Subjects, animal care and surgery
Thirty six adult female Long Evans rats (275 ± 25g; Simonsen, Gilroy, CA) were separated into six groups. These animals were housed (12:12 h light:dark cycle) and treated in accordance with the Laboratory Animal Welfare Act, and with Guidelines / Policies for Rodent Survival Surgery (Animal Studies Committee of Washington University in St. Louis). All animals received FES implants. ‘Short term’ refers to analysis 15 days after FES device implantation, 2h after the last BrdU injection; ‘long term’ refers to analysis 22 days after FES device implantation, 7 days after the last BrdU injection. The six groups were:
- SCI, NO FES activation, short-term survival (n = 6)
- SCI, FES activation, short-term survival (n = 6)
- SCI, NO FES activation, long-term survival (n = 6)
- SCI, FES activation, long-term survival (n = 6)
- NO SCI, NO FES activation, short-term survival (n = 6)
- NO SCI, FES activation, short-term survival (n = 6)
All SCI animals were fitted with 10.5 cm plastic collars to prevent autophagia (Ejay, Glendora, CA) and had their bladders expressed 3 times a day or until recovery of reflex urination. All rats were handled for 5 minutes per day and housed individually with absorbent bedding (ALPHA-Dri™, Shepherd, Kalamazoo, MI).
Spinal cord injury
Rats subjected to SCI (groups 1-4) were anesthetized (75 mg/kg Ketaset®, 0.5 mg/kg Domitor®, i.p.) and laminectomy was performed at T8-T10. A 1-mm section of the spinal cord at T9 was removed through a dural slit, using a BARON® suction tube (Roboz, Rockville, Maryland). The dural opening was covered with fascia, and the muscle and overlying skin were closed with layered sutures. Anesthesia was reversed with Antisedan® 1 mg/kg. This ‘suction ablation’ injury model helps maintain the integrity of the dura and major blood vessels, and therefore produces less bleeding and secondary ischemia than traditional blade transection injury.
Stimulator and electrode implantation
All animals were anesthetized as described above. Twenty-one days after SCI in groups 1-4, a 2-channel battery-powered electrical stimulator (Dr. J.C. Jarvis, University of Liverpool) (Salmons et al., 1991
) was implanted into each animal (). In groups 5-6, an FES device was implanted into each of the uninjured rats. To insert the FES device, an incision was made in the skin of the lower back at the midline from L1 to L5. A subcutaneous pocket was created around the incision and the FES device was inserted into it. A 0.5 cm incision was made on the lateral aspect of both legs overlying the common peroneal nerve. The stainless steel wire stimulating electrodes were tunneled bilaterally underneath the skin from the stimulator site to the incision in the leg. The wire electrodes were sutured into the tibialis anterior muscle, adjacent to the common peroneal nerve. The return electrode was sutured near the midline at the L4 paraspinal muscles. Intraoperative test stimulation was performed to ensure that the peroneal nerve was being activated during stimulation, as indicated by alternating flexion of the right and left hindlimbs.
Electrical stimulation paradigm
Three days after FES implantation the devices were activated in groups 2, 4, and 6. The FES unit was activated three times daily, for 1 hour at a time, between the hours of 8:00 AM and 5:00 PM (). The FES pattern consisted of 1 s stimulation of one common peroneal nerve followed by 1 s of rest; then the other common peroneal nerve was stimulated for 1 s followed by 1 s of rest and the cycle was repeated. Stimulus pulses were monophasic, 3 V, 200 μs long, and were delivered at 20 Hz. This configuration of stimulation preferentially activates large myelinated fibers within the common peroneal nerve (Gorman et al., 1983
; Grill et al., 1996
), and produced alternating flexion of the hindlimbs that crudely approximated bilateral stepping.
Bromodeoxyuridine injection paradigm
Beginning ten days after FES implantation, rats received daily injections of BrdU (50 mg/kg i.p.) for five consecutive days. BrdU labels new cells by incorporating into replicating DNA (Dolbeare 1996
; Horner et al., 2000
). To determine the amount of cell birth, groups 1, 2, 5, and 6 were sacrificed at the completion of BrdU injections. To determine the ability of newborn cells to survive in vivo
, groups 3 and 4 were sacrificed 7 days after the completion of BrdU injections.
