Effect Of Lycium Barbarum Polysaccharide On Crushing Spinal Cord Injury In Rat

Spinal cord injury (SCI) has long been a focus of research, from both basic scientificand clinical therapeutic points of view. We have previous studied early neurosurgery (4-14days after SCI) of spinal cord contusion on30ASIA-A (complete paralytic) patients,gaining satisfactory result. Translating back to the bench side, much work remains onmedicines that may further reduce the secondary injury or improve function recovery, inaddition to the early neurosurgery therapy. Immune response is one of the most importantpathological events following SCI, in particular the double-edged sword effect involvingboth beneficial and detrimental, for example M1and M2polarization of macrophage. Ingeneral, M1is detrimental, whereas M2is neuroprotective, which is our main field.Lycium barbarum polysaccharide (LBP), the main content of Lycium barbarum, wasfound to serve most beneficial functions of wolfberry. Mounting evidence suggested animmune regulation in mice and in human of LBP, as well as its direct neuroprotective andstrongly anti-oxidative effect on CNS. The effects of LBP are highly related to SCI, buthave not yet been studied. Besides, the pathological changes involved in SCI are often fairly subtle. Therefore,an animal model with high stability is crucial. The weight-dropping and compressioninjury model are most commonly used, which, however, often result in profuse bleeding,distributed longitudinally from injury site to distal. The distal hemorrhage is unpredictableboth in extent and severity, which stand in way of construction of a model with highstability. Additionally, there remains another problem that, it is still not certain about theformation of the distal hemorrhage.Since there is still no reasonable and wildly accepted explanation for the formation ofdistal hemorrhage, the first part in our study centered on the formation of distalhemorrhage in a compressive SCI model.In the second part, for resolution of the problem in contusion and compression injury,a modification on a crush injury model was conducted to minimize the interference ofdistal hemorrhage in model stability.Then in the third part, we studied the effects of LBP the on rats compromised withabove model, in particular its relationship with macrophage M1/M2polarization.Part one: Cause of distal hemorrhage in compressive spinal cord injuryCompressive model: a weight compression on cord was performed, with a resultantcomplete compression in dorsaventral direction.Morphology after compressive SCI: at6h and3d, the distal hemorrhage wasobvious in the center of posterior column, similar to the observation in other literature, andspecifically in the interface between the Cortical Spinal cord Tracts (CST) posterior to thegray commissure and the other part of posterior column.By dorsal hemisection at two adjacent sites rostral and caudal to the compression siteimmediately the model was completed, there was no distal hemorrhage at3d post-injury.And by instant injection of carbon powder into the compression site when SCI wascompleted, we observed carbon powder in distal hemorrhage at6h.The results showed that the distal hemorrhage results from the direct expansion of thehemorrhage in the compressive site, the specific distribution of which may related to the CST and the other dorsal column posterior to it. Therefore, it may potentially rely oncontrolling the central hemorrhage to avoid the distal hemorrhage.Part two: Avoiding the distal hemorrhage, a crushing spinal cord injury modelwith low variability in rat was obtained.Aiming to make a model avoiding the occurrence of distal hemorrhage, we adopt aforceps crush model for following two reasons:1) graded injury severity can be easilyachieved by adjusting the spacer between the forceps when they are fully closed;2)crushing force is exerted from lateral sides to avoid or diminish the involvement of CST.Comparison of the hemorrhage in crush injury of different severity showed, injurywith0.5mm spacer caused a hemorrhage focus of fusiform shape at injury site, withoutdistal extending, and the result was stable with low variability (coefficient of variation,CV=10.4%), whereas the severer injury may still result the distal hemorrhage, and mildercould not cause enough lesion and may increase relative weight of the systemic error witha resultant higher variability（CV:0.2mm,22%;0.8mm,55%）.In the0.5mm crush injury, hemorrhage predominantly existed in gray matter, with aresultant fact that the majority part of hemorrhage was in the lateral gray matter whereasthe minor in central canal gray matter. In different saggital planes, the CVs of hemorrhagearea between animals were calculated, and the CV (8%) in the plane containing main partof lateral gray matter was significantly smaller than that (17%) in the central canal plane.At7d post-crush of0.5mm, consistent secondary damage was observed, includinglesion area (CV=9.9%), spared neuron in para-lesion area (CV=10.2%), number ofapoptotic oligodendrocyte (CV=13.5%) and reactivated macrophage/microglia (CV=11%)Therefore, the0.5mm injury is a model with controllable injury size and lowvariability, and the lesion in lateral gray matter of cord is especially stable, which