Do neural signals remodel bone?

The hormone leptin is best known for its influence on body weight. But it also controls bone mass, and recent work in mice is beginning to uncover the neuroendocrine systems involved.

The hypothalamus is the part of the brain that maintains homeostasis, which includes regulating the autonomic nervous system — the unconscious system for monitoring and controlling the state of the body's internal organs. A key input to the hypothalamus is leptin, a hormone that is produced in fat cells to signal energy sufficiency and which affects food intake, body weight and the neuroendocrine systems1,2,3,4. Previous work from Gerard Karsenty and colleagues showed that leptin also affects bone mass5,6: leptin-deficient ob/ob mice have high bone mass, and when leptin is replaced, their bone mass goes down. On page 514 of this issue (Elefteriou et al.7), the team reports that leptin directly regulates bone physiology through control of the sympathetic nervous system, the arm of the autonomic nervous system best known for the fight-or-flight response.

The work is provocative for a number of reasons. First, the results establish that the sympathetic nervous system is required for the effects of leptin on bone mass. The authors examined mice lacking the β2-adrenergic receptor (β2-AR) — one of several adrenergic receptors that respond to noradrenaline, a neurotransmitter released by sympathetic neurons. The mutant mice are resistant to the bone-reducing effects that leptin exerts via actions on the central nervous system (CNS). Second, the mice have increased bone mass at baseline, compared with normal mice, indicating that the sympathetic nervous system helps to set the level of bone mass, even under normal conditions.

Finally, and perhaps most strikingly, the mice are resistant to bone loss induced by the removal of their ovaries, and the consequent decrease of circulating oestrogen. Oestrogen deficiency in postmenopausal women causes rapid bone loss and predisposes them to osteoporotic fractures. Does menopausal bone loss require sympathetic nervous activity? Interestingly, ovariectomy in mice causes a dramatic loss of the nerves that usually penetrate bone8. These observations suggest that oestrogen may be pivotal to the ability of the sympathetic nervous system to regulate bone formation. Moreover, if the sympathetic nervous system sets bone mass through β-adrenergic receptors, treatment with β-blockers (which are used to treat hypertension, for example) should improve bone mass. Whether this intriguing prediction is correct has not been settled by retrospective studies9,10,11 and will require carefully designed clinical trials.

Bone continuously renews itself by a process called remodelling: old bone is eaten away (resorbed) by cells called osteoclasts, and new bone is laid down by cells called osteoblasts. It has been known for some time that the hypothalamus integrates inputs from the body and transmits signals that alter bone remodelling. This was previously thought to be mediated by altering the secretion of hormones, especially gonadal steroids. Elefteriou et al.7 report that the sympathetic nervous system is also involved, and they provide a molecular mechanism linking leptin, the sympathetic nervous system and decreased bone mass.

Hormones that control bone resorption often act indirectly, by stimulating osteoblasts to promote the formation of osteoclasts, rather than directly on osteoclasts themselves. For example, Elefteriou et al. report that adrenergic signals cause osteoblasts to secrete RANK ligand, the principal physiological inducer of osteoclast formation (Fig. 1). This effect requires protein kinase A, an enzyme that activates its target proteins by adding a phosphate group to them. In this case, the target is ATF4, a factor that is essential for osteoblast development and function. Once activated, ATF4 stimulates the production of RANK ligand.

Figure 1: Brain and bone in sympathy.

Elefteriou et al.7 find that leptin exerts its effects on bone through the sympathetic nervous system, which releases the neurotransmitter noradrenaline. Activation of one noradrenaline receptor, the β2-adrenergic receptor (β2-AR), causes osteoblasts to produce RANK ligand, through activation of protein kinase A (PKA) and phosphorylation (P) of the gene regulator ATF4. RANK ligand stimulates the formation of osteoclasts, which eat away the bone, so that overall bone mass diminishes.

Collectively, the results suggest that leptin induces bone loss via actions on the CNS and by regulating input from the sympathetic nervous system to bone. This is confirmed by the absence of leptin effects in mice lacking β2-AR. However, leptin-deficient ob/ob mice have high rates of bone resorption (previously attributed to oestrogen deficiency), despite a reduction in β-adrenergic signalling that should lead to decreased resorption.

