Costello syndrome (CS) is a developmental disorder caused by germline mutations in HRAS gene, which largely affect the oncogenic codon 12, thus leading to a constitutive activation of HRAS. Patients show some prominent phenotypical characteristics of a premature-like aging and cases of osteoporosis/osteopenia have been reported frequently in these patients. To date, the cellular bases and molecular mechanisms underlying bone loss in CS are unclear. My study aimed at exploring whether CS HRAS mutations trigger bone loss in the HRAS G12V CS mouse model as described in patients and identify which bone cells are affected by these mutations.

CS HRAS G12V mice life and health span was previously assessed in a long-term study in our group and we observed that these mice have a reduced lifespan in a gene dosagedependent manner, they have reduced body wight throughout their lifespan. I monitored bone mass via micro-computed tomography in various age groups and I identified that mice indeed CS mice develop osteoporosis in long bones (femur and tibia), while L5 vertebrae were not affected. This bone loss was generally present in homozygous mutant mice (HRAS G12V ki/ki) and from all age cohorts studied mostly affected 12-month-old males and 15-month-old females. Surprisingly, CS mice that survived long beyond median life span had no differences when compared to wild type controls, which may be an indication that a better overall health span in CS mice reflects itself in the bone homeostasis. Ex vivo bone histomorphometry revealed that an increased osteoclastogenesis causes a bone loss especially in ki/ki CS mice. Furthermore, in vitro differentiation of osteoclast from bone marrow cells showed that in contrast to ex vivo histomorphometry, where only ki/ki mice showed increased osteoclast number, both +/ki and ki/ki mutant genotypes showed a cell autonomous increased osteoclastogenesis. Surprisingly, despite their increased number +/ki osteoclasts resorbing activities were not increased when compared to wild type controls, indicating that they may be defective and in part explain why +/ki CS mice generally do not undergo bone loss. Furthermore, enhanced osteoclastogenesis in mutant genotypes was independent of age of bone marrow donor mice, including from bone marrow collected from mice with normal bone mass. The latter further indicates that despite of an intrinsic high osteoclastic potential of CS progenitor cells, osteoclast-mediated bone loss needs an extrinsic crucial influence triggered by a yet elusive factor or mechanism (e.g., inflammation, hormonal changes, osteoclast/osteocytes secreted factors). The increased osteoclastogenesis in CS cells was rescued in vitro by MEK inhibition thus suggesting that HRAS-MAPK pathway is a major osteoclastogenesis controller and can be targeted for subsequent preclinical studies. Last but not least, preliminary bioenergetic studies revealed that CS ki/ki osteoclasts may have an increased glycolysis. Corroborating this data with previous studies indicating that resorbing osteoclasts are relying on glycolysis, I consider that ki/ki osteoclast enhanced activity may also contribute to bone loss in CS mouse model and possibly in patients too.

A surprising aspect was that generally osteoblast numbers were not affected in CS mice ex vivo and in in vitro differentiation of bone marrow stromal cells, indicating that HRAS mutations do not affect osteoblast differentiation. However, in in vitro differentiation of MC-3T3 that stably overexpress HRAS mutants in osteoblasts was severely suppressed. These observations indicate that a moderate expression and activation of HRAS G12V in mouse primary cells is not potent enough to block osteoblast differentiation.

Overall, I showed in my thesis for the first time that the bone loss in CS mice harboring HRAS G12V mutations is predominantly an osteoclast-driven pathophenotype and that the bone loss is influenced by the overall mouse health. Future work will focus on identifying molecular targets and pathways that are dysregulated in CS osteoclast differentiation and the identification of CS “systemic factors” that may exacerbate osteoclast activation and bone loss, followed by preclinical intervention and bone phenotype rescue.