SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

Alternative precursor messenger RNA (pre-mRNA) splicing plays an important role in the generation of functional diversity of the genome. The process of pre-mRNA splicing is regulated by cis- and trans-elements, and their deregulations result in aberrantly spliced individual variants and aberrant expression profiles. Accumulating evidence has revealed that aberrant splicing contributes to a number of diseases including human neoplasms. It is well known that germ line mutations in the cis-element of tumor suppressor genes such as mismatch repair (MMR) genes, the adenomatous polyposis coli (APC) gene and the E-cadherin (CDH1) gene are involved in Lynch syndrome, familial adenomatous polyposis and hereditary diffuse gastric cancer, respectively. In addition, somatic mutations in cis-elements also play a role in tumorigenesis. These genetic alterations including nonsense, missense or silent mutations in cis-elements led to aberrant transcripts by exon skipping, retention of the intron or introduction of a new splice site. The majority of erroneous transcripts with a premature termination codon are eliminated through nonsense-mediated mRNA decay. However, it is difficult to accurately predict the resulting transcripts with current in silico strategies. Correct interpretation of genetic alterations and the investigation of aberrant transcripts are crucial for genetic diagnosis of hereditary diseases and elucidation of the molecular characteristics of neoplasms from a clinical point of view. In this review we summarize the current knowledge of the regulatory mechanism underlying alternative pre-mRNA splicing and aberrant splicing, with particular focus on digestive tract malignancies. (Cancer Sci 2011; 102: 309–316)

Alternative precursor messenger RNA (pre-mRNA) splicing is the process by which the exons of pre-mRNA are spliced in different arrangements to produce structurally and functionally distinct mRNA and proteins (Fig. 1).(1) The completion of the Human Genome Project in 2004(2) uncovered that the human genome contains approximately 23 000 protein-coding genes, a much smaller number of genes than had previously been estimated. This finding suggests that alternative pre-mRNA splicing is one of the mechanisms that maintains functional diversity of the human genome. Notably, over 95% of the genes in mammalian genomes are predicted to have multiple isoforms.(3–5) Accumulating evidence has demonstrated the importance of aberrant alternative splicing in malignancy. Intensive studies on splice variants have facilitated the profound understanding of the regulatory mechanism of pre-mRNA splicing and alternative splicing. Recently, an enormous amount of information has been accumulated in public bioscientific databases, and challenges have been made to predict cancer-specific splice variants using the databases. The Alternative Splicing Annotation Project, published in 2003, described the alignment of expressed sequence tags and cDNA sequences and Lee et al.(4) predicted splice variants that were alternatively expressed in cancer tissues; their data on several of the splice variants were validated by RT-PCR in gastric cancer.(6)

image

Figure 1.  Alternative precursor messenger RNA (pre-mRNA) splicing. The green boxes indicate constitutive exons and the blue boxes indicate alternatively spliced exons.

Download figure to PowerPoint

Based on these backgrounds, we herein summarize the current knowledge of the regulatory mechanism of alternative pre-mRNA splicing and aberrant alternative splicing, with particular focus on digestive tract malignancy. In addition, we discuss the utilization of this information, so that future studies can continue to contribute to this field of research.

