Summary
The present protocol simulating central cord syndrome (CCS) in mice has improved repeatability and minimized operation damage to the experimental animals, avoiding disrupting the anatomical structure excessively. The strategy in this study is advantageous because it allows for research into injury mechanisms by producing consistent results.
Abstract
Animal models of central cord syndrome (CCS) could substantially benefit preclinical research. Identifiable anatomical pathways can give minimally invasive exposure approaches and reduce extra injury to experimental animals during operation, enabling the maintenance of consistent and stable anatomical morphology during experiments to minimize behavioral and histological differences between individuals to improve the reproducibility of experiments. In this study, the C6 level spinal cord was exposed using a spinal cord injury coaxial platform (SCICP) and combination with a minimally invasive technique. With the assistance of a vertebral stabilizator, we fixed the vertebrae and compressed the spinal cords of C57BL/6J mice with 5 g/mm2 and 10 g/mm2 weights with SCICP to induce different degrees of C6 spinal cord injury. In line with the previous description of CCS, the results reveal that the lesion in this model is concentrated in the gray matter around the central cord, enabling further research into CCS. Finally, histological results are provided as a reference for the readers.
Introduction
Recent years have witnessed a constantly rising incidence of spinal cord injury (SCI), with more injuries in older people from less violent tauma1. These injuries more frequently involve the cervical spine and more often lead to an incomplete neurologic dysfunction2.
In the twenty-first century, CCS is the most prevalent type of incomplete SCI, accounting for more than half of all SCI. Compared with conventional incomplete SCI, CCS is characterized by disproportionately more impairment of the upper than lower extremities3. It is characterized by predominantly upper extremity weakness with less significant sensory and bladder dysfunction. CCS is thought to be caused by post-traumatic central region hemorrhage and edema or, as recently proposed, by Wallerian degeneration from compression of the spinal cord in spinal canal stenosis. The management of CCS lacks high-level evidence to guide, which requires a comprehensive understanding of its pathophysiology4. However, models of CCS have not been reported. Suitable animal models are essential for the understanding of pathophysiology, which can provide a research base for clinical and preclinical studies5,6,7,8,9,10.
In this study, a CCS model in mice is established with a spinal cord injury coaxial platform (SCICP) and a minimally invasive operation plan, which allows for further research into and understanding of CCS. The model is proved valid in the course of the research process by histological, magnetic resonance imaging (MRI), and immunofluorescence analysis.
Protocol
Experiments were approved by the Laboratory Animal Ethical and Welfare Committee of Shandong University Cheeloo College of Medicine (approval number: 22021). They were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publications No. 85-23, revised 1996). All mice used in this study were 9-10 week old female C57BL/6J mice purchased from Jinan Pengyue Experimental Animal Company (Jinan, China). A total of 9 mice involved in this study were equally randomized to the control, mild, and severe groups. At 7, 28, and 70 days postinjury, one mouse from each group was sacrificed.
1. C6 laminectomy and spinal cord exposure
NOTE: The exposure was performed under a microscope. Bleeding can be avoided by paying attention to two aspects: (i) All blood vessels should be avoided. (ii) Muscles need to be separated at the origin and termination points of the muscle.
- Prepare surgical instruments and SCICP.
NOTE: The structure of SCICP has been reported in previous study11. The difference with respect to the previous study is that the present protocol accomplishes spinal cord injury by compression. Two different weights (10.4 g and 20.8 g) of this platform can produce compression of 5 g/mm2 and 10 g/mm2, respectively (Figure 1). The spinal cord exposure and compression steps are shown in Figure 2. - Administer isoflurane to the mouse by inhalation using a nose cone (induction: 3%-5%, maintenance: 1.5%-2%).
- After the anesthesia takes effect, explore a small bulge at the midline behind the neck of mice, which is the spinous process of the second thoracic vertebra (T2).
- Shave the hair around this bulge. Disinfect the skin with three alternating applications of an iodophor solution followed by skin antiseptics 75% ethanol.
- Position the mouse prone on the operation table. Apply eye ointment to protect the eyes.
- Lay a 3-4 mm-thick pad under the chest to allow an arching cervical spine curve, facilitating exposure of the inter-lamina space and an unobstructed airway during operation. Inject buprenorphine as pre-operative analgesia (0.05-0.1 mg/kg, SQ).
