Improvement of Contused Spinal Cord in Rats by Cholinergic-like Neuron Therapy

Background Disability in spinal cord injury is an important medical problem, and cell transplantation is considered as an option for the treatment. Objectives The purpose of this study is to use bone marrow stromal cells (BMSCs) derived cholinergic neuron-like cells (CNL) in order to ameliorate the contusion model of spinal cord injury in rats. Materials and Methods The CNLs were produced by pre inducing BMSCs with β-mercaptoethanol (BME) followed by inducing with nerve growth factor (NGF). The cells were immunoreactive to neurofilament 200, NeuN, synaptophysin, synapsin, microtubule associated protein-2 and choline acetyl transferase (ChAT). The CNL were transplanted in contused rats (CR), which were sacrificed after 12 weeks. Results The results showed that BBB test showed an improvement in the CR, while the quantitative analysis showed that the improvement rate was higher in the rats treated with CNL than those treated with BMSCs only or the untreated animals, similar results were noticed in the improvement index. Immunohistochemical analysis of the tissue section prepared from the CR showed that the transplanted cells were engrafted and integrated in the traumatized spinal cord. The morphometric analysis showed that the volume density of the cavity in the CNL treated rats was significantly lower than that of the untreated ones, while the spinal tissue regeneration index was significantly higher. Conclusions The conclusion of the study is that CNL can improve the injured spinal cord.


Background
McDonald et al. transplanted neural phenotype derived such as bone marrow stromal cells (BMSCs) and umbili from mouse embryonic stem cells (ESCs) into rats with spi-cal cord cells could be an alternative source for the treat nal cord injury, which improved their locomotors activi-ment of central nervous system injury models because of ties (1). However, Henon suggested using adult stem cells their ability to differentiate into neuron-like cells and to as an alternative source for cell therapy because of the ethi-engraft the traumatized tissues with functional recovery cal problems in the clinical use of ESCs (2). Newman et al. (3). Other investigators agreed with them (4)(5)(6)(7)(8). Moreover, mentioned that the hematopoietic system derived cells Romano emphasized the safety concerns of using ESCs in clinics and suggested the use of adult stem cells for transplantation (9). Li et al. (10) reported the adverse ef fects of ESC transplantation in different models of neu rodegeneration including tumor growth and immune rejections, similar concerns were addressed by Coutts and Keirstead (11). Chopp et al. transplanted BMSCs in an animal model of spinal cord injury, which improved the behavioral test results (12). Hofstetter et al. justified the use of BMSCs for their accessibility and expandability (13). One of the major advantages of BMSCs transplanta tion in spinal cord injury is its autologous feature (14). Intraspinal transplantation of undifferentiated BMSCs was reported to express neuron, astrocyte (15) and oligo dendrocyte markers (16). While Verdú et al. evaluated the injured spinal cord by measuring the size of the cystic cavities, which was larger in the untreated animals than those treated with olfactory ensheathing cells transplant (17). Zurita and Vaquero (18) confirmed that BMSCs could improve chronic paraplegia and reduce the size of spinal cord injury cavitation (14,19), where the transplanted cells integrated in the injured spinal cord tissues (20). In the transplantation of primed human NSCs in a contu sion model of rat spinal cord, the cells differentiated into cholinergic neurons, which improved the locomotion of the injured rats, however, the improvement depended on several variables such as post injury transplantation timing, transplantation site and the survival of the dif ferentiated cells (21). Moreover, partial improvement of the atomized Moto neurons in the newborn was also documented (22), Hou et al. used self-assembling peptide seeded with Moto neurons differentiated from fetal neu ral stem cells and transplanted in injured spinal cords of rats, which resulted in partial functional recovery (23).
One of the functions of the cholinergic neurons in the spinal cord is motor activities (24).

Objectives
In this study, the improvement of locomotion in the contused rats transplanted with cholinergic neurons de rived from autologous BMSCs has been evaluated by us ing quantitative methods.

