Abstract 2592

Poster Board II-568

Chronic Myelogenous Leukemia (CML) represents one of the best characterized models for studying cancer stem cell properties and resistance to tyrosine kinase inhibitors (TKIs). A substantial proportion of CML patients develop primary or secondary resistance to such agents, and in approximately ∼30% of the cases the resistance mechanism(s) remain(s) obscure. It is known that CML leukemic stem cells could remain viable and quiescent despite the presence of growth factors that seem to protect them from apoptosis, or drugs. By performing genome-wide microarray analysis of CML sensitive and resistant cell lines, a novel oncogene Twist-1 has been shown to be deregulated in CML cells innately resistant to imatinib through, an unknown mechanism (Tipping AJ et al. Exp Hematol 2003; 31(11): 1073–1080). In this study, we have observed similar results in the same BCR-ABL+ KCL22 cell line resistant and sensitive to imatinib by quantitative RT-PCR (qRT-PCR) using specific primers and at the protein level. Twist-1 is a key player during embryogenesis that represses mesoderm differentiation and retains undifferentiated cells into a proliferative state. In adult tissues, its expression has been reported in intestinal crypts where adult gut stem cells reside. In cancer, Twist-1 overexpression has been reported in a variety of solid tumors and hematopietic malignacies (Sezary syndrome, B-Cell leukemias) and its deregulation is always associated with poor prognosis and resistance to reagents targeting microtubules (Puisieux A et al. Br J Cancer 2006; 94 (1):13–17). In addition, functional studies show that Twist-1 plays a major role in tumor promotion and progression, by inhibiting differentiation, and interfering with the p53 tumor suppressor pathway, thereby favoring cell survival (Ansieau S et al. Cancer Cell 2008; 14(1):79–89). In this context, we have analyzed Twist-1 gene expression by qRT-PCR in both CD34+ (immature cells) and CD34 (mature compartment) blood or marrow cells from CML patients at diagnosis, at different disease phases, and once TKIs resistance has occur. Results are expressed as ratio to normal CD34 cells values. We observed a 74- and 39-fold increase in CD34+ and CD34 CML cells (n= 30 patients), respectively. When Twist-1 expression could be monitored during CML progression in a same patient, a regular expression increase in the CD34+ compartment was observed [8.5 fold (n=17), 18 fold (n=4) and ≥105 fold (n=8) in chronic, accelerated and blastic phases respectively]. Twist-1 was also overexpressed in the CD34 compartment along with disease progression [between chronic and later stages (11 fold), but with no significant differences between accelerated (88 fold) and blastic (73 fold) phases]. Our results then clearly identify Twist-1 gene expression as a unique predictive molecular marker of resistance, especially in patients with unidentified mechanism of resistance as we were able at the time of diagnosis to point out 2 patients/13 that present a higher Twist-1 level relatively to the mean value and which have develop a cytogenetic resistance within 6 months (ELN criteria) without any mutation, BCR-ABL overexpression or clonal evolution. The overexpression of Twist-1 is predominant in the immature compartment during CML progression and resistance, supporting the hypothesis that resistant cells that overexpress Twist-1 might emerge from CML stem cells. In samples from patients who became resistant to imatinib within the course of the disease, Twist-1 expression follow-up at the time of cytogenetic remission, and later resistance, indicate that Twist-1 expression was transiently down regulated by TKIs. Finally, using Twist-1 specific siRNA we confirmed in CML cell lines the direct involvement of this protein in CML resistance to imatinib. Altogether, our results suggest for the first time that Twist-1 may represent a useful molecular marker to identify at diagnosis patients prone to become resistant to imatinib. The exact interactions between Twist-1 and BCR-ABL proteins pathways remain to be determined yet.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution