Central injection of fibroblast growth factor 1 induces sustained remission of diabetic hyperglycemia in rodents

Abstract

Type 2 diabetes (T2D) is among the most common and costly disorders worldwide. The goal of current medical management for T2D is to transiently ameliorate hyperglycemia through daily dosing of one or more antidiabetic drugs. Hypoglycemia and weight gain are common side effects of therapy, and sustained disease remission is not obtainable with nonsurgical approaches. On the basis of the potent glucose-lowering response elicited by activation of brain fibroblast growth factor (FGF) receptors, we explored the antidiabetic efficacy of centrally administered FGF1, which, unlike other FGF peptides, activates all FGF receptor subtypes. We report that a single intracerebroventricular injection of FGF1 at a dose one-tenth of that needed for antidiabetic efficacy following peripheral injection induces sustained diabetes remission in both mouse and rat models of T2D. This antidiabetic effect is not secondary to weight loss, does not increase the risk of hypoglycemia, and involves a novel and incompletely understood mechanism for increasing glucose clearance from the bloodstream. We conclude that the brain has an inherent potential to induce diabetes remission and that brain FGF receptors are potential pharmacological targets for achieving this goal.

FIGURE 1

Diabetes remission induced by a single i.c.v. FGF1 injection in ob/ob mice. (a,b) Blood glucose levels during an intraperitoneal glucose tolerance test (ipGTT) performed in fasted ob/ob (B6) mice 6 h after (a) a single i.c.v. injection of either vehicle (Veh; open symbols; n = 8) or 3 μg of mFGF1 (black symbols: n = 9), or (b) a single s.c. injection of either Veh or the same dose of mFGF1 (Veh, n = 7; FGF1, n = 6). (c) Blood glucose values from an ipGTT performed in fasted ob/ob (B6) mice either 7 d (left), 4 weeks (middle), or 18 weeks (right) following a single i.c.v. injection of mFGF1 (3 μg). (d) Time course of blood glucose levels from the same cohort of ad-libitum (ad-lib)-fed ob/ob mice both prior to and after a single i.c.v. injection of mFGF1 (3 μg). (e) Food intake (left), body weight (middle), and fat mass (right) of ob/ob (B6) mice following i.c.v. injection of either mFGF1 or Veh. (f) Daily blood glucose levels from i.c.v. Veh-injected ob/ob mice that were fed either ad-lib (n = 10) or pair-fed to a separate cohort of mice that had received i.c.v. mFGF1 (3 μg; n = 10). Data are the mean ± s.e.m. *P < 0.05, **P < 0.01, ****P < 0.0001 for group (Veh vs. FGF1) by repeated measures designs by linear mixed model analyses.

FIGURE 2

Diabetes remission induced by a single i.c.v. FGF1 injection across multiple rodent models of T2D. (a) Daily blood glucose levels from ad-lib-fed ob/ob (B6) mice following a single i.c.v. injection of either hFGF1 (3 μg; n = 6; grey symbols), mFGF1 (3 μg; n = 6; black symbols), or Veh (n = 4; open symbols). (b) Fasting blood glucose values (left) from lean, wild-type (WT) mice 6 h after i.c.v. injection of either mFGF1 (3 μg; n = 5) or Veh (n = 5), and daily blood glucose levels (right) from WT mice ad-lib-fed (standard chow) following a single i.c.v. injection of mFGF1 (3 μg; n = 8), or Veh (n = 8). (c) Daily blood glucose levels from WT ad-lib-fed DIO mice following a single i.c.v. injection of mFGF1 (3 μg; n = 8), or Veh (n = 8). (d) Blood glucose values from ad-lib-fed ob/ob (B6) mice following a single s.c. injection of either mFGF1 (0.5 mg/kg body weight; n = 11) or Veh (n = 10). (e) Fasting (before and 1.5 h after i.c.v. injection) and ad-lib-fed blood glucose levels (120 h) from ob/ob (B6) mice receiving i.c.v. injection of either FGF19 (3 μg; n = 5) or Veh (n = 5). (f) Time course of blood glucose levels from ad-lib-fed db/db mice both prior to and following a single i.c.v. injection of mFGF1 (3 μg; n = 6) or Veh (n = 9). (g) Time course of blood glucose levels from ad-lib-fed DIO WT mice rendered diabetic with a low dose of STZ (DIO-LD STZ) both prior to and following a single i.c.v. injection of mFGF1 (3 μg; n = 8) or Veh (n = 8). (h) Daily blood glucose levels from ad-lib-fed ZDF rats following a single i.c.v. injection of either rFGF1 (3 μg; n = 10; black symbols) or Veh (n = 10; open symbols). Data are the mean ± s.e.m. *P < 0.05, **P < 0.01, ****P < 0.0001 for group (Veh vs. FGF1) by repeated measures designs by linear mixed model analyses. ^#P<0.05, FGF1 (or FGF19) vs. Veh as determined by two-tailed t-test.

