Expression of Voltage-Gated Potassium Channels in Human and Rhesus Pancreatic Islets Lizhen Yan,1 David J. Figueroa,2 Christopher P. Austin,3,4 Yuan Liu,5 Randal M. Bugianesi,1 Robert S. Slaughter,1 Gregory J. Kaczorowski,1 and Martin G. Kohler1 Voltage-gated potassium channels (Kv channels) are urea receptor SUR1, sets the ␤-cell resting membrane involved in repolarization of excitable cells. In pancre- potentials (Em) under low plasma glucose conditions (3,4). atic ␤-cells, prolongation of the action potential by Elevated plasma glucose concentration results in an in- block of delayed rectifier potassium channels would be crease in metabolic activity, which leads to closure of KATP expected to increase intracellular free calcium and to channels and to membrane depolarization (5,6). Voltage- promote insulin release in a glucose-dependent manner. gated calcium channels (Ca channels) then become acti- However, the specific Kv channel subtypes responsible vated, and the resultant rise in intracellular Ca2ϩ triggers for repolarization in ␤-cells, most importantly in hu- mans, are not completely resolved. In this study, we insulin secretion. Sulfonylureas, widely used insulin secre- have investigated the expression of 26 subtypes from Kv tagogues, bind to the SUR1 receptor and block the KATP subfamilies in human islet mRNA. The results of the channel, causing insulin secretion in the absence of glu- RT-PCR analysis were extended by in situ hybridization cose metabolism (7,8). ␤ and/or immunohistochemical analysis on sections from KCa in -cells consists of at least two different compo- human or Rhesus pancreas. Cell-specific markers were ␤ nents. There is a large conductance KCa channel in -cells, used to show that Kv2.1, Kv3.2, Kv6.2, and Kv9.3 are with no obvious physiological function (9). In nondissoci- ␤ expressed in -cells, that Kv3.1 and Kv6.1 are expressed ated ␤-cells, a second type of K current has been in ␣-cells , and that Kv2.2 is expressed in ␦-cells. This Ca study suggests that more than one Kv channel subtype described (10). This current is linked to depolarization- might contribute to the ␤-cell delayed rectifier current induced rhythmic electrical activity of ␤-cells, important and that this current could be formed by heterotetra- for insulin secretion (11). mers of active and silent subunits. Diabetes 53: The remaining potassium current is generated by Kv 597–607, 2004 channels that produce either a fast transient current, IA, or a slow inactivating, delayed rectifying current, IDR (12–14). ␤ Both currents exist in -cells, with IDR being the major contributor to the repolarization of these cells. Thus, he role of potassium channels in excitation- blockage of IDR should enhance Ca influx and therefore secretion coupling is well established (1). In lead to an increase in insulin secretion as has been ␤ pancreatic -cells, insulin secretion is modulated previously reported (15–17). Because IDR does not open Tby the activity of different ionic currents. Among until the membrane is depolarized above a threshold level these are the three main potassium currents found in of ca Ϫ20 mV, its activation would be glucose dependent. ␤ -cells: the ATP-sensitive (KATP), calcium-activated (KCa), Therefore, pharmacological interference with this mecha- and voltage-gated (Kv) currents. Each has a functional role nism may provide a novel way to treat type 2 diabetes at different stages in the process of glucose-induced insu- without causing the hypoglycemic adverse effect of sulfo- lin secretion (2). nylureas (18,19). IDR has also been shown to be part of the KATP, consisting of inward rectifier Kir6.2 and sulfonyl- signaling pathway of glucagon-like peptide-1–induced glu- cose-dependent insulin secretion (20). The function of IDR ␦ ␤ From the 1Department of Ion Channels, Merck Research Laboratories, Rah- currents in -cells may be similar to that of -cells because way, New Jersey; the 2Department of Molecular and Investigative Toxicology, action potential initiation is dependent on depolarization Merck Research Laboratories, West Point, Pennsylvania; the 3Department of through metabolism-dependent blockage of K .In Neuroscience, Merck Research Laboratories, West Point, Pennsylvania; ATP the 4National Human Genome Research Institute, Bethesda, Maryland; and the ␣-cells, the opening of Na channels apparently initiates the 5Department of Bioinformatics, Merck Research Laboratories, West Point, action potential, but the IDR may still be involved in Pennsylvania. Address correspondence and reprint requests to Dr. Lizhen Yan or Dr. repolarization (21). Martin G. Kohler, RY80N-C31, Department of Ion Channels, Merck Research Kv channels belong to the six-transmembrane (TM) Laboratories, Rahway, NJ 07065. E-mail: [email protected] or