Expression differences in mitochondrial and secretory chaperonin 60 (Cpn60) in pancreatic acinar cells

Abstract

In pancreatic acinar cells, chaperonin Cpn60 is present in all the cellular compartments involved in protein secretion as well as in mitochondria. To better understand the role Cpn60 plays in pancreatic secretion, we have evaluated its changes under experimental conditions known to alter pancreatic secretion. Quantitative protein A–gold immunocytochemistry was used to reveal Cpn60 in pancreatic acinar cells. Cpn60 immunolabelings in cellular compartments involved in secretion were found to decrease in acute pancreatitis as well as upon stimulation of secretion and in starvation conditions. A major increase in Cpn60 was recorded in diabetic condition. This was normalized by insulin treatment. Although in certain situations changes in secretory enzymes and in Cpn60 correlate well, in others, nonparallel secretion seemed to take place. In contrast, expression of mitochondrial Cpn60 in acinar cells appeared to remain stable in all conditions except starvation, where its levels decreased. Expression of Cpn60 in the secretory pathway and in mitochondria thus appears to behave differently, and Cpn60 in the secretory pathway must be important for quality control and integrity of secretion.

INTRODUCTION

The pancreatic acinar cells have the ability to synthesize and secrete one of the highest rates of proteins, which, once discharged into the pancreatic juice, are targeted to the duodenal lumen. The intracellular aspects of secretion have been well characterized through morphocytochemical and biochemical techniques (Palade 1975; Jamieson 1981; Bendayan 1984). These studies and others have identified several successive steps in the processing of secretion, from synthesis at the level of the rough endoplasmic reticulum (RER) to concentration and glycosylation in the Golgi and to maturation and packaging in the secretory granules. The proper folding of newly synthesized proteins relies on the assistance of molecular chaperones. Much work has now established that molecular chaperones are involved in protein folding, assembly, disassembly, and degradation. Their activity is modified in response to stress as well as in pathological disorders (Ellis and van der Vies 1991; Gething and Sambrook 1992; Ellis and Hartl 1996; Ranson et al 1998; Fink 1999).

In previous studies, we reported that Cpn60 (chaperonin with 60-KDa molecular weight, the bacterial homolog of GroEL) as well as Cpn10 (GroES) and Grp94 are involved in the process of protein secretion in exocrine cells; they codistribute and coimmunoprecipitate with various secretory proteins (Arias et al 1994; Velez-Granell et al 1994; Bruneau et al 1998, 2000). Cpn60 as well as Grp94 are preferentially associated to particular pancreatic lipolytic secretory enzymes, namely the colipase-dependent lipase and the bile salt–dependent lipase, respectively (LeGall and Bendayan 1996; Bruneau et al 1998). The association occurs in all the intracellular compartments involved in protein secretion with increasing concentration gradients from the RER to the Golgi and to the zymogen granules (Velez-Granell et al 1994; LeGall and Bendayan 1996; Bruneau et al 1998). It was further established that these enzyme-chaperone complexes are discharged as such into the pancreatic juice and are targeted to the duodenal lumen (Bruneau et al 2000). These results support the hypothesis that Cpn60 as well as Grp94 play important roles in the proper folding and assembly of pancreatic secretory enzymes as well as in their posttranscriptional processing, likely preventing aggregation of unfolded polypeptides and autoactivation of enzymes before their secretion by the acinar cells (Velez-Granell et al 1994).

On the other hand, it is well established that Cpn60 is also a mitochondrial resident protein playing totally different roles from those of the protein located in the secretory pathway (Jindal et al 1989; Itoh et al 1995; Soltys and Gupta 2000; Keskin et al 2002).

