| Perceptual learning refers to substantial and long-lasting improvements in discrimination and detection abilities through repetitive training or practice, which reflects plasticity in the adult visual system. In the visual system, the improved visual functions include visual acuity (VA), contrast sensitivity (CS), position perception and contours recognition. Visual perceptual learning (VPL) is increasingly used to improve visual functions of adult amblyopia. The relevance of VPL and adult cortical plasticity has been investigated in psychophysical and imaging studies, however, where VPL occurs and how it enhances cortical plasticity are still unclear. VPL may occur in the early visual areas, including the primary visual cortex (Ⅵ), and also in the higher cortical areas. Several neural network models on VPL propose that the excitatory-inhibitory balance is essential for VPL, which improves overall perceptual performance through signalenhancement, noise reduction, or both. The decrease of cortical inhibition facilitates the recovery of ocular dominance plasticity in adult visual cortex, however, the relevance of cortical inhibition and VPL-induced cortical plasticity is unclear.In this study, by mimicking the similar training protocol used in human, we built a new mice model of VPL:(1) Following 35 days’ near contrast threshold training in grating orientation discrimination task, mice showed 85% improvement in CS at the trained spatial frequency (SF) and 50% improvement in the untrained VA.(2) VA was also improved by more than 50% with near SF threshold training. The perceptual training induced a general improvement in the untrained CS functions.Thelearning exhibited interocular transfer.(3) The learning effect of training in grating discrimination task completely transferred to that in grating detection task, not vice versa; interestingly, the learning effect of both tasks failed to transfer to reverse task, respectively.(4) VPL completely recovered VA in adult amblyopia mice. Combining intrinsic optical imaging, morphology and cortical lesions experiments, we investigated the neural mechanisms underlying VPL(1) Intrinsicoptical imaging showed that cut-off SFs of V1 was increased 41% in VPL mice than those in control and naive mice. The cut-off SFs in V1 highly correlated with the behavioral VA in mice.(2) The dendriticspine density in layer 2-3 neurons of V1 in VPL mice was significantly higher than that in control mice; these results demonstrated that VPL induced functional and structural changes in V1.(3) Lesions of medial prefrontal cortex (mPFC) partially impaired VPL, suggesting that high areas also participated in the VPL proceeding. Lastly, using western blotting, pharmacological and optogenetic approaches, we studied the relevance of cortical plasticity, cortical GABAergic inhibition and VPL:(1) MK801 dose-dependently impaired the daily improvement, suggesting that this VPL was NMDA-dependent cortical plasticity;(2) The protein levels of GABA synthestic enzymes, GAD65 and GAD67 were reduced by the perceptual training, but not by swimming alone;(3) Intraperitoneal administration of GABAergic inhibition diazepam reversibly impaired the induction of VPL;(4) Local activation of inhibition neurons in the bilateral V1 of VGAT-ChR2-EYFP mice by laser during trainingalso reversibly inhibition the VA improvement.These results suggested that the induction of VPL is modulated by GABAergic inhibition in V1, which may alter the excitation and inhibition balance. These changes may promote adult cortical plasticity, which finally contributed the improvement of overall perceptual performance.Themouse model of VPL comining molecular experiment, transgenetic manipulation and optogenetic technology,will further provide a framework for VPL research to reveal the underlying neural and molecular mechanisms, which will lead to an increased understandingof plasticity in the adult visual system and provide theoretical basis for the application of VPL in adult amblyopia. |
PDF Full Text Request