Adult (180-200 g) female Fischer 344 rats (Animal Resource Centre, Murdoch, Western Australia) were housed in groups of three under controlled conditions of temperature and humidity, with free access to food and water. All procedures were approved by the Animal Ethics Committee of The University of Western Australia (approval number 99/008/E42). Fourteen days after injury, previously described injection procedures  were used to transplant OEG (n=6 short term 2 weeks, and n=11 long term 4 months) or SCs (n=6 short term 2 weeks and n=11 long term 4 months) into the lesion site controls received an injection of cell culture medium (n=6 short term 2 weeks and n=11 long term 4 months) or injury alone (n=6 short term 2 weeks and n=11 long term 4 months). Six additional rats were perfused 2 weeks after contusion without a transplant, to examine the lesion milieu into which transplanted cells would have been introduced. Total rats used n=74. All rats receiving a transplant and/or medium were perfused at 2 weeks or 4 months after transplantation. The numbers of animals were final numbers analyzed. In the long term groups 3 animals died and in the short term groups 6 rats.
Olfactory ensheathing glia and Schwann cell culture and purification
OEG were prepared from the olfactory bulbs of adult female Fischer 344 rats [8, 13, 33]. In brief, the bulbs were extirpated and the olfactory nerve layer dissected. Tissue was mechanically dissociated and treated with trypsin 0.5% w/v (Worthington Biochemical Corporation, Lakewood, NJ) in Hank’s Balanced Saline Solution (Sigma) for 60 minutes at 37°C in a CO2-free incubator. Tissue was pelleted by centrifugation at 1500 rpm for 5 minutes, resuspended in DF10S medium (1:1 DMEM/F-12 -Sigma), 10% (v/v) fetal calf serum - Hunter Antisera) supplemented with bovine pituitary extract (20 μg/mL Gibco) and forskolin (2 μM Sigma), then plated on poly-l-lysine-coated culture dishes (Corning, Acton, MA) for one week. OEG were purified by immunoaffinity to an antibody to the p75 receptor (192-IgG) [13, 33].
SCs were isolated from adult female Fischer 344 rats [19, 33]. Briefly, the sciatic nerves were dissected, placed in 35 mm uncoated tissue culture dishes in D10S medium consisting of DMEM (Sigma St Louis, MO) with 10% fetal calf serum (Hunter Antisera Jesmond, NSW, Australia) and cut into 1-2 mm3 pieces. Each week, nerve segments were transferred to a new culture dish, leaving behind fibroblasts that had migrated out of the explants. After three weeks of fibroblast depletion, explants were enzymatically and mechanically dissociated and transferred to new dishes for culture in the presence of bovine pituitary extract (20 μg/mL Gibco, Carlsbad, CA) and forskolin (2 μM Sigma) .
Lentiviral labelling of glia
Lentiviral vectors encoding the gene for DSRED-2 were made according to methods described by . Cultured OEG or SCs were labeled by adding the vector (using a multiplicity of infection of 100  to the culture medium for 16 hours, followed by 2 washes in culture medium and fresh media for growth.
Spinal cord injury and post-operative care
Anesthesia was induced with 5% (v/v) halothane (in 60:40 O2:N2O) then maintained at 2-3%. The skin and muscle layers were cut rostro-caudally to expose the vertebral column, and a T10 laminectomy performed  Using an New York University (NYU) impactor, a moderate contusion injury was induced by dropping a 10 g weight from a 12.5 mm height [99, 105] Gel-Foam (Pharmacia & Upjohn, Kalamazoo, MI) was used to stop bleeding in some animals. Animals were partially suspended from clamps to the dorsal processes of T8 and T11/12, to circumvent diaphragm-induced dorso-ventral movement of the spinal column. Immediately after the injury, the animal was removed from the clamps and the muscles sutured in layers with Vicryl (Johnson & Johnson, New Brunswick, NJ). Skin was closed with Michel suture clips (Fine Science Tools Inc., North Vancouver, BC, Canada). Immediately after surgery, each animal was injected with penicillin (Benacillin intramuscular), analgesic (Temgesic buprenorphine sub-cutaneous 0.1 mg/kg) and 2-4 mL 0.9% saline (sub-cutaneous). Animal cages were placed partially onto a heating pad for 24 hours, to assist with maintenance of body temperature. Antibiotics were continued on alternate days, with analgesic twice daily and saline once daily for one week. Manual bladder expression was performed twice daily for two to three weeks, by which time all animals had recovered bladder tone.
