CIMT促進腦缺血大鼠運動功能恢復與局部腦區葡萄糖代謝的相關性

李瑩瑩 華 艷 余克威 鮑偉奇 王瑜元 胡 健 胡世紅 白玉龍△

（1復旦大學附屬上海市第五人民醫院康復醫學科 上海 200240；2復旦大學附屬華山醫院康復醫學科，3PET中心 上海 200235）

Ischemic stroke is recognized as one of the main causes of long-term disability and death worldwide［1-2］.It is triggered by thrombotic or embolic occlusion of cerebral arteries，and subsequently results in“learned non-use”of the affected limb without the appropriate treatment.Constraint-induced movement therapy（CIMT）has gained considerable popularity as a treatment approach to stroke rehabilitation.The idea of this approach is to use the affected extremity intensively and repeatedly，while restraining the unaffected extremity［3］.Recent studies have shown that CIMT can reduce motor deficits and enhance fine movement of the affected limb for hemiplegic stroke patients［4-5］.CIMT has also been shown to improve behavioral performance in ischemic rat models［6-9］.Although CIMT can improve neuroplasticity and functional recovery after cerebral ischemia，the process and pattern of rehabilitation are still unclear［10］.

Micro18F-fluorodeoxyglucose positron emission tomography（18FDG-PET）can clearly visualize glucose metabolism at different levels in different brain regions.Glucose metabolism in the brain can reflect the activity of brain functions to a certain extent，and thus provide direct evidence for the activation of brain functional areas after compulsory exercise.

We previously reported the effect of CIMT on the behavior of rats with middle cerebral artery occlusion（MCAO）and the effect of cerebral glucose metabolism［11］.However，whether there is a correlation between the recovery of motor function and the activation of brain functional areas remains to be determined.This study is a continuation of our previous work.Its purpose is to further explore the correlation between motor function recovery and regional glucose metabolism induced by CIMT in rats with cerebral ischemia，so as to further understand the action mechanism of CIMT and lay a foundation for subsequent molecular mechanism research.

Materials and Methods

AnimalsTwenty-two male specific pathogenfree（SPF）Sprague-Dawley（SD）rats with body weights of 230-280 g were used in the experiment.They were provided by Sino-British Sippr/BK Lab Animal Ltd.Using a random number table，these rats were divided into an ischemic group treated with CIMT（CIMT，n=6），an ischemic group（Control，n=6），a sham-operated group（Sham，n=6），and a normal group（Normal，n=4）.All rats were kept in a room temperature environment with a 12-hour light/dark cycle.They were given plenty of food and water.In the experiment，all operations were carried out during the day.The study protocol was revised and approved by the Animal Experimental Committee of Fudan University（Approval No.201802173S）.The experiment was conducted according to the Guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals［12］.

Establishment of cerebral ischemia modelsIschemia models were established by occluding the left middle cerebral artery in the CIMT and Control groups，as previously described［13］.The middle cerebral artery was not blocked in the Sham group，and no special treatment was given to the Normal group.The operation was performed by the same surgeon using the same specification surgical nylon monofilament.The insertion depth of the monofilament was 18-20 mm.The Sham group received the same treatments as the Control group，but the middle cerebral artery was not blocked.Before，during，and after the MCAO procedure，Laser-Doppler flowmetry（moorVMS-LDF Vascular Monitoring System；Moor Instruments Ltd.，UK）was used to continuously monitor the regional cerebral blood flow（rCBF）.The blood flow in the middle cerebral artery decreased by 70%-80% after the insertion of the monofilament and the ischemic period lasted for 1.5 h.The same level of blood flow was maintained throughout the ischemic period and was slowly returned to the base value after the monofilament was removed.If the thrombe was met with the above monitoring standard，the MCAO model was deemed successful［14］.The results showed that there were no significant differences between the Control and CIMT groups.The specific operation steps and results were described in our previous study［11］.

