What is the consequence of alternative splicing of identical mRNA transcripts?

Processing of Primary RNA Transcripts

In prokaryotes, the primary mRNA transcript is functional as soon as it is synthesized. This is seen when ribosomes bind to the free 5′ end, even before the remainder of the molecule is transcribed. (Remember that synthesis is 5′ to 3′, so the 5′ end of mRNA is synthesized first.) In eukaryotes, however, the RNA transcript must undergo processing before it is a functional mRNA. This processing occurs in the nucleus and involves three steps: 5′ capping, 3′ polyadenylation (polyA tailing), and exon splicing.

Capping involves the addition of an inverted 7-methylguanosine triphosphate attached to the 5′ end of the primary transcript (Fig. 16-10). This produces a 5′ to 5′ phosphodiester bond, thereby providing a free 3′ hydroxyl at the 5′ end of the molecule. The polyA tail is attached at the 3′ end by polyA polymerase, using adenosine triphosphate as a precursor, and it extends to a length between 20 and 250 bases. Capping and polyadenylation serve a dual purpose. They make the mRNA stable byblocking access to the termini by exonucleases, and they participate in polypeptide chain initiation. Neither of these posttranscriptional elements has a counterpart in the DNA sequence.

Splicing of eukaryotic primary RNA transcripts removes the introns, leaving the exons connected together in a functional message. Primary transcripts are also called heterogeneous nuclear RNA (hnRNA), since they contain from zero to as many as 50 introns of variable length. The introns to be removed are bounded at both ends by specific base sequences called splice sites or splice junctions. Splice junctions usually begin with a consensus sequence GU (the donor site) and end with a consensus sequence AG (the acceptor site). The donor loops over to the acceptor, forming a lariat structure (Fig. 16-11) that is released when the RNA is cleaved at the acceptor site. The donor and acceptor sites are recognized by specialized small nuclear RNA particles (snRNPs; Fig. 16-12) that associate with nuclear proteins to form spliceosomes. The small nuclear RNA particles hold onto the 5′ exon and the 3′ exon so both ends can be rejoined.

Primary transcripts of ribosomal and transfer RNA also require processing by nucleases, and there are similarities between prokaryotes and eukaryotes. Ribosomal proteins begin to associate with the primary transcript as it undergoes processing. The tRNA precursor is also processed by nuclease cleavage of precursor RNA. An intron is removed from the anticodon loop, and trimming occurs at both ends of the molecule. Processing is completed by modification of selected bases and addition of the -CCA terminal at the 3′ end. The proportions of ribosomal and tRNA species are easily coordinated by processing the precursor transcripts.

Pathology

Key Points about RNA Transcription

Eukaryotic and prokaryotic transcription are similar, but the attachment of RNA polymerase to the promoter requires a more complex interaction of transcription factors for the former.

Eukaryotic hnRNA undergoes processing before it is a functional mRNA.

Alternative RNA Splicing and Editing

Most primary transcripts in eukaryotic cells derive from complete removal of all introns and complete joining of all exons. This results in only one species of mature mRNA being synthesized from each primary transcript. However, many eukaryotic genes can give rise to many different forms of mature mRNAs using different promoters and different polyadenylation sites (discussed previously), as well as by alternative splicing. This form of regulation of RNA processing is especially useful because a single gene can be expressed differently at various developmental stages or in different tissues of the same organism. One example is the troponin T gene that synthesizes the fast skeletal muscle protein. This gene consists of 18 exons; however, only 11 are found in all mature mRNAs. Five of the exons can be included or excluded, and two are mutually exclusive; if one is included, the other is excluded. Altogether, 64 different mature mRNAs can be produced by alternative splicing of the primary transcript of the troponin T gene.

