Cell reprogramming for skeletal dysplasia drug repositioning

Skeletal dysplasia describes a wide group of disorders that affect skeletal growth. In some, such as achondroplasia (ACH), patients have serious complications, while in others, such as thanatophoric dysplasia (TD), neonatal mortality is more common. The majority of skeletal dysplasias are associated with mutations, for example, ACH and TD are caused by gain-of-function mutations in the gene encoding fibroblast growth factor receptor 3 (FGFR3). FGFR3 is a transmembrane tyrosine receptor and activates STAT and MAPK pathways to suppress the proliferation and differentiation of cartilage cells, chondrocytes. Although it is known that several types of ligands activate FGFR3 and that the resulting signaling cascade both up-regulates and down-regulates genes accordingly to achieve the suppression of chondrocytes, the mechanism for the gain-of-function is poorly understood, which is one reason drug development for ACH and TD has been relatively ineffective.

Because of the difficulty in identifying or designing inhibitors specific to the FGFR3 isoform, molecules that instead act on downstream or upstream targets have been sought. One molecule, C-type natriuretic peptide (CNP), counters the mutation effect in mice by inhibiting the MAPK pathway to correct extracellular matrix synthesis and rescue bone growth.¹ Another molecule, soluble FGFR3, reduces the FGFR3 signal by acting as a decoy receptor to reduce the available number of ligands that bind to membrane-bound FGFR3 and thus prevent activation of the corresponding signaling pathways.² Yet results from mouse models are sometimes difficult to extrapolate to humans, especially for systems that endure physical stresses, because of the significantly different sizes and proportions of mouse and human cartilage. For example, the density of chondrocytes in cartilage is much less in humans than it is in mice, suggesting that these cells handle physical stress differently. Thus, drug discovery using patient cells is preferred, however, chondrocytes are extremely difficult to acquire, especially from child patients, and even then are difficult to expand and maintain in culture.

Like many other diseases, the study of skeletal dysplasia has benefited tremendously from the invention of induced pluripotent stem cells (iPSCs). One crucial advantage of iPSCs is that they can be generated from patient somatic cells. Differentiating these iPSCs into the desired cell type has provided a whole new source of cells for scientific study. Since the first human cells were reprogrammed into iPSCs, a long list of new disease models has emerged.³ Furthermore, because the cells of these models are human based, there is an expectation that they will significantly reduce the cost and time of drug discovery. For this same reason, it is expected that iPSC technology will be instrumental in drug repositioning, as it will provide an abundant source of cells on which existing drugs that have had their safety already measured can be tested. Indeed, in our iPSC-based disease model for skeletal dysplasia, we demonstrated potential drug repositioning of statin, as we found a positive effect on bone growth in diseased cells.⁴ Even though several studies have described auspicious effects by statins on human chondrocytes,^5,6 we would not have considered statin as a candidate for ACH or TD because of the negligible number of patient cells available had it not been for iPSCs. Additionally, because iPSCs can be derived from humans, they are less likely to suffer from false positives or negatives in drug testing, an unfortunate and frustrating outcome too commonly seen when using animal models. Coincidently, one excellent example of a false negative comes from a statin study where cholesterol levels did not improve in rats.⁷

We therefore investigated the effects of several molecules, including statins, on chondrocytes differentiated from iPSCs, which were reprogrammed from ACH and TD patient-fibroblasts in culture.⁴ One concern about iPSCs is whether they adequately recapitulate cellular properties upon differentiation. We found that compared with those from healthy-iPSCs, chondrocytes derived from patient-iPSCs had defective cartilage tissue formation, which is consistent with ACH and TD phenotypes (Fig. 1). Statin enhanced the degradation of FGFR3 in these cells, which diminished FGFR3 signaling and its downstream targets including the MAPK pathway. Importantly, this discovery is a rare instance when drug testing was done at the tissue level. Additionally, statin was found to induce bone growth in model mice that bore the ACH mutation, whereas wild-type mice showed no significant response to treatment. Overall, this report is the first to describe cartilage tissue generated from a completely iPSC-based system and demonstrated that statin could not only retard mutant FGFR3 activity, but also potentially recover cartilage malformation.

Figure 1.

Fibroblasts taken from a healthy subject (top) or patient (bottom) are reprogrammed into iPSCs and then differentiated into chondrocytes. The iPSC-derived chondrocytes from the healthy subject go on to form healthy cartilage. However, iPSC-derived chondrocytes from the patient show diseased chondrocytes that form defective cartilage (purple). Introducing statin during the differentiation stage of patient-iPSCs recovers the cartilage. Illustration by Aya Motomura.

Despite these encouraging findings, statin may not be an ideal candidate for treating skeletal dysplasia, especially in children, because of its effects on cholesterol, which is essential for development. It would be interesting, therefore, to determine if the mechanism of the FGFR3 action is independent of that on cholesterol. Nevertheless, statin could be a promising paradigm for drug compounds that could be further tested using iPSC-based models and eventually reach the clinic.