An anti-ACVR1 antibody exacerbates heterotopic ossification by fibro-adipogenic progenitors in fibrodysplasia ossificans progressiva mice

Abstract

Fibrodysplasia ossificans progressiva (FOP) is a rare genetic disease characterized by progressive and catastrophic heterotopic ossification (HO) of skeletal muscle and associated soft tissues. FOP is caused by dominantly acting mutations in the gene encoding the bone morphogenetic protein (BMP) type I receptor, ACVR1 (ALK2), the most prevalent of which results in an arginine to histidine substitution at position 206 (ACVR1[R206H]). The fundamental pathological consequence of FOP-causing ACVR1 receptor mutations is to enable activin A to initiate canonical BMP signaling in fibro-adipogenic progenitors (FAPs), which drives HO. We developed a monoclonal blocking antibody (JAB0505) against the extracellular domain of ACVR1 and tested its effect on HO in 2 independent FOP mouse models. Although JAB0505 inhibited BMP-dependent gene expression in wild-type and ACVR1(R206H)-overexpressing cell lines, JAB0505 treatment profoundly exacerbated injury-induced HO. JAB0505-treated mice exhibited multiple, distinct foci of heterotopic lesions, suggesting an atypically broad anatomical domain of FAP recruitment to endochondral ossification. This was accompanied by dysregulated FAP population growth and an abnormally sustained immunological reaction following muscle injury. JAB0505 drove injury-induced HO in the absence of activin A, indicating that JAB0505 has receptor agonist activity. These data raise serious safety and efficacy concerns for the use of bivalent anti-ACVR1 antibodies to treat patients with FOP.

Anti-ACVR1 antibodies exacerbate heterotopic ossification in fibrodysplasia ossificans progressiva (FOP) by activating FOP-mutant ACVR1

“…Moreover, other means of dimerization of ACVR1 have the same effect as the anti-ACVR1 antibodies, suggesting that FOP-mutant ACVR1 is activated by simple dimerization (whereas WT ACVR1 is not). Similar results have been concurrently reported, thereby corroborating our observations ( 14 ). Activation of FOP-mutant ACVR1 in response to antibody-mediated dimerization mimics the response to activin A.…”

“…Irrespective of the reasons for the discrepancy observed in signaling outcomes between our initial bioassay and our in vivo data, it is clear that in physiological settings ACVR1[R206H] is activated by anti-ACVR1 mAbs. This observation has been concurrently and independently corroborated using a different anti-ACVR1 antibody and a different mouse model of FOP ( 14 ). Furthermore, at least in vitro, it also holds for an additional FOP-causing variant of ACVR1, 258G; however, we have yet to test any other ACVR1 variants documented to cause FOP ( 4 ).…”

Twists in the fibrodysplasia ossificans progressiva story challenge and expand our understanding of BMP biology

“…With WT ACVR1, the anti-ACVR1/type 2 BMP receptor complexes did not signal, effectively mirroring the nonsignaling complex. However, as independently demonstrated by Aykul, Huang, et al and Lees-Shepard, Stoessel, et al, the FOP-mutant ACVR1/type 2 BMP heterotetramers resulted in complexes that induced phosphorylation of Smad1/5/8 and drove HO in FOP, much like activin A ( 10 , 11 ). However, antibody binding alone was insufficient to drive HO in FOP.…”

“…An important finding of the Aykul, Huang, et al ( 10 ) and Lees-Shepard, Stoessel, et al ( 11 ) papers is the implication of an additional factor causing FOP. While three conditions consisting of the R206H ACVR1 variant, activin A or anti-ACVR1 antibody binding, and R206H ACVR1 homodimerization are necessary for HO development in FOP, they are not sufficient.…”

“…The fascinating FOP story takes another twist in this issue of the JCI . Given the clear evidence that activin A binding to FOP-variant ACVR1 drives HO in FOP, both Regeneron and David Goldhamer’s laboratory developed anti-ACVR1 antibodies as potential therapeutics for FOP ( 10 , 11 ). In both cases, in vitro testing in cell lines overexpressing either WT or FOP-mutant ACVR1 confirmed that such antibodies blocked both WT and mutant ACVR1 signaling.…”