Cells were transplanted after no more than 2 weeks in vitro. Fourteen days post-contusion, adult female rats were anesthetized and the spinal cord exposed as described above. Either 5 μL of a cell suspension containing 5 × 105 OEG or SCs in culture medium, or medium alone, (1:1 DMEM/F-12 Sigma) was injected into the center of the lesion over a period of 3 minutes using a 5 μL syringe (Hamilton, Reno, NV) mounted in a stereotaxic frame. The needle was withdrawn 3 minutes later to encourage cell integration into the injection site. Wounds were sutured and post-operative care was administered as described above.
All testing of the BBB and ladder walk were carried out by observers blinded to the treatments. BBB testing  was performed on each animal on the day following contusion, to confirm the successful lesion (a score of 0). The BBB test was repeated on day 3, day 7 and day 14 after contusion thereafter every week after cell transplantation. 3 weeks post-injury, animals were consistently scoring BBB results of 10–11. A ladder rung walking test modified from  was introduced 1 week after injury and measured every week for 4 months. The ladder rung walking test provided 6 points (Table 2) for scoring, thus allowing a greater level of sensitivity than the two BBB points by which they were separated.
The ladder rung walking test  apparatus was 1.5 m in length and consisted of two clear Perspex sidewalls and a floor with removable metal rungs spaced 1 cm apart. Prior to surgery, rats were trained to walk along the ladder to reach their home cage. From 1 week after injury, trials were performed weekly and video recorded. The animals had 3 runs per day when tested. The video camera was positioned to capture the central portion of the ladder, approximately 30 cm in length, allowing analysis of approximately 8 steps of each hind limb. Each hind limb step in each trial was later scored (Table 2) the scores were averaged to give a single value, per rat, per trial.
Ten days prior to perfusion (4 months post-transplantation of cells), 11 rats from each of the long term (4 month) groups were anesthetized as above and subjected to a T12 laminectomy. While held in a stereotaxic frame, Fluorogold (1 μL Fluorochrome Inc., Denver, CO) was injected into the spinal cord at 4 points (0.25 μL each), 2 segments caudal (8 mm) to the contusion site. All injections were performed by the same individual, to maximize comparability between animals this individual was also blinded as to which groups had previously been injected with OEG, SCs or control cells. Muscles were sutured in layers and the skin closed with Michel suture clips (Fine Science Tools Inc., North Vancouver, BC, Canada). Animals were treated with the post-operative care regime detailed above.
At 2 weeks or 4 months after Schwann cell or OEG transplantation, rats were terminally anesthetized with Nembutal (intraperitoneal sodium pentobarbital Merial Australia, Parramatta, NSW, Australia) and perfused intracardially with approximately 150 mL of heparinized phosphate-buffered saline (Heparin 1 U/mL David Bull Laboratories, Melbourne, VIC, Australia) followed by approximately 150 mL 4% buffered paraformaldehyde solution. The spinal cord and brain were immediately removed and post-fixed in buffered paraformaldehyde overnight at 4°C. The following day, paraformaldehyde was replaced with 0.1 M phosphate buffer. Prior to tissue sectioning, each spinal cord or brain was placed into sucrose (30% w/v in phosphate buffer) overnight and embedded in gelatin (10% w/v in phosphate buffer Difco, Franklin Lakes, NJ). The gelatin block was fixed by immersion in buffered paraformaldehyde for 2 hours and placed into buffered sucrose solution overnight. Tissue was sectioned (spinal cord at 40 μm brain at 50 μm) using a freezing sledge microtome and placed into phosphate buffer containing 0.05% w/v sodium azide.
Tissue analysis - Quantification
Numbers of animals per group were 8–12 and based on: (i) a prior power analysis study to ascertain experimental numbers in animal models of spinal cord injuries, in order to appropriately assess statistical differences, and (ii) previous publications from the laboratory. All rats were tested with BBB, gridwalk, fluorogold traced, immunohistochemistry, and tissue analysis. These were randomly allocated to the groups without the knowledge of the surgeon or behaviorist. Animals were taken from the naïve cages and re-housed into post-surgery cages and numbered. This number was then combined by an independent person to identify the treatment. At every level the person was blinded to the treatment.
The perfusions were performed on the same day for all groups with the same fixative batch. The immunohistochemistry was also carried out between each of the groups reducing differences in staining between days and groups. The analysis was performed by a trained viewer who was blinded to the groups.
Tissue volume estimation
One in six sections from each spinal cord was mounted onto gelatin-coated microscope slides and stained with gold chloride solution for myelin followed by cresyl violet for Nissl substance . Slides were dehydrated through an ethanol/toluene series and mounted in DPX (Chem Supply, Gillman, Australia). Sections were visualized using a Leica DM RBE microscope and digitally photographed. Image Pro Plus (Media Cybernetics, Silver Spring, MD) was used to measure the tissue area of a 4.5 mm section of spinal cord with the lesion center at its midpoint.
