The protein kinase C beta II exon confers mRNA instability in the presence of high glucose concentrations.

Previous studies showed that short term exposure of cells to high glucose destabilized protein kinase C (PKC) betaII mRNA, whereas PKCbetaI mRNA levels remained unaltered. Because PKCbeta mRNAs share common sequences other than the PKCbetaII exon encoding a different carboxyl terminus, we examined PKCbetaII mRNA for a cis-acting region that could confer glucose-induced destabilization. A beta-globin/growth hormone reporter con struct containing the PKCbetaII exon was transfected into human aorta and rat vascular smooth muscle cells (A10) to follow glucose-induced destabilization. Glucose (25 mm) exposure destabilized PKCbetaII chimeric mRNA but not control mRNA. Deletion analysis and electrophoretic mobility shift assays followed by UV cross-linking experiments demonstrated that a region introduced by inclusion of the betaII exon was required to confer destabilization. Although a cis-acting element mapped to 38 nucleotides within the betaII exon was necessary to bestow destabilization, it was not sufficient by itself to confer complete mRNA destabilization. Yet, in intact cells antisense oligonucleotides complementary to this region blocked glucose-induced destabilization. These results suggest that this region must function in context with other sequence elements created by exon inclusion involved in affecting mRNA stability. In summary, inclusion of an exon that encodes PKCbetaII mRNA introduces a cis-acting region that confers destabilization to the mRNA in response to glucose.

Protein kinase C (PKC), 1 a serine/threonine kinase, comprises a family of 12 isozymes that have been implicated in signaling pathways affecting cellular processes such as cell proliferation and differentiation, apoptosis, tumor promotion, transcriptional activation, and hormone production (1). The PKC isozymes exhibit differential cellular distribution and substrate specificity. The conventional PKC isozymes, which are Ca 2ϩ -dependent and activated by phospholipid and diacylglycerol, include PKC␣, PKC␤I, PKC␤II, and PKC␥. PKC␤I and PKC␤II are encoded by the same gene, and are translated from alternatively spliced products of PKC␤ pre-mRNA. The inclusion of the PKC␤II exon in the 3Ј-region through alternative splicing results in the PKC␤II mRNA. This pattern of splicing generates a stop codon at the ␤II-␤I boundary such that the ␤I exon, with its coding sequence and 3Ј-UTR, now becomes the 3Ј-UTR of PKC␤II mRNA (Fig. 1). As a result, the PKC␤I and PKC␤II mRNAs differ only by the sequence of the included PKC␤II exon, and the proteins differ only by their carboxylterminal 50 -52 amino acids, respectively (1).
We have previously shown that acute hyperglycemia downregulates PKC␤II, but not PKC␤I, at the mRNA and protein levels in vascular smooth muscle cells (2). To understand the mechanism by which elevated glucose down-regulates PKC␤II expression in vascular smooth muscle cells, earlier studies were carried out to determine at what level suppression of PKC␤II expression occurred. These studies clearly demonstrated that high glucose (10 -30 mM), at levels that commonly occur in hyperglycemia associated with diabetes mellitus, exerted some regulation at the level of transcription, but that the most dominant level of regulation occurred post-transcriptionally through increased destabilization of PKC␤II mRNA via a cytoplasmic nuclease activity (2). PKC␤I and other isozyme-specific PKC mRNAs were not destabilized. The effect of elevated glucose (25 mM) was independent of osmotic controls, because mannitol did not down-regulate PKC␤II mRNA expression (2,3). We have also shown that PKC␤I and PKC␤II enzymes have opposite signaling roles in cell division, where PKC␤II signaling suppresses vascular smooth muscle cell proliferation by attenuating G 1 /S transition and high glucose treatment, which down-regulates PKC␤II mRNA and protein, and stimulates vascular smooth muscle cell proliferation (4,5). High glucose also suppressed insulin effects on glucose uptake, another PKC␤II-dependent process (3).
The regulation of mRNA stability has emerged as an important mechanism for controlling gene expression. Depending on the system, half-lives of mRNA range from a few minutes to days. The decay rates of many eukaryotic mRNAs are regulated by developmental or environmental stimuli. Most of the mechanisms that control mRNA stability share common features, and determinants of mRNA stability have been shown to reside in the 5Ј-cap, 5Ј-UTR, coding region, poly(A)-tail, and the 3Ј-UTR (or the AU-rich elements) such that each may play some role in regulating mRNA decay rates (6 -11).
Although it is now clear that the decreased stability of PKC␤II mRNA correlates with cell exposure to high concentrations of glucose, the molecular mechanism contributing to the regulation of PKC␤II expression by altering its mRNA stability remained to be defined. Here, using deletion and competition analyses, we identify a ϳ38-nucleotide sequence within a region of secondary structure in the alternatively spliced exon that binds factors in vitro in a manner dependent on the sequence and the availability of factors present only in the cytosol of glucose-treated cell extracts. The degree to which protein-RNA complex was formed was dependent on the presence of the 38-nt region. These data suggest that insertion of the ␤II exon specifies glucose responsive destabilization of PKC␤II mRNA. Because PKC␤II and PKC␤I mRNAs differ only by the inclusion of the PKC␤II exon through alternative splicing, we propose that a cis-acting region introduced by inclusion of the PKC␤II exon defines glucose-mediated mRNA destabilization. To our knowledge this is the first characterization of an element that is introduced by alternative splicing that allows for metabolite regulation of stability of a specific mRNA.

