High Uptake of myo-Inosit’ol by Rat Pancreatic Tissue in Vitro Stimulates Secretion*

A study by Hokin-Neaverson, Sadeghian, K., A. L., and Eisenberg, F. (1975) Biochem. Biophys. Res. Commun. 6’7, 1537-1544, demonstrates that free myo-inositol in the pancreas is significantly increased during intense cholinergic stimulation of secretion. Incubation of rat pancreatic tissue in medium with 100 mM myo-inositol increases lo-fold the endogenous content of free myo-inositol and elicits a prompt and sustained 50% increase in the rate of release of amylase activity. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis reveals that the electrophoretic pattern of the protein mixture released in the presence of 100 mM myo-inositol is the same as that of the secretory output released in the presence of 10 FM carbamylcholine. Microscopic examination of tissue pieces indicates that there is no significant decrease in the zymogen granule content of the pancreatic acinar cells during incubation in medium with 100 mM myo-inositol. Jamieson, J. D., and Palade, G. E. (1967) J. Cell Biol. 34, 597-615, have shown that pulse-labeled secretory proteins in guinea pig pancreas first appear in zymogen granules 1 hour postpulse, becoming maximally accumulated in these storage sites by 2 hours postpulse.

From the Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania A recent study by Hokin-Neaverson, M., Sadeghian, K., Majumder, A. L., and Eisenberg, F. (1975) Biochem. Biophys. Res. Commun. 6'7, 1537-1544, demonstrates that free myo-inositol in the pancreas is significantly increased during intense cholinergic stimulation of secretion. Incubation of rat pancreatic tissue in medium with 100 mM myo-inositol increases lo-fold the endogenous content of free myo-inositol and elicits a prompt and sustained 50% increase in the rate of release of amylase activity. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis reveals that the electrophoretic pattern of the protein mixture released in the presence of 100 mM myo-inositol is the same as that of the secretory output released in the presence of 10 FM carbamylcholine.
Microscopic examination of tissue pieces indicates that there is no significant decrease in the zymogen granule content of the pancreatic acinar cells during incubation in medium with 100 mM myo-inositol. Jamieson, J. D., and Palade, G. E. (1967) J. Cell Biol. 34, 597-615, have shown that pulse-labeled secretory proteins in guinea pig pancreas first appear in zymogen granules 1 hour postpulse, becoming maximally accumulated in these storage sites by 2 hours postpulse. myo-Inositol (100 mM) stimulates release of pulse-labeled secretory proteins only if incubation in medium with 100 mM myo-inositol is initiated anytime during the first 80 min postpulse. The findings thus indicate that a high uptake of myo-inositol by rat pancreatic tissue in vitro selectively stimulates the release of just those secretory proteins being packaged in newly forming zymogen granules.
In 1953, Hokin and Hokin first reported that there is a marked increase in phosphatidylinositol synthesis associated with cholinergic stimulation of amylase secretion by the pancreas (1). Subsequent studies have provided a detailed description of phosphatidylinositol metabolism in relation to pancreatic secretion. Secretagogues elicit not just a marked increase in phosphatidylinositol synthesis, but also a net decrease in the phosphatidylinositol content of the pancreas (2). Increased phosphatidylinositol synthesis occurs promptly (as soon as 2 min after stimulation) and almost exclusively in the rough ER' and Golgi complex (3). Calcium ion deprivation, which markedly reduces amylase secretion, only partially inhibits increased phosphatidylinositol synthesis (4). The concentrations of secretagogues required to stimulate phosphatidylinositol synthesis and degradation are higher than those sufficient to stimulate amylase secretion (5).
The generally accepted interpretation of these findings is that increased phosphatidylinositol turnover and stimulated enzyme secretion do not have a causal relationship and therefore are two phenomena which merely occur in parallel in response to different types of secretagogues. Lapetina  Michell recently suggested, however, that a physiologically significant consequence of enhanced phosphatidylinositol turnover may be the production and accumulation of intracellular cyclic inositol 1,2-monophosphate (6). In response to this proposal, Hokin-Neaverson et al. have found that free inositol is the sole water-soluble product of acetylcholine-stimulated phosphatidylinositol degradation; free inositol increases in amounts which are equal to the net loss of phosphatidylinositol content of the tissue (7). There is thus direct evidence that intense physiological stimulation of secretion by the pancreas is accompanied by a significant increase in the intracellular level of free inositol.
