Identication and subcellular localization of the NaC/HC exchanger and a novel related protein in the endocrine pancreas and adrenal medulla
Pierre Moulin, Yves Guiot, Jean-Christophe Jonas2, Jacques Rahier, Olivier Devuyst1 and Jean-Claude Henquin2
Units of Pathology, 1Nephrology, 2Endocrinology and Metabolism, Faculty of Medicine, Universite Catholique de Louvain, Avenue Hippocrate, B-1200 Brussels, Belgium (Requests for offprints should be addressed to J C Henquin; Email: henquin@endo.ucl.ac.be)

Abstract

NaC/HC exchangers (NHE) constitute a family of membrane antiporters that contribute to the regulation of cellular pH and volume in many tissues, including pancreatic islets. We investigated the molecular identity of NHE in rodent and human endocrine pancreas, and determined its cellular and subcellular localization. NHE1 was the most abundantly expressed isoform in rat islets, and was also expressed in mouse and human islets. By western blot, an antiserum raised against the C-terminus end of NHE1 conrmed the presence of a w100 kDa protein corresponding to NHE1 in islets and unexpectedly unveiled the existence of a w65 kDa cross-reactive NHE1-related protein. By immunohistochemistry, the antiserum labelled the membranes of pancreatic acini and ducts, but also diffusely stained the cytoplasm of insulin, glucagon and somatostatin cells as well as endocrine cells of the adrenal medulla. Electron microscopy localized the NHE1 immunoreactivity in the membrane of secretory granules, an unexpected nding supported by a decrease in immunohistochemical signal in degranulated b-cells. Islets of Slc9A1swe/swe mice, which lack full NHE1 protein, were found to express an mRNA corresponding to the end of NHE1 as well as the w65 kDa protein. They still showed the cytoplasmic labelling but no plasma membrane was stained. We conclude that both the full-length and the shorter-splice variant of NHE1 are expressed in all cell types of the endocrine pancreas and in the adrenal medulla of rodents and humans. The complete protein is addressed to the plasma membrane and the shorter one to the membrane of secretory granules where its function remains to be established. Journal of Molecular Endocrinology (2007) 38, 409422

Introduction

Pancreatic b cells adjust insulin secretion to the ambient concentration of glucose and other nutrients through changes in their metabolism (Newgard 2002, MacDonald et al. 2005a, Matschinsky et al. 2006). Oxidative glycolysis increases the ATP:ADP ratio, which closes ATP-sensitive KC channels in the plasma membrane, thereby causing Ca2C inux through voltage-dependent Ca2C channels and rise in the concentration of cytosolic Ca2C (Seino et al. 2000, Gilon et al. 2002, MacDonald et al. 2005b). This rise triggers exocytosis of insulin-containing granules. Simultaneously, but independently of its action on ATP-sensitive KC channels, the metabolism of glucose produces amplifying signals that augment secretion without further increasing Ca2C (Henquin 2000, Aizawa et al. 2002, Straub & Sharp 2002). During stimulation by nutrients, the b cell cytosolic pH (pHi; Lindstrom & Sehlin 1984, Best et al. 1988, Juntti-Berggren et al. 1991, Shepherd & Henquin 1995, Salgado et al. 1996, Shepherd et al. 1996) and volume
Journal of Molecular Endocrinology (2007) 38, 409422 09525041/07/038409 q 2007 Society for Endocrinology Printed in Great Britain
(Miley et al. 1997) increase. The functional signicance of these changes is still debated (Pace et al. 1983, Lindstrom & Sehlin 1986, Bertrand et al. 2002, Gunawardana & Sharp 2002) and the underlying mechanisms are incompletely elucidated. However, experiments using ionic substitutions in the extracellular medium or pharmacological tools (e.g. dimethyl-amiloride to block NaC/HC countertransport) have established that, besides HCO3K/ClK exchangers, a NaC/HC exchanger is implicated in the regulation of b cell pHi (Juntti-Berggren et al. 1991, Shepherd & Henquin 1995, Shepherd et al. 1996) and possibly volume (Miley et al. 1998). Similarly, NaC/HC exchange has been implicated in the control of pHi and volume of adrenal chromafn cells (Delpire et al. 1988, Kao et al. 1991). Sodiumproton exchangers (NHE) are widely distributed integral membrane proteins that regulate cellular volume and pH (Orlowski & Grinstein 1997, Ritter et al. 2001). The SLC9 family comprises many pseudogenes and genes that encode at least nine isoforms of the NHE proteins (Orlowski & Grinstein 2004, Nakamura et al. 2005). Several SLC9 genes are

DOI: 10.1677/jme.1.02164 Online version via http://www.endocrinology-journals.org

