313 UNSORTED, FRESHLY ISOLATED PORCINE ADIPOSE-DERIVED STEM CELLS ARE MORE EFFICACIOUS IN BONE HEALING COMPARED WITH PURIFIED CD34+ ADIPOSE-DERIVED STEM CELLS

We have previously shown that heterologous transplantation of porcine adipose-derived stem cells (ADSC) enhances bone healing. Freshly harvested ADSC are a heterogeneous population that contains several types of cells other than stem cells. The isolation of highly purified ADSC could be of clinical importance. In this study, we compared the in vitro growth characteristics and in vivo healing potential of ADSC unsorted or separated using CD34 as a marker. The ADSC were extracted from the back fat of 4 male pigs at 6 months of age. For the in vitro experiment, aliquots of the ADSC were sorted by magnetic beads (Miltenyi Biotec, Gladbach, Germany) into CD34-positive (CD34+) and CD34-negative (CD34–) cell populations. The unsorted ADSC (uADSC), plus the CD34+, CD34–, and a 50:50 mixture of CD34+ and CD34– (MIX) were plated in 24-well plates and differentiated into osteocytes. A robotic stage inverted microscope was used to photograph the entire culture well, and then number, dimension, and density of bone nodules were assessed. Alizarin red (AR) staining was performed and quantified. Cells were harvested before cell plating and then on several time points during expansion, at confluence, and at 3, 6, or 18 days [d] of differentiation for cell counting and RNA extraction. Real-time RT-PCR was performed for CD34, COL1A1, and SPARC genes. For the in vivo experiment, freshly isolated ADSC were sorted by flow cytometry into CD34+ and CD34– cell populations. Unsorted and sorted cells were transplanted, in duplicate, into 10- or 25-mm mandible osteoctomies. Mandibles were harvested after 8 weeks for evaluation of healing by DEXA scanning. In vitro data were statistically analysed using a mixed model (SAS) with time and cell type as fixed effect and pig as the random effect. The in vivo data were assessed by ANOVA with cell types as the fixed effect and pig as the random effect. Freshly harvested ADSC contained 42.3 ± 11.0% CD34+ cells. The uADSC reached confluence at 6 days after plating, whereas other cell types reached confluence at 16 days. Expression of CD34 decreased after plating but was similar between cell types. Among osteogenic genes, only expression of SPARC increased during differentiation. The number of osteogenic nodules was higher (P < 0.05) in uADSC than the in other cell types, but the area and nodule density were similar to CD34– and greater (P < 0.05) than CD34+ and MIX. The amount of AR was higher (P < 0.05) in uADSC compared with CD34– and MIX but similar to CD34+. In the in vivo trial, uADSC had a greater (P < 0.05) healing compared with sorted cells. Among those, CD34– cells appeared to increase healing compared with CD34+ cells. Results indicate that CD34+ cells do not differ significantly from CD34– in the in vitro osteogenic differentiation but have lower in vivo healing capacity; however, in vitro data were confused by a lack of pure CD34– cells. The freshly isolated ADSC have a greater healing capacity than sorted cells, as indicated by in vitro and in vivo experiments. Overall our data indicate that the sorting of ADSC CD34+ cells is not of clinical relevance.