Heterogeneity and contact-dependent regulation of amylase release by individual acinar cells

We have used a reverse hemolytic plaque assay to investigate the amylase release of single and aggregated pancreatic acinar cells. We have found that a minority of single acinar cells released detectable amounts of amylase under basal conditions and were modestly stimulated, in a dose-dependent manner, during a 30-min exposure to concentrations of carbamylcholine (CCh) ranging from 10^?8 to 10^?5 M. This stimulation was largely accounted for by the recruitment of additional secreting cells, rather than by a significant increase in their individual secretory output. We have also observed that aggregates comprising two to five acinar cells secreted more frequently and released more amylase than single acinar cells in the presence of each of the CCh concentrations tested. Under both basal conditions and following CCh stimulation, the proportion of secreting aggregates and their amylase output increased linearly with the aggregate size. Under basal conditions as well as in the presence of secretagogue concentrations in the 10^?8?10^?7 M range, individual cells contributed similarly to amylase secretion whether they were single or part of aggregates. By contrast, following stimulation by 10^?6?10^?5 M CCh, aggregated cells showed a much higher average secretion than single cells. Investigating the mechanism of this contact-dependent effect, we found that 10^?3 M heptanol did not significantly modify the secretion of single cells and markedly promoted the basal amylase release of acinar cell pairs. This effect was associated with a marked reduction in gap junctional communication between acinar cells, as evaluated by microinjection of Lucifer yellow, and was not observed during exposure to high concentrations of CCh, which also reduced junctional communication. These data show that pancreatic acinar cells are intrinsically heterogeneous in their ability to release amylase and that their basal as well as stimulated secretion are promoted by the establishment of direct intercellular contacts. Our experiments also suggest that junctional coupling contributes to the contact-dependent mechanism which enhances the recruitment of secreting cells and their individual output. These observations strengthen the view that direct interactions between acinar cells are essential in the control of pancreatic secretion. © 1994 Wiley-Liss, Inc.     

Heterogeneity and contact-dependent regulation of hormone secretion by individual B cells

A reverse hemolytic plaque assay was developed to visualize insulin release from individual adult pancreatic B cells. Cells obtained by mechanical dispersion of isolated rat islets of Langerhans were mixed with protein A-coated sheep red blood cells and incubated in the presence of an anti-insulin serum, under conditions known to affect insulin release. The cell mixture was further incubated with complement and finally fixed. Insulin release was revealed by the presence of hemolytic plaques which resulted from the complement-mediated lysis of red blood cells bearing insulin-anti-insulin complexes bound to protein A. Quantitation of hemolytic plaques around trypan blue-unstained and immunohistochemically identified B cells showed that stimulation of insulin release results in the recruitment of increasing numbers of secreting B cells as well as in the enhanced response of individual B cells. Reverse changes occur upon inhibition of insulin release. Comparison of freshly dispersed and one-day-cultured preparations did not reveal significant differences in the secretory response of undamaged B cells. In both preparations, single B cells responded to secretagogues in smaller proportions and to a lesser extent than clusters in which B cells had either maintained or restored contacts and junctional communication with their neighbours. However, the overall preponderant response of clusters was less than expected from the number of individually secreting B cells they contained. The data show that B cells are heterogeneous in terms of their ability to release insulin and provide evidence that cell-to-cell adhesion and/or junctional communication regulate hormone secretion from individual B cells.     

Glucose-stimulated insulin release by individual pancreatic beta cells: potentiation by glyburide.

To investigate the effect of glyburide on insulin secretion by individual beta cells from normal rats, we employed a reverse hemolytic plaque assay. Pancreata were harvested from female Wistar-Furth rats, the pancreatic islets isolated, and the latter dispersed into single cells. These cells were mixed with protein A-coated ox erythrocytes, the mixture was placed in a Cunningham chamber in the presence of insulin antiserum, and the cells were exposed to the various test substances. Having developed hemolytic plaques around the insulin-secreting cells with complement, the percentage of plaque-forming cells was determined and the plaque areas (reflecting the amount of insulin secreted) were quantitated. For the purpose of validation, we demonstrated that (i) plaque-forming (but not nonplaque-forming) cells could be identified as insulin secreting by an independent immunofluorescent technique, (ii), plaques did not form if insulin antiserum was deleted from the preparation, (iii) plaques failed to develop if insulin antiserum was preabsorbed with insulin, and (iv) incubation with non-protein A-coated RBC or omission of complement resulted in no plaque formation. In addition, both the percentage of plaque-forming cells and the mean plaque are increased upon exposure to glucose (0.75-20 mM) in a concentration-dependent manner at 5- and 60-min incubation times. Moreover, somatostatin suppressed the percentage of plaque-forming cells and diminished the mean plaque area of cells which continued to secrete insulin in response to glucose. Exposure of cells to 100 nM glyburide in the presence of 5 mM or 20 mM glucose had no effect on the percentage of plaque-forming cells present at 5 min or 60 min. Similarly, glyburide did not alter mean plaque area at 5 or 60 min when cells were co-incubated with 5 mM glucose. However, mean plaque area was markedly enhanced at 5 and 60 min in response to glyburide and 20 mM glucose. These results demonstrate that glyburide (i) does appear to enhance insulin secretion by an effect directly on the pancreatic beta cell; (ii) does not act by recruiting previously noninsulin-secreting cells into a secretory pool; (iii) does not potentiate the effect of glucose, at fed concentrations, on insulin secretion by individual cells; but (iv) does augment insulin secretion by beta cells stimulated with supraphysiologic concentrations of glucose.     