RNA editing involves processing of RNA in the nucleus by enzymes that change a single nucleotide (insertion, deletion, or substitution). One example is the apolipoprotein B (apo B) gene. In the liver, this gene produces a 4536-amino acid protein (apo B-100), whereas in the small intestine, the same gene produces a 2152-amino acid protein (apo B-48). The truncated protein is identical in amino acid sequence to the first 2152 amino acids in apo B-100. This occurs because in cells of the small intestine one nucleotide (at position 6666) is edited by deamination of a cytosine residue by a sequence-specific cytidine deaminase. The conversion of a cytosine to uracil at this position produces a stop codon that terminates translation and produces the truncated protein. Thus, selective editing of mRNAs prior to translation in specific tissues is used to produce different proteins.

III PTHrP structure and posttranslational processing

The primary transcript of PTHrP represents a polyprotein analogous to the primary transcript of the proopiomelanocortin gene. Therefore, most cells expressing the PTHrP gene actually secrete several different PTHrP peptides. The primary transcript has a typical prepro sequence encompassing amino acid residues −36 to −1, which confers the ability for secretion from the cell. The mature peptide is very well conserved across species with residues 1–111 being 98% homologous in chickens compared with humans. This area has several stretches of basic amino acids that allow proteolyic cleavage of the primary transcript by prohormone convertases within the Golgi and secretory granules to generate a series of peptides encompassing the amino-terminus, the midregion, and the carboxy-terminus of PTHrP.

The amino-terminus of PTHrP is the portion that is homologous to PTH. Of the first 13 aa of the two proteins, 8 are identical and 3 represent conservative changes. Furthermore, there is considerable similarity in the predicted conformation of the two peptides through residue 34. There are several species containing this amino-terminal region that appear to be secreted by various cells. Most cells produce PTHrP 1–36, which is equipotent with PTH 1–34 at binding and activating the type I PTH/PTHrP receptor (PTH1R). In addition, an “intact” PTHrP molecule that includes at least the first 74 aa circulates in patients with HHM. Finally, keratinocytes have been reported to make a glycosylated amino-terminal version of PTHrP with a molecular weight of 18 kDa.

Various cells have also been shown to produce midregion peptides that begin at Ala-38 and stretch to amino acids 94, 95, or 101. These peptides have been shown to circulate and their secretion appears to be regulated. This portion of PTHrP appears to function to facilitate placental calcium transport from mother to fetus. However, the receptor for this portion of PTHrP has not been identified. The midregion of the molecule also contains nuclear localization sequences (NLSs) and there is a growing body of literature that suggests that nuclear-targeted PTHrP may have effects on cellular proliferation, differentiation, and apoptosis. At this point, it is unclear whether the midregion portions that are secreted versus targeted to the nucleus are the same or represent distinct peptides, as there are potential processing sites between Ala-38 and the NLS between amino acids 87 and 106. Interestingly, in this regard, recent data have suggested that initiation of PTHrP translation may occur downstream of the signal peptide to generate a nonsecreted form of PTHrP. Other evidence suggests that longer peptides containing both the amino-terminal and the NLS sequences may be imported to the nucleus after binding to cell surface PTH1R. Obviously, there is much information concerning this portion of the molecule that remains to be elucidated.

Finally, carboxy-terminal portions of PTHrP also appear to circulate and have been detected in the urine of normal patients and in the serum of patients with renal failure. Peptides encompassing residues 107–111 and 107–139 have been shown to inhibit bone resorption in vitro and this portion of the molecule has been dubbed “osteostatin.” However, not all groups have demonstrated this effect and the biological significance of carboxy-terminal PTHrP remains undefined.

Primary Transcripts Are Modified to Form Mature Messenger RNA

The primary transcripts of RNA synthesized by RNA polymerase II (mRNA) are modified in the nucleus by three distinct reactions: the addition of a 5′ cap, the addition of a polyadenylic acid (poly-A) tail, and the excision of the noninformational intron segments. These modifications are required to form a mature RNA capable of supporting the translation of a protein, and the entire set of events is called RNA processing.

The 5′ end of the mRNA (the end that is synthesized first during transcription) is capped by the addition of a methylated guanosine nucleotide. The addition of the 5′ cap is the first modification of the mRNA primary transcript and occurs almost immediately with the onset of transcription (Fig. 5-33).

