| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute for Molecular Bioscience (B.J.M., A.J.C., G.P.M.), Queensland Bioscience Precinct, The University of Queensland, Brisbane, Queensland 4072, Australia; The Kovler Diabetes Center (C.S., C.A., B.L.W., K.Y., C.J.R.), Department of Medicine, Section of Endocrinology, Diabetes, and Metabolism, University of Chicago, Chicago, Illinois 60637
Address all correspondence and requests for reprints to: Christopher J. Rhodes, Kovler Diabetes Center, Department of Medicine, Section of Endocrinology, Diabetes, and Metabolism, University of Chicago, 5841 South Maryland Avenue, MC 1027, Room N138, Chicago, Illinois 60637. E-mail: cjrhodes{at}uchicago.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
0.001), which by further electron microscopy-tomography analysis was found to be mostly contributed by microautophagic activity. This increased autophagic activity in Rab3A–/– mouse islet ß-cells was associated with a specific decrease in islet lysosomal-associated membrane protein 2 gene expression (P 0.05), at both the mRNA and protein expression levels. Lysosomal-associated membrane protein 2 is a documented negative regulator of autophagy. These findings indicate that the up-regulation of degradative pathways provides secretory-deficient endocrine cells with a compensatory mechanism for regulating their intracellular hormone content in vivo. These data may also have implications for the ß-cell’s response to diminished insulin secretion during the pathogenesis of type 2 diabetes. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
However, there is another contributing factor to consider in control of the ß-cell’s insulin stores. Normally, a ß-cell contains a relatively constant number of ß-granules (2, 6), but these are continually turned over, having an estimated half-life of 3–5 d (6, 7). Older ß-granules in the ß-cell’s storage pool that do not undergo exocytosis are eventually retired by intracellular degradation. This intracellular digestion of insulin can be attributed to two mechanisms. The first mechanism is referred to as crinophagy, whereby a ß-granule can fuse directly with a large lysosomal-related vacuolar compartment (often referred to as a "crinophagic body") resulting in degradation of the ß-granule lumenal content (7, 8, 9, 10). The second mechanism is autophagy, which can be subdivided into macroautophagy and microautophagy in terms of degradation of whole organelles (11, 12, 13). Macroautophagy is regarded as the primary autophagic mechanism for intracellular protein degradation in mammalian cells (13, 14). In terms of insulin granule degradation, macroautophagy would involve the formation of a membrane around a ß-granule to form a double membrane structure termed an "autophagosome." The ß-granule-containing autophagosome subsequently docks with a lysosomal vacuole, and their outer membranes fuse to deliver the entire ß-granule into the vacuole interior for degradation (13, 14, 15, 16). Microautophagy refers to the process whereby a whole ß-granule is taken up into a lysosomal compartment by phagocytotic-like mechanism (13, 16). Although both crinophagy and autophagy have been presumed to occur in pancreatic ß-cells (7, 8, 9, 10), the relative contributions of macroautophagy and/or microautophagy in endocrine cell hormone degradation have not been documented.
Historically, crinophagy has been regarded as the prevalent mechanism for regulating intracellular insulin content in the ß-cells of the endocrine pancreas under both normal and abnormal conditions. A crinophagic body in the ß-cell contains numerous insulin crystals (1, 6, 8). The insulin crystal is notoriously difficult and slow to degrade by lysosomal proteases (17), explaining why insulin, but not the more soluble C peptide, is detectable in these degradative structures (8). In this study we find that autophagy, as well as crinophagy, significantly contributes to the intracellular ß-granule degradation process in ß-cells. It is noted that both autophagy and crinophagy use similar large lysosomal-related vacuolar compartments for ß-granule/insulin degradation, and as such the term "crinophagic body" to describe these degradation organelles is unfitting. In this text we shall refer to a "multigranular body," as an alternative to a crinophagic body.
The details behind the mechanisms that control intracellular hormone degradation in endocrine cells remain poorly understood (7, 8, 9, 18). However, for pancreatic ß-cells this is clearly a regulated process. Under conditions where there is little demand for insulin secretion, there is an increase in the intracellular degradation of insulin (6, 7). Conversely, when the rate of insulin secretion is high, there is a corresponding decrease in the rate of insulin degradation within the ß-cell (6, 10, 19, 20, 21). Thus, the intracellular digestion of insulin granule contents serves as a longer-term regulatory mechanism for keeping insulin content at optimal levels while maintaining the ß-cell’s capacity to secrete (7, 22).
