Effect of Glucose on Initiation and Elongation Rates in Isolated Rat Pancreatic Islets*

SUMMARY Isolated rat pancreatic islets incubated in media containing glucose at 16.6 mM incorporate [3H]leucine into protein at twice the rate of islets incubated in glucose at 2.8 mu. Increased incorporation was independent of new RNA synthesis, so the effects of glucose on initiation and elongation rates were determined. The relative rates of elongation at 2.8 and 16.6 mM glucose were determined by pulsing islets in [3H]leucine for 6 min and measuring the ratio of incorporation into soluble (S) and polysome associate (P) peptides. At 2.8 mM glucose, S/P = 3.41 f 0.28, and at 16.6 mu glucose, S jP = 4.72 f 0.27, an apparent 28% increase in elongation at the higher glucose. The ratio method assumes that the size of the proteins synthesized is the same. Islets proteins were labeled with [14C]leucine at 16.6 mM glucose and with [3H]leucine at 2.8 mM glucose and subjected to co-electrophoresis on sodium dodecyl sulfate polyacrylamide gels, and the average molecular weight of proteins synthesized at the higher glucose concentration was found to be about 26% smaller. After correction for differences in the size of proteins synthesized, there was no significant effect of glucose on elongation. at the level of initiation includes: (a) increased [3H]leucine incorporation into nascent peptide with the distribution on heavier polysome aggregates indicating an increased number of ribosomes per mRNA,

Glucose regulation of proinsulin-insulin synthesis appears to be more complex. When elongation was partially blocked, it was determined that in 16.6 mM glucose about 2 to 3 times more proinsulin mRNA was available for translation. There was a preferential utilization of the available mRNA for proinsulin in addition, since there was about a 5 to 6-fold increase in over-all translation of proinsulin mRNA. Methods are being developed to determine whether the glucose effect on translation of proinsulin mRNA is at the level of initiation, elongation, or both.
While mammalian cells deprived of glucose have diminished rates of protein synthesis (l-4), alterations of glucose concentrations within the physiological range of 2 to 20 mM increase rates of protein synthesis only in the islets of Langerhans of the pancreas (5-10).
Previous studies demonstrated that isolated rat islets incorporated about twice as much [3H]leucine into protein when incubated in 15.3 mM glucose compared to 2.8 mM glucose (S-10).
This glucose stimulated increase in incorporation was associated with marked changes in islet RNA metabolism (11) and with an increase in the number of ribosomes active in protein synthesis (i.e. polyribosomes) (12). Regulation could therefore occur at a number of levels, including synthesis of islet mRNA, or at initiation and elongation of peptide synthesis.
It was previously demonstrated that glucose activation of islet ribosomes occurs in the presence of actinomycin D at a concentration sufficient to inhibit islet RNA synthesis by 90% (13). Since RNA synthesis was not rate-limiting, this study was undertaken to determine the effects of varying glucose concentrations on initiation and elongation rates in isolated rat pancreatic islets. , and sonicated in 300 ~1 of HZ0 for 10 s as described (lo), then 50 ~1 of each preparation were mixed, sodium dodecyl sulfate buffer (100 ~1) was added, and the sample was subjected to electrophoresis and prepared for counting as described under "Experimental Procedure." (l%]Leucine incorporated at 16.6 mM glucose, O---0; [aH]leucine incorporated at 2.8 mM glucose, O---O. Bovine proinsulin and insulin were not separable by this method and migrated in Slices 5 to Y on separate gels subjected to electrophoresis under identical conditions. Note that the maximal glucose effect, the highest W:3H ratio, occurred in this same region of the gel. The average molecular weights of proteins synthesized was estimated by migration on sodium dodecyl sulfate gels as described by Palmiter (19). Standard curves were derived by performing electrophoresis on proteins of known molecular weight.
were minced finely in cold Hanks' buffer (140 rnM NaCl, 6 mM KCl, 1.3 mM CaC12, 1.3 mM MgSOd, 0.3 mM NazHPOa, 0.4 mM KH~POI, and NaHC03, 7.5yo w/v, added to adjust pH to 7.4), the fat was decanted, then mixed with 50 mg of crude collagenase in 5 ml of Hanks' buffer, and digested with vigorous stirring with a magnetic stirring bar at 37" for 16 to 20 min. The method was modified by centrifuging the digest at 500 X g for 0.5 min, the collagenase containing supernatant was discarded, and the pellet was vigorously resuspended in fresh buffer by hand shaking. This was repeated four to six times with aliquots observed under the microscope until islets free of acinar tissue predominated.
