Engineered zinc finger protein-mediated VEGF-a activation restores deficient VEGF-a in sensory neurons in experimental diabetes

Abstract

Objective: The objectives of the study were to evaluate retrograde axonal transport of vascular endothelial growth factor A (VEGF-A) protein to sensory neurons after intramuscular administration of an engineered zinc finger protein activator of endogenous VEGF-A (VZ+434) in an experimental model of diabetes, and to characterize the VEGF-A target neurons.

Research design and methods: We compared the expression of VEGF-A in lumbar (L)4/5 dorsal root ganglia (DRG) of control rats and VZ+434-treated and untreated streptozotocin (STZ)-induced diabetic rats. In addition, axonal transport of VEGF-A, activation of signal transduction pathways in the DRG, and mechanical sensitivity were assessed.

Results: VEGF-A immunoreactivity (IR) was detected in small- to medium-diameter neurons in DRG of control rats. Fewer VEGF-A-IR neurons were observed in DRG from STZ-induced diabetic rats; this decrease was confirmed and quantified by Western blotting. VZ+434 administration resulted in a significant increase in VEGF-A protein expression in ipsilateral DRG, 24 h after injection. VEGF-A was axonally transported to the DRG via the sciatic nerve. VZ+434 administration resulted in significant activation of AKT in the ipsilateral DRG by 48 h that was sustained for 1 week after injection. VZ+434 protected against mechanical allodynia 8 weeks after STZ injection.

Conclusions: Intramuscular administration of VZ+434 increases VEGF-A protein levels in L4/5 DRG, correcting the deficit observed after induction of diabetes, and protects against mechanical allodynia. Elevated VEGF-A levels result from retrograde axonal transport and are associated with altered signal transduction, via the phosphatidylinositol 3'-kinase pathway. These data support a neuroprotective role for VEGF-A in the therapeutic actions of VZ+434 and suggest a mechanism by which VEGF-A exerts this activity.

FIG. 1.

VEGF-A is downregulated in the DRG of STZ-induced diabetic rats. Western blot analysis of VEGF-A levels in pooled L4/5 DRG obtained from control (C) rats or diabetic rats 3 weeks after STZ injection (D) show a dramatic decrease in VEGF-A in diabetes (A). Densitometric analysis of VEGF-A protein levels (normalized to total ERK to correct for protein loading) shows a significant decrease after STZ injection (***P < 0.0001, t test, n = 5 per group, B). Representative micrographs show that VEGF-A-IR is present in sensory neurons in L4/5 DRG. A population of small- to medium-diameter neurons in DRG from control rat DRG shows VEGF-A-IR (arrows, C and E). Fewer VEGF-A-IR neurons were observed in DRG obtained from STZ-induced diabetic rats (D and E). Data for individual animals (n = 4 per group, **P < 0.01, t test) are shown as circles and the mean value is shown as a horizontal bar (E). Scale bar, 50 μm.

FIG. 2.

Intramuscular administration of VZ+434 increases VEGF-A in the ipsilateral DRG of STZ-induced diabetic rats. Diabetic rats (3 weeks after STZ injection) received two intramuscular injections of the VEGF-A–activating zinc finger protein-transcription factor VZ+434 (250 μg total) into the gastrocnemius muscle and were killed at the indicated time points after injection (n = 4 per time point). VEGF-A levels in right (uninjected, contralateral limb) and left (injected, ipsilateral limb) were assessed using Western blotting (A) and quantified using densitometry (normalized to total AKT to correct for protein loading; B). Intramuscular VZ+434 caused a rapid and significant increase in VEGF-A protein expression in ipsilateral L4/5 DRG compared with contralateral by 24 h after injection (A and B, *P < 0.05, **P < 0.01 paired t test). Increased VEGF-A-IR (C and D, arrows) was clearly detected within a population of small- to medium-diameter sensory neurons in the ipsilateral DRG (left, D) compared with the contralateral DRG (right, C). Cell size distribution analysis shows a significant increase in VEGF-A-IR neurons in small-diameter neurons in the ipsilateral (F, *P < 0.05, ANOVA) compared with contralateral DRG (E). Scale bar, 50 μm.

FIG. 3.

Phenotypic characterization of VEGF-A-IR neurons in L4/5 DRG of diabetic rats after VZ+434. Representative micrographs show that VEGF-A-IR is present in sensory neurons in L4/5 DRG. Sections have been dual labeled with (A, C, and E) anti–VEGF-A and anti-CGRP (B); IB4 (D); and anti-NF200 (F). A large proportion of neurons that express VEGF-A-IR (A and C, arrows) coexpress the phenotypic markers CGRP (B, arrows) and IB4 (D, arrows). Note, very few VEGF-A-IR neurons colocalize with NF200 (E and F, asterisk). Scale bar, 50 μm.

