Possible New Strategies for the Treatment of Congenital Hyperinsulinism

Abstract

Objective: Congenital hyperinsulinism (CHI) is a rare disease characterized by persistent hypoglycemia as a result of inappropriate insulin secretion, which can lead to irreversible neurological defects in infants. Poor efficacy and strong adverse effects of the current medications impede successful treatment. The aim of the study was to investigate new approaches to silence β-cells and thus attenuate insulin secretion.

Research design and methods: In the scope of our research, we tested substances more selective and more potent than the gold standard diazoxide that also interact with neuroendocrine ATP-sensitive K⁺ (K_ATP) channels. Additionally, K_ATP channel-independent targets as Ca²⁺-activated K⁺ channels of intermediate conductance (K_Ca3.1) and L-type Ca²⁺ channels were investigated. Experiments were performed using human islet cell clusters isolated from tissue of CHI patients (histologically classified as pathological) and islet cell clusters obtained from C57BL/6N (WT) or SUR1 knockout (SUR1^-/-) mice. The cytosolic Ca²⁺ concentration ([Ca²⁺]_c) was used as a parameter for the pathway regulated by electrical activity and was determined by fura-2 fluorescence. The mitochondrial membrane potential (ΔΨ) was determined by rhodamine 123 fluorescence and single channel currents were measured by the patch-clamp technique.

Results: The selective K_ATP channel opener NN414 (5 µM) diminished [Ca²⁺]_c in isolated human CHI islet cell clusters and WT mouse islet cell clusters stimulated with 10 mM glucose. In islet cell clusters lacking functional K_ATP channels (SUR1^-/-) the drug was without effect. VU0071063 (30 µM), another K_ATP channel opener considered to be selective, lowered [Ca²⁺]_c in human CHI islet cell clusters. The compound was also effective in islet cell clusters from SUR1^-/- mice, showing that [Ca²⁺]_c is influenced by additional effects besides K_ATP channels. Contrasting to NN414, the drug depolarized ΔΨ in murine islet cell clusters pointing to severe interference with mitochondrial metabolism. An opener of K_Ca3.1 channels, DCEBIO (100 µM), significantly decreased [Ca²⁺]_c in SUR1^-/- and human CHI islet cell clusters. To target L-type Ca²⁺ channels we tested two already approved drugs, dextromethorphan (DXM) and simvastatin. DXM (100 µM) efficiently diminished [Ca²⁺]_c in stimulated human CHI islet cell clusters as well as in stimulated SUR1^-/- islet cell clusters. Similar effects on [Ca²⁺]_c were observed in experiments with simvastatin (7.2 µM).

Conclusions: NN414 seems to provide a good alternative to the currently used K_ATP channel opener diazoxide. Targeting K_Ca3.1 channels by channel openers or L-type Ca²⁺ channels by DXM or simvastatin might be valuable approaches for treatment of CHI caused by mutations of K_ATP channels not sensitive to K_ATP channel openers.

Keywords: KATP channels; KCa3.1 channels; L-type Ca2+ channels; NN414; congenital hyperinsulinism; diazoxide.

Figure 1

Nifedipine and diazoxide reduce [Ca²⁺]_c in human CHI islet cell clusters. (A) Representative recording showing inhibition of glucose-induced oscillations of [Ca²⁺]_c by nifedipine (5 µM) in the presence of 10 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a patient with diffuse CHI (Table 1: 7, depicted as “human, Pt. #7” in the figure). (B) Summary of all experiments recorded in the presence of 10 mM glucose comprising 11 islet cell clusters isolated from pancreatic tissue of one CHI patient (Table 1: 7). (C) Representative recording showing the influence of diazoxide (250 µM) on glucose-induced oscillations of [Ca²⁺]_c in the presence of 10 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a patient with focal CHI (Table 1: 4). (D) Summary of all respective experiments from two patients, one with focal and one with mosaic form of CHI (Table 1: 1, black circles; 4, white circles) (n = 5). *p ≤ 0.05 and ***p ≤ 0.001.

Figure 2

Effects of NN414 on [Ca²⁺]_c and mitochondrial membrane potential ΔΨ. (A) Representative recording showing the reduction of glucose-induced oscillations of [Ca²⁺]_c by NN414 (5 µM) in the presence of 10 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a patient with diffuse CHI (Table 1: 2). (B) Summary of all respective experiments from three patients, one with diffuse, one with focal, and one with mosaic form (Table 1: 1, black circles; 2, white circles; and 4, gray circles) (n = 30). (C) Representative recordings showing the effect of NN414 (5 µM) on glucose-induced oscillations of [Ca²⁺]_c in islet cell clusters from WT (dashed curve) and SUR1^-/- (gray curve) mice. NN414 significantly reduced [Ca²⁺]_c in islet cell clusters from WT mice, but not in islet cell clusters from SUR1^-/- mice. (D) Summary of all respective experiments (n = 10 for each genotype, three different mouse preparations for each series). (E) Typical recordings showing measurement of ΔΨ in islet cell clusters obtained from WT (dashed curve) and SUR1^-/- (gray curve) mice. The switch from 0.5 to 10 mM glucose hyperpolarizes ΔΨ. The addition of NN414 has no influence on ΔΨ in WT and SUR^-/- islet cell clusters, respectively. (F) Summary of all experiments made under these conditions (n = 13, three different mouse preparations for each series). ***p ≤ 0.001.