The formation of the cap involves the condensation of the triphosphate moiety of a GTP molecule with the diphosphate group of the nucleotide at the 5′ end of the initial transcript. The enzymes responsible for the capping reaction(s) are thought to reside within the subunit structure of RNA polymerase II. The addition of the 5′ cap structure is critical for mRNA to be translated in the cytoplasm and appears to be needed to protect the growing RNA chain from degradation in the nucleus.

The second modification of an mRNA transcript occurs at its most 3′ end, the addition of a poly-A tail. The 3′ end of most polymerase II transcripts is not defined by the termination of transcription, but by the specific cleavage of the RNA molecule and the addition of adenosine residues to the cleaved molecule by a separate polymerase, poly-A polymerase. The signal for cleavage is the appearance of the sequence AAUAAA in the growing RNA chain, with the actual cleavage occurring about 10 to 30 nucleotides away from this signal sequence. Immediately on cleavage, poly-A polymerase adds 100 to 500 residues of adenylic acid to the 3′ end of the cleaved RNA molecule. The RNA polymerase II appears to continue transcription well beyond the cleavage site, with the subsequent RNA being rapidly degraded, presumably because they lack 5′ cap structure. The exact functions of the poly-A tail are not welldefined; however, experimental evidence suggests that it plays an important role in the export of mature mRNA from the nucleus to the cytoplasm. In addition, it may serve a regulatory function, in that some genes contain multiple sites for poly-A addition.

After modifications of the 5′ and 3′ ends of the primary transcript, the noninformational intron segments are removed, and the coding exon sequences are joined together by RNA splicing. The specificity of exon joining is conferred by the presence of signal sequences marking the beginning (called the 5′ donor site) and the end (called the 3′ acceptor site) of the intron segment (Fig. 5-34).

These signal sequences are highly conserved (they are approximately the same in all known intron segments), and as might be predicted, alterations in these sequences lead to aberrant mRNA molecules. For example, a group of genetic diseases, collectively called the βthalassemia syndromes (characterized by the abnormally low expression of hemoglobin), are directly attributable to single-base changes in the genome at splice junctions of the β-globin gene that disrupt the appropriate joining of exon segments. Therefore, splicing reactions must occur with exquisite precision to ensure that a functional RNA molecule is formed.

The excision of an intron segment from RNA is conducted by a ribonucleoprotein complex called the spliceosome. The spliceosome is formed from a set of undefined proteins complexed with a series of small RNA molecules referred to as U1 through U12. The splicing reaction occurs in steps that include: (1) recognition of consensus 5′ donor and 3′ acceptor sequences, (2) cleavage of the 5′ splice site and formation of a looped RNA structure termed the lariat, and (3) the cleavage of the 3′ place site and subsequent ligation of the RNA molecule. The exact role of individual components of a spliceosome is currently being studied; however, it is known that the excision of introns requires the energy of ATP hydrolysis. The excised intron is degraded almost immediately after its release from the primary RNA transcript.

Although it would be logical that the splicing of RNA proceeds by the removal of the most 5′ intron to the last or 3′ intron, experimental evidence has demonstrated that the removal of introns from any given transcript follows a preferred path, often beginning with introns internal to the transcript. This appears to be an inefficient mechanism, and initially, it was viewed with skepticism; however, the recent discovery that a single gene may express multiple different proteins by the selected joining of exon sequences to form different mRNA has shed some new light on the pathways of intron removal. For example, a single gene encoding the protein troponin T can produce at least 10 distinct forms of the molecule by simply joining different combinations of encoded exon segments. This variability can be influenced by cell type or by factors extrinsic to a cell, enabling the expression of protein isoforms needed to compensate for alterations in cellular metabolism. The ability of certain genes to form multiple proteins by joining different exon segments in the primary trans-cript is called alternate splicing and has caused a reexamination of the concept of “one gene, one protein.”