A final interesting component of the work by Elefteriou and colleagues may help to explain this apparent paradox. They report that mice lacking the leptin-regulated neuropeptide CART (for ‘cocaine- and amphetamine-regulated transcript’), have a low bone mass at baseline. Because levels of the RNA encoding CART are very low in leptin-deficient mice, one might predict that CART mediates the effects of leptin. By contrast, the authors find that mice without CART have increased bone resorption in response to leptin. These observations suggest that neurons expressing CART act to inhibit bone loss, including that mediated by leptin.

The CART neurons involved in bone loss have yet to be identified. One obvious target is the CART neurons in a region of the hypothalamus called the arcuate nucleus12. These neurons are stimulated by leptin and also express POMC, the prototypic CNS target of leptin. Interestingly, these neurons directly target sympathetic preganglionic neurons (the first neurons in the autonomic neuron chain)13. However, it is unclear whether they mediate leptin's effects on bone, as mice that lack leptin receptors only in POMC/CART neurons do not seem to have bone abnormalities14.

CART expression is widespread in the brain and spinal cord, including sympathetic preganglionic neurons themselves, and in several other regions of the hypothalamus including the ‘ventral premammilary nucleus’, which has neurons that express high levels of receptors for sex steroids15. Given the proposed interaction of oestrogen and the autonomic nervous system, it is intriguing to speculate that hypothalamic cell groups (including CART neurons) might integrate a number of signals, including leptin and oestrogen, to coordinate the CNS control of bone mass through its output to sympathetic preganglionic neurons.

Several points remain to be reconciled with this model of the regulation of bone mass. For example, obese women have high leptin levels and high bone mass; lean women with very low leptin levels (such as marathon runners) have low bone mass. In addition, whether leptin has direct effects on bone cells remains to be settled5,16. Nonetheless, Elefteriou and colleagues' findings provide support for the notion that the hypothalamus and its control of the autonomic nervous system are key to the regulation of bone homeostasis.


  1. 1

    Ahima, R. S., Saper, C. B., Flier, J. S. & Elmquist, J. K. Front. Neuroendocrinol. 21, 263–307 (2000).

    CAS  Article  Google Scholar 

  2. 2

    Friedman, J. M. Nutr. Rev. 60, S1–S14 (2002).

    Article  Google Scholar 

  3. 3

    Flier, J. S. Cell 116, 337–350 (2004).

    CAS  Article  Google Scholar 

  4. 4

    Zigman, J. M. & Elmquist, J. K. Endocrinology 144, 3749–3756 (2003).

    CAS  Article  Google Scholar 

  5. 5

    Takeda, S. et al. Cell 111, 305–317 (2002).

    CAS  Article  Google Scholar 

  6. 6

    Ducy, P. et al. Cell 100, 197–207 (2000).

    CAS  Article  Google Scholar 

  7. 7

    Elefteriou, F. et al. Nature 434, 514–520 (2005).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Burt-Pichat, B. et al. Endocrinology 146, 503–510 (2005).

    CAS  Article  Google Scholar 

  9. 9

    Pasco, J. A. et al. J. Bone Miner. Res. 19, 19–24 (2004).

    CAS  Article  Google Scholar 

  10. 10

    Rejnmark, L. et al. Calcif. Tissue Int. 75, 365–372 (2004).

    CAS  Article  Google Scholar 

  11. 11

    Schlienger, R. G., Kraenzlin, M. E., Jick, S. S. & Meier, C. R. J. Am. Med. Assoc. 292, 1326–1332 (2004).

    CAS  Article  Google Scholar 

  12. 12

    Kristensen, P. et al. Nature 393, 72–76 (1998).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Elias, C. F. et al. Neuron 21, 1375–1385 (1998).

    CAS  Article  Google Scholar 

  14. 14

    Balthasar, N. et al. Neuron 42, 983–991 (2004).

    CAS  Article  Google Scholar 

  15. 15

    Elias, C. F. et al. J. Comp. Neurol. 432, 1–19 (2001).

    CAS  Article  Google Scholar 

  16. 16

    Reseland, J. E. & Gordeladze, J. O. FEBS Lett. 528, 40–42 (2002).

    CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Elmquist, J., Strewler, G. Do neural signals remodel bone?. Nature 434, 447–448 (2005).

Download citation

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.