Thapsigargin-sensitive cationic current leads to membrane depolarization, calcium entry, and insulin secretion in rat pancreatic beta-cells

Glucose-induced insulin secretion by pancreatic beta-cells depends on membrane depolarization and [Ca2+]i increase. We correlated voltage- and current-clamp recordings, [Ca2+]i measurements, and insulin reverse hemolytic plaque assay to analyze the activity of a thapsigargin-sensitive cationic channel that can be important for membrane depolarization in single rat pancreatic beta-cells. We demonstrate the presence of a thapsigargin-sensitive cationic current, which is mainly carried by Na+. Moreover, in basal glucose concentration (5.6 mM), thapsigargin depolarizes the plasma membrane, producing electrical activity and increasing [Ca2+]i. The latter is prevented by nifedipine, indicating that Ca2+ enters the cell through L-type Ca2+ channels, which are activated by membrane depolarization. Thapsigargin also increased insulin secretion by increasing the percentage of cells secreting insulin and amplifying hormone secretion by individual beta-cells. Nifedipine blocked the increase completely in 5.6 mM glucose and partially in 15.6 mM glucose. We conclude that thapsigargin potentiates a cationic current that depolarizes the cell membrane. This, in turn, increases Ca2+ entry through L-type Ca2+ channels promoting insulin secretion.     

Beta cell contact and insulin release

Insulin secretion by intact islets, dispersed islet cells and dispersed cells allowed to reaggregate was compared in perifusion. Although single cells and aggregates showed basal insulin secretion and a prompt response to glucose challenge, basal secretion, peak insulin secretion and total insulin secretion during a 60 minute stimulation were profoundly less than those activities of intact islets. These results suggest that dispersed beta cells are responsive to glucose as a secretagogue, but the magnitude of the response is greatly diminished and not restored by simple cell contact.     

Co-ordinated Ca(2+)-signalling within pancreatic islets: does beta-cell entrainment require a secreted messenger

Isolated beta-cells are heterogeneous in sensory, biosynthetic and secretory capabilities, however, to enable efficient and appropriate secretion, cellular activity within the intact islet is synchronised. Historically, the entrainment of activity to a common pattern has been attributed to gap-junction mediated cell-to-cell communication. Although clearly influential, the possibility remains for other local synchronising mechanisms. In this study, we have used small clusters of insulin-secreting MIN6 cells to assess how contact-dependent, homotypic interactions between cells influences nutrient- and non-nutrient- evoked Ca(2+)-handling and insulin secretion, and to determine whether a secreted product plays a role in the synchronisation of oscillatory activity. Tolbutamide evoked a concentration-dependent recruitment of active cells within cell clusters, both in terms of numbers of cells and amplitude of the evoked Ca(2+)-response. The change in [Ca(2+)](i) was characteristically oscillatory above a mean elevated plateau, and was in phase between member cells of an individual cluster. Even at maximal concentrations (100 microM) some cells within a cluster responded before their immediate neighbours. Subsequent oscillatory behaviour then became entrained between member cells within that cluster. Inhibiting exocytosis using the microtubule inhibitors vincristine and nocodazole, or the adrenergic agent noradrenaline, did not prevent tolbutamide-evoked oscillatory changes in [Ca(2+)](i) but did reduce the probability of obtaining synchronous activity within an individual cluster. Above a threshold glucose concentration, the number of cells secreting insulin increased, without a commensurate change in secretory efficiency. This recruitment of cells secreting insulin mirrored Ca(2+) data that showed a glucose-dependent increase in cell number, without a change in the mean basal-to-peak change in [Ca(2+)](i). Together these data suggest that synchronised behaviour in MIN6 cells is dependent, in part, on a secreted factor that acts in a local paracrine fashion to recruit heterogeneous individual cellular activity into an organised group response.     