FIG. 4.

Expression of VZ+434 DNA plasmid copy number in ipsilateral and contralateral tissues. At 24 h after unilateral intramuscular injection of VZ+434 (250 μg), tissues were harvested and DNA was extracted. The background PCN of VZ+434 was set at 2,700 (derived from mean + 2 SD from contralateral tissue samples). Data from individual animals are shown as circles, with the mean as a horizontal bar (n = 8 for muscle and n = 4 for DRG). Ipsilateral muscle contained significantly more copies of VZ+434 per 100 ng DNA than contralateral muscle, whereas there was no significant left-right tissue difference in VZ+434 in DRG (*P < 0.05, Mann-Whitney U test).

FIG. 5.

VEGF-A undergoes axonal transport in the sciatic nerve in vivo. Adult rats received a unilateral sciatic nerve ligature at mid-thigh level, proximal to the site of VZ+434 injection (250 μg, intramuscularly) Longitudinal sections of sciatic nerve, 24 h after ligation, were immunostained for VEGF-A (red) and S100 to label Schwann cells (green; A–G). VEGF-A-IR can clearly be seen to accumulate within axons (A, arrows) at the ligature site, proximal and distal to the ligature site, indicating bidirectional axonal transport. A region highlighted by a rectangle in (A) is shown at increased magnification in (B). VEGF-A-IR was restricted to axons and not Schwann cells (B and C). Little VEGF-A-IR was observed in unligated contralateral nerve (D) or untreated ligated nerve (E). In contrast, VEGF-IR accumulated both proximal and distal to ligatures in both untreated (F) or treated (G) control rats. Scale bars = 1 mm. Ligated sciatic nerves from VZ+434-injected STZ-induced diabetic rats (I) or control rats (J) were cut into 5-mm segments proximal and distal to the ligature and samples processed for Western blotting to assess VEGF-A levels (see schematic diagram, H). VEGF-A accumulated both proximal and particularly distal to the ligature in VZ+434-injected diabetic rats (I). Samples prepared from control rats injected with VZ+434 showed the greatest accumulation of VEGF in proximal samples (J). Data for individual animals are shown as circles; the mean value is shown as a horizontal bar (I and J). (A high-quality digital representation of this figure is available in the online issue.)

FIG. 6.

AKT is phosphorylated in ipsilateral L4/5 DRG after VZ+434 injection. Diabetic rats (3 weeks after STZ injection) treated with VZ+434 were killed at the indicated time points after injection (24 h, 48 h, 1 week, 2 weeks; n = 4 per time point). Phosphorylated (p-AKT) and total (t-AKT) AKT levels in right (uninjected, contralateral limb) and left (injected, ipsilateral limb) DRG were assessed using Western blotting (A) and quantified using densitometry (B). Intramuscular administration of VZ+434 caused a rapid and significant ipsilateral increase in pAKT levels in L4/5 DRG (A and B) by 48 h after injection (*P < 0.05) and remained significantly elevated for 1 week (***P < 0.001). Immunocytochemical analysis with antibodies against VEGF-A and pAKT showed colabeling of VEGF-A-IR (C) and pAKT (D) neurons in the DRG (arrows), however, pAKT-IR neurons that did not colocalize with VEGF-IR neurons were also observed (asterisk). Reduced pAKT-IR was observed in contralateral DRG (E). Scale bar, 50 μm.

FIG. 7.

VZ+434 protects against mechanical hyperalgesia. Sensory testing was conducted prior to treatment (baseline measures) and at 4, 8, and 12 weeks after STZ injection. VZ+434-treated rats received unilateral intramuscular injections at 4, 6, and 8 weeks after STZ injection. At the 4- and 8-week time points, mechanical thresholds were assessed 3 days after VZ+434. At 8 weeks after STZ injection, a significant decrease in the paw withdrawal force (PWF) was observed in STZ-induced diabetic rats and in the contralateral hind paw of VZ+434-treated STZ-induced diabetic rats compared with their baseline measures (A, mean + SD, *P < 0.05, n = 7–9, two-way ANOVA, Bonferroni multiple comparison test). At 8 weeks, STZ-induced diabetic rats (B) and the contralateral hind paw of VZ+434-treated rats (C) compared with nondiabetic controls (B and C). VZ+434 treatment (D) provided significant protection against this allodynia in the ipsilateral limb (B–D, data shown for individual animals *P < 0.05; **P < 0.01, n = 7–9, one-way ANOVA, Dunnett post hoc test).