517 Impact of uncoupling protein-2 overexpression on processing

Narudee Kashemsant and Catherine B Chan Department of Biomedical Sciences, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3

(Requests for offprints should be addressed to C B Chan; Email: [email protected])

Abstract

Hyperproinsulinemia is observed in type 2 diabetic patients. We hypothesized that the induction of uncoupling protein-2 (UCP2) would impair processing of proinsulin to mature and potentially contribute to hyperproinsulinemia, based on the evidence that hormone processing is an ATP-dependent process and UCP2 up-regulation can suppress cellular ATP production. UCP2 was overexpressed (UCP2-OE) by twofold in INS-1 cells by means of plasmid transfection. Although UCP2-OE reduced glucose-stimulated insulin secretion and cellular ATP content, no effects on proinsulin processing, as measured by western blotting, were observed. To increase the demand for insulin, we then cultured UCP2-OE and control INS-1 cells in medium containing 20 mM KCl for 24 h. High KC markedly reduced glucose- stimulated insulin secretion from control cells, indicating inability of cells to meet secretory demand. Independent of UCP2 expression, high KC reduced preproinsulin mRNA expression but had no effect on ATP content despite increasing ATP synthase expression. In UCP2-OE cells, high KC decreased total cellular insulin species content and increased the ratio of proinsulin to insulin, indicating an impairment of processing. We conclude that UCP2-OE can negatively impact C proinsulin processing, possibly by ATP-dependent alteration of the granule environment or reduction of Ca2 availability, particularly when cells are chronically stimulated to secrete insulin. Journal of Molecular Endocrinology (2006) 37, 517–526

Introduction In granules, addition of ATP stimulates proinsulin conversion, whereas use of an ATPase inhibitor Type 2 diabetic patients secrete an elevated fraction of reduces conversion (Rhodes et al. 1987). Therefore, proinsulin relative to insulin (Kahn & Halban 1997) changes in b-cell metabolism that impair ATP gener- and an increased proinsulin to insulin ratio in ation may decrease maturation of insulin. apparently healthy individuals predicts future develop- In b-cells, ATP production can be decreased by an ment of type 2 diabetes (Hanley et al. 2002, Pradhan increase in uncoupling protein-2 (UCP2) expression or w et al. 2003, Zethelius et al. 2003). Proinsulin has 5% of activity (Chan et al. 2004). UCP2 is an inner mito- the biological activity of mature insulin (Goodge & chondrial membrane protein expressed widely in many Hutton 2000) and therefore its excessive secretion tissues including (Fleury et al. 1997, Gimeno contributes to worsening of glucose tolerance et al. 1997). UCP2 transports either unprotonated free (Mykkanen et al. 1999). fatty acids (Garlid et al. 1998) or protons (Klingenberg Processing of insulin to its mature form involves 1999) to dissipate the proton motive force and, prohormone-mediated cleavage of C-peptide from consequently, reduce the potential energy available for proinsulin by subtilisin-like protein convertases conversion of ADP to ATP (Schrauwen & Hesselink (SPCs)-2 and -3 (Irminger et al. 1996). Factors that 2002). Overexpression of UCP2 inhibits glucose-stimu- influence the function of SPCs, such as granule pH or lated insulin secretion (Chan et al. 1999, 2001, Hong et al. 2C [Ca ], might be altered in the b-cell of diabetic 2001), whereas its absence promotes enhanced insulin 2C patients. The P-domain of SPCs confers both Ca and release (Zhang et al. 2001). The role of endogenous pH-dependence (Steiner 1998). Thus, basification of UCP2 is less clear; for example, its up-regulation 2C pH or reduction in [Ca ] in the granule impairs following overexpression of its transcription factor maturation (Rhodes et al. 1987, Davidson et al. 1988). peroxisome proliferator-activated receptor-a is reported C Regulation of granule pH and [Ca2 ] is an ATP- to decrease (Tordjman et al. 2002) or not be a factor in dependent process and intracellular energy homeo- (Ravnskjaer et al. 2005) glucose-stimulated insulin stasis is an important determinant of both insulin secretion. Here, we explore the relationship between biosynthesis and processing (Goodge & Hutton 2000). UCP2 overexpression and proinsulin processing.

Journal of Molecular Endocrinology (2006) 37, 517–526 DOI: 10.1677/jme.1.02091 0952–5041/06/037–517 q 2006 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access 518 N KASHEMSANT and C B CHAN . UCP2 and proinsulin processing

