Bone versus immune system

Article metrics

A molecule on activated T cells triggers bone loss by switching on bone-resorbing cells. Fortunately, it seems that this mechanism is kept in check by another molecule, secreted by the T cells.

The importance of bone to the immune system is well known: immune cells form in the bone marrow. But the importance of immune cells to bone is less clear. Little is known beyond the fact that normal bone growth and restructuring are disrupted in disorders such as autoimmune diseases. Several years ago, however, a particular molecule on the surface of activated T cells was found to activate bone-resorbing cells1. This raised a problem: T cells are working constantly to fight off the universe of foreign particles in which we live, so, at any point in time, some T cells are activated. What prevents these T cells from causing extensive bone loss? On page 600 of this issue, Takayanagi and colleagues2 provide an answer: T cells also secrete a molecule that inhibits the development and activation of the bone-resorbing cells. This finding underscores the dynamic relationship between bone and immune system.

Often thought of as a rigid, unchanging entity, skeletal bone is actually the result of a dynamic process involving the secretion and resorption of the bone matrix. These opposing actions are carried out by two cell types — osteoblasts and osteoclasts, respectively — and must be kept in balance to maintain skeletal integrity and calcium metabolism, as bone is the main source and repository of the body's calcium. The most common problem with this balancing act occurs when the rate of resorption exceeds the rate of mineral deposition. This results in a loss of bone mass, as seen in osteoporosis, many inflammatory diseases, such as rheumatoid arthritis, and many cancers3. One way of maintaining the balance is to keep in check the resorption of bone by osteoclasts.

Over the past few years, the molecular mechanisms underlying the maturation and activation of osteoclasts have been worked out from studies of genetic alterations resulting in skeletal abnormalities in mice, rats and humans. Osteoclasts are formed from bone-marrow-derived precursor cells in response to the coordinated actions of several protein factors. The study of one of these proteins, RANKL (also known as TRANCE, OPGL and ODF), and its binding partners has led to the tales of the immune system and bones becoming more closely intertwined. It seems that the cells of the immune system and the cells that remodel bone regulate each other's function.

RANKL is normally expressed on osteoblasts and activated T cells (Fig. 1). It binds to RANK, which is expressed on osteoclasts and dendritic cells1 (antigen- presenting cells needed for T-cell-dependent immune responses). When RANK on osteoclasts is activated it sends signals into the cells through 'adapter proteins'. One of these, TRAF6, allows RANKL to activate several signalling pathways (including the transcription factor NF-κB and the protein kinases JNK and c-Src)4. These signalling pathways allow osteoclasts to mature and develop their bone-resorbing functions. A receptor called OPG also binds to RANKL, and acts as a decoy to prevent it from binding to RANK and activating target cells5.

Figure 1: The interface between the immune system and bones.

During inflammation, activated T cells express more of the RANKL protein and secrete the interferon-γ molecule. RANKL is important for communication between T cells and dendritic (antigen-presenting) cells. It is also essential for the differentiation, activation and survival of osteoclasts — the cells that break down bone. It binds to RANK on osteoclasts to trigger intracellular signalling pathways that promote osteoclast differentiation and bone resorption. The TRAF6 protein is a key intermediate in these pathways. However, if osteoclasts were activated every time T cells were activated, bone loss would be more common than it is. Takayanagi et al.2 propose a solution: interferon-γ promotes the degradation of TRAF6, thereby preventing activated T cells from triggering massive bone destruction during inflammation. The effects of RANKL can also be blocked by OPG5, a soluble decoy receptor. Together, then, RANKL, interferon-γ and OPG maintain the balance between bone deposition and bone resorption. Note that RANKL is also found on other cells, such as osteoblasts.

So, RANKL is expressed on activated T cells, binds to RANK on osteoclasts, and thereby induces these bone-resorbing cells to mature and become active. Given all this, one might expect massive bone destruction under most inflammatory conditions — for example, in autoimmune diseases, cancer, allergy, infection and even, perhaps, injury. Indeed, RANKL on activated T cells can activate osteoclasts and trigger bone loss6. In arthritic rats, use of the decoy receptor OPG prevents bone destruction (but not inflammation)6. Moreover, in mice engineered to lack RANKL, overexpression of this protein on T cells restores some of the normal functions of osteoclasts7, suggesting that T cells are important in osteoclast function under non-pathological conditions. But why isn't bone loss more extensive? This is where Takayanagi et al.2 come in.

These authors have discovered a crucial counter-regulatory mechanism, by which activated T cells can inhibit the RANKL-induced maturation and activation of osteoclasts. Although the T cells involved in inflammatory conditions express RANKL, they also secrete another protein, interferon-γ. Takayanagi et al. have found that interferon-γ blocks RANKL-induced osteoclast differentiation in vitro. They also show that, in mice engineered to lack the interferon-γ receptor, the induction of inflammatory arthritis results in much more bone destruction than in normal mice. In a nutshell, it seems that interferon-γ — probably amongst other factors — prevents uncontrolled bone loss during inflammatory T-cell responses.

Takayanagi et al. suggest that interferon-γ activates the ubiquitin–proteasome pathway — involved in protein degradation — within the osteoclasts, resulting specifically in the degradation of the adapter protein TRAF6 (Fig. 1). They have several reasons for making this proposal. First, osteoclast maturation in the presence of interferon-γ can be 'rescued' by overexpression of TRAF6 in early osteoclast cells. Second, the suppressive effect of interferon-γ on osteoclast formation was markedly decreased in mice lacking proteasome components.

Third, signals downstream of CD40 — a protein that is important in regulating B cells and dendritic cells, and which also signals through TRAF6 — were also attenuated by treatment with interferon-γ. This result also suggests a broader role for interferon-γ. Moreover, TRAF6 is involved in yet more signalling pathways, such as those requiring lipopolysaccharide and interleukin-1 (ref. 8), which are proteins involved in the innate — rather than the adaptive — immune response. It will be interesting to find out whether interferon-γ has effects on these pathways, too.

The fact that there are complex regulatory interactions between bone-remodelling cells and immune cells has broad implications for researchers studying either system. We suggest that the term 'osteoimmunology' could be used to describe the interface between these disciplines. It is becoming clear that, without a better understanding of this interface, it will be difficult to prevent or treat many common diseases that affect both bones and the immune system.


  1. 1

    Wong, B. R., Josien, R. & Choi, Y. J. Leuk. Biol. 65, 715– 724 (1999).

  2. 2

    Takayanagi, H. et al. Nature 408, 600–605 (2000).

  3. 3

    Rodan, G. A. & Martin, T. J. Science 289, 1508–1514 (2000).

  4. 4

    Wong, B. R. et al. Mol. Cell 4, 1041– 1049 (1999).

  5. 5

    Hofbauer, L. C. et al. J. Bone Mineral Res. 15, 2– 12 (2000).

  6. 6

    Kong, Y. Y. et al. Nature 402, 304–309 (1999).

  7. 7

    Kim, N., Odgren, P. R., Kim, D. K., Marks, S. C. Jr & Choi, Y. Proc. Natl Acad. Sci. USA 97, 10905–10910 (2000).

  8. 8

    Lomaga, M. A. et al. Genes Dev. 13, 1015– 1024 (1999).

Download references

Author information

Correspondence to Joseph R. Arron or Yongwon Choi.

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.