- Make a 1-1.5 cm longitudinal incision with a sterile scalpel centered on the 2nd thoracic vertebra spinous process to expose the fascial layer (Figure 2A).
- Remove a portion of the adipose tissue above T2 with sterile micro scissors to find the T2 spinous process.
- Separate the bilateral trapezius and rhomboid muscles from C5-T2 along the midline with micro scissors (Figure 2B).
- Separate the muscles on the lamina of C5-T2 vertebrae with micro scissors, and retract the muscle layer to the sides with sterile micro-retractors (Figure 2C).
- Cut the multifidus and cervical spinal muscles on the surface of the vertebrae.
- Locate T2 according to the highest point of the spinous processes. Probe the spinous processes successively toward the rostral end from T2 to locate C6 (Figure 3).
- Lift the C6 lamina with forceps, cut off the lamina, and the spinal cord is exposed (Figure 2D).
2. Cervical spinal cord compression injury
- Clamp the C6-7 facet joints with the vertebral stabilizator, and lock it (Figure 2E).
- Aim the sterile weight tip at the exposed spinal cord, and ensure the flat bottom of the tip is positioned parallel to the dorsal surface of the spinal cord (Figure 2F).
- Adjust the sleeve to make the weight compress the spinal cord. Stop adjusting when the weight keeps a constant relative position with the spinal cord (Figure 2G).
NOTE: Do not make this process too violent or quick in case the weight exerts contusion force on the spinal cord. - Remove the weight and the vertebral stabilizator after a 5 min compression.
- Observe the color alterations of the spinal cord after compression under the microscope (Figure 2H).
- Rinse with sterile PBS and use suction to clean the operation site.
- Suture the muscles and skin in layers using polypropylene nonabsorbable suture (size: 6-0).
- Disinfect the surgical area, place the mouse on a warm pad until the mouse restores full consciousness, and then return the mouse to the mouse cage.
- Inject buprenorphine for analgesia (0.05-0.1 mg/kg, SQ) every 8-12 h for 3 days.
3. Histological analysis
- Anesthetize the mouse by intraperitoneal injection of 1.25% tribromoethanol (0.02 mL/g body weight) on days 7, 28, or 70 after injury. Transcardially infuse the mouse with 60 mL of phosphate-buffered saline (PBS) and 20 mL of 4% paraformaldehyde11.
- Transect the spinal cord at 0.5 cm from the lesion center from both sides with micro scissors, and preserve the 1 cm-long section.
- Immerse the preserved spinal cord section in 30% sucrose at 4 ˚C for 48 h.
- Embed the tissues with OCT, slice the tissues into 6 µm thick sections with a cryotome, and collect the sections on a glass slide.
- Hematoxylin and eosin staining
- Rinse the 6 µm sections with 1x PBS for 5 min 3 times to remove residual OCT.
- Immerse the sections in hematoxylin for 90 s. Wash the sections under running water for 3 min.
- Immerse the sections in eosin for 4 min. Soak in 95% alcohol for 30 s to remove excess eosin.
- Finally, dehydrate the slides with alcohol (95% alcohol and 100% alcohol twice, successively) for 30 s and put the slides in a xylene bath for clearing for 2 min. Then, seal the sections with a cover glass and resin gel.
- Prussian blue staining
- Submerge the slides for 20 min in an equal mixture of potassium ferrocyanide (10%) and hydrochloric acid (10%).
- Rinse 3 times with distilled water, and counterstain for 5 min with Nuclear Fast Red.
- Rinse three times with distilled water, followed by one rinse with 95% alcohol and two rinses with 100% alcohol for 5 min.
- Clear the sections in xylene twice for 3 min each and then seal with resin gel12.
- Immunofluorescence staining
- Incubate the slides with the following primary antibodies for 1 h at 37 °C: rabbit anti-Ionized calcium-binding adapter molecule 1 (Iba-1) (1:500), which was up-regulated in microglia after nerve injury; mouse anti-glial fibrillary acidic protein (GFAP) (1:300), which is expressed in astrocytes in the central nervous system; rabbit anti-neurofilament-200 (NF-200) (1:2000), which is expressed in neurofilament.
- Incubate with secondary antibodies for 1 h at room temperature (RT): Alexa Fluor488 goat anti-mouse and Alexa Fluor594 goat anti-rabbit (1:1,000).