couldserve as a good model.Part three: Effect of LBP on crushing spinal cord injury MethodsIn vivo study was carried out in the above crush model of0.5mm, and two LBPadministration plans were performed for different purpose: LBP-pre, since there is a longinterval between LBP intake and pharmacodynamic action, rats were treated with LBPsince7d pre-SCI, following most LBP study in past.2) LBP-aft, LBP was given since7dafter injury as a therapeutic medicine. Water was given as vehicle.Histological observation and Western blotting were performed, at7and14d inLBP-pre experiment and14in LBP-aft. By western blotting, levels of ED1, iNOS (markerof macrophage M1polarization), Arginase1(Arg1, marker of macrophage M2polarization)were assessed. In morphology, we measured the lesion area flanked by glial fibrillaryacidic protein (GFAP) hyper-positive glial scar, and in the para-lesion area,immunofluorescence intensity of ED1and GFAP were measured and number ofED1+iNOS+(M1) or ED1+Arg1+(M2) were counted.In vitro: primary astrocytes and murine N9microglial cell line were cultured, forfurther evidence of the in vivo result in astroglial scar and macrophage/microgliapolarization.Primary astrocytes were incubated with different medicines: LBP versus vehicle, LBPplus TNFα+IFNγ+LPS versus TNFα+IFNγ+LPS. And GFAP expression was assessed byimmunocytochemistry and western blotting. The TNFα+IFNγ+LPS are known stimuli forastrocyte activation.N9cell experiment generally include two parts:1) To investigate the effect of LBP and LBP associated with LPS+IFNγ (activator formacrophage M1polarization) or IL-4(activator for macrophage M2polarization), cellswere incubated with following medicines for one day: vehicle versus LBP, LPS+IFNγversus LPS+IFNγ+LBP or IL4versus IL4+LBP.2) To mimic the animal experiment in LBP-pre and LBP-aft, we treated the N9cellsfor total2days with two sequential and different stimuli, with each stimulus for one day.“LBP-pre”:1st d, LBP and2nd d, LBP plus LPS+IFNγ+IL-4, with correspondingtreatment free of LBP as control.“LBP-aft”:1st d, LPS+IFNγ+IL-4and2nd d, LBP plus LPS+IFNγ+IL-4, with corresponding treatment free of LBP as control. iNOS and Arg1levels were evaluated by WB.Results:The lesion areas results were beyond our expectation. In LBP-pre animals, lesionarea were increased at both7d (1.30versus0.97mm2) and14d (1.17versus0.75mm2)compared to the vehicle at corresponding time point, respectively; in LBP-aft group it wasreduced in comparison to the control at14d (0.56versus0.80mm2), conversely to theLBP-pre group.To explain the above phenomenon, following experiments were conducted:Reactive astrocyte is the main component in glial scar, which directly influence thelesion expansion. Result on primary astrocyte and in animal showed no differencebetween all LBP groups and corresponding controls.ED1: Immunostaining intensity of ED1in LBP-pre group was increased than control,at both7d and14d, and the ED1intensity level significantly correlated with lesion areachange. WB result also showed higher ED1level in LBP-pre animals compared to vehiclecontrol. Between the LBP-aft and their control, no significant difference was found at14d,by either Immunostaining or WB.The ED1result could not completely explain the effect of LBP on lesion area, then,we turned to the M1/M2polarization (iNOS/Arg1).iNOS and Arg1:In in vivo study, WB: LBP-pre up-regulated iNOS level and down-regulated Arg1,with a consequent increased ratio of iNOS/Arg1; whereas, LBP-aft induced increase inboth iNOS and Arg1, with the latter to a greater extent, resulting a lower iNOS/Arg1ratio.Since it has been reported that iNOS can be expressed in neurons and neutrophils,immunohistochemistry was used to confirm if the iNOS/Arg1changes represent theM1/M2ratio. The LBP-pre animals at7d and14d showed increased M1and reduced M2cells, exhibiting significantly higher M1/M2ratio, relative to the corresponding controls.In LBP-aft study, no measurable changes was found in M1cell number between the twogroups, however, treated animals displayed an increased number of M2cells, with a resultant lower M1/M2ratioIn N9cells stimulated for1d, LBP alone or allied with IL4or resulted in increasediNOS and decreased Arg1, in comparison to their corresponding controls, whereas nodifference between LPS+IFNγ and LBP plus LPS+IFNγ group. In N9cells stimulated for2d,“LBP-pre” caused augmented iNOS and suppressed Arg1, however,“LBP-aft”induced decrease in iNOS and increase in Arg1, compared to their controls, respectively.In summary, LBP could serve a protective property in SCI, when given since7d afterinjury. LBP also played a detrimental role if the ad ministration begins at7d before injury.Further, allowing for that the M1macrophages exhibit detrimental effect, whereas the M2beneficial, with the ratio of M1/M2determining the role of macrophage/microglia in SCI,the converse modulation on microglia/macrophages polarization of LBP in our resultprobably serve as the mechanism underlying its divergent effects on cord damage.Although the LBP-pre administration is not beneficial for SCI, from clinical applicationpoint of view LBP is definitely a safe and effective medicine to reduce the secondaryinjury after spinal cord contusion or other types of spinal cord injury.