Cordycepin administration ameliorates neurological deficits after TBI, and reduces both GMI and WMI.
Since TBI causes multiple sensorimotor function defects, we performed the cylinder test, grid walking test, wire hanging test, and the rotarod test to evaluate the effect of cordycepin on neurological function after TBI (Fig. 1a). No obvious difference was observed between cordycepin and saline administration in sham mice. However, one-week tests after TBI showed that cordycepin reduced forelimb asymmetry in the cylinder test, decreased foot fault rate in the grid walking test, improved scores in the wire hanging test. Rotarod test was continued to 28d after TBI for tracking long-term outcomes. Cordycepin prolonged duration on rotarod even on 28d after TBI, suggesting a long-term pro-recovery effect of cordycepin. These data indicated that cordycepin administration demonstrated superior sensorimotor functional recovery after TBI compared with vehicle mice.
Immunohistochemical staining of MAP2 on the serial coronary sections was conducted for assessing the neuronal tissue loss (Fig. 1b). Cordycepin significantly decreased the volume of neuronal tissue loss compared with vehicle mice (Fig. c).
Besides grey matter, white matter is also the main target of TBI closely linked to neurological impairment[28, 29]. Firstly, MBP and NF-200 were stained to observe myelin and neurofilament changes in the striatum after TBI (Fig. 1d). The fluorescence intensity of NF-200 significantly reduced after TBI, whereas cordycepin increased the NF-200 fluorescence intensity significantly (Fig. 1e, up-left panel). White matter fiber tracts in the striatum were perpendicular to the coronal sections we chose, constructing bundles in the sections. Although no significant difference was found between the groups by measuring the fluorescence intensity of MBP staining (Fig. 1e, down-left panel), we found that the bundles became loose and the space among the bundles decreased in perilesional striatum after TBI. Further analysis revealed that the area proportion of MBP staining bundles in vehicle mice significantly increased because of the atrophy of striatum, while cordycepin reversed the increase (Fig. 1e, down-right panel). The data manifested the effect of cordycepin on maintaining white matter structure.
It was reported that TBI led to axonal degeneration and the myelin debris could exist for a long time. So MBP staining may be difficult to reflect the factual conditions of myelin. Secondly, regions of corpus callosum (CC) were dissected from mice in all groups on 35 days after TBI to examine the ultrastructure of axon and myelin via transmission electron microscope (Fig. 1f). After TBI, no substantial variety were observed in the myelinated axon number (Fig. S1a) or degenerating myelinated axon number (Fig. S1b). But the number of nonmyelinated axons were significantly decreased by TBI and restrained by cordycepin significantly (Fig. h left panel). Cordycepin administration tendentiously abolished the TBI-induced degenerating nonmyelinated axons (Fig. 1h, right panel, p = 0.0507). Interestingly, although no obvious changes were observed in the two types of myelinated axon numbers, g-ratio of myelinated axons was reduced by TBI, whereas cordycepin ameliorated the reduction (Fig. 1g). These data indicated cordycepin protecting the nonmyelinated axons in number and the myelin in structure.
CAPs were usually measured to assess the conductive capacity of white matter fibers. Due to the saltatory conduction of nervous impulse between Ranvier nodes, the nerve conductive velocity on myelinated axons is faster than nonmyelinated axons. The difference of conductive velocity can be presented on the N1 and N2 peaks in CAPs. Thirdly, mice were sacrificed 35d after TBI and sections were prepared to record CAPs in CC with electrophysiology technique (Fig. S1c). The amplitudes of N1 and N2 were both deceased by TBI (Fig. S1d, Fig. 1i), while only N2 amplitude was significantly rescued by cordycepin (Fig. 1i). The results suggested cordycepin improved nerve conductive capacity of nonmyelinated fibers.
Taken together, cordycepin administration ameliorated TBI-induced neurological deficits and brain tissues injury, including grey matter and white matter injury.