In this study, we have examined a glucose-intolerant, insulin secretion-deficient animal model, the Rab3A–/– null mouse. Rab3A is a low molecular weight GTP-binding protein that has been shown to direct vesicular traffic in neuroendocrine cells (23). In the ß-cell, Rab3A resides on the ß-granule membrane, where it plays a role in directing transport of ß-granules to the cell surface for subsequent exocytosis (24). In pancreatic islets isolated from Rab3A–/– mice, ß-cells secrete approximately 70% less insulin compared with wild-type animals in response to secretagogue stimulation (26). Despite this net overproduction of insulin, insulin content does not markedly increase in Rab3A–/– mouse islets. Here we provide evidence that excessive insulin production, in this instance, is compensated for by a marked increase in insulin degradation facilitated by both autophagic and crinophagic ß-granule degradation. Thus, regulating the intracellular digestion of ß-granules is a key mechanism for maintaining insulin stores at constant levels in the face of insulin secretory dysfunction in the ß-cell. This, in turn, may have consequences for understanding the ß-cell’s management of intracellular hormone content in other settings of insulin secretory dysfunction, such as type 2 diabetes.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). After chronic (18 h) exposure to 16.7 mM glucose, there was a similar 2-fold increase in preproinsulin mRNA levels in both Rab3A–/– and Rab3A+/+ islets compared with levels observed at 2.8 mM glucose (Fig. 1A). This reflected a specific glucose-induced increase in preproinsulin mRNA levels, because control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels did not change. In contrast to preproinsulin mRNA levels, there was a specific 4- to 5-fold increase in proinsulin biosynthesis after the 1 h incubation at 16.7 mM glucose compared with that at basal 2.8 mM glucose (P 0.01) (Fig. 1B). This rapid, glucose-induced increase in proinsulin biosynthesis above that of general protein synthesis observed in both Rab3A–/– and Rab3A+/+ islets, in the absence of any change in preproinsulin mRNA levels, underlines the specific translational control of proinsulin biosynthesis by glucose (3, 4).
|
0.01) in both Rab3A–/– and Rab3A+/+ mouse islets (Fig. 1B
). Thus, the approximate 2-fold increase in the magnitude of glucose-induced proinsulin biosynthesis after an 18-h exposure to stimulatory (16.7 mM) glucose over that observed after a 1-h incubation (Fig. 1B) in both Rab3A–/– and Rab3A+/+ islets reflects the specific glucose-induced increase in preproinsulin mRNA levels (Fig. 1A), as previously indicated (4). Taken together, these results demonstrate that mechanisms that control proinsulin biosynthesis, at both the translational and transcriptional levels, remain normal in the Rab3A–/– mouse islet ß-cell.
Proinsulin Processing to Insulin Is Normal in Rab3A–/– Mouse Islet ß-Cells
Another important mechanism affecting insulin production in the pancreatic islet ß-cell is the efficient conversion of proinsulin to insulin and C peptide by limited proteolysis (3). Short-term pulse-chase radiolabeling studies revealed that normal proinsulin conversion in Rab3A–/– mouse islets was indistinguishable from that in control Rab3A+/+ mouse islets, with more than 90% of proinsulin converted to insulin within 3 h after initiating proinsulin biosynthesis (Fig. 1C) (25).
Secretagogue-Stimulated Insulin Secretion Is Deficient in Isolated Rab3A–/– Mouse Islets
Glucose-regulated insulin secretion is significantly deficient in Rab3A–/– islet ß-cells (26). Here, we examined whether potentiation of glucose-induced insulin secretion by the incretin, glucagon-like peptide 1 (GLP-1), was also diminished in Rab3A–/– islets. A typical sigmoidal response curve for glucose-stimulated insulin secretion for control Rab3A+/+ islets was observed, with a threshold at 5–6 mM that reached a maximal rate at 12–16 mM glucose (Fig. 2A). The addition of GLP-1 significantly potentiated glucose-induced insulin secretion between 8 and 16 mM glucose in normal Rab3A+/+ islets (Fig. 2A), as expected (27). In contrast, glucose-stimulated insulin secretion from isolated Rab3A–/– islets was significantly decreased (
70%) compared with control islets (Fig. 2A; P 0.05). Moreover, GLP-1-mediated potentiation of glucose-induced insulin secretion was absent in Rab3A–/– islets (Fig. 2A; P 0.02), further indicating that the functional secretory capacity of Rab3A–/– islet ß-cells was severely restricted (26). It should be noted that this lack of GLP-1-potentiation of glucose-induced insulin secretion was not due to Rab3A–/– islets being unresponsive to GLP-1. GLP-1-stimulated activation of protein kinase-A, as judged by increased phosphorylation of the transcription factor cAMP responsive element binding protein, was still apparent in Rab3A–/– islets and equivalent to that in Rab3A+/+ islets (data not shown). The insufficient insulin-secretory response to glucose/GLP-1 in Rab3A–/– islets was more likely reflective of deficient ß-granule transport in Rab3A–/– ß-cells (24, 26).
|
), yet defective insulin secretion (Fig. 2A) (26), the insulin content of Rab3A–/– islets was not significantly different from that of control Rab3A+/+ islets (Fig. 2B).
Intracellular Insulin Degradation Is Enhanced in Rab3A–/– Islets
An indication of insulin degradation activity in Rab3A–/– and control islets was examined by a longer-term pulse-chase radiolabeling strategy. A window of [35S]protein/[35S](pro)insulin synthesized in a 90-min pulse radiolabeling period was chased in the islets for up to 96 h in culture. The chase period was conducted at a basal 2.8 mM glucose to keep (pro)insulin secretion to a minimum and retain the majority of the [35S](pro)insulin intracellularly (25). Indeed, negligible [35S](pro)insulin was detected in the media of these experiments (data not shown). As such, any difference between detectable [35S](pro)insulin in Rab3A+/+ and Rab3A–/–islets was most likely reflective of intracellular insulin degradation activity (7).
