Effects of estrogen on Survival and Neuronal Differentiation of adult human olfactory bulb neural stem Cells Transplanted into Spinal Cord Injured Rats

In the present study we developed an excitotoxic spinal cord injury (SCI) model using kainic acid (KA) to evaluate of the therapeutic potential of human olfactory bulb neural stem cells (h-OBNSCs) for spinal cord injury (SCI). In a previous study, we assessed the therapeutic potential of these cells for SCI; all transplanted animals showed successful engraftment. These cells differentiated predominantly as astrocytes, not motor neurons, so no improvement in motor functions was detected. In the current study we used estrogen as neuroprotective therapy before transplantation of OBNSCs to preserve some of endogenous neurons and enhance the differentiation of these cells towards neurons. The present work demonstrated that the h-GFP-OBNSCs were able to survive for more than eight weeks after sub-acute transplantation into injured spinal cord. Stereological quantification of OBNSCs showed approximately a 2.38-fold increase in the initial cell population transplanted. 40.91% of OBNSCs showed differentiation along the neuronal lineages, which was the predominant fate of these cells. 36.36% of the cells differentiated into mature astrocytes; meanwhile 22.73% of the cells differentiated into oligodendrocytes. Improvement in motor functions was also detected after cell transplantation.

3 This subsequently stimulates excitotoxic neuronal death cascade events associated with activation of caspase, astrocytes and microglial cells (Chen et al., 2005 andRavizza et al., 2005). McDonald et al., (1999) indicated that not only the immature spinal cord environment, but also the adult one may permit neuronal differentiation under certain conditions. They were first to achieve substantial neuronal differentiation of embryonic NSCs in the adult spinal cord.
Several types of stem cells have been investigated in SCI. Each cell type has a different therapeutic potential, varying according to their cellular behavior, post-implantation survival, proliferation and specific differentiation. Although each cell type has a specific mechanism for the treatment of neurodegenerative disease, they all fall into broad categories: axonal remylination, lost neuron regeneration, immune modulation, extracellular environment modification or a combination thereof (Vawda et al., 2012).
Human adult OBNSCs are an ideal autologous source and interesting tool for cell -based therapy of neurodegenerative disorders. They (1)  SCI is a complex event and many obstacles persist after injury that would not be fully overcome by stem cell transplantation alone. Thus combining complementary strategies would be required to advance stem cells-based treatments to the clinical stage (Teng et al., 2002). In the current study, to further enhance the survival, proliferation and migration of transplanted adult OBNSCs in the hostile microenvironment of SCI, we treated the animals with estrogen as neuroprotective, immunomodulatory and anti-inflammatory agent (Siriphorn et al., 2012 andsamantaray et al., 2016.) This was done rapidly after KA injury to prevent the acceleration of primary and secondary tissue damage and to prevent microglial invasion to the injured cord.
In vitro and in vivo studies show that estrogen prevents Ca 2+ influx (Nilsen et al., 2002), decreases the calpain activity , and decreases proteolytic and apoptotic markers (Linford and Dorsa 2002; Sribnick et al. 2007b). The attenuation of these parameters with estrogen therapy following SCI is important and essential for neurological recovery.

Animals
Twenty -eight male rats (weighing 250: 300 g) were housed in autoclaved plastic cages under a temperature-controlled environment of 24 ± 1 0 c with access to rat-chow and water ad libitum and a 12-hour light-dark cycle. The rats were acclimated to the environment for 7 days prior to the experiment. We followed the procedures adopted by the relevant IACUC at Mansoura University, Egypt.

Experimental Design
The animals were randomly divided into five groups: The normal control group: (n=2) was left under observation without treatment, the sham control group: ( n=4) was bilaterally injected intraspinally with cereprospinal fluid (CSF). The Excitotoxic lesion group: (n=6) was bilaterally intra spinal injected with K.A (1.5 μL of 2.5 Mm), the KA+ Estrogen group: (n=6) given estrogen after KA excitotoxic injury (at 15 min and 24 h after KA injury, 4 mg/kg BW estrogen was given intravenously in tail-vein). These initial doses were followed by intraperitoneal injection of five daily doses of 2 mg/kg BW, and the treated group: (n=6) given estrogen as the fourth group followed by transplantation of adult human olfactory bulb neural stem cells (h-OBNSCs) 7 days after excitotoxic injury.