Tissue processing and immunohistochemistry
Rats were perfused intracardially with 0.1M PBS for 5 min followed by 4% paraformaldehyde (Sigma) for 15 min. We selected every 6th section (40 μm frozen) from spinal cord levels C2, T1, T7, T11, L1, and L5 for anti-BrdU immunohistochemistry (Horner et al., 2000
). The sections were incubated in 2 N HCl for 60 min at 37°C, transferred to 0.1M borate buffer (pH 8.5) for 20 min, and rinsed with PBS. Nonspecific labeling was blocked with 0.1% BSA in 0.1% Triton X-100/PBS for 60 min. A mouse monoclonal anti-BrdU antibody (1:600; Roche, Mannheim, Germany) was incubated with the tissue overnight at 4°C. Then the tissue was treated with a CY3-conjugated secondary antibody (1:2000; Jackson, West Grove, PA) in 2% normal goat serum (NGS) for 60 min.
To co-label the BrdU+ cells, we fixed the sections with 4% paraformaldhyde for 30 min after applying the CY3-conjugated antibody. Sections were permeabilized, as needed, with 0.1% Triton-X-100 for 60 min and blocked with 2% NGS for 60 min. Primary antibodies diluted in 2% NGS were applied to the sections for 2 h. Antibodies included: rabbit anti-NG2 (1:250, Chemicon, Temecula, CA), mouse anti-Nestin (1:8, Developmental Studies Hybridoma Bank - DSHB), rabbit anti-GFAP (1:4, Diasorin, Stillwater, MN), mouse anti-APC-CC1 (1:20, Oncogene, Cambridge, MA), mouse anti-ED1 (1:100, Serotec, Raleigh, NC), mouse anti-OX42 (1:100, Serotec), sheep anti-Glut-1 (1:30, Biodesign, Saco, MN), rat anti-CD31 (1:20, BD Pharmingen, Lexington, KY), mouse anti-NeuN (1:200, Chemicon), mouse anti-TUJ1 (1:200, Babco, Richmond, CA), guinea pig anti-doublecortin (1:3000, Chemicon), and mouse anti-PSA-NCAM (1:8, DSHB). Secondary antibodies (1:300, Molecular Probes, Eugene, OR) were conjugated to Alexa 488. Primary and secondary control slides were included with each stain series to control for secondary cross-reactivity and non-specific labeling. Note that interpretation of immunological phenotypic assessment in vivo should be tempered by the fact that such analyses have largely been deciphered in vitro.
In a separate procedure, slices from three rats from each of the cell birth groups were stained with Hoechst 33342 (1:100,000, Molecular Probes, Eugene, OR) as an index of overall cell density.
Confocal microscopy and quantification of BrdU- and neural phenotype-positive cells
At each spinal level sampled, we quantified the number of 1) BrdU-labeled cells as an index of ‘newly born cells’; 2) Hoechst 33342-labeled nuclei as an index of ‘total cells’; and 3) BrdU-labeled cells co-labeled with the phenotypic markers noted above. Scanning laser confocal microscopy was used to identify doubly-labeled cells. Object counts in 40 μm frozen sections were completed using an unbiased optical fractionator procedure (Peterson 1999
; West et al., 1991
), with the assistance of a semiautomatic stereology system (Stereoinvestigator™, Microbrightfield Inc., Brattleboro, VT) driving a Ludl X-Y motorized stage (Ludl Electronics Products, Ltd., Hawthorn, NY). Fluorescent Images were acquired with a Magnafire® camera. Rough boundaries of each cord cross-section were drawn under low magnification to delimit the optical fractionator areal sampling fraction. Samples were then marked under higher magnification (20×, UplanApo, 0.8 N.A.). The software randomly superimposed a sampling grid with an unbiased counting frame (x = 122 μm, y = 110 μm) and moved in a raster pattern to provide 12-15 counting frames within the cross-section of the cord. The top of the section was marked and a guard focus height of 2 μm was set in the software. The optical thickness of each image was approximately 0.7 μm and the total optical disector height was 20 μm.
Cells meeting the sampling criteria were marked interactively during the session. For each BrdU+ or Hoechst+ cell, the complete cell nucleus was followed through the z-axis, and only cells with a well-circumscribed, labeled nucleus and/or immunopositive cell body contained entirely within the counting volume frame were considered positive. Based on the numbers of marked cells and the volume of the slices, the software calculated the total numbers and densities of BrdU+ nuclei and Hoechst+ nuclei. Six slices per spinal level were quantified per animal in each procedure. Each animal's mean value was then calculated, and the final N used in each statistical analysis.
To avoid the confounds associated with immunolabeling cells within and immediately surrounding the SCI site, we performed our analyses on spinal levels more distant from the injury, where tissue necrosis and immune cell infiltrates are minimized, and where the total numbers of cells and cord circumference are unchanged from uninjured conditions.
Comparison between experimental groups was conducted by two-way ANOVA using SigmaStat® and applying Tukey post-hoc tests. Comparisons were made for the total number of BrdU+ cells and phenotypic markers between groups and within groups by spinal cord level. For all analyses significance was accepted at p <0.05 with Bonferroni correction for multiple tests.