| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INSULIN-GLUCAGON-GI PEPTIDES-DIABETES MELLITUS |
Transplantation and Autoimmunity Branch and Genetics and Biochemistry Branch (B.A.P.), National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20889-5603
Address all correspondence and requests for reprints to: Dr. Anath Shalev, Transplantation and Autoimmunity Branch, National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, AFFRI, Building 43, Room 3222, 8901 Wisconsin Avenue, Bethesda, Maryland 20889-5603. E-mail: . AnathS{at}intra.niddk.nih.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Increasing evidence suggests that alterations in RNA processing can lead to a variety of human diseases, including cancer (2, 3). That said, alternative splicing allows the production of multiple mRNA forms from one gene and therefore is crucial for protein diversity (1). Indeed, alternative splicing occurs in at least 34% of human genes and plays a major role in several endocrine systems (1). However, its role may still be underestimated, especially in light of the relative low number of genes found in the human genome that cannot account for the multitude of known human proteins. In addition, alternative splicing contributes to tissue and developmental stage-specific expression of protein isoforms. Besides yielding different proteins, alternative splicing in the 5'-noncoding region will not affect the gene product, but may influence gene expression and translation (3, 4).
The peptide hormone insulin is produced in the ß-cells of pancreatic islets and is responsible for maintaining glucose homeostasis by allowing glucose disposal in peripheral tissues. Glucose induces insulin biosynthesis mainly at the translational level (5), but the mechanisms are not well understood. Based on peptide sequence homology, insulin belongs to the insulin gene superfamily including IGF-I, IGF-II, and relaxin. Except for insulin, all of these genes have been shown to be alternatively spliced (6, 7, 8). The human (pre)proinsulin gene consists of three exons and two introns. The coding region starts in exon 2, so that exon 1 and part of exon 2 comprise the 5'-untranslated region (5'UTR) (9, 10). Both introns are flanked by canonical splice sites and encode the typical consensus sequences of GT-AG introns. However, analysis of the whole 5'UTR sequence revealed a cryptic 5'-splice site in intron 1. This raised the possibility that the human insulin gene may also be alternatively spliced. To test this hypothesis we studied isolated human pancreatic islets and found that indeed alternative splicing occurs at this newly described cryptic splice site. We further investigated how the expression of this insulin gene splice variant (SPV) was regulated metabolically and studied its function with regard to translation efficiency in vitro and in vivo. Our data suggest a novel mechanism that links metabolic stimuli to the expression of 5'UTR splice variants and the regulation of translation efficiency.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The native proinsulin gene (NAT) and the SPV were cloned from human islet cDNA using primers 1 and 3 or 2 and 3, respectively. The PCR products were directly subcloned under the T7 promoter into the pcDNA3.1/V5-His-TOPO vector (Life Technologies, Inc., Invitrogen, Carlsbad, CA), yielding the plasmids pcDNA-NAT and pcDNA-SPV, respectively. The 5'UTRs of the NAT and SPV were amplified with primers 10 and 11 using the plasmids pcDNA-NAT and pcDNA-SPV as templates. PCR products were cut with HindIII at the introduced restriction sites and subcloned under the simian virus 40 promoter upstream of the firefly luciferase reporter gene into the HindIII site of the pGL3 control vector (Promega Corp., Madison, WI), resulting in pGL3-natU and pGL3-spvU, respectively. The orientation and sequence of all constructs were verified by sequencing.
|
Quantitative RT-PCR
Total RNA was extracted from human islets with TRIzol reagent (Life Technologies, Inc.) according to the manufacturer’s instructions. RT was performed using the First Strand cDNA Synthesis Kit (Roche, Indianapolis, IN) after the samples were treated with deoxyribonuclease (DNase; Life Technologies, Inc.). The efficiency of the DNase treatment was confirmed by the lack of amplification in samples without reverse transcriptase. We thereby ruled out potential genomic contamination interfering particularly with the amplification of the intronic sequences. While DNase-treated samples were used for the results shown, similar results were also obtained when using non-DNase-treated samples. Quantitative real-time PCR was performed on a TaqMan apparatus (PE Applied Biosystems) using the 18S ribosomal subunit as an internal standard as recommended. The fold change in expression level was calculated using the formula 2-(
CT), where CT represents the cycle difference between the culture conditions corrected for 18S. Primers 4 and 5 detected only native proinsulin gene expression, whereas primers 7 and 8 amplified the SPV exclusively.
