Interleukin-1 Reduces the Glycolytic Utilization of Glucose by Pancreatic Islets and Reduces Glucokinase mRNA Content and Protein Synthesis by a Nitric Oxide-dependent Mechanism*

Culture of rat pancreatic islets with interleukin-1 (IL-1) results in up-regulation of the inducible isoform of nitric oxide synthase and overproduction of nitric oxide (NO). This is associated with reversible inhibition of both glucose-induced insulin secretion and islet glucose oxidation, and these effects are prevented by the inducible nitric oxide synthase inhibitorN G-monomethylarginine. IL-1 also induces accumulation of nonesterified arachidonic acid in islets by an NO-dependent mechanism, and one potential explanation for that effect would involve an IL-1-induced enhancement of islet glycolytic flux. We have therefore examined effects of IL-1 on islet glycolytic utilization of glucose and find that culture of islets with IL-1 in medium containing 5.5 mm glucose results in suppression of islet glucose utilization subsequently measured at glucose concentrations between 6 and 18 mm. The IL-1-induced suppression of islet glucose utilization is associated with a decline in islet glucokinase mRNA content, as determined by competitive reverse transcriptase-polymerase chain reaction, and in glucokinase protein synthesis, as determined by immuoprecipitation experiments, and all of these effects are prevented byN G-monomethylarginine. These findings suggest that IL-1 can down-regulate islet glucokinase, which is the primary component of the islet glucose-sensor apparatus, by an NO-dependent mechanism. Because reductions in islet glucokinase levels are known to cause a form of type II diabetes mellitus, these observations raise the possibility that factors which increase islet NO levels might contribute to development of glucose intolerance.

of these effects are prevented by the inducible nitric oxide synthase inhibitor N G -monomethylarginine (NMMA) (1,2,4), indicating that they occur through NO-dependent mechanisms.
We have recently reported that IL-1 also induces accumulation of nonesterified arachidonic acid in islets by an NO-dependent mechanism (6). Our findings suggested that this reflected suppression of re-esterification of arachidonic acid released during phospholipid turnover, but others have found that NO stimulates arachidonic acid release from macrophagelike cells by a mechanism involving accelerated glycolytic flux (13). This has been attributed to activation of a macrophage phospholipase A 2 enzyme that is regulated by an isoform of the glycolytic enzyme phosphofructokinase (13)(14)(15)(16). Because islets express a similar phospholipase A 2 enzyme (17)(18)(19), it seemed possible that NO-induced acceleration of glycolytic flux might also contribute to IL-1 induced accumulation of nonesterified arachidonic acid in islets. We have therefore examined effects of culturing islets with IL-1 on islet glycolytic utilization of glucose, as reflected by production of [ 3 H]OH from [5-3 H]glucose (20 -28), and on expression of glucokinase mRNA by competitive PCR. Media-Media included KRB (Krebs-Ringer bicarbonate buffer: 25 mM HEPES, pH 7.4, 115 mM NaCl, 24 mM NaHCO 3 , 5 mM KCl, 2.5 mM CaCl 2 , 1 mM MgCl 2 ), nKRB (KRB supplemented with 3 mM D-glucose), cCMRL-1066 (CMRL-1066 supplemented with 10% heat-inactivated fetal bovine serum, 1% L-glutamine and 1% (w/v) each of penicillin and streptomycin), and Hank's balanced salt solution supplemented with 0.5% penicillin-streptomycin.

Materials-Male
Isolation of Pancreatic Islets-Islets were isolated aseptically from male Sprague-Dawley rats by a described procedure (29) involving collagenase digestion of excised, minced pancreas, density gradient isolation, and manual selection under microscopic visualization. Isolated islets were transferred into Falcon Petri dishes containing 2.5 ml of cCMRL-1066, placed under an atmosphere of 95% air, 5% CO 2 , and cultured at 37°C with or without IL-1 or other additives.
Incubation of Islets with IL-1 and Measurement of Insulin Secretion-Islets (400 per condition) were placed in Petri dishes (10 ϫ 35 mm); suspended in cCMRL medium (1 ml) containing no additives, IL-1 (5 units/ml) alone, or both IL-1 and NMMA (0.5 mM); and incubated (2-48 h, 37°C). In some experiments, islets were then removed from the incubation medium, and their secretion of insulin, oxidation of [U- 14 C]glucose to [ 14 C]O 2 , or production of [ 3 H]OH from [5-3 H]glucose was examined in a subsequent incubation. In such experiments, islets were washed 3 times in nKRB, transferred to siliconized test tubes (12 ϫ 75 mm), and preincubated (30 min, 37°C) in nKRB (0.2 ml). For insulin secretion experiments, islets were placed in fresh in KRB medium containing various concentrations (3-18 mM) of D-glucose and incubated (30 min, 37°C, under 95% air, 5% CO 2 ). Aliquots of medium were then removed for measurement of insulin by radioimmunoassay.