EXPERIMENTAL PROCEDURES
Cell Culture-The vascular smooth muscle cell line (A10, ATCC CRL 1476), derived from rat aorta, was grown in Dulbecco's modified Eagle's medium (with 5.5 mM glucose) containing 10% fetal bovine serum, 100 units penicillin G, and 100 g of streptomycin sulfate/ml, at 37°C in a humidified 5% CO 2 , 95% air atmosphere in either 6-well or 100-mm plates. Cells were grown to Ͼ90% confluency and medium was changed every 4 days. Cell synchronization was achieved by serum deprivation (0.5% fetal bovine serum) for 48 h as demonstrated previously (12). Primary cultured human aortic smooth muscle cells (Clonetics, San Diego, CA) were grown in smooth muscle growth medium (Clonetics) containing 5.5 mM glucose, 5% fetal bovine serum, 10 ng/ml human recombinant epidermal growth factor, 390 ng/ml dexamethasone, 50 g/ml gentamicin, and 50 ng/ml amphotericin-B at 37°C in a humidified 5% CO 2 , 95% air atmosphere. Cells were grown to Ͼ90% confluency and medium was changed every 5 days.
Lipofectin Transfections-A10 cells were grown and synchronized by serum deprivation for 48 h in either 6-well or 100-mm plates. The p␤G or p␤G-PKC␤II expression vectors were transfected in serumfree medium with the Lipofectin-DNA complex. Lipofectin reagent was purchased from Invitrogen. Following 4 h of incubation, the Lipofectin-DNA complex was washed off with 1ϫ Dulbecco's phosphate-buffered solution and replaced with fresh medium containing 2% serum. For stably transfected A10 cell selection, 0.7 mg/ml G-418 was added to the media. It was changed every 4 days, and 10 -14 days later the colonies were pooled and maintained in Dulbecco's modified Eagle's medium (10% fetal bovine serum) with 0.2 mg/ml G-418.
mRNA Half-life Determination-p␤G-PKC␤II stable transfectants and p␤G stable transfected A10 cells were plated into 100-mm dishes. 50 g/ml DRB (5,6-dichloro-1-␤-D-ribofuranosylbenzimidazole) dissolved in 95% ethanol was added to the plates 30 min prior to the treatments and the 0-h RNA sample was isolated. 25 mM glucose was added to the glucose-treated plates. RNA samples were isolated from normal (5.5 mM glucose) and glucose-treated (25 mM glucose) p␤G-PKC␤II and p␤G dishes at 2, 4, 6, 10, and 18 h. In a separate control, an equivalent amount of 95% ethanol was added.
Isolation of RNA and Northern Blot Analysis-Total cellular RNA was isolated from 100-mm plates using Tri-Reagent (Molecular Research Center, Inc.). RNA samples (10 g) were prepared in formamide, formaldehyde, and 1ϫ MOPS, and fractionated on 1.2% agarose-formaldehyde gels. Ethidium bromide was added in the loading buffer for visualization and quantitation of 18 S and 28 S RNA. After fractionation, the integrity and loading of RNA was assessed under UV light (12)(13)(14). The size-fractionated RNA was then capillary transferred to Hybond membranes (Amersham Biosciences), and cross-linked to membranes by baking at 80°C in a vacuum oven for 2 h. Membranes were hybridized overnight at 42°C with 2 ϫ 10 7 cpm of the ␤-globin probe (labeled with [␣-32 P]dCTP by nick translation as described (16)) per ml of hybridization buffer. Membranes were washed with high stringency conditions; label was detected and quantitated using a Amersham Biosciences PhosphorImaging system.
Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)-Total RNA was isolated from control or glucose-treated A10 cells or aortic smooth muscle cells and 2 g was used to synthesize first strand cDNA using an oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen pre-amplification kit). The upstream sense primer for ␤-globin was (5Ј-GCATCTGTCCAGTGAGGAGAA-3Ј) and the downstream antisense primer was (5Ј-AACCAGCACGTTGC-CCAGGAG-3Ј). Sense and antisense primers for ␤-actin (number 5402-3) were obtained from Clontech. To detect PKC␤I and -␤II, the upstream sense primer corresponded to the C4 kinase domain common to both PKC␤I and PKC␤II (5Ј-CGTATATGCGGCCGCGTTGT-GGGCCTGAAGGGG-3Ј) and the downstream antisense primer was specific for PKC␤I (5Ј-GCATTCTAGTCGACAAGAGTTTGTCAGT-GGGAG-3Ј) (16). PCR was performed using platinum Taq DNA polymerase (from Invitrogen) on 10% of the reverse transcriptase reaction product. Following amplification in a Biometra Trioblock thermocycler (␤-globin: 94°C, 1 min; 58°C, 1 min; and 72°C, 3 min for 30 cycles; PKC␤I and -␤II: 95°C, 30 s; 68°C, 2 min for 35 cycles), 20% of the amplified products were resolved on a 1.2% agarose gel or 5% of products were resolved on 6% PAGE gels and detected by silver staining. The PCR reaction was optimized for linear range amplification to allow for quantification of products.
cDNA Probe Preparation-The pRSV-␤G vector (obtained from Dr. Norman C. Curthoys, Colorado State University) was digested with HindIII and BglII to obtain a 507-bp fragment, which was isolated from agarose gel using the QIAquick gel extraction kit (Qiagen). The ␤-globin probe was labeled with [␣-32 P]dCTP using Prime-a-Gene (Promega) to a specific activity of 10 9 disintegrations/min.
