Lipid determination in bone marrow and mineralized bone tissue: From sample preparation to improved high-performance...

In view of their key roles in the bone physiology (e.g., in the biomineralization process) and their potential implication in bone pathologies, an approach to study lipids in situ is needed. The aim of the present paper is to propose an original procedure to characterize lipids in both bone marrow (BM) and mineralized tissue (MT) compartments, taking into consideration sample preparation, lipid extraction and analytical issues, when using small sample size (≤ 0.5g of rat femurs). The potential contamination of the MT by marrow lipids and the poor accessibility of certain lipids from the MT - two major issues in bone handling - were taking care, respectively by performing two cleaning steps after BM removal and by adding a demineralization step to the overall lipid extraction protocol. For lipid analyses, a multi-one-dimensional HP-TLC method was developed to analyze the major neutral and polar lipids at once and showed an excellent resolution (for 15 standards) and a good precision (inter-day RSD<13%). When subjected to the entire "lipid extraction-HP-TLC" protocol, spike recoveries of the standards ranged between 76 and 122%. This HP-TLC method was suitable for lipid determination in both BM and MT [e.g., the MT had 5-times lesser lipids and a lower TG/phospholipid ratio than the BM (P <0.05)], and was quite reliable in term of lipid quantification. The demineralization step allowed to extract additional phosphatidylserine and esterified cholesterol from the MT, suggesting that these two species were associated to the mineralized matrix possibly in relation to their physiological role in the bone. Moreover, a reverse phase HPLC method for fatty acid determination as naphthacyl esters was set up to study fatty acids in bone samples and was used to validate the HP-TLC data. The fatty acid profile of the MT exhibited lower linoleic acid (18:2 n-6) and linolenic acid (18:3 n-3+n-6) levels and higher arachidonic acid (20:4 n-6) and docosahexaenoic acid (22:6 n-3) levels (P<0.05, compared to BM), suggesting that the MT is more metabolically active than the BM in term of long chain fatty acid production. In sum, the present work should contribute to facilitate future studies in the bone lipid field in view to understand better their implication in the marrow fat expansion-associated bone pathologies, such as osteoporosis.