| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Section on Cellular Neurobiology, Laboratory of Developmental Neurobiology, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Dr. Y. Peng Loh, Building 49, Room 5A38, 49 Convent Drive, MSC 4490, Bethesda, Maryland 20892-4490. E-mail: ypl{at}codon.nih.gov.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A defect in the CPE gene has been identified as the cause of the fat/fat phenotype of the CPEfat/fat mouse, which exhibits multiple endocrine disorders, including obesity, infertility, hyperproinsulinemia, and diabetes (6). The mutation in the CPE gene results in the expression of a mutant form of the CPE protein harboring a Ser to Pro mutation (6) that results in an enzymatically inactive and unprocessed protein [proCPE(202)] (7, 8) that is differentially degraded in the tissues of these mice. For example, in the pituitary, no immunoreactive CPE can be seen (6), whereas in the pancreas (6) and brain (9) CPE is present, although at lower levels than in their normal littermates. In an immortalized pancreatic ß-cell line (NIT3) derived from the CPEfat/fat mouse there are also high levels of proCPE(202) present (8), but when transfected into a mouse anterior pituitary cell line, AtT20, the mutant CPE is completely degraded (7).
The absence of CPE in the pituitary of the CPEfat/fat mouse correlates well with the lack of regulated secretion of proopiomelanocortin (POMC). For POMC, binding to CPE is through its sorting signal motif encoded in the N terminus of the POMC protein (3, 10). In the absence of CPE, either in primary cultures of the CPEfat/fat pituitaries (11) or by antisense down-regulation of CPE in a neuroendocrine cell line (3), POMC was missorted from the regulated to the constitutive secretory pathway in vivo, demonstrating the reliance of POMC on the presence of CPE in the TGN for sorting to the RSP. However, the involvement of CPE in the sorting of proinsulin to the RSP has been drawn into question by two studies based on the observation that (pro)insulin was secreted in a regulated manner from primary cultures of pancreatic islets of the CPEfat/fat mouse (12) or from NIT3 (8) cells, indicating that its sorting to the RSP was independent of CPE. This conclusion was based on the assumption that proCPE(202) was retained and fully degraded in the endoplasmic reticulum (ER), such that it was completely absent from the RSP.
The substantial presence of proCPE(202) in the pancreas of CPEfat/fat mice and in NIT3 cells as well as our recent study showing that the proCPE(202) mutant itself can bind proinsulin with similar efficiency as wild-type CPE (13) prompted us to investigate the possibility that in NIT3 cells, some proCPE(202) could escape degradation, exit the ER, and reach the TGN, where it could act as a sorting receptor.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Pulse-chase analysis of CPE in NIT3 cells
Pulse-chase experiments were performed essentially as described by Varlamov et al. (8) except with more time points and inclusion of 0.2 mM phenylmethylsulfonylfluoride in the chase medium. In a previous experiment (not shown) we found that the presence of 0.2 mM phenylmethylsulfonylfluoride in the culture medium of NIT3 cells had no effect on the level of cellular proCPE(202). The cells were harvested in 50 mM Tris-Cl (pH 7.4), 0.1% Triton X-100, and protease inhibitors (1x Complete mini protease inhibitor cocktail, Roche Molecular Biochemicals, Indianapolis, IN) and extracted twice at 4 C. The CPE in the supernatants was immunoprecipitated as described previously (3) using antisera to both the amino and carboxyl termini of CPE (3). The immunoprecipitates were analyzed by SDS-PAGE using 12% polyacrylamide Tris-glycine gels (Invitrogen, San Diego, CA) followed by autoradiography. In a similar experiment, NIT3 cells were preincubated for 2 h with either 50 µM lactacystin or 0.2 mM chloroquine before the pulse-chase experiment. The cells were then chased for 4 h in the presence of these compounds, after which the cellular content of proCPE(202) was determined by immunoprecipitation.
Western blot analysis of CPE and proCPE(202)
NIT1 and NIT3 cells were grown to a similar confluence in 60-mm dishes, rinsed three times with Hanks’ balanced salt solution (Life Technologies, Inc.), and then extracted in lysis buffer [50 mM HEPES (pH 7.4), 150 mM NaCl, 10 mM EDTA, 10 mM sodium pyrophosphate, 100 mM sodium fluoride, 1% Triton X-100 and 1x Complete mini protease inhibitor cocktail] at 4 C. The soluble extracts, obtained after centrifugation of the total cell lysates at 13,000 rpm for 20 min, were assayed for protein content, and 10 µg from each extract were analyzed by Western blot with a new antibody that was raised in guinea pigs against an internal sequence of mouse CPE. The peptide, C345EKFPPEETLKSYWEDNK362, was custom-synthesized by Covance Laboratories, Inc. (Denver, PA) and conjugated to keyhole limpet hemocyanin via the amino-terminal Cys residue (CPE numbering starts from the translational start site). The blots were probed with the new anti-CPE antiserum (no. 1415; 1:5,000) overnight at 4 C. Visualization of the Western blot signal was by enhanced chemiluminescence (Pierce Chemical Co., Rockford, IL) using goat antiguinea pig IgG coupled to horseradish peroxidase (1:10,000; Sigma, St. Louis, MO) as the secondary antibody. As controls, parallel lanes were probed with 1415 preimmune serum (not shown) or with 1415 immune serum preabsorbed with the antigenic peptide at a concentration of 10 µg/ml.
