Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (24)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Surono, A.
Right arrow Articles by Matsuo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Surono, A.
Right arrow Articles by Matsuo, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics Pages 493-500  

Circular dystrophin RNAs consisting of exons that were skipped by alternative splicing
Introduction
Results
Discussion
Materials And Methods
   Amplification of scrambled exons
   Amplification of authentic transcript
   Examination of cap structure or poly(A) tail sequence
   Sequencing of the PCR product
   Muscle biopsy samples from DMD cases
Acknowledgements
References

Circular dystrophin RNAs consisting of exons that were skipped by alternative splicing

Circular dystrophin RNAs consisting of exons that were skipped by alternative splicing

Agus Surono1, Yasuhiro Takeshima2, Tri Wibawa1, Makoto Ikezawa3, Ikuya Nonaka3 and Masafumi Matsuo1,*

1Division of Genetics, International Center for Medical Research and 2Department of Pediatrics, Kobe University School of Medicine, 7-5-1, Kusunokicho, Chuo, Kobe 650-0017, Japan and 3Department of Ultrastructural Research, National Institute of Neuroscience, National Center of Neurology and Psychiatry, Ogawahigashicho, Kodaira, Tokyo 187-0031, Japan

Received November 1, 1998; Revised and Accepted December 11, 1998

Exon skipping by alternative splicing and circular RNA formation are proposed to be interrelated events. Since multiple patterns of alternative splicing have been demonstrated in both the 5[prime] and 3[prime] regions of the dystrophin gene, the dystrophin transcript in skeletal muscle cells provides a model system in which this idea is tested. Nine circular RNAs that were expected to result from known exon skipping patterns in the 5[prime] region of this gene were in fact identified, but three other circular RNAs expected to result from other known exon skipping reactions in this region could not be detected. The identification of two unexpected circular RNAs led to the discovery of two novel alternative splicing reactions. One circular RNA originating in the 3[prime] region of the gene was identified but it lacked one small sized exon compared with the expected exon structure. Circular RNAs from the 5[prime] region of the dystrophin gene could not be detected in Duchenne muscular dystrophy patients who have deletions of one or more exons in this segment of the gene, even though alternative splicing products were identified. These results showed that circular RNA formation is not necessarily linked to exon skipping and suggest that an undetermined factor regulates circular RNA formation.

INTRODUCTION

Pre-mRNA undergoes several modifications, such as splicing, cap site addition and poly(A) tail addition, on its way to becoming mature mRNA. Pre-mRNA splicing, which releases the intron as a lariat, proceeds with remarkable accuracy, and an ordered array of exons is maintained in mature mRNA. Splicing is carried out by the spliceosome, a multi-component ribonuclear protein complex. In addition to authentic transcripts, scrambled transcripts, characterized by joining exons in an order different from that in the genome, have been observed in transcripts from a limited number of human genes. Some scrambled transcripts contain the donor splice site of a 3[prime] exon joined to the acceptor site of a 5[prime] exon, and have properties that indicate that they are circular (1).

Circular RNA was first identified in transcripts of the DCC (deleted in colorectal cancer) gene expressed in a human cell line (2). Since then, circular RNAs have been identified in transcripts of the human ETS-1 gene and the cytochrome P450 2C18 gene expressed in a cell line or in the epidermis, respectively (3-5). Transcripts of the rat kidney cytochrome P450 2C24 gene (6) and the rat testis androgen-binding protein gene (5) were also shown to produce circular RNAs. These circular transcripts were found to be substantially less abundant than their corresponding authentic mRNAs, which makes it technically difficult to study these molecules. Since circular RNA formation is not a phenomenon restricted to a single gene or cell lines, these circular molecules have been assumed to play important roles in human gene expression and development (1). However, the biological function of the circular RNA molecules remains unknown.

The exact mechanism by which circular RNA is produced has not been elucidated. The presence of the skipped exons of the mRNA in the circular RNA led to the hypothesis that exon skipping and circular RNA formation are mechanically interrelated phenomena (the exon skipping-circular RNA hypothesis) (6). Furthermore, it was proposed that alternative pre-mRNA splicing has the potential to generate not only mRNAs that lack one or more exons but also circular RNA molecules that are composed of the skipped exons. Circular RNA conceivably could be generated from a loop structure of the lariat released from the pre-mRNA (6). Indeed, spliceosome-mediated exon circularization has been shown to occur in vitro (7). However, the absence of a tight correlation between alternative splicing and circular RNA formation was noted in studies of transcripts of the rat cytochrome P450 2C24 gene (5).

The dystrophin gene, which is mutated in Duchenne/Becker muscular dystrophy (DMD/BMD) patients, is the largest human gene, spreading over >3000 kb of the X chromosome (8) and consisting of 79 exons that form a 14 kb transcript. Deletion mutations that remove several kilobases including one or more exons have been identified in ~60% of DMD/BMD cases, and two deletion hot spots have been identified in the 5[prime] and central regions of the dystrophin gene. In addition to its huge size, the gene is characterized by the presence of multiple alternative splicing events that result in exon skipping (9). At least 12 patterns of alternative splicing have been identified in the 5[prime] region of the dystrophin gene in normal skeletal muscle (10), and the possibility that alternative splicing could modify the clinical phenotype of DMD/BMD by editing the translational reading frame has been pointed out (11-13). In addition, alternative splicing of exons 71-74 at the 3[prime] end of the gene produces tissue-specific transcripts in brain cells (14) and in cardiac Purkinje fibers (15).