Materials and methods 5 min. The cells then were added to the substrate buffer agitating again for 5 min. The samples were kept in the Cell culture, UCP2 transfection, or oligomycin dark for 10 min before measuring luminescence on an treatment I450 Micro Beta Scintillation and Luminescence detector (Perkin-Elmer). The remainder of each cell sample was Rat insulinoma cells (INS-1 cells (Asfari et al. 1992), a measured for protein concentration using the Lowry kind gift of Dr Bruce Verchere, Vancouver, Canada) method (Sigma-Aldrich) and data were expressed as were cultured overnight in RPMI 1640 medium (Fisher nmol/mgprotein. Scientific, Ottawa, Canada) with 11.1mM D-glucose, 10 mM HEPES, 10% fetal bovine serum (FBS), 2 mM glutamine, 100 U/ml penicillin, 100 U/ml strepto- Western blotting mycin, 1 mM sodium pyruvate, and 50 mM 2-mercaptoethanol. The cells were plated in 24-well Proinsulin and insulin plates (250 000 cells/well) and cultured at 37 8Cina INS-1 cell lysates were prepared using 3% acetic acid. humidified 95% air–5% CO atmosphere for 24 h 2 The protein concentration was determined using the before conducting UCP2 plasmid transfection. Lowry method. Protein samples (100 mg) were separ- Plasmid (pcDNA 3.1, Invitrogen) containing the full- ated on 15% polyacrylamide gels under non-reducing length cDNA sequence for human UCP2 was transfected conditions, then electrophoretically transferred onto into overnight cultured INS-1 cells as described (Chan nitrocellulose membranes (Trans-Blot, Bio-Rad), et al. 2001). In selected control experiments, plasmid followed by blocking of non-specific binding in 5% encoding enhanced green fluorescence protein (EGFP) skim milk–1% Tween 20/PBS overnight at 4 8C with was transfected. After 24 h, both control (non-trans- agitation. Details of antibodies are given in Table 1. fected) and UCP2- or EGFP-transfected cells were Specific signals were detected using ECL Plus reagent cultured for a further 24 h in either RPMI with 10% FBS according to the manufacturer’s instructions (GE or RPMI supplemented with 20 mM KCl and 10% FBS Healthcare, Piscataway, NJ, USA). The intensity of before use. Addition of KCl changed the total osmolarity, reactive bands was quantified using Kodak Image directly measured by an osmometer, by !10 mosmol/kg. . station 440 (Perkin-Elmer). Subsequently, after strip- In all cases, the RPMI contained 5 5mMglucose. ping, membranes were incubated with goat anti-C- To lower ATP by inhibiting ATP synthase, oligomycin . peptide (Table 1) in order to resolve the identity of (in DMSO; final concentration 0 1 ng/ml) was added mature insulin versus proinsulin bands. to INS-1 cells cultured in RPMI. Control cells were treated with an equal volume of DMSO. The cells were harvested for use in assays 12 h later. Prohormone convertase, ATP synthase, and UCP2 The cells were lysed with N-tris[hydroxymethyl] methyl- 2-aminoethane-sulfonic acid buffer containing 0.1% Insulin secretion assay and ATP content SDS, and aprotinin (1 mg/ml) and bestatin (10 mg/ml) Transfected, oligomycin-treated and control INS-1 cells as protease inhibitors. Total protein was separated by were washed and preincubated twice for 30 min in Krebs 12% SDS-PAGE. In some experiments, cell fractionation Ringer bicarbonate buffer (KRBB; pH 7.4), supple- to enrich mitochondrial and cytosolic fractions was mented with 10 mM HEPES and 0.1% BSA. After a carried out using the Qproteome Compartment kit second brief wash, the cells were incubated for 2 h in (Qiagen). The proteins were electrotransferred onto KRBB containing 2.8, 11, or 22 mM glucose. Following nitrocellulose membranes and blocked with 5% skim incubation, the KRBB was transferred to microcentrifuge milk containing 1% Tween 20 in PBS for 45 min. Details tubes and the adherent cells were extracted with 3% of antibodies are given in Table 1. Specific signals were acetic acid, with both fractions frozen at K20 8Cuntil detected and quantified as above. Protein loading was assayed for insulin. Insulin secretion in response to normalized using mouse anti-b-actin antibody (Table 1). glucose was quantified as a percent of total cellular insulin content using a RIA with rat insulin as standard. Quantitative RT-PCR INS-1 cells were incubated exactly as described above for insulin secretion before measuring ATP using ATPlite Total RNA was extracted from 106 cells of either control (Perkin-Elmer, Woodbridge, ON, Canada), which is or UCP2 overexpressing (UCP2-OE) cells using TRIzol based on the production of light caused by the reaction reagent (Invitrogen) according to the manufacturer’s of ATP with added luciferase and D-luciferin. Briefly, after protocol. cDNA synthesis was carried out using Cloned trypsinization and washing in 0.1M BSA-free PBS AMV First-Strand cDNA Synthesis kit (Invitrogen). All (pH 7.4), one portion of INS-1 cells was disrupted with cDNA samples were amplified in PCRs using specific mammalian lysis solution with agitation at 700 r.p.m. for primers as listed in Table 1.

Journal of Molecular Endocrinology (2006) 37, 517–526 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access UCP2 and proinsulin processing . N KASHEMSANT and C B CHAN 519

Table 1 Antibodies for immunoblotting and primer sets for quantitative PCR

Primary antibody Forward primer (50 to 30) Reverse primer (50 to 30)

Gene product (Pre)(pro)insulin Guinea pig anti-insulin 1:1000, 4 8C, GGAACGTGGTTTCTTCTACAC GGGAGTGGTGGACTCAG ON (R A Pederson, Vancouver, BC, (preproinsulin I) GACCAGCTA- (preproinsulin I) TCCACGATG- Canada) CAGTCGGAAAC (preproinsulin II) CCGCGCTTCTG preproinsulin II) (Toriumi & Imai 2003) C-peptide Goat anti-C-peptide 1:1000, RT, 1 h N/A N/A (Linco, St Charles, MI, USA) SPC3 Rabbit anti-SPC3 (N P Birch, N/A N/A Auckland, New Zealand) 1:1000, 4 8C, ON ATP synthase Mouse anti-ATP synthase subunit a N/A N/A 1:2000, 4 8C, ON (Invitrogen) UCP2 Goat anti-UCP2 1:2500, 4 8C, ON CAGCCAGCGCCCAGTACC CAATGCGGACGGAGGCAAAGC (Santa Cruz Technologies, (Joseph et al. 2002) Santa Cruz, CA, USA) b-Actin Mouse anti-b-actin 1:5000, RT, 1 h N/A N/A 18S rRNA N/A CTTTGGTCGCTCGCTCCTC CTGACCGGGTTGGTTTTGAT (Proudnikov et al. 2003)

ON, overnight; RT, room temperature. All secondary antibodies were from Sigma-Aldrich. Membranes were incubated in appropriate secondary antibody at a dilution of 1:15 000 for 1 h at room temperature.