These recent data have stimulated interest in further investigation of the function of Cpn60 and, in particular, of its involvement in the pathogenesis of some secretion-related problems. We have carried out an observational study using a limited number of rats under 4 different experimental conditions related to exocrine pancreatic function: acute pancreatitis, diabetes, cholinergic stimulation of secretion, and inhibition of secretion as induced by starvation. Changes in the secretory capacity of the exocrine pancreas under these conditions have been established previously (Ermak and Rothman 1981; Bendayan et al 1985; Gregoire and Bendayan 1986; Aughsteen and Cope 1987; Nagy et al 1989; Willemer et al 1989; Andrzejewski et al 1996; de Dios et al 1999). By means of quantitative immunocytochemistry, we found that levels of Cpn60 in cellular compartments involved in secretion vary significantly according to the condition studied. On the other hand, levels of mitochondrial Cpn60 remained rather stable, suggesting that both forms of Cpn60, the secretory and the mitochondrial, are regulated by independent factors.

MATERIALS AND METHODS

Animals

Eighteen Sprague-Dawley male rats (Charles River, St. Constant, Quebec, Canada) were divided into 6 groups according to experimental conditions: normal control, acute pancreatitis, diabetes, insulin-treated diabetes, stimulation of secretion, and starvation. All the experiments were conducted with the approval of the Institutional Committee of Deontology for experimentation on animals. The experimental models were prepared as follows:

(1) Acute pancreatitis: 3 rats weighing 250 g were fasted overnight with free access to water. They were anesthetized by an intraperitoneal injection of urethane (1 g/kg body weight). Acute pancreatitis was induced by retrograde injection of sodium deoxycholate (Sigma Aldrich Canada, Oakville, ON, Canada) in the biliopancreatic duct, using a modified method of Aho et al (1980) (Zhang and Li 2000). Briefly, after a small median laparotomy the biliopancreatic duct was temporarily closed at the liver hilum with a soft microvascular clamp to prevent reflux of the infused material into the liver. A retrograde injection of 4% sodium deoxycholate into the distal biliopancreatic duct was given (40 mg/kg body weight). The clamp was removed 5 minutes after the injection. The abdomen was closed, and the rats were kept for 5 hours.

(2) Diabetes: an experimental chronic hyperglycemic state was induced in 6 rats weighing 100 g by a single intraperitoneal injection of streptozotocin dissolved in 10 mmol/L of citrate buffer, pH 4.5, at 70 mg/kg body weight. The hyperglycemic state developed in the first 72 hours after the injection and remained throughout the 3 months of the experiment. Blood glucose levels averaged 29.4 ± 1.0 mmol/L vs 6.6 ± 0.3 mmol/L for the age-matched controls. After 2 months, 3 of the hyperglycemic rats were treated subcutaneously by a daily injection of human biosynthetic insulin (Eli Lilly Co., Indianapolis, IN, USA) for 1 month to lower their glycemic levels as described previously (Doucet et al 2000). The amounts of insulin ranged from 15 to 20 units/day depending on the animal and were adjusted according to blood glucose monitoring. Blood glucose levels as determined at 10:00 AM averaged 4.9 ± 0.8 mmol/L in these insulin-treated diabetic animals.

(3) Stimulation of secretion: a 10-mmol/L phosphate-buffered saline solution (PBS) of carbamyl β-methylcholine chloride (Sigma) in a final concentration of 12 mg/kg body weight was injected intraperitoneally into 3 rats weighing 250 g. The animals were killed 5 minutes after the induction of secretion (Bendayan et al 1985).

(4) Starvation: 3 rats weighing 250 g were placed in individual cages. They were deprived of food for 48 hours but had free access to drinking water (Bendayan et al 1985).

Tissue processing

At the end of the experiments, all the rats were anesthetized with urethane, and small fragments of the splenic part of the pancreas were fixed by immersion with 1% glutaraldehyde in 100 mmol/L phosphate buffer, pH 7.4, for 2 hours at 4°C. The tissue samples were washed in phosphate buffer, dehydrated in a series of graded methanol solutions, and embedded in Lowicryl K4M at −20°C as described previously (Bendayan 1995). Ultrathin sections were cut, mounted on Parlodion-carbon–coated nickel grids, and processed for immunocytochemistry.