- Snap photographs and further analyze with a fluorescence microscope13.
4. Magnetic resonance imaging
- Anesthetize the mouse at 7 days post-injury with Isoflurane anesthesia (1%-2% isoflurane, 20%-30% O2) administered through a mini mask.
- Scan the cervical spinal cord in sagittal orientation. Use the following settings for MRI imaging: Spin-echo (SE) sequence in multislice and interleaved fashion with TR/TE = 2500/12 ms, acquisition matrix =256 x 128 matrix over the field of view (FOV) = 12 x 8 mm2, slice thickness = 1 mm, and the number of excitations (NEX) = 2.
NOTE: Keep the mouse's respiration rate at 10-15/min during scanning to eliminate respiration-related image artifacts14.
Representative Results
The sagittal HE section suggests that even though the damaged area in the gray matter was wider in the severe group, continuity over the white matter was present. In addition, the difference in damaged gray matter area between the severe and mild groups supports the reasonableness of the group setting in the protocol (Figure 4).
The coronal HE sections show that the lesion mainly exists in the gray matter in both groups. In the severe group, the structure of the white matter surrounding the gray matter was more likely to be impacted, but the outline of the white matter was still maintained (Figure 5). NF-200 immunofluorescence suggests that even though the white matter surrounding gray matter was affected in the severe group, the white matter was still relatively intact. These results are consistent with the characteristics described for CCS in the previous study4 (Figure 6).
No red blood cells were found in sagittal HE sections at 7 days post-injury in either the mild or the severe group. The Prussian blue staining revealed no hemosiderosis in the mild group but in the severe group. These results indicate the inducement of hemorrhage may require a relatively severe degree of damage (Figure 7).
Immunofluorescence revealed areas of elevated GFAP and Iba-1 expression in both mild and severe injury, suggesting an inflammatory response and the formation of a glial scar in the lesion. Also, the severe group exhibited a larger lesion area than the mild group (Figure 8).
MRI is a relatively minimally invasive method to observe the spinal cord. The results suggest that in both the mild and severe groups, there is a hypointense signal change in the lesion with a high signal outline. The severe group showed a significantly larger hypointense signal area (Figure 9). The hypointense signal suggests a precipitate from the reticulocyte lysate in this area, and the surrounding hyperintense signal suggests an inflammatory response. We did conduct several behavioral tests in our previous study. For instance, the grip strength test in the forelimbs reveals a significant difference15.
Figure 1: The sleeve and weights of SCICP. The surface area of the tip was designed to be 1.3 mm x 1.6 mm based on the exposed area of the spinal cord measured after the C6 laminectomy. The weight is coated with PTFE, which effectively reduces friction between the sleeve's inner wall and the weight. Please click here to view a larger version of this figure.
Figure 2: Exposure and compression of the spinal cord. (A) Longitudinal incision of the skin; (B) Separate the muscles rostrally from T2 spinous process; (C) Separate the muscles above the laminae; (D) C6 laminectomy; (E) Fixing the vertebral body; (F) Determining the location of compression; (G) Compression of the spinal cord; (H) No significant damage to the white matter above the spinal cord after spinal cord compression. Please click here to view a larger version of this figure.
Figure 3: Mouse cervical skeleton anatomy. The site indicated by the arrow is the T2 spinous process. Please click here to view a larger version of this figure.
Figure 4: The sagittal HE stained sections. (A) Cervical spinal cord sagittal section. (B,C) The Severe group had more severe damage than the Mild group, but both focused on the gray matter around the central cord. The 7, 28, and 70 dpi images suggest no significant difference in the expression of injury in the same injury group at different periods and that the continuity of the white matter in the superior and inferior spinal cord is maintained. Scale bar: 1 mm. Please click here to view a larger version of this figure.
Figure 5: Cervical spinal cord injury coronal HE stained sections. (A-C) The injury primarily affects the gray matter surrounding the central cord, as seen in panels B and C. The severe injury group suffers from a more extensive range of damage than the mild injury group, which is more likely to affect the white matter. Scale bar: 400 µm. Please click here to view a larger version of this figure.
Figure 6: NF-200 coronal immunofluorescence after injury. NF-200 response with no significant difference at white matter outline. Scale bar: 400 µm. Please click here to view a larger version of this figure.