Regulatory mechanism of alternative pre-mRNA splicing

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

Pre-mRNA splicing is a mechanism that removes introns from pre-mRNA and connects exons, eventually generating mature mRNA (Fig. 2), and the process is regulated by cis-elements and trans-elements. The cis-elements include consensus splice sites comprised of a 5′ splice site (5′SS), a branch point motif, a poly-pyrimidine tract ([Y]n), and a 3′ splice site (3′SS) (Fig. 2). Known as a splice donor site, the 5′SS includes the normally invariant GU nucleotide sequence at the 5′ end of the intron. Known as a splice acceptor site, the 3′SS terminates the intron with the normally invariant AG nucleotide sequence that is preceded by a poly-pyrimidine tract (Y)n, where Y denotes a pyrimidine (U or C), and a branch point motif containing a nucleotide A residue. These consensus sequences are termed the U2 type consensus splice site because they are recognized by spliceosomes containing a U2 non-coding uridine-rich small nuclear ribonucleoprotein (snRNP). Burge et al.(7) reported the frequency of nucleotides that appeared in the U2-type consensus splice site from the information of 1683 human introns. The cis-regulatory elements include splice enhancers and silencers, both of which play an important role in the recognition of the 5′SS and 3′SS regions. Splice enhancers are essential for exon inclusion and are required for constitutive splicing, and splice silencers are important for exon skipping. Depending on their localization within the genome, splice enhancers and silencers are subclassified into exonic splice enhancers (ESE), intronic splice enhancers (ISE), exonic splice silencers (ESS) or intronic splice silencers (ISS). The ESE are present in the majority of exons, and several programs such as ESEfinder, RescueESE and PESE can predict the ESE consensus sequences.

image

Figure 2.  Regulatory mechanism of alternative precursor messenger RNA (pre-mRNA) splicing and its alteration in malignancies. Cis-elements are indicated with rectangles and trans-elements are indicated with ellipses. In the nucleotide sequences, Y denotes a pyrimidine (U or C) and R denotes a purine (G or A). ESE, exonic splice enhancer; ESS, exonic splice silencer; hnRNP, heterogeneous nuclear ribonucleoprotein; ISE, intronic splice enhancer; ISS, intronic splice silencer; snRNP, small nuclear ribonucleoprotein; SRp, serine/arginine-rich protein; SS, splice site; U2AF, U2 small nuclear ribonucleoprotein auxiliary factor.

Download figure to PowerPoint

The trans-elements includes spliceosomes, multicomponent complexes that are essential for pre-mRNA splicing.(8,9) Spliceosomes consist of five non-coding uridine-rich snRNP including U1, U2, U4, U5 and/or U6, and over 200 different splicing factors that interact with the snRNP. snRNP U1 associates with the 5′SS, and a heterodimer of U2 small nuclear ribonucleoprotein auxiliary factor (U2AF) with the poly-pyrimidine tract and the 3′SS. Splicing factor 1 (SF1) interacts with the branch point motif and is replaced by snRNP U2. Members of serine/arginine-rich proteins (SR proteins, symbolized by SRp) and SR-like proteins such as alternative splicing factor/splicing factor 2 (ASF/SF2), SC35, SRp20, SRp40, SRp55 and SRp75 specifically bind to ESE.(9) On the other hand, heterogeneous nuclear ribonucleoproteins (hnRNP)(10) such as hnRNP A1, hnRNP A2/B1, hnRNP C1, hnRNP F and hnRNP G bind to both ESS and ISS. In many cases, hnRNPs block an assembly of spliceosomes, thus resulting in exon skipping.

Proteins produced through alternative pre-mRNA splicing from the same gene show different biological properties.(5,11) In genomic regions with multiple alternative splice sites, SR proteins and hnRNP influence the selection of the splice sites in a concentration-dependent manner,(12) and the concentrations vary across cell and tissue types. It has been proposed that the differential expression of trans-elements between tissue types are the major contributing factor to tissue-specific alternative pre-mRNA splicing,(13) and several tissue-specific factors have recently been identified. For example, epithelial splicing regulatory protein 1 (ESRP1) and ESRP2 are trans-elements specifically expressed in epithelial cells, including the stomach and colorectum.(14) Alterations in the concentration, localization and activity of SR proteins and hnRNP can induce aberrant pre-mRNA splicing.(15) Therefore, comprehension of altered expression of trans-elements should be useful for the understanding of molecular mechanisms and characteristics of human diseases.

Alternative pre-mRNA splicing altered in malignancy

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

Human neoplasms involve not only aberrant splice variants but also altered expression of splice variants (Fig. 2). The primary causes of these two changes include alterations in cis-elements and trans-elements. To resolve the mechanism of these changes, investigations have been performed using different methodologies (Table 1).