Materials and Methods
The trans differentiation of the cholinergic neuron-like cells from the BMSCs was done according to Naghdi et al. (25), briefly, the BMSCs were prepared from 6-8-week old Sprague-Dawely rats by removing and cutting the ends of the femurs and the tibias, they were flashed out with 5 ml of αMEM medium (Gibco) supplemented with 10% FBS (Gibco) and cultured in the same medium (supplement ed with 10% FBS, penicillin and streptomycin) for 24 hrs. The adherent cells were used for differentiating the BM-SCs after the fifth passage, and the features of the BMSCs were evaluated using anti-fibronectin, anti-CD 45, anti-CD 44 and Oct-4 expression using RT-PCR. The differentia tion protocol is consisted of a pre induction stage using β-mercaptoethanol for 24 hours (BME: 1mM) followed by an induction stage using the nerve growth factor for 2, 4 and 6 days (3, 5 and 7 days from the beginning of the differentiation protocol) (NGF: 100 ng/ml). A cocktail of antibodies (primary antibodies) against neuronal mark ers was used in order to evaluate the differentiation by immunocytochemistry, including nestin, NF-68, NF-200, MAP2, NeuN, synaptophysin and the cholinergic neu ron marker (choline acetyl transference: ChAT), which was also detected by immunocytochemistry. The BMSCs derived cholinergic neuron-like cells (CNL) used for the in vivo study were labeled with bromodeoxyuridine (BrdU: 0.1 mM, Sigma) by adding BrdU into the culture 72 hours before the pre induction. The female Sprague-Dawley rats (230-250 g) were purchased from Razi In stitute, Tehran, Iran. The animals were divided into five groups: sham operated (S), contused without treatment (C), placebo (P: contused animals treated with 9 µL nor mal saline only, which was used as the vehicle and was injected intraspinally at the epicenter, rostral and caudal of the impact site), the BMSCs treated group (B: 300,000 BMSCs in vehicle injected as above) and the cholinergic trans differentiated neurons from the BMSCs (N: 300,000 CNL in vehicle injected as above). The contusion was carried out by using the New York Weight drop device (NYW) (26). Ketamine (80mg/kg) and xylazine (10mg/ kg) were used in order to anesthetize the animals, which were laminectomies at T13, and a 10 gm weight rod was dropped from a height of 2.5 cm onto the exposed spinal cord, then the muscles and skin were sutured over the laminectomies vertebra. Postoperative care was done us ing Ringer lactate (subcutaneous: 5 ml) and ceftazoline (50 mg/kg) twice a day for 3 days, and Tramadol (20 mg) for 2 days. The animals were maintained for 12 weeks and an open field test was done on all of them (in the experi mental and control groups) according to Basso-Beattie-Bresnahan scale (BBB scale). They were subjected to pre surgical training for 10 days, then BBB test was done on the day of the surgery (day 0) and days 4,7,8,11,14,21,28,35,42,49,56,63,70, 77 and 84. The data were analyzed by non-linear regression using the Logistic model, the "c" coefficient was considered as the improvement rate (26,27). The immunostaining was done on the cultured cells as follows: they were washed with phosphate buffer saline (PBS), fixed in acetone, rewashed with PBS, perme ated with 0.3% triton X-100, blocked with 10% normal goat serum and incubated with a primary antibody, that was followed by FITC conjugated secondary antibody. The im munolabeled cells (200 cells) were counted at 200 X from random fields on the immunostained coverslips. Double labeling immunohistochemistry was done on the spinal tissues cut with cryostat from the groups, then they were immunolabeled with anti-BrdU primary antibody, which was followed by incubation with Rhoda mine conju gated secondary antibody. The tissues were then double labeled with anti-ChAT primary antibody and incubated with FITC conjugated secondary antibody. The numeri cal density per area, the mean of the numerical density of the transplanted BMSCs per area and the mean of the numerical density of the transplanted trans differentiat ed cholinergic neuron-like cells per area were evaluated. Similar groups were perfused with paraformaldehyde, processed for paraffin, cut and stained with hematoxylin and eosin. The mean of the volume density of the spinal cavity at the epicenter, rostral and caudal of the impact site were calculated. The spinal tissue regeneration index (STRI) was calculated as follows: ( in treated animals -in untreated animals) STRI = 1 -V ̅ vcavity in treated animals -V ̅ vcavity in un treated animals / V ̅ vcavity in untreated animals in untreated animals The data were statistically analyzed by using SPSS pack age (www.spss.com) and the normality of the data was evaluated by the one-sample Kolmogorov-Smirnov test (SK test) and the one way analysis of variance (ANOVA). Turkey's test post Hoc was used for analyzing the results.  