Since islets have a connective tissue capsule which is destroyed on prolonged incubation with collagenase, it was found that this modification gave more reproducible islet yields, about 250 to 500 islets per preparation.
Islets were removed with a drawn out capillary pipette, placed into Petri dishes, and washed 10 to 20 times with fresh buffer to minimize nuclease contamination from acinar tissue and collagenase.
Islets were incubated in small siliconized Dounce homogenizers in 250 to 500~1 of Krebs-Ringer bicarbonatebuffered media (107 mM NaCl, 5.9 mM KCl, 1.1 mM MgClt, 1 mM CaC12, NaHC03, 7.5yo w/v, added to adjust pH to 7.4; and bovine albumin, 1% w/v) at 37" (unless otherwise stated) in a Dubnoff shaker at 40 rpm, continually gassed with 5% CO2-95yo 02 to . _--Polypeptide Elongation-Islet polyribosomes are rapidly labeled with a short pulse of radioactive amino acids (12). That the label is in nascent peptide was demonstrated by the fact that puromycin removed greater than 90% of the radioactivity from polysomes labeled for 5 min. On the other hand, when polysomes were labeled for 15 min, puromycin removed significantly less radioactivity, suggesting that the labeled amino acid may be incorporated into ribosomal protein. 1 The radioactivity in the soluble protein increases linearly after a lag (16,17). This is due to labeling of nascent peptides on polysomes followed by their release into soluble proteins.
The ratio of released peptides to polysome-associated peptides at 2.8 versus 16.6 mM glucose, is a measure of the relative rate of elongation (18,19). As seen in Table I, islets incubated for 45 min in 16.6 mM glucose then pulsed with [3H]leucine incorporated about twice as much label into both nascent and soluble peptide as did islets similarly treated but in 2.8 mM glucose. The ratio of incorporation during the 6-min pulse into soluble versus nascent proteins at 16.6 mM glucose was 4.7 compared to a ratio of 3.4 at 2.8 mM glucose. Increased glucose appeared to increase the relative rate of elongation by about 28 %. Measurement of elongation rates by this method must be corrected for differences in molecular weights of proteins synthesized under different experimental conditions  of Polyribosomes-Polyribosomes were prepared and analyzed on 10 to 40% sucrose gradients as previously described (12). [3H]Leucine incorporation into nascent peptide and completed peptide was determined by precipitating polysomes and supernatant, respectively, with trichloroacetic acid (5%) and counting in a liquid scintillation counter as described (12).
Electrophoresis-Sodium dodecyl sulfate acrylamide gel electrophoresis on 15% gels in 0.1% (w/v) sodium dodecyl sulfate was performed as described (15). The gel in Fig. 1 was sliced into lto 2-mm sections, digested in 200 ~1 of 30yo He02 at 22" overnight, then mixed with 1 ml of NCS, and counted in a liquid scintillation mixture as described (10). Spillover of 14C counts into the 3H channel was 7.4yo, and spillover of 3H counts into the 1% channel was 1.3% and was corrected in each case. The gel slices in Fig. 4 were solubilized with NCS (500 ~1) and I&O (50 ~1) overnight at 37" before mixing with the liquid scintillator.
Effect of aurintricarboxylic acid on distribution of nascent peptide on polysomes from islets. Isolated islets (25 per tube in each of four tubes) were incubated in Krebs' buffer with the glucose concentration indicated f ATA (10-b M) for 45 min and then pulsed with [3H]leucine (final concentration, 625 bCi per ml) for 6 min, cycloheximide was added (100 pg per ml) for 1 min, and the polysomes were prepared and analyzed on sucrose gradients as described under "Experimental Procedure." The contenuous line (-) represents the absorbance at 260 nm of carrier liver polysomes. The bars with solid lines represent [3H]leucine in nascent peptide from control islets; the bars with dashed lines represent nascent peptide from ATA-treated islets. The results of the separate preparation are plotted on the same graph with the polysome profile of carrier rat liver which was the same in each.
