Human and rodent spinal cords differ in size and the anatomical location of the circuits.The functional recovery pattern following spinal cord injury (SCI) is also vastly different between them.Currently,most SCI experiments on mammals use rodent models,but that data is not applicable to the clinical situation.Old World monkeys are the closest living animal relatives of humans,and would likely provide better screening model information than rodent experiments.Verification of rat model results using non-human primates would provide additional confidence and would be more readily translatable to clinical applications.However,compared with the development of rodent model,monkey models of SCI lag behind,still using century-old methods to injury the spinal cord without vertebral stabilization.We introduced a new SCI method in non-human primates,with the level of tissue displacement as the injury determinant used on stabilized animals.Among the eight adult male monkeys included,two rhesus monkeys were used to test the safety of the spine stabilizer for targeting spinal cord exposition without injury,two crab-eating monkeys had 1.0 and 1.5 mm deep spinal cord contusions at T9,and four rhesus monkeys had 1.0 or 1.5 mm deep spinal cord contusion at T9.The functions of the animals were evaluated electrophysiologically by magnetic stimulation-induced motor-evoked potentials (MEP) and posterior tibial nerve stimulation-induced cortical somatosensory evoked potentials (SSEP),and behaviorally by a novel walking corridor test and in-cage hind-limber digital function observations.The injury was also morphologically characterized by quantitative histological assays: hematoxylin and eosin,luxol fast blue,glial fibrillary acidic protein (GFAP),and SMI-31.We demonstrated that our vertebral stabilizing technique was a safe way to secure the target vertebra,allowing for improved surgical manipulation,spinal cord exposure,and injury execution.The contusion was produced using a plunger to hit the spinal cord via a rod that rested on the dura.The push-off cerebral spinal fluid in the subarachnoidal space helped to increase the injury accuracy.Both 1.0 mm and 1.5 mm tissue displacement produced mild SCI in the monkeys with distinct neurological deficits.Fortunately,these selected levels of injury did not induce the need for assisted urination.However,the additional 0.5 mm increment of displacement in the 1.5 mm depths noticeably increased the injury and tissue damage.The central injury produced bilateral symmetric paralysis.The paracentral contusion also mimicked the Brown-squirrel syndrome in the clinic that has never been reproduced in rodent models.Both SSEP and MEP were affected by the mild SCI,though MEPs were more useful to depict the function loss of the animals.The walking corridor test was a safe tool to effectively determine the deficits of locomotor function in the untrained animals.The results from the electrophysiological and behavioral assays were closely correlated.Histological results also showed that at the epicenter,cavities of volume 3.65 * 109 urn3 in the 1.0 mm and 5.46 x 1 (f um3 in the 1.5 mm groups appeared after injury.Compared to sham,the cross sectional area was reduced by 31.08% and 46.65%,while the myelination was reduced by 52.33% and 67.10% in the 1.0 mm and 1.5 mm groups,respectively.The numbers of SMI-31-IR axons were markedly decreased either at the lesion border or in the peripheral white matter in both injury groups.The percentage of GFAP-IR astrocytes decreased at the lesion border in both injured groups,though they were increased in the periphery of the white matter in both injury groups.These histological results support the behavioral assays.The importance of infrequent SCI using non-human primates is clear.Our results suggest that SCI models using non-human primates with vertebral stabilization are useful for evaluating repair strategies before they are translated to clinical trials for human SCI.