Islet Glucose Utilization-As in previously described procedures (20,25), triplicate batches of 10 islets per incubation condition were placed into Microfuge tubes (0.5 ml). Medium was then removed, and radioactive mixture (15 l) was added. This mixture was prepared by placing  H]glucose (7 l, Amersham, specific activity 1 mCi/ml) into silanized glass test tubes and concentrating the solution to dryness under nitrogen to evaporate any [ 3 H]OH. The desired final glucose concentrations (3-18 mM) of the radioactive mixtures were achieved by adding various volumes (0 -167 l) of glucose-free KRB containing 1% bovine serum albumin and an appropriate corresponding volume (33-200 l) of KRB containing 18 mM glucose and 1% bovine serum albumin. The total initial amount of [ 3 H] added to each condition was determined by adding 15 l of each radioactive mixture directly to scintillation vials containing water (0.5 ml). For blank incubations, radioactive mixture was added, but no islets were present. After all additions were complete, tubes containing the islets were placed in scintillation vials (20 ml) containing water (0.5 ml). The vials were then flushed with 95% air, 5% CO 2 , capped with Teflon/silicone septum lids, and incubated (1 h, 37°C, shaking water bath). A Hamilton syringe was then used to introduce 1 N HCl (20 l) into the islet-containing tubes to prevent further catabolism of [5-3 H]glucose. Vials containing these tubes were then incubated (24 h, 37°C) to permit [ 3 H]OH formed by the islets to evaporate and equilibrate with water in the vials. Vials were then cooled to room temperature. The Microfuge tubes were removed and their exterior surfaces rinsed with scintillation fluid (12 ml), which was placed in the vial containing the water in which the Microfuge tube had been immersed. The scintillation vial was then capped with the original septum-lid and mixed. The vials were then equilibrated in the dark, and their 3 H-content was measured by liquid scintillation spectrometry. Average disintegrations/min in blank tubes was subtracted from experimental measurements, and glucose utilization was calculated as:  (30,31), islets (30 from each incubation condition) were placed into Beckman polyallomer tubes. After centrifugation (Beckman Microfuge, 5 s, 10,000 ϫ g), supernatant was discarded, and the islets were resuspended in fresh medium (nKRB, 0.2 ml) and preincubated (30 min, 37°C). Islets were then collected by centrifugation, supernatant discarded, and KRB medium (0.15 ml) containing various concentrations (3-18 mM) of [U-14 C]glucose was added. After resuspension of the islets, the polyallomer tubes were placed in scintillation vials containing filter paper covering the bottom of the vial. The vials were then equilibrated with 95% air, 5% CO 2 , sealed with lids containing gas-tight Teflon/silicone septa, and incubated (2 h, 37°C, shaking water bath) to permit islet metabolism of [ 14  to escape from the incubation solution and react with hyamine in the filter paper. The polyallomer tubes were then removed from the vials and their exterior surfaces rinsed with scintillation fluid (1 ml, ACS, Amersham), which was placed inside the scintillation vial. Additional scintillation fluid (9 ml) was added to each vial, and the 14 C-content was measured by liquid scintillation spectrometry. Total 14 C-content of the stock [U-14 C]glucose solution and blank conversion of [U-14 C]glucose to [ 14 C]O 2 without islets were also determined.
