The diabetes-associated K+ channel TALK-2 controls human beta cell endoplasmic reticulum Ca2+ handling, which promotes basal insulin release and limits glucose-stimulated insulin secretion.

AIMS/HYPOTHESIS: The two-pore domain K+ channel TWIK1-related alkalinisation-activated K+ channel 2 (TALK-2) is encoded by KCNK17, which is one of the most abundant beta cell K+ channel transcripts that also shows high islet expression specificity. Polymorphisms that increase islet KNCK17 expression or result in TALK-2 gain-of-function are associated with a predisposition for developing type 2 diabetes. However, there is a gap in knowledge of the beta cell function(s) of TALK-2. As K+ channels typically control beta cell Ca2+ handling, we aimed to examine the TALK-2 channel control of beta cell Ca2+ homeostasis and the resulting impact on insulin secretion.

METHODS

Localisation of TALK-2 was evaluated with immunofluorescent staining as well as TALK-2-GFP construct co-expressed with intracellular markers. TALK-2 function was evaluated by measuring changes in cytoplasmic Ca2+ (Ca2+C), endoplasmic reticulum Ca2+ (Ca2+ER), ER membrane potential (Vm), K+ currents and insulin secretion in a TALK-2 inducible cell line and/or primary human beta cells with adenoviral-mediated shRNA knockdown (KD) of TALK-2 or scramble shRNA.

RESULTS

TALK-2 protein localised to the plasma membrane and ER membrane, and formed functional channels on the ER membrane. Ca2+ER release was accelerated by TALK-2 (slope for TALK-2-expressing cells vs controls: 14.8 ± 0.7 vs 8.9 ± 1.3, respectively, shown as mean ± SE), which reduced Ca2+ER storage (ΔCa2+ER amplitude: TALK-2-expressing cells reduced by 25 ± 5%) and increased basal relative Ca2+C (fold increase by 12 ± 2%). Furthermore, TALK-2 diminished ER membrane hyperpolarisation following Ca2+ER release (Accelerated Sensor of Action Potentials [ASAP3ER] amplitude decreased by 20 ± 0.8% in TALK-2-expressing cells), suggesting that TALK-2 strengthens the electrical driving force for Ca2+ER leak. In human beta cells, TALK-2-KD increased Ca2+ER stores by reducing Ca2+ER leak (2.30 ± 0.12 vs controls 2.65 ± 0.14). Moreover, TALK-2-KD reduced beta cell Ca2+C at euglycaemic conditions (2.88 ± 0.36 vs controls 3.16 ± 0.36) and increased beta cell Ca2+C influx in response to hyperglycaemic conditions (4.07 ± 0.55 vs controls 3.45 ± 0.48). Human pseudoislets with beta cell-specific TALK-2-KD displayed reduced basal insulin secretion (0.266 ± 0.065 vs controls 0.432 ± 0.073) and enhanced glucose-stimulated insulin secretion (GSIS; 85.01 ± 13.96 vs controls 42.53 ± 5.52).

CONCLUSIONS/INTERPRETATION: These data support the notion that TALK-2 functions on the human beta cell ER membrane to increase the electrical driving force for beta cell Ca2+ER release, reduces glucose-stimulated Ca2+ influx and limits GSIS. Furthermore, TALK-2-mediated amplification of Ca2+ER leak likely enhances basal insulin secretion by increasing Ca2+C. Therefore, polymorphisms in KCNK17 that increase TALK-2 activity or expression would be predicted to increase type 2 diabetes risk by blunting beta cell glucose-stimulated Ca2+ influx, limiting GSIS, promoting Ca2+ER leak and elevating basal insulin secretion.