In vitro translation
The plasmids pcDNA-NAT and pcDNA-SPV were linearized using the restriction enzyme StuI, and the DNA was used in the RiboMAX RNA production system (Promega Corp.) according to the supplier’s protocol. Samples were treated with DNase, and the transcribed RNA was extracted and quantified with a spectrophotometer. Six micrograms of purified uncapped NAT or SPV RNA were used in a nonradioactive 25-µl wheat-germ translation reaction (Promega Corp.). The reaction was incubated for 1 h at 30 C. Five microliters of the reaction were run on a SDS-PAGE and analyzed by Western blotting using the mouse antihuman proinsulin antibody 3A1 (Advanced ImmunoChemical, Inc., CA). Bands were detected with the ECL system (Amersham Pharmacia Biotech, Arlington Heights, IL) and visualized and quantified on an Image Station 440 (Kodak, Rochester, NY).
Transfection studies
The human kidney cell line 293 was transiently transfected with 1 µg pGL3-natU or pGL3-spvU and 100 ng of the control plasmid pRL-TK (Promega Corp.) per 35-mm well using FuGENE 6 transfection reagent (Roche). The transfection was performed in serum-free, low glucose DMEM (Life Technologies, Inc.), but after 3 h the medium was changed to high glucose DMEM (Life Technologies, Inc.) supplemented with 10% heat-inactivated fetal calf serum. Cells were harvested 20 h after transfection, and luciferase activity was measured with the Dual-Luciferase Reporter Assay System (Promega Corp.). All experiments were performed in triplicate, and all results were corrected for transfection efficiency using the Renilla luciferase of the pRL-TK vector.
Prediction of secondary RNA structure
Michael Zucker’s algorithm and the m-fold program were used to compare the predicted secondary RNA structure of the 5'UTR of the NAT and SPV. Statistical analysis was performed using two-sided t tests.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Any changes involving intron 1 will alter the 5'UTR without disrupting the gene product. Intron 1 of the human proinsulin gene belongs to the major class of GT-AG introns, and as such, it has the typical 5'-splice site sequence AG/GT at the exon 1, intron 1 border (Fig. 1A). Careful analysis of the intron 1 sequence revealed a putative cryptic 5'-splice sequence present within the intron. Interestingly, we observed that this cryptic 5'-splice site is conserved in chimpanzee, monkey, and mouse (Fig. 1B). This suggested to us that the proinsulin gene may be alternatively spliced. To test this hypothesis, we performed RT-PCR on cDNA obtained from human pancreatic islet cells using the primers shown in Fig. 1A. We found an amplicon migrating at a slightly larger size compared with the native proinsulin gene, consistent with the predicted 26-bp size difference of the proposed SPV (Fig. 1C). To detect the SPV a specific 5'-primer had to be used because the SPV mRNA makes only approximately 1% of the very abundant NAT mRNA as calculated from quantitative real-time RT-PCR results (data not shown). As the 3'-primer was designed complementary to the 3'-end of the coding region, any larger SPVs, including the whole intron 1 (179 bp) or intron 2 (786 bp), would also have been amplified. The absence of these amplicons supports two conclusions. One, larger splice intermediates or precursors are very unlikely, and two, the apparent SPV is probably not secondary to genomic contamination of the RNA. To analyze whether the slightly larger SPV was due to alternative splicing at the cryptic 5'-splice site, RT-PCR products of both the native insulin gene (NAT) and the SPV were subcloned and sequenced. While sequence analysis of NAT showed a correctly spliced proinsulin gene, all three SPV clones had retained the 26 bp, confirming that splicing had occurred at the second AG/GT site (data not shown). RT-PCR performed on cDNA from different islet cell donors also demonstrated the presence of the alternatively spliced SPV message, excluding the possibility of an individual variation.