Table 1.   Analytical strategy of alternative precursor messenger RNA (pre-mRNA) splicing altered in malignancy
AlterationAnalytical targetRepresentative methods
  1. hnRNP, heterogenous nuclear ribonuculeoproteins; IHC, immunohistochemistry; SR proteins, serine/arginine-rich proteins.

Cis-elementsConsensus splice sites, splice enhancers and silencersGenomic DNA sequencing, conversion analysis
Trans-elementsSpliceosomes, SR proteins, hnRNP, RNA componentsWestern blotting, IHC, proteomic analysis
Aberrant splicing
 Aberrant splice variantsIndividual transcripts (mRNA)RT-PCR, cDNA sequencing, transcription assay with minigene constructs
 Aberrant expression profileAn entire set of transcripts (mRNA)Microarray, Northern blotting, RT-PCR

In the early 1990s, the importance of analyzing genetic alterations in non-coding regions had already been emphasized. Recently, increasing evidence suggests that genetic alterations in exon boundaries or consensus splice sites cause aberrant splicing. Transcripts yielded by aberrant splicing often harbor a premature termination codon (PTC), and are detected through nonsense-mediated mRNA decay (NMD), a mRNA surveillance system that eliminates the majority of the truncated or erroneous proteins.(16) Activation or haplo-insufficiency of tumor suppressor genes and/or acquired oncogenic activity of several aberrantly-spliced pre-mRNA. which may be overlooked by the NMD, might contribute to carcinogenesis.

Regarding the trans-elements, overexpression of several SR proteins and hyperphosphorylation of SR proteins has been observed in malignant tissues.(15) The hyperactivation of SR protein kinases (SRpK) leads to the hyperphosphorylation of SR proteins. Therefore, the SRpK inhibitor is considered to be a possible approach for the development of new drugs. The spliceosome is also regarded to be a target for anticancer drugs,(17) which will be described in other parts of the present review.

Considering the current research, we propose that aberrant splicing in malignancies should be subclassified into two categories for further discussion: (i) the generation of an aberrant splice variant as an individual transcript, which is not normally expressed but is expressed in malignant cells; and (ii) an aberrant expression profile of splice variants as an entire set of transcripts in malignant cells (Fig. 2). The former category is primarily induced by alterations in cis-elements but can also be induced by the alteration of trans-elements. The latter category, the aberrant expression profile that is defined by the comparison of transcripts between malignant cells and the corresponding normal cells, is primarily induced through alterations in trans-elements as well as those in cis-elements (Fig. 2). Aberrant splice variants are also components of the expression profile of variants. To comprehensively understand the mechanism of human carcinogenesis, studies on the two categories of splice variants are crucial. This subclassification will be helpful to investigate the complex aberrant splicing mechanisms.

Detection of aberrant transcripts in malignant tissues requires causative aberrations of cis- and trans-elements in tumors. The identification of genetic changes in cis-elements is essential for confirmation of the resulting splice alteration of the transcripts as well, and the Human Genome Variation Society (HGVS) supports the systematical description of sequencing information by providing guidelines for a standardized nomenclature.(18) The sequencing analyses, both at the whole genome and transcripts in normal and tumor cells, will be a powerful tool to understand alternative pre-mRNA splicing in malignancy.

Alteration of cis-elements

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

The alteration of cis-elements in digestive tract malignancies has already been reported for various genes. The information of genetic alterations in digestive tract malignancies is already available in public databases, and selected resources are listed in Table 2. For example, genes in which genetic alterations were reported in gastric cancer include CDH1 (E-cadherin), FGFR2, MUTYH (MYH), TP53 (p53), FHIT, AKT2, PTEN (MMAC1), CASP6 and CHEK2 (CHK2). Among these genes, alteration of cis-elements and the resultant aberrant splice variants are summarized in Table 3.

Table 2.   Public databases
TitleWeb address
  1. IARC, International Agency for Research on Cancer; MMR, mismatch repair.