Synthesis of p␤G-PKC␤II Chimeric Minigene-The parent vector p␤G contains the strong viral promoter derived from the long terminal repeat of the Rous sarcoma virus followed by the transcriptional start site, the 5Ј-nontranslated region, the entire coding sequence, and two introns from the rabbit ␤-globin gene, a multicloning site containing four unique restriction sites, and the 3Ј-nontranslated region and polyadenylation site of the bovine growth hormone (17). It carries neomycin and ampicillin resistance. The 404-bp PKC␤II product corresponding to the 216-bp ␤II exon and flanking regions was obtained by PCR amplification using sense primer to the upstream PKC␤ common C4 domain containing the SpeI site (5Ј-CGTATATA-CTAGTGTTGTGGGCCTGAAGGGGAACG-3Ј) and antisense primer to the ␤IV5 domain containing the XbaI site (5Ј-TGCCTGGTGAAC-TCTTTGTCGAGAAGCTCT-3Ј) such that the exon-included PKC␤II mRNA was amplified. The insert contained 70 bp of the C4 exon and 118 bp of the ␤I exon common to both PKC␤I and PKC␤II in addition to the 216-bp exon specific for PKC␤II. After size fractionation it was Splicing of the last common domain C4 to PKC␤IV5 domain results in mature PKC␤I mRNA whereas inclusion of the PKC␤II-specific exon results in PKC␤II mRNA. A STOP codon is generated when ␤I exon splices onto the ␤II exon to produce mature PKC␤II mRNA in which the ␤I exon serves as an extended 3Ј-UTR. extracted from the gel (Qiagen QIAquick gel extraction kit), digested with SpeI and XbaI, and purified. The PKC␤II cDNA was ligated into the SpeI and XbaI sites of the multicloning region of p␤G vector. The construct, p␤G-PKC␤II, was verified by restriction mapping and dideoxynucleotide sequencing.
Using PCR primers for C4 (last common domain) as sense primer with the SpeI site synthesized upstream (5Ј-CGTATATACTAGTGTT-GTGGGCCTGAAGGGGAACG-3Ј) and for ␤II exon antisense primer with the XbaI site synthesized downstream (5Ј-CGGAGGTCTACACA-TCTACTTTCTAGAAGCTCT-3Ј), PKC␤II exon without the -␤I exon (286 bp product) was amplified. This insert contained 70 bp corresponding to the C4 exon and the entire PKC␤II exon. The product was gel purified and ligated into the SpeI and XbaI sites of the multicloning region of p␤G vector. The construct, p␤G-PKC␤II⌬␤I, was verified by restriction mapping.
The region of instability comprising 38 nucleotides was synthesized with the SpeI and XbaI restriction enzyme sites to facilitate cloning into the p␤G vector: 5Ј-AACTCTACTAGTGAATTTTTAAAACCCGAAGTC-AAGAGCTCTAGATAGTA-3Ј. The construct, p␤G-PKC␤II38, was verified by restriction mapping.
As a control, the analogous carboxyl-terminal region of PKC␦ (C4-V5/␦) was digested using BstXI and XbaI (450 bp), purified, and ligated into the multicloning region of the ␤G vector. The construct, p␤G-PKC␦, was verified by restriction mapping.
In Vitro Transcription Vectors-The 404-bp PKC␤II product corresponding to the 216-bp ␤II exon and flanking regions described above was obtained by PCR amplification using sense primer to the upstream PKC␤ common C4 domain (5Ј-CGTATATGCGGCCGCGTTGT-GGGCCTGAAGGGG-3Ј) and antisense primer to ␤I exon (5Ј-GCAT-TCTAGTCGACAAGAGTTTGTCAGTGGGAG-3Ј) such that the exonincluded PKC␤II mRNA was amplified. This PKC␤II cDNA piece was cloned into the pCR-Blunt vector (Invitrogen) such that transcripts could be generated from the upstream T7 RNA polymerase promoter.
In Vitro Transcript Preparation-The RNA probes were generated by consecutive restriction digestion of the pCR-Blunt-PKC␤II vector. Riboprobe A (RpA) was the full-length PKC␤II insert described above, linearized with BamHI; riboprobe B (RpB) was the PKC␤II insert linearized at 175 bp with BglII within the PKC␤II exon such that the PKC␤I-specific exon was eliminated; riboprobe C (RpC) was linearized at 137 bp with HpaI, which cut within the PKC␤II-specific exon; riboprobe D (RpD) was linearized at 102 bp with SspI, which cut within the PKC␤II-specific exon. After digestion, the riboprobes were purified and their sizes and linearity were confirmed following size fractionation on agarose gels. One g of each linearized plasmid DNA was further used for in vitro transcription with the Ampliscribe kit (Epicentre) for competitor unlabeled probes or with Riboscribe kit (Epicentre) for transcribing labeled RNA probes using T7 RNA polymerase at 37°C for 2 h in the presence of nucleotides, RNase inhibitor, and buffer according to the manufacturer's instructions. The nonspecific unlabeled probes were from analogous fragments of PKC␦ (described above), PKC␤I (15), and ␤-actin, and were prepared in a similar manner. For labeled RNA probes, 5 l of [␣-32 P]CTP (3000 Ci/mmol) was used in the reaction and 1 l of RNase-free DNase I was then added and incubated for 15 min at 37°C to remove the template DNA. The transcripts were precipitated using 5 M ammonium acetate and incubated for 15 min on ice followed by a 70% ethanol wash. The pellet was resuspended in RNase-free water. The integrity of RNA probes was confirmed and probes were purified by 6% native polyacrylamide gel electrophoresis.
