CD99-positive "Ewing's sarcoma" from mouse-bone marrow-derived mesenchymal progenitor cells?

One of the unsolved enigmas of Ewing’s sarcoma family of tumors (ESFT) is its histogenetic origin. Lately, evidence has been accumulating that boneor bone marrow–derived mesenchymal progenitor cells (MPC) can be transformed by EWS-FLI1 causing small-round cell-like tumors in mice. This observation was extended recently in a report by Riggi et al. (1). Riggi et al. (1) report that EWS-FLI1 is capable of efficient transformation of MPCs with high tumorigenic potential. The tumors were interpreted as ‘‘Ewing’s sarcoma.’’ This conclusion requires a thorough comparison of gene expression in MPC-derived tumors with human ESFT. Riggi et al. (1) restricted their analysis to NSE and CD99, a diagnostic hallmark of ESFT in humans. The authors used O13 antibody (2) from Signet Laboratories (Dedham, MA) on tumor sections and show in Fig. 3C a striking induction of O13 reactivity in MPC. This result is provided without any comment, indicating that the authors overlooked the lack of a confirmed functional CD99 homologue in rodents. Recently, a CD99 orthologous protein has been identified in mice (3) with only 45% identity with human CD99, mainly due to similarities in the transmembrane domain. The immuno reactive extracellular CD99 domain shows significant divergence between mice and men (4). The epitope of O13 antibody has been mapped to a DGEN motif between residues 49 and 64 of human CD99 (5). This motif is completely lacking in mouse CD99. Thus, it is unlikely that the O13 reactivity detected by Riggi et al. (1) is derived from the established murine CD99 orthologue. The possibility remains that the EWS-FLI1-induced O13 reactivity in mouse MPCs derives from other proteins. Therefore, this question arises: What is the identity of the O13 reactive protein in MPC? Human ESFT show a tight association between the presence of a EWS-ETS gene rearrangement and high CD99 but their interrelationship is still unclear. When introduced into human neuroblastoma or rhabdomyosarcoma cells, EWS-FLI1 readily induced CD99 expression. However, when EWS-FLI1 is knocked down in ESFT cells, CD99 expression remains invariably high, favoring the hypothesis of a CD99-positive precursor cell. If a murine CD99 homologue overexpressed in murine ‘‘Ewing’s sarcoma’’ existed, one would expect that its precursor is also positive for this marker. With respect to O13 reactivity, this was not the case for MPCs. Thus, we believe that without further evidence for similarity between human ESFT and MPC-induced tumors in mice, MPCs cannot be considered ESFT progenitors.