Real-time PCRs were carried out using 1 ml cDNA culture medium, 18S rRNA expression was not changed C (1 mg/ml) and SYBR green (Invitrogen). Reaction by UCP2 transfection or K (Fig. 1B). A representative conditions included 40 cycles of melting at 95 8C for immunoblot for UCP2 (32 kDa) and b-actin (43 kDa) is 15 s, annealing at 53 8C for 45 s, and extension at 72 8C shown in Fig. 1C. UCP2 protein expression was increased for 20 s (for UCP2 and 18S); 40 cycles of melting at approximately twofold by plasmid transfection (P!0.01, 95 8C for 20 s, annealing at 55 8Cfor20s,and Fig. 1D). Expression of UCP2 was limited to the extension at 72 8C for 30 s (for preproinsulin I); and mitochondrial compartment in both control (not 40 cycles of melting at 95 8C for 20 s, annealing at 58 8C shown) and UCP2-OE cells (Fig. 1E). for 20 s, and extension at 72 8Cfor30s(for preproinsulin II). The PCR products (expected sizes UCP2 – 127 bp, rat preproinsulin I – 216 bp, rat Glucose-stimulated insulin secretion and ATP preproinsulin II – 326 bp and 18S – 130 bp) were run content on 1% agarose gels to verify that the PCRs generated To confirm the inhibitory effect of UCP2-OE on the expected products. The expression of UCP2 and glucose-stimulated insulin secretion, control or UCP2- preproinsulin mRNAs was normalized to 18S for OE INS-1 cells were incubated in various glucose comparison between treatment groups. concentrations (2.8, 11, or 22 mM) in KRBB (Fig. 2A). In control cells, insulin release increased . ! Statistical analysis significantly between 2 8 and 11 mM glucose (P 0.001) and the effect of glucose was similar in EGFP- All data are presented as meanGS.E.M. Significant expressing cells, which acted as a transfection control differences were determined using two-way ANOVA with (Fig. 2B). In UCP2-OE cells, glucose did not stimulate Bonferroni post-tests unless otherwise stated. Differences any increase in insulin secretion, even at 22 mM glucose were considered significant if P!0.05 or better. (PO0.05). The ATP content of UCP2-OE cells was significantly reduced (P!0.0001) compared with control cells (Fig. 3A). This reduction is probably Results largely cytosolic because secreted ATP (as an indicator of ATP sequestered in granules) did not change after mitochondrial uncoupling (M Collins & C Chan, Quantification of UCP2 mRNA and protein unpublished data). The ATP synthase expression was Figure 1AshowsaO200-fold increment of UCP2 mRNA similar in control and UCP2-OE cells (Fig. 3B and C). expression in UCP2-OE cells relative to controls cultured To promote b-cell exhaustion and thereby force new in RPMI (24.49G6.74 vs 0.12G0.06, P!0.0001) and insulin synthesis and processing, control and UCP2-OE RPMI-K (21.38G2.98 vs 0.017G0.001, P!0.0001). In the INS-1 cells were cultured for 24 h in RPMI www.endocrinology-journals.org Journal of Molecular Endocrinology (2006) 37, 517–526

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access 520 N KASHEMSANT and C B CHAN . UCP2 and proinsulin processing

supplemented with 20 mM KCl (Fig. 2A). RPMI-K culture reduced insulin content by O85% (176.4G 20.9vs22.4G2.4pmol/well, nZ6, P!0.001). In control cells, RPMI-K did not affect basal insulin secretion. Although 11 and 22 mM induced glucose- stimulated insulin release, the effect was blunted compared with cells cultured in RPMI (P!0.001). In UCP2-OE cells, RPMI-K did not alter the pattern of insulin secretion observed in RPMI. The ATP content of UCP2-OE cells cultured in RPMI-K was reduced to a similar extent as cells cultured in RPMI (Fig. 3A). The ATP synthase expression was significantly elevated by threefold after RPMI-K culture (P!0.05) in both control and UCP2-OE cells (Fig. 3B).

Figure 1 Effect of transfecting a UCP2-expressing plasmid into INS-1 cells on UCP2 expression. (A) UCP2 mRNA expression, relative to 18S rRNA expression, in control (open) and UCP2-OE (closed) cells cultured with RPMI or RPMI-K. A significant increase (P!0.0001) was detected in UCP2-OE cells under both C K conditions (nZ10 for all groups). (B) 18S rRNA expression in control and UCP2-OE cells. (C) Representative blot of UCP2 Figure 2 (A) Effects of UCP2-OE on insulin secretion in INS-1 protein expression in UCP2-OE cells. (D) The cells cultured in cells cultured in either RPMI or RPMI-K. Insulin release as a RPMI had significantly increased UCP2 expression 48 h after percentage of total cell insulin content was calculated. (B) Control UCP2 transfection (PZ0.0076 using Student’s t-test, nZ9) (untransfected) or EGFP-transfected (EGFP) INS-1 cells had compared with control cells (nZ10). (E) UCP2 was expressed in similar glucose responsiveness. *P!0.05, †P!0.01, ‡P!0.001 the mitochondrial fraction (M) of INS-1 cells, as indicated by its compared with 2.8 mM glucose concentration in the same enhanced detection in the ATP synthase-enriched component treatment; §P!0.001 compared with control cells at the same compared with the cytosolic fraction (C). glucose concentration (nO25 in each group).

Journal of Molecular Endocrinology (2006) 37, 517–526 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access UCP2 and proinsulin processing . N KASHEMSANT and C B CHAN 521

0.80G0.15, nZ11, PO0.05) when cells were cultured in RPMI or RPMI-K medium for 24 h (Fig. 4D and E). Higher molecular weight proenzymes (80–90 kDa) were also detected and were not significantly affected by UCP2-OE (data not shown). However, comparing RPMI with RPMI-K medium showed that SPC3 was increased in RPMI-K control but not UCP2-OE cells (Fig. 4E, P!0.05).