Figure 7: Prussian blue staining. (A-C) Hemosiderosis was observed in the severe group but not the mild group. Scale bar: 400 µm. Please click here to view a larger version of this figure.
Figure 8: Sagittal GFAP and Iba-1 immunofluorescence after injury. (A-C) As the degree of injury increases, the area of GFAP and Iba-1 response increases. Scale bar: 1 mm. Please click here to view a larger version of this figure.
Figure 9: Sagittal MRI after cervical spinal cord injury (T2-weighted images). The injury region was observed as a hypointense signal in the mild and severe injury groups, with a significantly wider area of hypointense signal in the severe injury group. Please click here to view a larger version of this figure.
Discussion
Of the numerous types of spinal cord injury, CCS is one of the most potentially treatable types of injury3,4. Due to the lack of laboratory research models, research on CCS from the 1950s focused on clinical studies and cadaveric dissection investigations3,16,17. The present study shows using compatible tools and minimally invasive procedures to establish mice's CCS model. From a technical perspective, this platform has strong operability and good reproducibility. Given that experiment results demonstrate the validity, our technique to establish the model closest to the standard previous studies have defined for CCS4.
Previous studies of compression injury have mainly employed aneurism clips, balloons, and calibrated forceps9,10,18. Moreover, most injuries occurred at the level of the thoracic spinal cord18. The spinal cord at the C6 level was chosen as the injured segment in this study to investigate the characteristics of CCS. It is worth the attention that the survival rate of the CCS model is also an essential factor in ensuring experimental consistency. The present study reports causing bilateral compression injury to the mouse cervical spinal cord, while high-level spinal cord traumatic injury, especially bilateral injury, could be fatal for experiment animals if too serious. According to El-Bohy, the C4/5 spinal cord is more likely to affect the descending bulbospinal tract and respiratory-related motoneurons, which leads experiment animals to respiratory depression and death18,19,20,21,22,23., In this study, mice with different degrees of compression on the C6 cervical spinal cord have significantly differentiated injury characteristics suggested by histological tests. Although there were significant behavioral and histological differences in the mouse cervical spinal cord clamping model reported by Forgione, disruption of the pedicles, articular processes, laminae, and even nerve roots were required in order to clamp the spinal cord with the modified clamps, which had a significant influence on the stability of the cervical structures24. Another study of cervical injuries reported using the transverse process as a fixation site5. Even though the articular processes were prevented from damage, over-muscular tissue breakdown could likewise lead to an impact on the stability of the spinal cord. In the present study, only the 6th cervical lamina was resected to maintain the stability of the cervical spinal cord, with the adjoining articular joints preserved and excessive muscular damage avoided. At the same time, compression from above the spinal cord prevents damage to the nerve roots.
The HE results suggest that the area of damage to the cervical spinal cord of the mice in each group was mainly in the gray matter near the central cord, which characterized CCS, with significant differences in the scope of injury between the different groups. Notably, the pathological sections we displayed may have alleviated injury manifestation because the specimens were collected at a few days post-injury. Immunofluorescence (NF-200) showed less damage to the neural tracts in the white matter region of the spinal cord, which also confirmed that the damage in CCS was mainly concentrated around the central cord. Immunofluorescence result was compounded by previous histological results of pathology. Previous studies have shown that CCS leads to edema near the central cord, leading to hematoma and, ultimately, dysfunction in the medial portion of the lateral corticospinal tract3. Hemorrhage has been reported as a typical component of CCS but is rarely seen in subsequent imaging and autopsy studies17. In this study, HE results at 7 days post-injury suggested signs of tissue edema in all groups; however, no residual red blood cells were found in the injury area. Therefore, Prussian blue was used to examine the injury area for hemorrhage, and the results corresponded with hemosiderosis observed in the injury area of the severe injury group at 7 days post-injury, whereas the mild group did not. MRI T2 images showed that both mild and severe injuries had low signal areas in the damaged area of the injury at 7 days post-injury, indicating the deposition of reticulocyte lysate here. These results provide circumstantial evidence that the discrepancy between the previously reported findings is probably due to the MRI test being potentially more sensitive than the histological test14, in addition to the severity of the injury, which may also influence the amount of hemorrhage in the area of injury. GFAP was also expressed extensively in the damaged area. At the same time, Iba-1 expression was also seen in intact areas, suggesting the persistence of an inflammatory response, consistent with the MRI results, where a ring of hyperintense signal around the hypointense signal area in the lesion suggests the presence of an inflammatory response. Ultimately, based on the results of the present study, the area of injury in the model was focused on the gray matter around the central cord, which is generally consistent with the descriptions previously reported13. Unfortunately, we did not perform MRI repetitively in every experiment animal to show how the injury site dynamically changes with time. Future researchers can include this in their work for better investigation into CCS. Also, Immunolabeling with neuronal markers like NeuN, which define the gray matter can be included in the study.