P MOULIN

and others. NaC/HC exchanger in the endocrine pancreas
also known to give rise to multiple transcripts or partial mRNA (Orlowski & Grinstein 2004). The rst ve isoforms (NHE1 to NHE5) are well characterized and display distinct physiological and pharmacological properties. NHE1 is ubiquitous. In polarized cells, it is usually inserted in the basolateral domain of the plasma membrane where it fulls housekeeping regulation of cell volume and pH (Orlowski & Grinstein 2004). NHE2 and NHE3 are mainly found at the apical pole of epithelial cells in kidney (Chambrey et al. 1998), intestine (Chu et al. 2002) and duct cells of salivary glands and pancreas (Lee et al. 1998, 2000), where they play a role in NaC and uid absorption, and secretion of protons (Orlowski & Grinstein 2004). NHE4 has been identied in the macula densa of the kidney (Peti-Peterdi et al. 2000) and in the stomach (Rossmann et al. 2001). NHE5 has mainly been found in the brain where it seems to behave like NHE3 (Baird et al. 1999). Although NHE1 to NHE5 are usually localized in the plasma membrane, NHE3 and NHE5 have also been observed in recycling vesicles (Kurashima et al. 1998, Szaszi et al. 2002). In contrast, the ubiquitously distributed NHE6 to NHE9 have not been localized to the plasma membrane but to intracellular organelles (endosomes, trans-Golgi network or mitochondria; Nakamura et al. 2005). The primary aims of the present study were to determine the molecular identity of NHE in human and rodent islets and to establish its subcellular localization. The initial results unexpectedly led to the discovery of a novel protein related to NHE1 and heavily concentrated in the membrane of secretory granules in all cell types of the endocrine pancreas, and also in the adrenal medulla.
of glibenclamide (5.0 mg/kg body weight) or saline at 12 h interval. Animals were fed ad libitum and decapitated 4 h after the second injection. C57BL/6.SJL, C/swe (slow wave epilepsy) mice (Cox et al. 1997) were purchased from the Jackson Laboratory (Bar Harbor, MI, USA). These heterozygous mice were mated and the resulting homozygous NHE1 null mutant (Slc9A1swe/swe) and wild-type (Slc9A1C/C) were studied at 57 weeks of age. Homozygous mutant mice exhibited a neurological phenotype including ataxia in the hind limbs, and seizures. Genotyping was performed as described (Cox et al. 1997) to conrm the phenotype. Rat islets were obtained by collagenase digestion of the pancreas (Jonas et al. 1998), and were used freshly or after 7 days of culture in RPMI 1640 medium containing 10 mM glucose and 0.5 g/100 ml BSA. A similar procedure was used to obtain mouse islets except that they were cultured for 1842 h in RPMI 1640 medium containing 10% heat-inactivated FCS, 100 IU/ml penicillin, 100 mg/ml streptomycin and 10 mm glucose.

Immunoblot analysis

Parafn-embedded specimens were cut into 3 mm thick sections and processed as described elsewhere (Sempoux et al. 1998) including, when necessary, an antigen retrieval treatment. Primary antibodies were diluted in Tris (pH 7.4) supplemented with 1% BSA and applied overnight at 4 8C. All subsequent incubations lasted 1 h at room temperature. For double immunouorescence experiments, anti-hormone and NHE1 antisera incubations were carried out sequentially. When necessary, a tyramine amplication step was added (Sempoux et al. 2003). The peroxidase activity was revealed by 3,-diaminobenzidine hypochloride (DAB: 50 mg/100 ml, pH 7.4; Fluka Chemie, Buchs, Switzerland) for 10 min. Antibodies and detailed conditions are described in Table 1. Specic optical density of the immunohistochemical signal was measured as described previously (Rahier et al. 1989).
Electron microscopy and immunogold labelling
SDS-PAGE and immunoblotting were performed as described (Combet et al. 1999). The extracts were solubilized by heating at 95 8C for 3 min in sample buffer. Proteins (1040 mg/lane) were separated by electrophoresis through 7.5% acrylamide slabs and transferred to nitrocellulose. Membranes were blocked for 30 min at room temperature in blotting buffer, followed by incubation with the primary antibody (antiNHE1 at 1/2000). The membranes were then washed and incubated for 1 h at room temperature with antirabbit Fab peroxidase-labelled antibody (Dako). After washing, immunoblots were visualized with ECL-Plus reagent (Amersham).
After 24-h xation in paraformaldehyde, small pancreas blocks were cryoprotected in PBS containing 15% sucrose for 48 h, before being frozen into liquid nitrogen. Forty micrometre thick cryosections were cut and incubated with NHE1 antiserum (1/50) for 45 min at 4 8C. After rinsing in PBS, NG-Ig (1/40) was applied for 1 h. After short xation in 2.5% glutaraldehyde, the signal was amplied by a silver enhancer solution according to the manufacturers instructions. The sections were rinsed and embedded into Epon 812 and processed for electron microscopy (Rahier et al. 1989).
Radioactive RT-PCR analysis of rat NHE isoforms mRNA
The primers are indicated in Table 2. Radioactive PCR was performed as described previously (Jonas et al. 1999), with a thermal cycle prole consisting of a