Indirect immunofluorescent microscopy
NIT3 cells were grown in two-chambered plastic slides, fixed in 2% paraformaldehyde/1x PBS for 30 min, and then permeabilized by 0.25% Triton X-100/1x PBS for 5min. After blocking with 1% BSA/1x PBS for 2 h at room temperature the cells were then incubated for 24–36 h at 4 C with primary antisera diluted as indicated in 1% BSA/1x PBS. The antisera used were guinea pig anti-CPE antiserum (no. 1415-2; 1:5000) raised against an internal sequence of CPE, rabbit antiprohormone convertase 2 (anti-PC2) antiserum (1:2000) provided by Iris Lindberg (St. Louis, MO), and rabbit anticalnexin (1:2000; Sigma). After washing with 1x PBS, the cells were then incubated with their respective secondary antibodies for 1 h at room temperature. For CPE detection, goat antiguinea pig IgG-Alexa-488 was used, and for calnexin and PC2 detection, goat antirabbit IgG-Alexa-568 was used. The secondary antibodies were obtained from Molecular Probes, Inc. (Eugene, OR), and used at a concentration of 1:1000 in 1% BSA/1x PBS. Preimmune serum, omission of primary antibody, and absorption of the CPE antiserum with the CPE peptide were used as negative controls. Immunofluorescence was observed on an MRC-1024 Laser Scanning Confocal Imaging System using LaserSharp software (both from Bio-Rad Laboratories, Inc., Richmond, CA). The laser source was a krypton/argon mixed gas laser emitting energy at 488, 568, and 647 nm. An optical magnification of x300 was obtained using a x100 objective lens and x3 magnification, and the images were captured through filters HQ598/40 for Alexa-568 emission and 522/35 for Alexa-488 emission. For preimmune and absorption control slides (see Fig. 3
), the images were acquired using identical parameters as the positively immunostained slides. Images were processed identically and arranged using Adobe Photoshop 6.0 (Adobe Systems, Inc., San Jose, CA).
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows that after the 30-min pulse, cellular proCPE(202) decreased at a relatively slow rate, with about 77% still remaining after 90 min. This result is similar to that described previously, where little if any degradation of proCPE(202) was observed after 90 min (8). In contrast to that report, however, we found that approximately 15% of the initial labeled proCPE(202) was still present in the cells after 6 h, which probably represents levels approaching steady state. ProCPE(202) was confirmed to be at least partially degraded by the proteosome, as demonstrated by the prevention of its loss by about 41% within 4 h when the proteosome was irreversibly inhibited by lactacystin (Fig. 1B). Chloroquine treatment had no effect on the loss, demonstrating and reaffirming previous results (8) that proCPE(202) is not degraded in an acidic compartment such as the lysosome.
|
). The sizes of these immunoreactive CPEs were consistent with that of mature CPE (
55 kDa) in NIT1 and proCPE (57 kDa) in NIT3 and are similar to those reported previously (8). In control experiments no immunoreactivity was detected when preimmune serum was used (data not shown) or when the immune serum was preabsorbed with the antigenic peptide (Fig. 2, right), demonstrating the specificity of this new antiserum.
|
). Double-label indirect immunofluorescent microscopy showed that the punctate staining of proCPE(202) did not colocalize with calnexin (Fig. 4, upper panel). This was most obvious in areas close to the cell boundaries where there was an absence of calnexin staining. In addition, staining of calnexin in the perinuclear region was commonly devoid of proCPE(202) staining. Both of these observations indicated that a portion of the proCPE(202) in these cells is not retained in the ER. In double-labeling with prohormone convertase 2 (PC2), a secretory granule marker protein, proCPE(202), was colocalized with PC2, indicating that the proCPE(202) that was not in the ER was present in granules of the RSP.