Total protein degradation in Rab3A+/+ and Rab3A–/– islets was similar, with approximate general protein half-lives of 29.6 ± 2.4 h and 32.1 ± 4.2 h, respectively (Fig. 3A). In comparison, [35S](pro)insulin degradation was significantly longer, with an estimated half-life in excess of 3 d (Fig. 3B). This finding is consistent with previous observations of an estimated 3- to 5-d half-life of a ß-granule (6, 7), and that the insulin crystal contained in mature ß-granules is relatively resistant to lysosomal protease degradation (6, 17). The rate of insulin breakdown in Rab3A–/– islets was observed to be slightly faster than that of Rab3A+/+ islets by this pulse-chase radiolabeling analysis at more than 48 h (Fig. 3B). The residual [35S](pro)insulin in Rab3A–/– islets was significantly lower than that in Rab3A+/+ control islets at 96 h (P < 0.05; Fig. 3B), indicative of increased intracellular insulin degradation activity.
|
) (29). Secretory granules in islet non-ß-cells have larger, more rounded, dense cores with negligible haloes (Fig. 4B) (30). In the present study, EM analysis revealed that degradative compartments, referred to here as "multigranular bodies," were much more numerous in Rab3A–/– mouse islet ß-cells than in Rab3A+/+ control mouse islet ß-cells (Fig. 4A). It is noteworthy that multigranular bodies were also more numerous in the non-ß-cells of Rab3A–/– mouse islet compared with Rab3A+/+ mouse islet non-ß-cells (Fig. 4B). This is not surprising, given that this animal model represents a total knockout of the Rab3A gene in all tissues (26).
|
-, 
-, pp-, and 
-cells at the EM level by granule morphology alone, non-ß-cell data were grouped together into a single category. A large number of 5-µm2 surface areas in islet cell electron micrographs were counted to minimize sectioning artifacts.
On average, control Rab3A+/+ islets contained 0.36 ± 0.04 multigranular bodies/5-µm2 ß-cell section surface area. In contrast, Rab3A–/– mouse islet ß-cells contained significantly more 4.41 ± 0.20 multigranular bodies/5-µm2 ß-cell section surface area, implicating a 12.3-fold increase relative to control mice (P 0.001; Fig. 5A). Moreover, multigranular bodies were found to be significantly larger in Rab3A–/– vs. Rab3A+/+ islet ß-cells (analysis of 1522 and 233 multigranular bodies, respectively; P 0.001; Fig. 5A), and contained more residual insulin crystals (P 0.001; Fig. 5A) compared with control animals. Thus, intracellular insulin degradative activity was higher in Rab3A–/– vs. Rab3A+/+ mice.
|
), Rab3A–/– mouse islet non-ß-cells revealed a significant 8.7-fold increase in the number of degradative compartments relative to control animals at 2.08 ± 0.22 multigranular bodies/5-µm2 non-ß-cell section surface area (P 0.001; Fig. 5B). In contrast to ß-cells, the average size of multigranular bodies did not differ between control and Rab3A–/– mice (Fig. 5B). Residual secretory granule cores were not consistently observed in non-ß-cell multigranular bodies and thus could not be quantified, presumably due to the fact that their primary hormone products (i.e. glucagon, somatostatin, pancreatic polypeptide, and ghrelin in -, -, pp-, and -cells, respectively) are more soluble and thus more readily degraded than the paracrystalline insulin core within mature ß-granules (17).
Although two-dimensional EM surveys of thin sections indicated increased degradation of ß-granules/insulin in Rab3A–/– islet ß-cells mediated by an apparent 12-fold increase in the number of multigranular bodies per ß-cell surface area (Fig. 5), we wanted to rule out the possibility that this increase was due to the limitations of conventional EM. For example, some multigranular bodies resided in close proximity to each other, which might have suggested a single larger tubular degradative compartment that had been cut in several places in the plane of view to reveal an apparent increase in the number of individual multigranular bodies. Thus, we analyzed the structure of several multigranular bodies using a three-dimensional (3D) EM technique referred to as electron tomography (ET), to unambiguously address the nature of these compartments at high resolution (6 nm) and in 3D (31, 32). Representative 3D reconstructions of multigranular bodies in mouse islet ß-cells are presented in QuickTime Movies 1–4 (see supplemental data published on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). The tomographic analysis indicated that although often multilobed, each multigranular body visible within a plane of section indeed represented a single degradative compartment rather than a portion of a large, continuous tubular system.
The number of ß-granules per ß-cell surface area (excluding the nucleus) was also examined in the same electron micrographs. In Rab3A+/+ control mouse islets, an average of 74.8 ± 2.6 ß-granules/5-µm2 ß-cell section surface area was observed, whereas in Rab3A–/– mouse islets a significantly less 52.8 ± 1.8 ß-granules/5-µm2 ß-cell section surface area was found (P 0.001). This represents a 29.4% decrease in ß-granules in Rab3A–/– vs. Rab3A+/+ control mouse islets. It should be noted that this ß-granule decrease is not reflected in the islet insulin content of Rab3A–/– mouse islets (Fig. 2B), because this is also contributed by the residual insulin crystals found in the increased numbers of multigranular bodies in Rab3A–/– mouse islet ß-cells (Fig. 5). Moreover, considering that islet ß-cells secrete only 1–2% of their intracellular insulin stores per hour under stimulatory conditions (Fig. 2A), it is unlikely that this approximately 30% decrease in ß-granules in Rab3A–/– mouse islet ß-cells impinges on insulin-secretory capacity.
