Pharmacological characterization of the first in class clinical candidate PF-05190457: a selective ghrelin receptor competitive antagonist with inverse agonism that increases vagal afferent firing and glucose-dependent insulin secretion ex vivo

Abstract

Background and purpose: Ghrelin increases growth hormone secretion, gastric acid secretion, gastric motility and hunger but decreases glucose-dependent insulin secretion and insulin sensitivity in humans. Antagonizing the ghrelin receptor has potential as a therapeutic approach in the treatment of obesity and type 2 diabetes. Therefore, the aim was to pharmacologically characterize the novel small-molecule antagonist PF-05190457 and assess translational pharmacology ex vivo.

Experimental approach: Radioligand binding in filter and scintillation proximity assay formats were used to evaluate affinity, and europium-labelled GTP to assess functional activity. Rat vagal afferent firing and calcium imaging in dispersed islets were used as native tissues underlying food intake and insulin secretion respectively.

Key results: PF-05190457 was a potent and selective inverse agonist on constitutively active ghrelin receptors and acted as a competitive antagonist of ghrelin action, with a human Kd of 3 nM requiring 4 h to achieve equilibrium. Potency of PF-05190457 was similar across different species. PF-05190457 increased intracellular calcium within dispersed islets and increased vagal afferent firing in a concentration-dependent manner with similar potency but was threefold less potent as compared with the in vitro Ki in recombinant overexpressing cells. The effect of PF-05190457 on rodent islets was comparable with glibenclamide, but glucose-dependent and additive with the insulin secretagogue glucagon-like peptide-1.

Conclusions and implications: Together, these data provide the pharmacological in vitro and ex vivo characterization of the first ghrelin receptor inverse agonist, which has advanced into clinical trials to evaluate the therapeutic potential of blocking ghrelin receptors in obesity and type 2 diabetes.

Figure 1

(A) The in vitro potency of PF‐05190457 against the human ghrelin receptor using filter binding at 90 min and 24 h. The concentration–response curves represent the average ± SD for % inhibition at each concentration tested where n = 86–103 replicates over 52 (90 min) or 44 experiments (24 h). (B) The quantitative association and dissociation kinetics of PF‐05190457 using [¹²⁵I]ghrelin binding to human ghrelin receptor membranes in the presence or absence of 1, 10 or 100 nM PF‐05190457. The specific binding plot represents the arithmetic mean ± SEM at each time point tested over eight experiments. (C) The in vitro potency of PF‐051907457 against the ghrelin receptor across species. The concentration–response curves represent the mean ± SD for the % inhibitation values at each concentration tested where n = 16–103 replicates over 52 (human), 9 (dog), 8 (primate), 49 (rat) or 15 experiments (mouse). (D) The association and dissociation kinetics of [³H]PF‐05190457 a (3 and 10 nM) over 7 h. Each specific binding point represents the arithmetic mean ± SEM at each time point tested over seven experiments.

Figure 2

(A) The in vitro functional activity of PF‐05190457 against the human ghrelin receptor as measured by europium‐labelled guanosine‐5′‐triphosphate (Eu‐GTP). The agonist mode measured the ability of PF‐05190457 to act as an agonist, partial agonist or inverse agonist receptor, and the observed negative % effect values were characteristic of an inverse agonist. The antagonist mode measured the ability of increasing concentrations of PF‐05190457 to block ghrelin‐stimulated changes (B). The concentration–response plot is a representative experiment from n = 16 with 1–2 replicates per concentration and plotted as the mean % effect ± SD. (C) is the K _b potency of PF‐05190457. Data represent the mean ± SEM at each concentration of PF‐05190457 tested. Concentration ratio = EC₅₀ of ghrelin with PF‐05190457/EC₅₀ of ghrelin without PF‐05190457 (n = 10 experiments).

Figure 3

(A) An example trace of PF‐05190457 E _max 10⁻⁶ M increasing vagal afferent firing. (B) and (C) Concentration–response curves of PF‐05190457 and compound 2 respectively, on rat gastric vagal afferent firing ex vivo. Data represent mean ± SEM (PF‐05190457 0.1–1 nM n = 4; 10–1000 nM n = 5; compound 2 n = 4).

Figure 4

(A) Example raw trace of PF‐5109457 (1 × 10^–7 M) in the presence of 9 mM glucose on intracellular Ca^{2 +} in rat dispersed islets (A: 9 mM glucose buffer baseline; B: 1 × 10^–7 M PF‐05190457 in 9 mM glucose buffer; C: 9 mM glucose buffer wash; D: unrecorded wash to prevent bleaching and E: 1 × 10^–7 M Glibenclamide in 9 mM glucose wash). (B) The same as (A) but with compound 2. (C) The average glucose‐dependent effect of PF‐05190457 on intracellular Ca^{2 +} in rat dispersed islets in the presence of 3 and 9 mM glucose. All data are normalized against the positive control glibenclamide. Data represent the mean ± SEM (n = 4). Comparison of compound 2 to glibenclamide in the presence of 9 mM glucose (n = 3; D). Concentration–response curve of PF‐05190457 showing the mean ± SEM, where n = 3–7 pending concentration (E). (F) Additive effect of GLP‐1 EC₅₀ and PF‐05190457 IC₅₀ on intracellular calcium in rat dispersed islets in the presence of 9 mM glucose. All data are normalized against glibenclamide. Data represent the mean ± SEM (n = 6; ***P < 0.05 two‐tailed t‐test with Dunnett adjustment).

Figure 5

In vitro and ex vivo potency of PF‐05190457 across all assays against time. Good correlation in potency of PF‐05190457 is observed using assays with the same exposure time, with an observed increase in potency with longer incubation times as equilibrium is achieved around 4 h.