|
). Hyperglycemia for 1 h did not significantly alter either NAT or SPV expression (Fig. 2, A and D). Hyperglycemia for 24 h, however, increased SPV expression significantly by more than 2-fold, whereas NAT remained unchanged (Fig. 2, B and E). After 72 h of hyperglycemia, NAT expression was 2-fold higher compared with islets exposed to low glucose (Fig. 2C); however, SPV expression was increased more than 10-fold in response to glucose (Fig. 2F).
|
The translation efficiency of the proinsulin gene SPV is increased in vitro and in vivo
As inclusion of the 26-bp intronic sequence in the SPV alters the 5'UTR, and the 5'UTR often regulates translation, we wanted to compare the translation efficiencies of SPV and NAT. We first compared in vitro translation from the pcDNA-NAT and pcDNA-SPV plasmids using the combined transcription and translation wheat-germ assay. [Due to the absence of ubiquitination, the wheat-germ system is superior to the reticulocyte lysate system when small proteins less than 10 kDa (i.e. preproinsulin) are being produced.] For the same amount of plasmid used, the SPV yielded 2 times more (pre)proinsulin protein, as determined by Western blotting (data not shown). To exclude the possibility that the 26-bp intronic sequence affected transcription rather than translation, we in vitro transcribed NAT and SPV, treated the samples with DNase, and purified the RNA. This enabled us to add exactly the same amounts of NAT and SPV RNA to the in vitro translation reactions in the wheat-germ system. Again the SPV was more efficiently translated into (pre)proinsulin compared with NAT (Fig. 3, A and B).
|
). A similar trend with much smaller, but significant, effects was observed in the hamster insulinoma cell line HIT-T15 (data not shown). These results suggest that species-specific differences may play an important role in the regulation of translation efficiency.
The secondary RNA structure of the proinsulin gene SPV 5'UTR is dramatically altered by inclusion of the 26-bp intronic sequence
As in vitro wheat-germ extract does not contain any mammalian proteins, which would uniquely recognize the proinsulin mRNA, the improved translation efficiency conferred by the 26-bp intronic sequence was not likely due to trans-acting factors specifically binding to the proinsulin mRNA. We therefore asked whether the 26-bp sequence was directly responsible for the increased translation efficiency, potentially by altering the 5'UTR secondary structure. Interestingly, analysis of the secondary structure of the 5'UTR of NAT and SPV using m-fold (12, 13) showed that inclusion of the 26-bp sequence resulted in a dramatic change in the lowest energy predicted structure for the SPV 5'UTR. This suggests that the higher translation efficiency of the SPV may be directly correlated with the structural changes induced by the presence of the 26-bp sequence (Fig. 4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In addition, we found that the critical sequence involved in this alternative splicing pathway is evolutionary conserved (Fig. 1B), suggesting that this regulatory mechanism may also be conserved. Furthermore, our findings demonstrate that SPV expression is regulated by glucose, the major stimulus for insulin production and secretion. The SPV, in turn, is translated more efficiently than NAT into native (pre)proinsulin protein. This is consistent with the idea that glucose-induced insulin biosynthesis is regulated at the translational level (14). This correlation suggests that increased SPV expression can add to glucose-induced insulin production. Although it is unlikely that the SPV can account for the entire glucose-induced effect on insulin biosynthesis based on the low expression level found under the restricted conditions of our experiments, it may still contribute to 10% or more of the produced insulin, given the dramatic increase in translation in our in vivo studies. In addition, we cannot exclude that SPV preferentially is exported out of the nucleus or is more efficiently bound by ribosomes to render it more readily available for translation than NAT. It is also possible that the expression of SPV is much higher during different stages of development or disease. The time-dependent increase in SPV expression observed in human islets exposed to high glucose supports this idea and suggests that chronic hyperglycemia, as found in patients with poorly controlled diabetes, may modulate the expression of SPV in an attempt to maintain insulin production. We are planning experiments to test these possibilities. The fact that the cryptic 5'-splice site used in the SPV is conserved in the proinsulin gene of several species suggests that it has universal physiological importance, as nonfunctional intronic sequences are typically not maintained (1). Perhaps more