In conclusion, the characteristics of the findings on pathology and MRI scans have close similarities with those described for CCS in previous studies4. The present protocol feasibly modeling CCS enables further research into and understanding of CCS.
Disclosures
The authors have nothing to disclose.
Acknowledgments
This study was supported by the National Key Research and Development Project of Stem Cell and Transformation Research (2019YFA0112100) and the State Key Program of National Natural Science of China (81930070).
Materials
| Name | Company | Catalog Number | Comments |
| 4% fixative solution | Solarbio | P1110 | 4% |
| Anti-Neurofilament heavy polypeptide antibody | Abcam | ab8135 | Dilution ratio (1:2000) |
| Eosin Staining Solution (water soluble) | Biosharp | BL727B | |
| Ethanol | Fuyu Reagent | ||
| Fluorescent microscope | KEYENCE | BZ-X800 | |
| Frozen Slicer | Leica | ||
| GFAP (GA5) Mouse mAb | Cell Signaling TECHNOLOGY | #3670 | Dilution ratio (1:600) |
| Goat anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor Plus 488 | ThermoFisher SCIENTIFIC | A32723TR | Dilution ratio (1:1000) |
| Goat anti-Rabbit IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor Plus 594 | ThermoFisher SCIENTIFIC | A32740 | Dilution ratio (1:1000) |
| Hematoxylin Staining Solution | Biosharp | BL702A | |
| Mice | Jinan Pengyue Experimental AnimalCompany | C57BL/6J | |
| Microsurgery apparatus | Shandong ULT Biotechnology Co., Ltd | All the surgey instruments are custom-made | Ophthalmic scissors, micro mosquito forceps, microsurgery forceps, micro scissors |
| Normal sheep serum for blocking (working solution) | Zhong Shan Jin Qiao | ZLI-9022 | working solution |
| O.C.T. Compound | SAKURA | 4583 | |
| Phosphate buffered solution (PBS) | Solarbio | P1020 | pH 7.2–7.4 |
| Prussian Blue Iron Stain Kit (With Eosin) | Solarbio | G1424 | |
| RWD Laboratory inhalation anesthetic station | RWD Life Science Co., Ltd | R550 | |
| Small animal in vivo microCT imaging system | PerkinElmer | Quantum GX2 | |
| Spinal cord injury coaxial platform | Shandong ULT Biotechnology Co., Ltd | Custom-made(Feng's standard) | https://shop43957633.m.youzan.com/wscgoods/detail/367x5ovgn69q18g?banner_id=f.81386274~goods.7~ 1~b0yRFKOq&alg_id= 0&slg=tagGoodList-default%2COpBottom%2Cuuid% 2CabTraceId&components_ style_layout =1&reft=1659409105184&spm= g.930111970_f.81386274&alias= 367x5ovgn69q18g&from_uuid= 1362cc46-ffe0-6886-2c65-01903 dbacbba&sf=qq_sm&is_share= 1&shopAutoEnter=1&share_cmpt =native_wechat&is_silence_auth=1 |
| Surgery microscope | Zumax Medical Co., Ltd. | zumax, OMS2355 | |
| Tris Buffered Saline+Tween (TBST) | Solarbio | T1082 | Dilution ratio (1:19) |
| Xylene | Fuyu Reagent |
References
- Liu, C., et al. Survival in 222 Patients With Severe CSCI: An 8-Year Epidemiologic Survey in Western China. Archives of Physical Medicine and Rehabilitation. 100 (10), 1872-1880 (2019).
- Qi, C., Xia, H., Miao, D., Wang, X., Li, Z. The influence of timing of surgery in the outcome of spinal cord injury without radiographic abnormality (SCIWORA). Journal of Orthopaedic Surgery and Research. 15 (1), 223 (2020).