Catalogue of somatic mutations in cancerhttp://www.sanger.ac.uk/genetics/CGP/cosmic/
The IARC TP53 mutation databasehttp://www-p53.iarc.fr/index.html
The Roche cancer genome databasehttp://rcgdb.bioinf.uni-sb.de/MutomeWeb/header.html
The p53 web sitehttp://p53.free.fr/index.html
Mismatch repair variant databasehttp://www.med.mun.ca/MMRvariants/
MMR gene unclassified variants databasehttp://www.mmruv.info
Locus specific mutation databaseshttp://www.hgvs.org/dblist/glsdb.html
Mutation database (The University of Tokyo)https://reseq.lifesciencedb.jp/resequence/registration_j.html
Finnish disease heritagehttp://www.findis.org/
The Singapore human mutation and polymorphism databasehttp://shmpd.bii.a-star.edu.sg/
Table 3.   Alteration of cis-elements in digestive tract malignancy
Cis-element (Gene)Cancer typeAlteration typeMolecular aspectsAnalytical methodsReferences
  1. DPD, dihydropyrimidine dehydrogenase; FAP, familial adenomatous polyposis; HDGC, hereditary diffuse gastric cancer; HNPCC, hereditary non-polyposis colorectal cancer; IHC, immunohistochemistry; NMD, nonsense-mediated mRNA decay; SNP, single nucleotide polymorphism; 3′SS, 3′ splice site.

CDH1HDGCGermline mutationc.1135 del8ins5 (IVS8+5 del8ins5) of CDH1 caused three aberrant transcriptsGenomic and cDNA sequencing, IHC20
FHITGastric cancer (sporadic)Somatic mutationAberrant transcripts of FHIT were diverseLOH analysis, cDNA sequencing, western blotting21
MYHGastric cancer (familial), KATO-III and other cell linesGermline mutationc.892–2A>G (IVS10–2A>G) at 3′SS of MYH caused a truncated proteinGenomic and cDNA sequencing, transfection assay22
KLK12Gastric cancer (sporadic)Genetic polymorphismKLK12 protein was absent in individuals with c.457+2C/C in intron 4, but not in those with the T/T or T/CGenomic and cDNA sequencing, western blotting23
MMR genesHNPCC, COS-7 cell lineGermline mutationThe nonsense mutation within exon 12 of MLH1 caused exon skipping in three unrelated familiesIn vitro translation analysis, in vitro transcription assay24
HNPCC, COS-1 cell lineGermline mutationDisruption of an ESE at the 5′ end of exon 3 of MLH1 caused exon skippingGenomic and cDNA sequencing, in vitro transcription assay25
HNPCCGermline mutationSome of the MLH1 or MSH2 single-base substitutions led to exon skipping, but others did notGenomic and cDNA sequencing, RT-PCR26
HNPCC, COS-7 and other cell linesGermline mutationComputer predictions do not always correlate with in vivo splicing defects of MLH1 or MSH2In silico splicing analysis, in vitro transcription assay27
APCFAPGermline mutationDifferent single-base substitutions at or close to splice sites of APC were systematically evaluatedGenomic and cDNA sequencing, RT-PCR28
FAP, Caco-2 cell lineGermline mutationIn vitro experiments supported the importance of NMD in alternative splicing of APCRT-PCR, cDNA sequencing, in vitro NMD assay with cycloheximide29
DPDColon cancer (sporadic)Genetic polymorphismThe splice site polymorphism IVS14+1G->A of DPD caused a reduction in DPD activityGenomic sequencing, DPD activity assay30

