Quantitative Expression Analysis of WT1 Main Isoforms in AML

Abstract 1469

The Wilms Tumor 1 (WT1) gene was first described as a tumour suppressor gene, but its accurate role in leukemia development has not been completely elucidated. Some authors support the role of WT1 as a prognostic marker in acute myeloid leukemia (AML) based on the assessment of its expression at the mRNA level. However, the prognostic value of the main isoforms of WT1 has been less studied.

The aim of this study was to develop a specific quantitative assay to estimate the ratio of expression of the four major WT1 isoforms (A, 5-/KTS-; B, 5+/KTS-; C, 5-/KTS+; D, 5+/KTS+) and to evaluate their prognostic impact.

WT1 expression was analyzed in bone marrow samples from 108 patients with AML at diagnosis (65 male/46 female, median age: 61 yr, range: 17 – 91). Likewise, peripheral blood samples of 20 healthy donors and 6 samples of cord blood CD34+ cell selection were analyzed as normal controls. We performed a new method to quantify the ratios of the four major isoforms of WT1. Briefly, to amplify all isoforms within a PCR reaction, specific WT1 primers flanking exon 4 to exon 10 were used in cDNA samples, followed by capillary electrophoresis with laser-induced fluorescence analysis on an ABIPRISM 310 DNA Analyzer (Applied Biosystems, Foster City, CA) and lastly analyzed with the Gene Mapper 4.2 software (Applied Biosystems). The amount of each isoform was calculated by the area under the curve. Subsequent comparisons of isoform ratios were made by standardized calculation of percentage. All values are given as the mean of duplicate PCRs. In parallel, RQ-PCR for total WT1 detection was performed as previously described by Barragan et al. (Haematologica 2004; 89: 926–933). GUS gene was used as housekeeping gene.

Eighteen patients (17%) did not express WT1, while 90 patients (83%) overexpressed WT1 above background levels. The median value of each WT1 isoform was: 18% (range: 2 – 73) for A isoform; 16% (range: 7 – 63) for B isoform; 24% (range: 2 – 52) for C isoform; and 33% (range: 3 – 55) for D isoform. None of healthy donors had detectable WT1 levels in peripheral blood. All samples of CD34+ cells expressed the four isoforms of WT1: 21% (range: 2 – 26) for A isoform; 16% (range: 1 – 64) for B isoform; 24% (range: 1 – 47) for C isoform; and 36% (range: 25 – 44) for D isoform. These data reveal that, in our series, the most predominant isoform was +5/+KTS, both in AML and in cord blood CD34+ cell selection samples. There were no significant differences when comparing the proportion of each isoform between the cord blood CD34+ cell selection samples and the cohort of AML patients. There was not significant correlation between the overexpression of total WT1 with the ratio of each isoform, and we were unable to demonstrate that the overexpression of WT1 is due to a particular isoform overexpression. A significant lower event-free survival (EFS) was observed in those patients overexpressing total WT1, taking a cut-off value of 3000 WT1 copies/ GUS copies × 10⁴ (75^th percentile, P =.001). However, when the same cut-off as well as the median value for each one of the isoforms was used, we found no significant differences in EFS and in overall survival.

To sum up, none of the isoforms were correlated with overexpression of total WT1 or survival. We were unable to find differences between the expression of each isoform of WT1 in CD34+ cells from normal cord blood and in AML patients. Further studies including larger controls need to be carried out.