Erythropoietin is neuroprotective in a rat hemisection model of spinal cord injury

V.R. King, L. Torup#, S. A. Averill, J.V. Priestley and A.T. Michael-Titus Neuroscience Centre, Institute of Cell and Molecular Science, St. Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary, University of London, 4 Newark Street, Whitechapel, London E1 2AT, Great Britain #H. Lundbeck A/S, Ottiliavej 9, 2500 Valby, Denmark

The cytokine erythropoietin (EPO) has been shown to be neuroprotective in several disorders, including stroke, peripheral nerve injury and spinal cord compression and contusion (Gorio et al., 2002; Campana and Myers, 2003). In this study we investigated the effects of EPO in a hemisection model of spinal cord injury. This model is well-suited not only for the exploration of neuroprotection, but also for the investigation of neuroregeneration. Experiments were carried out in adult male Wistar rats (150-200 g). Animals received a spinal cord hemisection under isoflurane anaesthesia at mid-thoracic level (T10-T11 vertebrae), followed by an intraperitoneal injection of vehicle (saline) (n=6 rats) or EPO (40 μg/kg) 30 minutes and 24 hours following injury (n=6 rats). Three days after injury, animals were anaesthetized and perfused with 4% paraformaldehyde. The tissue was post-fixed and cryoprotected, and horizontal sections (12 μm) were processed for immunohistochemistry to label macrophages and astrocytes, using ED1 (1:1000; Serotec, UK) and glial fibrillary acidic protein (GFAP) (1:1000; Chemicon, UK) antibodies, to assess the inflammatory reaction and the lesion size, respectively. Cell death was measured using the TUNEL labelling method (DeadEnd Colorimetric TUNEL kit; Promega, UK), to detect apoptotic nuclei. In addition, 15 μm sections taken from the upper lumbar spinal cord of two uninjured control animals (perfused as described above) were processed for immunohistochemistry using a rabbit antibody to the EPO receptor (1:500; Upstate Biotechnology Inc., USA). Results were expressed as means ± s.e. mean. Lesion size was analysed using Students’s t-test. A two-way ANOVA for position (within the lesion vs. outside the lesion) x treatment (saline vs. EPO) was used to analyse counts of TUNEL-positive nuclei and ED1 labelling. Intense labelling for the EPO receptor was detected in the spinal cord both in fibres and in neuronal and glial cells. GFAP staining showed that treatment with EPO resulted in significantly smaller lesions compared to vehicle treatment (-29%; p<0.05). The number of TUNEL-labelled nuclei was counted within the lesion site and in the intact adjacent tissue. Results indicated that, although there was no difference between EPO- and vehicle-treated animals in the number of TUNEL-labelled nuclei within the lesion site, EPO treatment significantly reduced TUNEL-labelling in the intact tissue adjacent to the lesion, compared to control animals (-31%; p<0.05). EPO- and vehicle-treated animals did not differ significantly in the degree of macrophage infiltration within the lesion site, or in the adjacent intact tissue. ED1 immunoreactivity in the intact tissue in EPO-treated animals decreased, compared to vehicle-injected animals, although the effect did not reach statistical significance. Our data show that EPO reduces significantly the size of the lesion after injury, and also the cell death, and thus confers marked neuroprotection after this type of spinal cord trauma. The effects may be mediated by the EPO receptor. These results are comparable to the observations reported after spinal cord compression or contusion (Gorio et al. 2002) and extend the neuroprotective profile of this cytokine to a new type of injury. Future work will explore the neuroregenerative potential of this cytokine after spinal cord hemisection.

Gorio, A et al. (2002) Proc. Natl.Acad. Sci., 99, 9450-9455.
Campana, W.M., Myers, R.R. (2003) Eur. J. Neurosci., 18, 1497-1506.