Although multiple alternative splicing is known to occur in the dystrophin transcript, circular RNAs have not been studied. In this report, we identified circular dystrophin transcripts in normal skeletal muscle for the first time. The identification of certain species of the circular RNA led to the discovery of novel alternative splicing reactions. Studies of DMD patients showed that disruption of normal exon organization by deletion mutations resulted in the disappearance of all possible circular RNAs. These results shed doubts on the proposed correlation between alternative splicing and circular RNAs.

RESULTS

In a previous report, we identified 12 patterns of alternative splicing in the 5[prime] region of the dystrophin gene covering exons 1-18, on the basis of the absence of exons from the mature dystrophin transcript (10). Since it has been proposed that exon skipping and circular RNA formation are interrelated events [exon skipping-circular RNA hypothesis (6)], we studied the correlation between exon skipping and the formation of circular RNAs in the dystrophin transcript. First, we attempted to detect circular transcripts originating from the 5[prime] region of the dystrophin gene in skeletal muscle samples. RNA in which the 3[prime] and 5[prime] ends were from exons 17 and 2, respectively, was expected to be present at relatively high amounts as a result of the known alternative splicing of exons 1-18 (10). Thus, we attempted to amplify this RNA by nested PCR from cDNA synthesized using random primers. A fragment spreading from exon 15 to exon 6 was amplified by using an outer set of primers (c15F and c6R) (Fig. 1a, reaction 1). This amplification reaction did not lead to a product that could be detected by gel electrophoresis, although the authentic dystrophin transcript could be amplified by PCR amplification from the same cDNA sample (data not shown) (16).


Figure 1. Schematic description of the strategy to amplify scrambled exons. (a) Scrambled exons in the 5[prime] region of the dystrophic gene. A fragment spreading from exon 16 to exon 3 or exon 5 was obtained by nested PCR (reaction 1). The first PCR was carried out using the c15F and c6R primers (bold arrowheads). To obtain the exon 16 to exon 3 or exon 16 to exon 5 fragments, a second PCR was carried out using either the c16F and c3R primers (arrows) (reaction 1a) or the c16F and c5R primers (arrows) (reaction 1b), respectively. The fragment spreading from exon 13 to exon 3 was obtained by using an outer set of primers (c12F and c4R; bold arrowheads) and an inner set of primers (c13F and c3R; arrows) (reaction 2). The fragment spreading from exon 7 to exon 3 was amplified by nested PCR using primer sets c6F and c5R (bold arrowheads) and c7F and c3R (arrows), respectively (reaction 3). (b) Scrambled exons in the 3[prime] region of the dystrophin gene. A fragment spreading from exon 74 to exon 72 was obtained by nested PCR using an outer set of primers (c73F and c72R2; bold arrowheads) and an inner set of primers (c74F and c72R3; reaction 4), respectively. Boxes with numbers within represent exons. The figure is not drawn to scale.

A fragment extending from exon 16 to exon 3 was then amplified from the first amplified product by using an inner set of primers (c16F and c3R) (reaction 1a). As a result, several products of different sizes were obtained (Fig. 2a, lane 1). All of them were sequenced. Sequence analysis of a 469 bp product showed that it contained the 3[prime] end of exon 16 joined precisely to the 5[prime] end of exon 17, the 3[prime] end of exon 17 directly and precisely joined to the 5[prime] end of exon 2, and the 3[prime] end of exon 2 joined to the 5[prime] end of exon 3 (exon 17-2 transcript ) (Fig. 3a). The presence of exons 4-15 in this transcript was confirmed by sequencing separate amplified products covering exons 2-17 (data not shown). The sequenced regions did not differ from the authentic dystrophin cDNA sequence.


Figure 2. Amplification products obtained by reverse transcription-nested PCR. (a) Amplified products spreading from exon 16 to exon 3, as revealed by the gel electrophoresis pattern of products obtained in reaction 1a (Fig. 1a). The fragments corresponding to 469, 407 and 231 bp with unexpected fragments of 691 and 531 bp were obtained from cDNA prepared by using random or c6R primers (lanes 1 and 2, respectively). The size of each fragment is indicated on the right. For negative control, cDNA synthesized without reverse transcriptase was used (lane 3). Mk is a DNA size marker of HaeIII-digested [phis]X174 DNA (Takara, Shuzo, Japan). Additional minor bands detected on the gel represent misprimed PCR amplification products. (b) Amplified fragment spreading from exon 16 to exon 5, as obtained in reaction 1b (Fig. 1a). The expected fragment corresponding to 426 bp was obtained from cDNA prepared by using random or c6R primers (lanes 1 and 2, respectively). The additional minor band detected represents a misprimed PCR amplification product. Other details are as in (a). (c) Amplified product spreading from exon 13 to exon 3. The expected fragment corresponding to 203 bp was obtained from cDNA prepared by using random and c6R primers (lanes 1 and 2, respectively). The additional minor bands detected in the gel represent misprimed PCR amplification products. Other details are as for (a). (d) Amplified product spreading from exon 7 to exon 3. Four major and weak bands were visualized in the product of reaction 3. The same amplified products were obtained from cDNAs synthesized by using random or exon-specific primers (lanes 1 and 2). Other details are as for (a).(e) Amplified product spreading from exon 74 to exon 72. In reaction 4, a product of an unexpected size corresponding to 190 bp was obtained from cDNA generated by random and c72R1 primers (lanes 1 and 3, and 2 and 4, respectively) from human skeletal muscle and brain (lanes 1 and 2, and 3 and 4, respectively). cDNA synthesized without reverse transcription was used as a negative control (lane 5). Other details are as for (a).