Germline mutations in the CDH1 gene were identified in patients with hereditary diffuse gastric cancer (HDGC)(19) that is dominated by diffuse-type gastric adenocarcinoma exhibiting signet ring cell morphology (MIM #137215). Oliveira et al.(20) systematically analyzed genetic alterations and transcripts of the CDH1 gene in HDGC tumors. As a result, they found a mutation of c.1135 del8ins5 (IVS8+5 del8ins5) at the 5′SS of intron 8, and at least three types of aberrant transcripts. However, the reason why three different variants were simultaneously produced from a single-type mutation remains unknown. In addition to CDH1, germline mutations in the FGFR2 gene have been identified in several patients with hereditary gastric cancer and these mutations also included splice site alterations. Lee et al.(21) characterized 48 aberrant-sized FHIT transcripts in 35 cases of gastric cancer tissues, and found that the aberrant transcripts were derived from exon skipping, partial retention of intron, cryptic splice acceptor site activation and other alterations at the mRNA level. These reports suggested that different types of aberrant transcripts might be generated during the multi-step splice processing.(21) Other mutations identified in familial gastric cancer include MYH c.892–2A>G (IVS10–2A>G) at a 3′SS in two of 20 Japanese familial gastric cancer patients, which was not identified in 128 sporadic digestive tract cancer patients.(22) The mutation was confirmed to produce aberrant MYH transcripts by RT-PCR analysis and revealed to induce abnormal subcellular localization of MYH protein by immunohistochemistry (IHC).(22) Another study reported that cells with homozygosity for the single nucleotide polymorphism (SNP) of KLK12 c.457+2C/C in the 5′SS region of intron 4 showed loss of KLK12 protein expression, but cells with T/T or T/C at the site expressed the protein,(23) suggesting that SNP in cis-elements might lead to alteration of transcripts and/or their expression.

Regarding familial colorectal tumors, germline mutations in mismatch repair (MMR) genes including MLH1, MSH2 and MSH6 (hereditary non-polyposis colorectal cancer [HNPCC] or Lynch syndrome), APC (familial adenomatous polyposis [FAP] of the colon), SMAD4 (juvenile polyposis syndrome) and STK11/LKB1 (Peutz–Jeghers syndrome) are involved in their tumorigenesis. Somatic mutations in TP53, KRAS, BRAF, PIK3CA, AKT2, FHIT, IGFBP3, CHEK2, PTPRT, alkaline sphingomyelinase and other genes have been reported and registered (Table 2).

HNPCC (MIM #114500) is a hereditary cancer syndrome characterized by an increased risk of the early onset of colorectal cancer and related cancer in the endometrium, renal pelvis, small intestine, stomach and so on. Germline mutations in six MMR genes including MLH1, MSH2, MSH6, PMS2, MLH3 and EXO1 are involved in the syndrome. Among the mutations, alterations in the cis-elements and the resulting aberrant splice variants are summarized in Table 3. Stella et al.(24) analyzed 58 HNPCC families and detected an AAG to TAG nonsense mutation at codon 461 within exon 12 of the MLH1 gene, which caused skipping of exon 12 in three unrelated HNPCC families, and this alteration was functionally confirmed in vitro.(24) McVety et al.(25) reported that disruption of an ESE site also induced exon skipping in HNPCC, and the authors confirmed the exon skipping using in vitro minigene assay in COS-1 cells.(25) Pagenstecher et al.(26) reported frequent splice alterations in unclassified variants of MSH2 and MLH1. These data indicated that analyses of transcripts should precede functional tests for the characterization of uncharacterized variants. Although several in silico algorithms for the prediction of splice variants have been developed, the prediction does not always correlate with results obtained through an in vitro and/or in vivo assay.(27) Further progress in the development of in silico algorithms is needed to apply the prediction in clinics.

Germline mutation in the APC gene (MIM #175100) is responsible for FAP, an inherited autosomal dominant disease characterized by several hundred adenomatous polyps in the colon and rectum. In 2004, Aretz et al.(28) reported the first systematic evaluation of several single-base substitutions in the APC gene at the splice sites or close to splice sites at the transcript level. In this study, one exonic mutation in exon 4 (c.423G>T) and three in exon 14 (c.1956C>T, c.1957A>G and c.1957A>C) led to complete exon skipping due to aberrant splicing, although they had been predicted to result in missense or silent mutations. One possible explanation for this effect may be the disruption of ESE motifs.(28) De Rosa et al.(29) disclosed the importance of NMD degradation in alternative splicing of the APC gene using cycloheximide, a chemical inhibitor of translation that is also known to inhibit NMD. Furthermore, it was reported that a SNP of the dihydropyrimidine dehydrogenase (DPD) gene IVS14 + 1G>A at the 5′SS was associated with reduced DPD activity(30) and severe toxicity in colorectal cancer patients treated with 5-fluorouracil.