martin_ family of K channels, where Kv1 to Kv11 subfamilies exist, [email protected]. Received for publication 13 April 2003 and accepted in revised form 14 although Kv7 is only found in Aplysia (22,23). Members of November 2003. the Kv1 to Kv4 subfamilies form tetrameric functional L.Y. and D.J.F. contributed equally to this article. channels, homomultimers or heteromultimers, usually DAPI, 4,6-diamidino-2-phenylindole; Em, membrane potential; IA, fast tran- sient current; IDR, delayed rectifying current; IHC, immunohistochemistry; with members from the same subfamily. Members of the ISH, in situ hybridization; KATP current, ATP-sensitive potassium current; KCa Kv5 to Kv11 families code for “silent subunits” that do not current, calcium-activated potassium current; Kv current, voltage-gated po- tassium current; TM, transmembrane. express as functional homomultimers. In heterologous © 2004 by the American Diabetes Association. expression systems, silent subunits can coassemble with DIABETES, VOL. 53, MARCH 2004 597 Kv CHANNELS IN PRIMATE ISLETS TABLE 1 PCR primers used for RT-PCR amplification of Kv channels Subtype Accession no. Sense primer (5Ј to 3Ј) Antisense primer (5Ј to 3Ј) Kv1.1 L02750 CATCTGGTTCTCCTTCGAGC GTTAGGGGAACTGACGTGGA Kv1.2 L02752 TCCGGGATGAGAATGAAGAC TTGGACAGCTTGTCACTTGC Kv1.3 M55515 GTTCTCCTTCGAACTGCTGG CTGAAGAGGAGAGGTGCTGG Kv1.4 M55514 CCCCAGCTTTGATGCCATCTTG TGAGGATGGCAAAGGACATGGC Kv1.5 M55513 TGCGTCATCTGGTTCACCTTCG TGTTCAGCAAGCCTCCCATTCC Kv1.6 X17622 TCAACAGGATGGAAACCAGCCC CTGCCATCTGCAACACGATTCC Kv1.7 AJ310479 TGCCCTTCAATGACCCGTTCTTC AAGACACGCACCAATCGGATGAC Kv2.1 L02840 TACAGCCTCGACGACAACG ACCACGCGGCGGACATTCTG Kv2.2 U69962 AACGAAGAACTGAGGCGAGAG ACTCCGCCTAAGGGTGAAAC Kv3.1 S56770 AACCCCATCGTGAACAAGACGG TCATGGTGACCACGGCCCA Kv3.2 AI363404 CTGCTGCTGGATGACCTACC TGTGCCATTGATGACTGGTT Kv3.3 AF055989 TTCTGCCTGGAAACCCATGAGG TGTTGACAATGACGGGCACAGG Kv3.4 M64676 TTCAAGCTCACACGCCACTTCG TGCCAAATCCCAAGGTCTGAGG Kv4.1 AJ005898 ATCTCGAGGAGATGAGGTTC TTCTTTCGGTCCCGATAC Kv4.3 AF048712 TGGCTTCTTCATCGCTGTCTCG CCGAAGATCTTCCCTGCAATCG Kv4.4 NM_012283 AGCCAAGAAGAACAAGCTG AGGAAGTTTAGGACATGCC Kv5.1 AF033382 TCCACATGAAGAAGGGCATCTGC TCACGTAGAAGGGGAGGATG Kv6.1 AF033383 TGCACCAACTTCGACGACATCC GGAACTCCAGGGAGAACCAGCC Kv6.2 AJ0111021 AAGCTCTTCGCCTGCGTGTC CAGCAGCAGCGACACGTAGAAC Kv6.3 NM_172347 ATGCCCATGCCTTCCAGAGA AGAGCTGCACGATCTCCTCG Kv8.1 AF167082 TTCCACAGCTGCCCGTATCTTTG TTTTGCCTGTGGTGGTGTCTGG Kv9.1 AF043473 TTTGAGGACTTGCTGAGCAGCG TTGCTCCAGGCACACCAACAAG Kv9.2 XM_043106 GTACTGGGGCATCAACGAGT CCACGGAGAGGTAGAGCAAG Kv9.3 AF043472 CTCTGTGGGCATTTCCATTT AGAAACAGGCACAAACACCC Kv10.1 AF348982 GCTTGCCCGTCACTTCATTGGTC TTCTTCCAGGCACTGTGATAGGA Kv11.1 AF348983 AGCCATGCTCAAACAGAGTG CTCCTCGTAGTCGTCGCACA Kv2 and Kv3 subunits and modulate the biophysical char- subtype-specific primer pair for each of the 26 members of the Kv1–Kv11 acteristics of the latter subunits (24–27). There are several families. To obtain efficient and specific primer pairs, we used the Vector NTI difficulties that obscure the correlation of any particular program (Informax, Frederick, MD) to select sequences. Each primer se- Kv subunit with a specific physiological function: the high quence then was submitted to a basic local alignment search tool search degree of sequence homology results in many Kv channels against GenBank to ensure specificity of the selected sequence. Specific having similar pharmacological and biophysical proper- primer pairs were then used to amplify Kv channel subtypes from human fetal ties, and most excitable cells express more than one Kv brain cDNA. Finally, the most effective primer pairs for each subtype were channel gene. used to study the expression of Kv channels in human islets (Table 1). Antibodies. A rabbit polyclonal antibody for Kv1.6 was raised against the Previous studies have identified Kv2.1 and Kv3.2 in peptide RRSSYLPTPHRAYAEKRM, corresponding to residue 509–526 of the rodent ␤-cells and insulinoma cells (16,28–30), and block rat Kv1.6 (34). Human Kv1.6 shares 17 of 18 amino acid residues with the rat of Kv2.1 has been implicated in eliciting glucose-depen- channel in this region. Rabbit polyclonal antibodies against Kv2.1 and Kv3.2 dent insulin secretion (16,31,32). While Kv2.1 has been proteins were purchased from Alomone Labs (Jerusalem, Israel). The Kv2.1 detected in human islets (33), no studies have yet been antibody was raised against the peptide HMLPGGGAHGSTRDQSI, corre- ␤ sponding to residue 837–853 of rat Kv2.1. Human Kv2.1 shares 15 of 17 amino attempted in human or primate -cells to define the acids with this region of the rat channel. The Kv3.2 antibody was raised molecular components of IDR. Although a molecular basis against the peptide DLGGKRLGIEDAAGLGGPDGK(C), corresponding to res- for the A-type current has been reported