4.2.3.1.2.2 Splicing

The primary transcript will contain introns as well as exons. In becoming a mature mRNA, the cell will remove the introns through a process known as splicing. Figure 4.20 shows the process with two levels of detail. On the left side of the figure, it can be seen that a specific adenylate residue is brought to the 5′ border of the intron. Hydrolysis at the 5′ border is immediately followed by creation of a new phosphodiester bond between the exposed 5′ intron phosphate and the 2′ carbon on the ribose of the adenylate residue. This creates a loop structure in the intron. The newly created 3′ hydroxyl at the end of the previous exon is then available to form a phosphodiester bond with the 5′ phosphate of the next exon after another hydrolytic reaction frees the intron. The intron is released from the pre-mRNA as a loop structure with a tail, as structure known as a lariat.

The act of splicing out introns as lariats is a very complex process, mediated by over 50 proteins and 5 additional RNA molecules. The five RNA molecules—U1, U2, U4, U5, and U6—are known as small nuclear RNAs (snRNAs). Each snRNA is < 200 nucleotides long and participates in recognition of intron/exon boundaries as well as remodeling of phosphodiester bonds. They do not work alone, though, and associate with several proteins to create complexes known as small nuclear ribonucleoproteins (snRNPs). The snRNPs come together to form the structure known as a spliceosome. The right-hand side of Figure 4.20 shows some basic interactions between the snRNPs. First, the branchpoint adenine nucleotide is recognized by the branchpoint binding protein and a helper peptide (not shown), which recruit snRNP U2 to the site. At the same time, snRNP U1 identifies and associates with the upstream border between the intron and the adjacent exon. The other three snRNPs (U4/U6-U5) then become involved as a unit to help bring the 5′ splice site to the branchpoint and facilitate hydrolysis with the aid of additional proteins and ATP. During the process, there are rearrangements and dissociations of snRNPs in the spliceosome, resulting in only the U2, U5, and U6 snRNPs remaining with the lariat as it is released.

Alternative Splicing

Processing of the primary transcript of a gene to give mature mRNA requires accurate splicing. It also provides possibilities for producing more than one type of mRNA (and potentially therefore more than one protein) from a single gene. Such alternative processing pathways occur for a number of polypeptide hormone genes. For example, in the case of human GH it is known that some of the heterogeneity found in the hormone extracted from pituitary glands is due to alternative splicing of the parent gene. The normal (22K) form of human GH comprises a single polypeptide chain of 190 amino acid residues. A naturally occurring variant (20K) which makes up 5–15% of the GH in the human pituitary is identical to 22K human GH except that it lacks 15 residues (residues 32-46), with retention of a single polypeptide chain (Lewis et al., 1980). The explanation of the 20K variant appears to be that it arises from an alternative splicing event in which the nucleotide sequence coding for residues 32-46 is treated as part of exon 3 in normal (22K) human GH, but as part of intron 2 in the 20K variant (Figure 6) (Wallis, 1980; DeNoto et al., 1981). This has been confirmed from the sequences of the mRNAs for the 22K and 20K human GH that have now been determined. Whether the 20K variant of human GH has a specific function, distinct from that of the 22K form, is not yet clear. Variants equivalent to 20K human GH have not been found in non-primates.

Another example of alternative splicing of a hormone gene occurs in the case of the calcitonin gene. In the C cells of the thyroid gland, the product of this gene is spliced to give mRNA coding for calcitonin, but in the hypothalamus the same initial transcript is spliced differently, to give mRNA for CGRP (calcitonin gene-related peptide) which has biological actions quite distinct from those of calcitonin (Amara et al., 1982). The mechanisms underlying the different splicing pathways are being studied intensively; for example, in C cells exon 4 in the mRNA precursor is recognized, and completes the calcitonin mRNA, but in hypothalamus this exon is “skipped” and two further exons downstream of it are used to complete the CGRP mRNA (Lou et al., 1994). Thus two quite distinct regulatory peptides are produced, in different tissues, from the same gene.