As summarized above, alterations in cis-elements associate with hereditary cancer syndromes through disruption of splicing. However, the analyses describing the alterations in cis-elements and their resulting transcripts in sporadic cancer cases are still limited. In addition, the current prediction algorithms for altered splicing need to be improved. Therefore, further studies are necessary for the correct interpretation of alterations in cis-elements.

Alteration of trans-elements

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

Alterations of trans-elements in gastrointestinal cancer are summarized in Table 4. The current data of deregulated trans-elements in tumorgenesis are still limited compared with the alterations in cis-elements. Ghigna et al.(31) analyzed the expression of SR proteins and hnRNP in colon cancer tissues using a Northern blot analysis. They consequently demonstrated that expression of ASF, SRp40, SRp55 and other elements was more severely reduced in tumors showing a more altered CD44 splicing pattern.(31) Regarding CD44 splice variants in human cancers, there are a number of discrepancies in the published data and thus it appears to be difficult to reconcile all of those results.(32) Reportedly, hnRNP K, an RNA-binding protein that plays a role in RNA editing, alternative splicing and many other processes is upregulated in cancer tissues and involved in tumorigenesis through the modulation of gene expression in response to mitogenic stimuli.(33) It is of note that Klimek-Tomczak et al.(33)identified a mutation of hnRNP K c.274G>A in tumors and the surrounding mucosa, but mutation of hnRNP K was not found in individuals that were tumor free. This observation might suggest that the substitution is involved in the development of colorectal cancer through the deregulation of RNA editing. Matos et al.(34) found that RAC1b, an alternative splice variant lacking exon 3b, of the RAC1 gene was overexpressed in a subset of colorectal tumors and that the expression was required to sustain tumor cell viability. Using an in vitro splicing assay with a RAC1 minigene construct in HT29 cells, they additionally showed that SRp20 increased RAC1 expression, and that ASF/SF2 acted as an enhancer of endogenous RAC1b splicing.(35) Consistently, induction of ASF/SF2 by the inhibition of the phosphatidylinositol 3-kinase (PI3K) pathway promoted RAC1b expression, whereas induction of SRp20 by the activation of β-catenin/TCF4 signaling inhibited the expression of RAC1b.(35) Recently, a proteomic approach together with IHC and other analyses revealed that hnRNP A2/B1, hnRNP F and other elements are upregulated in gastric cancer, and that SR-A1, hnRNP A1, hnRNP K and others are upregulated in colorectal cancer. Proteomic approaches have facilitated to identify the deregulation of trans-elements in tumorigenesis. The discovery will open a new avenue to study the association between expression of trans-elements and splicing alteration. As the sequence technologies are rapidly developing, we hope that involvement of trans-elements in tumorigenesis will be resolved in the near future.

Table 4.   Alteration of trans-elements in digestive tract malignancy
Trans-elementCancer typeAlteration typeMolecular aspectsAnalytical methodsSelected references
  1. ASF/SF2, alternative splicing factor/splicing factor 2; ESE, exonic splice enhancer; FAP, familial adenomatous polyposis; hnRNP, heterogenous nuclear ribonuculear protein; IHC, immunohistochemistry; SRp, serine/arginine-rich protein.