EM Evidence of Increased ß-Granule Crinophagy and Autophagy in Rab3A–/– Islets
The intracellular degradation of insulin and turnover of ß-granules in ß-cells is presumed to be facilitated by both crinophagy and autophagy (6, 10). However, these processes have not been clearly demonstrated to occur simultaneously in pancreatic ß-cells. Because of the high incidence of ß-granule/insulin degradation in Rab3A–/– mouse islet ß-cells, we were able to capture examples of crinophagy in which ß-granules were apparently docked with multigranular bodies (Fig. 6A). An area of darkened shading on the electron micrographs at the point of contact between the ß-granule and multigranular body membranes (as indicated by the arrow on Fig. 6A), suggested that these organelles were indeed in physical contact. In addition, rare images of ß-granules already fused with multigranular bodies were acquired (Fig. 6, B and C), with their membranes continuous in some cases (Fig. 6C), indicating that the process of crinophagy had occurred.
|
) to form a double-membraned vesicle encapsulating an entire ß-granule (Fig. 6E); this early organization of autophagic membranes is known as an "autophagosome" (13, 14). The docking of autophagosomes to multigranular bodies to facilitate the delivery of ß-granules in toto into degradative compartments was also frequently observed (Fig. 6F). This process results in digestion/breakdown of the ß-granule membrane inside the multigranular body, as well as the granule’s luminal contents such as insulin (Fig. 6G). Furthermore, it was also possible to capture images of the process of microautophagy, whereby degradative/multigranular structures selectively phagocytose and engulf entire ß-granules (Fig. 6, H–J). This process, like macroautophagy yet in contrast to crinophagy, results in the degradation of the ß-granule membrane in addition to the granule cargo (Fig. 6K), but in a selective rather than bulk autophagic manner. Tomographic analyses of the 3D organization of ß-cell organelles revealed that the majority of multigranular bodies in Rab3A–/– mouse islet ß-cells have invaginations, indicative of microautophagic activity. A Quicktime movie (supplemental Movie 5) shows a multigranular body with invaginations revealed to be continuous with the cytosol by EM-tomography, clearly indicating its microautophagic activity. A 3D surface-rendered model generated by segmentation of the tomographic reconstruction of this particular multigranular body highlights such invaginations and microautophagic uptake (supplemental Movie 6).
Examination of Autophagic Gene (ATG) Expression in Rab3A–/– vs. Rab3A+/+ Isolated Mouse Islets
A family of genes (known as ATGs) has been implicated in the control of macroautophagy (12, 14, 16, 33). However, it essentially remains unclear how crinophagy and microautophagy are differentially regulated (16). Nonetheless, we examined whether there was an alteration in the expression of key ATGs in isolated islets of Rab3A+/+ vs. Rab3A–/– mice by semiquantitative RT-PCR analysis using ß-actin mRNA expression as a reference that did not alter between Rab3A+/+ vs. Rab3A–/– mouse islets (Fig. 7A). As observed previously (Fig. 1), there was no change in preproinsulin gene expression between Rab3A+/+ vs. Rab3A–/– mouse ß-cells (Fig. 7A). For most of the ATGs examined there was no significant difference in their expression between islets from Rab3A+/+ vs. Rab3A–/– mice (Fig. 7, A and B). There was a trend for ATG5 to be slightly increased in Rab3A–/– mouse islets, but this was not statistically significant (Fig. 7, A and B). Likewise, there was no change in the expression of the lysosomal protein, LAMP-1 (Fig. 7, A and B). However, in contrast, there was a specific 67% significant decrease in the related LAMP-2 gene expression (P < 0.01; Fig. 7, A and B). Although a semiquantitative RT-PCR approach can reveal large qualitative changes in mRNA expression levels, it is not necessarily a reliable quantitative method. As such, the expression levels of LAMP-1 and LAMP-2 in Rab3A+/+ vs. Rab3A–/– mouse isolated islets were reexamined by fluorescence-based real-time quantitative RT-PCR, following a method previously described using ß-actin mRNA levels as a reference (34). This more quantitative approach also indicated that expression levels of LAMP-1 were unchanged, but there was a significant approximately 80% decrease in LAMP-2 mRNA levels in Rab3A+/+ vs. Rab3A–/– mouse isolated islets (P 0.05; Fig. 5C).
|
0.05; Fig. 7D). LAMP-1 protein expression and that of the p85 regulatory subunit of phosphatidylinositol-3'-kinase (PI3Kp85; used as a loading control) did not change between Rab3A+/+ vs. Rab3A–/– mouse islets (Fig. 7D), emphasizing a specific decrease of LAMP-2 expression in Rab3A–/– mouse islets. It should be noted that LAMP-2 is a negative regulator of autophagy (35, 36), and this reduced expression LAMP-2 in Rab3A–/– mouse islets correlated with the increased autophagy observed. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Yet, despite the net overproduction of insulin in Rab3A–/– islet ß-cells, intracellular insulin levels were apparently not significantly different from controls due to the marked up-regulation in intracellular insulin/ß-granule degradation by autophagy and crinophagy.