6.06.2.3.1 Small RNAs

MicroRNA genes encode primary transcripts (pri-miRNAs), which are processed into short stem-loop structures called pre-miRNAs and then modified into mature miRNAs. These 19–25 nucleotide (nt) single-stranded mature miRNAs bind to the 3′-untranslated regions of target mRNAs to destabilize or inhibit their translation. miRBase (www.mirbase.org) catalogs all ~1500 identified human miRNAs and provides each with a name in the format ‘mir-#’ for the pre-miRNA and ‘miR-#’ for the mature miRNA. The HGNC provide a parallel gene nomenclature, so that the MIR100 gene encodes the mir-100 stem-loop, which is modified to create the miR-100 mature transcript. miRNAs have been found to affect many cellular functions and have been implicated in diseases; for instance, mutations in MIR96 have been linked to progressive hearing loss (Soldà et al., 2012).

tRNAs are ~80  nt molecules with an essential role in protein translation, transporting specific amino acids to the ribosomes to elongate the peptide chain. As they display a high degree of structural conservation, they can be accurately predicted within a genome. The Genomic tRNA database (gtrnadb.ucsc.edu/) predicts over 500 tRNA genes and 110 tRNA pseudogenes in the human genome. The HGNC has approved symbols for all of these tRNA loci in the format ‘TRNA + single letter code for amino acid isotype + number,’ with the anticodon type also specified in the name, for example, transfer RNA alanine 1 (anticodon UGC) (TRNAA1). There are also 22 tRNAs encoded in the human mitochondrial genome, as mentioned previously. Each of these has been provided with an approved name in the format ‘MT-T + single letter code for amino acid isotype + number,’ again with the specific anticodon type included in the name, for example, mitochondrially encoded tRNA serine 1 (UCN) (MT-TS1).

The spliceosome is a large ribonucleoprotein made up of over 200 proteins and five small nuclear RNAs (snRNAs: U1, U2, U4, U5, and U6), which are highly conserved across genomes. The minor spliceosome also contains four different snRNAs known as U11, U12, U4atac, and U6atac, where the ‘atac’ suffix references the unusual AT/AC splice sites processed by the minor spliceosome. snRNA nomenclature follows the format ‘RN + snRNA species + number’ (e.g., RNU1-1 for ‘RNA, U1 small nuclear 1’). In total, so far, HGNC have named over 100 spliceosomal snRNA loci.

Small nucleolar RNAs (snoRNAs) fall into two main types: H/ACA box snoRNAs direct pseudouridylation, and C/D box snoRNAs direct methylation. In collaboration with snoRNABase (www-snorna.biotoul.fr) and experts in the field, HGNC have named snoRNA genes using the root symbol SNORA# for the H/ACA box genes and SNORD# for the C/D boxes. Another class of snoRNAs is the small Cajal body-specific RNAs (scaRNAs), named after the suborganelles within the nucleus where they are located, which are approved as SCARNA#. Over the past 5 years, HGNC have named all 414 snoRNA genes currently characterized in the snoRNABase.

There are four types of rRNA found in eukaryotic ribosomes: 18S rRNA in the small subunit and 28S, 5.8S, and 5S rRNAs in the large subunit. The 18S, 5.8S, and 28S rRNA genes are arranged in tandem repeats that produce one precursor transcript, which is then cleaved to give the three types of rRNA. These repeats are found in five clusters in the human genome, and HGNC have assigned a name to each cluster in the form RNA18S1–5, RNA5-8S1–5, and RNA28S1–5. As large amounts of RNA are required to make ribosomes, there are hundreds of copies of this rRNA repeat in each cluster, and the number of copies varies considerably between individuals (Stults et al., 2008). Such repetitive regions are problematic when sequencing genomes, and hence, they are not represented in the current human genome build (GRCh37). 5S rRNA genes are also found in tandem repeats scattered throughout the genome, with several major clusters at human chromosome 1q42. Again, the number of 5S rRNA genes in each individual's genome is likely to be highly variable, but to date, HGNC have named the only 5S rRNA loci annotated in the current genome assembly (GRCh37), comprising 17 loci at 1q42, in the format RNA5S1–17.