Our experimental approach to the hypothesis that the significance increase in free inositol which occurs during intense physiological stimulation plays a role in intense stimulated secretion has been to determine the rate of uptake of [3H]myo-inositol by rat pancreatic tissue in vitro as a function of the concentration of myo-inositol in the medium and the rate of release of pulse-labeled secretory proteins as a function of the amount of myo-inositol taken up. The results of our studies demonstrate that high extracellular concentrations of myo-inositol, in the absence of physiological secretagogues, (a) increase free inositol in rat pancreatic tissue to levels which are attained only during exhaustive physiological stimulation and (b) stimulate secretion.

Tissue Preparation
Pancreatic tissue for each experiment was obtained from a single male albino rat, weighing 200 to 300 g, killed by a blow to the head after lipid-Tissue pieces (total wet weight 15 to 25 rug) were kept for 15 min in 2.5 ml of ice-cold oxygenated KHBS containing radio-labeled myo-inositol at a given concentration before initiating incubation at 37". After a selected interval of incubation at 37", the medium was removed and the tissue pieces quickly washed with 2.5 ml of fresh KHBS before being homogenized with a Teflon homogenizer in 5.0 ml of ice-cold 10% (W/U) trichloroacetic acid. The homogenate was centrifuged at 2500 x g for 10 min and the radioactivity of an aliquot of the supernatant determined as a measure of the free myo-inositol in the tissue pieces. The ice-cold trichloroacetic acid-precipitable material was washed once with ice-cold 5% trichloroacetic acid and then suspended in 5.0 ml of ice-cold chloroform/methanol (2/l, u/u) After 2 to 3 hours at 4', the suspension was blended on a Vortex mixer with 1.05 ml of 0.05 M KC1 and centrifuged at 30,000 x g for 10 min.

Pulse-labeling Secretory
Proteins-Our protocol was a modification of that employed by Jamieson and Palade (11) in their studies of the intracellular transport of secretory proteins in the pancreatic exocrine cell. All the tissue pieces for each experiment were kept in ice-cold oxygenated KHBS before being transferred to 2.0 ml of fresh KHBS for incubation at 37" for 5 min. Thirty microcuries of L-[3H]leucine (0.5 to 0.8 pM) were added and incubation at 37" continued for another 2 min, after which the tissue pieces were washed three times with 5.0 ml of fresh KHBS containing 0.4 IIIM L-leucine (postpulse medium) and finally divided among a number of 25-ml Erlenmeyer flasks containing 5.0 ml of fresh postpulse medium for further incubation at 37". The washing of the tissue pieces and their transfer to postpulse medium generally consumed 2 to 4 min. Additions of myo-inositol or carbamylcholine did not occur until the tissue pieces were transferred to postpulse medium. After selected intervals of postpulse incubation, 0.5 ml of postpulse medium was removed from each set of tissue pieces and the pulse-labeled secretory proteins released in this aliquot precipitated by 0.75 ml of ice-cold 30% trichloracetic acid after the addition of 1.0 ml of ice-cold albumin solution (4 mg of albumin/ml) as carrier. The trichloroacetic acid precipitates were pelleted and the pellets washed twice with ice-cold 5% trichloroacetic acid; the pellets were then dissolved in 0.1 ml 0.5 N NaOH and 0.5 ml of NCS tissue solubilizer for determination of radioactivity. At the end of the postpulse incubation, all the postpulse medium was removed from each set of tissue pieces and the tissue pieces homogenized with a Teflon homogenizer in 5.0 ml of ice-cold 10% trichloroacetic acid. These trichloroacetic acid precipitates were pelleted and the pellets washed twice with ice-cold 5% trichloroacetic acid; the pellets were then dissolved in 0.2 ml of 0.5 N NaOH and 1.0 ml of NCS tissue solubilizer for determination of the amount of pulse-labeled secretory proteins still retained by the tissue pieces. with Siliclad, in order to minimize adsorption of secretory proteins to glass. Every hour during the 3.hour incubation the medium was entirely replaced with fresh medium; the medium collected was chilled and supplemented with benzamidine to a final concentration of 1 rn~ (to prevent zymogen activation (13)) The media collected from the three l-hour intervals were mixed and cleared of cellular debris by centrifugation at 100,000 x g for 1 hour at 4'. The cleared supernatant was dialyzed against 50 IIIM ammonium bicarbonate and 1 rn~ benzamidine (pH 7.6) for 4 to 5 hours at 4" and lyophilized. The resulting white powder was dissolved in 1% SDS, 1% B-mercaptoethanol, 10% glycerol, 0.01% bromphenol blue (tracking dye), and 60 mM Tris-HCl (pH 6.7), boiled for 2 min and subjected to gel electrophoresis in 5 mm inner diameter glass tubes, using Maizel's (14) discontinuous "sodium dodecyl sulfate-disc" system and 13% acrylamide in the running gel. Gels were run at room temperature at a constant 80 V and then stained at room temperature in 0.2% Coomassie brilliant blue R, dissolved in 25% Isopropyl alcohol and 10% acetic acid. Background destaining was accomplished in 10% isopropyl alcohol and 10% acetic acid.