interesting than the relevance of these findings in regard to the particular effects of glucose on insulin production are the unusual mechanisms involved. Alternative splicing is known to be regulated according to developmental stage, in a tissue-specific manner, or by different signaling pathways, including cytokines, growth factors, and insulin (1). This was one of the reasons to test the effects of insulin alone on the expression of SPV. However, we did not detect any insulin-dependent changes (data not shown), so we conclude that the observed effects were primarily glucose mediated. Hyperglycemia induced the expression level of the proinsulin gene SPV, which could be due to glucose increasing the SPV mRNA stability. However, although glucose has been shown to enhance insulin gene message stability (10, 15), recent studies have demonstrated that this is achieved via binding of the polypyrimidine tract-binding protein to the 3'UTR (15). Such a mechanism is unlikely in our case, as the glucose effects would have to be mediated via the retained 26 bp of the 5'UTR to enhance SPV message stability selectively. Alternatively, glucose may indirectly affect the splicing process and lead to preferential use of the cryptic splice site. Whether glucose needs to be metabolized via mitochondrial oxidation or via the hexosamine or pentose phosphate pathway to achieve the described effects remains to be determined. Testing glucose metabolites or other metabolic stimuli, such as FFA, may further broaden our understanding of the metabolic regulation of alternative splicing. Interestingly, the generated SPV is translated more efficiently into protein. Recent studies have shown that glucose can regulate the translation of target genes, e.g. CD36 (16), but in our study glucose affected translation efficiency via induction of an alternative splicing product. As translation control of glucose-induced insulin biosynthesis occurs within minutes in the absence of transcription (5), it cannot be attributed to an acute increase in the expression of SPVs. However, over longer time frames glucose enhances the expression of the SPV and may thereby augment translation efficiency. This novel observation suggests that metabolic stimuli can regulate the expression of 5'UTR SPV, which have distinct translation efficiencies and lead to altered protein synthesis from target genes.
Major types of alternative splicing include inclusion or exclusion of entire exons or introns. The usage of an alternative splice site, as described in the present study, is not as common (1), particularly the use of an intronic cryptic 5'-splice site. The one example of intronic splicing found in the human double-stranded RNA-dependent protein kinase was due to alternate usage of 3'-splice sites (4).
In general, most studies focus on alternative splicing affecting the open reading frame and thereby the gene product. Only a few examples of alternative splicing in the noncoding region have been described (3, 4, 17, 18). Interestingly, they all affect translation efficiency. In the case of the GLI1 oncogene found in human basal cell carcinomas, alternative splicing eliminates three upstream ATG codons and thereby increases translation (18). In some squamous cell carcinomas an intron containing eight upstream ATGs is retained in the HYAL1 tumor suppressor gene mRNA, which blocks its translation (17). An upstream open reading frame (uORF) was detected in the 5'UTR of RNA-dependent protein kinase, but was found not to affect translation efficiency (4). Nevertheless, one alternatively spliced form was less efficiently translated compared with the other two.
Mammalian 5'UTR sequences that regulate translation are typically several hundreds base pairs long, may have complicated structures, and may contain uORFs (19). The 5'UTR of the CD36 gene, the translation of which is regulated by glucose, follows this pattern. In contrast, the insulin gene 5'UTR spans only 59 bp, and no uATGs were found in either the native 5'UTR or the 26-bp intronic sequence, making regulation via an uORF very unlikely. Specific binding of RNA-binding proteins to the 26 bp does not appear to be the mechanism behind the observed translational regulation, considering our in vitro findings of increased SPV translation (relative to NAT) in the absence of any mammalian proteins in the wheat-germ extract. The most likely remaining mechanism is therefore the 26-bp RNA sequence and the induced structural alterations. Indeed, comparison of the predicted secondary RNA structure of the NAT and SPV 5'UTR revealed a dramatic structural change caused by retention of the 26-bp intronic sequence. This altered 5'UTR structure may lead to increased mRNA stability, induced nuclear export, or preferential binding of ribosomes and result in enhanced translation.
