Neutralizing intraspinal nerve growth factor blocks autonomic dysreflexia caused by spinal cord injury

Abstract

Autonomic dysreflexia is a condition that develops after spinal cord injury in which potentially life-threatening episodic hypertension is triggered by stimulation of sensory nerves in the body below the site of injury. Central sprouting of small-diameter primary afferent fibers in the dorsal horn of the spinal cord occurs concurrently with the development of this condition. We propose a model for the development of autonomic dysreflexia in which increased nerve growth factor (NGF) in the injured cord stimulates small-diameter primary afferent fiber sprouting, thereby magnifying spinal sympathetic reflexes and promoting dysreflexia. We identified this population of afferent neurons using immunocytochemistry for calcitonin gene-related peptide. Blocking intraspinal NGF with an intrathecally-delivered neutralizing antibody to NGF prevented small-diameter afferent sprouting in rats 2 weeks after a high thoracic spinal cord transection. In the same rats, this anti-NGF antibody treatment significantly decreased (by 43%) the hypertension induced by colon stimulation. The extent of small-diameter afferent sprouting after cord transection correlated significantly with the magnitude of increases in arterial pressure during the autonomic dysreflexia. Neutralizing NGF in the spinal cord is a promising strategy to minimize the life-threatening autonomic dysreflexia that develops after spinal cord injury.

Fig. 1.

PC12 cell neurite response assay. PC12 cells were cultured for 48 hr without NGF treatment (a) or with 10 ng/ml NGF (b), with 10 ng/ml NGF plus anti-NGF Ab (1:1000; c) or with 10 ng/ml NGF plus anti-NGF Ab (1:2000; d).

Fig. 2.

Digital photomicrographs of horizontal sections of spinal cord at the injury site in the thoracic segment immunostained using the anti-NGF Ab. Sections were taken from a rat 7 d after SCT at T4. Panels a and c depict NGF-IR macrophages at the transection site. Panels b andd illustrate NGF–IR Schwann cells at the dorsal root entry zone subpial rim and near an arteriole. The boxeson panels a and b delineate the region depicted in panels c and d, respectively. Scale bars: a, b, 100 μm;c, d, 50 μm. Arrowspoint to examples of immunoreactive cells.

Fig. 3.

Digital photomicrographs demonstrating that the neutralizing antibody (anti-NGF Ab) penetrated the spinal cord after intrathecal administration of the anti-NGF Ab to a cord-injured rat for 2 weeks. The spinal cord had been transected at the fourth thoracic segment (T4). The presence of the anti-NGF Ab (made in rabbit) was visualized by fluorescence immunocytochemistry using an anti-rabbit antibody. Immunoreactivity to the anti-NGF Ab in the dorsal two-thirds of a cord section at T12 of the injured rat is shown in panela. Lack of immunoreactivity in T12 of an intact, untreated rat (b) and in the cerebellum of the anti-NGF Ab-treated rat (c) demonstrates that the labeling is specific. Scale bar, 100 μm (refers to all panels).

Fig. 4.

Digital photomicrographs of CGRP-IR fibers in the dorsal horn of control intact rats (a, d), rats 2 weeks after cord injury that received an intrathecal infusion of nonimmune rabbit IgG (b,e), and rats 2 weeks after cord injury that received intrathecal infusion of anti-NGF Ab (c, f). Tissue sections are from T6. Panelsd–f are magnifications of laminae III-V from panelsa–c. The boxes on panelsa–c indicate the regions shown in the higher magnification photomicrographs. Scale bars: a–c, 100 μm; d—f, 50 μm.

Fig. 5.

Mean areas of CGRP-IR fibers in laminae III–IV of the thoracic (T) and lumbar (L) spinal segments in rats with no spinal cord injury (intact) and in rats 2 weeks after spinal cord transection (SCT). After SCT, rats received nonimmune rabbit IgG or anti-NGF Ab intrathecally. The CGRP-IR areas in rats treated with rabbit IgG increased significantly (asterisk) compared with intact controls; the CGRP-IR areas in rats treated with anti-NGF Ab were significantly smaller than areas in the same cord segments of rats treated with rabbit IgG (open diamond); the CGRP-IR areas in T1–3 and L1–5 of some spinal rats were larger than the area in T6–9 within the same treatment group (open cross). Each treatment group contained four rats. In the rabbit IgG–treated group, samples could not be obtained for T10-L5 in one rat because of catheter-induced damage to the cord.

Fig. 6.

Pulsatile arterial pressure (AP), mean arterial pressure (MAP), and heart rate (HR) measurements in two rats 2 weeks after cord transection that received nonimmune IgG or anti-NGF Ab. Colon distension for 1 min (between arrows) stimulated an increase in AP and MAP and a decrease in HR.