Effects of oligomycin on cellular ATP, insulin secretion, and proinsulin processing Oligomycin inhibits the ability of ATP synthase to lower cellular ATP content. The concentration and time of exposure to oligomycin of INS-1 cells were titrated to achieve w25% reduction in cellular ATP content (Fig. 5A), similar to that observed in UCP2-OE cells. Oligomycin had no effect on basal insulin secretion but significantly reduced the response to 11 and 22 mM glucose (Fig. 5B). Oligomycin-treated INS-1 cells had 60% higher proinsulin:insulin ratio than control cells but this was not statistically significant (0.057G0.012 vs 0.036G0.007, nZ4, PZ0.19).

Insulin gene transcription and cellular content Figure 3 (A) ATP content (nmol/mg protein) in control and UCP2- In RPMI cultured cells, preproinsulin I gene expression OE cells. UCP2-OE cells had significantly lower ATP content than was significantly higher in UCP2-OE than controls, control cells after culture in both RPMI and RPMI-K (*P!0.05 and †P!0.001, nZ24). (B) Representative immunoblot of ATP while preproinsulin II mRNA expression was not synthase expression. (C) Quantification of ATP synthase significantly different (Fig. 6A and B). Exposure to expression in control and UCP2-OE cells cultured in either RPMI RPMI-K for 24 h significantly suppressed preproinsulin or RPMI-K (*P!0.05, nZ8–9). I and II transcription in UCP2-OE cells only. A trend to reduced transcript expression was seen in control cells Proinsulin processing and SPC3 expression for both preproinsulin I and II. The RIA utilized in these experiments detects both To determine whether proinsulin processing was proinsulin and insulin with similar affinity (unpub- altered in UCP2-OE cells, western blot analysis for lished data). Total insulin species content was increased proinsulin and insulin was performed. In addition to in UCP2-OE cells cultured in RPMI (Fig. 6C, P!0.001). using guinea pig anti-insulin antibody for detecting Total insulin content was not changed in control cells both proinsulin and insulin bands, the proinsulin cultured in RPMI-K but was reduced by 15% in UCP2- bands were differentiated from mature insulin bands OE cells (P!0.01). using goat anti-C-peptide antibody. In these control experiments, rat insulin (which includes insulin I and insulin II), porcine insulin, and bovine proinsulin were used as standards (Fig. 4A). The proinsulin:insulin Discussion ratios were similar in control and UCP2-OE cells cultured in RPMI (0.18G0.02; nZ11 vs 0.22G0.02; Elevated proinsulin secretion from b-cells in diabetic nZ10, PO0.05; Fig. 4B). The cells transfected with human patients (Kahn & Halban 1997, Hanley et al. EGFP had similar proinsulin:insulin ratios as control 2002, Pradhan et al. 2003, Zethelius et al. 2003) and cells (data not shown). This ratio was significantly rodent models (Alarcon et al. 1995, Leibowitz et al. 2001, increased in the UCP2-OE group (0.30G0.02, nZ9, Guest et al. 2002) is well-documented. Recently, the role P!0.01) compared with the control cells (0.22G0.02, of mitochondria in b-cell dysfunction has received nZ6) when cells were cultured in RPMI-K for 24 h attention (Maechler 2003) and we hypothesized that (Fig. 4B and C). the reduction of ATP production would impair insulin The SPC3 (64 kDa) expression in control and UCP2- secretion and disrupt other cellular functions (Chan OE groups was similar (0.68G0.17, nZ10 vs et al. 2004). Here, we show that reducing b-cell ATP, by www.endocrinology-journals.org Journal of Molecular Endocrinology (2006) 37, 517–526

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access 522 N KASHEMSANT and C B CHAN . UCP2 and proinsulin processing

Figure 4 Effect of UCP2-OE on the proinsulin:insulin ratio. (A) Representative control blots showing the distinction between proinsulin and insulin bands as detected by C-peptide antibody (Ab) and insulin antibody. (B) Representative blots of proinsulin and insulin expression from control and UCP2-OE cell lysates. (C) Quantification of proinsulin:insulin ratios from densitometric measurement of proinsulin and insulin immunoblots (nZ9–11, *P!0.001 compared with RPMI). Significant effects of UCP2-OE were detected (†P!0.05). (D) Representative blot of immunostaining for SPC3. (E) Quantification of SPC3 expression from immunoblots (nZ8–11, ‡P!0.05 compared with RPMI).

overexpressing UCP2 by twofold or treating cells with Low cellular ATP could also affect operation of ATP- C oligomycin, is correlated with reduced glucose-stimu- dependent Ca2 pumps. SPC2 and SPC3 are calcium- lated insulin secretion but, unless the b-cells are dependent proteases of the subtilisin family (Halban & C exhausted, there is no significant impact on proinsulin Irminger 1994). When the supply of intracellular Ca2 processing. is reduced, proinsulin processing is impaired (Guest C Proinsulin biosynthesis (Alarcon et al. 2002) and et al. 1997). UCP2-OE reduces glucose-stimulated Ca2 maturation to insulin are energy-dependent processes influx (McQuaid et al. 2006), which may directly impact C (Rhodes & Halban 1987, Slee et al. 1990). However, Ca2 availability in the granule. insulin transcription and total cellular content were Interestingly, in this study, proinsulin processing was increased by UCP2-OE in INS-1 cells, proving that the not altered solely by UCP2-OE and consequent suppression of glucose-stimulated insulin secretion was reduction of ATP. Possibly, cells unable to secrete by not due to inadequate supply. virtue of UCP2-OE effects on KATP and voltage- C Maintenance of a granule microenvironment that dependent Ca2 channel activity (Chan et al. 2001, supports proinsulin conversion to insulin is ATP- McQuaid et al. 2006) maintained a normal proinsuli- dependent (Rhodes & Halban 1987, Slee et al. 1990). n:insulin ratio because of prolonged retention of K ATP stimulates proton and Cl entry into the granule insulin granules despite a sub-optimal granule environ- to promote acidification (Rhodes & Halban 1987, Barg ment. Therefore, a strategy to depolarize INS-1 cells et al. 2001) and the activity of SPC3 is highly and promote b-cell exhaustion was devised. Previously, pH-sensitive (Orci et al. 1994). By reducing cellular insulin secretion from UCP2-OE islets was stimulated by ATP content by w25% with UCP2-OE, proton pump acute KCl application (Chan et al. 1999). INS-1 cells activity was predicted to be impaired, creating sub- were therefore treated for 24 h with a depolarizing optimal conditions for SPC3 activity. concentration of KCl, which reduced the insulin

