LX-2931

Osteoporosis is a worldwide epidemic but pharmacological agents to stimulate new bone formation are scarce. We have shown that increasing tissue levels of sphingosine-1-phosphate (S1P) by blocking its degradation by the S1P lyase has pronounced osteoanabolic effect in mouse osteoporosis models by stimulating osteoblast differentiation through the S1P receptor 2 (S1P2). However, S1P lyase inhibitors have side effects complicating potential clinical use. Here, we tested whether direct S1P2 engagement by the S1P2 agonist CYM5520 exerted osteoanabolic potential in estrogen deficiency-induced osteopenia in mice. We compared its efficacy to LX2931, a novel S1P lyase inhibitor currently tested in rheumatoid arthritis.

CYM5520, LX2931 or vehicle were administered to ovariectomized mice for 6 weeks beginning 5 weeks after ovariectomy, Bone mass, cellular composition and mechanical strength were assessed by microCT, histomorphometry and three point bending tests. Plasma markers of bone metabolism were analyzed by ELISA.

Therapeutic treatment with CYM5520 and LX2931 clearly increased long bone and vertebral bone mass to impressive 3-5 fold over vehicle in osteopenic ovariectomized mice. As expected, lymphopenia was a side effect of LX2931, whereas none occurred with CYM5520. Consistent with an osteoanabolic effect, CYM5520 increased osteoblast number, osteoid surface and alkaline phosphatase area 2-3 fold over vehicle. Plasma concentrations of the osteoanabolic marker procollagen I C-terminal propeptide were also elevated by CYM5520 and LX2931. LX2931 but not yet CYM5520 increased cortical thickness and mechanical strength without affecting mineral density.

Treatment with a pharmacological S1P2 agonist corrected ovariectomy-induced osteopenia in mice by inducing new bone formation thus constituting a novel osteoanabolic approach to osteoporosis.

Progressive lung disease with early onset is the main cause of mortality and morbidity in cystic fibrosis patients. Here we report a reduction of sphingosine-1-phosphate (S1P) in the lung of unchallenged Cftr tm1EUR F508del CFTR mutant mice. This correlates with enhanced infiltration by inducible nitric oxide synthase (iNOS)-expressing granulocytes, B cells, and T cells. Furthermore, the ratio of macrophage-derived dendritic cells (MoDC) to conventional dendritic cells (cDC) is higher in mutant mouse lung, consistent with unprovoked inflammation. Oral application of a S1P lyase inhibitor (LX2931) increases S1P levels in mutant mouse tissues. This normalizes the lung MoDC/cDC ratio and reduces B and T cell counts. Lung granulocytes are enhanced, but iNOS expression is reduced in this population. Although lung LyC6+ monocytes are enhanced by LX2931, they apparently do not differentiate to MoDC and macrophages. After challenge with bacterial toxins (LPS-fMLP) we observe enhanced levels of proinflammatory cytokines TNF-α, KC, IFNγ, and IL-12 and the inducible mucin MUC5AC in mutant mouse lung, evidence of deficient resolution of inflammation. LX2931 does not prevent transient inflammation or goblet cell hyperplasia after challenge, but it reduces MUC5AC and proinflammatory cytokine levels toward normal values. We conclude that lung pathology in homozygous mice expressing murine F508del CFTR, which represents the most frequent mutation in CF patients, is characterized by abnormal behavior of infiltrating myeloid cells and delayed resolution of induced inflammation. This phenotype can be partially corrected by a S1P lyase inhibitor, providing a rationale for therapeutic targeting of the S1P signaling pathway in CF patients.