Figure 3. Schematic representation of the results of sequencing of PCR-amplified products. The junction sequences of the 3[prime] ends of the 3[prime] exons and the 5[prime] ends of the 5[prime] exons are shown. All of the junctions were shown to be perfect, i.e. the sequences were precisely those predicted for the exact joining of the two exons. The presumed RNA structure is represented schematically on the right.

In the 407 and 231 bp products, the 3[prime] end of exon 17 or 16 was found to be joined directly to the 5[prime] end of exon 3 (exon 17-3 and exon 16-3 transcripts, respectively) (Fig. 3b and c). All of these transcripts corresponded to the predicted products of alternative splicing reactions previously shown to occur in the 5[prime] region of the dystrophin gene (10).

More surprisingly, sequencing of the 691 and 531 bp products disclosed that the 3[prime] end of exon 19 or 18 was joined directly to the 5[prime] end of exon 3 (exon 19-3 and exon 18-3 transcripts, respectively) (Fig. 3d and e). These two products were not expected to occur as a result of any of the alternative splicing reactions identified in our previous study (10). Therefore, we specifically looked for splicing products that might give rise to the exon 19-3 and exon 18-3 transcripts by re-examining amplified products spreading from exon 1 to exon 20 (10). In addition to the fragments identified previously (10), two barely visible bands were identified near the electrophoresis front (data not shown). Sequencing of these two products disclosed that they corresponded to two novel alternative splicing products in which the 3[prime] end of exon 2 was joined directly to the 5[prime] end of exon 20 in one and of exon 19 in the other, respectively (Fig. 4). Thus, the detection of scrambled exons led to the identification of novel alternative splicing reactions in the dystrophin gene.


Figure 4. Sequences of the junction regions of exons produced by novel alternative splicing reactions. Both of the two barely visible bands spreading from exon 1 to exon 20 were sequenced. Results of sequencing disclosed that the 3[prime] end of exon 2 joined directly and precisely to the 5[prime] end of exon 20 in one product (a) and of exon 19 in the other (b). No nucleotide changes were detected in the sequenced region.

Sequencing of all of the other weak bands detected after the PCR reaction 1a (Fig. 2a, lane 1) revealed them to be non-specific. The same amplification products were obtained when cDNA synthesized by using antisense primer to exon 6 was used as a template for nested PCR amplification (Fig. 2a, lane 2). As expected, a product was not obtained when cDNA synthesized without reverse transcription was used as a template (Fig. 2a, lane 3), indicating that the products were derived from RNA. Furthermore, the products identified above could not be amplified when antisense primer to exon 16 was used at the time of cDNA synthesis (data not shown), suggesting that they were not derived from tandemly arranged exons.

A fragment spreading from exon 16 to exon 5 was then amplified from the same first product spreading from exon 15 to exon 6 (Fig. 1a, reaction 1b). Examination of the products of this reaction by gel electrophoresis revealed one strong band and one weak band (Fig. 2b). Sequencing of these products showed that the weak band was non-specific, while the strong band consisted of exons 16, 17 and 5 (exon 17-5 transcript) (Fig. 3f). To look for other scrambled exons, the amplified fragment spreading from exon 13 to exon 3 was amplified following amplification of a fragment spreading from exon 12 to exon 4 (Fig. 1a, reaction 2). This time, one abundant amplified product and several non-specific products were obtained (Fig. 2c). The abundant product was shown to be a scrambled exon consisting of exons 12, 13, 3 and 4 (exon 13-3 transcript) (Fig. 3g). In a different reaction, a region spreading from exon 7 to exon 3 was obtained after first amplifying a fragment spreading from exon 6 to exon 5 (Fig. 1a, reaction 3). In this pair of reactions, two clear and two weak bands were obtained (Fig. 2d). Sequence analysis showed that these bands corresponded to four scrambled products designated, respectively, the exon 7-3, exon 7-2, exon 9-3 and exon 9-2 transcripts (Fig. 3h-k). Each of these products would be expected to result from alternative splicing reactions that are known to occur in the 5[prime] region of the dystrophin gene pre-mRNA.