Reverse Transcriptase-Polymerase Chain Reaction Analyses of Islet Content of mRNA Species-After incubation of islets under various conditions, total RNA was isolated after solubilization in guanidinium thiocyanate by phenol/chloroform/isoamyl alcohol extraction and isopropyl alcohol precipitation (32). First strand cDNA was transcribed from total RNA with avian myeloblastosis virus reverse transcriptase (reverse transcriptase, Boehringer Mannheim). Polymerase chain reactions (PCR) were performed on a Perkin-Elmer DNA Thermal Cycler 480. Primer pairs used to amplify fragments of cDNA encoding glucokinase (33) in competitive PCR reactions are described below. Primer pairs used for other gene products were: inducible nitric oxide synthase (34), sense 5Ј-TGCTTTGTGCGGAGTGTCAG and antisense 5Ј-AGAT-GCTGTAACTCTTCTGG (expected fragment 650 bp); and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (35), sense 5Ј-TAGAC-AAGATGGTGAAGG and antisense 5Ј-TCCTTGGAGGCCATGTAG (expected fragment length 1006 bp). Amplification steps (30 cycles) included denaturation (95°C, 1 min), annealing (50 to 60°C, 1 min), and extension (72°C, 2 min) and were performed in Taq-polymerase buffer (Life Technologies, Inc.) containing 1.5 mM MgCl 2 , 1 M of each primer, 200 M each of dATP, dGTP, dCTP, and dTTP, and 25 units/ml Taq DNA polymerase (Life Technologies, Inc.). PCR products were analyzed by 1% agarose gel electrophoresis and visualized by ethidium bromide staining (32). Band intensities of PCR products were measured with an IS-1000 Digital Imaging System (Alpha Innotech Corp.). Fluorescence was recorded, and the areas of the peaks were measured. Increased levels of inducible nitric oxide synthase mRNA in islets could be detected by RT-PCR after 4 h exposure to IL-1. Levels of islet GAPDH mRNA, as detected by RT-PCR, were constant in control and in IL-1-treated islets for up to 48 h in culture. To examine the potential amplification of genomic DNA in PCR reactions, a control amplification was performed in which the reverse transcriptase step was omitted. In no case was a PCR product of the expected length obtained under these conditions.
Examination of the Abundance of Glucokinase mRNA by Competitive RT-PCR-Competitive RT-PCR (72,73) was used to determine the abundance of glucokinase mRNA. In this approach, a competitor DNA species is prepared which contains the same primer template sequences as the target cDNA but which contains an intervening sequence which differs from the target in size or in restriction sites so that PCR products from the target and competitor can be distinguished (72,73). Using the competitor as an internal control, amounts of target cDNA can be determined by allowing known amounts of the competitor to compete with the target for primer binding during amplification (72,73).
To prepare the competitor DNA, two composite primers were synthesized (sense 5Ј-TCACAAGTGGAGAGCGACTCACTGGCATGGCCTTC-CG-3Ј and antisense 5Ј-ATTTGTGGTGTGTGGAGTCCTTGGAGGCC-ATGTAGGC-3Ј). These primers contain the glucokinase primer sequence (underlined) attached to sequences which hybridize to rat GAPDH cDNA (35). This pair of primers was then used in PCR reactions with rat GAPDH cDNA as template. In these reactions, the glucokinase primer sequences are incorporated into the PCR product during amplification, and the intervening sequence derives from GAPDH. The resultant PCR product (360 bp in length) was analyzed by agarose gel electrophoresis, isolated with a QIAEX gel extraction kit (QIAGEN), and used as the competitor DNA species in subsequent PCR experiments. In these experiments, the primer pair (sense 5Ј-TCACAAGTG-GAGAGCGACTC-3Ј and antisense 5Ј-ATTTGTGGTGTGTGGAGTCC-3Ј) was used. These primers hybridize to the glucokinase cDNA sequence and to the competitor DNA sequence. The PCR product derived from the glucokinase cDNA is 450 bp in length, and that from the competitor DNA is 360 bp in length. The products were then analyzed by 1% agarose gel electrophoresis and visualized with ethidium bromide. Product band intensity was then determined with an IS-1000 Digital Imaging System.
The primer set selected for amplification of glucokinase cDNA will not yield a product of the appropriate size with genomic DNA as template. In the rat glucokinase gene (75), the sequence recognized by the sense primer is interrupted by the intron between exons 8 and 9, and the sequence recognized by the antisense primer occurs in exon 10. The amplified sequence of the cDNA therefore includes 11 bp of exon 8, the entirety of exon 9, and a fragment of exon 10. The intervening intron sequences would cause any product amplified from glucokinase genomic DNA to be far larger than that from glucokinase cDNA. In addition, no products of the expected size were observed in control competitive PCR reactions in which the the reverse transcriptase step was omitted. The target glucokinase RT-PCR product was also subcloned and sequenced. The sequence corresponded exactly to the appropriate region of the glucokinase cDNA.
With a fixed amount of target and competitor, varying the PCR cycle number from 19 to 31 was found to yield a constant relative intensity of the target and competitor PCR product bands. In subsequent experiments, 28 PCR cycles were used. To examine the relationship between the amount of input DNA and the ratio of the signals for the target and competitor, in one set of experiments the competitor DNA solution was serially diluted, and aliquots of each dilution were added to a reaction mixture containing a fixed amount of glucokinase cDNA. In a second set of experiments, a solution of target glucokinase cDNA was serially diluted, and aliquots of each dilution were added to a reaction mixture containing a fixed amount of competitor DNA. After PCR amplification, products were analyzed by agarose gel electrophoresis, and product band intensity was determined as above. The ratio of target to competitor product band intensities was found to correspond to the amount of input DNA over a wide range of concentrations. To determine the relative abundance of glucokinase mRNA in islets after incubation under various conditions, total RNA was isolated as described above, and its concentration was determined spectrophotometrically (260 nm). Equal measured amounts of RNA from each incubation condition were then used in competitive RT-PCR reactions with a fixed amount of competitor DNA, and the target to competitor ratio was determined as described above.
