The neuronal K+Cl− co-transporter 2 (Slc12a5) modulates insulin secretion

Intracellular chloride concentration ([Cl−]i) in pancreatic β-cells is kept above electrochemical equilibrium due to the predominant functional presence of Cl− loaders such as the Na+K+2Cl− co-transporter 1 (Slc12a2) over Cl−extruders of unidentified nature. Using molecular cloning, RT-PCR, Western blotting, immunolocalization and in vitro functional assays, we establish that the “neuron-specific” K+Cl− co-transporter 2 (KCC2, Slc12a5) is expressed in several endocrine cells of the pancreatic islet, including glucagon secreting α-cells, but particularly in insulin-secreting β-cells, where we provide evidence for its role in the insulin secretory response. Three KCC2 splice variants were identified: the formerly described KCC2a and KCC2b along with a novel one lacking exon 25 (KCC2a-S25). This new variant is undetectable in brain or spinal cord, the only and most abundant known sources of KCC2. Inhibition of KCC2 activity in clonal MIN6 β-cells increases basal and glucose-stimulated insulin secretion and Ca2+ uptake in the presence of glibenclamide, an inhibitor of the ATP-dependent potassium (KATP)-channels, thus suggesting a possible mechanism underlying KCC2-dependent insulin release. We propose that the long-time considered “neuron-specific” KCC2 co-transporter is expressed in pancreatic islet β-cells where it modulates Ca2+-dependent insulin secretion.

Full size blots and gels: Shown are original blots and gels that were used to construct cropped Figures 1F, 1H and 1i in the main text.

Supplementary Table 1:
The primer-sets used in RT-PCR experiments are named after the target transcript (KCC1-4, GAPDH) or variants (KCC2a-b, KCC3a-d), followed by numbers indicating amplicon sizes in base pairs (bp). Sense and antisense primer sequences encompass transcript positions, which are numbered relative to start codon A 1 TG. Primers were designed according to mouse (m), rat (r) or human (h) cDNAs and/or RefSeq nucleotide sequences as indicated.

Supplementary Figure 2: Localisation of NKCC1, KCC2 and GLUT2 in the plasma membrane of islet cells. A-C.
Mouse pancreatic sections were immunolabeled with antibodies recognising all cadherin members (pan-Cadherin) as plasma membrane marker and co-labeled against NKCC1 (A), KCC2 (B) or GLUT2 (C). Also shown in C, are insulin-labeled endocrine cells. D-F. Images obtained by using the Colocalization Finder tool of ImageJ and representing co-localisation (green) of pan-Cadherin and NKCC1 (D), KCC2 (E) and GLUT2 (F). Also shown in F, co-localization of pan-Cadherin and GLUT2 in in the plasma membrane of insulin producing !-cells. G-i. Shown are RGB profiles (signal intensities) obtained by using the ImageJ RGB Profiler tool along the yellow lines depicted in A, B and C. Green and red traces represent pan-Cadherin and NKCC1 (G), KCC2 (H) or GLUT2 (i) fluorescence intensities, respectively, whereas the blue trace represents that of insulin-labeled cells (i).

Supplementary Figure 3: KCC1, KCC3 and KCC4 mRNAs expression in MIN6 !-cells. A.
Representative RT-PCR experiments demonstrating transcript expression of the main splice variants of KCC1, KCC3 and KCC4. Shown are amplicons of expected lengths for KCC1 (440bp, 529bp and 531bp), KCC3 (589bp), KCC3a (401bp), KCC3b (695bp), KCC3c (675bp), KCC3d (356bp) and KCC4 (663bp and 691bp). As a positive control, RT-PCR was directed by using GAPDH primers (555 bp). As a negative control of RT-PCR reactions, water was used instead of cDNA and performed with GAPDH primers. B. Quantitative PCR of KCC1, KCC2, KCC3 and KCC4 mRNAs. Results are expressed relative to KCC2 (total) transcripts. C. Schematic representation of KCC1, KCC3 and KCC4 transcripts of reference (RefSeqs) along with accession numbers. The coding regions and exons are indicated as black and grey arrowed boxes, respectively. The 5'-and 3'-untranslated regions (UTRs) are represented as filled lines and the relative positions of the primer sets used for transcript screening in RT-PCR experiments as opposite arrowheads. Amplicons quantified by qPCR primers are shown as colored horizontal bars.

Supplementary Figure 4: Expression of mKCC2-S25 in medullary adrenal cells. A-B.
Representative immunofluorescence microscopy images of the rat adrenal gland (20x, A or 100x, B) obtained using rabbit polyclonal KCC2 antibodies (07-432) and Cy3-conjugated secondary antibodies in 5 #m thick sections. KCC2 expression is shown in the medulla of the gland. Nuclei were counterstained with DAPI. Scale bars represent 50#m. C. Representative RT-PCR experiment showing KCC2 mRNA expression (KCC2a + KCC2b) detected by using the KCC2-565 primer set. GAPDH-555 primer set was used in the positive RT-PCR control reaction. As negative control, water instead of total cDNA was used in RT-PCR reactions with the KCC2-565 primer set. D. MspI digestion of KCC2-565 PCR products showing a restriction pattern consistent with the absence of exon 25 (362 bp). E. MspI digestion control of KCC2-565 products obtained using as PCR templates pCMV-myc.KCC2a and pCMV-myc.KCC2a-S25 cloning plasmids. F-G. Overexpression of myc-tagged mKCC2a-S25 in COS7 (F) and MIN6 !cells (G) demonstrating that KCC2a-S25 cloned from MIN6 directs protein translation.

Supplementary Figure 5: KCC2 expression in glucagon-positive "-cells. A-C.
Immunofluorescence microscopy images of pancreatic tissue obtained from an STZ-diabetic mouse immunolabeled against KCC2 (A, red), glucagon (B, green) and insulin (C, blue) using Cy3-, Alexa-fluor 488-or DyLight405-labeled secondary antibodies, respectively. D-E. Shown are superimposed images of A and C (D) and of A and B (E). F. Shown is a high magnification (100x) image of STZ pancreatic tissue to visualise and identify immunoreactive KCC2 in glucagon-positive and -negative cells of the islet (white arrows). G-J. KCC2 expression in "TC6 cells at the protein and mRNA levels. An original Western blot analysis of 100-50#g "-TC6 protein extracts developed with validated KCC2 antibodies (07-432) is shown (G) along with representative RT-PCR (H) and RT-qPCR using the KCC2-147 primer set which does not distinguish among KCC2 variants (see Table 2) (i) and a confocal immunofluorescence microscopy image of "TC6 cells attached to glass coverslips where white arrows indicate immunoreactive KCC2 near or at the plasma membrane region (J).