Constraint of movement in the unaffected limbStarting from day 8（d8）after cerebral ischemia，the ipsilateral（left）forelimbs and upper torsos of rats in the CIMT and Sham groups were subjected to constrained movement for 14 days using a plaster cast with a smooth cotton inner lining.The rats in the Control group received no CIMT training and the rats in the Normal group received no treatment.All animals could move freely in the cage.CIMT training consisted of fine grip training and daily activities.The foot-fault test（FFT）was used to evaluate grip training for the affected limb.Daily training sessions consisted of 16 to 20 grips repeated 5 times with an interval of 1 to 2 minutes.In the rest of the time，the rats engaged in daily activities，grooming，eating，etc.inside the cage.After 2 weeks of CIMT training，the cast was removed and a behavioral assessment and micro-PET scan were performed.For detailed methods please refer to our previous study［11］.

Behavioral assessmentIn the present study，FFT was used to assess the fine motor function in the affected limb.A horizontal ladder with 34 rods separated from each other by 2 cm-intervals was used as previously described［15］. Under normal circumstances，rats could walk the whole distance by taking 9 to 12 steps in a row.Falling off or slipping through the rod once for the forelimb on the affected side was considered as one faulty step.The fall ratio for the affected limb was calculated from the ratio of faulty steps to the total number of steps.FFT was mainly used to measure the fine grasping ability of rat forelimbs.The gross motor function and balance ability was evaluated with the BBW.The apparatus for BBW was a beam with a width of 2.5 cm and a length of 150 cm，elevated 55 cm above the ground.The animals were rated using a 5-point scale.Higher scores indicated more severe neurological deficits.FFT and BBW tests were performed on d0 before surgery，and on d7 and d22.The results showed that CIMT promoted the improvement of rats’movement，grasping，and coordination ability.

Micro-PET/CT imagingOn d7 and d22 after cerebral ischemia，micro-PET/CT imaging was performed using the Siemens Inveon MM PET/CT scanner（SIEMENS，USA）with a lutetium oxyorthosilicate（LSO）crystal detector.18FDG was provided by the PET Center，Huashan Hospital of Fudan University.Before micro PET scan，1.0 mCi of pyrogen-free18FDG was injected through the tail vein，and then the rats were returned to the cage for about 40 minutes so that the tracer could be completely absorbed and metabolized［16］.The rats were anesthetized with 10% chloral hydrate（350 mg/kg，intraperitoneum），and then were placed in the micro PET scanner in the prone position.After a 5-minute CT scan for positioning and attenuation correction，a 15-minute dynamic PET image acquisition was performed.The acquisition mode was emission acquisition with an energy window of 350-650 kV and a time window of 3.432 ns.The collected original data were divided into single frames，and the image reconstruction method was used to obtain a multiple planar reconstruction（MPR）image with a pixel size of 0.776×0.776×0.796 mm3and an image matrix size of 128×128×159.Each PET image was spatially normalized into the space of an18FDG template using the brain normalization function in the PMOD v3.3（PMOD Technologies，Zurich，Switzerland）［17］.In the Fusion module，PET images of rats were overlapped with PET modules in the system，and standardized processing was performed.The standardized regions of interest（ROI）in the template were superimposed on the transformed PET images，and the standardized uptake values（SUVs）of brain regions were obtained for homogenization.This was done through the PMOD software package in conjunction with the W.Schiffer rat brain template and atlas.

Statistical analysisStatistical analysis was performed using the SPSS 19.0（IBM SPSS statistics，IBM Corp.，USA）.Data analysis was performed using the GraphPad Prism 5.0（LA Jolla，CA，USA）.Data were expressed inx±s.Pearson correlation coefficient（r）was used to analyze the correlation between behavioral score and glucose metabolism level in local brain region.P＜0.05 indicated that the difference was statistically significant.

Results

Correlation between the behavioral score of MCAO rats and the level of glucose metabolism in activated brain regions before CIMT（d7）Micro-PET/CT was performed in all the four groups.The Sham group was designed to eliminate the effect of plaster fixation method used in CIMT intervention on regional cerebral glucose metabolism.The Normal group was to perform as control.Because these two groups of rats have no cerebral ischemia injury，micro-PET/CT results showed that there were no significant differences in local glucose metabolism and their behavior was also normal.Therefore，we only selected the CIMT and Sham groups for correlation analysis.The same was true for the following.

