Glepaglutide, a novel glucagon-like peptide-2 agonist, has anti-inflammatory and mucosal regenerative effects in an experimental model of inflammatory bowel disease in rats

Glepaglutide, formerly called ZP1848, was synthesized batchwise by means of a standard solid-phase procedure using the Fmoc strategy. The crude material was purified using preparative high-performance liquid chromatography. The identity of the pure material was confirmed by electrospray mass spectrometry [28]. Immediately before use, glepaglutide was dissolved in vehicle (15 mM histidine and 271 mM mannitol) and administered subcutaneously (SC) at a dose of 5 ml/kg. INDO (Sigma-Aldrich, Denmark, product number I7378-5G; 50 mg) was freshly dissolved in 14.3 ml of 0.9% NaCl supplemented with 27.1 mg NaH₂PO₄, H₂O, and 12 mg NaOH. Prior to administration at a dose of 2 ml/kg, the pH was adjusted to pH 6.8–7.3 with HCl.

Naive Wistar rats (Taconic, Ll. Skensved, Denmark), weighing 170–230 g, were used in the experiments. The animals were allowed at least 7 days of acclimatization in Makrolon type III cages before study start. Animals were housed (two per cage) under a 12:12-h light/dark cycle. Food (standard Altromin type 1321, Brogården, Denmark) and drinking water (domestic quality tap water with added citric acid to approximately pH 3) were provided ad libitum.

Glepaglutide (80 nmol/kg or 400 nmol/kg, SC, once daily) or vehicle (SC, once daily) were administered for 14 days to naive Wistar rats to assess the intestinotrophic properties of the peptide, before initiating studies in the small intestinal inflammation model (Fig. 1A). Overall, 18 rats were included per arm, with 6 rats each sacrificed on day 0, day 7, and day 14.

Small intestinal inflammation was induced by administration of INDO (7 mg/kg, SC, once daily) on 2 days consecutively (day 0 and day 1) in male Wistar rats (Fig. 1B and C). Two treatment regimens were studied: glepaglutide (400 nmol/kg, SC, twice daily) was administered in a co-treatment regimen (Fig. 1B – administered at inflammation onset and throughout), and a post-treatment regimen (Fig. 1C – administered after inflammation onset and throughout).

This is a well-established model of CD, because the small intestinal inflammation induced by INDO shares a number of pathological similarities with CD [29]. INDO produces small intestinal inflammation in a dose-dependent manner [29, 30]. Yamada et al. showed that one injection SC of INDO at 7.5 mg/kg caused acute damage to the intestine with mucosal erosions and ulcerations in the distal jejunum and proximal ileum. Inflammation was localized primarily on the mesenteric side of the mid-small intestine and persisted throughout the first 3 days, then gradually resolved within 1 week [29]. Two injections SC of INDO 7.5 mg/kg 24 h apart resulted in more severe chronic intestinal inflammation with multiple mucosal ulcers on the mesenteric side of the small intestine in the injected rats [29, 31]. With this regimen, the INDO-induced chronic damage was more pronounced in the region of the mid-small intestine and the inflammation persisted for approximately 2 weeks [29]. In this study, a dose of INDO 7 mg/kg was used based on the results of an INDO-induced small intestinal inflammation model validation study conducted by Zealand Pharma. The study showed that administration of INDO at 7 mg/kg (SC, once daily) on 2 days consecutively (day 0 and day 1) in male Wistar rats was sufficient and appropriate to introduce macroscopic features of inflammation (adhesions, intestinal wall thickening and ulcerations [32]), while 3 mg/kg was insufficient for inducing inflammation and 15 mg/kg led to exaggerated mortality (data not shown).

Two groups of controls, INDO controls (administered with INDO and treated with vehicle) and controls (administered with saline solution and treated with vehicle), were included in the studies (Fig. 1B and C). Groups of rats (n = 7–15) were sacrificed during the active inflammation phase (2–3 days after inflammation onset), on day 3 for the co-treatment regimen (Fig. 1B) and on day 4 for the post-treatment regimen (Fig. 1C), to determine small intestinal length, jejunal and ileal mass, and α-1-acid glycoprotein (α-1-AGP) and myeloperoxidase (MPO) concentrations in the jejunum and ileum. Equivalent groups of rats were sacrificed during the body weight recovery phase (14–15 days after inflammation onset), on day 14 for the co-treatment regimen (Fig. 1B), and on day 16 for the post-treatment regimen (Fig. 1C), to determine jejunal and ileal mass. Body weight and survival were recorded daily.

Animals were anesthetized with carbon dioxide and then sacrificed by cervical dislocation. The small intestine (from the pylorus to the ileocecal junction) was dissected away from the abdomen and the surrounding mesenteric fat. Small intestinal length was measured after weighing the small intestine with a 1 g weight to obtain uniform tension. Thereafter, the small intestine was divided into two segments of equal length labeled as the jejunum and ileum. Segments were gently flushed with 0.9% saline, padded dry, and weighed. A 5 cm biopsy was collected 10 cm distal from the start of the jejunal segment, and another 5 cm biopsy was collected 10 cm distal from the start of the ileal segment [29‐31]. The biopsies were immediately frozen in liquid nitrogen and stored at -80 °C for later analysis of concentrations of α-1-AGP and MPO using commercially available enzyme-linked immunosorbent assay kits (Life Diagnostics [2510–2] and Hycult Biotech [HK 210], respectively). The preparation method was validated in house.