Cytoplasmic Extract (S100) Preparation-In brief, A10 cells incubated in 5.5 mM glucose (low) or 25 mM glucose (high glucose) for 4 h were washed with 1ϫ phosphate-buffered saline, gently scraped, collected, and centrifuged at 1850 ϫ g for 10 min. The packed cells were re-suspended in hypotonic buffer (10 mM HEPES, pH 7.9, at 4°C, 1.5 mM MgCl 2 , 10 mM KCl, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol, 10 M leupeptin, 10 M antipain), allowed to swell for 10 min on ice, and homogenized in a Dounce homogenizer with 60 strokes using type B pestle. The nuclei were pelleted by centrifuging at 3300 ϫ g for 15 min, and the cytoplasmic extract was mixed with 0.11 volume of cytoplasmic buffer (0.3 M HEPES, pH 7.9 at 4°C, 1.4 M KCl, 0.03 M MgCl 2 ). After centrifugation for 1 h in a Beckman type 50 rotor at 40,000 rpm, the supernatant (S100) was aliquoted, frozen in liquid nitrogen, and stored at Ϫ80°C. An aliquot was used to determine the protein concentrations using the Bradford protein assay (18).
RNA Electrophoretic Mobility Shift Assay-S100 extracts from A10 cells exposed to low (5.5 mM) glucose and high (25 mM) glucose concen-trations containing 3 g of protein were incubated with 3 g of yeast tRNA and 10 units of RNase inhibitor in a final volume of 10 l of RNA shift buffer (12 mM HEPES, pH 7.9, 10 mM KCl, 10% glycerol, 5 mM EDTA, 5 mM dithiothreitol, 5 mM MgCl 2 ) for 10 min at room temperature. 100 -300-Fold excess specific cold competitors or 100-fold excess nonspecific cold competitors quantified spectrophotometrically using a Pharmacia Gene Quant were added to the binding reactions and incubated for 5 min at room temperature. Labeled RNA probes RpA, RpB, RpC, or RpD (described above) to an activity of ϳ1 ϫ 10 6 cpm were added and incubated for 20 min at room temperature. RNase T1 was added to the binding reaction to digest the unbound RNA. Because the cytoplasmic extracts may contain proteins that bind to negatively charged polyanions, like nucleic acids, heparin was added to the reaction to suppress nonspecific binding (19,20). 100 Units of RNase T1 was added and further incubated for 15 min at room temperature, followed by the addition of 5 mg/ml heparin to reduce nonspecific binding, for 10 min on ice. Samples were separated on a 10% polyacrylamide gel in 0.5ϫ TBE buffer. Gels were dried and exposed to Amersham Biosciences PhosphorImaging screen.
UV Cross-linking of RNA-Protein Complexes-RNA-protein binding reactions were carried out as described above for the RNA electrophoretic mobility shift assay. To demonstrate specificity, the 38-nucleotide sequence (50 nM) described above, or 50 nM of a 20-nucleotide antisense 2Ј-O-(2-methoxy)ethyl (MOE) oligonucleotide (AS 25647) targeting a portion of the 38-nucleotide region, 5Ј-CTTGACTTCGGGT-TTTAAAA-3Ј, were added to binding reactions. As a control, another 20-nucleotide antisense MOE oligonucleotide (AS 25649), 5Ј-GAAGTT-GGAGGTGTCTCGCT-3Ј, upstream of this region or a scrambled 20mer MOE oligonucleotide was used. Following heparin addition, the samples were transferred to a 96-well plate, and irradiated for 10 min in Stratalinker (Stratagene) on ice. Laemmli's buffer was added to the sample, which was then boiled for 5 min and separated on a 10% SDS-polyacrylamide gel. Gels were dried and exposed to Amersham Biosciences PhosphorImaging screens. To control for nonspecific protein/RNA interactions, 1 g of micrococcal nuclease (Sigma) was added to control and glucose-treated cell extracts with 70 mM EDTA in addition to the labeled probe. There was no difference in the UV crosslinking in the presence of micrococcal nuclease to either control or high glucose-treated cytosolic extracts.