Crinophagy has been previously deemed to be the major mechanism of intracellular insulin degradation in the ß-cell (6, 8, 10), but this study has revealed a critical role for autophagy in management of ß-granule/insulin content in vivo, under conditions where insulin secretion is impaired. Previous studies that used diazoxide to inhibit insulin secretion without affecting insulin production reported a significant increase in intracellular insulin content concomitant with increased lysosomal enzyme activity and crinophagy, but not autophagy (9, 10). This increase in insulin content observed in diazoxide-treated islets reflected that a transient increase in crinophagy alone was insufficient to redress the net oversupply of intracellular insulin. Here, in contrast, we find that in an in vivo model of insulin-secretory deficiency, the Rab3A–/– mouse, excess insulin production in ß-cells is counterbalanced by a up-regulation of autophagic degradative pathways, as well as crinophagy, to maintain intracellular insulin stores at near-normal levels over the long term. Consistent with other studies, increased intracellular insulin breakdown in Rab3A–/– mouse ß-cells was indicated by a marked increase in degradative compartments (7, 8, 9, 10, 22). These compartments are historically referred to as "crinophagic bodies" because crinophagy has been documented as the major mechanism of secretory granule degradation in the ß-cell (6, 22). Indeed, EM evidence in this study indicated that crinophagy does contribute to increased insulin/ß-granule degradation in ß-cells of Rab3A–/– islets (Fig. 6). However, autophagy, particularly microautophagy, was also markedly increased in Rab3A–/– islet ß-cells compared with Rab3A+/+ control ß-cells. It is this augmentation in autophagic activity over the long term that likely compensates for the overproduction of insulin in Rab3A–/– islet ß-cells in vivo to maintain intracellular insulin stores at near-normal levels.
Autophagy is subdivided into different mechanisms, referred to as "macroautophagy," "microautophagy," and "chaperone-mediated autophagy" (13, 14, 16). Both macroautophagy and microautophagy were observed in Rab3A–/– islets (Fig. 6). Chaperone-mediated autophagy is a selective mechanism by which certain cytosolic proteins are delivered to and degraded in lysosomal compartments (16, 37). As such, it is quite unlikely to be involved in the selective degradation of a whole organelle, such as ß-granule, as found in Rab3A–/– mouse ß-cells. Likewise, proteosomal degradation mechanisms are not involved in intracellular degradation of insulin in ß-cells (38). Due to the high levels of ß-granule/insulin degradation in Rab3A–/– islet ß-cells, multiple stages of macroautophagy were documented, including an autophagosome (Fig. 6, D–F) (13, 14, 16, 39), and, more commonly, microautophagy in which individual ß-granules are imbibed into a multigranular degradation compartment by a phagocytotic mechanism (Fig. 6, H–J, and supplemental Movies 5 and 6). These data provide the first evidence that macroautophagy and microautophagy act in concert with crinophagy to regulate ß-granule/insulin content in ß-cells, under conditions in which regulated secretion is impaired and insulin production is excessive.
There is an important functional implication between crinophagy and autophagy in terms of ß-granule degradation in the ß-cell. Because crinophagy involves the fusion of the ß-granule and multigranular body outer membranes, a ß-granule membrane protein is incorporated into the multigranular body membrane and is probably not subject to lysosomal degradation. Presumably, ß-granule membrane constituents are recycled back to the Golgi for use in subsequent rounds of insulin granule packaging (40). In contrast, the autophagic mechanisms documented here result in delivery of the entire ß-granule to the lumen of the multigranular body/lysosomal compartment (12, 13, 14), so that both lumenal cargo and membrane components of the ß-granule are proteolytically degraded. As such, autophagy plays a much more prominent role in ß-granule turnover than previously thought, particularly under conditions where a marked increase in intracellular degradation is required. Moreover, because the insulin contained in the multigranular bodies will be irreversibly degraded [albeit slowly (17)], it no longer is able to contribute to the ß-cell’s insulin-secretory capacity. Therefore, measurements of islet total insulin content do not always provide an accurate measurement of ß-cell-secretory capacity, where a sizeable proportion of intracellular insulin terminally resides in multigranular body/degradative compartments. Nonetheless, most likely because of increased ß-granule autophagic activity in the Rab3A–/– islet ß-cells, a 30% reduction in ß-granule number was observed. However, it is unlikely that this moderate decrease in ß-granules will contribute to insulin-secretory dysfunction in this model. A ß-cell’s ability to secrete insulin only reflects a minor subpopulation of ß-granules that remain capable of readily undergoing exocytosis (2, 41). This is reflected in a 1–2% of the total ß-granule population secreted/hour under marked stimulated conditions (Fig. 2A).