To investigate whether a cap structure is present in these scrambled exons, cDNA enriched in cap-containing transcripts was prepared as described in Materials and Methods. A fragment spreading from exon 1 to exon 6 was PCR amplified from this preparation of cDNA and the authentic splicing product was obtained (data not shown). However, no amplified products were obtained from cap-containing cDNA by using the nested PCR reactions 1, 2 or 3 (data not shown). The observed lack of a cap structure in the scrambled exons indicated that they were not linear trans-spliced products of two authentic pre-mRNAs. These observations indicated that scrambled transcripts are indeed circular RNAs, as represented schematically in Figure 3.

As a result of these studies, we identified 11 circular transcripts in the 5[prime] region of the dystrophin gene (Fig. 3a-k). We repeatedly failed to detect the two single exon transcripts corresponding to what would be expected to result from the demonstrated skipping of exon 4 or exon 9, although these products might have been too small to detect. More importantly, the scrambled transcript that would be expected to result from the known alternative splicing reaction that skips exons 2-16 could also not be identified, despite repeated attempts. This result implies that alternative splicing does not necessarily result in the production of circular RNA.

Exons 71, 72, 73 and 74 in the 3[prime] region of the dystrophin gene are subject to alternative splicing in brain and cardiac Purkinje fibers but not in skeletal muscle (14,15). In fact, amplification of an authentic fragment spreading from exon 70 to exon 75 resulted in the production of only one fragment from skeletal muscle but two fragments from brain tissue, indicating brain-specific alternative splicing of exons 71-74 (data not shown). We therefore attempted to amplify a scrambled exon spreading from exon 74 to exon 72 by nested PCR (Fig. 1b, reaction 4). Remarkably, the expected PCR product was obtained not only from brain tissue but also from skeletal muscle (Fig. 2e). Sequencing of the product disclosed that the 3[prime] end of exon 74 was joined directly to the 5[prime] end of exon 72 (exon 74-72 transcript) (Fig. 3i). However, circular RNA retaining exon 71 was not obtained even from brain tissue, in which skipping of exons 71-74 has been clearly demonstrated to occur. The detection of the exon 74-72 transcript in skeletal muscle was surprising because alternative splicing from exon 70 to exon 75 has not been reported in this tissue except in one BMD patient (17). Thus, we looked for alternative splicing products in normal muscle by employing nested PCR designed to amplify a fragment corresponding to exons 70-75 (17). A very weak additional band was visualized after the second round of PCR amplification. Sequencing of this weak product disclosed that it contained exon 70 joined directly to exon 75 (data not shown). This result indicates that the alternative splicing reaction that jumps from exon 70 to exon 75 occurs even in skeletal muscle, albeit at an extremely low level. The presence of a poly(A) tail in the resulting exon 74-72 transcript was examined by amplifying this transcript from cDNA synthesized by using a poly(T) primer. Although the authentic dystrophin transcript could be amplified from this cDNA, no product could be obtained in PCR reaction 4 (data not shown). This result confirmed that the exon 74-72 transcript lacks a poly(A) tail, supporting the idea that the scrambled exon is circular RNA.

In an attempt to determine whether circular RNAs have a pathophysiological role, we tried to detect them in skeletal muscle of DMD patients. Remarkably, formation of circular RNAs differed according to the location of deletion mutations in the dystrophin gene. Circular RNAs derived from the 5[prime] region of the dystrophin gene could not be obtained from four DMD patients carrying a deletion of at least one exon in this region of the gene (deletions of exons 3-6, 4-6, 6 or 8-13), even though the presence of alternatively spliced mRNA from the 5[prime] region could be demonstrated (Table 1). In contrast, circular RNAs from the 5[prime] region could be detected in three DMD cases carrying a deletion of exons in the central region of the dystrophin gene (deletions of exons 48-52, 46-53 or 45-52) as alternatively spliced products were detected (Table 1). Circular RNA derived from the 3[prime] region of the dystrophin gene could be detected in all seven cases (Table 1). These results indicated that disruption of the normal exon constitution by deletion of one or more exons prevents the formation of circular RNAs in the region affected by the deletion.

Table 1. Circular RNAs identified in DMD cases
Deletion
 5[prime] region
Alternative splicing		  
Circular RNA		 3[prime] region
Alternative splicing		  
Circular RNA
In the 5[prime] region        
Exons 3-6 + - + +
Exons 4-6 + - + +
Exon 6 + - + +
Exons 8-13 + - + +
In the central region        
Exons 48-52 + + + +
Exons 46-53 + + + +
Exons 45-52 + + + +
+, identified; -, not identified.

DISCUSSION

In this report, we first identified dystrophin gene-derived scrambled exons in which the splice donor site of the 3[prime] exon is joined precisely to the splice acceptor site of the 5[prime] exon (Fig. 3). This was achieved by amplifying skeletal muscle cDNA by nested PCR using a forward primer based on the sequence of the 3[prime] exon and a reverse primer complementary to the 5[prime] exon (Fig. 1). These scrambled exons could not be amplified from either cDNA corresponding to cap-containing transcripts or from cDNA obtained by using a poly(T) primer. These results indicated that scrambled exons identified in this study were circular RNAs, as reported in other human genes (2-4). The following findings led to the conclusion that circular RNAs were not artificial: (i) without a reverse transcription step, no amplified product corresponding to circular RNA was obtained; (ii) nucleotide sequences at the borders of exons were completely conserved in circular RNAs (Fig. 3); and (iii) circular RNAs from the 5[prime] region were detected in the limited cases of DMD, but not in others (Table 1). Our findings extend the original findings of exon circularization to the largest human gene, which is characterized by the large sizes of the introns and by the presence of a large number of exons. This finding suggested that formation of circular RNA is a universal phenomenon in transcript splicing.