Immunocytochemistry

For the labeling procedure, the thin tissue sections mounted on nickel grids were incubated by floating them on a drop of 1% ovalbumin in PBS, pH 7.2, for 10 minutes at room temperature and then transferred to a drop of the rabbit anti-Cpn60 diluted 1:10 with PBS (Stressgen, Victoria, BC, Canada). The grids were rinsed with PBS, transferred to the ovalbumin solution for 10 minutes, and incubated for 30 minutes on a drop of the protein A–gold complex 10 nm in diameter (Bendayan 1995). They were then washed thoroughly with PBS and rinsed with distilled water. After staining with uranyl acetate, the sections were examined with a Philips 410 electron microscope. The specificity of the immunolabeling was assessed by various control experiments, including incubation with the protein A–gold solution alone, omission of the primary antibody step, and preadsorption of the antibody with its corresponding antigen, followed by the protein A–gold complex.

Evaluation

Morphometrical evaluation of the labelings was performed using a Videoplan 2 image processing system (Carl Zeiss, Toronto, ON, Canada). At least 30 micrographs, recorded at 28 000× final magnification, were analyzed for each animal tissue. Labeling densities were evaluated as described in detail previously (Bendayan 1995) and reported as mean number of gold particles per square micrometer ± standard error to the mean. Statistical analysis of the results were carried out using the Student's t-test.

RESULTS

In normal control animals, the pancreatic acinar cells (Fig 1) displayed large spherical nuclei and a basal cytoplasm occupied by extensive parallel arrays of RER and occasional mitochondria. Lysosomes were rare. A large Golgi complex and numerous zymogen secretory granules at various stages of maturation were present on the apical third of the cells. In acute pancreatitis, a large number of cells appeared with damaged membranes, swollen organelles, and extracted cytoplasm. Our study focused on those cells that remained unbroken with intact plasma membrane (Fig 2). However, despite their integrity, these cells displayed increased number of zymogen granules of smaller sizes and numerous lysosomes of various forms and states of maturation. However, no fusion of lysosomes with secretory granules was observed in these cells. The Golgi apparatus appeared to be rather fragmented, the cisternae giving rise to small vacuolar profiles (Fig 2). In animals with diabetes (Fig 3), in condition of stimulation of secretion, and in starvation (not illustrated), the cells maintained their shape and morphological aspect. However, the zymogen granules decreased during stimulation of secretion and starvation, both in number and in size, as reported previously (Bendayan et al 1985).

Fig 1.

Immunocytochemical labeling of Cpn60 in pancreatic acinar cells from normal animals. (a) Anti-Cpn60 followed by protein A-gold. The gold particles revealing Cpn60 antigenic sites are mainly located over the zymogen granules (g). Some are also present over the rough endoplasmic reticulum (RER) and the Golgi apparatus (G). Mitochondria (m) also display a significant labeling for Cpn60. Nucleus (N) is devoid of labeling. Bar = 0.5 μm. (b) Anti Cpn60 + Cpn60, followed by protein A-gold. Upon adsorption of the antibody with its antigen, the labeling by gold particles is almost abolished. Bar = 0.5 μm. (c) Omission of the antibody step. Very few gold particles are seen on the different cellular compartments. Bar = 0.5 μm

Fig 2.

Pancreatic acinar cell in acute pancreatitis condition. The morphological aspect to the Golgi apparatus (G) is quite altered with the presence of many small vacuoles and vesicles. The labeling for Cpn60 is present in the vacuolated Golgi cisternae and related vesicles as well as in the zymogen granules (g). However, the labeling appears much lower than in tissues of the normal animals. Mitochondria (m) retained normal levels of labelings. Bars = 0.5 μm

Fig 3.