A diffusion-adsorption model for the computation of the amount of hormone around a secreting cell, detected by the reverse hemolytic plaque assay

The reverse hemolytic plaque assay enables the detection of secretion products from individual cells in cultures by visualizing the plaques formed after complement-mediated hemolysis around the secreting cells. However, the precise quantitation of the amount of secretion remains problematic. In this study we propose a computation model for estimating the spreading of the secreted molecules, based on the underlying processes of diffusion and antigen adsorption by immobilized antibodies. The translational diffusion coefficient of rat prolactin at 37 degrees C, determined by laser light scattering, was 9.89 x 10(-7) cm2/s. The time-dependent concentration distribution around a constantly secreting cell in a flat quasi infinite layer, was derived from the diffusion equation, using an analytical approach based on Laplace transformation. The relations between plaque size, incubation time and secretion level were expressed as a function of the threshold concentration of secretion product that can be detected and the effective diffusion coefficient, taking antigen adsorption into account. We obtained very good agreement between observed and predicted results for plaque formation by dispersed prolactin secreting cells of 14-day-old female rat pituitaries. This study confirms the validity of the assumptions underlying the reverse hemolytic plaque assay, provided that the cell density is low, the incubation time is moderately long and the concentration of specific antiserum is sufficiently high.     

Cellular interaction between mouse pancreatic alpha-cell and beta-cell lines: possible contact-dependent inhibition of insulin secretion

The endocrine cells in the pancreatic islet have cellular communication between the heterotypic cells as well as the homotypic cells. The present study was conducted to elucidate the cellular interaction between pancreatic alpha cells and beta cells utilizing differentiated mouse cell lines (i.e., alphaTC clone 6 and betaTC cells). Co-culture of these two cell lines on a gyratory shaker generated numerous cellular aggregates of homogenous size within 48 h. Immunohistochemical staining for insulin and glucagon demonstrated that betaTC cells were located in the central core of each aggregate, while alphaTC cells formed a mantle layer surrounding the betaTC cells. This segregation was observed regardless of the ratios of the two cell types employed. Although glucagon at concentrations of 10(-8) M or higher stimulated insulin secretion from betaTC cells in both monolayer and aggregates, an increase in the ratio of alphaTC/betaTC cells in aggregate cultures was accompanied by a decrease in secreted insulin and a rise in intracellular insulin content of the betaTC component. The inhibitory effect of alphaTC cells on betaTC insulin secretion was not limited to aggregate culture, since insulin secretion from betaTC cells was also suppressed, and intracellular insulin content increased, by co-culture of alphaTC with betaTC cells in monolayer. On the other hand, the secreted and intracellular insulin of betaTC cells was not affected by alphaTC cells in a Transwell co-culture system in which direct cell-to-cell contacts were prevented by a semipermeable membrane that permitted chemical communication via medium metabolites. These data suggest that the insulin secretion from betaTC cells may be inhibited possibly as a result of the contact with alphaTC cells.     

Glucose refractoriness of beta-cells from fed fa/fa rats is ameliorated by nonesterified fatty acids

The aim of this study was to characterize the glucose responsiveness of individual beta-cells from fa/fa rats under ad libitum feeding conditions. Enlarged intact islets from fed fa/fa rats had a compressed insulin response curve to glucose compared with smaller islets. Size-sorted islets from obese rats yielded beta-cells whose glucose responsiveness was assessed by reverse hemolytic plaque assay to determine whether glucose refractoriness was caused by a decreased number of responsive cells or output per cell. In addition, the effects of palmitic acid on glucose-stimulated insulin secretion were assessed because of evidence that nonesterified fatty acids have acute beneficial effects. Two- to threefold more beta-cells from >250 microm diameter (large) islets than <125 microm diameter (small) or lean islets responded to low glucose. Increasing the glucose (8.3-16.5 mM) induced a >10-fold increase in recruitment of active cells from small islets, compared with only a 2.6-fold increase in large islets. This refractoriness was partially reversed by preincubation of the cells in low glucose for 2 h. In addition, secretion per cell of the large islet beta-cell population was significantly reduced compared with lean beta-cells, so that the overall response capacity of large but not small islet beta-cells was significantly reduced at high glucose. Therefore, continued near-normal function of the beta-cells from small islets of fa/fa rats seems crucial for glucose responsiveness. Incubation of beta-cells from large islets with palmitic acid normalized the secretory capacity to glucose mainly by increasing recruitment and secondarily by increasing secretion per cell. In conclusion, these studies demonstrate refractoriness to glucose of beta-cells from large islets of fa/fa rats under ad libitum feeding conditions. When acutely exposed to nonesterified fatty acids, islets from fa/fa rats have a potentiated insulin response despite chronic elevation of plasma lipids in vivo.     

