The characterization of the underlying population level transcriptomic changes of bone tissue and cells that accompany age associated decreases in BMD, may provide considerable insight of the molecular mechanisms associated with osteoporosis. Two major technical challenges to obtaining transcriptomic data from human bone tissues and cells are related to the inherent difficulty of purifying RNA from bone tissues and the variability of isolated cell populations from bone.

In this study, these two separate technical problems were addressed. The first problem that was addressed was the difficulty of obtaining un-degraded RNA from human bone samples once they have been removed from the patient. In order to assess the best way to extract RNA from a bone sample, 6 different conditions were run on samples from 5 patients undergoing total hip arthroplasty at Boston Medical Center. The six conditions compared different methods of inhibition of RNase, storage of bone tissues after removal for later purification, and the purification procedures. The methods included same day extraction followed by direct glass filter purification (D2C), processing in QIAzol the same day (NSD), processing in QIAzol at a later day after saving the bone by flash freezing in liquid nitrogen (NLD), saving ground bone in QIAzol for processing at a later day by flash freezing in liquid nitrogen (QLD), saving whole bone chips in RNAlater to be processed in QIAzol on a later day (RLWB), and saving ground bone chips in RNAlater to be processed in QIAzol on a later day (RLGB). RNA quality was assessed by capillary electrophoreses with calculations of RIN values, total yields of RNA from comparable amounts of starting bone tissue and qPCR and assessment of CT values for 18S ribosomal RNA and COL1A1 genes. It was found that overall it is best to process the bone on the same day using QIAzol (NSD). If the bone cannot be processed the same day, then it was best to powder the bone in liquid nitrogen and save it in QIAzol (QLD) as this had a better outcome than any other storage methods.

The second question assessed was related to the consistencies of osteogenic cellular phenotyping following the marrow stromal cell osteoinduction. For this study, osteogenic induction was carried out on marrow stromal cells obtained from 10 patients undergoing total hip arthroplasty at BMC. These studies compared the cell culture mineralization and alkaline phosphatase (ALP) levels to collagen production. Five samples that showed good mineralization and five that showed poor mineralization were chosen to have their RNA extracted and assessed. For each patient qPCR was run to determine COL1A1 levels. The levels were normalized to the levels of 18S, a house keeping gene. It was found that ALP and mineralization levels greatly corresponded to COL1A1 gene expression. Cultures that had good mineralization also showed increased ALP levels and greater expression of COL1A1. Meanwhile, cultures that had poor mineralization had low ALP levels and had lower expression of COL1A1.

Overall, our study demonstrates that the best method to extract RNA from human bone is to process it the same day in QIAzol (NSD) or if same day processing is not possible to grind in liquid nitrogen and save it in QIAzol (QLD). It also demonstrated that COL1A1 gene expression corresponds to ALP and mineralization levels. Through these findings, two issues with obtaining transcriptomic data from human bone tissues and cells have been addressed. Further studies with larger sample sizes are needed which will provide more statistically significant conclusions and will allow for further genetic analysis.