Premature Growth Plate Closure Caused by a Hedgehog Cancer Drug Is Preventable by Co-Administration of a Retinoid Antagonist in Mice.

The growth plates are key engines of skeletal development and growth and contain a top reserve zone followed by maturation zones of proliferating, prehypertrophic, and hypertrophic/mineralizing chondrocytes. Trauma or drug treatment of certain disorders can derange the growth plates and cause accelerated maturation and premature closure, one example being anti-hedgehog drugs such as LDE225 (Sonidegib) used against pediatric brain malignancies. Here we tested whether such acceleration and closure in LDE225-treated mice could be prevented by co-administration of a selective retinoid antagonist, based on previous studies showing that retinoid antagonists can slow down chondrocyte maturation rates. Treatment of juvenile mice with an experimental dose of LDE225 for 2 days (100 mg/kg by gavage) initially caused a significant shortening of long bone growth plates, with concomitant decreases in chondrocyte proliferation; expression of Indian hedgehog, Sox9, and other key genes; and surprisingly, the number of reserve progenitors. Growth plate involution followed with time, leading to impaired long bone lengthening. Mechanistically, LDE225 treatment markedly decreased the expression of retinoid catabolic enzyme Cyp26b1 within growth plate, whereas it increased and broadened the expression of retinoid synthesizing enzyme Raldh3, thus subverting normal homeostatic retinoid circuitries and in turn accelerating maturation and closure. All such severe skeletal and molecular changes were prevented when LDE-treated mice were co-administered the selective retinoid antagonist CD2665 (1.5 mg/kg/d), a drug targeting retinoid acid receptor γ, which is most abundantly expressed in growth plate. When given alone, CD2665 elicited the expected maturation delay and growth plate expansion. In vitro data showed that LDE225 acted directly to dampen chondrogenic phenotypic expression, a response fully reversed by CD2665 co-treatment. In sum, our proof-of-principle data indicate that drug-induced premature growth plate closures can be prevented or delayed by targeting a separate phenotypic regulatory mechanism in chondrocytes. The translation applicability of the findings remains to be studied.