proteins then were solubilized and electrophoresed on sodium dodecyl sulfate polyacrylamide gels where the mobility on the gels is proportional to the log of the molecular weight. Islet proteins synthesized in 16.6 mM glucose have a different molecular weight profile on sodium dodecyl sulfate polyacrylamide gel electrophoresis than proteins synthesized in 2.8 mM glucose (Fig. 1). The over-all pattern of proteins synthesized appeared to be qualitatively similar at both glucose concentrations, but there was a relative increase in synthesis of lower molecular weight peptides at 16.6 mM glucose. The average molecular weight of proteins synthesized at 16.6 mM glucose was 34,000, compared to 46,000 at 2.8 mM glucose (average of three gels each). When the apparent increased elongation rate (28%) at 16.6 mM glucose is corrected (19) for synthesis of peptides which are on the average 26% smaller, the relative rates of elongation are virtually the same. A change in the elongation rate, therefore, is not the explanation for the doubling in incorporation seen in Table I. Polypeptide Initiation-A number of observations suggest that the effect of glucose on islet protein synthesis is through increased initiation of translatable mRNA. It was previously demonstrated that the increased protein synthesis at 16.6 mM glucose is associated with an increase in [3H]leucine incorporation into nascent peptide which corresponds to the number of ribosomes active in protein synthesis (Ref. 13; see also min, and the polysomes were prepared and displayed as in Fig. 2. shifted to heavier polysome aggregates, suggesting that at 16.6 mM glucose the average mRNA is loaded with more ribosomes than at 2.8 mM glucose. Further evidence that initiation is rate limiting at 2.8 mM glucose was obtained by incubating islets in aurintricarboxylic acid (ATA), an antibiotic which inhibits initiation in mammalian cells (20). ATA at 1OV M produced about a 50% reduction in protein synthesis at 16.6 mM glucose associated with a decrease in [3H]leucine incorporation into nascent peptide ( Fig. 2A). ATA also shifted the distribution of nascent peptide to smaller polysomes.
ATA at 10e5 M had negligible effects on protein synthesis in islets incubated in 2.8 mM glucose (Fig. 2B).
Islet mRNA: Amount and Rates of Utilization-Several investigators have recently been able to estimate the amount of mRNA and its utilization rates under various experimental conditions by partially inhibiting elongation (6, 21, 25). Since the amount of protein synthesized over a period of time bficiently greater than the translation time) is proportional to the amount of mRNA and the initiation of the messenger, when elongation is slowed ribosomes build up, and the amount of protein synthesized is no longer limited by initiation.
If the mRNAs are saturated with ribosomes, the amount of protein synthesized is proportional to the amount of available mRNA. Low doses of cycloheximide were used to partially inhibit elongation of islet protein synthesis which was associated with a buildup of ribosomes on mRNA in islets incubated at 2.8 mM glucose, but not at 16.6 mM glucose (Fig. 3), which is further evidence that initiation is rate-limiting at low glucose. Partial inhibition of elongation with low dose cycloheximide * The abbreviation used is: ATA, aurintricarboxylic acid.
was performed in isolated islets to determine the effects of glucose on the mRNA activity for total islet proteins, and specifically for proinsulin and insulin. Islets incubated in 16.6 mM glucose incorporated about twice as much [3H]leucine into total protein as did islets in 2.8 mM glucose (Table II).
Incorporation of [31-I]leucine at both concentrations of glucose in the presence of low dose cycloheximide (2 pg per ml) was not significantly different.
This observation suggests that the amount of mRNA available for translation of total islet proteins is the same at low and high glucose, and the difference in incorporation is due to different rates of translation of the mRNA.
Since the over-all rate of elongation is the same, the doubling of protein synthesis in high glucose is most likely due to a doubling of the over-all rate of initiation.
The [311]leucine-labeled islet proteins synthesized at 2.8 and 16.6 rnrvf glucose f low dose cycloheximide were electrophoresed on sodium dodecyl sulfate polyacrylamide gels, and an estimate of the rate of utilization of individual islet messengers was obtained (Fig. 4). Non-globin proteins in reticulocytes (22), and adenovirus infected HeLa cells (23) have been similarly studied but, unfortunately, islet proteins are heterogeneous and cannot be clearly resolved into discrete protein bands. Nevertheless, if it is assumed that,the rate of translation of individual proteins is the same in low dose cycloheximide, it appears that the amount of mRNA present for larger molecular weight proteins (Fig. 4C) is as abundant as for smaller proteins, but the increased synthesis of smaller proteins at 16.6 mM glucose is due to increased utilization of smaller messengers (Fig. 40). Even at 2.8 mM glucose there is evidence for differential rates of messenger utilization (Fig. 4, compare A to B). The experiment demonstrates that the effect of glucose on proinsulin-insulin synthesis is more complex than the effect on total islet protein synthesis.