The increase of autophagy in Rab3A–/– islet ß-cells compensates for the abnormal imbalance between insulin secretion and insulin production. This implies that intracellular insulin degradation activity is a highly regulated process, and that maintenance of insulin stores is an important aspect of ß-cell function. A question arises as to what controls the increase in multigranular bodies in Rab3A–/– islet ß-cells selective for ß-granule degradation. An initial examination of the expression of key genes implicated in multigranular body degradation mechanisms between isolated islets from Rab3A+/+ and Rab3A–/– mice indicated that there was no significant difference in the expression of ATGs implicated in the up-regulation of macroautophagy (12, 14, 16, 33). However, increased ATG expression is more associated with a response to starvation or pathogenic infections and directed at the process of aggregating soluble proteins destined for macroautophagic degradation, rather than degradation of whole organelles (33, 42). This suggests that macroautophagy, although present, was not a major contributor to ß-granule degradation in Rab3A–/– mouse islet ß-cells, consistent with the observation that microautophagy was more common. In contrast to ATGs, the expression of the lysosomal associated membrane protein, LAMP-2, but not that of the related LAMP-1, was significantly decreased in Rab3A–/– vs. Rab3A+/+ mouse islets (Fig. 7). LAMP-2 has been shown to be a negative regulator of autophagy (35, 36), and disabling mutations in LAMP-2 are associated with Danon disease in humans (43). In Danon disease and LAMP-2–/– mice there is a marked increase in autophagic vacuoles in many tissues causing cardiomyopathy, myopathy, and neuronal degeneration (33, 35, 42, 44). As such, it is quite possible that the specific decreased expression of LAMP-2 in Rab3A–/– mouse islets causes an increase in multigranular body degradation compartments in ß-cells and increased ß-granule/insulin degradation. This implies that LAMP-2 might be dynamically regulated in ß-cells to control autophagic degradation of organelles, like ß-granules. Because no change in ATG expression was observed in Rab3A–/– mouse islets, it is likely that the increased autophagic degradation of ß-granules was more commonly contributed by increases in microautophagy and less so by macroautophagy. However, it should also be considered that a multigranular body in the ß-cell may originally be derived from a lysosome (20, 21), and thus the significant increase in the number of these degradation compartments in Rab3A–/– mouse islet ß-cells implies that the regulation of lysosomal biogenesis will also contribute to controlling insulin/ß-granule degradation in the ß-cell (45, 46). Because LAMP-2 is a lysosomal-membrane protein, a decrease in its expression may favor formation of mircoautophagic multigranular bodies from lysosomes in ß-cells. However, in general, the mechanisms that regulate microautophagy in mammalian regulated secretory cells remain undetermined (16), and further experimentation will be required to determine whether the specific decrease in LAMP-2 expression triggers the up-regulation of (micro)autophagic activity in ß-cells.
It is possible that the findings in this study might be peculiar to the Rab3A–/– mouse model. However, it is unlikely that deletion of Rab3A leads to increased ß-granule degradation. Of more than 40 members of the Rab family of proteins involved in directional vesicular trafficking, only Rab7 has been implicated in autophagy (47). Considering that an imbalance between insulin production and secretion in ß-cells leads to adaptive ß-granule degradation under other circumstances (7, 10, 19, 21, 22), we favor this reason as the cause of increased crinophagic and autophagic activity in Rab3A–/– mouse islet ß-cells. However, further experimentation is required to substantiate this idea, and for the moment alternative explanations cannot be ruled out.
Notwithstanding, it should be noted that increased autophagy has been associated with the pathogenesis of human type 2 diabetes (48), and our findings here might have relevance. In certain models of type 2 diabetes, both insulin gene expression and proinsulin biosynthesis can be unchanged or even up-regulated in contrast to insulin secretory dysfunction and deficiency in the same ß-cell population, acquiring an imbalance between insulin production and secretion reminiscent of the Rab3A–/– mouse ß-cell phenotype (26). This implies that ß-granule degradation could well be up-regulated in the ß-cell during the progression of type 2 diabetes, to counter excessive insulin production. Initially, this may be a protective effect to avert over accumulation of ß-granules within the ß-cell. Although autophagic ß-granule degradation may be initially selective for ß-granules, if continued chronically, it could become less discriminatory and other ß-cell organelles might be degraded resulting in autophagic mediated-death of ß-cells (16, 33), which in turn would contribute to decreased ß-cell mass that marks the onset of type 2 diabetes (49). Indeed, there is precedence of this in that chronic autophagic programmed cell death type II (also known as nonapoptotic cell death) has been shown to be a significant contributing factor to certain myopathies and the later stages of neuronal degenerative diseases (16, 33). Although the current study in Rab3A–/– mouse islet ß-cells newly reveals the potential of ß-cell autophagy as a contributing factor in the pathogenesis of type 2 diabetes, future experiments will be needed to reaffirm its part in either the increased incidence of ß-cell death and/or secretory dysfunction that marks the onset/progression of the disease in humans.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-32P]trisphosphate (3000 Ci/mmol) was purchased from Amersham Pharmacia Biotech, Inc. (Piscataway, NJ). GLP-17–36 was purchased from Bachem, Inc. (King of Prussia, PA). All other reagents were of analytical grade and obtained from either Sigma Chemical Co. (St. Louis, MO) or Fisher Scientific, Inc. (Pittsburgh, PA).
Animals
The Rab3A+/+ on a B6 background (B6129SF2/J) and Rab3A–/– mice were obtained from The Jackson Laboratory (Bar Harbor, ME) (26). Mice were housed on a 12-h light, 12-h dark cycle and were allowed free access to standard mouse food and water. Mice were used at 18–24 wk of age. Animal experimentation was conducted in accord with accepted standards of humane animal care as outlined by the National Institutes of Health Guide for the Care and use of Laboratory Animals and approved by the Institutional Animal Use and Care Committees at the University of Chicago.
