The development of stem-cell-based models of two diseases that cause dwarfism reveals that statins — drugs that are used to treat high levels of blood cholesterol — may also promote cartilage formation and bone growth. See Article p.507
Many medical conditions can cause short stature, but a faulty gene encoding the protein FGFR3 is responsible for two-thirds of all forms of dwarfism in humans. FGFR3 normally controls a brake signal in the molecular machinery that regulates the growth of limb bones during childhood and adolescence. In 1 in every 10,000–30,000 births, genetic mutations cause FGFR3 to become overactive and so brake too hard. Although our understanding of the cellular processes that go awry in dwarfism is good, development of treatments has been hampered by a lack of efficient methods for screening and testing potential drugs. In this issue, Yamashita et al1 page 507. report a major step forward in solving this problem, establishing a human-disease-based system for screening potential drugs to treat skeletal-growth defects.
In humans, the most common FGFR3 mutation results in achondroplasia, a disorder that causes short extremities, increased curvature of the spine and distortion of skull growth, resulting in substantial health problems2. More-severe mutations in FGFR3 can cause thanatophoric dysplasia, in which a small chest, and respiratory problems, may cause death either at or shortly after birth3. In both dysplasias, skeletal defects are caused by decreased proliferation and impaired maturation of cartilage-forming cells called chondrocytes within growing regions of bone4. It has not previously been possible to obtain chondrocytes from patients, but Yamashita and colleagues took advantage of improved cell-reprogramming techniques5 to do just that.
The authors isolated skin cells from three individuals with thanatophoric dysplasia and converted the cells to induced pluripotent stem cells, which can give rise to any cell type in the body. Next, Yamashita and co-workers stimulated the stem cells to become chondrocytes, which had the same genetic make-up as the original patients. They then took advantage of the chondrocytes' ability to aggregate into cartilage-forming particles6 to generate a system for analysing particles formed by thanatophoric dysplasia chondrocytes and by controls without the FGFR3 mutation. The authors compared the particles' similarities and differences as the different cells grew and matured over several weeks in culture (Fig. 1).
A major difference was that, compared to controls, thanatophoric dysplasia particles exhibited impaired maturation associated with degradation of cartilage. Remarkably, reducing FGFR3 levels or adding antibodies to block FGFR3 activity in dysplasia cultures restored growth and maturation of the cartilage-forming particles to normal levels. Yamashita and co-workers used their culture system to assay several molecules that affect either the response of cells to FGFR3 signals or the formation of chondrocytes from stem cells, to determine which could promote cartilage development in dysplastic cells. Molecules that had positive effects included C-type natriuretic peptide (CNP) and several statins, including lovastatin and rosuvastatin.
CNP has a positive effect on bone formation and growth4, and its overexpression in chondrocytes counteracts dwarfism in a mouse model of achondroplasia7. As a result, CNP has been pursued as a potential achondroplasia treatment, although it is not an ideal candidate. A major obstacle is that the peptide, which must be injected, is degraded within minutes of being administered. A more stable version is effective in mouse models of achondroplasia and is currently in clinical trials, but still requires daily injections8. In addition, the effects of CNP on the cardiovascular system and the central nervous system raise the possibility of undesirable side effects if the drug is used long-term in children.
Statins — cholesterol-lowering drugs that are available in tablet form — provide an interesting alternative. Their safety has been evaluated in children with inherited high cholesterol9, and evidence10 suggests that early statin treatment improves the chances of children with this condition reaching the age of 30 without having a heart attack. In addition to their cholesterol-related properties, the drugs stimulate production of chondrocyte molecules that make up the structure of cartilage11, and repress production of cartilage-degrading enzymes12. In a series of compelling experiments, Yamashita et al. demonstrated that lovastatin stimulates production of cartilage components in thanatophoric dysplasia chondrocytes, and promotes the formation of chondrocytes from stem cells. It also restores cartilage formation by chondrocytes derived from patients with achondroplasia. Finally, the authors showed that injecting rosuvastatin into mice with an achondroplasia-causing defect in FGFR3 partially restored bone growth in the limbs and head.
What are the mechanisms underlying these striking effects? Yamashita and colleagues' study does not provide the full answer. However, the authors do find that high levels of FGFR3 protein, but not messenger RNA, are reduced to normal levels when lovastatin is added to cultures of particles derived from people with either form of dysplasia. This suggests that statins stimulate degradation of FGFR3. Cellular protein-degradation machines called proteasomes might be involved, because adding a proteasome inhibitor to lovastatin-containing cultures increased levels of FGFR3. The researchers speculate that this is related to the ability of statins to lower cholesterol in cells, and to destabilize cell membranes so that FGFR3 (which spans the membrane) is more easily internalized and degraded, but this remains to be determined.
If the ability of statins to restore cartilage-particle growth is found to be independent of their cholesterol-lowering properties, it may be possible to modify the drugs such that these two effects are separated. However, if the cartilage-promoting effect of statins is a direct consequence of a decrease in cholesterol, extreme care is needed before using the drugs to treat children with achondroplasia. It will be crucial to ensure that cholesterol levels in these children are maintained at reasonable levels.
Between the ages of 25 and 35, mortality related to heart disease is more than 10 times higher in people with achondroplasia than in the general population13,14. The reasons for this are not understood, and limited data suggest that serum cholesterol levels in children with achondroplasia are in the high normal range15. Whether statin treatment would help to reduce this mortality is therefore unclear.
In summary, Yamashita et al. have established a disease model of achondroplasia and related dysplasias based on pluripotent stem cells. The results of the study raise the possibility that statins might be effective in treating children with these disorders. Furthermore, the authors' system allows screening of additional compounds in the search for even safer drugs.
Yamashita, A. et al. Nature 513 507–511 (2014).
Shiang, R. et al. Cell 78, 335–342 (1994).
Tavormina, P. L. et al. Am. J. Hum. Genet. 64, 722–731 (1999).
Laederich, M. B. & Horton, W. A. Curr. Opin. Pediatr. 22, 516–523 (2010).
Okita, K. et al. Nature Methods 8, 409–412 (2011).
Koyama, N. et al. Stem Cells Dev. 22, 102–113 (2013).
Yasoda, A. et al. Nature Med. 10, 80–86 (2004).
Lorget, F. et al. Am. J. Hum. Genet. 91, 1108–1114 (2012).
Eiland, L. S. & Luttrell, P. K. J. Pediatr. Pharmacol. Ther. 15, 160–172 (2010).
Braamskamp, M. J. et al. Circulation 128, A17837 (2013).
Hatano, H., Maruo, A., Bolander, M. E. & Sarkar, G. J. Orthop. Sci. 8, 842–848 (2003).
Simopoulou, T., Malizos, K. N., Poultsides, L. & Tsezou, A. J. Orthop. Res. 28, 110–115 (2010).
Hunter, A. G., Hecht, J. T. & Scott, C. I. Jr Am. J. Med. Genet. 62, 255–261 (1996).
Wynn, J., King, T. M., Gambello, M. J., Waller, D. K. & Hecht, J. T. Am. J. Med. Genet. A 143A, 2502–2511 (2007).
Collipp, P. J., Sharma, R. K., Thomas, J., Maddaiah, V. T. & Chen, S. Y. Am. J. Dis. Child. 124, 682–685 (1972).