Osteoclast Isolation: New Developments and Methods


To the Editor:

In a recent issue of the Journal, David et al.(1) have presented a method for in vitro generation of large numbers of osteoclasts from the bone marrow of newborn rabbits.

We have previously been informed of this procedure by Dr. David and Dr. Baron and have established the method in our laboratory, confirming all the findings described in their work, including 1,25-dihydroxyvitamin D2–induced up-regulation of genes relevant to the osteoclast function, i.e., c-Fos and c-Src. (A. Teti et al.,(2 unpublished observation). In addition to this, we have performed further observations and would like to provide additional information to those who are interested in establishing such a method in their own laboratory.

First, the method can be applied not only to rabbits but also to small rodents (mice and rats). The yield is high and allows the obtaining of adequate material for molecular studies from a limited number of animals. This would not be a substantial advantage for rabbits but can be relevant for small rodents. To give an example, from a purified osteoclast population obtained from 10 newborn mice, an average of 600 μg of cell protein can be extracted. Similar results can be obtained from rats. The method can be applied to adult animals as well, albeit with a lower yield. When cultures are established from the bone marrow fraction of adult rats, particular attention must be paid to the density of cells initially plated. To our surprise, in fact, excess cell numbers inoculated in culture resulted in a paradoxically poor cell viability. Therefore, if one wants to prepare osteoclasts from adult rats, a titration cell density curve is recommended to identify correct working conditions. We have successfully applied this method also to humans and obtained excess formation of dysfunctional osteoclasts in a case of osteopetrosis.(2)

A weakness of the method seems the impossibility of dissecting effects on osteoclastogenesis from those on osteoclast function. We have addressed this concern by applying a simple modification. This consists of a standard trypsin procedure to release and transfer the purified osteoclast population to appropriate substrates, where they can be treated with test agents. Substrates also include bone/dentin/ivory slices which are efficiently resorbed (Fig. 1), and bone resorption is so efficient that excess inoculation of osteoclast should be avoided. Once again a titration curve is recommended. This finding abrogates the dogma that osteoclasts are trypsin resistant. Although this requires further investigation, apparently the mononuclear late precursors are capable of escavating pits as well, with an efficiency similar to that exhibited by multinucleated cells. The procedure was found to be applicable to all the species tested.

Figure FIG. 1..

Bone resorption by trypsinized osteoclasts. Osteoclasts were generated in vitro from the adherent bone marrow cell fraction of newborn mice treated for 7 days with 10−8 M 1,25-dihydroxyvitamin D3. Osteoclast population was then purified by flushing off the stromal cells, trypsinizing, transferring to bone slices, and further incubating with vitamin D3 for 2 days. Cells were then removed by ultrasonication, and sections were stained with toluidine blue. Resorption pits are marked with an arrow. Magnification, ×400.

Applying this method, we have made interesting observations concerning divergences in the cell survival pattern of osteoclasts among species. It may be worthy to disclose this concept to allow correct experimental time-frame design. Rabbit osteoclasts develop in approximately 10 days, and their formation continues for several days. The mature osteoclasts survive weeks or months without apparent signs of phenotypic alterations. In contrast, mouse osteoclasts suddenly appear at day 6 or 7 of culture, dramatically increase up to day 10, then rapidly undergo apoptosis, leaving characteristics and easily recognizable “ghosts.” Therefore, if experiments are planned with mouse cells, this timing must be strictly considered.

In conclusion, David et al.(1 have presented an innovative method that, in our opinion, will substantially contribute to future advances in the understanding of important aspects of osteoclast biology.


The original work was supported by grants from Telethon (E.456), Associazione Italiana per la Ricerca sul Cancro (AIRC), Cofinanziamento Ministero dell'Università e della Ricerca Scientifica-Università dell'Aquila 1997, Hoechst Marion Roussel, Eli Lilly and Dompè S.p.A. Dr. Anna Taranta is a recipient of a fellowship from Telethon. Dr. Silvia Migliaccio is supported by an Eli Lilly grant.