Figure 3

Diazoxide and NN414 open mutated K_ATP channels. (A) Representative trace showing the activation of SUR1_E1507K/WT Kir 6.2 channels expressed in HEK-293 cells by the channel agonists diazoxide and NN414. This mutation is associated with CHI. At the beginning of the experiment the patch was pulled in nucleotide-free medium, which activates numerous channels as inhibitory nucleotides leave the pore. Addition of ATP rapidly inhibits almost all channel activity. Concurrent application of diazoxide (340 µM) or NN414 (5 µM) enhances channel activity. (B, C) Summary of all experiments with diazoxide (n = 7) and NN414 (n = 7), respectively. *p ≤ 0.05. The inset shows single channel openings at extended scales. The channel has an amplitude of about 4 pA, giving, at 50 mV driving force, a conductance of 80 pS, which is typical for K_ATP channels under these conditions. Four min under control conditions of the continuous recording were taken out for the clarity of the figure.

Figure 4

K_ATP channel-dependent and -independent effects of VU0071063. (A) Representative recording showing the reduction of [Ca²⁺]_c by VU0071063 (30 µM) in the presence of 10 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a CHI patients with a focal lesion (Table 1: 4). The star depicts the nadir after wash-out of VU0071063. (B) Summary of four experiments obtained from two patients, one with focal and one with diffuse form of CHI. VU0071063 rapidly reduced [Ca²⁺]_c in all 4 experiments, but due to the low number of experiments, the effect is not significant. (Table 1: 2, black circles; 4, white circles). (C) Representative recordings showing the effect of VU0071063 (30 µM) on oscillations of [Ca²⁺]_c induced by 10 mM glucose in islet cell clusters from WT (dashed curve) and SUR1^-/- (gray curve) mice. VU0071063 significantly reduced [Ca²⁺]_c in islet cell clusters from SUR1^-/- mice, revealing K_ATP channel-independent effects of the compound. Note the drop in [Ca²⁺]_c after removal of VU0071063 (black star: WT, gray star: SUR1^-/-). (D) Summary of all respective experiments; n = 45 and 29 for WT and SUR^-/- islet cell clusters. (E) Representative recordings showing the effect of VU0071063 (30 µM) on the mitochondrial membrane potential (ΔΨ) in islet cell clusters obtained from WT (dashed curve) and SUR1^-/- (gray curve) mice. (F) Summary of all respective experiments; n = 42 and 39 for WT and SUR^-/- islet cell clusters. Cell cluster were isolated from three WT and three SUR1^-/- mice. *p ≤ 0.05 and ***p ≤ 0.001.

Figure 5

The K_Ca3.1 channel opener DCEBIO reduces [Ca²⁺]_c in islet cell clusters isolated from SUR1^-/- mice and in human islet cell clusters. (A) Representative recording showing rapid inhibition of glucose-induced oscillations of [Ca²⁺]_c by DCEBIO (100 µM) in the presence of 10 mM glucose in islet cell clusters from SUR1^-/- mice. (B) Summary of all respective experiments; n = 30. Islet cell clusters were obtained from three different SUR1^-/- mice preparations. ***p ≤ 0.001. (C) Representative recording showing the reduction of glucose-induced oscillations of [Ca²⁺]_c by DCEBIO (100 µM) in the presence of 10 mM glucose in a human islet cell cluster isolated from pancreatic tissue affected by diffuse CHI (Table 1: 2). (D) Summary of all respective experiments from biopsies of two CHI patients, one with diffuse, one with mosaic form (Table 1: 1, black circles; 2, white circles) (n = 27). ***p ≤ 0.001.

Figure 6

DXM lowers [Ca²⁺]_c in islet cell clusters lacking functional K_ATP channels. (A) Representative recording showing a rapid decrease of [Ca²⁺]_c by DXM (100 µM) in the presence of 10 mM glucose in an islet cell cluster from a SUR1^-/- mouse. (B) Summary of all respective experiments (n = 13) with different cell cluster from three SUR1^-/- mice. ***p ≤ 0.001. (C) Representative recording showing reduction of [Ca²⁺]_c by DXM (100 µM) in the presence of 15 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a patient with focal CHI (Table 1: 5). (D) Summary of all respective experiments obtained from biopsies of two patients with focal and two patients with diffuse CHI (Table 1: 3, gray circles; 5, black circles; 6, white circles; and 7, hatched circles) (n = 16). ***p ≤ 0.001.

Figure 7

Simvastatin as a potential strategy to silence islet cell clusters affected by CHI. (A) Representative recording showing rapid inhibition of glucose-induced oscillations of [Ca²⁺]_c by simvastatin (7.2 µM) in the presence of 10 mM glucose in an islet cell cluster from SUR1^-/- mice. (B) Summary of all respective experiments (n = 43) with islet cell clusters obtained from three SUR1^-/- mice. ***p ≤ 0.001. (C) Representative recording showing the reduction of glucose-induced oscillations of [Ca²⁺]_c by simvastatin (7.2 µM) in the presence of 15 mM glucose in a human islet cell cluster isolated from pancreatic tissue of a patient affected by focal CHI (Table 1: 5). (D) Summary of all respective experiments from biopsies of two patients with focal and two patients with diffuse CHI (Table 1: 5, black circles; 6, white circles; 7, gray circles; and 8, hatched circles) (n = 18). ***p ≤ 0.001.