The exon skipping-circular RNA hypothesis proposes that lariats containing skipped exons undergo additional splicing events that result in the circularization of the skipped exons (6). In support of this hypothesis, we identified 12 circular RNAs consisting of exons of the dystrophin transcript that were skipped due to alternative splicing. In particular, the identification of two additional circular RNAs (exon 19-3 and exon 18-3 transcripts) prompted us to study two minor products that resulted from two previously undetected novel splicing reactions, from exon 2 to exon 20 and from exon 2 to exon 19 (Fig. 4). Thus, the identification of circular RNA was an important clue that led to the identification of new splicing reactions and further supported the idea that formation of circular RNA is linked with alternative splicing. However, we failed to detect three products that were expected to result from alternative splicing reactions known to occur in the 5[prime] region of the dystrophin gene and one product expected to result from an alternative splicing reaction known to occur in the 3[prime] region of the gene (10,17). In the latter case, we failed to detect the expected circular RNA containing exon 71 even though skipping of exons from 71 to 74 was shown to occur. This could presumably be due to a failure of the spliceosome to recognize the unusually small exon 71 (39 bp). This could also be the case for two single exon transcripts that we failed to detect in the 5[prime] region of the gene, but the exon 16-2 transcript that we expected but failed to detect would almost certainly be sufficiently large (1691 bp) to be detected by the spliceosome. In conclusion, our results show that circular RNA formation in the dystrophin transcript does not always result from alternative splicing, implying that other unknown mechanisms are involved. A loose correlation between alternative splicing and circular RNA formation was also reported for the human cytochrome P450 2C18 and rat androgen-binding protein genes (5). These results call into question the validity of the exon skipping-circular RNA theory.

The number of exons that are circularized has been reported to range from one to five. For example, a circular RNA from the testis-determining Sry gene was reported to be composed of a single exon of 1231 nucleotides that had been joined head to tail (18). Also, a circular RNA consisting of five exons was reported in the human c-ets-1 gene transcript (3). Our results show that the highest number of exons detected in a circularized transcript can now be extended to 17 (as in the 2287 nucleotide exon 19-3 transcript; Fig. 3). Since exons 3-19 span a region of >500 kb of the dystrophin gene, our result suggest that a 500 kb long lariat can be spliced efficiently by the spliceosome to produce circular RNA. Thus, RNA circularization proceeds regardless of the number of exons or the size of the introns.

Circular RNAs could only be amplified by nested PCR (Fig. 1), while authentic transcripts could be obtained by one round of PCR amplification. This fact indicated that circular RNAs were present at much lower levels than the exon-skipped authentic transcripts, even though a 1:1 correlation between exon-skipped and circularized transcripts would be expected. Differential stabilities of these molecules could explain the observed differences in the proportions of the two products. In fact, circular transcripts of the androgen-binding protein gene in rat testis have been reported to be less stable than the exon-skipped transcript (5). In this study, the exon 16-2 transcript could not be detected even though alternative splicing skipping of exons 2-16 does occur. The inability to detect the exon 16-2 transcript might be due to either extreme instability of this molecule or to the fact that it is not produced.

In vitro studies have shown that circularization is enhanced when an intrinsic complementary sequence is introduced into the RNA substrate to form a stem that brings the splice sites into proximity to promote splice site pairing (7). Therefore, we must consider the effect of secondary structure of pre-mRNA of the dystrophin gene on circular RNA formation. However, the dystrophin gene is too big to analyze for the secondary structure of pre-mRNA (19) and we cannot discuss this effect further. Instead, the effect of deletion mutations on circular RNA formation was studied in skeletal muscle from DMD patients. The results showed that the absence from the gene of an exon normally present in the circular RNA resulted in a failure to detect the circular RNA (Table 1). This suggested that the structure of the pre-mRNA influences circular RNA formation.

To date, at least three possible roles have been proposed for circular RNA exons in vivo. First, circular RNAs could be protein-encoding entities much like linear RNAs (20,21). Second, circular RNA exons might not be translated in vivo but nevertheless might play important roles in regulating gene expression (22). Third, all circular RNAs might simply be functionless products of aberrant splicing events (23). In this study, we attempted to clarify the pathophysiological role of circular RNAs in patients with mutations of the dystrophin gene by analyzing circular RNAs in skeletal muscle. In the DMD patients who had deletions of one or more exons in the 5[prime] region, none of the circular RNAs that normally are obtained from the 5[prime] region of the primary transcript could be shown to be produced even though alternative splicing proceeded normally (Table 1). In contrast, the formation of these circular RNAs was not affected by deletions of exons located further downstream in the transcript (Table 1). From these observations, we conclude that circular RNA formation is not related to clinical severity of DMD. We currently are analyzing more samples from DMD cases to determine whether there is any correlation between circular RNA formation and clinical or pathological findings.