Animal Surgery and KA injection
The rats were anaesthetized by an intraperitoneal injection of Ketamine (60 mg/kg body weight) and Xylazine (20 mg/kg body weight) mixture. With electric clippers, the dorsal area of the rat 5 was shaved from the lower back to the neck, an area extending 2 cm bilaterally from the midline. Animals were placed in a stereotaxic frame with a vertebral clamp, and the vertebral column was immobilized. An incision was made between thoracic-13 (T-13) and lumbar-3 (L-3) vertebrae according to the method of (Braga-Silva et al., 2007). The KA was injected bilaterally and spinally via a stereotaxic apparatus similar to that used by Yezierski et al. (1993).
Injections were made at 1.6 mm from dorsal surface and approximately 0.5 mm from the mid line (David et al., 1999). The needle was held for 2 minutes to prevent oozing of the KA solution. Muscles and the overlying skin were closed after KA microinjection. Each rat was singly housed and supplied with adequate food and clean water. Prophylactic analgesics and antibiotics were not used, to prevent any possible interaction with the experimental therapy.

Behavioral Assessment
The Basso, Beattie and Bresnahan (BBB) score: The hind limb locomotor function of each animal was evaluated using a 21-point scale (Basso et al. 1995). A Zero BBB score was given when there was no movement of the hind limbs, while a score of 21 was given when trunk stability, plantar stepping, toe clearance, an erect tail and coordinated limb movement was detected (Basso et al. 1995). The tests were done by two independent blinded examiners. Motor function of hind limbs of each rat was assessed for 4 minutes weekly till the end of the study. The final average score of the rats in all groups was then compared.

Human Olfactory Bulb NSC Isolation and Culturing
This step was described with details in our previous study published as Marei et al., (2016).

Transfection and Infection
Human embryonic kidney (HEK)-293T cells in log-phase growth were rapidly transfected by using the standard Lipofect Amine reagent (Invitrogen), with GFP-Vector plus helper plasmids to develop virions (Cenciarelli et al., 2006). Two days after cell transfection media containing virions was collected and transferred directly onto OBNSCs. Lentiviral infection was performed 6 in the presence of polybrene solution (8 mg/ml) (Sigma-Aldrich). During the OBNS/PC selection and maintenance time, the antibiotic G418 (Euroclone) was added to the cells at 400mg/ml.

OBNSCs Transplantation
Animals in the treated group received cyclosporine (10 mg/kg/daily, s.c) one day prior to transplantation till the end of the study. Transplantation surgeries occurred at 7 days after KA excitotoxic injury. Immediately before transplantation, cell viability was assessed by the trypan blue exclusion test (only populations with a 95 % viability were transplanted) and the total number of cells was measured using a hemocytometer. After that they were suspended at 40,000cells/µL in artificial CSF (Sigma-Aldrich), and were kept on ice during transplantation surgery. Rats were re-anesthetized as before and the site of surgery was re-opened. Each rat was stereotaxically and bilaterally injected with 4 µL (2 µL for each side) of cell suspension (160,000 cells) at a depth of 1.5mm, 4mm cranial to KA injected site. The needle was removed after 5 min to avoid cell suspension oozing. The muscle and skin were sutured again.

Sample collection
The animal samples in time course experiments are listed in Table (  Section was incubated with anti GFP (the same antibody that was used in fluorescence but at 1:100 dilution) for 3hrs at room temperature, then washed. Secondary antibody (ready to use) (Biotinylated Goat Anti-Rabbit IgG, ab64256,) (Abcam) was applied to each section for 15 min at room temperature followed by washing with PBS. Freshly prepared horseradish peroxidase substrate was applied to the sections at room temperature until suitable staining developed (about 5 min); then they were rinsed with PBS and counterstained with hematoxylin for 3 minutes Slides were washed, dehydrated and examined under a light microscope.

Stereological Quantification
An enzyme stained section was examined under an Olympus CX41 light microscope. The approximate percentage of GFP-Positive differentiated cells (according to nuclear morphology) was assessed by a hematoxylin nuclear counterstain and GFP immunolabeling. For quantification of GFP positive cells, starting sections were selected randomly, and the four serial sections that contained the most human cells were used for fate analysis. The number of labeled cells was counted for each type (neurons or astrocytes or oligodendrocytes) using Image J Version 10.2 software with a cell counter plug-in.This number is expressed as a percentage of the total number of GFP-positive cells counted in each individual type.

Statistical analysis
Statistical analysis of all groups was performed using one-way ANOVA analyses P<0.05 was considered statistically significant.

Control group
10 The spinal cord of of the rat is composed of two distinct regions; the outer white matter and the inner gray matter. The shape of gray matter was similar to that of a butterfly or H shape; within it a central canal is located. The ventral horns of H shape were broader than the dorsal horns (figure1a). The gray matter of each ventral horn is formed primarily of the large neuronal cell bodies, a mixture of glial cells, several blood capillaries, and a fine amorphous eosinophilic background called the neuropil (figure1b). The cytoplasm of neuronal cell bodies and dendrites containing basophilic granules called Nissl granules (figure1c).

Sham control group
This sham group was examined to confirm that the degeneration that occurred in spinal cord cells was certainly a result of KA injection, not from the Hamilton needle. There was no histological alteration after CSF injection.