Materials
All chemicals were of reagent grade and were purchased from the following sources: ( tissue for every pmol of myo-inositol per ml in the medium. Since there is 0.66 pmol of myo-inositol per g of tissue, wet weight, in the unstimulated rat pancreas (15), we find that free inositol in the rat pancreas is increased by approximately 10, 100, and 1000% during incubation in media containing myoinositol at concentrations of 1, 10, and 100 mM, respectively. Stimulation of Secretion by High myo-Inositol Concentrations-To examine any effects of extracellular myo-inositol concentration on the secretory process, tissue pieces were pulse-labeled with L-[3H]leucine and then transferred to postupulse medium containing varying concentrations of myoinositol. This protocol ensured that in each separate experiment the secretory proteins were pulse-labeled in the tissue pieces under identical conditions and that the stimulatory effect of carbamylcholine could be directly compared with those of varying concentrations of myo-inositol.
The kinetics of release of pulse-labeled secretory proteins during 4 hours postpulse stimulation with 10 pM carbamylcholine are shown in Fig. 3. Data collected from six separate experiments show that 10 pM carbamylcholine elicits through-out the postpulse period a 2.6 * 0.5-fold increase in the rate of release of incorporated radiolabeled leucine. Fig 4 shows the per cent release of pulse-labeled secretory proteins after 4 hours postpulse stimulation with varying concentrations of carbamylcholine.
This dose-response curve compares very favorably with that determined by Scheele and Palade (16) using their pancreatic lobule system, and shows that the concentration of carbamylcholine which elicits maximal stimulation of release of pulse-labeled secretory proteins is 10 PM. Fig. 5 shows the per cent release of label after 4 hours incubation in postpulse medium containing varying concentrations of myo-inositol.
The per cent of label released in postpulse medium with 1 /.LM to 40 mM myo-inositol is the same as that released under control conditions. At a concentration of 60 mM myo-inositol there is a 1.9.fold increase in the label released and at concentrations of 80 and 100 mM there is a 2.25-fold increase. The per cent of label released in the presence of both 100 mM myo-inositol and 10 pM carbamylcholine is approximately a third greater than that released by either agent alone. We thus find that the extracellular myoinositol concentrations which result in a minimum lo-fold increase in the endogenous intracellular level of free inositol are those which stimulate secretion.
The stimulation of pancreatic secretion elicited by high extracellular concentrations of myo-inositol appears similar in several respects to the stimulation elicited by carbamylcholine. First, there is no difference in the composition of the secretory output released in the presence of 100 mM myo-inositol from that released under control or secretagogue-stimulated conditions (17). Fig. 6 shows that the electrophoretic pattern of the protein mixture released in the presence of 100 mM myoinositol is the same as those for the mixtures released in the presence of 10 pM carbamylcholine or in the absence of any stimulatory agent. Furthermore, the relative proportions at which the major protein components are released in the presence of 100 mM myo-inositol appear similar to those released in the presence of 10 ELM carbamylcholine.