Islet Isolation and in Vitro Insulin Secretion
Pancreatic islets of Langerhans were isolated by collagenase digestion followed by Histopaque-Ficoll density gradient centrifugation as previously described (50). For static incubations, 20–25 freshly isolated islets were preincubated in 500 µl Krebs-Ringer buffer (KRB) containing basal 2.8 mM glucose and 0.1% BSA (wt/vol) for 60 min at 37 C, and then incubated for a further 60 min in the same volume of KRB/0.1% BSA containing basal 2.8 mM glucose, stimulatory 16.7 mM glucose, or 16.7 mM glucose ± 1 nM GLP-1. After this second incubation at 37 C, the media were removed and the islets placed in 200 µl lysis buffer (50 mM HEPES, pH 7.5; 1% Nonidet P-40; 2 mM Na vanadate; 100 mM NaF; 4 mM EDTA; 1 µM leupeptin; 10 µg/ml aprotinin; 100 µM phenylmethylsulfonylfluroride), followed by sonication (25 W; 10 sec). The media and lysate were then analyzed for insulin content by mouse insulin RIA (Linco Research, Inc., St. Charles, MO) (4, 26).
RNA Analysis
The preproinsulin mRNA levels in islets were analyzed by the RNase protection assay, using the Direct lysis kit (Ambion, Inc., Austin, TX), as described previously (4). Essentially, RNA was protected by [32P]uridine-labeled antisense RNA fragments corresponding to the coding region of preproinsulin, or to part of the GAPDH coding region (Ambion). Samples were resolved on denaturing 5% acrylamide/Tris-borate EDTA-buffered gels and analyzed by autoradiography.
An initial indication of the steady-state mRNA levels for certain ATGs and lysosomal membrane-associated proteins (LAMP-1 and -2) was by semiquantitative RT-PCR analysis following methodology previously described for isolated islets (51). Essentially, total RNA was extracted from freshly isolated islet preparations from individual Rab3A+/+ or Rab3A–/– mice using RNAeasy kits (QIAGEN, Chatsworth, CA) followed by DNase digestion to remove any contaminating DNA. The RNA was then converted to cDNA (Ready-To-Go T-Primed First-Strand Kit, Amersham Pharmacia Biotech (Piscataway, NJ). These cDNAs were used as templates for PCRs using specific primers of annealing temperatures between 60–65 C in the presence of MgCl2 and deoxynucleotide triphosphates. Amplification was generally between 24–28 cycles to be in the linear range. PCR products were resolved on agarose gels and quantified using PerkinElmer Optiquant Analysis Software (PerkinElmer, Wellesley, MA). The apparent mRNA levels were relative to ß-actin mRNA control. The mouse nucleotide sequences of the primers used were as follows. Preproinsulin-1: TGGTGCACTTCCTACCCCT (forward), GCCTTA GTTGCAGTAGTTCTCC (reverse); Preproinsulin-2: CTGTGGATGCGCTTCCT (forward), TTGCAGTAGTTCTCCAGCTGGTA (reverse); ß-actin: AGGCGGACTGTTACTGAGC (forward), AACACCTCAAACCACTCCC (reverse); ATG4: ATGGGAGTTGGCGAAGGCAAGTCTA (forward), CGTTAAAGTCTTCCTCCGTCTTAC (reverse); ATG5: CATCCACTGGAAGAATGACAG (forward), CATCCAGAGCTGCTTGTGGTCT (reverse); ATG6 (also known as beclin-1): ACTGGACACGAGCTTCAAGAT (forward), TTCCTCCTGGTCCAATCACA (reverse); ATG7: CAGCAAATGAGATCTGGGAAGCC (forward), GCTTTAGGACAATCTGGGCTA (reverse); ATG8 (also known as LC3): ATTACGTGAACATGAGCGAGC (forward), AGGAAGAGACTGCTCCTGAC (reverse); ATG10: TCGACACCACATGTGGGAA (forward), GGTGATGTGAGATATTGCAGTCCCA (reverse); ATG12: CCAAGGACTCATTGACTTCATC (forward), AATAGAGGTCCCCTGAATAAGC (reverse); ATG16: AGTCCTGGACATGATGGTGCGT (forward), TCTCTTCTGTGTGGTCTGAGGT (reverse); LAMP-1: GAAAATGTTTCTGACCCCAGCCT (forward), GTGTCATTTGGGCTGATGTTGAACGC (reverse); LAMP-2: CCATTGGATGTCATCTTTAAGTGC (forward), GTTGAAAGCTGAGCCATTAG (reverse).
To confirm the specific decrease in LAMP-2 expression in Rab3A–/– islets, a fluorescence-based quantitative real-time RT-PCR method was applied as previously described (34), using the same forward and reverse primers for ß-actin, LAMP-1, and LAMP-2 mRNA.
Immunoprecipitation
For analysis of proinsulin biosynthesis, isolated mouse islets were preincubated at 37 C in 200 µl KRB/0.1% (wt/vol) BSA, containing basal 2.8 mM glucose for 60 min, and for a further 60 min at either basal 2.8 mM or stimulatory 16.7 mM glucose. The last 20 min of the incubation was carried out in the additional presence of 250 µCi/ml [35S]methionine to monitor protein/proinsulin biosynthesis as described elsewhere (4, 50). Islets were washed then lysed in lysis buffer with sonication (10 sec/25 W), and the lysates were subjected to immunoprecipitation using guinea pig antibovine insulin (Sigma) as described previously (4, 50). Immunoprecipitated (pro)insulin was resolved by alkaline-urea gel electrophoresis. Gels were fixed in 50% methanol/10% acetic acid, dried, and quantified by phosphor imager analysis. Total protein synthesis was analyzed by 10% (wt/vol) trichloroacetic acid precipitation of 5 µl aliquots of the same islet lysates as described elsewhere (4, 50). The extent of specific glucose-induced proinsulin biosynthesis was corrected for the smaller effect of glucose on total protein synthesis.