|
). As a control, mature CPE from NIT1 cells was effectively released in a stimulated manner, as was PC2 from both NIT1 and NIT3 cells (Fig. 5). Five secretion experiments of this type were performed, giving similar results.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although CPE enzymatic activity is apparently absent in these mice (7, 9), the protein itself is present at differing levels in different tissues. As noted previously, the pituitary appears to be devoid of CPE protein (6); however, levels of the mutant CPE are readily detectable in the pancreas (6) and brain (9) of these mice, indicating that differential degradation of the mutant CPE was occurring. In all other studies using tissues of the CPEfat/fat mouse and where the presence of CPE immunoreactivity was tested, it was only tested on affinity-purified CPE-like activity, resulting in the complete absence of signal. Unfortunately, as proCPE(202) does not bind to this affinity column (7), it is unclear what levels of the protein are actually present in these tissues. In vitro binding and chemical cross-linking studies (15), an antisense knockout study (16), and crystal structure analysis (13) have indicated that proinsulin can bind to the membrane form of CPE in a similar manner to that of N-POMC1–26, a peptide that contains the POMC RSP sorting signal. These studies are the basis of the hypothesis that CPE may be involved in the sorting of proinsulin to the RSP in vivo. In an effort to address this question, Irminger et al. (12) performed secretion experiments on primary cultures of pancreatic islets from CPEfat/fat mice. In these experiments the researchers showed that there was no difference in the stimulated secretion of (pro)insulin between cultures from control and CPEfat/fat islets and concluded that CPE was not involved in the sorting of proinsulin to the RSP. Unfortunately, pulse-chase experiments to determine the sorting efficiency of proinsulin were not performed using the primary cultures. In addition, levels of CPE protein were not measured in these cultures, allowing for the possibility that the mutant CPE was present, as in pancreatic tissue, and might have acted in some way to facilitate the sorting of the proinsulin to the RSP, as previous experiments suggested (13, 15, 16).
However, this possibility was not supported by subsequent work of Varlamov et al. (8). In an immortalized pancreatic ß-cell line (NIT3) derived from the CPEfat/fat mice, the researchers showed that the mutant CPE, which was present in significant amounts, was not converted to mature CPE, a step that normally occurs in a post-Golgi compartment (17) and that it had a higher turnover compared with wild-type CPE from a control pancreatic ß-cell line, NIT1 (18). In addition, as they did not detect it in the culture medium, and the mutant had a general ER staining pattern by ICC, the researchers concluded that the mutant CPE was expressed, retained, and completely degraded by the ER. They also showed stimulated secretion of (pro)insulin from NIT3 cells, albeit at two third less than from NIT1 cells, suggesting that CPE played no role in proinsulin sorting to the RSP in these cells.
In the present work, we provide new and contrasting evidence showing that in NIT3 cells, some of the mutant CPE can escape degradation, exit the ER and enter a PC2-containing granular compartment consistent with its identification as either immature secretory granules (ISG) and/or MSG of the RSP. Our evidence stemmed originally from secretion studies of NIT3 cells, in which we observed proCPE(202) in the medium, in contrast to the report by Varlamov et al. (8). Further investigation revealed that proCPE(202) could be secreted in a secretegogue-dependent manner by GLP-1. This result alone demonstrated that some of the proCPE(202) in NIT3 cells escaped degradation and was trafficked through the secretory pathway. It is possible that our use of protease inhibitors, our concentration procedure, and our new antibody allowed us to detect proCPE(202) in the medium, in contrast to Varlamov et al. (8).
We further investigated the stability of proCPE(202) in NIT3 cells by pulse-chase analysis. We found that, similar to Varlamov et al. (8), radiolabeled proCPE(202) was relatively stable up to 90 min after labeling and indeed was quite long-lived for a mutant protein (t1/2 = 3 h). Even at 6 h, about 15% of the labeled proCPE(202) remained, suggesting that this was approaching steady state levels. To determine whether the slow rate of disappearance was solely due to proteosomal degradation, we performed the pulse-chase experiment on NIT3 cells treated with 50 µM lactacystin, a potent, irreversible inhibitor of the 20S subunit of the proteosome. In untreated NIT3 cells, 64% of the proCPE(202) was lost after 4 h, whereas in lactacystin-treated NIT3 cells, approximately 38% was lost. This represents about 41% efficacy in the prevention of loss that occurred within 4 h. As previous experiments by others have shown that use of 20 µM or more lactacystin or its ß-lactone derivative on a number of cell lines results in 75–95% inhibition of the proteosome (19, 20), this suggested the possibility that the loss of radiolabeled proCPE(202) from the cell lysate was not exclusively by the proteosome. If in the worst case scenario that only 75% of the proteosome was inhibited under our experimental conditions, we can extrapolate that had it been 100% inhibited we would have prevented the loss of proCPE(202) by only about 55% within 4 h, suggesting that up to 45% of the labeled proCPE(202) was not degraded by the proteosome. However, as approximately 15% of the labeled proCPE(202) in the pulse-chase experiment remained after 6 h, a time point considered to be approaching steady state, it is likely that further intracellular degradation and/or constitutive secretion and degradation occurred after 4 h to account for the lower levels of proCPE(202) observed at steady state. Also, similar to Varlamov et al. (8), we showed that the proCPE(202) was not degraded by the lysosome, as evidenced by the lack of prevention of degradation in the presence of chloroquine, a compound that neutralizes the acidic environment of the lysosome, thus preventing the activity of degradation enzymes.