Pancreatic acinar cell in diabetic condition. Labeling by gold particles, reflecting Cpn60 antigenic sites, is intense over the zymogen granules (g) and the Golgi apparatus (G). It is also present over the rough endoplasmic reticulum (RER). Labeling is also present over mitochondria (m). Bar = 0.1 μm

The gold particles revealing Cpn60 antigenic sites were present over the RER, Golgi cisternae, and zymogen granules (Figs 1–3). Few were over lysosomes and nuclei. In addition, Cpn60 immunolabeling was detected in significant amounts over mitochondria (Figs 1–3). Labeling experiments carried out to assess the specificity of these results yielded very low to no labeling in all cellular compartments when the first antibody was preadsorbed with its corresponding antigen (Fig 1b) and when it was omitted (Fig 1c).

Quantitative evaluation of the labelings for Cpn60 revealed increasing intensities along the RER-Golgi-granules secretory pathway both in control and in most of the experimental conditions (Table 1). However, these intensities did vary according to the experimental condition and the cellular compartment examined.

Table 1.

Cpn60 immunolabelings in intracellular compartments of rat pancreatic acinar cells under different experimental conditions

The intensities of labeling through the secretory pathway declined under conditions of acute pancreatitis, stimulation of secretion, and starvation (Table 1). In diabetes, on the other hand, they increased significantly (Table 1). The changes in the granules were the most significant. Compared with the control condition, Cpn60 immunolabeling in zymogen granules decreased about 40% in acute pancreatitis, 60% under stimulation of secretion, and 75% in starvation. It increased 3-fold in diabetes (Table 1). Treatment of diabetic animals with insulin normalized the levels of Cpn60 in the secretory pathway (Table 1). Cpn60 labelings in mitochondria, on the other hand, remained rather constant in all conditions except starvation, where a 55% decrease was registered (Table 1).

DISCUSSION

Molecular chaperones including chaperonins and heat shock proteins (Hsps) are large families of specialized proteins present in various cellular compartments (Ellis and van der Vies 1991; Hendrick and Hartl 1993; Fink 1999). They play important roles in enabling polypeptides to reach biologically active forms. Chaperonins, including Cpn60 and Cpn10, are known to be involved in successful folding, sorting, transmembrane transport, and assembly of oligomeric polypeptide complexes (Ranson et al 1998; Keskin et al 2002). In addition, Cpn60 as well as Cpn10 were reported to follow the regulated secretory pathway in exocrine as well as endocrine cells (Brudzynski et al 1992; Arias et al 1994, 2000; Velez-Granell et al 1994; Cechetto et al 2000; Soltys and Gupta 2000; Sadacharan et al 2002). In pancreatic acinar cells, Cpn60 follows the well-characterized RER-Golgi-granules secretory pathway, being concentrated sequentially along this route and finally discharged into the pancreatic juice (Velez-Granell et al 1994). The demonstration of close associations between Cpn60 and some pancreatic secretory enzymes in an earlier work (LeGall and Bendayan 1996) prompted us to investigate its fate under experimental conditions known to alter protein secretion.

It is well known that acute pancreatitis is a multiple-stage disease with high levels of morbidity and mortality. Independent from different etiologic causes, it results in an intracellular activation of trypsinogenic and lipolytic enzymes, which in turn injure the acinar cells and obliterate their integrity (Andrzejewski et al 1996; Whitcomb 1999; Lerch and Gorelick 2000; Mayer et al 2000). Stress induced by water immersion of animals induces an increase in Cpn60 expression in pancreatic cells, preventing intrapancreatic trypsinogen activation and protecting against cerulein-induced rat pancreatitis (Otaka et al 1997; Lee et al 2000; Rakonczay et al 2002). Thus, Cpn60 may play key roles in preventing early enzyme autoactivation. Along this line, Strowski et al (1997) have reported that the pancreas reacts to various types of stress factors with different inductions of Hsp messenger ribonucleic acids. They proposed that during cerulein-induced acute pancreatitis, failure to appropriately increase Hsp levels leads to the development of pancreatitis. In our study, pancreatic tissue in acute pancreatitis condition displayed a major decrease in Cpn60, which may have led to insufficient protein quality control along the secretory pathway. This in turn could have allowed aggregation and autoactivation of some pancreatic proteases. It is of interest to underline that pancreatic secretory proteins, such as amylase and lipase, that are not in zymogen-inactivated forms were those found to be associated with Cpn60 (LeGall and Bendayan 1996).