Splicing of RNA in Eukaryotes

A distinguishing feature of most primary transcripts of higher eukaryotes is the presence of untranslated intervening sequences ( introns ) that interrupt the coding sequence and are excised from the primary RNA transcript. In the processing of RNA in higher eukaryotes, the amount of discarded RNA ranges from 30% to nearly 90% of the primary transcript. The remaining coding segments ( exons ) are joined together by splicing enzymes to form translatable mRNA molecules. The excision of the introns and the formation of the final mRNA molecule by joining of the exons is called RNA splicing. The 5' segment (the cap) of the primary transcript is never discarded and hence is always present in the completely processed mRNA molecule; the 3' segment is also usually retained. Thus, the number of exons is usually one more than the number of introns. The number of introns per gene varies considerably (Table 25-1). Furthermore, within different genes the introns are distributed differently and have many sizes (Figure 25-9), and introns are usually longer than exons.

TABLE 25-1. Translated Eukaryotic Genes in Which Introns Have Been Demonstrated

*At present the histone and interferon genes are the only known translated genesin the higher organisms that do not containin trons

The splicing reaction is remarkably precise: cuts are made at unique positions in transcripts that contain thousands of bases. The fidelity of the excision and splicing reaction is extraordinary, for if an error of even one base were made, the correct reading frame would be destroyed. Such fidelity is achieved by recognition of particular base sequences by splicing enzymes.

Base sequence studies of the regions adjacent to several hundred different introns indicate that common sequences can be found at each end of an intron. The sites at which cutting occurs are always 5' to GU and 3' to AG. The rule is that the base sequence of an intron begins with GU and ends with AG.

Introns are excised one by one, and ligation occurs before the next intron is excised; thus, the number of different nuclear RNA molecules present at any instant is huge. Translation does not occur until processing is complete.

Before RNA splicing was discovered, the nucleus was observed to contain a significant amount of seemingly untranslated RNA. The collection of RNA molecules of widely varied sizes was given the name heterogeneous nuclear RNA (hnRNA), a term that is still sometimes used for nuclear RNA.

4.4.6 3′-Untranslated Sequences and Transcriptional Termination

The 3′ ends of primary transcripts are determined by transcriptional termination signals located downstream of the ends of each coding region. However, the 3′ ends of mature mRNA molecules are created by cleavage of each primary precursor RNA and the addition of a several hundred nucleotide polyadenylate [poly(A)] tails (see Fig. 4.6). The cleavage site is marked by the sequence 5′-AAUAAA-3′ located 15–20 nucleotides upstream of the poly(A) site and by additional GU-rich sequences 10–30 nucleotides downstream. Histone mRNAs, which do not have poly(A) tails, have stem-loop structures instead with cleavage of the primary transcript mediated by a distinct protein complex that includes the U7 snRNP [35].

Some complex transcriptional units contain several potential polyadenylation and/or transcription termination sites. It is often difficult to distinguish the latter from the former as the product available for analysis (mRNA) has lost the portion of the 3′-terminus originally transcribed by RNA polymerase. Alternative polyadenylation (or termination) sites can determine final protein structure if the longer precursor RNA contains an exon not found in the shorter precursor RNA. In a simple case, two proteins with different carboxyl termini are formed. But if alternative exon splice sites are made available in the longer precursor RNA, proteins with entirely different sequences can be produced.

The region from the translation termination codon to the poly(A) addition site may contain up to several hundred nucleotides of a 3′-UTR, which includes signals that affect mRNA processing and stability. Many mRNAs that are known to have a very short half-life contain AU-rich elements, 50- to 150-nucleotide sequences containing AUUUA motifs that regulate mRNA stability [36]. Other, less well-characterized sequences can have similar effects. Removal or alteration of these sequences can prolong the half-life of mRNA, indicating that such elements represent a general regulatory feature of mRNAs whose level of expression can be rapidly altered.