周围神经损伤
-
Figure 1｜Effect of metformin on the survival of SCI mice.
Following SCI, the survival of injured mice in three groups (control, 0 dpi and 3 dpi groups) was recorded daily. No death occurred in any of the three groups on the first day post-injury. At 3 days after SCI, the mortality rates were 20% in the control group, 0% in the 0 dpi group, and 10% in the 3 dpi group. There were additional deaths in the following weeks in each of the groups, but the mortality rates were stabilized by 10 dpi, i.e. 50% in the control group, 30% in the 0 dpi group, and 20% in the 3 dpi group (Figure 1). No further death was observed in each group after 14 dpi. The difference in survival rates of control animals between the pilot study (10%) and the later experimental study (50%) was likely due to different batches of animals. While the nonsignificant difference in mortality among different treatment groups indicated that the use of 100 mg/kg metformin in the subacute phase of severe SCI tended to reduce mortality.
Figure 2｜Effect of metformin on the hind limb motor function of SCI mice.
BMS locomotor function tests were scored as baseline for all experimental mice before injury, and all mice had normal motor function (9 points). On 1 dpi, BMS scores of mice in all groups fell to 0–1 point, and then began to recover overtime. On 7 dpi, BMS scores of all animals recovered to 1–4 points, but there were no statistical differences among the three groups (P > 0.05). On 14 dpi, the BMS score in the 3 dpi group began to be higher than that of the control group (P = 0.0055). On 28 dpi, the BMS score in the 3 dpi group was significantly higher than that in the 0 dpi (P < 0.05) and control (P < 0.01) groups. By the end of the experiment (42nd day post-injury), the BMS score in the 3 dpi group was significantly higher than that in the control group (P < 0.05), and kept a trend slightly, but not significantly, better than that in the 0 dpi group (P = 0.17; Figure 2).
Figure 3｜Effect of metformin on the open-field locomotor activity of SCI mice.
Four weeks after SCI, PAS open-field locomotor function test was conducted to determine rearing activity, rest time, average speed and travel distance of mice. Figure 3A shows the chamber for performing the PAS open-field locomotor function test. The 3 dpi group tended to have slightly, but not significantly better improvements in rearing activity compared to the control group (Figure 3B) (P = 0.063). No differences in the rest time were observed among the three groups (P > 0.05; Figure 3C). The 3 dpi group tended to have slightly, but not significantly, better improvement in average speed (P = 0.083; Figure 3D). However, the 3 dpi group has significantly longer travel distance compared to the control group (P < 0.05; Figure 3E). Taken together, our results demonstrated that metformin treatment in the subacute phase had a better functional recovery compared with treatment at other time points studied.
Figure 4｜Effect of metformin on the spinal cord lesion size in SCI mice at 14 days post-injury.
Reactive astrogliosis occurs after SCI. The primary benefits of glial scars include confining the inflammatory injury area and minimizing the extent of secondary damage following SCI (Gaudet and Fonken, 2018). Several studies demonstrated that elimination of astrogliosis early after SCI resulted in greater lesion area and worse functional outcomes (Sims and Yew, 2017). To compare the treatment effect of metformin at different phases, the area of the lesion site in each treatment group was measured. GFAP immunofluorescence staining was employed to outline the lesion site and GFAP-positive astrocytes were considered the boundary of the glial scar at the lesion site in mice (Wanner et al., 2013). Since our data showed that metformin treatment significantly improved motor function as early as 14 dpi, we performed morphological studies in the spinal cord tissues collected at this time point. Spinal cord sections were immunostained with the GFAP antibody. As shown in Figure 4A, in the control group, obvious tissue atrophy around the lesion center was observed. In the 0 dpi group, the overall morphology of the spinal cord tissue was relatively preserved (Figure 4B). The area of injury around the lesion center was significantly reduced in the 0 dpi group compared to the control group (P < 0.01). In the 3 dpi group, the morphology of the spinal cord was well preserved (Figure 4C). The white matter and gray matter were well arranged without a large amount of glial hyperplasia around the epicenter. Quantitative analyses showed that the average area of the GFAP-defined lesion site in the 3 dpi group was significantly smaller than that of the control group (P < 0.001; Figure 4D).