Table 1 Immunohistochemical staining Antibody; dilution Antigen retrieval Yes Yes No No No No No Amplication system Detection system
Polyclonal rabbit anti-NHE1; 1/2000 Polyclonal rabbit anti-NHE1; 1/1000 Monoclonal mouse anti-insulin; 1/1000 Polyclonal rabbit anti-NHE1; 1/50 Monoclonal mouse anti-insulin; 1/80 Monoclonal mouse anti-glucagon; 1/80 Monoclonal mouse anti-somatostatin; 1/80
EV EV 2B-SP 2B-SP-BT-STR 2F 2F 2F
DAB DAB DAB Fluorescence Fluorescence Fluorescence Fluorescence
EV, En-Vision; 2B, biotinylated secondary antibody (1/500); 2F, secondary antibody FITC conjugate (1/20); BT, biotinylated tyramine; SP, streptavidinperoxydase complex (1/500); STR, streptavidin-Texas Red conjugate (1/50).
www.endocrinology-journals.org Journal of Molecular Endocrinology (2007) 38, 409422
and others. NaC/HC exchanger in the endocrine pancreas Reference or GenBank Accession no.
CAG TGG GTC TGA GCC TAT GC CGG TTT AAG CTG TTG TCC TTC CTA GAA AGT CGC TTG ATT CCC TGT ATC TTC CGC AAA TAT CTG TGG TCA AAC GCC ACT GCG TAC TGT GTC AAT C ATA GGC CAG TGG GTC TGA GC GGC TCC TTG CTC CGA ATC ATG GGC TCC TTG CTC CGA ATC ATG TCA TGC CCT GCA CAA AGA CG TCA TGC CCT GCA CAA AGA CG
NM_012652 NM_012653 M85300 M85301 NM138858 M81768 Bell et al. 1999; NM_012652 NM_012652 Cox et al. 1997 Cox et al. 1997
10 min denaturing step at 95 8C followed by 30 cycles of amplication (1 min steps at 94, 60 and 72 8C each) and a nal extension step of 10 min at 72 8C. TATAboxbinding protein (TBP) was used as control gene and amplied by a 24 cycles PCR (Jonas et al. 1999). The amplimers were then separated on a 6% polyacrylamide gel in Tris borate EDTA buffer, in parallel with a 100-bp DNA ladder. The gel was dried, and the amount of [a-32P]dCTP incorporated in each amplicon was quantied with a Cyclone Storage Phosphor System (Packard, Meriden, CT, USA). The ratio of specic product/control gene was then calculated for each sample.
Amplication of NHE1 species-specic cDNA and human NHE1 probe production for in situ hybridization
Table 2 Sequences of oligonucleotide primers used for PCR analysis of NHE isoforms
Gene Rat NHE1 Rat NHE2 Rat NHE3 Rat NHE4 Rat NHE5 Human NHE1 Mouse NHE1: ex2P4 Mouse NHE1: P1P4 Slc9A1C/C: MR0975MR0977 Slc9A1swe/swe: MR0976MR0977
The primers used for human and rodent NHE isoforms are described in Table 2. PCRs were performed in a thermocycler 2400 (Applied Biosystems, Foster City, CA, USA) with a total volume of 25 ml mixture containing GeneAmp PCR buffer, 1.5 mM MgCl2, 200 mM deoxyNTP, 0.5 mM primers and 1 U Taq Gold polymerase (Applied Biosystems). The thermal cycle prole was 10 min denaturation at 94 8C followed by 35 cycles (30 s at 94 8C, 45 s at 62 8C and 1 min at 72 8C) and a nal extension of 10 min at 72 8C. Abelson protooncogene or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were taken as controls. The amplied DNA samples were electrophoresed on ethidium bromide agarose gel and quantied by GelDoc 2000 scanning device (Bio-Rad). The identity of NHE1 PCR product was conrmed by DNA sequence analysis using the dye terminator sequencing system on a Genetic analyser 3600 (Applied Biosystems). The PCR procedure was repeated with a no-reverse transcription control to exclude genomic DNA contamination and carry-over. The probe production procedure for in situ hybridization was similar to that described previously for insulin-like growth factor-II including negative and sequence controls (Sempoux et al. 2003) except that a human NHE1 hybridization probe was produced from human heart using specic primers (Table 2).