Journal of Molecular Endocrinology (2006) 37, 517–526 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access UCP2 and proinsulin processing . N KASHEMSANT and C B CHAN 523

Figure 5 Effect of oligomycin on INS-1 cell function. (A) ATP content (nmol/mg protein) in control and oligomycin-treated cells. Oligomycin-treated cells had w25% lower ATP content than control cells (PZ0.063, nZ10). (B) Effects of oligomycin on insulin secretion in INS-1 cells. Insulin release as a percentage of total cell insulin content was calculated. *P!0.05, compared with 2.8 mM glucose (Glc) concentration in the same treatment; †P!0.05 compared with control cells at the same glucose concentration (nO20 in each group). content by nearly 90%. Under this condition, the UCP2-OE cells had an increased proinsulin:insulin Figure 6 Effect of UCP2-OE and RPMI-K on preproinsulin mRNA ratio. The mechanism for this impairment of insulin expression and cellular content of insulin species. (A) Quantitative processing did not involve any further decrease in PCR of preproinsulin I mRNA. *P!0.05, effect of UCP-OE; C cellular ATP. Although K treatment can affect b-cell †P!0.0001, effect of RPMI-K, nZ10 for all groups. (B) Prepro- ‡ ! . Z ATP content in the short term (5–60 min) (Ainscow & insulin II mRNA quantification. P 0 01, effect of RPMI, n 10 for all groups. (C) Effect of UCP2-OE and RPMI-K on total insulin Rutter 2001, Elmi et al. 2001), longer-term exposure species content in INS-1 cells. †P!0.001, effect of UCP2-OE; appears to allow compensations to occur. Moreover, §P!0.01, effect of RPMI-K. SPC3 expression was increased (in control cells) or C unchanged (in UCP2-OE cells) by high K .We speculate that limitations in ATP availability leading to because the electrophoresis was run under non- C altered granule pH and Ca2 environment, coupled reducing conditions. Others have used high per- with the increased secretory demand led to the formance liquid chromatography to separate insulin impaired proinsulin processing. Unlike in cells cul- species (Leahy et al. 1991, Neerman-Arbez et al. 1993) tured in RPMI, granules in RPMI-K-cultured cells were but comparison of different molecular forms by western no longer retained within the b-cell for a sufficient time blot has been an accepted strategy for quantifying for processing in a sub-optimal environment. processing of other hormones, such as islet amyloid We quantified proinsulin processing by identifying polypeptide (Marzban et al. 2004). In future, this proinsulin versus insulin by sequential use of insulin technique could be combined with radiolabeling to and C-peptide antibodies. The peptides could not be detect processing of newly synthesized proinsulin identified simply on the basis of molecular weight relative to total. www.endocrinology-journals.org Journal of Molecular Endocrinology (2006) 37, 517–526

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access 524 N KASHEMSANT and C B CHAN . UCP2 and proinsulin processing

