HSV-mediated gene transfer of vascular endothelial growth factor to dorsal root ganglia prevents diabetic neuropathy

Abstract

We examined the utility of herpes simplex virus (HSV) vector-mediated gene transfer of vascular endothelial growth factor (VEGF) in a mouse model of diabetic neuropathy. A replication-incompetent HSV vector with VEGF under the control of the HSV ICP0 promoter (vector T0VEGF) was constructed. T0VEGF expressed and released VEGF from primary dorsal root ganglion (DRG) neurons in vitro, and following subcutaneous inoculation in the foot, expressed VEGF in DRG and nerve in vivo. At 2 weeks after induction of diabetes, subcutaneous inoculation of T0VEGF prevented the reduction in sensory nerve amplitude characteristic of diabetic neuropathy measured 4 weeks later, preserved autonomic function measured by pilocarpine-induced sweating, and prevented the loss of nerve fibers in the skin and reduction of neuropeptide calcitonin gene-related peptide and substance P in DRG neurons of the diabetic mice. HSV-mediated transfer of VEGF to DRG may prove useful in treatment of diabetic neuropathy.

Figure 1

Schematic representation of the replication-incompetent HSV vector T0VEGF. VEGF expression is driven by the HSV ICP0 promoter, in a recombinant defective in ICP4, ICP22, and ICP27.

Figure 2

(a) Analysis of transgene (VEGF) production in vitro by ELISA. DRG cells infected with T0VEGF at a multiplicity of infection of 1 for 1 h released 17–23 ng/ml of VEGF in supernatant collected at 24, 48, and 72 h after transfection. (b) Detection of viral genomes. Viral genome was detected from the DRG of the animals, 7 days after footpad inoculation by PCR. (c) The amount of VEGF accumulating proximal to a ligature on sciatic nerve over 24 h was substantially greater in animals inoculated 7 days earlier with T0VEGF than in control animals.

Figure 3

The amplitude of the evoked sensory response was reduced substantially in animals with diabetes, and this reduction was not affected by control vector inoculation. Inoculation of T0VEGF resulted in substantial preservation of the evoked sensory amplitude. Mean±s.e.m.; *P<0.005 compared to diabetic or diabetic inoculated with Q0ZHG; n = 8 animals per group.

Figure 4

Thermal pain sensation was evaluated with a hotplate. Diabetic animals showed sensory nerve impairment manifested by an increased latency to withdraw. Diabetic mice inoculated with T0VEGF showed near-normal thermal pain sensation. Mean±s.e.m.; * P<0.005 compared to diabetic or diabetic inoculated with Q0ZHG; n = 8 animals per group.

Figure 5

Sudomotor function after pilocarpine injection was evaluated by silicone imprint. The number of sweat droplets per sweat gland was substantially reduced in diabetic compared to control animals (illustrated top, first and second bars in graph). Diabetic animals inoculated with T0VEGF (third bar) showed substantial preservation of sweating compared to diabetic alone or Q0ZHG-inoculated diabetic mice. Mean±s.e.m.; * P<0.005 compared to diabetic or diabetic inoculated with Q0ZHG; n = 5 animals per group.

Figure 6

Nerve fibers in the footpad visualized by PGP 9.5 immunoreactivity. The nerve area, quantitated as proportional area occupied by the nerve in the footpad, was substantially reduced in diabetic animals but preserved in diabetic animals inoculated with T0VEGF. Mean±s.e.m.; * P<0.005 compared to diabetic or diabetic animals inoculated with Q0ZHG; n = 5 animals per group.

Figure 7

CGRP-immunoreactivity in dorsal horn was reduced in diabetic animals, but the area of immunoreactive terminals was substantially preserved in animals inoculated with T0VEGF compared to Q0ZHG. Mean±s.e.m.; * P<0.0001 compared to diabetic or diabetic inoculated with Q0ZHG; n = 5 animals per group.

Figure 8

SP immunoreactivity was reduced in diabetic animals compared to control (right); diabetic animals inoculated with T0VEGF showed substantial preservation of SP-IR (P<0.0001) compared to diabetic or diabetic inoculated with Q0ZHG. Mean±s.e.m.; * P<0.0001 compared to diabetic or diabetic inoculated with Q0ZHG; n = 5 animals per group.

Figure 9

Histogram of the number of blood vessels in control, diabetic, and diabetic T0VEGF-inoculated animals, by vessel cross-sectional area. The number of vessels in the diabetic animals was reduced compared to controls, but it was preserved in the vector-treated group (11.3±3.1 versus 21.0±1.1 vessels per cross-section in diabetic mice versus T0VEGF-treated mice, respectively; P<0.05).