-Antisense primer-3 0

Size (bp) -Sense primer-3 0

200 200

GAA CAT CCA CCC CAA GTC TG AAA CCA ACC CAA GTC TAG CAT TGT AGA GCT TCA CAT CCG TCT TAT GG CAG CGT GTT TAC CCT CTT CTA TGT GCG GTC AGC CTA TCG TAT CC ACC ACC ACT GGA AGG ACA AG ACG TCT TCT TCC TCT TCC TGC TG AGG ACA TCT GTG GTC ATT ATG GC CCT GAC CTG GTT CAT CAA CA CCT GAC CTG GTT CAT CAA CT

In silico analyses

Electronic database searches for mouse Slc9A1 gene or mRNA matching sequences, structure and theoretical alternative splicing or start sites were available at http:// www.ensembl.org/Mus_musculus/geneview? geneZ ENSMUSG00000028854&dbZcore (Stalker et al. 2004). RT-PCR primers were tested using blastn against all
nucleotide databases available from the National Center for Biotechnology Information (NCBI) at http://www. ncbi.nlm.nih.gov/BLAST/ (Altschul et al. 1990). Alternate splicing sites were searched with online tool available at http://www.fruity.org/cgi-bin/seq_tools/splice.pl.

Results

Identication of NHE isoforms expressed in the endocrine pancreas
Different isoforms of NHE (NHE1NHE5) were searched by semi-quantitative radioactive RT-PCR analysis of isolated rat islets. We found a major expression of NHE1 (Fig. 1A), whereas NHE2 was expressed in lower amounts. No signal was seen for NHE3 and NHE4, and only a weak signal was obtained for NHE5. Similar results were obtained with several preparations of both fresh and cultured islets. Control experiments carried out with a no-reverse transcription control did not yield any signal.
Since NHE1 was the predominant isoform in rat islets, we studied its expression in the human pancreas. RT-PCR showed expression of NHE1 mRNA in heart and kidney (Fig. 1B, lanes 1 and 2) and in extracts of total pancreas from three subjects (lanes 35). The PCR products obtained in pancreatic samples were sequenced and shown to correspond to human NHE1. By western blot, a polyclonal antiserum raised against the C-terminus part of NHE1 (CtNHE1) detected a major protein band at w100 kDa in pancreas membrane extracts (Fig. 1C) from three different subjects, with a similar pattern as in kidney extracts (positive control). The band pattern was also similar in rat and mouse total pancreas. However, in isolated human, rat or mouse islets, the w100 kDa band was weaker and the predominant signal corresponded to a lower molecular weight (low MW, w65 kDa). This low MW band was also visible in total pancreas extracts at least from rat and mouse, but with a low intensity (Fig. 1C). This inverse pattern is compatible with a greater abundance of the w100 kDa protein in the exocrine pancreas (98% of
Figure 1 Expression of NHE isoforms in the pancreas and control tissues. (A) Radioactive RT-PCR analysis of NHE isoforms and TBP after 30-cycles amplication. RT-, non-reverse transcript negative control; Islets, cDNA library from isolated rat islets (representative of three experiments). Kidney and brain, cDNA libraries taken as positive controls. (B) RT-PCR for NHE1 on human tissues including heart, kidney and three different pancreas extracts. (C) Western blot analysis (20 mg/lane) of human and mouse pancreas extracts. Preimmune, membrane fraction of human pancreas incubated with preimmune serum. Cytosol, cytoplasmic fraction of the same human pancreas incubated with immune serum. Human kidney, membrane extracts from human kidney cortex as positive control for the CtNHE1 antiserum. Human pancreas, membrane extracts from human pancreas incubated with the CtNHE1 antiserum (representative of three different subjects). Human islets, membrane extracts from one human islet preparation. Rat, rat pancreas and isolated islets extracts respectively. Mouse, mouse pancreas and isolated islets extracts respectively. Low MW band (approximately 65 kDa, arrow) is present with a low intensity in total pancreas extract and high intensity in islet extract.