Expression of preproinsulin, SPC3, and SPC2 is A secondary hypothesis was that UCP2-OE would coordinately regulated (Schuppin & Rhodes 1996). impair gene transcription because of the ATP-depen- We did not measure SPC2 expression but predict that dence of chromatin organization and remodeling its expression in UCP2-OE cells would be relatively (Orphanides & Reinberg 2002) and transcription normal. Moreover, SPC2 is quantitatively less important initiation (Woychik & Hampsey 2002). Conversely, the than SPC3 and not rate-limiting for proinsulin proces- 20% reduction in ATP in UCP2-OE cells increased sing (Furuta et al. 1998, Guest et al. 2002). Interestingly, preproinsulin I mRNA and did not alter preproinsulin SPC3 expression in INS-1 cells is reported to be reduced II mRNA expression. The concentration of ATP in by w75% relative to native rat b-cells, leading to slower normal b-cells is w1mM (Erecinska et al. 1992) w proinsulin conversion (Neerman-Arbez et al. 1993). In compared with a Km of 10 mM for RNA polymerase control cells used in this study, the ratio of proinsulin to II, the initiator of transcription, and 200 mM for helicase insulin was w0.18. Reports from normal mouse (Jiang & Gralla 1995). Thus, the reduction in cellular pancreas give a lower ratio of 0.06 (Furuta et al. 1998) ATP content was unlikely to markedly affect gene suggesting some differences between cell lines and transcription in INS-1 cells. However, the possibility native islets. remains that UCP2-OE could affect transcription of Changes in proinsulin processing after RPMI-K certain genes by other means. culture might also be attributed to non-secretory effects In several models of diabetes, UCP2 expression is C of K on INS-1 cells. Exposing INS-1 cells to RPMI-K up-regulated (Kassis et al. 2000, Zhang et al. 2001, strongly suppressed preproinsulin I and II transcription Laybutt et al. 2002, Joseph et al. 2004) and impairment by 50–70% and was more pronounced in UCP2-OE of proinsulin processing is reported (Leahy et al. 1991). cells. Likewise, total insulin species content of INS-1 UCP2 may also be induced in b-cells of some type 2 cells was significantly reduced only in UCP2-OE cells, diabetic patients (Sasahara et al. 2004). In an environ- although the decrement was smaller (w25%) than for ment of hyperlipidemia and hyperglycemia, increased preproinsulin mRNA, suggesting a reduced supply of secretory demand is placed on the b-cell. The present newly synthesized insulin. Although this would contrib- study raises the possibility that one consequence of ute to lack of response to glucose over time, it is unclear UCP2 induction, secondary to reduced ATP pro- that reduced biosynthesis would impact upon proces- duction, is decreased proinsulin processing, a con- sing. However, the expression of other genes involved tributing factor to glucose intolerance. in insulin biosynthesis and processing might also have C been altered by RPMI-K culture. Interestingly, K - dependent gene expression has been described in Acknowledgements plants (Armengaud et al. 2004), but it is apparently not C known if K directly regulates gene transcription in The technical assistance of Monique Saleh and Mark C mammals. Alternatively, prolonged K exposure and Collins is greatly appreciated. consequent b-cell depolarization increase intracellular C C Ca2 . INS-1 cells exposed to 30 mM K for only 6 h had a sevenfold increase in insulin reporter gene activity, Funding C which was partially blocked by a Ca2 channel antagonist (Lawrence et al. 2001). This difference This research was supported by grants to C B C from the Canadian Diabetes Association in honor of the late Neil from the present findings may reflect a time-dependent C M Miller and from the Canadian Institutes of Health effect of K exposure. The transcription factor Pdx-1 Research MOP 43978. The authors declare no conflict was shown recently to influence proinsulin processing of interest. independent of glucose metabolism (Wang et al. 2005). Moreover, Pdx-1 expression is sensitive to changes in 2C Ca /calmodulin-dependent protein kinase II D2 References expression, the activity of which reflects intracellular 2C C Ca fluxes (Osterhoff et al. 2003). Culturing rat islets Ainscow EK & Rutter GA 2001 Mitochondrial priming modifies Ca2 in 20 mM glucose for 72 h evoked a persistent increase oscillations and insulin secretion in pancreatic islets. Biochemical C in intracellular Ca2 and an increase in insulin Journal 353 175–180. Alarcon C, Leahy JL, Schuppin GT & Rhodes CJ 1995 Increased production (Xu et al. 2000). Chronically increased secretory demand rather than a defect in the proinsulin conversion demand for insulin requires glucose-regulated mechanism causes hyperproinsulinemia in a glucose-infusion rat expression of the proinsulin gene (Leibowitz et al. model of non-insulin-dependent diabetes mellitus. Journal of 2002). Of note, RPMI-K culture decreased insulin Clinical Investigation 95 1032–1039. Alarcon C, Wicksteed B, Prentki M, Corkey BE & Rhodes CJ 2002 transcript abundance but increased expression of ATP Succinate is a preferential metabolic stimulus-coupling signal for synthase demonstrating differential effects on genes glucose-induced proinsulin biosynthesis translation. Diabetes 51 important to insulin secretion. 2496–2504.

Journal of Molecular Endocrinology (2006) 37, 517–526 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access UCP2 and proinsulin processing . N KASHEMSANT and C B CHAN 525