Secondly, 100 mM myo-inositol and 10 pM carbamylcholine each evoke similar kinetics of stimulated discharge of amylase activity, the only difference being in the magnitude of the stimulation (Fig. 7). The rate of amylase release in the presence of both 100 mM myo-inositol and 10 pM carbamylcholine is the same as that which occurs in the presence of 10 FM carbamylcholine alone. Thirdly, the stimulatory effect of 100 mM myo-inositol exhibits the same Ca2+ dependency as that exhibited by 10 WM carbamylcholine.
The results in Table I show that incubation of tissue pieces, pulse-labeled with L-[3H]leucine, in Ca'+-free KHBS with EDTA during the first 30 min of the postpulse period, completely inhibits stimulated secretion by both agents.
There are, however, certain characteristics of myo-inositol stimulation which distinguish it from carbamylcholine stimulation. First, we find that 100 mM myo-inositol-stimulated release of pulse-labeled secretory proteins does not begin until 2 hours postpulse, in both the presence ,and absence of carbamylcholine (Fig. 8). In both unstimulated and maximally physiologically stimulated tissue, the initial traces of a packet of pulse-labeled secretory proteins do not appear in zymogen granules until 30 min postpulse; an hour is required for the majority of the packet to accumulate in these storage sites (18). The 2-hour delay in the onset of 100 mM myo-inositolstimulated release suggests that such a high extracellular  . 3 (center). Kinetics of release of pul se-l abel ed proteins under myo-inositol concentration alters the kinetics of the intracellular transport of secretory proteins through the rough ER and Golgi complex, extending their normal transit time through these compartments. Secondly, we find that 100 mM myo-inositol cannot stimulate the release of pulse-labeled secretory proteins after they have accumulated in mature zymogen granules. Jamieson and Palade have shown that secretory proteins in pancreatic exocrine cells are transported from the site of their synthesis in the rough ER to condensing vacuoles of the Golgi complex where they are intensively concentrated, the condensing vacuoles being converted to zymogen granules (11,18,19). They have shown by electron microscope autoradiography that the time course of the passage of pulse-labeled secretory proteins through these intracellular compartments is such that during the first 10 min postpulse, most of the pulse-labeled secretory proteins are confined to regions containing rough ER; by 20 min to the periphery of the Golgi complex; by 40 min to the condensing vacuoles, and finally from 60 to 120 min postpulse, they accumulate in the zymogen granules. Accordingly, we measured the stimulatory effects of 100 mM myo-inositol and 10 pM carbamylcholine when applied at increasing times after initiation of the postpulse incubation. Data collected from four separate experiments show that if transfer to media with 10 PM carbamylcholine or 100 mM myo-inositol occurs at any time from 40 to 80 minutes after the pulse-label period, the amount of label released above controls is 90 to 110% that released if the transfer is made immediately (5 min) after the pulse-label period (Table II). There is, however, a dramatic decrease in the capacity of 100 mM myo-inositol to stimulate the release of a packet of pulse-labeled secretory proteins during the period 80 to 100 min after their synthesis. In the case of 10 pM control and carbamylcholine-stimulated conditions. carbamylcholine, the decrease in its stimulatory effect is less severe, as the amount of label released above controls is 70-80% that released by tissue pieces transferred to medium with 10 pM carbamylcholine immediately after the pulse-label period.
Light microscopic examination supports this evidence that 100 mM myo-inositol stimulates the release of secretory proteins from a pool other than that of mature zymogen granules. Figs. 9 and 10 show that in tissue pieces which have been incubated for 4 hours in medium with 100 mM myo-inositol, both tissue architecture and cellular ultrastructure remain well preserved; there is, however, no significant decrease (certainly not more than 25%) in the zymogen granule content of the cells. We find, as expected, that there is a marked depletion of zymogen granules in most cells in tissue pieces which have been incubated for 4 hours in medium with 10 pM carbamylcholine, either in the presence or absence of 100 mM myo-inositol (Fig.  11 a and b).
Thirdly, stimulation of secretion by 100 mM myo-inositol is associated with an inhibition of [3H]myo-inositol incorporation into phosphatidylinositol.