endocrine cells, no specic labelling could be observed in exocrine cells (not shown). Since insulin granules showed a major immunoreactivity against CtNHE1, we investigated the impact of a strong stimulation of insulin secretion on the signal distribution in b cells. We compared the pancreas of control rats with that of rats treated with high doses of glibenclamide. The immunoreactivity of insulin in islets from test animals was markedly decreased (by 56G8%, S.D., Fig. 6A and B), reecting b-cell degranulation (Rahier et al. 1989). As compared with controls, the cytoplasmic CtNHE1 staining was fainter in degranulated islets but the plasma membrane labelling was stronger so that the overall signal intensity was only slightly reduced (by 20G9%, S.D.; Fig. 6C and D).
Identication of a novel NHE1-like protein
In control Slc9A1C/C mice, as in humans and control rats, the islets were diffusely stained by CtNHE1 antiserum, whereas only membrane staining was observed in acinar cells (Fig. 7A). In Slc9A1swe/swe mice that lack functional NHE1 protein, a diffuse labelling unexpectedly persisted in islets, while exocrine cells were negative (Fig. 7B).
Figure 5 Ultrastructural localization of CtNHE1 immunohistochemical signal. (A) Portion of b cell cytoplasm (80 000!). The signal is restricted to the peripheral zone of the insulin granules. (B) Immunogold particles distribution between endocrine granules and other sub-cellular structures. Values represent the mean number (GS.E.) of gold particles by mm2, measured on 27 micrographs of islets from 3 animals. The signal is ve- to tenfold higher over endocrine granules than in the rest of the cytoplasm. Gold particles are twice more abundant around insulin than glucagon granules. (C and D) Distribution of distances between gold particles and centre of insulin (C: nZ347) and glucagon (D: nZ200) granules. Values correspond to the proportion of gold particles found at a given distance interval from the centre of the closest insulin or glucagon granule. The electron micrograph below each histogram shows two secretory granules at a magnication matching the geometric scale of the X-axis in the histogram. The origin of this axis (zero value) is projected on the centre of the granule on the left.
By contrast, hepatocytes showed no cytoplasmic staining and the selective labelling of their plasma membranes in Slc9A1C/C mice (Fig. 7C) was completely absent in Slc9A1swe/swe mice (Fig. 7D). Adrenal medulla from Slc9A1C/C and Slc9A1swe/swe mice showed a diffuse cytoplasmic staining pattern by CtNHE1 antiserum (Fig. 7E and F). In adrenal cortex from Slc9A1C/C mice, the antiserum distinctly stained the cell membranes but not the cytoplasm (Fig. 7G). CtNHE1 did not stain the

membranes from Slc9A1swe/swe mice adrenal cortex (Fig. 7H). The persistence of a labelling in islets or adrenal medulla from Slc9A1swe/swe mice was surprising, particularly since the exocrine pancreas, the liver and the adrenal cortex were negative, as expected. Therefore, to determine whether this reected a non-specic binding or revealed the presence of a fragment of NHE1 in Slc9A1swe/swe mice, RT-PCR was used to amplify
Figure 6 Inuence of b-cell degranulation on CtNHE1 detection in rat pancreas. (A and B) The insulin content was markedly decreased in islets from glibenclamide-treated rats as compared with control animals. (C and D) CtNHE1 staining of the same islets (step sections). Islets from control rats (C) showed a dense, diffuse labelling whereas those of glibenclamide-treated rats (D) displayed a fainter cytoplasmic labelling with a strong membrane pattern. These experiments were performed on 24 h-xed tissue, which explains why little signal was observed in acini. Bar: 50 mm.
dened fragments of NHE1 mRNA from liver, isolated islets and adrenals. The strategy was to seek out expression variations between full-length mRNA and a short-length mRNA downstream of the mutation point (Fig. 8A). A common primer was designed at the end of the epitope-coding region (P4). A short-end primer was chosen between the swe mutation point and the epitope-coding region (P1), and a full-length forward primer was chosen at the -end exon 2 as described previously (Bell et al. 1999). The largest amplied segment of 1834 bp (ex2/P4) encompasses a large portion of the full-length NHE1 mRNA (Fig. 8B). It was only observed in tissues from Slc9A1C/C animals. A smaller segment of 706 bp (P1/P4) corresponding mostly to the -end of the mRNA includes the region coding for the C-terminus epitope. It was observed in liver, islets and adrenals from Slc9A1C/C as well as Slc9A1swe/swe mice (Fig. 8B). A RT-PCR using primers designed for genotyping conrmed the mutated or wild-type sequence of mRNA extracted from these tissues (Fig. 8B: 200 bp). These results indicate that the mRNA sequence which encodes the epitope recognized by CtNHE1 antiserum is present in both Slc9A1C/C and Slc9A1swe/swe mice even if the latter lack full-length mRNA and NHE1 protein (Cox et al. 1997). By western blot, isolated islets and adrenals from both Slc9A1C/C and Slc9A1swe/swe mice showed a low