- Brooks, N. P.
- Avila, M. J., Hurlbert, R. J.
- Forgione, N., Chamankhah, M., Fehlings, M. G. A mouse model of bilateral cervical contusion-compression spinal cord injury. Journal of Neurotrauma. 34 (6), 1227-1239 (2017).
- López-Dolado, E., Lucas-Osma, A. M., Collazos-Castro, J. E. Dynamic motor compensations with permanent, focal loss of forelimb force after cervical spinal cord injury. Journal of Neurotrauma. 30 (3), 191-210 (2013).
- Allen, L. L., et al. Phrenic motor neuron survival below cervical spinal cord hemisection. Experimental Neurology. 346, 113832 (2021).
- Reinhardt, D. R., Stehlik, K. E., Satkunendrarajah, K., Kroner, A. Bilateral cervical contusion spinal cord injury: A mouse model to evaluate sensorimotor function. Experimental Neurology. 331, 113381 (2020).
- Ropper, A. E., Ropper, A. H.
- Sun, G. D., et al. A progressive compression model of thoracic spinal cord injury in mice: function assessment and pathological changes in spinal cord. Neural Regeneration Research. 12 (8), 1365-1374 (2017).
- Elzat, E. Y., et al. Establishing a mouse contusion spinal cord injury model based on a minimally invasive technique. Journal of Visualized Experiments. (187), 64538 (2022).
- Lu, J., Xu, F., Lu, H. LncRNA PVT1 regulates ferroptosis through miR-214-mediated TFR1 and p53. Life Sciences. 260, 118305 (2020).
- Zeng, H., et al. Lentivirus-mediated downregulation of α-synuclein reduces neuroinflammation and promotes functional recovery in rats with spinal cord injury. Journal of Neuroinflammation. 16 (1), 283 (2019).
- Bilgen, M., Al-Hafez, B., Berman, N. E., Festoff, B. W. Magnetic resonance imaging of mouse spinal cord. Magnetic Resonance in Medicine. 54 (5), 1226-1231 (2005).
- Yilihamu, E. E., et al. A novel mouse model of central cord syndrome. Neural Regeneration Research. 18 (12), 2751-2756 (2023).
- Chikuda, H., et al. Effect of early vs delayed surgical treatment on motor recovery in incomplete cervical spinal cord injury with preexisting cervical stenosis: A randomized clinical trial. JAMA Network Open. 4 (11), e2133604 (2021).
- Jimenez, O., Marcillo, A., Levi, A. D. A histopathological analysis of the human cervical spinal cord in patients with acute traumatic central cord syndrome. Spinal Cord. 38 (9), 532-537 (2000).
- Menezes, K., et al. Human mesenchymal stromal/stem cells recruit resident pericytes and induce blood vessels maturation to repair experimental spinal cord injury in rats. Scientific Reports. 10 (1), 19604 (2020).
- Hutson, T. H., Di Giovanni, S. The translational landscape in spinal cord injury: focus on neuroplasticity and regeneration. Nature Reviews. Neurology. 15 (12), 732-745 (2019).
- El-Bohy, A. A., Schrimsher, G. W., Reier, P. J., Goshgarian, H. G. Quantitative assessment of respiratory function following contusion injury of the cervical spinal cord. Experimental Neurology. 150 (1), 143-152 (1998).
- El-Bohy, A. A., Goshgarian, H. G. The use of single phrenic axon recordings to assess diaphragm recovery after cervical spinal cord injury. Experimental Neurology. 156 (1), 172-179 (1999).
- Gonzalez-Rothi, E. J., Lee, K. Z. Intermittent hypoxia and respiratory recovery in preclinical rodent models of incomplete cervical spinal cord injury. Experimental Neurology. 342, 113751 (2021).
- Locke, K. C., Randelman, M. L., Hoh, D. J., Zholudeva, L. V., Lane, M. A. Respiratory plasticity following spinal cord injury: perspectives from mouse to man. Neural Regeneration Research. 17 (10), 2141-2148 (2022).
- Forgione, N., et al. Bilateral contusion-compression model of incomplete traumatic cervical spinal cord injury. Journal of Neurotrauma. 31 (21), 1776-1788 (2014).