Alternative splicing of the 5'UTR of the proinsulin gene seems to be an appropriate adaptation to a metabolic stimulus, glucose, leading to increased translation of the hormone, insulin, necessary to correct the hyperglycemia. This raises the possibility that this type of mechanism may be used by other pathways and constitutes an additional way to regulate protein synthesis in response to the changing metabolic needs (Fig. 5). It therefore might represent a physiological mechanism that should be considered when looking at regulated translation or alternative splicing. These findings underline the importance of including the noncoding region and intronic sequences in any type of splicing analysis, especially in genes with similar structures containing introns in the 5'UTR of the gene pre-mRNA.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received February 21, 2002.
Accepted for publication March 29, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. Evans-Molina, J. C. Garmey, R. Ketchum, K. L. Brayman, S. Deng, and R. G. Mirmira Glucose Regulation of Insulin Gene Transcription and Pre-mRNA Processing in Human Islets Diabetes, March 1, 2007; 56(3): 827 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Marcucci, F. E. Baralle, and M. Romano Complex splicing control of the human Thrombopoietin gene by intronic G runs Nucleic Acids Res., January 12, 2007; 35(1): 132 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. Nackley, S. A. Shabalina, I. E. Tchivileva, K. Satterfield, O. Korchynskyi, S. S. Makarov, W. Maixner, and L. Diatchenko Human Catechol-O-Methyltransferase Haplotypes Modulate Protein Expression by Altering mRNA Secondary Structure Science, December 22, 2006; 314(5807): 1930 - 1933. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Narendran, A. M. Neale, B. H. Lee, K. Ngui, R. J. Steptoe, G. Morahan, O. Madsen, J. A. Dromey, K. P. Jensen, and L. C. Harrison Proinsulin is encoded by an RNA splice variant in human blood myeloid cells PNAS, October 31, 2006; 103(44): 16430 - 16435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kralovicova, T. R. Gaunt, S. Rodriguez, P. J. Wood, I. N.M. Day, and I. Vorechovsky Variants in the Human Insulin Gene That Affect Pre-mRNA Splicing: Is -23HphI a Functional Single Nucleotide Polymorphism at IDDM2? Diabetes, January 1, 2006; 55(1): 260 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Wang, X. Guo, and J. Floros Differences in the translation efficiency and mRNA stability mediated by 5'-UTR splice variants of human SP-A1 and SP-A2 genes Am J Physiol Lung Cell Mol Physiol, September 1, 2005; 289(3): L497 - L508. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. H. Minn, C. Hafele, and A. Shalev Thioredoxin-Interacting Protein Is Stimulated by Glucose through a Carbohydrate Response Element and Induces {beta}-Cell Apoptosis Endocrinology, May 1, 2005; 146(5): 2397 - 2405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. H. Minn, H. Lan, M. E. Rabaglia, D. M. Harlan, B. A. Peculis, A. D. Attie, and A. Shalev Increased Insulin Translation from an Insulin Splice-Variant Overexpressed in Diabetes, Obesity, and Insulin Resistance Mol. Endocrinol., March 1, 2005; 19(3): 794 - 803. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) were assigned the value 1. 
, Fold change in expression at high glucose compared with low glucose for NAT (A–C) and SPV (D–F) after 1, 24, and 72 h incubation at 37 C. Islets from three donors were analyzed. Due to interindividual differences, comparisons were only made within the same donor. All experiments were performed in triplicate. The mean ± SD of one representative experiment are shown.