ChAT at the induction stage was more than that of the pre induc tion one, while MAP2 expression was variable. The result of BBB test in the CNL treated group was significantly higher than that of the BMSCs one at 11thday and the 2th, 3th and 4th weeks, while the score was not significant in the other time points (Figure 5 A) and the score of the sham-operated animals were significantly higher than that of the other groups. ANOVA showed that BBB score was significantly higher in the animals treated with CNLs and BMSCs than the untreated animals and those treated with normal saline. Figure 5 (B-E) shows the nonlinear re gression by using the Logistic model for BBB score versus time (in days) in the untreated animals, accordingly the curve fitting was done for the normal saline group, the animals treated with BMSCs and those treated with CNLs. Table 1 shows the improvement rates and the improve ment indices in the groups. The lowest index was noticed in the NS group, while the highest index was in the CNL group, however, the latter shows higher improvement rate than the BMSCs group. The improvement rate is the "c" coefficient in the logistic model, while the improve ment index is the ratio of the improvement rate in any group to that of the sham operated group. Figure 6 A shows the sustained decline in the percentage of nest ing immunoreactive cells while the reverse is correct for the values of NeuN, synaptophysin and ChAT. The nu merical density of the transplanted BMSCs per area was significantly higher than those of the CNLs (Figure 6 B). The results of the volume densities of the cavitation in the contused spinal cord are presented in Table 2, which shows that the volume density in the C and NS groups is higher than that in the CNL and BMSCs ones. The volume density of the CNL treated group is higher than that of the BMSCs group. Similar results were noticed in the spi nal tissue regeneration index. The BrdU labeled cells were injected intraspinally and the animals were sacrificed af ter 3 months.

Discussion
The percentage of the cholinergic neuron phenotypes pre induced with β-mercaptoethanol and induced with NGF was consistent with a previous investigation (25). From the in vitro study, the time point for their trans planting should be decided from the time course, the selection of this time point was based on two criteria: the differentiation property and the integrity of the transplanted cells. The differentiation criterion is the highest number of the induced cells immunoreactive to the neuroprogenitor cell marker (nestin). The reason for selecting the neuroprogenitor cell is its being the best choice for replacement therapy for neurodegenerative diseases (28). Moreover, the induced undifferentiated stem cells into neural phenotypes could engraft, survive     (30). Therefore, the best option is to select the neuro progenitor cells committed to cholinergic phenotypes, where the percentages of immunoreactive cells to  neuron-like cells into the spinal cord is that the in vivo which are essentially injured neurons with less chance of new microenvironment may not provide sufficient NGF survival as transplants (25). The transplantation of these to these cells, thus they start to die depending on their cells in the injured spinal cord resulted in improvement maturity (35) and the possible reason for cholinergic of BBB score compared with the untreated animals. The neuron-like cells death is deprivation of NGF (36). The trend of improvement with significantly higher BBB experiment was terminated as the improvement reached score was noticed in the cholinergic phenotype treated a plateau (12 weeks) (37). The non-linear regression of group more than in those treated with BMSCs at the first BBB score using the logistic model showed that initially two weeks, while in the following weeks (3-12 weeks), the rate of improvement was higher in the cholinergic the data showed no significant differences. However, phenotype treated group than those treated with the numerical density per area in the cholinergic the BMSCs, which may indicate that the cholinergic phenotype treated group at the end of the experiment phenotypes could improve the functional activity better was about 50% of that of the group treated with BMSCs. than the BMSCs. The possible reason for the lowest The possible explanation for the reduction in the number improvement index in the normal saline treated group of cholinergic transplants is that they were unable to is that the contusive rats were subjected to reoperation survive the transplantation. The induction protocol of at one week old surgical wound, where the wound was the BMSCs into cholinergic phenotypes by NGF, used reopened and the spinal cord was injected with the in this study, could result in their dependency on NGF vehicle and then closed again, these procedures may with conditioning of survival on the presence of NGF in traumatized the already contused spinal cord. Similar procedures were carried out in the other groups such as the BMSCs and the cholinergic phenotype treated ones. This may explain the low result of the improvement index in the BMSCs animals compared to those subjected to the contusion injury without reoperation. The explanation is that the improvement index was calculated by taking the ratio of the improvement rate of a given group to