The following study was approved by the Institutional Animal Care and Use Committee at Johns Hopkins University School of Medicine (RA14M347). Experimental animals were housed in a specific pathogen-free facility and fed a standard diet. All animals used in this study were 6-9-week female Lewis rats (100-180 g).
Single-level bilateral posterolateral intertransverse process lumbar spinal fusion surgery was performed, as described previously [19,20], on 53 host rats divided into three experimental groups: [A] iliac crest autograft (n=16); [B] syngeneic iliac crest allograft (n=19); and, [C] syngeneic femoral allograft bone (n=18). Briefly, host rats were anesthetized via intraperitoneal (IP) injection of ketamine (36 mg/kg) and xylazine (4 mg/kg). The surgical site was shaved and prepped with 70% ethanol and povidone-iodine and sterile gloves and masks were used by all surgical personnel. The surgical procedure was performed using an operating microscope or surgical loupes at 2.5 x to 10 x magnification. The L4 to L5 vertebral levels were identified by palpation and anatomical landmarks. A dorsal midline skin incision was made centered over the L4-L5 spinous process, and a self-retaining retractor was utilized to retract skin edges. Two paramedian fascial incisions were then made through the lumbar fascia. The intermuscular plane was established between the multifidus and longissimus muscles to expose the transverse processes of L4 to L5 as well as the inter-transverse membrane. Decortication of the transverse processes and lateral pars/facet joints was performed with a motorized burr. The appropriate graft was placed over the entire fusion bed space (L4 to L5) on each side of the spine.
In the autograft group [A], the midline incision was extended to the sacrum and dissection was taken to the iliac crest on each side, and following subperiosteal dissection over the posterior aspect of the iliac crest, a small volume of corticocancellous autograft was harvested with a rongeur. In the syngeneic allograft groups, hips (the ilium down to the acetabulum) [B] and femurs [C] were sterilely isolated from freshly euthanized littermate donors and placed on ice, immediately prior to fusion surgery. In most cases, harvested bone grafts were weighed on a sterile scale. In all cases, isolated bones were morselized with a rongeur prior to implantation to create a homogenous distribution of corticocancellous graft material over the fusion bed.
Fascia and skin were closed in layers with 5-0 absorbable sutures (Polysorb, Medtronic, Minneapolis, MN, USA). Normal saline was administered intraperitoneally after wound closure as needed. Rats were maintained on a heating pad until spontaneous ambulation was observed. Buprenorphine (0.01mg/kg) was administered subcutaneously every 24 hours for two days. After surgery, rats were closely monitored for any sign of nerve palsy, hemiparesis, or infection as well as any changes in general condition. All animals were euthanized eight weeks post-surgery and spines were harvested.
Harvested lumbar spines were imaged using a nanoSPECT/CT Small Animal Imager (Mediso Medical Imaging Systems, Budapest, Hungary). The coronal CT images were evaluated by two authors in a blinded fashion. Each fusion mass (i.e. two per animal, one on each side) was graded as either fused (i.e. robust fusion between L4-5 was observed), partially fused (i.e. some narrowing of the fusion mass between L4-5 was present), or non-fused (i.e. there was a discontinuity of fusion mass between L4-5). Axial cross sections were also generated to quantitatively calculate fusion mass volume via ImageJ software (US National Institutes of Health, Bethesda, MD) and the Volumest plugin as previously described [20–22].
Soft tissue surrounding the L3-L6 region of harvested spines was gently removed. Two blinded independent observers manually palpated the fusion site (L4–L5) to evaluate biomechanical fusion. Palpation was scored as either bilaterally fused (i.e. no segmental motion compared to the adjacent levels), or non-fused (i.e. similar segmental motion across the index level compared to the adjacent levels).
Harvested spine samples were fixed in 4% paraformaldehyde overnight, decalcified in Rapid Bone Decalcifier (American MasterTech Scientific, Lodi, California) for 8 hours, dehydrated by ethanol series (70%, 95%, 100%), and embedded in paraffin. Serial coronal sections (10 µm thick) across the level of the fusion masses were cut, deparaffinized in xylene, and subsequently rehydrated in a descending ethanol series (100%, 95%, and 70%). Hematoxylin and eosin (H&E) staining and Masson’s trichrome staining were conducted to evaluate the formation of bone, cartilage, and osteoid.
