A novel drug-like water-soluble small molecule Focal Adhesion Kinase (FAK) activator promotes intestinal mucosal healing

Abstract

Non-steroidal anti-inflammatory drugs (NSAIDs) injure the proximal and distal gut by different mechanisms. While many drugs reduce gastrointestinal injury, no drug directly stimulates mucosal wound healing. Focal adhesion kinase (FAK), a non-receptor tyrosine kinase, induces epithelial sheet migration. We synthesized and evaluated a water-soluble FAK-activating small molecule, M64HCl, with drug-like properties. Monolayer wound closure and Western blots measured migration and FAK phosphorylation in Caco-2 ​cells, in vitro kinase assays established FAK activation, and pharmacologic tests assessed drug-like properties. 30 ​mg/kg/day M64HCl was administered in two murine small intestine injury models for 4 days. M64HCl (0.1-1000 ​nM) dose-dependently increased Caco-2 FAK-Tyr 397 phosphorylation, without activating Pyk2 and accelerated Caco-2 monolayer wound closure. M64HCl dose-responsively activates the FAK kinase domain vs. the non-salt M64, increasing the V_max of ATP-binding. Pharmacologic tests suggested M64HCl has drug-like properties and is enterally absorbed. M64HCl 25 ​mg/kg/day continuous infusion promoted healing of ischemic jejunal ulcers and indomethacin-induced small intestinal injury in C57Bl/6 mice. M64HCl-treated mice exhibited smaller ulcers 4 days after ischemic ulcer induction or indomethacin injury. Renal histology and plasma creatinine were normal. Mild hepatic inflammatory changes and ALT elevation were similar among M64HCl-treated mice and controls. M64HCl was concentrated in kidney and gastrointestinal mucosa and functional nephrectomy studies suggested predominantly urinary excretion. Little toxicity was observed in vitro or in single-dose mouse toxicity studies until >1000x higher than effective concentrations. M64HCl, a water-soluble FAK activator, promotes epithelial restitution and intestinal mucosal healing and may be useful to treat gut mucosal injury.

Keywords: Focal adhesion kinase; Mucosal healing; Non-steroidal anti-inflammatory drugs; Small intestine; Ulcer.

Graphical abstract

Fig. 1

(a) Structure of M64. (b) Structure of M64HCl. (c) M64 purity by LC/MS. (d) M64HCl purity by LC/MS.

Fig. 2

The effect of M64HCl on phosphorylation of FAK-Tyr-397 in human Caco-2 ​cells. (a) Representative blots and Tyr-397/FAK fold change in Caco-2 ​cells in suspension at 100pM and 1000pM concentration (n ​= ​4, ∗p ​< ​0.05). (b) Representative blots and Tyr-397/FAK fold change in Caco-2 ​cells in suspension at 0–1000 ​nM after treatment with M64HCl (n ​= ​6, ∗p ​< ​0.05). (c) ZN40099027 as a positive control activates FAK at 100 ​nM (n ​= ​8, p ​< ​0.05). (d) Representative blots and Tyr-397/FAK fold change in AGS cells treated with M64HCl at 100 ​nM.

Fig. 3

(a) The non-salt version of the molecule M64 does not activate FAK at 10–1000 ​nM. (b) M64HCl at 10–10,000 ​nM does not stimulate the phosphorylation of Pyk2 at Tyr-402. (n ​= ​4), or of Src (c) at Tyr-419 (n ​= ​4).

Fig. 4

M64HCl stability in simulated fluid. (a) M64HCl concentration in gastric simulated fluid at 0–4 ​h compared to stability in saline. (b)) M64HCl concentration in intestinal simulated fluid at 0–4 ​h compared to saline.

Fig. 5

Inhibition of CYP2C9, CYP2C19, CYP2D6 and CYP3A4 by M64HCl in vitro in human liver microsomes.

Fig. 6

The effect of M64HCl on Caco-2 ​cell monolayer wound closure. (a) Typical wound images treated with H2O or M64HCl at 10 ​nM or 100 ​nM. Images were taken at 0 and 24 ​h after wounding. (b) Typical wound images treated with H2O or M64HCl at 10 ​nM in the presence of 4 ​mM hydroxyurea. (c) 10 ​nM or 100 ​nM M64HCl accelerates wound closure in Caco-2 monolayers. (n ​= ​16, pooled from 4 separate studies, ∗p ​< ​0.05). (d) M64HCl enhances Caco-2 ​cell monolayer wound even when proliferation is blocked by 4 ​mM hydroxyurea (n ​= ​16, ∗p ​< ​0.05).