Evaluation of bottlenecks in the late stages of protein secretion in Bacillus subtilis.

Despite a high capacity for secretion of homologous proteins, the secretion of heterologous proteins by Bacillus subtilis is frequently inefficient. In the present studies, we have investigated and compared bottlenecks in the secretion of four heterologous proteins: Bacillus lichenifomis alpha-amylase (AmyL), Escherichia coli TEM beta-lactamase (Bla), human pancreatic alpha-amylase (HPA), and a lysozyme-specific single-chain antibody. The same expression and secretion signals were used for all four of these proteins. Notably, all identified bottlenecks relate to late stages in secretion, following translocation of the preproteins across the cytoplasmic membrane. These bottlenecks include processing by signal peptidase, passage through the cell wall, and degradation in the wall and growth medium. Strikingly, all translocated HPA was misfolded, its stability depending on the formation of disulfide bonds. This suggests that the disulfide bond oxidoreductases of B. subtilis cannot form the disulfide bonds in HPA correctly. As the secretion bottlenecks differed for each heterologous protein tested, it is anticipated that the efficient secretion of particular groups of heterologous proteins with the same secretion bottlenecks will require the engineering of specifically optimized host strains.     

Myosin 5a controls insulin granule recruitment during late-phase secretion

We have examined the importance of the actin-based molecular motor myosin 5a for insulin granule transport and insulin secretion. Expression of myosin 5a was downregulated in clonal INS-1E cells using RNAinterference. Stimulated hormone secretion was reduced by 46% and single-cell exocytosis, measured by capacitance recordings, was inhibited by 42% after silencing. Silencing of Slac-2c/MYRIP, which links insulin granules to myosin 5a, resulted in similar inhibition of single-cell exocytosis. Antibody inhibition of the myosin 5a-Slac-2c/MYRIP interaction significantly reduced the recruitment of insulin granules for release. The pool of releasable granules independent of myosin 5a activity was estimated to approximately 550 granules. Total internal reflection microscopy was then applied to directly investigate granule recruitment to the plasma membrane. Silencing of myosin 5a inhibited granule recruitment during late phase of insulin secretion. In conclusion, we propose a model where insulin granules are transported through the actin network via both myosin 5a-mediated transport and via passive diffusion, with the former playing the major role during stimulatory conditions.     

Enzyme production by recombinant Trichoderma reesei strains.

The production of both homologous and heterologous proteins with the cellulolytic filamentous fungus Trichoderma reesei is described. Biotechnically important improvements in the production of cellulolytic enzymes have been obtained by genetic engineering methodology to construct strains secreting novel mixtures of cellulases. These improvements have been achieved by gene inactivation and promoter changes. The strong and highly inducible promoter of the gene encoding the major cellulase, cellobiohydrolase I (CBHI) has also been used for the production of eukaryotic heterologous proteins in Trichoderma. The expression and secretion of active calf chymosin is described in detail.     

Increase of gap junctions between pancreatic B-cells during stimulation of insulin secretion

The development of gap junctions between pancreatic B-cells was quantitatively assessed in freeze-fracture replicas of isolated rat islets under different conditions of insulin secretion. The results show that in resting B-cells, gap junctions are small and scarce but that these junctions increase when insulin secretion is stimulated. Both a short (90 min) stimulation by glucose in vitro and a prolonged (2.5 d) stimulation by glibenclamide in vivo raise the number of gap junctions; in addition, the glibenclamide stimulation causes an increase in the size of individual gap junctions. As a consequence, the total area occupied by gap junctions on the B-cell membrane and the ratio of this area to the cell volume were found significantly increased in the latter condition. The slight increase of these values observed after the glucose stimulation did not reach significance. These data indicate a change of gap junctions during the secretory activity of the pancreatic B-cells. The possibility that the coupling of the cells is affected by the treatment is discussed.     