Table 2. Nucleotide sequence of primers used in this study
Primer Sequence
c3R 5[prime]-CCTGTCAGGCCTTCGAGGAG-3[prime]
c4R 5[prime]-GTTCAGGGCATGAACTCTTG-3[prime]
c5R 5[prime]-TGCCAGTGGAGGATTATATTCCAA-3[prime]
c6R 5[prime]-AGCTCAGGAGAATCTTTTCA-3[prime]
c6F 5[prime]-TGAAAAGATTCTCCTGAGCT-3[prime]
c7F 5[prime]-GCCAGACCTATTTGACTGGAA-3[prime]
c12F 5[prime]-ACGCCAAGTACAACAACATAA-3[prime]
c13F 5[prime]-GTGCTTCAAGAAGATCTAGAAC-3[prime]
c15F 5[prime]-ACACAACTGGCTTTAAAGAT-3[prime]
c16F 5[prime]-ATCCATGGGCAAACTGTATT-3[prime]
c72R1 5[prime]-CTGCTAGCATAATGTTCAAT-3[prime]
c72R2 5[prime]-GCGTGAATGAGTATCATCGT-3[prime]
c72R3 5[prime]-TATCATCGTGTGAAAGCTGAG-3[prime]
c73F 5[prime]-ATGATAGCATCTCTCCTAATG-3[prime]
c74F 5[prime]-GATGAACATTTGTTAATCCAGC-3[prime]

MATERIALS AND METHODS

Amplification of scrambled exons

Skeletal muscle was obtained from human quadriceps at the time of orthopedic surgery after informed consent was obtained. Total RNA from muscle was extracted as described before (16), and total RNA of whole brain was purchased from Sawady (Tokyo, Japan). Three micrograms of total RNA was used for cDNA synthesis as previously described (16) using random or antisense hexamer oligonucleotides as primers. As a control, cDNA prepared without reverse transcriptase was employed.

Nested PCR was employed to detect circular RNA (24). Basically, the first PCR amplification was performed using an outer set of primers (antisense to the 5[prime] exon and sense to the 3[prime] exon). Then, a portion of the product was subjected to a second amplification reaction using an inner set of primers (antisense sequence of the 5[prime] exon and sense sequence of the 3[prime] exon). The primers used for the PCR reactions are indicated in Table 2 and their positions in the dystrophin gene are shown in Figure 1. The PCR conditions were essentially as previously described (16,24). In order to examine all circular transcripts corresponding to alternative splicing in the 5[prime] region, three different nested PCR reactions were performed (reactions 1, 2 and 3). Furthermore, the first PCR product in reaction 1 was subjected to a second, different PCR amplification by using two sets of primers (Fig. 1a, reactions 1a and 1b). For PCR amplification of circular RNA from the 3[prime] region of the dystrophin gene, a fragment spreading from exon 73 to exon 72 through exon 74 was amplified in the first PCR reaction and a fragment spreading from exon 74 to exon 72 was amplified in the second PCR reaction (reaction 4) (Fig. 1b).

Amplification of authentic transcript

To analyze authentic dystrophin transcripts, fragments spreading from exons 1 to 6, 1 to 20 and 70 to 75 were amplified using a forward primer corresponding to the sense strand of the 5[prime] exon and a reverse primer corresponding to the antisense strand of the 3[prime] exon (10,17).

Examination of cap structure or poly(A) tail sequence

In order to study the cap structure of scrambled dystrophin transcripts, a SMART PCR cDNA synthesis kit (formerly known as Cap Finder) (Clontech Laboratories, Palo Alto, CA) was used. Briefly, muscle cDNA was synthesized and ligated with the SMART II oligonucleotide (Clontech) and subsequently was amplified using a primer complementary to exon 6 in combination with a primer on the SMART II oligonucleotide under the conditions recommended by the supplier. The amplified DNA, which was enriched in transcripts containing a cap structure, was then examined for scrambled dystrophin transcripts.

In order to examine poly(A) tail sequence in scrambled transcripts, muscle cDNA was synthesized by using poly(T) primer (Clontech). Then, the scrambled exon 74-72 transcript was amplified from this cDNA by PCR reaction 4 (see above).

Sequencing of the PCR product

PCR products were separated by electrophoresis on 2% agarose gels and then stained by ethidium bromide. For sequencing, each amplified product was excised from the gel. After purification, DNA was subcloned into pT7 blue T vector (Novagen, Madison, WI) and sequences of the inserted DNA were determined by automated DNA sequencing using a Taq dye terminator cycle sequencing kit (Perkin Elmer Applied Biosystems, Norwalk, CT) with an automatic DNA sequencer (model 373 A; Perkin Elmer Applied Biosystems) as described previously (24).

Muscle biopsy samples from DMD cases

Skeletal muscle biopsy samples were obtained from seven DMD patients. Four patients had a deletion mutation in the 5[prime] region of the dystrophin gene (deletions of exons 3-6, exons 4-6, exon 6 and exons 8-13, respectively). Three patients had a deletion in the central region of the gene (exons 48-52, exons 46-53 and exons 45-52, respectively). cDNA was prepared from biopsy muscular samples and examined for scrambled exons as described above.