hnRNP A2/B1, hnRNP F and othersGastric cancer (sporadic)Somatic alterationThe trans-elements are upregulated in gastric cancersIHC, proteomic analysis 
ASF/SF2, SRp40, SRp55, SRp75, hnRNP A1, SRp20 and othersColon cancer (sporadic)Somatic alterationThe expressions of ASF/SF2, SRp40 and others were correlated to alternative CD44 splicingRT-PCR, northern blot analysis31
hnRNP KColon cancer (sporadic)Somatic alterationThe hnRNP K mutation c.274G>A reflected an RNA editing in cancerGenomic and cDNA sequencing, in vitro phospholyration assay33
SRp20, ASF/SF2Colon cancer (sporadic), HT29 cell lineSomatic alterationSRp20 increases inclusion of exon 3b of RAC1, whereas ASF/SF2 increases its skippingIn vitro transcription assay35
SR-A1, hnRNP A1, hnRNP K and othersColon cancer (sporadic)Somatic alterationThe trans-elements are upregulated in colorectal cancersRT-PCR, IHC, proteomic analysis 

Aberrant expression profile of splice variants

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

The aberrant expression profile of splice variants in digestive tract malignancies has been discussed in various genes; altered expression was reported in genes including CDH1 (E-cadherin), CD82 (KAI1), WISP1, BIRC5 (survivin), CD44, FGFR4, FHIT, MUTYH (MYH), FGFR2, MUC1 and CDCA1(6) in gastric cancer, and in genes including MLH1, CCND1 (cyclin D1), APC, VEGF (VEGFA), RAC1, MST1R, TCF4, CTNNB1 and TP53 (p53) in colorectal cancer. Among these genes, the function of the splice variants has been precisely analyzed for survivin (BIRC5) and RAC1 (Fig. 3).

image

Figure 3.  Splice variants in digestive tract malignancy. (a) Splice variants of the survivin gene. (b) Splice variants of the RAC1 gene. A GenBank accession number or an Emsembl transcript ID number for each variant is indicated on the left. Commonly used symbols for each variants are indicated in parentheses. The green boxes indicate constitutive exons and the blue boxes indicate alternative exons. a.a., amino acid.

Download figure to PowerPoint

Survivin was initially identified as an anti-apoptotic oncogene, and its expression is enhanced in various tumors.(36) Survivin is classified into the IAP family that includes cIAP1, cIAP2 and XIAP, and the members have been considered to be potential therapeutic targets of several types of malignancies.(37) The elevated expression of survivin might play a role in tumorgenesis through increased tumor cell viability, and may render the anti-apoptotic property for cancer cells to overcome the cytotoxic effects of anticancer drugs.(37) Two alternative splice variants of survivin have been reported, Sur-DeltaEx3 (NM_001012270.1) and Sur2B (NM_001012271), by Mahotka et al.(38) (Fig. 3a). The authors additionally revealed that Sur2B lacks its anti-apoptotic potential and acts as an antagonist against naturally occurring survivin. In good agreement with the apoptotic role of Sur2B, the variant was downregulated in metastatic lesions of gastric cancer.(39) A clear understanding of the interplay of the pro- and anti-apoptotic functions of the survivin splice variants is required before successful anti-survivin therapies can be fully developed.(40)

As described in an earlier paragraph, the small GTPase RAC1 gene has two splice variants: RAC1WT (NM_018890) and RAC1b (NM_006908). The RAC1b variant lacks alternative exon 3b and is shorter than RAC1WT, a wild-type variant with the in-frame inclusion of exon 3b, by 57 nucleotides in size (Fig. 3b). The splice variant RAC1b is overexpressed in a subset of colorectal tumors, and is required to sustain tumor cell viability.(34) Additionally, another report showed that activation of the tumor-specific splice variant RAC1b decreased the adhesion of colorectal cancer cells.(41) On the other hand, RAC1WT but not RAC1b, stimulated RelB-mediated gene transcription in cells.(42)

The association of VEGF splice variants with colorectal cancer has been reported by several groups.(43,44) VEGF(xxx) are the pro-angiogenic isoforms and VEGF(xxx)b are the anti-angiogenic isoforms, where xxx denotes the amino acid number, and the isoform VEGF(xxx) or VEGF(xxx)b is defined by the alternative splicing of a mutually exclusive exon at the 3′ end, namely exon 8a or exon 8b. SRp55, ASF/SF2 and SRpK are shown to alter the selection of splice sites of the VEGF gene in colorectal cancer cells.(43) The elucidation of regulatory mechanisms of the splice variants will be helpful for the development of cancer-specific VEGF inhibitors.(44)