The effect of extracellular myoinositol concentration on the rate of incorporation of [3H]myoinositol into phopholipid was examined during the same experiments conducted to study [3H]myo-inositol uptake. Fig.  12 shows that incorporation proceeds linearly during 2 hours incubation at 37" in KHBS with myo-inositol concentrations ranging from 1 pM to 10 mM; the rate of incorporation decreases as the myo-inositol concentration is increased. There is no significant incorporation in medium with 100 mM myo-inositol. It is difficult to assess from these results the absolute rates of incorporation in the tissue pieces since we do not know the specific activities (&i/~mol) of the intracellular myo-inositol of myo-Inositol Stimulate Secretion pools. We can, however, estimate relative rates of incorporation. The pancreas maintains in uiuo a rather considerable permeability barrier to intracellular free inositol (15); it is therefore likely that the tissue pieces maintain their endogenous content of free inositol and any increases resulting from uptake of myo-inositol from the extracellular medium. We adjusted the specific activities of ['H]myo-inositol in the media so that they were inversely proportional to the myo-inositol concentration (i.e. the specific activity in medium with 1 Jo myo-inositol was adjusted to be 1000 times greater than that in medium with 1 mM myo-inositol).
We followed this procedure so that the amount of microcuries of radiolabeled myoinositol taken up per g of tissue would be independent of the myo-inositol concentration in the medium, since myo-inositol uptake is directly proportional to myo-inositol concentration. We would therefore expect that if the amount of radiolabeled myo-inositol taken up is small relative to the endogenous content, the rate of incorporation of [3H]myo-inositol into phospholipid should not vary significantly with the concentration of myo-inositol in the medium. The data shown in Fig. 12 indicate, however, that there is a 70% decrease in the rate of incorporation in tissue pieces incubated in medium with 1 mM myo-inositol compared to that by tissue pieces incubated in 1 PM myo-inositol; in 1 mM myo-inositol the intracellular free inositol increases by only approximately 10%. We thus are led to conclude that the absence of any significant incorporation of ['H]myo-inositol into phospholipid by tissue pieces incubated in medium with 100 mM myo-inositol is not a consequence of the relatively low specific activity of the intracellular myo-inositol pools in these tissue pieces, but rather a result of a marked inhibition of phosphatidylinositol synthesis. Finally, we find that if atropine, a specific pharmacological antagonist of acetylcholine (21, is added at a concentration of and KHBS with 10 PM carbamylcholine (Gel 2l, 100 mM myo-inositol (Gel 3). and 10 PM carbamylchbline and 100 rni myo-inosittol (Gel 4) were prepared for electrophoretic fractionation as described in the text.
The arrows indicate the six major protein components common to all of the mixtures; their approximate molecular weights are 67.000. 53.500.
38,000, 33.000, 29,000, and 19,500. The 53.500-molecular weight component most likely represents amylase (20, 211. The dark marks at the lower ends of the gels represent the points where the gels were speared with India ink to mark the dye front at the end of the electrophoretic run. 10 PM to postpulse medium, it has no effect on 106 mM myo-inositol-stimulated secretion. As expected, atropine completely inhibits the stimulatory effect of carbamylcholine. DISCUSSION We believe that the experimental findings cited here and those which have been reported by Hokin and Hokin and their colleagues (l-5) not only tentatively account for myo-inositolstimulated secretion, but, more importantly, implicate a central role of enhanced phosphatidylinositol turnover in secretagogue-stimulated secretion. If we compare the doseresponse curves determined by ourselves and Scheele and Palade (16) for the stimulated discharge of pulse-labeled secretory proteins by carbamylcholine, with that determined by Hokin (51 for the stimulated discharge of amylase, we find the data is consistent in demonstrating that the carbamylcholine concentrations which maximally stimulate the discharge of pulse-labeled secretory proteins are the same as those which stimulate phosphatidylinositol turnover; they are, however, 1 to 2 orders of magnitude greater than those eliciting maximal stimulation of amylase discharge. We suggest that these  myo-inositol (m-m) was .determined as described in the text. The results are given as per cent of the sum of enzymatic activity in tissue pieces and medium and represent the average of three experimerits.
The  incubation (18). After 30 min postpulse, the medium for each set of ______~~ relationships are most simply understood by postulating that there occur within pancreatic acinar cells two pools of secretory proteins available for immediate discharge, and the regulatory system governing the stimulus-secretion coupling of one pool is activated by enhanced phosphatidylinositol turnover, while the regulatory system of the other pool is not.