MW band at w65 kDa, which was absent in the liver. In contrast, the w100 kDa protein observed in Slc9A1C/C tissues was absent in Slc9A1swe/swe mice (Fig. 8C). These observations were conrmed by reprobing western blot membranes with the 4E9 monoclonal antibody directed against the same region of NHE1. In islets, the w100 kDa band was fainter than that of w65 kDa. In adrenals, the w100 kDa band appeared much more abundant than that of w65 kDa. We attribute this inverse pattern to the high proportion of cortical cells in the total adrenal protein extract. These ndings show that although NHE1 is effectively absent from Slc9A1swe/swe islets and adrenals, the low MW protein is still present. This is consistent with the persistence of an immunohistochemical signal (Fig. 7B) and the presence of a -end of NHE1 mRNA (Fig. 8B) in Slc9A1swe/swe islets and adrenal medulla. In contrast, no similar protein (Fig. 8C) and cytoplasmic labelling (Fig. 7D) were observed in liver cells.

Discussion

This study identied NHE1 as the major isoform of the NaC/HC exchanger in rat pancreatic islets. Expression of NHE1 was also established by RT-PCR in isolated
Figure 7 CtNHE1 immunostaining in control and Slc9A1 mutant mice. (A and B) Pancreas. Both diffuse insular and membrane acinar immunohistochemical stainings were observed in the pancreas from Slc9A1C/C mice (A). Only the diffuse islet labelling was observed in Slc9A1swe/swe mice (B). (C and D) Liver. The membrane labelling observed in Slc9A1C/C mice (C) was absent in Slc9A1swe/swe mice (D). (EH) Adrenals. A diffuse cytoplasmic staining pattern was observed in medulla from both control (E) and Slc9A1swe/swe (F) mice. A distinct membrane labelling was observed in adrenal cortex of control (G), but not of Slc9A1swe/swe (H) mice. Bars: AD: 50 mm; EH: 25 mm.

swe/swe

Figure 8 (A) Graphical representation of NHE1 mRNA. Exons are represented with alternate pattern. Introns are not represented. The mutation point in Slc9A1swe/swe (AAG-TAG: STOP) and the CtNHE1 epitope-coding sequence (Ct 147aa) are indicated. Arrow heads below represent the primers used in (B). (B) RT-PCR products using the selected primers on cDNA of liver, isolated islets and adrenals from Slc9A1C/C and Slc9A1swe/swe mice. ex2-P4 and P1-P4 correspond to the full-length and the -end of NHE1 mRNA respectively. MR0975MR0977 and MR0976 MR0977 detect wild-type and mutated sequences respectively. (C) Western blot using polyclonal or monoclonal antibodies with liver, isolated islets and adrenal extracts (40 mg/lane) from Slc9A1C/C and Slc9A1swe/swe mice. The w100 kDa band corresponds to NHE1. The low MW band is similar to that shown by the arrow in Fig. 1C (mouse islets).

Acknowledgements

P M was supported by grant FIRST 415795 from the Walloon Region of Belgium. J C J is Senior Research Associate from the Fonds National de la Recherche Scientique, Brussels, Belgium. This work was supported by Grants (to J R, J C J and O D) from the Fonds de la Recherche Scientique Medicale, Brussels, and grant ARC 05/10-328 from the Direction de la Recherche Scientique de la Communaute Francaise de Belgique. We are grateful to Dr M Donowitz for kindly providing the CtNHE1 antiserum and to Dr D Dufrane for providing human islets. We are also indebted to E Riveira Munoz for her expertise in genetic analysis and to Ph Camby, Y Cnops, H Debaix, M Nenquin and L Wenderickx for their skilful help. The authors declare that there is no conict of interest that would prejudice the impartiality of this scientic work.