For pulse-chase radiolabeling studies to examine proinsulin conversion, isolated mouse islets were incubated in batches of 50 at 37 C in 200 µl KRB/0.1% (wt/vol) BSA for 30 min at 16.7 mM glucose, and then for another 30-min pulse under the same conditions but in the presence of 250 µCi/ml [35S]methionine. Islets were washed, and then chased for 0, 30, 60, 90, 120, or 180 min in 500 µl KRB/0.1% (wt/vol) BSA at 37 C containing 1 mM methionine at basal 2.8 mM glucose to minimize (pro)insulin secretion as previously established. At each chase time point, islets were collected, washed, lyzed by sonication, and subsequently immunoprecipitated for (pro)insulin, with [35S]proinsulin/insulin processing analyzed by alkaline-urea gel electrophoresis as described above. For longer (pro)insulin degradation pulse-chase radiolabeling experiments a similar approach was taken, except that a longer 90-min pulse in the presence of 250 µCi/ml [35S]methionine was used. In addition, as the chase incubation periods were much longer, these were conducted at 37 C in RPMI 1640 tissue culture medium (Invitrogen, Carlsbad, CA) containing 2.8 mM glucose, 10% fetal bovine serum (Hyclone Laboratories, Inc., Logan, UT), 100 U/ml penicillin, and 100 µg/ml streptomycin (Invitrogen) for 5, 24, 48, or 96 h, where the media was changed every 24 h.
High-Pressure Freezing and Freeze Substitution
After overnight culture in RPMI medium containing fetal bovine serum 10% (vol/vol) and 7 mM glucose (Sigma), islets were kept warm in culture medium buffered by the addition of HEPES (10 mM) on a humidified heating block at 37 C before freezing experiments. The sample holders used for high-pressure freezing experiments were also prewarmed to 37 C. Depending on size, 10–30 islets were gently manipulated into one half of the holder prefilled with HEPES-buffered RPMI containing fetal bovine serum 10% (vol/vol) and 7 mM glucose (Sigma). All manipulations were carried out on Parafilm placed on top of an inverted heating block warmed to 37 C under a dissecting microscope. The second half of the sample holder, filled with RPMI containing 10% dialyzed Ficoll (molecular mass, 70 kDa) and 0.5% Type IX ultra-low temperature gelling agarose (Sigma) as extracellular cryoprotectants, was then placed on top of the half of the sample holder containing the islets (32). Islets were frozen within approximately 10 msec under high pressure (2100 atmospheres) using a Balzers HPM 010 high-pressure freezer (BAL-TEC AG, Liechtenstein, Germany) and stored under liquid nitrogen. Specimens were freeze substituted and plastic embedded essentially as described previously (31).
Microtomy and Microscopy
Ribbons of serial thin (40–60 nm) sections cut from plastic-embedded islets using a Leica UltraCut-UCT microtome (Leica, Inc., Deerfield, IL) and collected onto Formvar-coated copper slot grids were poststained with 2% aqueous uranyl acetate and Reynold’s lead citrate to increase contrast when viewed in the EM. Grids were surveyed on a conventional electron microscope operating at 80 keV. Grids containing 300–400 nm-thick sections for tomographic analysis were lightly carbon coated to minimize charging in the EM. Colloidal gold particles (10 or 15 nm) were deposited on both surfaces of these sections for use as fiducial markers during subsequent image alignment. Tilt series data were digitally recorded as the sample grid was serially tilted at 1.5° angular increments over a range of 120° (±60°) about two orthogonal axes, and dual-axis tomograms (3D reconstructions) were generated as detailed elsewhere (31, 32).
Statistical Analysis
Results are expressed as means ± SE. Statistical analysis was performed by unpaired Student’s t test or repeated measure ANOVA, where P < 0.05 was considered significant.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Current Address for K.Y.: Department of Physiology and First Department of Internal Medicine, Kagoshima University School of Medicine, 8–35-1 Sakuragaoka, Kagoshima 890, Japan.
Disclosure Statement: B.J.M., C.S., C.A., B.L.W., K.Y., A.J.C., and G.P.M. have nothing to declare. C.J.R. has consulted for Amylin Pharmaceuticals, Inc./Eli Lilly partnership, Merck Inc., Takeda Pharmaceuticals Inc., and Metabolex Inc.
First Published Online June 19, 2007
Abbreviations: ATG, Autophagic gene; 3D, three dimensions/three-dimensional; EM, electron microscope/microscopy; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GLP-1, glucagon-like peptide-1; KRB, Krebs-Ringer buffer; LAMP, lysosomal-associated membrane protein; PI3Kp85, p85 regulatory subunit of phosphatidylinositol-3'-kinase.
Received for publication February 9, 2007. Accepted for publication June 15, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
4), expressed as a fold-increase above that measured at basal (2.8 mM) glucose. Representative alkaline-urea gel fluorography analyses for proinsulin biosynthesis are also shown. Panel C shows a representative alkaline-urea gel fluorography analysis of proinsulin to insulin conversion from a pulse-chase radiolabeling experiment in Rab3A+/+ and Rab3A–/– mouse islets (as described in Materials and Methods), with the electrophoretic migration of proinsulin and insulin indicated.
