Figure-S5 The associations of AMACR expression with cyclin D1, cyclin T2, and Ki-67 levels in myxofibrosarcoma tissues. In contrast to the primary myxofibrosarcoma specimens with a lower level of AMACR immunoexpression (left column), those with AMACR overexpression (right column) significantly increases expression of cyclin D1, cyclin T2, and Ki-67.
ARTICLE ABSTRACT
Purpose: Myxofibrosarcomas frequently display arm-level gains on 5p. We characterized the pathogenetic and therapeutic relevance of the α-methylacyl coenzyme A racemase (AMACR) at 5p13.3.Experimental Design:AMACR mRNA expression in myxofibrosarcomas was analyzed using the public transcriptome and laser-microdissected sarcoma cells. We performed florescence in situ hybridization (FISH) and immunohistochemistry in independent samples for clinical correlates. In AMACR-overexpressing myxofibrosarcoma cells and xenografts, we elucidated the biologic function of AMACR using RNA interference and explored the therapeutic effect and mechanism of an AMACR inhibitor, ebselen oxide.Results: AMACR protein overexpression and gene amplification were significantly associated with each other (P < 0.001), with higher tumor grades (both P ≤ 0.002), and univariately with worse metastasis-free survival (MFS; both P < 0.0001) and disease-specific survival (DSS; P = 0.0002 for overexpression; P = 0.0062 for amplification). AMACR protein overexpression also independently portended adverse outcome (DSS, P = 0.007; MFS, P = 0.001). However, 39% of AMACR-overexpression cases did not show gene amplification, implying alternative regulatory mechanisms. In myxofibrosarcoma cell lines, stable AMACR knockdown suppressed cell proliferation, anchorage-independent growth, and expression of cyclin D1 and cyclin T2. These growth-promoting attributes of AMACR were corroborated in the AMACR-silenced xenograft model and AMACR-underexpressed myxofibrosarcomas, showing decreased labeling for cyclin D1, cyclin T2, and Ki-67. Compared with fibroblasts, AMACR-expressing myxofibrosarcoma cells were more susceptible to ebselen oxide, which not only decreased viable cells, promoted proteasome-mediated degradation of AMACR protein, and induced cellular apoptosis in vitro, but also dose-dependently suppressed xenografted tumor growth in vivo.Conclusions: Overexpressed AMACR in myxofibrosarcomas can be amplification-driven, associated with tumor aggressiveness, and may be relevant as a druggable target. Clin Cancer Res; 20(23); 6141–52. ©2014 AACR.