Effects of the PKC␤II Exon Insertion on the Stability of ␤-Globin
Reporter mRNA-Previous experiments had suggested that insertion of the 3Ј-PKC␤II exon via alternative splicing introduced a glucose-responsive instability region that was not present in the PKC␤I mRNA (2). To test this, a corresponding region consisting of 70 nucleotides in the last domain common to both PKC␤I and PKC␤II (C4), the PKC␤II-specific exon (V5) (216 nucleotides), and 119 nucleotides of the flanking ␤I exon were recloned into an expression vector (p␤G) to form a vector expressing ␤-globin-PKC␤II mRNA (see Fig. 2, a and b). The resulting chimeric vector, p␤G-PKC␤II was under the regulation of a viral promoter and contained a polyadenylation sequence, eliminating regulatory effects on transcription and nuclear decay. The p␤G-PKC␤II vector was transiently transfected into A10 cells, a rat clonal vascular smooth muscle cell line, and also into human aorta smooth muscle cells. Aorta smooth muscle cells respond to glucose-induced destabilization of PKC␤II mRNA in an analogous manner to A10 cells and provide a primary cell model for corroborating glucose effects in human cells (2). As demonstrated by RT-PCR analysis (Fig. 2c), acute exposure (2 h) to high glucose (25 mM) resulted in a decrease (Ͼ80%) in the chimeric ␤-globin-PKC␤II mRNA, when compared with the same cells exposed to control levels of low glucose (5.5 mM). In A10 cells transfected with the parent vector, p␤G, ␤-globin mRNA levels remained unaltered after exposure to high glucose concentrations. Levels of ␤-actin mRNA remained unchanged in low glucose and high glucosetreated p␤G-PKC␤II and p␤G-transfected cells.
To further rule out the possibility of nonspecific effects caused by inserting the fragment of PKC␤II cDNA into the p␤G vector, an analogous fragment corresponding to the exon for the COOH-terminal domain of PKC␦ (C4-V5/␦) was subcloned into the parent p␤G vector creating a chimeric reporter expressing ␤-globin-PKC␦ mRNA (Fig. 3a). This chimeric plasmid was transiently transfected into A10 cells. As shown by RT-PCR analysis (Fig. 3b), exposure of cells to high glucose had no effect on the stability of ␤-globin-PKC␦ mRNA.
The Effect of High Glucose on the Half-life of ␤-Globin-PKC␤II mRNA-Half-life determinations of ␤-globin-PKC␤II mRNA were carried out in the presence of the transcriptional inhibitor, DRB. DRB was used rather than actinomycin D, because actinomycin D, commonly used to inhibit RNA polymerase II, has also been reported to inhibit translation (21), and in some instances to also inhibit mRNA degradation (22,23). Stable A10 cell transfectants of p␤G-PKC␤II or p␤G were pretreated with DRB, and then incubated with low (5.5 mM) or high (25 mM) glucose for various times up to 6 h. Northern blot analysis of total RNA showed that only 20% of the ␤-globin-PKC␤II mRNA remained after exposure to high (25 mM) glucose concentrations, compared with ␤-globin-PKC␤II mRNA in the presence of control (5.5 mM) glucose concentrations within 2 h (Fig. 4). This is in agreement with the results presented in Fig. 2. High glucose concentrations reduced the half-life of the ␤-globin-PKC␤II mRNA to 45 min, effectively decreasing the amount of the mRNA.
The Effect of Glucose-induced Destabilization by the ␤II Exon in the Absence of the ␤I Exon-Inclusion of the PKC␤II-specific exon in many cell types results in mature PKC␤II mRNA in which PKC␤I exon functions as an extended 3Ј-UTR (Fig. 1). Exon inclusion produces new C4-␤II and ␤II-␤I exon junctions in the PKC␤II mRNA. Hence, to examine the influence of glucose on mRNA stability that may be mediated by insertion of the ␤II exon, a new construct without the 3Ј-␤I exon was examined. The ␤-globin-PKC␤II⌬␤I vector was transiently transfected into A10 cells, and cells were exposed to high or low glucose (control). As shown (Fig. 5), high glucose also destabilized the ␤-globin-PKC␤II⌬␤I mRNA. The extent of destabilization was less than that observed for the ␤-globin-PKC␤II chimeric mRNA shown in Fig. 2c. Approximately 50% of the ␤-globin-PKC␤II⌬␤I mRNA was degraded compared with Ͼ80% degradation of the ␤-globin-PKC␤II chimeric mRNA in cells treated with high glucose. These results suggested that inclusion of the ␤II exon was necessary to confer full glucosesensitive mRNA degradation, but the C4-␤II exon junction alone was not sufficient to produce the same level of glucoseinduced destabilization as observed for the complete 3Ј PKC␤II mRNA sequence, including both the C4-␤II and ␤II-␤I exon junctions as shown in Fig. 2c.

Cytoplasmic Extracts from A10 Cells Treated with High Glucose Concentrations Specifically Retard a Labeled PKC␤II Exon
Transcript-Initial studies demonstrated that a nuclease activity present in the cytosolic extracts from glucose-treated cells mediated PKC␤II mRNA destabilization (2). To further investigate regions important for destabilization within the PKC␤II exon, cytoplasmic extracts from A10 cells were exam- , and 118 bp of the ␤I exon, subcloned into the parent vector p␤G at the SpeI and XbaI sites. The p␤G (obtained from N. P. Curthoys, Colorado University) is a chimeric gene containing a strong viral promoter derived from the long terminal repeat of the Rous sarcoma virus followed by the transcriptional start site, the 5Ј-nontranslated region, the entire coding sequence, and two introns from the rabbit ␤-globin genomic DNA; a multicloning site containing four unique restriction sites into which 404 bp from COOH-terminal PKC␤II cDNA was subcloned; and the 3Ј-untranslated region and polyadenylation site of the bovine growth hormone (GH-poly(A)). c, total RNA was extracted from human aorta smooth muscle cells (AoSMC) transiently transfected with p␤G-PKC␤II chimeric or a p␤G empty vector stability reporter system and exposed to either normal (5.5 mM) or high (25 mM) glucose for 2 h. 2 g of RNA was used in the RT-PCR analysis using primers for ␤-globin. A, 100-bp DNA marker; B, normal glucose; C, high glucose. The experiment was repeated five times and similar results were obtained.