Therapeutic strategies targeting aberrant alternative splicing

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

There are two strategies to target alternative splicing: (i) target aberrant splice variants or their resulting products; and (ii) target trans-elements. Products that are generated from alternative pre-mRNA splicing and are playing a vital role in malignancy are potential therapeutic targets. If such products are identified, it will be possible to treat cancer patients more selectively and with individualized therapies by regulating the altered splice variants of the target gene rather than regulating the entire target gene. Therefore, the identification of functionally relevant cancer-specific splice variants as well as that of cancer-specific deregulation in trans-elements are important tasks for the development of future therapies.

Targeting aberrant splice variants may be achieved through conventional small molecules or RNA-based therapeutics including synthetically modified oligonucleotides, RNA interference,(45,46) ribozymes, aptamers and other strategies. Novel strategies targeting splice variants of survivin and VEGF have also been under investigation.(40,44) Since RNA-based molecular therapeutics target specific nucleotide sequences, it should have a wide range of targets and high selectivity. However, one of the most important issues to be resolved is the development of a drug delivery system suitable for the therapeutics.

Targeting trans-elements that act as spliceosomes(17) or splicing modulators is another option. Subunits in activated spliceosomes, hyperphosphorylation of SR proteins and upregulated splicing modulators in tumors are potential targets for cancer treatment. Recently, two natural products have been isolated for this purpose, the pladienolide derivatives(47) and spliceostatin A,(48) which bind to SF3b and inhibit spliceosomal function. Benzothiazole, a Clk1/Sty inhibitor,(49) inhibits ASF/SF2-dependent splicing through the suppression of Clk-mediated phosphorylation. The development of RNA therapeutics targeting the SRpK using the siRNA strategy has been ongoing, because activation of SRpK leads to the hyperphosphorylation of SR proteins, thereby activating splicing.

Conclusions

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References

In the present review, we have summarized the current knowledge of the regulatory mechanisms involved in alternative pre-mRNA splicing and aberrant alternative splicing. In particular, we have focused on aberrant splicing and altered variant expression in gastrointestinal malignancies. As shown in this review, some of the silent/missense/nonsense mutations lead to exon skipping, retention of the intron or introduction of a new or cryptic splice site, although they are generally considered to result in no change, amino acid alteration or termination in amino acid sequence, respectively. Considering that in silico computer predictions do not always correlate with in vitro and in vivo splicing defects, we have to be careful of the interpretation of nucleotide changes in cis-elements. We also need to keep in mind that the regulatory mode of alternative pre-mRNA splicing might change in organ-, tissue- or cell-dependent manners. Therefore, accumulation of data on alternative splicing in different normal as well as malignant tissues is of great importance. In addition, information on the changes in trans-elements in neoplasms is so far limited. Since changes in splicing cannot be explained by the information of cis-elements alone, we have to accumulate knowledge on the changes of trans-elements in normal and malignant tissues and their roles in each type of cells. Acquisition of a huge body of human genome and transcript information has started to decipher the splicing codes and unveil the insights of splicing mechanisms.(1,50) The integrated information of the cis- and trans-elements and that of splice variants will help us develop more accurate prediction algorithms of the aberrant splicing in each type of tumors. As splice variants and altered trans-elements specifically in malignant tissues are promising targets for diagnosis and anticancer drugs, we hope that the integration of genome, transcriptome, proteome, functional and clinical data will make rapid progress in the development of new diagnostic and therapeutic strategies.

References

  1. Top of page
  2. Abstract
  3. Regulatory mechanism of alternative pre-mRNA splicing
  4. Alternative pre-mRNA splicing altered in malignancy
  5. Alteration of cis-elements
  6. Alteration of trans-elements
  7. Aberrant expression profile of splice variants
  8. Therapeutic strategies targeting aberrant alternative splicing
  9. Conclusions
  10. Disclosure statement
  11. References