The results showed that the BBW score was negatively correlated with the level of glucose metabolism in the left insular cortex and the auditory cortex（r=-0.585，P=0.046；r=-0.658，P=0.020，respectively），and was positively correlated with the level of glucose metabolism in the right hippocampus，superior colliculus，and inferior colliculus（r=0.624，P=0.030；r=0.615，P=0.033；r=0.609，P=0.036，respectively）.The results also showed that a higher glucose metabolism level in the left insular cortex and auditory cortex was associated with a lower balance beam score and better balance function in rats.Over-activation of the right hippocampus，superior colliculus，and inferior colliculus was detrimental to the recovery of balance function.It was observed that the recovery of balance function in MCAO model rats mainly depended on the activation of the affected brain region during the short-term natural recovery process（Fig 1）.

Fig 1 Correlation analysis diagram for balance beam score and glucose metabolism level in related activated brain regions on d7

The results showed that the fall ratio was negatively correlated with the glucose metabolism level in the left somatosensory cortex，insular cortex，and orbitofrontal cortex（r=-0.680，P=0.015；r=-0.717，P=0.009；r=-0.648，P=0.023，respectively）and was positively correlated with the glucose metabolism level in the right midbrain（r=0.664，P=0.019）.The results also showed that rats with a higher level of glucose metabolism in the left somatosensory cortex，insular cortex，and orbitofrontal cortex had a lower fall ratio and better fine grasping function.Over-activation of the right midbrain was detrimental to behavior.During the short-term natural recovery process 7 days after operation，the recovery of fine function in rats with cerebral ischemia was related to the activation of the affected brain region（Fig 2）.

Fig 2 Correlation analysis between the fall ratio and the glucose metabolism level in related activated brain regions on d7

Correlation between the behavioral score of MCAO rats and the level of glucose metabolism in activated brain regions after CIMT（d22）The results showed that the BBW score was positively correlated with the level of glucose metabolism in the left amygdala，caudate putamen，insular cortex，and entorhinal cortex（r=0.684，P=0.014；r=0.775，P=0.003；r=0.852，P=0.000；r=0.713，P=0.009，respectively），and was negatively correlated with the level of glucose metabolism in the Acb core shell and the caudate putamen in the contralesional hemisphere（r=-0.661，P=0.019；r=-0.883，P＜0.001）.The results suggested that a lower level of glucose metabolism in the ipsilateral amygdala，caudate putamen，insular cortex，and entorhinal cortex was associated with a higher level of glucose metabolism in the contralesional Acb core shell and caudate putamen，a lower balance beam score and better balance function in rats with cerebral ischemia.After CIMT intervention，the improvement of balance function in the ischemia group was related to the activation of bilateral brain regions.This suggested that the activation of both brain regions may play an effective role in the recovery 22 days after surgery.Compared with the Control group，CIMT promoted the recovery of motor function in rats with cerebral ischemia mainly through activation of Acb core shell in the contralesional hemisphere（Fig 3）.

Fig 3 Correlation analysis between balance beam score and glucose metabolism level in related activated brain regions on d22

The results showed that the fall ratio was negatively correlated with the level of glucose metabolism in the right entorhinal cortex（r=-0.587，P=0.045），which indicated that over-activation of the right entorhinal cortex contributed to the decrease in the fall ratio，thus promoting the improvement of the fine gripping function，but there was no significant statistical difference between the groups.However，the glucose metabolism level in the experimental group（0.017±0.001）was higher than that in the Control group（0.016±0.001），suggesting that the higher the glucose metabolism level in the right entorhinal cortex，the lower the fall ratio and the better the fine gripping function of rats with cerebral ischemia would be（Fig 4）.

Fig 4 Correlation analysis between step rate and glucose metabolism level in the right entorhinal cortex on d22

Discussion

Micro-PET uses radionuclide-labeled molecules forin vivoimaging and can provide a noninvasive，dynamic，and quantitative view of neuronal activity at the molecular level.It has become an important research method in the field of neurofunctional imaging.The basic principle of micro PET is roughly the same as that of clinical PET［18］.At present，the most widely used molecule for micro-PET，18FFDG，is an isomer of glucose［19］that can directly reflect the metabolic changes of neurons in the brain and can be applied to the study of the central reorganization mechanism.Traditional electrophysiological or molecular biological methods can selectively visualize functional changes in only a few brain regions，while PET technology can show the metabolic changes in the entire brain.Micro-PET can clearly display the glucose metabolism activity at different levels in different brain regions and the metabolism in brain regions can be observed from the coronal plane，horizontal plane，and sagittal plane，providing direct evidence for the activation of brain functional areas after CIMT.The plasticity of neuronal synapses depends on energy metabolism.Glucose metabolism is the main energy metabolism mode in the brain and the level of glucose metabolism is closely related to the activity of neurons［20-21］.Therefore，glucose metabolism in the brain reflects brain function activity to a certain extent.