As noted previously (lo), the glucose stimulatory effect is much greater for proinsulin-insulin synthesis, so that proinsulin and insulin account for about 2% of total islet protein synthesis at 2.8 mM glucose and for 18% at 16.6 mM glucose. When elongation was partially inhibited, there was still a significant 3-fold difference in proinsulin-insulin synthesis. This suggests that, in contrast to total mRNA activity, mRNA available for translation of proinsulin is increased S-fold by high glucose. Since mRNA activity accounts for only part of the glucose stimulatory effect, the remainder must be due to increased translation of mRNA for proinsulin. The effect of glucose on the relative elongation rate for proinsulin has not been determined, and it is therefore not possible to say with this data whether the translational effect is due to increased initiation, elongation, or both.

DISCUSSION
The purpose of this study was to determine how alterations of glucose in the media, within the physiological range of 2 to 20 mM, affect the rate of protein synthesis in isolated islets. A number of levels of regulation could esist., including the synthesis of islet mRNA, activation of pre-existing mRNA, increased rate of initiation, or increased elongation rates. Previous studies with actinomycin D showed that glucose stimulated islet protein synthesis was associated with polysome aggregation which was independent of new RNA synthesis (13). An increase in the number of translatable messengers which were previously synthesized, however, remains a possibility (24). This seems unlikely because an increase in the number of available mRNAs without a change in the rate of initiation would increase nascent peptide formation but not produce the shift to synthesis on larger polysome aggregates (25). The experiment where elongation was partially inhibited with cycloheximide (2 pg per ml) also supports the hypothesis that glucose control of total islet protein synthesis is post-transcriptional.
With partial inhibition of elongation, the messengers become saturated with ribosomes and the amount of protein synthesized, although reduced, should be proportional to the amount of messenger available for translation. Under these conditions the incorporation of [3H]leucine into total islet protein was not significantly different at 2.8 or 16.6 mM glucose (Table II).
Measurement of relative rates of elongation in tissues using the ratio method has been accomplished in tissue culture cells at various temperatures (18) and undergoing mitosis (16) and in the chicken oviduct which is estrogen-primed (19). The movement of radioactive material from nascent peptide to soluble chains, the average transit time, is defined by Fan and Penman (16) as the length of time required for a ribosome, after attachment to mRNA, to complete and release the nascent peptide. This average transit time is independent of initiation. Palmiter polysotncs should bc undcgraded; the rate of amino acid ittcorporatiott should be proportional in the tissues studied for the duratiott of labeling titne; attd the molecular weights of the protcitts syttthcsized should bc sitnilar.
Since alterations in glucose had 110 effect on total islet mRNA activity or on elongation, the rate limitittg event at 2.8 tntit glucost appeared to be ittitiation.
Evidcttcc supporting this hypothesis iucludes: (a) increased ttasccttt peptidc formation on larger polysome aggregates in 16.6 111~1 glucose indicated an increased number of ribosomes per mcsscngcr (13) ; (b) partial inhibition of initiation by ATA in 16.6 rnAr glucose decreased the number of ribosomes per messenger, but not itt 2.8 mht glucose; and (c) loading of ribosomcs on tnRNA in 2.8 rnhl glucose when elongation was partially blocked with low doses of cycloheximide, which did not occur in 16.6 1x1~ glucose.
Alterations of glucose in the ittcubatiou media have a specific effect, 011 proinsulin-insulin biosynthesis relative to total islet protein as proiusulitt-insulin accounts for about 2% at 2.8 tnht glucose aud 18% at 16.6 tnht glucose (Table II).
Previous studies (10) using actinomycin 11 suggested that glucose alters the amount of proinsulin tnRKA sgttthcsizcd during a 2-hour incubation.
The esperimcttt presented in Table II, nhcrc low dose cycloheximide partially inhibited elongation, also suggests that at 16.6 rnaf glucose there is 2 to 3 titnes more proittsulin mRNA available for translation, itt contrast to the cffcct of glucose on tnRNA for total islet protein.
When the ribosomes saturate the mRNA, and the atnourtt of proinsulin synthesized should be proportional to the amount of proittsulin tnRNA available, there is still a specific effect of glucose, so that proinsulin accounts for 6 to 7 y0 of islet protein syttthesis. 111 addition to a 2-to 3-fold increase in proittsuliu mRNA, this csperiment confirms previous actittomycin I> studies that there is a 5-to B-fold increase in translation of available proinsulitt mRNA. It cannot be detertnined whether the trattslational effect is on initiation or elongation of proittsulitt, but, specific immuttoprccipitation of nascent and soluble proinsulin, as described by I'almitcr for determination of initiation and elongation rates of chick oviduct proteins (19), is currently being evaluated.