Immunochemical Analyses of Islet Glucokinase-To generate an antiglucokinase antiserum, a fusion protein was prepared which contained the sequence of Schistosoma japonicum glutathione S-transferase joined to that of human islet glucokinase. The fusion protein was prepared in Escherichia coli transformed with pGEX vector and purified from E. coli homogenates by affinity chromatography on glutathione-agarose (36). The protein was injected subcutaneously on multiple occasions into a female New Zealand White rabbit as an emulsion in Freund's adjuvant. Immunoprecipitation experiments with this antiserum were performed essentially as described previously (37,38). After incubation under various conditions, islets were washed 3 times in methionine-deficient MEM (9 parts MEM without methionine per 1 part MEM with methionine) and incubated (1 h, 37°C) in methioninedeficient MEM. [ 35 S]Methionine (100 Ci/ml) was then added, and the islets were incubated (3 h, 37°C), harvested by centrifugation, washed (3 times, 0.1 M PBS), and lysed (1 h, 4°C) in PBS (1 ml) containing 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 1 g/ml aprotinin, 100 g/ml phenylmethanesulfonyl fluoride, 1 g/ml leupeptin, and 1 mM iodoacetamide. Cell debris was removed (centrifugation, 15 min, 10,000 ϫ g, 4°C), and the protein content of the supernatants was measured. Aliquots of supernatants from each condition containing identical measured amounts of protein were then diluted with lysis buffer to achieve a final volume of 800 l. These solutions were then preincubated (2 h) with preimmune serum (10 l) and precleared by treatment (1 h) with 100 l of stapylococcal protein A (Immunoprecipitin, Life Technologies, Inc.) and centrifugation (1 min, 4°C, 200 ϫ g). Anti-glucokinase antiserum (5 l) was then added to the supernatants, and the mixture was incubated (overnight, 4°C, with shaking). Staphylococcal protein A (100 l) was then added, and the mixture was incubated (2 h, 4°C, with shaking). Staphylococcal protein A-antibody complexes were then isolated by centrifugation and washed four times with PBS (1 ml) containing 0.5% Triton X-100 and 0.05% SDS. The immunoprecipitates were then washed twice with 10 mM PBS, reconstituted in SDS sample mixture (30 l, 0.25 M Tris-HCl, 20% ␤-mercaptoethanol, 4% SDS), and boiled (5 min). After centrifugation, proteins in the supernatants were analyzed on 10% SDS-polyacrylamide gels and visualized by autoradiography (39).

Incorporation of [ 35 S]Methionine into Trichloroacetic Acid-precipitable Proteins by Pancreatic
Islets-Isolated islets were incubated under conditions described above with no additions (control), with IL-1 (5 units/ml) alone, or with IL-1 plus NMMA (0.5 mM) for 24 h. The islets were then washed three times with MEM without methionine containing 5% fetal bovine serum and, after 1 h of incubation in methioninedeficient medium, were incubated (3 h) with [ 35 S]methionine under conditions described above. The islets were then placed in 15-ml conical test tubes, washed three times with PBS to remove unincorporated [ 35 S]methionine, and homogenized by sonication (Vibracell probe sonicator, 0.5-s bursts at 12% amplitude for 20 s, Sonics & Materials, Inc., Danbury, CT) in buffer A (50 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 1 g/ml leupeptin, 1 g/ml aprotinin, 1 mM phenylmethanesulfonyl fluoride). The homogenates were transferred to 1.5-ml Microfuge tubes and centrifuged (15,000 rpm, 4°C, 15 min, Beckman Microfuge). Aliquots (10 l) of the supernatants were then placed on 2-mm square sections of Whatman filter paper and allowed to dry. The filter paper pieces were then boiled in 10% trichloroacetic acid for 10 min, rinsed twice with ice-cold 5% trichloroacetic acid, rinsed once with ice-cold absolute ethanol, allowed to dry, and then placed in scintillation vials to which 12 ml of ACS scintillation mixture (Amersham) was added. The 35 S-content was then determined by liquid scintillation spectrometry. Other aliquots of the supernatants were used to determine total protein content or were subjected to SDS-PAGE and autoradiography under conditions described above.