Tissue cavitation was measured in all animals to gain an estimate of the effect of the transplanted cells upon the preservation of tissue at the lesion site the volume of tissue remaining at the injury site was measured in a similar manner to that described previously [8, 14, 108] The combined area of any cysts was calculated and subtracted from the total tissue area of the section, to give a value for the total tissue remaining. Areas showing signs of degeneration, such as microcysts and loss of neuronal profiles, were subtracted from the values for total tissue remaining to obtain a measure of intact tissue. The values for the centre four sections of each spinal cord were averaged to give a single semi-quantitative value of remaining tissue volume value for each rat. Values from four equivalent sections of an uninjured spinal cord were used to compute the tissue remaining as a proportion of uninjured spinal cord.
Series of one in six tissue sections (in 2 week or 4 month time points) were immunostained free-floating with mouse monoclonal or rabbit polyclonal antibodies. Sections were incubated overnight at 4°C in primary antibody diluted in phosphate buffer containing 10% normal goat serum and 0.2% (v/v) Triton X-100. Mouse monoclonal antibodies used were: i) anti-chondroitin sulfate proteoglycans (CS-56 1:50 Sigma) to identify proteoglycan deposition at the lesion site, ii) low affinity nerve growth factor receptor p75 (supernatant 192-IgG) to stain for OEG and Schwann cells (both transplanted and endogenous) and iii) anti medium/high neurofilaments RT-97 (supernatant Developmental Hybridoma Bank) to stain for spared and regenerated axons. Rabbit polyclonal antibodies were used for: i) p75 (1:200 Promega, Madison, WI), Glial fibrillary acidic protein (GFAP) (1:500 DakoCytomation, Denmark) to stain for astrocytes, unmyelinated Schwann cells and olfactory glia, ii) anti-collagen IV (1:200 Rockland Inc., Gilbertsville, PA) to ascertain the distribution of basal lamina and blood vessels and iii) anti-laminin-1 (1:200 Sigma) for basal lamina and blood vessels.
The following day, sections were washed three times with phosphate buffer and incubated with secondary antibody (diluted in the same solution as for primary antibodies) for 30 minutes at room temperature. Sections were washed three times with phosphate buffer, mounted onto gelatin-coated glass microscope slides, air dried and then coverslipped in a glycerol-based mounting medium.
Semi-quantitative analysis of proteoglycan, matrix and cell survival in vivo
One in 6 series of sections at 4 months post transplantation (and controls) were analyzed. Low magnification photomicrographs (10× objective) were taken using an IX70 Olympus microscope. Pixel fluorescence of immunostaining for CS-56, laminin and collagen IV was measured at rostral, middle and caudal locations of the contusion/injury sites and the average pixel fluorescence minus background readings of the three areas compared between medium control and OEG or SCs groups [37, 57]. In addition, one in 4 series of sections were analyzed for the presence of DSRED-2 labeled cells in the 2 week and 4 month survival groups. Pixel fluorescence was then compared between the 2 survival points to estimate survival of OEG and SCs at 4 months expressed as a percentage of those present at 2 weeks.
Fluorogold photography and counts
A 1 in 6 series of brain sections were mounted onto gelatin-coated glass microscope slides, air-dried and mounted in a glycerol-based mounting medium. Sections were photographed with a Hitachi HV-C20M digital camera as a large tiled image using a fluorescence microscope (Leica DM RBE) with a motorized stage controlled with Image Pro Plus (5.1) software (Media Cybernetics). The optical dissector method was used to photograph and count fluorogold-labeled neurons, in order to avoid inaccuracies due to double counting  Photographs were taken in the centre of the z-axis of each section. Those cell bodies in focus were identified, then tagged and counted using Image Pro Plus. Seven areas were analyzed: the reticular formation, vestibular nucleus, trigeminal and dorsal column nuclei, raphe nucleus, red nucleus, hypothalamus and motor cortex.
To quantify sparing/regeneration of descending propriospinal neurons, the labeled cell bodies were counted using an identical method in the spinal cord region rostral to the lesion site. Six spinal cord sections (1 cm in length) from each animal were analyzed.
The methods of statistical analysis were chosen that reflected the animal numbers used in this study. For Fluorogold labeling of supraspinal projecting nuclei in the brain, experimental groups were compared using a one-way analysis of variance (ANOVA) followed by the Dunnetts methods of multiple comparisons versus a single control group (injury only) and also the Tukey test for multiple comparison procedures between all experimental and control groups. Fluorogold analysis of propriospinal projections into the distal spinal cord were analyzed using a one way analysis of variance (ANOVA). This was also the case for the tissue sparing and proteoglycan and cell survival analysis. In the analysis of the BBB scores, groups were compared using the non parametric Kruskal Wallace ANOVA on ranks. This was followed by Dunnetts method of multiple comparisons versus a single control group.