Figure 5｜Effect of metformin on the neuron and astrocytes in the spinal cord of SCI mice.
Activated astroglia in the injured spinal cord produce a variety of pro-inflammatory cytokines, including interleukin-1β, tumor necrosis factor-α, proteases and other cytotoxic factors (Klusman and Schwab, 1997). Excessive glial cell proliferation can further induce apoptosis, enlarge the area of the injured area and lead to excessive glial scar formation, thus hinder the regeneration of functional neurons and functional recovery (Tran et al., 2018). In this study, the spinal cord tissues of mice in each group were obtained at 14 dpi, and NeuN/GFAP double-labeled staining was performed to investigate the activation of glial cells and the preservation of neurons. The NeuN staining without typical cell morphology (Figure 5) in the epicenter is most likely related to dying neurons. In the control group, only a few neurons could be observed around the lesion site (Figure 5A and C). A high magnification revealed a large amount of astroglial cells intermingled with a very small number of red neurons (Figure 5D). In the 0 dpi group, more neurons were observed around the lesion site with a reduced number of astrocytes (Figure 5E–H). In the 3 dpi group (Figure 5I–L), the intensity of GFAP immunoreactivity was lower compared to both control and 0 dpi groups, and a large number of neurons were observed around the lesion site. Statistical analyses indicated that the number of neurons around the lesion site in the 3 dpi group was significantly higher than that of the control group (P = 0.0207; Figure 5M). The immunopositivity of GFAP around the lesion site of spinal cord in both 3 dpi group (P = 0.0017) and 0 dpi (P = 0.0156) groups were significantly lower than that in the control group (Figure 5N).
Figure 6｜Effect of metformin on the microglia/macrophage activation in the spinal cord of SCI mice.
Ibal-1 immunofluorescence staining was performed to observe the hypertrophy, proliferation or infiltration of microglia/macrophages at 14 dpi (Nakamura et al., 2013). As shown in Figure 6A–C, in the control group, Iba-1 immunoreactivity around the lesion epicenter was strong, and a larger number of hypertrophic microglia were observed under a higher magnification (Figure 6C). No significant changes of Iba-1 immunoreactivity were observed in 0 dpi group (Figure 6D–F). In the 3 dpi group, the immunoreactivity of Iba-1 at the lesion site was drastically reduced (Figure 6G–I). Quantitative analysis confirmed that Iba-1 immunofluorescence intensity in the 3 dpi group was significantly lower than that in the control group (P = 0.0271).
Figure 7｜Effect of metformin on the neutrophil infiltration in the spinal cord longitudinal sections around the lesion site of SCI mice.
Neutrophils start to infiltrate into the lesion site about 2 hours after injury. Infiltration of neutrophils at the early stage of injury can have a positive influence on the pathophysiological process after SCI, and depletion of neutrophils can worsen neurological outcome (Stirling et al., 2009). However, a previous study has shown that activated neutrophils can cause inflammation by releasing various inflammatory factors (Smith, 1994), are involved in ischemia/reperfusion and induce further secondary injury (Kaminski et al., 2002). Here, we chose Ly-6G/Ly-6C (Gr-1) to identify infiltrated neutrophils since Gr-1 is predominantly found on neutrophils in the peripheral blood (Conlan and North, 1994). Thus, immunofluorescence staining with the Gr-1 antibody was performed on the spinal cord tissue at 14 dpi. As shown in Figure 7A–C, Gr-1-labeled neutrophils accumulated in a large number at the lesion center in the control group. In the 0 dpi group (Figure 7D–F), the number of neutrophils in the epicenter tended to decrease without statistical significance (P = 0.184) compared to the control group (Figure 7J). In the 3 dpi group, the number of neutrophils was significantly reduced compared to the control group (P = 0.0396; Figure 7G–J).