RESULTS
The initial question that motivated this study was whether treatment of islets with IL-1 would enhance glycolytic flux. Because culture of islets with IL-1 has been reported not to affect glycolytic utilization of 16.7 mM glucose (10, 11), we measured glucose utilization at several concentrations between 3 and 18 mM to determine whether enhanced utilization occurred at lower concentrations. Freshly isolated islets were incubated for 24 h in medium containing 5.5 mM glucose and no other additions, IL-1 alone, or both IL-1 and NMMA. The islets were then placed in fresh medium and incubated with various concentrations of [5-3 H]glucose. Glycolytic utilization of this substrate was determined by measurement of [ 3 H]OH production, which is generated at the triose-phosphate isomerase and enolase reactions reactions (20 -28) and provides a quantitative measure of islet glycolytic flux (27).
With control islets, glucose utilization increased with the medium glucose concentration until a maximal rate was achieved at 15-18 mM glucose (Fig. 1). About half of that rate was achieved at 8 mM glucose, as previously reported (20,21,23). In contrast, islets that had been incubated with IL-1 exhibited little increase in glucose utilization as the medium glucose concentration was increased within the range of 3-18 mM. With islets that had been incubated with both IL-1 and the nitric oxide synthase inhibitor NMMA, glucose utilization was similar to that of control islets at all tested glucose concentrations (Fig. 1). Incubation of islets with NMMA alone was found not to influence islet glucose utilization significantly at 3 mM glucose ( 0.646). These findings suggested that IL-1 induced a reduction in islet glucose utilization by an NO-dependent mechanism. Because these findings differed from the reported lack of effect of IL-1-treatment on islet utilization of 16.7 mM glucose (10, 11), we examined effects of IL-1 on insulin secretion and glucose oxidation to determine whether these parameters were affected under our incubation conditions in a manner similar to that previously reported. These experiments were conducted at multiple glucose concentrations to supplement information from previous studies performed at basal and near-maximally stimulatory glucose concentrations but not at intermediate concentrations (1, 2, 4, 7, 8, 10 -12, 31, 40).
With control islets that had been incubated for 24 h at 5.5 mM glucose without IL-1, subsequent incubation with various concentrations of glucose induced a progessive rise in insulin secretion as the medium glucose concentration increased (Fig. 2). With islets that had incubated with IL-1, there was little rise in insulin secretion as the medium glucose concentration increased within the range of 3-18 mM, in agreement with previous reports (7,31,40). With islets that had been incubated with IL-1 and NMMA, insulin secretion was similar to that of control islets, as previously reported (2,4). Incubation of islets with NMMA alone did not influence islet insulin secretion significantly at 3 mM glucose (1.19 Ϯ 0.83 fmol/islet ϫ min versus control value of 1.04 Ϯ 0.13, p value 0.61) or at 18 mM glucose (3.71 Ϯ 0.83 fmol/islet ϫ min versus control value of 3.58 ϩ 0.53, p value 0.90).
The influence of IL-1 treatment on islet oxidation of [U-14 C]glucose to [ 14 C]O 2 was examined next (Fig. 3). Control islets produced increasing amounts of [ 14 C]O 2 as the medium [U-14 C]glucose concentration increased, with near maximal levels at 15-18 mM glucose and half-maximal levels at about 7 mM glucose, in agreement with previous reports (22). Comparison of data in Figs. 1 and 3 indicated that control islets oxi-datively metabolized 30 Ϯ 1% of the glucose that had been utilized, and this value did not vary with the glucose concentration, as previously reported (20,27). IL-1-treated islets produced amounts of [ 14 C]O 2 similar to that of control islets at 3 mM glucose but produced smaller amounts than control islets at glucose concentrations between 6 and 18 mM. The 46% suppression in [ 14 C]O 2 production at 15 mM glucose approximates the reported 49% suppression at 16.7 mM glucose (9). Islets incubated with both IL-1 and NMMA exhibited rates of glucose oxidation that were statistically indistinguishable from those of control islets, consistent with previous reports (1, 2). Incubation of islets with NMMA alone had little influence on glucose oxidation at 3 mM glucose (2.97 pmol/islet ϫ 2 h versus control value of 3.79) or at 18 mM glucose (23.37 pmol/islet ϫ 2 h versus control value of 26.67).