References

Aizawa T, Sato Y & Komatsu M 2002 Importance of nonionic signals for glucose-induced biphasic insulin secretion. Diabetes 51 (Suppl 1) S96S98. Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ 1990 Basic local alignment search tool. Journal of Molecular Biology 215 403410. Anderie I, Blum R, Haase W, Grinstein S & Thevenod F 1998 Expression of NHE1 and NHE4 in rat pancreatic zymogen granule membranes. Biochemical and Biophysical Research Communications 246 330336. Baird NR, Orlowski J, Szabo EZ, Zaun HC, Schultheis PJ, Menon AG & Shull GE 1999 Molecular cloning, genomic organization, and functional expression of NaC/HC exchanger isoform 5 (NHE5) from human brain. Journal of Biological Chemistry 274 43774382. Barg S, Huang P, Eliasson L, Nelson DJ, Obermuller S, Rorsman P, Thevenod F & Renstrom E 2001 Priming of insulin granules for exocytosis by granular ClK uptake and acidication. Journal of Cell Science 114 21452154. Bell SM, Schreiner CM, Schultheis PJ, Miller ML, Evans RL, Vorhees CV, Shull GE & Scott WJ 1999 Targeted disruption of the murine Nhe1 locus induces ataxia, growth retardation, and seizures. American Journal of Physiology 276 C788C795. Bertrand G, Ishiyama N, Nenquin M, Ravier MA & Henquin JC 2002 The elevation of glutamate content and the amplication of insulin secretion in glucose-stimulated pancreatic islets are not causally related. Journal of Biological Chemistry 277 3288332891. Best L, Yates AP, Gordon C & Tomlinson S 1988 Modulation by cytosolic pH of calcium and rubidium uxes in rat pancreatic islets. Biochemical Pharmacology 37 46114615. Bonner-Weir S 1988 Morphological evidence for pancreatic polarity of beta-cell within islets of Langerhans. Diabetes 37 616621. Cavet ME, Akhter S, de Medina FS, Donowitz M & Tse CM 1999 NaC/HC exchangers (NHE1-3) have similar turnover numbers but different percentages on the cell surface. American Journal of Physiology 277 C1111C1121. Chambrey R, Warnock DG, Podevin RA, Bruneval P, Mandet C, Belair MF, Bariety J & Paillard M 1998 Immunolocalization of the NaC/HC exchanger isoform NHE2 in rat kidney. American Journal of Physiology 275 F379F386. Channaoui K & Rahier J 1992 Inuence of the decrease of intracellular antigenic content on morphometric analysis: effect of the type and dilution of the rst antibody. Histochemistry 97 389395.

Lindstrom P & Sehlin J 1984 Effect of glucose on the intracellular pH of pancreatic islet cells. Biochemical Journal 218 887892. Lindstrom P & Sehlin J 1986 Effect of intracellular alkalinization on pancreatic islet calcium uptake and insulin secretion. Biochemical Journal 239 199204. MacDonald MJ, Fahien LA, Brown LJ, Hasan NM, Buss JD & Kendrick MA 2005a Perspective: emerging evidence for signaling roles of mitochondrial anaplerotic products in insulin secretion. American Journal of Physiology Endocrinology and Metabolism 288 E1E15. MacDonald PE, Joseph JW & Rorsman P 2005b Glucose-sensing mechanisms in pancreatic beta-cells. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 360 22112225. Matschinsky FM, Magnuson MA, Zelent D, Jetton TL, Doliba N, Han Y, Taub R & Grimsby J 2006 The network of glucokinase-expressing cells in glucose homeostasis and the potential of glucokinase activators for diabetes therapy. Diabetes 55 112. Miley HE, Sheader EA, Brown PD & Best L 1997 Glucose-induced swelling in rat pancreatic beta-cells. Journal of Physiology 504 191198. Miley HE, Holden D, Grint R, Best L & Brown PD 1998 Regulatory volume increase in rat pancreatic beta-cells. Pugers Archiv 435 227230. Miyazaki E, Sakaguchi M, Wakabayashi S, Shigekawa M & Mihara K 2001 NHE6 protein possesses a signal peptide destined for endoplasmic reticulum membrane and localizes in secretory organelles of the cell. Journal of Biological Chemistry 276 4922149227. Nakamura N, Tanaka S, Teko Y, Mitsui K & Kanazawa H 2005 Four NaC/HC exchanger isoforms are distributed to Golgi and postGolgi compartments and are involved in organelle pH regulation. Journal of Biological Chemistry 280 15611572. Newgard CB 2002 While tinkering with the beta-cell.metabolic regulatory mechanisms and new therapeutic strategies. Diabetes 51 31413150. Numata M & Orlowski J 2001 Molecular cloning and characterization of a novel (NaC,KC)/HC exchanger localized to the trans-Golgi network. Journal of Biological Chemistry 276 1738717394. Numata M, Petrecca K, Lake N & Orlowski J 1998 Identication of a mitochondrial NaC/HC exchanger. Journal of Biological Chemistry 273 69516959. Orci L, Halban P, Perrelet A, Amherdt M, Ravazzola M & Anderson RGW 1994 pH-independent and -dependent cleavage of proinsulin in the same secretory vesicle. Journal of Cell Biology 126 11491156. Orlowski J & Grinstein S 1997 NaC/HC exchangers of mammalian cells. Journal of Biological Chemistry 272 2237322376. Orlowski J & Grinstein S 2004 Diversity of the mammalian sodium/proton exchanger SLC9 gene family. Pugers Archiv 447 549565. Ozaita A, Martone ME, Ellisman MH & Rudy B 2002 Differential subcellular localization of the two alternatively spliced isoforms of the Kv3.1 potassium channel subunit in brain. Journal of Neurophysiology 88 394408. Pace CS, Tarvin JT & Smith JS 1983 Stimulus-secretion coupling in beta-cells: modulation by pH. American Journal of Physiology 244 E3E18. Peti-Peterdi J, Chambrey R, Bebok Z, Biemesderfer D, St John PL, Abrahamson DR, Warnock DG & Bell PD 2000 Macula densa NaC/HC exchange activities mediated by apical NHE2 and basolateral NHE4 isoforms. American Journal of Physiology. Renal Physiology 278 F452F463. Rahier J, Stevens M, de Menten Y & Henquin JC 1989 Determination of antigen concentration in tissue sections by immunodensitometry. Laboratory Investigations 61 357363. Ritter M, Fuerst J, Woll E, Chwatal S, Gschwentner M, Lang F, Deetjen P & Paulmichl M 2001 NaC/HC exchangers: linking osmotic dysequilibrium to modied cell function. Cellular Physiology and Biochemistry 11 118.