ACKNOWLEDGEMENTS

We thank Dr A. Pugsley for advice and for critically reading the manuscript. We acknowledge Ms N. Kageyama for secretarial help. This work was supported by grants from the Ministry of Education, Science and Culture of Japan and the Research Grant (8A-1) for Nervous and Mental Disorders from the Ministry of Health and Welfare of Japan.

REFERENCES

1. Zaphiropoulos, P.G. (1998) Mechanisms of pre-mRNA splicing: classical versus non-classical pathways. Histol. Histopathol., 13, 585-589. MEDLINE Abstract

2. Nigro, J., Cho, K., Fearon, E., Kern, S., Ruppert, J., Oliner, J., Kinzler, K. and Vogelstein, B. (1991) Scrambled exons. Cell, 64, 607-613. MEDLINE Abstract

3. Cocquerelle, C., Daubersies, P., Majerus, M.-A., Kerchaert, J.-P. and Bailleul, B. (1992) Splicing with inverted order of exons occurs proximal to large introns. EMBO J., 11, 1095-1098. MEDLINE Abstract

4. Cocquerelle, C., Mascrez, B., Hetuin, D. and Bailleul, B. (1993) Mis-splicing yields circular RNA molecules. FASEB J., 7, 155-160. MEDLINE Abstract

5. Zaphiropoulos, P.G. (1997) Exon skipping and circular RNA formation in transcripts of the human cytochrome P-450 2C18 gene in epidermis and of the rat androgen binding protein gene in testis. Mol. Cell. Biol., 17, 2985-2993. MEDLINE Abstract

6. Zaphiropoulos, P.G. (1996) Circular RNAs from transcripts of the rat cytochrome P450 2C24 gene: correlation with exon skipping. Proc. Natl Acad. Sci. USA, 93, 6536-6541. MEDLINE Abstract

7. Pasman, Z., Been, M.D. and Garcia-Blanco, M.A. (1996) Exon circularization in mammalian nuclear extracts. RNA, 2, 603-610. MEDLINE Abstract

8. Nishio, H., Takeshima, Y., Narita, N., Yanagawa, H., Suzuki, Y., Ishikawa, Y., Ishikawa, Y., Minami, R., Nakamura, H. and Matsuo, M. (1994) Identification of a novel first exon in the human dystrophin gene and of a new promoter located more than 500 kb upstream of the nearest known promoter. J. Clin. Invest., 94, 1037-1042. MEDLINE Abstract

9. Ahn, A.H. and Kunkel, L.M. (1993) The structural and functional diversity of dystrophin. Nature Genet., 3, 283-291. MEDLINE Abstract

10. Surono, A., Takeshima, Y., Wibawa, T., Pramono, Z.A.D. and Matsuo, M. (1997) Six novel transcripts that remove a huge intron ranging from 250 kb to 800 kb are produced by alternative splicing of the 5[prime] region of the dystrophin gene in human skeletal muscle. Biochem. Biophys. Res. Commun., 239, 895-899. MEDLINE Abstract

11. Chelly, J., Gilgenkrantz, H., Hugnot, J., Hamard, G., Lambert, M., Recan, D., Akli, S., Cometto, M., Kahn, A. and Kaplan, J. (1991) Illegitimate transcription: application to the analysis of truncated transcripts of the dystrophin gene in nonmuscle cultured cells from Duchenne and Becker patients. J. Clin. Invest., 88, 1161-1166. MEDLINE Abstract

12. Winnard, A.V., Klein, C.J., Coovert, D.D., Prior, T., Papp, A., Snyder, P., Bulman, D.E., Ray, P.N., McAndrew, P., King, W., Moxley, R.T., Mendell, J.R. and Burghes, A.H.M. (1993) Characterization of translational frame exception patients in Duchenne/Becker muscular dystrophy. Hum. Mol. Genet., 2, 737-744. MEDLINE Abstract

13. Winnard, A.V., Mendell, J.R., Prior, T.W., Florence, J. and Burghes, H.M. (1995) Frameshift deletions of exons 3-7 and revertant fibers in Duchenne muscular dystrophy: mechanisms of dystrophin production. Am. J. Hum. Genet., 56, 158-166. MEDLINE Abstract

14. Feener, C.A., Koenig, M. and Kunkel, L.M. (1989) Alternative splicing of human dystrophin mRNA generates isoforms at the carboxy terminus. Nature, 338, 509-511. MEDLINE Abstract

15. Bies, R.D., Friedman, D., Roberts, R., Perryman, M.B. and Caskey, C.T. (1992) Expression and localization of dystrophin in human cardiac Purkinje fibers. Circulation, 86, 147-153. MEDLINE Abstract

16. Matsuo, M., Masumura, T., Nishio, H., Nakajima, T., Kitoh, Y., Takumi, T., Koga, J. and Nakamura, H. (1991) Exon skipping during splicing of dystrophin mRNA precursor due to an intraexon deletion in the dystrophin gene of Duchenne muscular dystrophy Kobe. J. Clin. Invest., 87, 2127-2131. MEDLINE Abstract

17. Patria, S.Y., Alimsardjono, H., Nishio, H., Takeshima, Y., Nakamura, H. and Matsuo, M. (1996) A case of Becker muscular dystrophy resulting from the skipping of four contiguous exons (71-74) of the dystrophin gene during mRNA maturation. Proc. Assoc. Am. Phys., 108, 308-314.