Data in Figs. 2 and 3 therefore conform to reported effects of IL-1 on islet glucose oxidation and insulin secretion, but the effect of IL-1 to reduce islet glucose utilization in Fig. 1 differs from previous studies (10,11). This discrepancy may be attributable to differences in conditions under which islets were exposed to IL-1. One difference is that, in our study, islets were exposed to IL-1 in medium containing 5.5 mM glucose, while in previous studies of IL-1 effects on islet glucose utilization, the medium glucose concentration during IL-1 exposure was considerably higher (10,11). As illustrated in Fig. 4, when islets were incubated for 24 h without IL-1 in medium containing 5.5, 11, or 23 mM glucose, subsequently measured utilization of 15 mM [5-3 H]glucose was similar for all three groups of islets. In contrast, suppression of glucose utilization was observed with islets that had been incubated for 24 h with IL-1 at 5.5 mM glucose, but this effect was reduced or absent with islets that had been incubated with IL-1 at 11 or 23 mM glucose, respectively. The effect of IL-1 to reduce islet glucose utilization therefore occurs when islets are cultured at a glucose concentration corresponding to a euglycemic state, but this effect is attenuated or prevented when IL-1 exposure occurs at higher glucose concentrations.
The shape of the control curve in Fig. 1 for glucose utilization as a function of glucose concentration over the range of 3-18 mM is highly characteristic of pancreatic islets (20,21,27) and is attributable to the kinetic properties of the enzyme glucokinase, which governs the overall rate of islet glycolytic flux (26,27,(41)(42)(43). The diminished rise in glucose utilization with increasing glucose concentrations in IL-1-treated islets suggested that IL-1 might reduce islet glucokinase expression under our culture conditions. To evaluate this possibility, islet glucokinase mRNA content was examined by competitive RT-PCR using cDNA prepared from islet RNA as template, oligonucleotide primers designed from the rat glucokinase cDNA sequence, and a competitor DNA. The competitor DNA prepared for these experiments shares with glucokinase cDNA the sequences recognized by the primers in the PCR reactions, but the competitor yields a smaller product (360 bp) than that derived from glucokinase cDNA (450 bp) (Fig. 5A). The ratio of signals from the target and competitor was found to correspond to the input DNA over a wide range of concentrations (Fig. 5B).
As illustrated in Fig. 6, when such competitive RT-PCR reactions were performed with equal measured amounts of input RNA from islets that had been incubated with IL-1 for various periods, the glucokinase target to competitor ratio declined substantially after 8 h of incubation and continued to fall at longer incubation intervals. In contrast, when equal measured amounts of input RNA from control islets incubated without IL-1 were used as template in the competitive RT-PCR reactions, the glucokinase target to competitor ratio declined relatively little as a function of time in culture (Fig. 6). As illustrated in Fig. 7, when islets were incubated with both IL-1 and NMMA for 24 h, RNA from the islets yielded a glucokinase target to competitor ratio similar to that of control islets and substantially greater than that of islets incubated with IL-1 alone for 24 h.
The data in Figs. 6 and 7 therefore suggest that culture of islets with IL-1 in medium containing 5.5 mM glucose induces a decline in islet content of glucokinase mRNA by an NO-dependent mechanism. Consistent with this possibility, IL-1 was found to induce an increase in islet production of NO, as reflected by nitrite accumulation in the incubation medium, with a time course similar to that of the IL-1-induced decline in islet glucokinase mRNA content (Fig. 8). Nitrite is produced from NO by spontaneous oxidation in aqueous solutions, and its production by islets has previously been demonstrated to correlate with IL-1-induced synthesis of the inducible isoform of nitric oxide synthase and to be prevented by NMMA (37).
To determine whether the IL-1-induced reduction in islet glucokinase mRNA was associated with a reduction in synthesis of glucokinase protein, islets were incubated at 5.5 mM glucose for 24 h without IL-1, with IL-1 alone, or with both IL-1 and NMMA, and the islets were then incubated with [ 35 S]methionine. Immunoprecipitation was then performed with antiglucokinase antiserum or with preimmune serum, and the immunoprecipitates were analyzed by SDS-PAGE and autoradiography. Immunoprecipitates obtained from control islets with anti-glucokinase antibody but not with preimmune serum For glucose utilization at 3 mM glucose, Student's t test p values for the difference between the control group and IL-1 treatment were 0.019, 0.506, and 0.699 for islets that had been cultured at 5.5, 11, or 23 mM glucose, respectively, and, for glucose utilization at 15 mM glucose, p values for the difference between the control group and IL-1 treatment were 0.002, 0.022, and 0.876 for islets that had been cultured at 5.5, 11, or 23 mM glucose, respectively. contained a radiolabeled band of the 50-kDa size expected for glucokinase (44) (Fig. 9). The intensity of the glucokinase band was greatly reduced in immunoprecipitates from islets that had been incubated with IL-1, but the intensity of the band from islets that had been incubated with both IL-1 and NMMA was similar to that from control islets, suggesting that IL-1 induces a decline in islet synthesis of glucokinase protein by an NO-dependent mechanism.