Cordycepin administration inhibits pro-inflammatory microglia/macrophage and promotes anti-inflammatory microglia/macrophage
Neuroinflammation is one of the main causes for secondary WMI in TBI. Our previous study has proved the close correlation between microglia/macrophage polarization mediated neuroinflammation and WMI. According to the temporal characteristic of microglia/macrophage polarization, we double stained CD16/Iba1 (pro-inflammatory) and CD206/Iba1 (anti-inflammatory) on 3d and 7d after TBI (Fig. 2a). Cordycepin inhibited pro-inflammatory microglia/macrophage both in the cortex and striatum on 3d and 7d (Fig. 2b), and promoted anti-inflammatory microglia/macrophage in the striatum on 7d after TBI (Fig. 2c). Besides, the polarization markers were detected on 7d after TBI using q-PCR. Markers related to pro-inflammatory microglia/macrophage such as CD16 and IL17a were significantly decreased by cordycepin (Fig. 2d). Meanwhile, markers related to anti-inflammatory microglia/macrophage such as CD206 and IL-10 were increased in the cordycepin administration group (Fig. 2e). The results demonstrated cordycepin inhibited pro-inflammatory microglia/macrophage and promoted anti-inflammatory microglia/macrophage polarization after TBI.
Cordycepin administration protects BBB integrity after TBI.
Normally, BBB prevents toxic substances in the blood from leaking into the brain parenchyma to maintain homeostasis in the brain. Preservation of BBB integrity contributes to reducing pro-inflammatory microglia/macrophage activation. Therefore, we investigated the effects of cordycepin on BBB integrity in TBI mice on 3d after TBI. Firstly, the tracers of sulfo-NHS-biotin (MW 443.4 Da) and Evans blue (Evans blue-Albumin, MW 68,500 Da) were injected into the femoral vein to examine the BBB permeability. Compared with TBI + vehicle group, sulfo-NHS-biotin or Evans blue leakage volume were significantly reduced by cordycepin (Fig. 3a-b, Fig. S2). The data collaboratively indicated that cordycepin reduced the leakage of both large and small molecules in TBI mice.
Endogenous IgG in the blood can leak into brain parenchyma through dysfunctional BBB, which makes IgG an important endogenous biomarker to assess the BBB leakage. We stained 3d-post-TBI frozen section with anti-mouse IgG antibody to detect the leakage of endogenous substance. IgG positive area was remarkably reduced in the cordycepin administration group, compared with the vehicle TBI mice (Fig. 3c and d, up panel). Additionally, cordycepin reduced fluorescence intensity of IgG staining in the striatum suggesting the reduction of IgG leakage in amount (Fig. 3d, down panel).
Tight junction proteins, such as ZO-1, are the main components of tight junctions between cerebrovascular endothelial cells supporting the BBB integrity. TBI distinctly reduced ZO-1 expression detected by western blot, whereas, cordycepin reduced ZO-1 protein loss after TBI (Fig. 3e-f). We also observed the ultrastructure of BBB with a transmission electron microscope (Fig. 3g). TBI led to edema in astrocyte end-feet, basement membrane thickening, and loss in tight junction proteins, while cordycepin reversed these degenerations.
MMPs are responsible for tight junction protein degradation in brain injury. Our previous study showed knocking out or pharmacological inhibition for MMPs reduced BBB permeability after brain injury. MMP-2 and MMP-9 are the main two MMPs involved in BBB disorder, and they mediate early reversible and late irreversible BBB dysfunction respectively. Accordingly, we observed the effects of cordycepin on MMP-9 and MMP-2 at 24 h and on 3d after TBI. Western blot showed that cordycepin significantly reduced the protein level of MMP-9 on 3d after TBI, but not MMP-2 (Fig. 3h-i). Many factors participate in MMPs post-transcriptional control. For example, endogenous tissue inhibitors of metalloproteinases (TIMPs) inhibit MMPs activity by high-affinity binding to the catalytic structural domain of MMPs in a non-selective manner. Consequently, it’s necessary to pay attention to the activity when evaluating the effects of MMPs. Based on the properties of MMP-9 and MMP-2 to catalytic degrade gelatin, we employed gelatin zymography to assess their activity. Zymography results showed that the activity of MMP-9 and MMP-2 was both reduced by cordycepin on 3d after TBI (Fig. 3j-k).