Excitotoxic lesion group
The microscopic examination of spinal cord sections revealed that KA lesions were mainly located in the ventral horn. The dorsal horn showed relative preservation of tissue architecture and its neurons were almost completely spared by K.A. injection from 1 day to eight weeks post injury. The white matter remained morphologically intact (fine meshwork-like appearance) after KA injection during the whole period of study. A variety of necrotic alterations was observed in the ventral horn of the gray matter from the first day of KA injection to the 6 th week after injection. The affected neurons were characterized by cell body shrinkage, intensely stained eosinophilic cytoplasm (red dead neuron) (figure2a), and loss of Nissl substance (chromatolysis) (figure 2b,c).The neuropil adjacent to the necrotic neurons was finely vacuolated or edematous with markedly astrocytic reactions(figure2a).

Estrogen treated group
Histopathological examination of spinal cord sections showed that the administration of estrogen rapidly after spinal cord injury dramatically improved the histological alterations following KA injury. A small number of neurons displayed signs of recovery and regeneration on the 1 st day of estrogen treatment (figure3a). After that the number of preserved neurons with clear integrated nissl granules was increased (figure3c). Only small numbers of necrotic neurons were still noticeable and the edematous or vacuolated neuropil wasn't detected (figure3b

Quantitative real time PCR
In the KA group the up-regulation of caspase-3, GFAP, Iba-1 and IL-1β expressions was significantly visible (P-Value<0.05) on the 1 st day of injury (1.6 fold, 2.19, fold, 1.7 fold and 2.4 fold changes) respectively till the end of the study (2.6 fold, 6.8 fold, 2.7fold and 3.4 fold changes) respectively. In the estrogen treated group the expression of caspase-3 activity and GFAP at one day (2 fold, 6.37 fold) respectively was significantly higher than the KA group (P-Value<0.05). Administration of estrogen led to obvious and significant (P-Value<0.05) down regulation of caspase-3 activity (1.02 fold) and GFAP expression (0.19 fold change) at 6 th week.
The down regulation of Iba-1 (0.44 fold change) and IL-1β (1.07 fold change) after estrogen treatment was significantly detected at 1 st week (P-Value<0.05) till the end of the study (0.36 fold change for Iba-1 and 1 fold change for IL-1β).

Immunohistochemical assessment for the engrafted h-GFP-OBNSCs
GFP was mainly used to trace the transplanted OBNSCs, and to differentiate between exogenous (our engrafted), and endogenous (non-human, non-GFP-positive) neuronal and glial elements.
Immunohistochemical assessment was performed to evaluate the ability of the engrafted

Locomotor Function Assessment (BBB score)
The data obtained from all groups were statistically analyzed by the one way ANOVA test.
Assessment of hind limb motor function displayed no difference between the control and sham control groups in whole time of the study.
The average BBB score for the KA excitotoxic group, estrogen treated group, and estrogen+ OBNSCS group are collectively shown in table (3)

KA excitotoxic group
The BBB scale showed that KA excitotoxic injury caused bilateral locomotor deficits in hind limbs that persisted to the last day of the study. In comparison to control or sham control groups (BBB score 21, exhibiting normal locomotion), the average BBB score at the 1 st and 2 nd week after injury was 12 (frequent to consistent plantar step with both hind limbs and occasional coordination between fore and hind limb). From 3 rd to 5 th weeks, the animals were unable to coordinate fore and hind limbs during locomotion (average BBB score 11). At the 6th week the average BBB score was decreased again. The rats were unable to make frequent or consistent 13 plantar steps with both hind limbs and there was no coordination between fore and hind limbs (final average BBB score 10). This indicated the success of the SCI model establishment.

Estrogen treated group
The average BBB score gradually increased after estrogen injection. In comparison to the KA group, the improvement in locomotor function was significant (P-Value<0.05) at the 3rd week of estrogen injection (average BBB score, 12). The average BBB score at the 4th week was 13, after that there was no further improvement. Functionally, this score indicated that the estrogen treated rats were frequently supporting their own body weight with plantar stepping and frequent coordination of forelimbs and hind limbs.

Estrogen + OBNSCs group
Assessment of hind limb motor function in the estrogen +OBNSCs group revealed consistently higher BBB scores than those of the estrogen treated group. In comparison to the KA groupthe improvement in BBB score became significant (P-Value<0.05) two weeks after OBNSCs transplantation (earlier than the estrogen group). By the 3rd week, the combination of estrogen and OBNSCs therapy resulted in a two-point improvement on the BBB scale compared to the estrogen group. The final average BBB score at the 6 th week was 17. This BBB value means that rats in this group were able to move both hind limbs with consistent plantar stepping and consistent fore/hind limb coordination. Toe clearance occurred frequently during forward limb advancement.  Table(4): showing the BBB score for KA, estrogen and OBNSCs+estrogen groups there is no significant difference between three groups at 1 st week(P-Value>0.05). At 2 nd week there is a significant improvement(P-Value<0.05) in OBNSCs+estrogen compared to KA and estrogen groups .After that there is a significant improvement in estrogen and OBNSCs+estrogen groups compared to KA ,but this improvement was more pronounced in OBNSCs+estrogen group than estrogen group.

