TNF-alpha in nucleus pulposus induces sensory nerve growth: a study of the mechanism of discogenic low back pain using TNF-alpha-deficient mice.

STUDY
DESIGN: We used retrograde neurotracing with fluoro-gold to investigate the relationship between tumor necrosis factor (TNF-á) and nerve growth into the nucleus pulposus (NP) of wild-type and TNF-alpha-deficient mice.
OBJECTIVE: To clarify mechanisms underlying nerve growth into the NP and the role of TNF-á in this process. SUMMARY OF BACKGROUND DATA: Degeneration of lumbar intervertebral discs is a cause of low back pain. Pathogenesis may involve sensory nerve ingrowth into the inner layers and NP of degenerating discs. We hypothesized that TNF-á in the NP is a major inducer of nerve ingrowth and investigated this hypothesis in vivo using wild-type and TNF-á-deficient mice.
METHODS: NP was harvested at the L4/5 level from 10 wild-type and 10 TNF-deficient mice. These 20 samples of wild-type NP or TNF-deficient NP were mixed with fluoro-gold and injected into the left quadriceps muscle of 20 other wild-type mice (1 sample per mouse). Five control mice underwent sham operations in which they received similar injections of NP-free fluoro-gold into their left quadriceps muscles to detect whether neurons innervating the muscle establish contact with injected NP. Seven and 14 days after surgery, left L4 dorsal root ganglions were removed and incubated with antibodies against growth-associated protein 43 (GAP43), a marker of axonal growth. We evaluated fluoro-gold-labeled and GAP43-immunoreactive dorsal root ganglions neurons.
RESULTS: Within the set of fluoro-gold-labeled neurons, 10% were positive for GAP43 in sham-operated animals, 22% positive in the TNF-deficient NP group, and 38% positive in the wild-type NP group. These intergroup differences in the percentage of GAP43-positive neurons were statistically significant (sham vs. TNF-deficient NP group: P = 0.009; TNF-deficient NP group vs wild-type NP group: P = 0.026). CONCLUSION.: The percentage of fluoro-gold-labeled GAP43-immunoreactive neurons significantly increased after injections of NP harvested from both mouse types. Furthermore, the percentage of GAP43-immunoreactive neurons was significantly higher in mice receiving wild-type NP compared with mice receiving TNF-deficient NP. These findings suggest that TNF-alpha acts as an inducer of axonal growth into degenerated discs, as evidenced by decreased GAP-43 immunoreactivity in mice receiving TNF-deficient NP injections and even lower GAP-43 immunoreactivity in control mice receiving NP-free fluoro-gold injections.