In our previous study［11］，we reported the behavioral performance and changes in cerebral glucose metabolism after CIMT treatment in a rat model of cerebral ischemia.We found that CIMT improved the behavior of rats with cerebral ischemia.Furthermore，rats in the CIMT group showed changes in glucose metabolism at specific regions in the ipsilateral and contralateral hemispheres after 2 weeks of intervention.Upper limb injury is one of the most common disabilities experienced by survivors of cerebral ischemia.Clinical and experimental studies have shown that during the CIMT intervention，due to the inhibition of the healthy limbs，the use and high-intensity autonomous movement of the affected limbs can promote the recovery of motor function and improve motor disorders after cerebral ischemia［22-24］.Based on our previous research，we further explored the correlation between behavioral changes and glucose metabolism levels in specific brain regions in rats with cerebral ischemia in this study and provided evidence for the mechanism underlying motor function improvements in rats by CIMT.

In this study，during the natural recovery process 7 days after surgery，the BBW score was negatively correlated with glucose metabolism in the left insular cortex and auditory cortex，and positively correlated with glucose metabolism in the right posterior hippocampus，superior colliculus，and inferior colliculus.The results showed that higher glucose metabolism levels in the left insular cortex and auditory cortex were associated with lower balance beam scores and better balance function in rats.Over-activation of the right hippocampus，superior colliculus，and inferior colliculus was detrimental to the recovery of balance function（Fig 1）.The fall ratio was negatively correlated with glucose metabolism in the somatosensory cortex，insular cortex，and orbitofrontal cortex in the left cerebral region and positively correlated with glucose metabolism in the midbrain of the right cerebral region.The results suggested that higher levels of glucose metabolism in the left somatosensory cortex，insular cortex，and orbitofrontal cortex were associated with lower fall ratio，and better fine grasping function in rats.Excessive activation of the right midbrain was detrimental to behavior（Fig 2）.These findings suggested that the balance beam score for rats in the ischemic group was lower than that in the Sham and Normal groups 7 days after surgery and glucose metabolism was reduced in the surgeryrelated brain regions in the left cerebral cortex.Due to the lack of behavioral and PET/CT scan results 24 hours after surgery，it was impossible to compare the values before and after surgery.However，it was not difficult to speculate that the behavioral scores of rats 24 hours after surgery and the level of glucose metabolism in operation-related areas of the affected brain region may be lower.Therefore，during the first 7 days of natural recovery after MCAO，the recovery of balance function in rats with cerebral ischemia mainly played a role through the activation of the insular cortex and visual cortex（visual compensation）on the affected side.The improvement of the fine grasping function may mainly depend on the increase of the metabolic level in the somatosensory cortex，insular cortex，and orbitofrontal cortex.In conclusion，the motor function recovery in MCAO model rats during the first 7 days of natural recovery mainly depended on the activation of the affected cortex.

The results of this study suggested that 22 days after surgery，lower levels of glucose metabolism in the left amygdala，caudate putamen，insular cortex，and entorhinal cortex were related to higher glucose metabolism levels in the right Acb core shell and caudate putamen，lower balance beam scores，and better balance function in rats with cerebral ischemia（Fig 3）.These results indicated that both cerebral hemispheres were involved in the recovery of balance function in cerebral ischemia rats，and the activation of the right Acb core shell may play a major role.Compared to d7，glucose metabolism in the left caudate putamen and entorhinal cortex decreased at d22 in the Control group，while metabolism in the right caudate putamen increased.This indicated that the recovery of balance function during the natural recovery process after 22 days of MCAO in rats with cerebral ischemia was related to the activation of the bilateral cerebral hemispheres，which mainly played a role through activation of the contralateral basal ganglia.The fall ratio was negatively correlated with glucose metabolism in the right entorhinal cortex（Fig 4）.There was no significant difference in the glucose metabolism in the right entorhinal cortex between the two groups，and there was an increasing trend in the CIMT group compared with the Control group.In conclusion，the improvement of fine motor function in rats with cerebral ischemia by CIMT may be related to the increase of glucose metabolism in the right entorhinal cortex.This conclusion needed to be verified with a larger sample size.

