Objective To observe effects of glial cell-derived neurotrophic factor(GDNF)target muscle injection on functional recovery,neuromorphology and GDNF expression in central facial neurons in rats after crushed facial nerve injury,so as to investigate the feasibility and mechanism of action of GDNF target muscle injection in the treatment of peripheral facial paralysis.Methods SD rats were randomly divided into sham-operated group(right facial nerve trunk exposed only),model group(facial nerve trunk pressed),experimental control group(facial nerve trunk pressed+buccal muscle injected with saline)and experimental group(facial nerve trunk pressed+buccal muscle injected with GDNF).The neurological recovery of rats was observed by buccal muscle electrophysiology,animal facial palsy score,morphological changes of buccal muscle fibers by Masson staining,morphological changes of facial nerve by toluidine blue staining and GDNF expression in facial neurons by Western Blot.Results ①Immediately after modelling surgery,SD rats of the sham-operated group showed no facial paralysis with scores of 0,while peripheral facial paralysis appeared in those of the other groups with average scores of 5.Twenty-eight days after surgery,the rats in the experimental group had basically recovered from facial paralysis while those in the model and control groups had improved to different degrees but not fully recovered from facial palsy,with scores of more than 3.②Buccal muscle electrophysiology revealed that the peak latency was prolonged to different degrees and the maximum amplitude decreased to different degrees in each group compared with the sham-operated group.The peak latency was significantly shortened(P<0.05)and the maximum amplitude increased(P<0.05)over time in the experimental group compared with the model group and the experimental control group.③Toluidine blue staining showed irregular nerve fiber morphology,discontinuous and unclear outer membrane and reduced number of axons after surgery in the model group,the experimental control group and the experimental group.Twenty-eight days after surgery,the facial nerve morphology of the experimental group was not significantly different from that of the sham-operated group,and was significantly recovered compared with the model group and the experimental control group.④Masson staining demonstrated decreased number of muscle fibers and areas occupied by muscle tissue in each group after modelling.The morphology of muscle fibers in the experimental group recovered faster than those of the model group and the experimental control group,and were close to normal by 28 days postoperatively(P<0.05).⑤Western Blot showed decreased expression of GDNF protein in central facial neurons of each group after modelling,and gradual increased expression during the observation period.Compared with the model group and the experimental control group,the protein expression of GDNF was significantly enhanced in the experimental group at each time point after surgery(P<0.01).Conclusion Target muscle injection of GDNF can be used as one of the effective modes of administration to improve peripheral nerve injury,promot nerve fibre repair,recover target muscle function and repair of injured peripheral nerves by central nervous system.