FIGURE 3

Effect a single i.c.v. injection of FGF1 on whole-body glucose kinetics in ob/ob mice. ob/ob (B6) mice underwent a basal glucose turnover study followed by a frequently sampled intravenous glucose tolerance test (FSIGT) 7 d after a single i.c.v. injection of mFGF1 (3 μg, black symbols; n = 13) or Veh (open symbols; n = 9). (a) Fasting blood glucose levels (left), and delta area under the glucose curve (Δ AUC) during the FSIGT (after correcting for differences of basal glucose; middle); plasma insulin levels (middle), and the acute insulin response to glucose (AIR_g) during the FSIGT (right). (b) Mean basal glucose turnover rate (GTR; left) and basal glucose clearance rate (right). (c) Insulin sensitivity (S_I; left), insulin-independent glucose disposal (S_G; right) (d) Liver glycogen content (left) and levels of mRNA encoding liver glucoregulatory genes from samples obtained at study termination (middle); basal plasma lactate levels obtained prior to the FSIGT (right). (e) Glucose clearance rate (Kg) determined from steady-state iv infusion of [2-¹⁴C]DG in tissues including skeletal muscle (left), and brain, heart, BAT, EWAT and s.c. WAT (right) in ob/ob mice 7 d after a single i.c.v. injection of mFGF1 (3 μg, n = 7) or Veh (n = 6). Data are mean ± s.e.m. *P<0.05, FGF1 vs. Veh as determined by two-tailed t-test. n.s., non-significant.

FIGURE 4

HSP25 expression in whole-mounts of the 3rd ventricle wall in response to i.c.v. FGF1. (a) Diagram illustrating the en-face whole-mount perspective and the relative location of the posterior ventral 3rd ventricle in a mid-sagittal view of the mouse brain. The square outlines the region imaged in (b) and (e). (b,e) Representative images of confocal z-stacks (b; of n = 4 images, c; of n = 4 images) spanning 80 μm from the ependymal surface deep into the parenchyma of posterior ventral 3rd ventricle from i.c.v. mFGF1- (3 μg) (b; n = 6) and i.c.v. Veh-treated (e; n = 4) WT mice. Inset letters show regions where the corresponding high power images were taken. Arrows point to a small subset of multi-ciliated ependymal cells seen in FGF1- and Veh-treated animals. m.e., median eminence. (c,f) Representative images of confocal z-stacks (c; of n = 30 images, f; of n = 12 images) of ventricular surface of 3rd ventricle showing HSP25 expression specifically following i.c.v. mFGF1 in vimentin (+) cells with elongated morphology, corresponding to tanycytes. (d,g) Representative images of confocal z-stacks (d; of n = 33 images, g; of n = 12 images) taken within the subependymal zone, showing HSP25 expression in stellate-shaped GFAP⁺ astrocytes specifically following i.c.v. mFGF1. Scale, 100 μm (a), 0.25 mm (b,e), and 10 μm (c,d,f,g).