Expression Analysis of WT1 Splicing Variants In CML and AML: −5/+KTS as a Novel Candidate for Early Marker of Relapse

Abstract 2749

Total WT1 expression is currently used for monitoring of AML patients and is of prognostic value also for CML patients according to our experience. Total WT1 level brings together levels of enormous amount of WT1 variants. Four major splicing variants reffered to as −5/−KTS, −5/+KTS, +5/−KTS and +5/+KTS are produced by combining two main independent splicing sites – exon 5 and KTS sequence. Short WT1 (sWT1) variant originates from usage of an alternative first exon E1a (Dalloso et al., 2001). Interestingly, exon 5 splicing was associated with resistance to apoptosis (Pritchard et al., 2006). The sWT1 was reported as overexpressed specifically in leukemias (Hossain et al., 2006). In view of literature data, we have suggested that splicing variant ratio or particular variant levels might have impact to leukemia pathogenesis, influence responsiveness to therapy and thus be a candidate for novel leukemic markers giving additional or more precise information about the disease state besides total WT1 levels. This idea was enhanced by our previous observation of differences of the four major variants levels associated with risk groups and different achieved responses to therapy among diagnostic AML patients samples (ASH 2009). In the current study, we extended our field of interest also on sWT1 vs. full length WT1. We aimed to find out whether there might be any advantage of monitoring of any of the WT1 variant in AML and CML patients undergoing therapy for more precise, sure and optionally earlier relapse detection/prediction.

We have designed a novel method for quantification of sWT1 and full length WT1 separately. Quantification is performed by two separated real-time PCRs based on discriminating forward primers. One of those primers hybridizes onto exon 1 the other onto alternative exon 1a which enables to quantify full length WT1 and sWT1, respectively. For quantification of the four major variants, we used a method that we previously developed and reported (ASH, 2009). By introducing LNA nucleotides into the sequence of the primers with lowest hybridization efficiency we further improved reaction sensitivity, efficiency and robustness. Fourteen CML and ten AML patients undergoing therapy and developing haematological or molecular relapse were retrospectively analysed so far.

We found out that the ratio of the four variants was changed in relapse as compared to diagnosis. The most abundant variant at diagnosis was +5/−KTS. In relapse, it became to be −5/+KTS in both AML and CML. The −5/+KTS was the first one to be increased in most of the cases and usually its expression kinetics was the most expressive of all four variants. As compared to BCR-ABL transcript, both total WT1 and −5/+KTS increases were more expressive or even earlier in some of CML patients and thus clearly predicted upcoming relapse. Short WT1 was detected in all AML samples, in CML we found it only in blast crisis not in the chronic phase samples. In all cases the proportion of sWT1 from total WT1 was very low, the majority of total WT1 was represented by full length WT1.

Taken together, we showed that mainly −5/+KTS WT1 variant is a good marker of CML and AML and well characterizes the course of the diseases. Our preliminary data indicated that both kinetics and actual levels of −5/+KTS might specify information done by total WT1 levels and in case of CML by BCR-ABL making relapse prediction more sure or possibly even earlier.

Supported by MZOUHKT2005.