Fig 2: Growth factor deprivation-mediated increase in IGF-1R and INSR is not regulated by phosphorylation of the homodimeric INSR. (A) hTCEpi cells were cultured in growth or basal media and treated with 1, 10, 100, 1000 nmol/l of insulin. At the physiological concentrations (1 nmol/l), treatment with insulin reduced INSR expression but increased IGF-1R expression levels. Treatments with increasing concentrations of insulin from 10 nmol/l up to 1000 nmol/l showed a decrease in both IGF-1R and INSR levels. (B) hTCEpi cells were cultured in growth or basal media. For cells in basal media, cells were treated with the insulin mimic MA-20 or an irrelevant IgG control. Treatment with MA-20 did not induce phosphorylation of INSR/IGF-1R, but did reduce expression of IGF-1R and INSR back to growth levels.

Fig 3: FGF19-activated Src kinase mediates Y67-FXR phosphorylation. a PMH from FXR-KO were transfected with siRNA for Src, ALK, JAK2, and INSR and infected with Ad-flag-FXR, and then the cells were treated with 50 ng ml-1 FGF19 for 10 min. Levels of indicated proteins were detected by IB. b PMH were treated with vehicle or FGF19 for the indicated times, and activated p-Y416-Src levels were determined by IB. c Mice were fed a ND or 0.5% CA chow for 3 h, and the interaction of FXR with Src was detected by CoIP (n = 3). d Fragments of FXR (left) and c-Src (right) that were fused to GST (top). Binding of Src or flag-FXR to GST-FXR or GST-Src fusion proteins, respectively, was detected by IB. e Flag-FXR-WT and siRNA for Src were expressed in hepatocytes which were treated with FGF19 for 10 min. p-Y-FXR and p-Y67-FXR levels were detected by IP/IB. Relative p-Y-FXR levels are shown on the bottom (n = 3). f C57BL/6 mice were infected with lenti-shRNA for Src or control RNA for 1 month, and mice were briefly fed a 0.5% CA chow, and then, p-Y67-FXR levels in liver extracts were detected by IB (n = 2 mice). g Flag-FXR-WT or Y67F-FXR was expressed in HepG2 cells, purified FXR proteins were incubated with ATP and recombinant Src (ab60884, Abcam) as indicated, and p-Y67-FXR levels were detected by IB. h Liver sections from C57BL/6 mice or FGF15-KO mice that had been fasted (Fs) or refed (Fd) were performed by IHC using the p-Y416-Src antibody. Scale bar, 50 µm. i–m C57BL/6 mice were infected with lentivirus expressing shRNA for Src for 2 weeks and treated for 48 h with ANIT or vehicle. i Experimental outline. j Protein levels of Src detected by IB. k Serum samples. l Liver sections stained with H&E or F4/80 antibody. Scale bar, 300 µm (H&E) and 100 µm (F4/80). m Levels of the hepatic and serum BAs and total bilirubin, AST, and ALT in serum. All values are presented as mean ± SD. Statistical significance was measured using the e one- or m two-way ANOVA with the Bonferroni post-test. *P < 0.05, **P < 0.01, and NS, statistically not significant

Fig 4: Schematic of the proposed hypothesis. (A) Insulin is secreted by the healthy lacrimal gland and maintains insulin homeostasis in the precorneal tear film. The presence of insulin downregulates INSR and IGF-1R through the homodimeric INSR to maintain the corneal epithelium. (B) In the presence of hyperglycemia, inflammation and oxidative stress, there is a reduction in insulin secretion as a result of a loss of corneal sensory nerves combined with direct damage to the lacrimal gland. The absence of adequate insulin levels results in an upregulation of INSR, IGF-1R and Hybrid-R and the subsequent accumulation of Hybrid-R in the corneal epithelial cell nucleus. In the nucleus, Hybrid-R directly binds DNA and functions to modulate the expression of genes involved in regulating cell growth and survival. Together, growth impairment and loss of neurotropic support leads to thinning of the corneal epithelium and abnormal wound healing.

Fig 5: Nuclear accumulation of the IGF-1/insulin Hybrid R occurs de novo in response to growth factor deprivation. (A) hTCEpi cells were cultured in growth, basal or basal media supplemented with 1 μg/ml cycloheximide (Cx) overnight. Cells were lysed and fractionated into cytosolic, membrane, soluble and insoluble nuclear fractions. Consistent with whole cell lysates, immunoblotting for IGF-1R showed an increase in IGF-1R expression in basal media compared to controls. Treatment with cycloheximide blunted the basal increase and reduced expression levels to below those seen in growth media. There was a large increase in nuclear accumulation of IGF-1R in both the soluble and insoluble fractions. Similar to IGF-1R, INSR was also increased in basal media. Cycloheximide treatment of samples in basal media blocked nuclear accumulation of INSR. Immunoblotting for GAPDH, Calnexin, SP1 and Histone H3 were used for cytosolic, membrane, soluble nucleus, and insoluble nucleus fractionation controls. (B) hTCEpi cells were cultured overnight in basal or growth media and fractionated into nuclear and non-nuclear (cytosolic) components. Samples were immunoprecipated for IGF-1R and immunoblotted for INSR and IGF-1R to confirm the presence of nuclear Hybrid-R. There was an increase in Hybrid-R expression in the nuclear compartment in basal media. GAPDH and SP1 were used for cytosolic and nuclear fraction contamination controls. (C) Immunofluorescence for IGF-1R (green) and nuclei (red) shows an increase in IGF-1R expression and nuclear accumulation in basal media compared to growth media.