Cordycepin administration inhibits neutrophil infiltration, without affect peripheral immune system
The massive influx of circulating leukocytes is the biggest contributors to MMPs surging in the early stage of brain injury, especially for neutrophil, which has been proved to be the main source of MMP-9[38, 39]. The main kinds of immune cells, including microglia, macrophage and its inflammatory subset, neutrophil, T cell, B cell, and dendritic cell, were analyzed in the brain on 3d after TBI by flow cytometry (Fig. 4a). The proportion of microglia in total cells decreased, while the proportions of macrophage, inflammatory macrophage, and neutrophil increased after TBI. Cordycepin made no significant differences in the amounts of microglia, macrophage, and inflammatory macrophage in TBI mice. However, the proportion of neutrophils was decreased by cordycepin administration. The proportion of T cell, B cell, and dendritic cell didn’t change obviously in each group (Fig. 4b).
Furthermore, immunofluorescent staining of Ly6G showed that the neutrophils adhered to blood vessels or went deep in the brain parenchyma indicating that neutrophils did infiltrate after TBI (Fig. 4c). Most infiltrated neutrophils surrounded the lesion, whereas few were found in other places (Fig. 4d). The number of neutrophils were remarkably reduced by cordycepin in the total or only the cortex (Fig. 4e, up panel) and had a downtrend in the striatum (Fig. 4e, down-left panel, p = 0.0501). More interestingly, we noticed that the number of neutrophils was significantly associated with the number of pro-inflammatory microglia/macrophage (Fig. 4e down-right panel) suggesting that infiltrated neutrophils might promote microglia/macrophage polarizing to pro-inflammatory type.
The immune cells in the blood were also detected by flow cytometry (Fig. 5). However, macrophages, inflammatory macrophages, neutrophils, T cells, B cells, and dendritic cells in the blood weren’t significantly affected by TBI or cordycepin at 3d after TBI. The results indicated cordycepin did not aggravate immunosuppression induced by brain injury obviously, moreover, its inhibition to neutrophils was more likely due to the chemotaxis in CNS.
Inhibiting of neutrophil infiltration by cordycepin ameliorates BBB disruption and microglia/macrophage pro-inflammatory polarization
To investigate the relationships among three events (neutrophil infiltration, BBB disruption, and microglia/macrophage polarization), anti-Ly6G monoclonal antibody (anti-Ly6G) was intraperitoneal injected twice for neutrophil depletion at 24 h pre- and post-TBI. By flow cytometry analyzing the blood, anti-Ly6G specifically depleted neutrophils without significant effect on macrophages (Fig. S3, Fig. 6a-b). Neutrophil depletion remarkably reduced neutrophils (Fig. 6c-d), IgG leakage (Fig. 6e-f), pro-inflammatory microglia/macrophage (Fig. 6g-h) in the brain, while cordycepin combined with neutrophil depletion didn’t exert a further influence on them (Fig. 6c-h). Moreover, neutrophil depletion significantly reduced the foot-fault rate on the grid test and prolonged the duration on rotarod at 3 day after TBI, whereas neutrophil depletion and cordycepin-combined group did not show further effect on neurological deficits (Fig. 6i). Interesting, the foot fault rate showed a significant positive correlation with neutrophil in both the cortex and the striatum (Fig. 6j).
These data manifested that neutrophil infiltration contributed to BBB disruption and microglia/macrophage pro-inflammatory polarization. Neutrophil infiltration might have an upstream position among the three events in the effect pathway of cordycepin.
Cordycepin inhibits adenosine A2a receptor expression and ameliorates inflammatory cytokines expression after TBI.
Due to the remarkably similar molecular structure to adenosine (Fig. 7a), cordycepin was considered to probably act on adenosine receptors. And it has been proved that cordycepin has interactions with adenosine A1, A2a receptors[40–42]. After TBI, adenosine receptors begin to work within dozens of minutes. Furthermore, we examined the changes of adenosine A1, A2a receptors at 24 h after TBI. Firstly, their mRNAs were detected by qPCR. A1 receptor mRNA expression had no substantial difference between each group (Fig. 7b, left panel), but TBI remarkably upregulated mRNA of A2a receptor, which was inhibited by cordycepin (Fig. 7b, right panel). The expression of adenosine A2a receptor by immunofluorescence staining showed a consistent phenomenon (Fig. 7c-d).
It was reported that adenosine A2a receptor knockout reduced the expression of pro-inflammatory cytokines and chemokines after TBI. Like adenosine A2a receptor knockout, cordycepin administration also significantly reduced the expression of TNF-α, IL-1β and CCL3 mRNA in TBI mice (Fig. 7f). These data suggested that cordycepin might work through inhibiting adenosine A2a receptor.