Na channels and two types of Ca channels in rat pancreatic B cells identified with the reverse hemolytic plaque assay

The reverse hemolytic plaque assay (RHPA) was used to study the secretory properties of single rat pancreatic B cells, and to identify insulin-secreting cells for patch-clamp experiments. In secretion studies using the RHPA, we find that the percentage of secreting B cells and the amount of insulin secreted per B cell increase as the glucose concentration is raised from 0 to 20 mM. Using the whole-cell variation of the patch-clamp technique, we find that identified B cells have three types of channels capable of carrying inward current: (a) tetrodotoxin-sensitive, voltage-dependent Na channels, which are nearly completely inactivated at -40 mV, (b) fast deactivating (FD) Ca channels, and (c) slowly deactivating (SD) Ca channels. We have shown that Na channels are functionally significant to the B cell, because tetrodotoxin partially inhibits glucose-induced insulin secretion. The properties of FD and SD Ca channels differ in several respects. FD channels deactivate at -80 mV, with a time constant of 129 microseconds, they are half-maximally activated near +15 mV, they do not inactivate during 100 ms, they conduct Ba2+ better than Ca2+, and they are very sensitive to washout during intracellular dialysis. SD channels, on the other hand, deactivate with a time constant of 2.8 ms, they are half-maximally activated near -5 mV, they inactivate rapidly, they conduct Ba2+ and Ca2+ equally well, and they are insensitive to washout.     

Evaluation of Bottlenecks in the Late Stages of Protein Secretion in Bacillus subtilis

Despite a high capacity for secretion of homologous proteins, the secretion of heterologous proteins by Bacillus subtilis is frequently inefficient. In the present studies, we have investigated and compared bottlenecks in the secretion of four heterologous proteins: Bacillus lichenifomis α-amylase (AmyL), Escherichia coli TEM β-lactamase (Bla), human pancreatic α-amylase (HPA), and a lysozyme-specific single-chain antibody. The same expression and secretion signals were used for all four of these proteins. Notably, all identified bottlenecks relate to late stages in secretion, following translocation of the preproteins across the cytoplasmic membrane. These bottlenecks include processing by signal peptidase, passage through the cell wall, and degradation in the wall and growth medium. Strikingly, all translocated HPA was misfolded, its stability depending on the formation of disulfide bonds. This suggests that the disulfide bond oxidoreductases of B. subtilis cannot form the disulfide bonds in HPA correctly. As the secretion bottlenecks differed for each heterologous protein tested, it is anticipated that the efficient secretion of particular groups of heterologous proteins with the same secretion bottlenecks will require the engineering of specifically optimized host strains.     

Homologous pancreastatin inhibits insulin secretion without affecting glucagon and somatostatin release in the perfused rat pancreas.

The identification of pancreastatin in pancreatic extracts prompted the investigation of its effects on islet cell function. However, in most of the investigations to date, pig pancreastatin was tested in heterologous species. Since there is great interspecies variability in the amino acid sequence of pancreastatin, we have investigated the influence of rat pancreastatin on insulin, glucagon and somatostatin secretion in a homologous animal model, namely the perfused rat pancreas. During 5.5 mM glucose infusion, pancreastatin (40 nM) inhibited insulin secretion (ca. 40%, P less than 0.025) as well as the insulin responses to 10 mM arginine (ca. 50%, P less than 0.025) and to 1 nM vasoactive intestinal polypeptide (ca. 50%; P less than 0.05). Pancreastatin failed to significantly modify glucagon or somatostatin release under any of the above experimental conditions. In addition, a lower pancreastatin concentration (15.7 nM) markedly suppressed the insulin release evoked by 11 mM glucose (ca. 85%, P less than 0.05). Our present observations reinforce the concept that pancreastatin is an effective inhibitor of insulin secretion, influencing the B-cell function directly and not through an A-cell or D-cell paracrine effect.     

Cellular and subcellular expression of Golf/Gs and Gq/G11 alpha-subunits in rat pancreatic endocrine cells.

We studied the cellular and subcellular localization of Galpha-subunits in pancreas by immunocytochemistry. Golfalpha and G11alpha were specifically localized in islet insulin B-cells and glucagon A-cells, respectively. Gsalpha and Gqalpha labeling was more abundant in B-cells. The presence of Golfalpha in B-cells was confirmed by in situ hybridization. In B-cells, Golfalpha and Gsalpha were found in the Golgi apparatus, plasma membrane (PM) and, remarkably, in mature and immature insulin secretory granules, mainly at the periphery of the insulin grains. Gqalpha was detected on the rough endoplasmic reticulum (RER) near the Golgi apparatus. In A-cells, the Galpha-subunits were mostly within the glucagon granules: G11alpha gave the strongest signal, Gsalpha less strong, Gq was scarce, and Golf was practically absent. Gqalpha and Gsalpha immunoreactivity was detected in acinar cells, although it was much weaker than that in islet cells. The cell-dependent distribution of the Galpha-subunits indicates that the stimulatory pathways for pancreatic function differ in acinar and in islet B- and A-cells. Furthermore, the G-protein subunits in islet cell secretory granules might be functional and participate in granule trafficking and hormone secretion.