Our proposal is therefore as follows. There occur within pancreatic acinar cells two pools of digestive enzymes and zymogens which can be tapped for secretion by physiological secretagogues. They are the stream of newly forming zymogen granules and the reservoir of mature zymogen granules. The threshold of stimulation for the former pool is greater than that for the latter one. The regulatory system governing the coupling between stimulation and secretion of the pool of mature zymogen granules is maximally activated in uctro by acetylcholine or its chemical analogue, carbamylcholine, at a concentration of 0.1 PM. At such a relatively low concentration of these secretagogues, there does not occur (a) any significant alteration of phosphatidylinositol metabolism nor (b) any activated extrusion of newly forming zymogen granules; these granules continue to became included into the pool of mature zymogen granules in the same manner as they do in unstimulated tissue. If the acetylcholine or carbamylcholine concentration is raised to 1 to 10 NM, there occurs a marked enhancement of phosphatidylinositol turnover. The dramatic decrease in the phosphatidylinositol content of the cells results in a significant (approximately 100%) increase in the intracellular level of free inositol and this increase, in turn, elicits a rapid and selective stimulation of the extrusion of newly forming zymogen granules via a Caz+-dependent process. According to this proposal, physiological secretagogues, both the similarities and the differences between the two during maximal stimulation, increase secretion by tapping both pools of secretory proteins. myo-Inositol-stimulated secre-stimulatory effects. Stimulation of amylase secretion by myotion, however, draws upon the pool of newly forming zymogen inositol should be as prompt but not as great as that elicited by granules only. This distinction in the mechanics of carbamyl-carbamylcholine since in this instance the secretory protein occurs within both pools from the onset of stimulation onward. choline-and myo-inositol-stimulated secretion accounts for When we measure the release of secretory proteins pulse- FIG. 11. a and b, low power light micrographs of two sections of a tissue piece incubated for 4 hours in KHBS with 100 mM myo-inositol and 10 PM carbamylcholine.
Within the tissue pieces regions can be found in which (a) symogen granule extrusion from the cells has been almost complete and (b) zymogen granule extrusion has been less extensive.
Comparison of b and Fig. 9 (the tissue pieces shown in both this figure and Fig. 9 were obtained from the same pancreas), shows that even in these cells which still retain some zymogen granules after 4 hours incubation in medium with 100 msr myo-inositol and 10 PM carbamylcholine, there has nevertheless been a significant reduction in the symogen granule content elicited by the additional presence of 10 pM carbamylcholine.
x 630. labeled in vitro, however, we are monitoring a packet of secretory proteins which must be transported from the rough ER and processed through the Golgi complex before entering the pool of newly forming zymogen granules. Not until this temporally defined packet has begun to enter this first pool, a journey which on the average requires 1 hour, can both carbamylcholine and myo-inositol stimulate its release. If, now, the packet is allowed to progressively accumulate in the second pool of mature zymogen granules, a process which spans the second postpulse hour, then only carbamylcholine can stimulate its release. These predictive features of the proposal are consistent with the results shown in Fig. '7 and Table  II.
The proposal also accounts for the difference in the sensitivity of the two stimulatory effects to inhibition by atropine. Since 10 pM atropine has no effect on the IO-fold increase in the intracellular free inositol which occurs during incubation in medium with 100 mM myo-inositol,2 we would expect atropine to have no effect on myo-inositol-stimulated secretion. Carbamylcholine, on the other hand, elicits an increase in intracellular free inositol only indirectly through its capacity to stimulate phosphatidylinositol degradation, and this response, as Hokin-Neaverson (2) has recently shown, is blocked by atropine.
We believe that there is sufficient evidence to suggest at this time, but not yet prove, that enhanced phosphatidylinositol turnover diverts the inclusion of newly forming zymogen granules from the pool of mature zymogen granules and instead stimulates their immediate discharge, this action being mediated by free inositol. We are planning to confirm by electron microscope autoradiography that incubation of pancreatic tissue in medium with 100 mM myoinositol (a) alters the kinetics of the intracellular transport of secretory proteins and (5) preferentially stimulates the discharge of newly forming zymogen granules. It is our belief that such evidence will more directly validate the view that phosphatidylinositol metabolism plays important roles in the secretory process, especially with respect to the role we have discussed here.