In diabetes, Cpn60 content in the acinar cells increased almost 3-fold, along with significant increases in lipase and chymotrypsinogen (Gregoire and Bendayan 1986; Bendayan and Levy 1988). However, levels of amylase, an enzyme found to be associated with Cpn60, were reported to decrease markedly in hyperglycemic conditions (Gregoire and Bendayan 1986). All these changes were restored by insulin treatment and normalization of blood glucose levels (Gregoire and Bendayan 1986). Because the hyperglycemic condition in the present study was maintained for 2 months, the changes reported previously for the secretory enzymes (Gregoire and Bendayan 1986; Bendayan and Levy 1988) and in this study for Cpn60 must reflect adaptation processes to hyperglycemic conditions.

It is well known that pancreatic enzyme secretion is stimulated by the cholinergic system and by cholecystokinin (Adler et al 1991; Del Rosario et al 2000). Treatment of the rats with carbamyl β-methylcholine, an analog of acetylcholine, led to decreases in levels of Cpn60 that coincided with decreases in pancreatic enzymes (Bendayan et al 1985). On the other hand, starvation condition, which reduces enzyme secretion, also led to decreased amounts of Cpn60. Previous data reported that fasting for 48 hours results in a 34% reduction in total pancreatic proteins along with an almost parallel decrease of 41% in amylase (Bendayan et al 1985). The interesting observation in our study is the parallel changes in secretory enzymes and in Cpn60 in these 2 conditions, which implies that secretory Cpn60 and the pancreatic enzymes in these particular cases, could be regulated by the same factors. However, such parallel processing of secretory enzymes and Cpn60 was not encountered in the 2 other experimental conditions, namely pancreatitis and diabetes. The secretory proteins and Cpn60 behave differently. In diabetes, decreases in amylase and increases in lipase and chymotrypsinogen have been reported (Gregoire and Bendayan 1986; Bendayan and Levy 1988). Cpn60 undergoes significant increases. Furthermore, lowering blood glucose levels normalizes pancreatic enzymes secretion (Gregoire and Bendayan 1986; Bendayan and Levy 1988), along with the levels of secretory chaperonin. However, in acute pancreatitis, the decreases in Cpn60 were observed very shortly after the onset of the experiment. Changes in secretory proteins occur only at later stages of pancreatitis. Thus, the rapid changes of Cpn60 in pancreatitis must reflect nonparallel processing and discharge of the chaperonin. Nonparallel secretion is a well-described, poorly understood phenomenon (Rothman 1975; Adelson and Miller 1985). It was defined as the very rapid changes in secretion of 1 particular protein that is synthesized and discharged preferentially to the others while still being packaged into the same secretory zymogen granule and follows the regulated secretory pathway. Our present results on Cpn60 bring additional elements to this subject. The nonparallel processing of Cpn60 occurring in pancreatitis could in fact alter the steady-state equilibrium between some pancreatic secretory enzymes and chaperones, thus initiating premature enzyme autoactivation.

A second interesting observation made on the expression of Cpn60 by pancreatic cells under experimental conditions refers to mitochondrial Cpn60. In contrast to the Cpn60 present in the secretory pathway, mitochondrial Cpn60 remained rather stable in all the conditions tested except starvation, where it displayed a significant decrease. Such differences indicate that expression of mitochondrial Cpn60 and secretory Cpn60 in pancreatic acinar cells must be regulated by different mechanisms. This is not surprising considering the different roles played by chaperonins in cells. Mitochondrial metabolism and function not being directly involved in secretion are thus not significantly altered in the experimental conditions tested. Amounts of mitochondrial Cpn60 remained unchanged. Although no definite role was assigned to the secretory Cpn60, this study has established independent behavior for the secretory and the mitochondrial Cpn60.