ined for transacting components that specifically interact with a glucose-sensitive element within the ␤II exon using electrophoretic mobility shift assays. As shown in Fig. 6a, only the extract from A10 cells exposed to high glucose retained the labeled probe, suggesting a large protein-RNA complex was forming in these extracts. No complex formation was observed with the control (5.5 mM glucose) cytoplasmic extracts. To further demonstrate specificity of the complex, excess cold competitor RNA (corresponding unlabeled transcript) was added to the incubation and shown to eliminate binding whereas a nonspecific RNA competitor (an analogous region of PKC␦) did not (Fig. 6a). The bound probe observed here may represent cooperative assembly of large complexes as a result of protein-RNA interactions on this element. It is also noteworthy that the complexes are observed only in glucose-treated cytosolic extracts.
To identify the boundaries of sequence contributed by the PKC␤II exon required for retention of probe, deletion constructs were generated as shown in Fig. 6e. Probes RpB (BglII digestion), RpC (HpaI digestion), and RpD (SspI digestion) were transcribed and analyzed with cytoplasmic extracts from control (5.5 mM glucose) and glucose (25 mM)-treated cells using RNA gel shift analysis. Excess unlabeled RNA competitors and nonspecific competitors were added for each transcript as shown in Fig. 6b. Shifted material was observed for transcript RpB, where the restriction deletion removed all the sequence(s) encoding the PKC␤I-specific exon. Complexes were also noted as with RpA, but the intensity of the interaction was diminished with RpB. No complexes were observed using transcripts RpC or RpD, both transcripts resulting from deletion of coding sequence within the PKC␤II-specific exon (Fig. 6, c and d).
From the deletion analysis, a region of 38 nucleotides positioned between the HpaI and BglII sites, within the PKC␤II exon, was shown to form complexes with cytosolic extracts from glucose-treated cells. To further confirm that this area was responsible for the shift, excess cold RpC and cold PKC␤I (C4-␤I) probe were added to the labeled RpB reaction but failed to compete the shift (data not shown). ␤-Globin-PKC␤II-38 Chimeric Vector Lacks Glucose Sensitivity-To determine whether this 38-nucleotide region could, by itself, confer glucose-sensitive instability when taken out of the region of secondary structure provided by exon inclusion, it was cloned into the p␤G vector reporter system. The new construct encoding a ␤-globin-PKC␤II-38 nucleotide region was transfected into A10 cells. As shown in Fig. 7, high glucose failed to destabilize ␤-globin-PKC␤II-38 mRNA suggesting that the region of 38 nucleotides was not sufficient by itself to elicit the glucose response.
MOE Antisense Oligonucleotides Block Glucose-induced Destabilization of PKC␤II mRNA-Because of the limitations of the ␤-globin system to map glucose-responsive elements within the PKC␤II mRNA, a third approach was taken to further demonstrate that this 38-nt region was associated with destabilization. Antisense (AS) MOE oligonucleotides (20-mers) were designed to "walk" complementary to this putative 38-nt PKC␤II mRNA instability region. MOE modifications are resistant to exo-and endonuclease degradation and do not support cleavage of hybridized mRNA by RNase H. Furthermore, these oligonucleotides bind with high affin- FIG. 3. a, p␤G-PKC␦ chimeric reporter containing an analogous COOH-terminal region from PKC␦, was cloned. b, the p␤G-PKC␦ stability reporter system was transiently transfected into A10 cells exposed to either normal or high glucose for 2 h. 2 g of RNA was used in the RT-PCR analysis using primers for ␤-globin. A, 100-bp DNA marker; B, normal glucose; C, high glucose. The experiment was repeated four times to ensure reproducibility.
FIG. 4. Northern blot analysis of p␤G-PKC␤II. p␤G-PKC␤II A10 stable transfectants were pretreated with 50 g/ml DRB and incubated with normal (5.5 mM) or high glucose for 1-6 h. Total RNA was isolated and 10 g used in Northern blot analysis using a labeled ␤-globin probe. Densitometric scan analysis was performed and data were plotted as percent of ␤-globin-PKC␤II chimeric mRNA remaining versus time. The t1 ⁄2 was 45 min. The data are representative of an experiment repeated on five occasions with similar results.