Armengaud P, Breitling R & Amtmann A 2004 The potassium- Jiang Y & Gralla JD 1995 Nucleotide requirements for activated RNA dependent transcriptome of Arabidopsis reveals a prominent role of polymerase II open complex formation in vitro. Journal of Biological jasmonic acid in nutrient signaling. Plant Physiology 136 2556–2576. Chemistry 270 1277–1281. Asfari M, Janjic D, Meda P, Li G, Halban PA & Wollheim CB 1992 Joseph JW, Koshkin V, Zhang CY, Wang J, Lowell BB, Chan CB & Establishment of 2-mercaptoethanol-dependent differentiated Wheeler MB 2002 Uncoupling protein 2 knockout mice have insulin-secreting cell lines. Endocrinology 130 167–178. enhanced insulin secretory capacity after a high-fat diet. Diabetes 51 Barg S, Ma X, Eliasson L, Galvanovskis J, Gopel SO, Obermuller S, 3211–3219. Platzer J, Renstrom E, Trus M, Atlas D, Striessnig J & Rorsman P Joseph JW, Koshkin V, Saleh MC, Sivitz WI, Zhang CY, Lowell BB, 2C 2001 Fast exocytosis with few Ca channels in insulin-secreting Chan CB & Wheeler MB 2004 Free fatty acid-induced beta-cell mouse pancreatic B cells. Biophysical Journal 81 3308–3323. defects are dependent on uncoupling protein 2 expression. Chan CB, MacDonald PE, Saleh MC, Johns DC, Marban E & Wheeler MB Journal of Biological Chemistry 279 51049–51056. 1999 Overexpression of uncoupling protein 2 inhibits glucose- Kahn SE & Halban PA 1997 Release of incompletely processed stimulated insulin secretion from rat islets. Diabetes 48 1482–1486. proinsulin is the cause of the disproportionate proinsulinemia of Chan CB, De Leo D, Joseph JW, McQuaid TS, Ha XF, Xu F, Tsushima NIDDM. Diabetes 46 1725–1732. RG, Pennefather PS, Salapatek AM & Wheeler MB 2001 Increased Kassis N, Bernard C, Pusterla A, Casteilla L, Penicaud L, Richard D, uncoupling protein-2 levels in beta-cells are associated with Ricquier D & Ktorza A 2000 Correlation between pancreatic islet impaired glucose-stimulated insulin secretion: mechanism of uncoupling protein-2 (UCP2) mRNA concentration and insulin action. Diabetes 50 1302–1310. status in rats. International Journal of Experimental Diabetes Research 1 Chan CB, Saleh MC, Koshkin V & Wheeler MB 2004 Uncoupling 185–193. protein 2 and islet function. Diabetes 53 S136–S142. Klingenberg M 1999 Uncoupling protein – a useful energy dissipator. Davidson HW, Rhodes CJ & Hutton JC 1988 Intraorganellar calcium Journal of Bioenergetics and Biomembranes 31 419–430. and pH control proinsulin cleavage in the pancreatic beta cell via Lawrence MC, Bhatt HS, Watterson JM & Easom RA 2001 Regulation C two distinct endopeptidases. Nature 333 93–96. of insulin gene transcription by a Ca2 -responsive pathway Elmi A, Idahl L-A & Sehlin J 2001 Modulation of islet ATP content by C C involving calcineurin and nuclear factor of activated T cells. inhibition or stimulation of the Na /K pump. European Journal of Molecular Endocrinology 15 1758–1767. Pharmacology 426 139–143. Laybutt DR, Kaneto H, Hasenkamp W, Grey S, Jonas J-C, Sgroi DC, Erecinska M, Bryla J, Michalik M, Meglasson M & Nelson D 1992 Groff A, Ferran C, Bonner-Weir S, Sharma A et al. 2002 Increased Energy metabolism in islets of Langerhans. Biochimica et Biophysica expression of antioxidant and antiapoptotic genes in islets that may Acta 1101 273–295. contribute to b-cell survival during chronic hyperglycemia. Diabetes Fleury C, Neverova M, Collins S, Raimbault S, Champigny O, Levi- 51 413–423. Meyrueis C, Bouillaud F, Seldin MF, Surwit RS, Ricquier D et al. 1997 Leahy JL, Halban PA & Weir GC 1991 Relative hypersecretion of Uncoupling protein-2: a novel gene linked to obesity and proinsulin in rat model of NIDDM. Diabetes 40 985–989. hyperinsulinemia. Nature Genetics 15 269–272. Leibowitz G, Yuli M, Donath MY, Nesher R, Melloul D, Cerasi E, Gross Furuta M, Carroll R, Martin S, Swift HH, Ravazzola M, Orci L & Steiner DF DJ & Kaiser N 2001 b-Cell glucotoxicity in the Psammomys obesus 1998 Incomplete processing of proinsulin to insulin accompanied by model of type 2 diabetes. Diabetes 50 (Suppl 1) S113–S117. elevation of des-31,32 proinsulin intermediates in islets of mice Leibowitz G, Uckaya G, Oprescu AI, Cerasi E, Gross DJ & Kaiser N 2002 lacking active PC2. Journal of Biological Chemistry 273 3431–3437. Glucose-regulated proinsulin gene expression is required for Garlid KD, Jaburek M & Jezek P 1998 The mechanism of proton adequate insulin production during chronic glucose exposure. transport mediated by mitochondrial uncoupling proteins. FEBS Endocrinology 143 3214–3220. Letters 438 10–14. McQuaid TS, Saleh MC, Joseph JW, Gyulkhandanyan A, Manning-Fox JE, Gimeno RE, Dembski M, Weng X, Deng N, Shyjan AW, Gimeno CJ, Iris F, Ellis SJ, Woolf EA & Tartaglia LA 1997 Cloning and characterization of MacLellan JD, Wheeler MB & Chan CB 2006 cAMP-mediated an uncoupling protein homolog: a potential molecular mediator of signaling normalizes glucose-stimulated insulin secretion in uncou- b human thermogenesis. Diabetes 46 900–906. pling protein-2 overexpressing -cells. Journal of Endocrinology 190 Goodge KA & Hutton JC 2000 Translational regulation of proinsulin 669–680. biosynthesis and proinsulin conversion in the pancreatic b-cell. Maechler P 2003 Novel regulation of insulin secretion: the role of Seminars in Cell and Developmental Biology 11 235. mitochondria. Current Opinion in Investigational Drugs 4 1166–1172. C Guest PC, Bailyes EM & Hutton JC 1997 Endoplasmic reticulum Ca2 Marzban L, Trigo-Gonzalez G, Zhu X, Rhodes CJ, Halban PA, Steiner is important for the proteolytic processing and intracellular DF & Verchere CB 2004 Role of beta-cell prohormone convertase transport of proinsulin in the pancreatic beta-cell. Biochemical (PC)1/3 in processing of pro-islet amyloid polypeptide. Diabetes 53 Journal 323 445–450. 141–148. Guest PC, Abdel-Halim SM, Gross DJ, Clark A, Poitout V, Amaria R, Mykkanen L, Zaccaro DJ, Hales CN, Festa A & Haffner SM 1999 Ostenson CG & Hutton JC 2002 Proinsulin processing in the The relation of proinsulin and insulin to insulin sensitivity and acute diabetic Goto-Kakizaki rat. Journal of Endocrinology 175 637–647. insulin response in subjects with newly diagnosed Type II diabetes: Halban P & Irminger JC 1994 Sorting and processing of secretory the Insulin Resistance Atherosclerosis Study. Diabetologia 42 1060. proteins. Biochemical Journal 299 1–18. Neerman-Arbez M, Sizonenko SV & Halban PA 1993 Slow cleavage at Hanley AJG, D’Agostino R Jr, Wagenknecht LE, Saad MF, Savage PJ, the proinsulin B-chain/connecting peptide junction associated Bergman R & Haffner SM 2002 Increased proinsulin levels and with low levels of endoprotease PC1/3 in transformed beta cells. decreased acute insulin response independently predict the Journal of Biological Chemistry 268 16098–16100. incidence of type 2 diabetes in the Insulin Resistance Athero- Orci L, Halban P, Perrelet A, Amherdt M, Ravazzola M & Anderson RG sclerosis Study. Diabetes 51 1263–1270. 1994 pH-independent and -dependent cleavage of proinsulin in the Hong Y, Fink BD, Dillon JS & Sivitz WI 2001 Effects of adenoviral same secretory vesicle. Journal of Cell Biology 126 1149–1156. overexpression of uncoupling protein-2 and -3 on mitochondrial Orphanides G & Reinberg D 2002 A unified theory of gene expression. respiration in insulinoma cells. Endocrinology 142 249–256. Cell 108 439–451. Irminger JC, Meyer K & Halban P 1996 Proinsulin processing in Osterhoff M, Mohlig M, Schwanstecher M, Seufert J, Ortmann J, C the rat insulinoma cell line INS after overexpression of the Schatz H & Pfeiffer AF 2003 Ca2 /calmodulin-dependent protein endoproteases PC2 or PC3 by recombinant adenovirus. kinase II delta2 regulates gene expression of insulin in INS-1 rat Biochemical Journal 320 11–15. insulinoma cells. Cell Calcium 33 175–184. www.endocrinology-journals.org Journal of Molecular Endocrinology (2006) 37, 517–526

