ObjectiveTo investigate morphological and functional neural network alterations in the primary visual cortex （V1） during early-stage diabetes. Methods32 SPF male C57BL/6J mice were randomly divided into control and diabetic groups（n=16）. Type 1 diabetes was induced by intraperitoneal injection of streptozotocin （STZ）， with model validation via oral glucose tolerance test （OGTT） at week 4. Nissl staining was used to assess morphological changes in brain slices of the two groups. Immunohistochemistry was performed to detect the expression of calcium/calmodulin-dependent protein kinase Ⅱ （CaMKⅡ） and somatostatin （SST） in the V1 area. The co-localization density of SST-positive signals on CaMKⅡ+ neurons was then quantified by ImageJ software. In vivo electrophysiology was used for studying neuronal firing and functional connectivity. ResultsCompared with the control mice， nissl staining showed diabetic mice exhibited significant V1 thinning （P=0.02）， reduced neuronal density （P=0.01）. Immunohistochemistry found decreased SST integral optical density （P=0.02） and elevated mean fluorescence intensity （P=0.01）， and lower SST density on CaMKⅡ+ neurons in the diabetic group than in the control group. Compared with the control mice， in vivo electrophysiology studies indicated that diabetic mice showed reduced action potential peak-to-trough ratio and the absolute values of slope （P<0.01）， as well as impaired functional connectivity， including conditional firing， mutual information， and Granger causality （P<0.000 1）. ConclusionEarly diabetes induces V1 structural atrophy accompanied by SSTergic system disruption and multi-level functional degradation from single neurons to neural networks.

Objective:To investigate the damage of visual function in the early stage of diabetes mellitus and the response characteristics of primary visual cortex neurons in diabetic mice.Methods:Twenty 7-week-old SPF grade male C57BL/6J mice were randomly divided into diabetes group and normal control group by the random number table method, with 10 mice in each group.The diabetes model in diabetes group was established by the intraperitoneal injection of streptozotocin.Fasting blood glucose concentration and body mass of mice were measured before and 1, 2 and 3 weeks after modeling, and intraperitoneal glucose tolerance test was performed 4 weeks after modeling to evaluate the establishment of diabetes model.At 4 weeks after modeling, the electroretinogram (ERG) responses of mice were recorded at dark adaptation luminances of 0.01 and 3.0 cd·s/m 2, and the ERG response to light adaptation luminance of 3.0 cd·s/m 2 was recorded 10 minutes after light adaptation to evaluate the retinal function of mice.The fundus of mice was examined with an ultra-wide field laser scanning ophthalmoscope.The visual function of mice was evaluated via the cutoff frequency of grating discrimination detected by visual water maze.The spatial frequency tuning curves of primary visual cortex (V1) neurons were detected by in vivo electrophysiology technique and the maximum firing intensity, self-firing intensity, optimal spatial frequency, cutoff frequency and bandwidth of neurons were calculated to evaluate the neuronal function of mice.The research scheme was approved by the Experimental Animal Ethics Committee of Anhui Medical University (No.LLSC20230419). Results:The diabetic model was successfully established in 10 mice in the diabetes group.At 4 weeks after modeling, compared with the normal control group, the b-wave amplitudes of mice at a dark adaptation luminance of 0.01 cd·s/m 2, the a- and b-wave amplitudes of mice at a dark adaptation luminance of 3.0 cd·s/m 2 and the b-wave amplitudes of mice at a light adaptation luminance of 3.0 cd·s/m 2 after 10 minutes of light adaptation showed a downward trend in diabetes group, but the differences were not statistically significant (all at P>0.05).Ultra-wide field laser scanning ophthalmoscopy showed no obvious vascular changes in the retina of diabetic mice.The results of visual water maze detection showed that the cutoff frequency of diabetes group was (0.45±0.06)c/d, which was significantly lower than (0.58±0.05)c/d of normal control group ( t=5.10, P<0.05). In vivo electrophysiological results showed that the maximum firing intensity of neurons in V1 region in diabetes group 4.29(2.60, 8.33)spikes/second, which was significantly lower than 7.10(4.34, 11.6)spikes/second in normal control group( Z=-4.29, P<0.05).The optimal spatial frequency, cutoff spatial frequency and bandwidth were 0.03(0.02, 0.05), 0.07(0.05, 0.12) and 0.14(0.07, 0.22)c/d in diabetes group, which were significantly lower than 0.41(0.03, 0.05) and 0.10(0.07, 0.14), 0.14(0.10, 0.26)c/d of normal control group, and the differences were statistically significant ( Z=-3.22, -3.19, -2.19; all at P<0.05). Conclusions:The abnormal visual function may occur in the early stage of diabetes before the appearance of retinal vasculopathy, which is related to the damage of neurons in the V1 region.