Fig. 7

M64HCl interacts with the FAK kinase domain. (a) M64HCl stimulates the conversion of ATP to ADP by highly purified full-length 125 ​kDa human FAK in an in vitro kinase assay 2 independent experiments with 4 replicates in each experiment, ∗p ​< ​0.05). (b) M64HCl binding stimulates the conversion of ATP to ADP by the 35 ​kDa FAK kinase domain (Figure represents pooled data from 3 experiments with 4 replicates in each experiment, n ​= ​2, ∗p ​< ​0.01). (c) M64HCl dose dependently activates the FAK kinase domain compared to the non-salt M64 molecule. (d) M64HCl increases the Vmax of ATP-binding vs. H2O control (n ​= ​6, ∗p ​< ​0.05).

Fig. 8

M64HCl promotes mucosal healing in two murine small intestinal injury models. (a) M64HCl plasma concentration 0–12 ​h after a single intraperitoneal injection of 2 ​mg/kg in mice. (b) M64HCl plasma concentration after 10 ​mg/kg gavage administration. (c) Representative images of jejunal ischemic ulcers in saline-treated mice and M64HCl-treated mice 4 days after ulcer-induction. (d) Ischemic ulcer area after 4 days of M64HCl continuous infusion or saline treatment (3.28 ​± ​0.50 vs. 5.06 ​± ​0.37 ​mm², n ​= ​10, ∗p ​< ​0.05). (e) Representative images of indomethacin induced small intestine ulcers from saline-treated mice and M64HCl-treated mice. (f) Total ulcer size in indomethacin induced small intestinal injury treated with saline control, M64HCl, misoprostol or M64HCl plus misoprostol (3.92 ​± ​0.37 ​mm² in saline vehicle vs. 2.69 ​± ​0.29 ​mm² in M64HCl vs. 2.84 ​± ​0.32 ​mm² in misoprostol only vs.2.07 ​± ​0.29 ​mm² in synergistic treatment, n ​= ​12, ∗p ​< ​0.05). (g) Representative immunohistochemical images of FAK-397 phosphorylation in epithelium at the edge of the ulcer bed from saline control or M64HCl-treated mice. Arrows indicate epithelium at the edge of the ulcer. (h) Blinded phosphor-FAK immunoreactivity scores are increased in M64HCL-treated ulcers vs. saline-treated ulcers (n ​= ​9, ∗p ​< ​0.05).

Fig. 9

M64HCl tissue distribution. (a) M64HCl concentration in tissues after a four day continuous osmotic mini-pump infusion at 30 ​mg/kg/day. (b) M64HCl levels in mouse plasma vs. urine after a 4 day infusion at 30 ​mg/kg/day (n ​= ​8, ∗p ​< ​0.05). (c) M64HCl excretion was blocked by a bilateral functional nephrectomy. M64HCl drug levels were measured by LC-MS 2 ​h after bilateral functional nephrectomy and a subcutaneous injection of 30 ​mg/kg of M64HCl. C ​= ​non-surgical controls, N ​= ​nephrectomy (n ​= ​2, ∗p ​< ​0.05). (d) M64HCl in brain vs plasma 30 ​min or 2 ​h after 5 ​mg/kg was administered intraperitoneally.

Fig. 10

The toxic IC50 for M64HCl exceeds FAK activating concentrations 10⁷-fold. (a) % live IMR-90 ​cells after 48 ​h M64HCl exposure, analyzed by crystal violet cytotoxicity assay. Lower % numbers mean cell death. (b) % cytotoxicity to max LDH release in SH-SY5Y cells, analyzed via Promega CytoTox Assay. Higher % numbers mean cell death. X-axes are split from 2 to 10 and 40–50. FAK activation threshold is labelled “FAK”, in μM below, and with a dotted red line. N ​= ​8–9 for IMR-90 ​cells, N ​= ​6–8 for SH-SY5Y cells; ∗p ​< ​0.05.