Bone Marrow Cell Isolation and Culture
Nine donor rats were euthanized and femurs and ilia were isolated, dissected and cleaned in a sterile biological safety cabinet. Bones were cut into smaller pieces, crushed using a sterile mortar and pestle, and washed several times in chilled culture medium (Dulbecco’s Modified Eagle Medium (DMEM; high glucose; Gibco, USA) supplemented with 10% Fetal Calf Serum (FCS) and 1% (v:v) Penicillin–Streptomycin) to isolate bone marrow cells. The resulting cell suspension was passed through a 100-μm nylon mesh filter and subsequently underwent hemolysis and centrifugation. Recovered nucleated cells were enumerated, re-suspended and either used in limiting dilution assays or cultured at 37°C in humidified air with 5% CO2. The first media exchange was performed ∼72 h after plating, with subsequent media exchanges every 2–3 days.
Limiting Dilution Colony - Forming Unit Fibroblast (CFU‐F) Assay
The frequency of mesenchymal progenitor cells within hip-derived and femur-derived bone marrow was determined via limiting dilution colony‐forming unit fibroblast (CFU‐F) assays. Briefly, cultures were initiated with freshly isolated bone marrow cells at densities of 3 × 105, 1 × 105, 3 × 104, 2.5 × 106, and 5 × 106 cells/well in 96/well plates in culture media, with 10 replicate wells per density per rat (n = 7 for femur, n = 6 for hip). On day 10 of culture, plates were fixed with methanol and stained with Crystal Violet. The total number of wells containing at least 1 fibroblastic CFU‐F colony of 10 or more cells (representing at least three population doublings) at each density was enumerated using microscopic observation. CFU-F frequencies were calculated using the online ELDA tool available at http://bioinf.wehi.edu.au/software/elda/ .
Osteogenic Differentiation Assay
To assess the osteogenic differentiation capacity of hip-derived and femur-derived bone marrow, passage 1 (P1) cells from 8 donor rats were seeded in triplicate at a density of 50,000 cells/cm2 in either osteogenic media (DMEM (low glucose; Gibco, USA) supplemented with 10% FCS, 1% penicillin–streptomycin, 10 mM β-glycerophosphate (Sigma), and 50 μM L-ascorbic acid-2-phosphate) or in standard culture media (as a negative control) and cultured for 21 days. Mineralization was assessed via Alizarin Red S staining. Briefly, samples were washed twice with PBS, fixed with 3.7% formaldehyde for 20 min, washed again three times, subsequently incubated for 10 min with 40 mM Alizarin Red S (Sigma) and then washed extensively before imaging. Subsequently, Alizarin Red S was eluted from stained cultures for quantification via incubation with 10% acetic acid at room temperature with shaking for 30 min, followed by cell scraping and transfer to Eppendorf microcentrifuge tubes. The resulting samples were vortexed for 30 sec, heated at 85°C for 10 min, incubated on ice for 5 min, and subsequently centrifuged at 20,000 x g for 15 min. 200 ul of the resulting supernatant was transferred to a new microcentrifuge tube and neutralized with 75 ul of 10% ammonium hydroxide. One hundred fifty microliters was transferred to a 96-well plate, and absorbance was measured at 405 nm using a microplate reader (PerkinElmer VICTOR3).
In the case of three experimental groups (e.g. comparison of graft weights, CT volumes), intergroup comparisons of continuous variables were performed via one-way ANOVA (parametric) or Kruskal-Wallis tests (non-parametric), while categorical data sets were tested via the χ2 test. In the case of two experimental groups (e.g. comparison of cell yield, Alizarin Red staining), intergroup comparison was performed via paired Student’s T-Test. All reported p values are 2-sided and p values < 0.05 were considered to be statistically signiﬁcant. All statistical analyses were performed using GraphPad Prism 6.0 (La Jolla, CA, USA) with the exception of analysis of CFU-F frequency, which was performed using the online ELDA tool available at http://bioinf.wehi.edu.au/software/elda/ .