that of the sham operated group, the second trauma resulted from reopening the operation site and injecting the BMSCs, which caused the decline in the rate of improvement. However, the improvement rate in those animals injected with normal saline was lower than those injected with the BMSCs. The improvement of the contused spinal cords using cholinergic transplant is consist with the results of Gao et al. (22), who reported that cholinergic neurons differentiated from neural stem cells were able to innervate the target muscle and cause motor function improvement, while other investigators reported similar results from transplanting BMSCs in the injured spinal cord (38). Also Dezawa et al. (39) revealed that BMSCs could successfully integrate the injured spinal cord and cause behavioral improvement. Clinical trials by other investigators reported improvement of the traumatized spinal cord (40,41). The transplantation of neuronal lineages differentiated from bone marrow could promote the recovery of mice with spinal cord injury with significant improvement of the motor function; it may explain the higher improvement index compared with that of the BMSCs (42). Also, Tarasenko et al. revealed that the pre differentiation stage of the transplanted human neural stem cells was a determining factor in the outcome of the functional improvement of the contused rats (21). Table 1 shows the improvement rate and the improvement index in the sham operated (sham), the rats with contusion spinal injury without any treatment (contusion), the rats with contusion spinal injury injected with normal saline (normal saline), the rats with contusion spinal injury injected with bone marrow stromal cells (BMSCs) and the rats with contusion spinal injury injected with trans differentiated cholinergic neuron-like cells (cholinergic).   Figure 4: A)Immunostaining of the pre induced BMSCs with anti-synaptophysin monoclonal antibody, then la beled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide. B) Immunostaining of the induced BMSCs with anti-MAP2 monoclonal antibody, then la beled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide. C) Immunostaining of the induced BMSCs with anti-synaptophysin monoclonal antibody, then labeled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide. D) Immunostaining of the pre induced BMSCs with anti-ChAT monoclonal antibody, then labeled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide. E) Immunostaining of the pre induced BMSCs with anti-MAP2 monoclonal antibody, then labeled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide. F) Immunostaining of the pre induced BMSCs with anti-ChAT monoclonal antibody, then labeled with (secondary antibody) FITC conjugated rabbit anti-mouse antibody, the cells are counterstained with ethidium bromide (scale bar = 75 μm: all). ; standard error = 0.74102311, corre lation coefficient = 0.98836318. "a" means that the animal group treated with cholinergic neuron-like cells is sig nificantly higher than the other group at that time point. Figure 6: Shows a histogram of the percentages of immunoreactive cells to nestin (white solid column), synaptophysin (black solid column), NeuN (down cross hatched column) and ChAT (up cross-hatched column) in the preinduction phase (1st day) and induction phase at 3rd, 5th and 7th daysfrom the start of the differentiation protocol. B) Shows a histogram of the numerical density per area, which represents the number of the survived bone marrow stromal cells (white solid column) and the trans differentiated cholinergic neuron phenotypes (black solid column), after 12 weeks of transplantation surgery, at the epicenter the of impact site of the spinal cord (E), the rostral (R) and the caudal (C) of the site. "a" means that the PIC to this marker is significantly higher at this time point than the other time points. "b" means that the PIC to this marker is significantly higher at this time point than the other markers. "c" means the numer ical density of the transplanted BMSCs per area, which is significantly higher than those of CNLs. Figure 7: Cholinergic neuron phenotypes and endog enous ChAT positive cells simultaneously labeled with mouse anti-ChAT monoclonal antibody (primary anti body) and incubated with rabbit anti-mouse antibody conjugated with FITC. B) Cholinergic neuron phenotypes labeled with mouse anti-BrdU monoclonal antibody (primary antibody) and incubated with (secondary anti body) rabbit anti-mouse antibody conjugated with Rho da mine. C) The images A and B merged showing a double labeled transplanted cell (arrow), empty arrowhead rep resents a cell immunostained with anti-ChAT antibody which is the host neuron immunoreactive to anti-ChAT antibody; arrowhead represents a non-cholinergic neu ronal transplanted cell. D) A phase contrast image of A, B and C (X 400), this represents the region of the square in E (scale bar = 25 μm). F) A tissue section stained with H&E from a spinal cord of a rat used as negative control, it is injected with normal saline and shows large cavities in the contused spinal cord (scale bar = 5000 μm).