The Acb，located in the ventral striatum，is the main brain region that controls motivation and reward and is also involved in the regulation of exercise fatigue［25］.Acb dysfunction involves mental diseases such as depression，obsessive-compulsive disorder，or neurological diseases as well as obesity，drug abuse，addiction，etc.Interestingly，a similar behavioral pattern can be observed in healthy adults who receive a motivational stimulus.Functional magnetic resonance imaging（fMRI）confirms that the ventral striatum is responsible for encoding expectations and drives motor/cognitive behavior to improve behavioral performance［26］.Animal studies have shown that the Acb plays an important role in the acquisition and expression of incentive significance［27］.Restoration of motor function and the ability to perform daily activities after stroke can be used as a reward or intrinsic motivation for therapeutic exercise.Thus，the Acb helps to promote participation and reprogramming responses during motor recovery after stroke.This conclusion has also been confirmed in related studies by our research group［28］.

In the process of rehabilitation after stroke，bilateral neurons in the central nervous system are recruited to the neural network that innervates the paralyzed upper limb.An fMRI study of stroke patients with good recovery supported a model of enhanced bilateral hemispheric motor reproduction after subcortical stroke，confirming the possibility of functional integration of these regions into the reconstructed hand functional network［29］.Our previous study suggested that CIMT significantly improved the walking ability of the rats and stimulated more neurons to enter the neural network of the paralyzed forelimb in the motor cortex and red nucleus on the healthy side，and that the motor cortex and red nucleus on the healthy side may play a more important role in structural restructuring than the corresponding motor cortex on the affected side［30］.The results of this study also confirmed the activation of bilateral cerebral hemispheres in the promotion of motor recovery in cerebral ischemia rats by CIMT，which may occur mainly through the activation of the basal ganglia.Basal ganglia are subcortical structures involved in regulating motor sequencing，motor skills，and complex actions［31］，and the specific molecular mechanism for their activation by CIMT needs to be further studied.

There are several limitations to this study as follows：（1）Lack of evaluation time points，lack of behavioral scores and PET/CT scan results 24 h or 48 h after cerebral ischemia，mainly due to the severe impairment of motor function in rats during the acute period after MCAO and the inability to complete behavioral evaluation or PET/CT scans，leading to increased mortality.（2）Lack of follow-up after treatment to investigate the long-term effects of CIMT on cerebral glucose metabolism in rats with cerebral ischemia in order to observe the dynamic changes and rules for the activation of brain regions promoted by CIMT after cerebral ischemia and further explore the mechanism of action for CIMT.（3）The sample size is small，mainly due to the high cost of micro-PET scanning.

During the natural recovery 7 days after cerebral ischemia，the recovery of motor function mainly depended on the improvement of glucose metabolism in the cortex on the affected side，while the activation of the brain region on the healthy side may be detrimental to the recovery of motor function.

During the natural recovery 22 days after surgery，both cerebral hemispheres were involved in the recovery of balance function in rats with cerebral ischemia；the recovery of motor function promoted by CIMT was also related to the activation of bilateral cerebral hemispheres，and there was no obvious bias.However，compared with the Control group，further improvements in the balance function promoted by CIMT were mainly related to the activation of the right Acb core shell in rats with cerebral ischemia.

In conclusion，compulsory exercise training mainly promotes the recovery of balance function in rats with cerebral ischemia through the activation of the Acb core shell on the contralesional hemisphere.CIMT promotes the recovery of the fine griping function，which may be related to the increase of glucose metabolism in the right entorhinal cortex.

Authors’ContributionsLI Ying-ying designed the study，performed the experiments，collected and analyzed the data，prepared the manuscript.HUA Yan and YU Ke-wei acquired and analyzed the data.BAO Wei-qi performed micro-PET/CT scaning and image reconstruction.WANG Yu-yuan and HU Jian acquired the data and performed the literature survey.HU Shi-hong revised the manuscript.BAI Yu-long designed and supervised the study，and interpreted the data in the experiment.

Conflicts of InterestThe authors declare no conflict of interest.