As illustrated in Fig. 10, SDS-PAGE and autoradiographic analysis of identical loaded amounts of total protein from [ 35 S]methionine-labeled islets that had been incubated under control conditions (first lane), with IL-1 alone (second lane), or with IL-1 plus NMMA (third lane) did not indicate a global reduction in the synthesis of proteins induced by incubation with IL-1. Under these [ 35 S]methionine-labeling conditions, total incorporation of [ 35 S]methionine into trichloroacetic acidprecipitable protein in islets that had been incubated under control conditions (4.22 Ϯ 0.24 ϫ 10 4 dpm/g) for 24 h was similar to that of islets that had been incubated with IL-1 alone (3.68 Ϯ 0.11 ϫ 10 4 dpm/g, p value 0.12) or with IL-1 plus NMMA (5.12 Ϯ 0.42 ϫ 10 4 dpm/g, p value 0.12). The total cytosolic protein content of islets incubated under control conditions (0.291 Ϯ 0.035 g/islet) for 24 h was also similar to that of islets incubated with IL-1 alone (0.243 Ϯ 0.032 g/islet, p value 0.36) or with IL-1 plus NMMA (0.272 Ϯ 0.035 g/islet, p value 0.75).

DISCUSSION
Stimulation of pancreatic islets with concentrations of glucose exceeding 5 mM induces insulin secretion (27), and this requires that glucose be transported into islet ␤-cells by GLUT-2 facilitative transporters (45-47) and metabolized (27,48). The first step in glycolytic utilization of glucose is its conversion to glucose 6-phosphate, and in ␤-cells this step is catalyzed predominantly by glucokinase (26, 27, 41, 42, 49 -51). Glucokinase is expressed only in glucose-sensing cells, such as hepatocytes and ␤-cells (52), and its kinetic properties cause the rate of glucose entry into glycolysis within ␤-cells to rise with extracellular glucose concentrations in the physiologic range (27,42,52). Because glucokinase activity in ␤-cells is substantially lower than that of other glycolytic enzymes, it governs glycolytic flux (26,27,42). Overwhelming evidence indicates that glucokinase is the primary component of the ␤-cell glucose-sensor apparatus and accounts for the characteristic glucose concentration dependence of insulin secretion (27,42,52). This is supported by the fact that genetic mutations in glucokinase can cause a form of type II diabetes mellitus designated maturity-onset diabetes of the young (MODY) and that catalytic activities of mutant glucokinases encoded by MODY genes are reduced (53)(54)(55). MODY is an autosomal dominant disorder and occurs despite the presence of one normal glucokinase allele (53). Because glucokinase is active as a monomer, dominant negative effects of a mutant gene product are unlikely (54), suggesting that 50% reduction of ␤-cell glucokinase levels can produce diabetes (42,54). Similarly, ␤-cell-specific targeted disruption of the glucokinase gene in mice produces diabetes in heterozygotes, and islets isolated from such mice exhibit about 48% of normal glucokinase activity (56). Nearly normal levels of ␤-cell glucokinase therefore appear to be required to maintain euglycemia, and relatively modest reductions in glucokinase may result in diabetes (42).
Glucokinase gene transcription is thought to occur constitutively in ␤-cells (42), and its regulation is incompletely understood, although a 50-kDa factor that binds to the upstream promoter of the gene is preferentially expressed in ␤-cells (57). No compensatory overexpression of the normal glucokinase gene appears to occur in MODY patients or in mice heterozygous for a disrupted glucokinase gene (42,56). This suggests that mechanisms to up-regulate glucokinase gene transcription may not be available in ␤-cells, although insulin enhances transcription of this gene in hepatocytes (42,58,59), which employ a different promoter than that employed by ␤-cells (33,57,59). Some observations suggest that ␤-cell glucokinase mRNA expression is not entirely constitutive and can be regulated: 1) exercise training of rats reduces both islet glucokinase mRNA content (60) and insulin secretion (61); and 2) individual ␤-cells in rat islets differ in glucokinase mRNA content, with highly glucose-responsive cells containing twice the amount of less responsive cells (43).