An in vitro cyclic adenosine monophosphate (cAMP) assay was conducted to assess the stimulation of cAMP accumulation by glepaglutide [33]. Human GLP-2 (hGLP-2) was used as a reference agonist. A cell line stably expressing hGLP-2 receptor (hGLP-2R) was generated by Lipofectamine (Invitrogen #18,324–012) transfection of human embryonic kidney (HEK) 293 cells (ATCC #CRL-1573) with a mammalian expression vector containing the complementary DNA encoding the hGLP-2R (natural variant dbSNP:rs17681684) and a G418 resistance gene. Following 4 weeks in culture in selection with 0.5 mg/ml G418, single clones were picked and tested in a functional GLP-2 receptor potency assay as described below. One clone was selected for use in compound profiling. For assaying, HEK293 cells stably expressing hGLP-2R were seeded at 30,000 cells per well in 96-well microtiter plates coated with 0.01% poly-L-lysine. Cells were grown overnight in 200 µl growth medium (DMEM with Glutamax-I [Invitrogen 61965], containing 10% v/v FBS [Invitrogen 10,270–106], 1% v/v PenStrep [Invitrogen 15140], 0.5 mg/ml Geneticin [Invitrogen 10,131–027], 1 mM sodium pyruvate [Invitrogen 11,360–039], and 1 × NEAA [Invitrogen 11140]). On the day of assaying, the growth medium was removed, and the cells were washed once with 150 μl Tyrode buffer (Tyrode’s salts, Sigma-Aldrich T2145) supplemented with 10 mM HEPES (Invitrogen 15,630), pH 7.4. To initiate cAMP accumulation, cells were incubated in 100 μl Tyrode buffer supplemented with 0.1% w/v alkali-treated casein (Sigma-Aldrich C4765) and 100 μΜ IBMX (Sigma-Aldrich I5879) and increasing concentrations of glepaglutide or hGLP-2 for 15 min at 37 °C. The reaction was stopped by decanting off the compound/buffer and replacing it with 80 µl lysis/detection buffer consisting of deionized water supplemented with 0.1% w/v BSA (Sigma-Aldrich A9430), 5 mM HEPES, and 0.3% v/v Tween-20 (Sigma-Aldrich P7949). After incubation at room temperature for 10 min, the cAMP content of 10 µl of the resulting cell lysate was estimated using the AlphaScreen® cAMP assay kit (Perkin Elmer 6760635 M) according to the manufacturer’s instructions.

Compound potency (EC₅₀) and maximal effect (E_max) values were estimated by computer-aided curve fitting using a three-parameter logistic non-linear model. This was done after normalizing the compound response relative to a saturating concentration of the reference agonist (hGLP-2) alone and the baseline cAMP level of the unstimulated receptor.

Data are presented as mean ± standard error of the mean, unless otherwise stated. The intestinal growth-promoting effect of glepaglutide in naive rats was determined using a two-way analysis of variance (ANOVA, dose and time), and significant findings were further analyzed with the Bonferroni post hoc test. In the INDO model, a t-test was used to compare INDO control-treated rats with vehicle control-treated rats, and INDO control-treated rats with glepaglutide-treated rats. A p value of ≤ 0.05 was considered statistically significant. Grubbs’ test was used to test for outliers in the section, and outliers were not included in the analyses. Differences in survival between different groups were compared by the Cox–Mantel log-rank test.

The in vitro potency of glepaglutide was determined by its ability to stimulate cAMP formation in HEK293 cells recombinantly overexpressing hGLP-2R. Glepaglutide activated hGLP-2R with a similar potency to native hGLP-2, and the maximal agonist response of glepaglutide was likewise similar to that of hGLP-2 (Table S1).

Glepaglutide dose-dependently increased small intestinal mass in naive Wistar rats (Fig. 2), in line with observations made in a recent study in rats and dogs [34]. Small intestinal mass was dose-dependently increased after 7 days of daily dosing, with the 400 nmol/kg dose being significantly increased compared to the 80 nmol/kg dose (two-way ANOVA, Tukey multiple comparisons test, p < 0.05). In the 400 nmol/kg dose group, small intestinal mass rose to a plateau by day 7, and on day 14 small intestinal mass was similar in both dose groups (two-way ANOVA, Tukey multiple comparisons test, p = 0.5885). The 400 nmol/kg dose of glepaglutide was selected for use in the inflammation model because, at this dose, glepaglutide mediated a significant and sustained increase in small intestinal mass.

Co-treatment with glepaglutide significantly decreased acute body weight loss (p < 0.001; data not shown), reversed small intestinal shortening (p < 0.001; Fig. 3), and significantly decreased ileal α-1-AGP (p < 0.05; Fig. 4B) and ileal MPO (p < 0.001; Fig. 4D) concentrations. Post-treatment with glepaglutide had no effect on acute body weight loss (data not shown). However, it significantly reversed small intestinal shortening (p = 0.05; Fig. 3) and significantly decreased jejunal and ileal α-1-AGP (p < 0.01; Fig. 4A, B) and jejunal MPO (p < 0.05; Fig. 4C) concentrations in the active inflammation phase. Ileal mass was significantly increased in the glepaglutide co-treatment group during the active inflammation phase (p < 0.05; Fig. S2B). Both jejunal and ileal mass were significantly increased in the co- and post-treatment groups in the body weight recovery phase (Fig. S2A–D). Glepaglutide had no effect on survival in the regimens used.