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access 526 N KASHEMSANT and C B CHAN . UCP2 and proinsulin processing

Pradhan AD, Manson JE, Meigs JB, Rifai N, Buring JE, Liu S & Steiner DF 1998 The proprotein convertases. Current Opinions in Ridker PM 2003 Insulin, proinsulin, proinsulin:insulin ratio, Chemical Biology 2 31–39. and the risk of developing type 2 diabetes mellitus in women. Tordjman K, Standley KN, Bernal-Mizrachi C, Leone TC, Coleman T, American Journal of Medicine 114 438. Kelly DP & Semenkovich CF 2002 PPARa suppresses insulin Proudnikov D, Yuferov V, LaForge KS, Ho A & Jeanne Kreek M 2003 secretion and induces UCP2 in insulinoma cells. Journal of Lipid Quantification of multiple mRNA levels in rat brain regions using Research 43 936–943. real time optical PCR. Molecular Brain Research 112 182. Toriumi C & Imai K 2003 Altered expression of I and II and Ravnskjaer K, Boergesen M, Rubi B, Larsen JK, Nielsen T, Fridriksson J, their mRNAs in the islets of Langerhans in dexamethasone-induced Maechler P & Mandrup S 2005 Peroxisome proliferator-activated diabetic rats. Biomedical Chromatography 17 26–32. receptor a (PPARa) potentiates, whereas PPARg attenuates, Wang H, Iezzi M, Theander S, Antinozzi P, Gauthier B, Halban P & glucose-stimulated insulin secretion in pancreatic beta-cells. Wollheim C 2005 Suppression of Pdx-1 perturbs proinsulin Endocrinology 146 3266–3276. processing, insulin secretion and GLP-1 signalling in INS-1 cells. Rhodes CJ & Halban PA 1987 Newly synthesized proinsulin/insulin Diabetologia 48 720–731. and stored insulin are released from pancreatic B cells predomi- Woychik NA & Hampsey M 2002 The RNA polymerase II machinery: nantly via a regulated, rather than a constitutive, pathway. Journal of structure illuminates function. Cell 108 453–463. Cell Biology 105 145–153. Xu GG, Gao Z-Y, Borge PD, Jegier PA, Young RA & Wolf BA 2000 Rhodes CJ, Lucas CA, Mutkoski RL, Orci L & Halban PA 1987 Insulin regulation of b-cell function involves a feedback loop on Stimulation by ATP of proinsulin to insulin conversion in 2C isolated rat pancreatic islet secretory granules. Association with SERCA gene expression, Ca homeostasis, and insulin expression the ATP-dependent proton pump. Journal of Biological Chemistry and secretion. Biochemistry 39 14912–14919. 262 10712–10717. Zethelius B, Byberg L, Hales CN, Lithell H & Berne C 2003 Proinsulin Sasahara M, Nishi M, Kawashima H, Ueda K, Sakagashira S, Furuta H, and acute insulin response independently predict Type 2 diabetes Matsumoto E, Hanabusa T, Sasaki H & Nanjo K 2004 Uncoupling mellitus in men – report from 27 years of follow-up study. protein 2 promoter polymorphism-866G/A affects its expression in Diabetologia 46 20–26. beta-cells and modulates clinical profiles of Japanese type 2 diabetic Zhang CY, Baffy G, Perret P, Krauss S, Peroni O, Grujic D, Hagen patients. Diabetes 53 482–485. T, Vidal-Puig AJ, Boss O, Kim YB et al. 2001 Uncoupling Schrauwen P & Hesselink M 2002 UCP2 and UCP3 in muscle controlling protein-2 negatively regulates insulin secretion and is a major body metabolism. Journal of Experimental Biology 205 2275–2285. link between obesity, beta cell dysfunction, and type 2 diabetes. Schuppin GT & Rhodes CJ 1996 Specific co-ordinated regulation of Cell 105 745–755. PC3 and PC2 gene expression with that of preproinsulin in insulin- producing beta TC3 cells. Biochemical Journal 313 259–268. Slee DJ, Jones PM & Howell SL 1990 Proinsulin processing in electrically permeabilized rat islets of Langerhans. Journal of Received 17 July 2006 Molecular Endocrinology 5 275–280. Accepted 11 September 2006

Journal of Molecular Endocrinology (2006) 37, 517–526 www.endocrinology-journals.org

Downloaded from Bioscientifica.com at 09/24/2021 06:06:18AM via free access