Our findings provide another example of regulation of islet glucokinase mRNA content and indicate that a reduction in this mRNA species occurs by an NO-dependent mechanism when islets are incubated with IL-1 in the presence of a physiologic concentration of glucose. This is associated with corresponding reductions in glucokinase protein synthesis, glycolytic utilization of glucose, and glucose-induced insulin secretion. These observations raise the possibility that factors which increase islet NO production could contribute to the development of glucose intolerance by reducing islet glucokinase levels. Mutations in the glucokinase gene probably account for a minority of the type II diabetic population (42, 53-55, 62), and down-regulation of glucokinase levels in islets FIG. 9. Immunoprecipitation analyses of the effect of interleukin-1 on islet synthesis of glucokinase. Pancreatic islets were incubated for 24 h at 37°C in cCMRL medium containing 5.5 mM glucose and no other additions or in medium supplemented with IL-1 (5 units/ ml) alone or with both IL-1 and NMMA (0.5 mM). At the end of that incubation, islets from each condition were placed in fresh medium supplemented with [ 35 S]methionine and incubated for 3 h at 37°C under conditions described under "Experimental Procedures." Islets were then washed in [ 35 S]methionine-free buffer, collected by centrifugation, and lysed as under "Experimental Procedures." The lysates were then subjected to centrifugation, and the protein content of the supernatants was measured. Aliquots of lysate supernatants containing equal amounts of measured protein were precleared as described under "Experimental Procedures" and then incubated with either preimmune serum (left panel) or with anti-glucokinase antiserum (right panel). Staphylococcal protein A was then added, and immunoprecipitates were collected by centrifugation and washed as described under "Experimental Procedures." Radiolabeled proteins contained in the resulting immunoprecipitates were then analyzed by SDS-PAGE and visualized by autoradiography as described under "Experimental Procedures." This experiment is representative of three, and similar results were obtained in each case. At the end of that incubation, islets from each condition were placed in fresh medium supplemented with [ 35 S]methionine and incubated for 3 h at 37°C under conditions described under "Experimental Procedures." Lysates were then prepared from the islets and centrifuged as described in the legend to Fig. 9. The protein content of the supernatants was measured, and aliquots containing equal amounts of measured protein were analyzed by SDS-PAGE and autoradiography.
with normal glucokinase genes might be considered as a potential contributor to the evolution of diabetes. There has been considerable interest in the possibility that immunomodulatory cytokines such as IL-1 contribute to development of type I diabetes mellitus (63)(64)(65), but NO production can also be increased by non-immunologic stimuli in various target cells (66). Endogenous factors which increased islet NO production sufficiently to reduce glucokinase levels by 50% might impair glucose tolerance (42).
Previous reports that IL-1 does not affect islet glucokinase levels (9) or glucose utilization (10, 11) employed culture conditions that differed from those used here. We exposed freshly isolated islets to IL-1 for 24 h in medium containing 5.5 mM glucose. Previous studies involved culture of islets for 5 days in medium containing 11.1 mM glucose, followed by exposure to IL-1 for two additional days at 11.1 mM glucose (9 -11). When islets are cultured for several days in medium containing glucose concentrations exceeding 9 mM, glucokinase levels are up-regulated to values exceeding those of freshly isolated islets, although glucokinase mRNA levels are unaffected (41). It has been proposed that glucose increases islet glucokinase expression post-transcriptionally by stabilizing the enzyme and protecting it from proteolysis (42). Any effect of IL-1 to reduce islet glucokinase levels in previous studies may have been offset by the effect of prolonged culture in medium containing high glucose concentrations to increase glucokinase expression. Consistent with this possibility is our observation that culture of islets at 11 or 23 mM glucose attenuates or prevents, respectively, the effect of IL-1 to suppress the low-affinity component of islet glucose utilization attributable to glucokinase. Although glucokinase down-regulation may not be required for IL-1 to inhibit glucose-induced insulin secretion (9 -11), such inhibition would be an expected consequence of glucokinase down-regulation. It is of interest that culture of clonal HIT-T15 cells with IL-1 has been reported to reduce glucose-induced insulin secretion and to reduce both the glycolytic utilization of glucose and glucokinase activity in these cells (74), providing additional support for the possibility that IL-1 can down-regulate ␤-cell glucokinase levels under some conditions. IL-1-induced NO-dependent down-regulation of islet glucokinase mRNA content could reflect a reduction in gene transcription or in mRNA stability. There are several precedents for effects of NO on gene transcription. NO amplifies both IL-1induced transcription of the cyclooxygenase-2 gene in rat mesangial cells (67) and Ca 2ϩ -induced gene transcription in neuronal cells (68) through cGMP-dependent mechanisms that reflect activation of guanylate cyclase by NO (66). NO also exerts cGMP-independent effects to down-regulate early growth response-1 gene expression in macrophages (69) and to inhibit Epstein-Barr virus gene transcription in lymphocytes (70). These effects may reflect S-nitrosylation of critical thiol groups in transcription factors (70,71), which appears to account for the ability of NO to reduce activity of the transcriptional factor AP-1 (71). Determination of whether similar mechanisms are involved in the IL-1-induced NO-dependent decline in islet glucokinase mRNA content will require further study.