Shepherd RM, Gilon P & Henquin JC 1996 Ketoisocaproic acid and leucine increase cytoplasmic pH in mouse pancreatic B cells: role of cytoplasmic Ca2C and pH-regulating exchangers. Endocrinology 137 677685. Stalker J, Gibbins B, Meidl P, Smith J, Spooner W, Hotz HR & Cox AV 2004 The Ensembl Web site: mechanics of a genome browser. Genome Research 14 951955. Stiernet P, Guiot Y, Gilon P & Henquin JC 2006 Glucose acutely decreases pH of secretory granules in mouse pancreatic islets. Mechanisms and inuence on insulin secretion. Journal of Biological Chemistry 281 2214222151. Straub SG & Sharp GW 2002 Glucose-stimulated signaling pathways in biphasic insulin secretion. Diabetes/Metabolism Research and Reviews 18 451463. Szaszi K, Paulsen A, Szabo EZ, Numata M, Grinstein S & Orlowski J 2002 Clathrin-mediated endocytosis and recycling of the neuronspecic NaC/HC exchanger NHE5 isoform. Regulation by phosphatidylinositol -kinase and the actin cytoskeleton. Journal of Biological Chemistry 277 4262342632. Zerbini G, Maestroni A, Breviario D, Mangili R & Casari G 2003 Alternative splicing of NHE-1 mediates NaLi countertransport and associates with activity rate. Diabetes 52 15111518. Zhang T, Haws P & Wu Q 2004 Multiple variable rst exons: a mechanism for cell- and tissue-specic gene regulation. Genome Research 14 7989.
Rossmann H, Sonnentag T, Heinzmann A, Seidler B, Bachmann O, Vieillard-Baron D, Gregor M & Seidler U 2001 Differential expression and regulation of NaC/HC exchanger isoforms in rabbit parietal and mucous cells. American Journal of Physiology. Gastrointestinal and Liver Physiology 281 G447G458. Roussa E, Alper SL & Thevenod F 2001 Immunolocalization of anion exchanger AE2, NaC/HC exchangers NHE1 and NHE4, and vacuolar type HC-ATPase in rat pancreas. Journal of Histochemistry and Cytochemistry 49 463474. Salgado A, Silva AM, Santos RM & Rosario LM 1996 Multiphasic action of glucose and alpha-ketoisocaproic acid on the cytosolic pH of pancreatic beta-cells. Evidence for an acidication pathway linked to the stimulation of Ca2C inux. Journal of Biological Chemistry 271 87388746. Seino S, Iwanaga T, Nagashima K & Miki T 2000 Diverse roles of K(ATP) channels learned from Kir6.2 genetically engineered mice. Diabetes 49 311318. Sempoux C, Guiot Y, Dubois D, Nollevaux MC, Saudubray JM, NihoulFekete C & Rahier J 1998 Pancreatic B-cell proliferation in persistent hyperinsulinemic hypoglycemia of infancy: an immunohistochemical study of 18 cases. Modern Pathology 11 444449. Sempoux C, Guiot Y, Dahan K, Moulin P, Stevens M, Lambot V, de Lonlay P, Fournet JC, Junien C, Jaubert F et al. 2003 The focal form of persistent hyperinsulinemic hypoglycemia of infancy: morphological and molecular studies show structural and functional differences with insulinoma. Diabetes 52 784794. Shepherd RM & Henquin JC 1995 The role of metabolism, cytoplasmic Ca2C, and pH-regulating exchangers in glucoseinduced rise of cytoplasmic pH in normal mouse pancreatic islets. Journal of Biological Chemistry 270 79157921.

Received 3 November 2006 Accepted 16 November 2006