FIG. 5. The C4 and ␤II exons were cloned into the parent p␤G vector to produce p␤G-PKC␤IID␤I. p␤G-PKC␤IID␤I chimeric plasmid was transiently transfected into A10 cells and cells were exposed to either normal or high glucose for 2 h. 2 g of RNA was used in the RT-PCR analysis using primers for ␤-globin. A, 100-bp DNA marker; B, normal glucose; C, high glucose. The experiment was repeated three times with similar results.
ity to the complementary mRNA sequences (24). AS oligonucleotides 25646 and 25648 spanned consecutive regions while AS 25647 overlapped these sequences (Fig. 8a). AS 25645 and AS 25649 corresponding to upstream (not shown) and downstream sequences, served as controls because they are outside the 38-nt region. To evaluate specific binding and targeting of the antisense, as a separate control, a scrambled sequence (AS 25581) was used. A10 cells were transfected with the AS oligonucleotides (50 nM) and then exposed to 25 or 5.5 mM glucose. As shown in Fig. 8b, AS 25647 blocked high glucose-induced destabilization. AS 25646 and AS 25648 blocked destabilization to a lesser extent. The downstream AS 25649 did not block glucose effects. Scrambled control sequences did not block destabilization (data not shown). This approach further mapped the relevant element to 20 nucleotides within this 38-nt region and suggested that the sequence was important for protein interaction.
UV Cross-linking Detected Association of Proteins-To provide insight into the basis for the mobility shift and nature of the components that bind to the PKC␤II mRNA in response to high glucose exposure, UV cross-linking experiments were performed using transcripts RpA and RpB, which demonstrated specific binding by mobility shift assays. Binding assays were carried out in parallel with the RNA shift analyses, but were further subjected to UV light to generate covalent bonds between the 32 P-labeled RNA transcript and associated proteins, digested with RNase A, and separated by SDS-polyacrylamide gel electrophoresis. Under these conditions a single band was observed at 10 -14 kDa for the control extracts (Fig. 9), and extracts from high glucose-treated cells showed a 5-fold increase in intensity over that observed for extracts from control cells exposed to low glucose. Longer exposure times did not elucidate any other bands. An excess 38-nucleotide sequence, corresponding to the HpaI-BglII region identified in mobility shift assays, competed for protein binding with RpB (Fig. 9). Unlabeled probes RpA and RpB competed for protein binding, respectively, with labeled RpA and RpB, but unlabeled RpC, RpD, and PKC␤I probes did not compete for protein binding following UV cross-linking (data not shown). To control for possible nonspecific protein interactions, micrococcal nuclease (1 g) was added to cell extracts in the presence of excess EGTA. The affinity of this exogenous protein binding to labeled FIG. 7. ␤II38 chimeric plasmid contained the minimum region defined in electrophoretic shift assays. p␤G-PKC␤II38 was transiently transfected into A10 cells and exposed to either normal or high (25 mM) glucose. 2 g of RNA was used in the RT-PCR analysis using primers for ␤-globin. A, 100-bp DNA marker; B, normal glucose; C, high glucose. The experiment was repeated five times and similar results were obtained.
FIG. 6. In vitro RNA electrophoretic mobility shift assay demonstrated a cytosolic factor binding to a glucose-regulated element present in the PKC␤II coding region. a, cytoplasmic extracts from A10 cells (normal and high glucose) containing 3 g of protein were incubated with 32 P-labeled RNA probe RpA in a final volume of 10 l of RNA shift buffer. 100-and 300-fold excess cold competitors (comp) or a nonspecific competitor (NS comp) was added to the binding reaction. Sequential deletion of in vitro transcribed probes identified the binding region. b, 32 P-labeled RNA probe B (RpB), with excess cold competitors (comp), was added to the binding reaction described in a; or c, 32 P-labeled RNA probe C (RpC); or d, 32 P-labeled RNA probe D (RpD), in a final volume of 10 l of RNA shift buffer. Samples were separated on a 10% polyacrylamide gel in 0.5ϫ TBE buffer. Gels were dried and exposed to a Amersham Biosciences phosphorimaging screen. The bottom arrow indicates free probe. The assay was repeated four times with separate cytosolic extracts and similar results were obtained. e, schematic of the PKC␤II RNA probes generated using the T7 promoter and used in in vitro electrophoretic mobility assays. Binding was detected in cell extracts with RpA and RpB. BSA, bovine serum albumin.
RpB in control and high glucose-treated cell extracts remained unaffected (data not shown). Addition of proteinase K to the reactions abolished complex formation (data not shown). These results were taken to demonstrate that cell extracts from A10 cells contain a small molecular weight protein (10,000 -14,000) that specifically binds to a PKC␤II mRNA region containing the 38-nucleotide HpaI-BglII sequence, and that the efficiency of binding, as measured by UV cross-linking, increases in response to high glucose.
Antisense Corresponding to a Portion of the HpaI-BglII Region Blocked Protein Interaction and Glucose-induced Destabilization-To demonstrate whether the protein binding to the mRNA was involved in its destabilization, antisense MOE oligonucleotide ( Fig. 8) complementary to a portion of the HpaI-BglII region and shown to block destabilization, was used as a competitor in the UV cross-linking experiments. The 20-mer, AS 25647, was shown to compete for protein binding with RpB ( Fig. 9). Another antisense oligonucleotide upstream of this region, AS 25649, did not block the protein interaction with RpB or glucose-induced destabilization. Hence, the association of a low molecular weight protein with a specific sequence in the PKC␤II carboxyl terminus exon was required for glucoseinduced destabilization of the mRNA. DISCUSSION The previously reported observation that only PKC␤II and not PKC␤I was subject to glucose-induced mRNA destabilization by a cytoplasmic nuclease activity suggested that the sequence within the PKC␤II exon was responsible. The results of this investigation indicate that when this region is inserted into a ␤-globin reporter gene, glucose-dependent instability is introduced and the mRNA half-life is markedly reduced.