As it appeared that a portion of the proCPE(202) escaped proteosomal degradation, we attempted to detect proCPE(202) by ICC using a new antibody that we developed in our laboratory to an internal sequence of CPE. Our results show that the new antiserum was specific for CPE under nonnative and native conditions, as judged by Western blot analysis and ICC, respectively. In addition, we found that this new antiserum was much more sensitive than our previous CPE antisera (data not shown). In NIT3 cells, proCPE(202) was distributed in a diffuse pattern as well as in a distinct punctate pattern characteristic of secretory granules. This is in contrast to the observations reported by Varlamov et al. (8), which showed that proCPE(202) in NIT3 cells has only a diffuse and generally nonpunctate staining pattern using a C-terminal-specific antibody that was interpreted as ER localization because of its similarity to calnexin staining in these cells. However, in the absence of double-label ICC in their studies, it was difficult to assess whether all of the CPE staining was ER localized or whether a proportion of it was in another compartment. To answer this question we performed colocalization studies by double-label ICC of proCPE(202) with calnexin or PC2. Figure 4 shows that although there is colocalization of calnexin, a marker for the ER, and proCPE(202), as expected from the previous report (8), a significant portion of the proCPE(202) did not colocalize with calnexin, as evidenced by the green signal of proCPE(202) in the merged panel (Fig. 4, upper panel merge). Additionally, the presence of calnexin in the perinuclear region that was significantly devoid of proCPE(202) staining indicated that the mutant CPE did not colocalize with calnexin in this area either. The proCPE(202), however, was colocalized with PC2, a granule marker protein (Fig. 4, lower panel), demonstrating that some of the mutant CPE had exited the ER, traversed the secretory pathway, and entered granules of the RSP. This is not unlike a recent report of a misfolded yeast plasma membrane H+-adenosine triphosphatase that could escape degradation by the proteosome. In this case after a delay in the ER, the mutant was trafficked to the plasma membrane, where it was internalized and ultimately degraded by the lysosome (21).
Our results presented here provide new insights into the fate of proCPE(202) in NIT3 cells. Although our results concur with Varlamov et al. (7), with respect to the ER degradation of proCPE(202), we now provide new evidence that up to 45% of the protein may escape proteosomal degradation in NIT3 cells. We believe that one critical difference between their results and ours is due to our new antibody, which is extremely sensitive for detecting an internal sequence of CPE. The costaining of granules with both CPE-specific and PC2-specific antisera and the stimulated secretion of both of these proteins by GLP-1 indicate that they are granules of the RSP.
The presence of proCPE(202) in the ISG/MSG is significant because of its potential involvement in sorting proinsulin to the RSP. In its capacity as a sorting/retention receptor, CPE may be a component of the sorting by entry model at the TGN (22) or as a retention receptor for (pro)insulin in the ISG (23). The works of Irminger et al. (12) and Varlamov et al. (8) have been cited repeatedly as evidence for the noninvolvement of CPE in this process; however, as we showed previously that proCPE(202) can bind proinsulin (13) and now that potentially as much as 45% of the proCPE(202) can exist in the secretory pathway, we can say that this is not definitive. Although the present study does not address the exact role of CPE in proinsulin sorting, a more definitive study using a CPE knockout mouse is currently in progress.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 5, 2002.
Accepted for publication September 17, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Hosaka, T. Watanabe, Y. Sakai, T. Kato, and T. Takeuchi Interaction between secretogranin III and carboxypeptidase E facilitates prohormone sorting within secretory granules J. Cell Sci., October 15, 2005; 118(20): 4785 - 4795. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. X. Cawley, J. Zhou, J. M. Hill, D. Abebe, S. Romboz, T. Yanik, R. M. Rodriguiz, W. C. Wetsel, and Y. P. Loh The Carboxypeptidase E Knockout Mouse Exhibits Endocrinological and Behavioral Deficits Endocrinology, December 1, 2004; 145(12): 5807 - 5819. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Boulatnikov and R. C. De Lisle Binding of the Golgi Sorting Receptor Muclin to Pancreatic Zymogens through Sulfated O-linked Oligosaccharides J. Biol. Chem., September 24, 2004; 279(39): 40918 - 40926. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
