein from HeLa cells. Nucl. Acids Res. 18, 6989-6995.

Influence of 5' and 3' end structures on pre-mRNA metabolism 19

Patzelt, E., Blaas, D. and Kuechler, E. (1983). CAP binding proteins associated with the nucleus. Nucl. Acids Res. 17, 5821-5835.

Patzelt, E., Thalmann, E., Hartim ith, K., Blaas, D. and Kuechler, E. (1987). Assembly of pre-mRNA splicing complex is cap dependent. Nucl. Acids Res. 15, 1387-1399.

Raabe, T., Bollum, F. J. and Manley, J . L. (1991). Primary structure and expression of bovine poly(A) polymerase. Nature 353, 229-234.

Raju, V. S. and Jacob, S. T. (1988). Association of poly(A) polymerase with Ul RNA. J. Biol. Chem. 263, 11067-11070.

Rasmussen, E. and Lis, J. (1993). In vivo transcriptional pausing and cap formation on three Drosophila heat shock genes. Proc. Nat. Acad. Sei. USA 90, 7923-7927.

Rhoads, R. E. (1988). Cap recognition and the entry of mRNA into the protein synthesis initiation cycle. Trends Biochem. Sei. 13, 52-56.

Rosbash, M. and Séraphin (1991). Who’s on first? The U1 snRNP-5' splice site interaction and splicing. Trends Biochem. Sei. 16, 187-190.

Rozen, F. and Sonenberg, N. (1987). Identification of nuclear cap specific proteins in HeLa cells. Nucl. Acids Res. 15, 6489-6500.

Sachs, A. B. (1993). Messenger RNA degradation in eukaryotes. Cell 74 ,413421.

Saldit-Georgieff, M., Harpold, M., Chen-Kiang, S. and Darnell, J . (1980). The addition of the 5' cap structres occurs early in hnRNP synthesis and prematurely terminated molecules are capped. Cell 19, 69-78.

Scherly, D., Boelens, W., van Venrooij, W. J., Dathan, N. A., Hamm, J. and M attaj, I. W. (1989). Identification of the RNA binding segment of human U1A protein and definition of its binding site on U l snRNA. EMBO J. 8, 4163-4170.

Sheets, M. D. and Wickens, M. (1989). Two phases in the addition of a poly(A) tail. Genes Dev. 3, 1401-1412.

Sheets, M. D., Ogg, S. C. and Wickens, M. (1990). Point mutations in the AAUAAA and the poly(A) addition site: effects on the accuracy and efficiency of cleavage and polyadenylation in vitro. Nucl. Acids Res. 18, 5799-805.

Shibagaki, Y., Itoh, N., Yamada, H., Nagata, S. and Mizumoto, K. (1992). mRNA capping enzyme. J. Biol. Chem. 267,9521-9528.

Shimotohno, K., Kodama, Y., Hashimoto, J . and M iura, K. I. (1977). Importance of 5'-terminal blocking structure to stabilize mRNA in eukaryotic protein synthesis. Proc. Nat. Acad. Sei. USA 74, 2734-2738.

Shuman, S., Lui, Y. and Schwer, B. (1994). Covalent catalysis in nucleotidyl transfer reactions: essential motifs in Saccharomyces cerevisiae RNA capping enzyme are conserved in Schizosaccharomyces pombe and viral

capping enzymes and among polynucleotide ligases. Proc. Nat. Acad. Sei. USA 91, 12046-12050.

Sonenberg, N. (1988). Cap-binding proteins of eukaryotic messenger RNA: Functions in initiation and control of translation. Prog. Nucl. Acids Res. Mol. Biol. 35, 174-207.

Staknis, D. and Reed, R. (1994). SR proteins promote the first specific recognotion of pre-mRNA and are present together with the U1 small nuclear ribonucleoprotein particle in a general splicing enhancer complex. Mol. Cell. Biol. 14, 7670-7682.

Sun, Q., Mayeda, A., Hampson, R., Krainer, A. and Rottman, F. (1993). General splicing factor SF2/ASF promotes alternative splicing by binding to an exonic splicing enhancer. Genes Dev. 7,2598-2608.

Takagaki, Y., Ryner, L. and Manley, J . L. (1989). Four factors are required for 3'-end cleavage of pre-mRNAs. Genes Dev. 3, 1711-1724.

Takagaki, Y., Manley, J . L., MacDonald, C. C., Wilusz, J . and Shenk, T. (1990). A multisubunit factor, CstF, is required for polyadenylation of mammalian pre-mRNAs. Genes Dev. 4 , 2112-2120.

Tarn, W. and Steitz, J . (1994). SR proteins can compensate for the loss of U1 snRNP functions in vitro. Genes Dev. 8,2704-2717.

van Gelder, C. W. G., Gunderson, S. I., Jansen, E. J . R., Boelens, W. C., Polycarpou-Schwarz, M., M attaj, I. W. and van Venrooij, W. J. (1993). A complex secondary structure in U1A pre-mRNA that binds two molecules of U1A protein is required for regulation of polyadenylation. EMBO J. 12, 5191-5200.

Wahle, E. (1991). A novel poly(A)-binding protein acts as a specificity factor in the second phase of messenger RNA polyadenylation. Cell 66,759-768.

Wahle, E., M artin, G., Schiltz, E. and Keller, W. (1991). Isolation and expression of cDNA clones encoding mammalian poly(A) polymerase. EMBOJ. 10, 4251-4257.

W ahle, E. and Keller, W. (1992). The biochemistry of 3'-end cleavage and polyadenylation of messenger RNA precursors. Annu. Rev. Biochem. 61, 419-440.

Wahle, E. (1995). 3'-End cleavage and polyadenylation of mRNA precursors. Biochim. Biophys. Acta 1261, 183-194.

W assarm an, K. M. and Steitz, J . A. (1993). Association with terminal exons in pre-mRNAs: a new role for the U 1 snRNP? Genes Dev. 7, 647-659.

Wickens, M. (1990). In the beginning is the end: regulation of poly (A) addition and removal during early development. Trends Biochem. Sei. 15, 320324.

Zahler, A., Lane, W., Stolk, J . and Roth, M . (1992). SR proteins: a conserved family of pre-mRNA splicing factors. Genes Dev. 6, 837-847.