Tissue architecture following KA injury and estrogen treatment
The results of the present work have shown that the histopathological changes following KA

real-time PCR (q RT-PCR)
15 Examining the changes in mRNA expression after KA and estrogen injections is a powerful tool to correlate the pathophysiology of KA excitotoxicity and the histopathological findings. In the current study quantitative real-time RT-PCR with a fluorescent TaqMan probe was used to measure alterations in individual gene expression after SCI, particularly genes involved in apoptosis (caspase-3), astrocytosis (GFAP), inflammation (IL-1B) and microgliosis (Iba1). These genes were selected since they represent various cell types that are related to essential pathological processes that happen after SCI.
Caspase-3 activation has been confirmed during the process of secondary injury after SCI In our study, the expression of mRNA for IL-1β was significantly increased after KA injection.
Our results are also those obtained by previous investigators, who stated that primary cultures of microglia were significantly stimulated by KA, and showed increased IL-1β levels ( Astrocytes represent main target cells for estrogen in the CNS. The astrocyte also has the greatest potential to induce the neuroprotective effect of estrogen (Dhandapani and Brann, 2003). Our approach revealed that the expression of GFAP after estrogen therapy was variable.
At the 1st day post estrogen treatment, the expression of GFAP was markedly up-regulated compared to the KA excitotoxic group, but after that the expression was significantly decreased to levels nearly indistinguishable from control group. This result is in harmony with results of previous studies which revealed that a high dose of 17β-estradiol delivered immediately in rat after SCI and continued for 21 days, enhances the immunoreactivity of GFAP and vimentin followed by decrease of this immunoreactivity at 50 days after injury ( Ritz and Hausmann, 2008). This dramatic increase of GFAP on the 1st day is attributed to the increase of astrocyte numbers as the main target for estrogen to induce the neuroprotective effect and attenuate the inflammatory response after KA injury. The resolution of inflammation after that, resulting in decreases in the number of astrocytes was accompanied by steady attenuation of GFAP expression until it reached levels nearly indistinguishable from the control group.

Ogawa et al. (2002) mentioned that adult spinal cords are not completely non neurogenic for
transplanted NSCs. Rather, a brief therapeutic time window can permit successful implantation.
This narrow window may change since the host spinal cord's microenvironment quickly changes following injury.
But the chronic phase of SCI isn't likely to be suitable for stem cell implantation because of lack of inducing factors for neurogenesis, glial scarring, or enlarged cyst formation, any or all of which might inhibit the survival of engrafted cells and neuronal regeneration (Okano, 2002).
Most of transplantation studies after SCI are done within 1-2weeks post injury when the potential for cell replacement is expected to be optimum (Okano, 2002). Based on this we have

Immuno-histochemical assessment of hGFP-OBNSCs
Long-term survival of engrafted cell populations is very important to researche that demonstrates functional recovery correlated with cell replacement and/or incorporation within the host circuitry, as in the present study. Previous studies revealed that successful engraftment in transplantation strategies of the CNS is very difficult to achieve, particularly for xenografts.
Moreover, many studies failed to carry out suitable control of host immunorejection, making the evaluation of these parameters hard to do (Salazar et al., 2010). In that regard the modulation of an immune response with immune-suppressant medications (such as tacrolimus or cyclosporine, used in the current study) was critical to avoid host-mediated rejection against h-GFP-OBNSCS xenografts (Hooshmand et al., 2009).The present work exhibited that the h-GFP-OBNSCs were able to survive for 6 weeks after subacute transplantation into injured spinal cord.
Furthermore, no tumors or signs of immune rejection were recorded during the current study.
Engrafted cells not only migrated for considerable distances in the spinal cord, they also showed a tropism toward the lesion site. This result agrees with our previous results revealing that all transplanted rats exhibited successful engraftment with h-NGF-GFP-OBNSCs after more than 9 weeks without tumor formation or abnormal morphology in the spinal cord (Marei et al., 2016).
Similar results were also obtained after sub-acute transplantation of pre-labeled human embryonic stem cell-derived oligodendrocyte progenitor cells (

Motor activity Assessment
The locomotor deficits seen following KA injection are attributed to degeneration of primary Collectively, these data suggest that the combination of estrogen and hOBNSCs promotes a compensatory mechanism to reorganize neural networks in the central pattern generator for locomotion by generating mature neurons, oligodendrocytes and astrocytes.