18. Capel, B., Swain, A., Nicolis, S., Hacker, A., Walter, M., Koopman, P., Goodfellow, P. and Lovell-Badge, R. (1993) Circular transcripts of the testis-determining gene Sry in adult mouse testis. Cell, 73, 1019-1030. MEDLINE Abstract

19. Matsuo, M., Nishio, H., Kitoh, Y., Francke, U. and Nakamura, H. (1992) Partial deletion of a dystrophin gene leads to exon skipping and to loss of an intra-exon hairpin structure from the predicted mRNA precursor. Biochem. Biophys. Res. Commun., 182, 495-500. MEDLINE Abstract

20. Bohjanen, P.R., Colvin, R.A., Puttaraju, M., Been, M.D. and Garcia-Blanco, M.A. (1996) A small circular TAR RNA decoy specifically inhibits Tat-activated HIV-1 transcription. Nucleic Acids Res., 24, 3733-3738. MEDLINE Abstract

21. Puttaraju, M., Perrotta, A.T. and Been, M.D. (1993) A circular trans-acting hepatitis delta virus ribozyme. Nucleic Acids Res., 21, 4253-4258. MEDLINE Abstract

22. Mikheeva, S., Hakim-Zargar, M., Carlson, D. and Jarrell, K. (1997) Use of an engineered ribozyme to produce a circular human exon. Nucleic Acids Res., 25, 5085-5094. MEDLINE Abstract

23. Sharp, P.A. (1994) Split genes and RNA splicing. Cell, 77, 805-815. MEDLINE Abstract

24. Hagiwara, Y., Nishio, H., Kitoh, Y., Takeshima, Y., Narita, N., Wada, H., Yokoyama, M., Nakamura, H. and Matsuo, M. (1994) A novel point mutation (G -1 to T) in a 5[prime] splice donor site of intron 13 of the dystrophin gene results in exon skipping and is responsible for Becker muscular dystrophy. Am. J. Hum. Genet., 54, 53-61. MEDLINE Abstract

*To whom correspondence should be addressed. Tel: +81 78 341 7451; Fax: +81 78 362 6064; Email: matsuo@kobe-u.ac.jp

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 6 Feb 1999
Copyright©Oxford University Press, 1999.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J. Med. Genet.Home page
Y Habara, Y Takeshima, H Awano, Y Okizuka, Z Zhang, K Saiki, M Yagi, and M Matsuo
In vitro splicing analysis showed that availability of a cryptic splice site is not a determinant for alternative splicing patterns caused by +1G->A mutations in introns of the dystrophin gene
J. Med. Genet., August 1, 2009; 46(8): 542 - 547.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
H. Suzuki, Y. Zuo, J. Wang, M. Q. Zhang, A. Malhotra, and A. Mayeda
Characterization of RNase R-digested cellular RNA source that consists of lariat and circular RNAs from pre-mRNA splicing.
Nucleic Acids Res., January 1, 2006; 34(8): e63 - e63.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
R. J. Dixon, I. C. Eperon, L. Hall, and N. J. Samani
A genome-wide survey demonstrates widespread non-linear mRNA in expressed sequences from multiple species
Nucleic Acids Res., October 19, 2005; 33(18): 5904 - 5913.
[Abstract] [Full Text] [PDF]

Home page
 J. Med. Genet.Home page
E Renkonen, H Lohi, H J Jarvinen, J-P Mecklin, and P Peltomaki
Novel splicing associations of hereditary colon cancer related DNA mismatch repair gene mutations
J. Med. Genet., July 1, 2004; 41(7): e95 - e95.
[Full Text] [PDF]

Home page
 J. Med. Genet.Home page
F Gualandi, C Trabanelli, P Rimessi, E Calzolari, L Toffolatti, T Patarnello, G Kunz, F Muntoni, and A Ferlini
Multiple exon skipping and RNA circularisation contribute to the severe phenotypic expression of exon 5 dystrophin deletion
J. Med. Genet., August 1, 2003; 40(8): e100 - 100.
[Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
C. Finta and P. G. Zaphiropoulos
Intergenic mRNA Molecules Resulting from trans-Splicing
J. Biol. Chem., February 15, 2002; 277(8): 5882 - 5890.
[Abstract] [Full Text] [PDF]

Home page
 Mol Biol EvolHome page
S. C. Warner, C. Finta, and P. G. Zaphiropoulos
Intergenic Transcripts Containing a Novel Human Cytochrome P450 2C Exon 1 Spliced to Sequences from the CYP2C9 Gene
Mol. Biol. Evol., October 1, 2001; 18(10): 1841 - 1848.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (24)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Surono, A.
Right arrow Articles by Matsuo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Surono, A.
Right arrow Articles by Matsuo, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?