Abstract

Abstract We have previously shown that long term progression of human colorectal cancer in serial xenotransplantation is driven by a small sub-fraction of all tumor initiating cells (TIC). These long-term TIC (LT-TIC) self-renew extensively in vivo and are exclusively able to seed metastases in distant organs. Defined phenotypic surface markers of LT-TIC would be crucial to develop effective antibody-based targeting strategies. However, it remains elusive whether a fixed phenotype of long-term TIC is associated with their tumorigenicity thereby allowing prospective isolation of this most relevant cancer cell fraction. To address this question, three dimensional spheroid cultures in serum free medium supplemented with FGF and EGF were generated from primary human colon cancer specimens to enrich for TIC. To calculate the frequency of TIC in spheroid cultures, cells were transplanted in limiting dilution into cohorts of Nod/SCID-IL2RG-/- (NSG) mice. TIC frequency varied from 1 in 22 to 1 in 2x104 spheroid cells, depending on the respective patient sample. When the spheroid cells were transferred to culture conditions that favor their differentiation (gelatin coated plates in the presence of serum and withdrawal of cytokines) the phenotype changed dramatically. The primary colon cancer cells did no longer grow as spheroids but formed an adherent cell layer and up-regulated the colon epithelial differentiation markers CDX2, DEFA5, KRT80, Muc20 and TFF2. In addition, CD133, a widely used marker to enrich for TICs in various solid cancers, was strongly down-regulated upon serum treatment. Strikingly, when xenotransplantated into NSG mice, cells cultured under spheroid and differentiation conditions equally formed serially transplantable tumors, demonstrating that tumor initiating and self-renewal capacity of TIC was not restricted to phenotypically immature spheroid cells. Moreover, CD133 expression did not predict successful tumor formation in vivo. Sorted CD133+ and CD133- cells from 3 individual patients formed tumors with equal efficiency and regenerated both, CD133+ and CD133- cells in vivo. Importantly, clonal analyses of individual lentivirally marked TIC clones cultured under either culture condition revealed no systematic differences in TIC frequency, demonstrating that phenotypic differentiation does not lead to quantitative elimination of TIC clones. Our results demonstrate that phenotypic differentiation does not eliminate the tumor-initiating potential of human colorectal TIC. Moreover, expression of CD133 does not predict their tumor forming and self-renewal potential. This pronounced phenotypic plasticity of human colon cancer TIC poses a grave challenge for surface-targeted elimination of TIC in colorectal cancer. Citation Format: Taronish D. Dubash, Christopher M. Hoffmann, Felix Oppel, Klara Giessler, Sarah Bergmann, Sebastian M. Dieter, Wilko Weichert, Martin Schneider, Manfred Schmidt, Christof von Kalle, Hanno Glimm, Claudia R. Ball. Unstable phenotype of human colon cancer tumor initiating cells. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3057. doi:10.1158/1538-7445.AM2014-3057

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call