Deletion analysis of the PKC␤II exon monitored by RNA electrophoretic mobility assay showed that a 38-nt region near the middle of the exon was required for this interaction. Furthermore, the addition of an RNase H-resistant antisense MOE oligonucleotide that targeted 20 nt within the 38-nt region abolished the interaction. However, when inserted into the ␤-globin reporter gene, the 38-nt region by itself was not sufficient to confer glucose-dependent destabilization, indicating that other sequences within the PKC␤II exon are necessary. In fact, the destabilization was maximal when a portion of the PKC␤I exon was present in addition to the PKC␤II exon. This difference highlights the importance of the context in which the ␤II exon is placed and may reflect differences in factors such as translation rate, secondary structure, and intracellular localization, which may also influence mRNA stability (25). Computer modeling of the RNA sequence encoded by the C4-␤IIspecific exon indicates that the 38-nt region would likely form a stem-loop structure. The complementary antisense that spanned an AU-rich element and the sequence at the base of a putative stem-loop structure blocked glucose-induced destabilization of PKC␤II mRNA as well as protein interaction. This possible secondary structure may, in part, explain some structural features of the context necessary for the cis-acting element to function.
The antisense oligonucleotide that blocked destabilization also targeted an AU motif. One group of cis elements that may mediate mRNA instability are the AU-rich elements (AREs or AUREs) (see Ref. 25 for review). These elements have been associated with AU-rich element-binding proteins. Analysis of the PKC␤II complete mRNA indicates 46% AU content. However, the RpB probe containing the PKC␤II exon contains 58% AU and the 38-nt fragment contains 66% AU. Furthermore, the 38-nt fragment contains an extended pentamer sequence AUUUUUA that has been identified as a putative AU-rich element-binding protein target. It has also been identified as conferring mRNA instability in the epidermal growth factor receptor transcript (26). Other U-rich sequences identified by Levine et al. (27) as possible AU-rich element-binding protein targets are also present in the PKC␤II exon. It is interesting to note that the c-fos mRNA contains two domains: an AUUUA pentamer region and a U-rich region (28). Both regions appear were transiently transfected into A10 cells and exposed to either 5.5 or 25 mM glucose for 2 h in the presence of the 20-mer antisense oligonucleotides. 2 g of RNA was used in the RT-PCR analysis using primers for the C4 domain and ␤IV5 domain (indicated by arrows in a), such that PKC␤I and PKC␤II products were detected simultaneously. The experiment was repeated at least four times to ensure reproducibility. Incubations were UV cross-linked and separated on a 10% SDS-polyacrylamide gel. Gels were dried and exposed to a Amersham Biosciences phosphorimaging screen overnight. The assay was repeated three times with separate cytosolic extracts and similar results were obtained. BSA, bovine serum albumin.
to be necessary for maximal RNA destabilization. In the case of PKC␤II, however, these elements occur in the coding region and not the 3Ј-UTR. They are also introduced in a regulated fashion via exon inclusion.
Our results support the proposal that the PKC␤II exon, introduced by hormone-regulated alternative splicing of a common pre-mRNA (16), not only specifies a different 52-amino acid carboxyl-terminal end in the protein, but also confers glucose responsive instability to the PKC␤II mRNA in vascular smooth muscle cells. High glucose concentrations increased the levels of a low molecular weight protein that binds to this region and may target the mRNA molecule for degradation. To our knowledge, this is the first report that defines an mRNA instability element present within an exon that responds to regulation by an external stimulus, acute high glucose concentrations in a mammalian cell. The scheme shown in Fig. 10 illustrates how hormone-regulated alternative splicing inserts a region into the mRNA that under conditions of high glucose concentrations, results in the destabilization of the mRNA encoding an important regulatory protein.
Taken together with previous reports by this laboratory showing that the alternative splicing of PKC␤ mRNA is regulated by insulin (5,16,29), this study highlights the possible integration of metabolic regulation mediated between nutrient (glucose) and endocrine (insulin) controls on vascular smooth muscle cell gene expression. Significantly, it is known that hyperglycemia can further increase smooth muscle cell proliferation that may contribute to the development of atherosclerotic lesions in diabetic subjects (30). In view of the results presented here, hyperglycemic episodes could result in the rapid destabilization of PKC␤II mRNA. As a consequence, destabilization would remove PKC␤II signaling that has been shown previously to repress smooth muscle cell proliferation (31)(32)(33), and may therefore help explain the contribution of acute hyperglycemic incidences to the increased risk of vascular disease in diabetes. FIG. 10. Model describing the regulation of protein kinase C ␤II mRNA alternative splicing and destabilization by high extracellular glucose concentrations. Insulin induces the insertion of an exon encoding the carboxyl-terminal region of PKC␤II mRNA (5,29), which results in PKC␤II protein and activity regulating specific cellular events (4,6). In the presence of high glucose concentrations, a low molecular weight protein (10 -14 kDa) binds to a region within the exon and results in PKC␤II mRNA destabilization (2,34) and the reversal of cellular events attributed to PKC␤II activation.