Mutations in TP53, or less often its regulators, increases risk for malignant transformation. Murine double minute protein 2 (MDM2), an E3 ubiquitin ligase, targets p53 for proteasomal degradation and is the most well studied negative regulator of p53. Recent investigations have highlighted the emerging importance of p53 in MDS. Haploinsufficiency for ribosomal protein S14 in deletion 5q MDS liberates free ribosomal proteins that bind to and promote degradation of MDM2, thereby activating p53 in erythroid precursors. A single nucleotide polymorphism (SNP) in an MDM2 promoter (SNP309) is linked to younger age of onset of several solid tumors and an increased risk for acute myeloid leukemia (AML) [Knappskog and Lonning. 2011. Transcription 2:207, Xiang et al. 2009. Leuk Res. 33:1454]. The thymine (T) to guanine (G) substitution introduces an additional Sp1 transcription factor binding site causing upregulation of MDM2 transcription. A second SNP in this promoter (SNP285) has also been linked to cancer susceptibility, where a guanine (G) to cytosine (C) exchange is associated with decreased ovarian and breast cancer risk (Knappskog and Lonning. 2011. Oncotarget. 2:251). The C-allele of SNP285 has diminished Sp1 promoter binding compared to the G-allele decreasing MDM2 expression. In this study we investigated genotype distribution of MDM2 SNPs in del(5q) and non-del(5q) MDS patients and compared results to healthy controls.
Using Sanger sequencing, we compared allele and genotype frequencies for SNP285 and SNP309 in 155 healthy controls, 97 non-del(5q), and 119 del(5q) MDS patients. For SNP285, we found no significant difference in genotype or allele frequency among non-del(5q) or del(5q) cases compared to controls (p=0.25 and 0.26, respectively). Although there was no difference in age at diagnosis by genotype in del(5q) MDS (p=0.82), there was a significant difference among non-del(5q) MDS cases [p<0.01, mean (range) for GC:80.8y (75–89) and GG:68.9y (27–91)], however, the frequency of the GC genotype was low [n=5, non-del(5q); n=11, del(5q)] with no CC genotype cases. For SNP309, there was no difference in allele frequency (p=0.68), however genotype frequency differed between controls, non-del5q, and del5q MDS (p=0.06). The genotype distribution was significantly different between non-del(5q) and del(5q) MDS (p=0.01). SNP309 genotype frequencies for controls, non-del(5q), and del(5q) MDS were GG:17.3%, 22.7%, 10.1%; TG: 42.9%, 37.1%, 53.8%; TT: 39.7%, 40.2%, 36.1%, respectively. We found no difference in age of disease onset by SNP309 genotype in either non-del(5q) or del(5q) cases (p=0.08 and 0.97, respectively). There was no significant relationship between SNP285 genotype and IPSS (p=1.0), cytogenetic risk (p=0.66), or WHO classification (p=0.16) in non-del(5q) or del(5q) cases (p=1.0, 0.78, and 0.60, respectively). Similar results were observed for SNP309 [p=0.85, 0.39, and 0.68 for non-del(5q); p= 0.06, 0.98, 0.27 for del(5q), respectively]. For SNP285 there was no difference in overall survival (OS) by genotype in either non-del(5q) (p=0.65) or del(5q) MDS (p=0.72). Progression free survival (PFS) also did not differ by genotype in non-del(5q) (p=0.82) or del(5q) (p=0.58) patients. There was a significant difference in LEN response rate in del(5q) MDS (p=0.04, non-responders: 23.8% GC and 76.2% GG, responders: 4.9% GC and 95.1% GG), however, genotype did not influence response duration (p= 0.40). For SNP309, there were no significant difference in OS by genotype (p=0.42), PFS (p=0.78), LEN response rate (p=0.17), or response duration in del(5q) MDS (p=0.65). In non-del(5q), there was no difference in OS (p=0.42), LEN response rate (p=0.91), or response duration (p=0.47). However, we found a significant difference in PFS by genotype (p=0.03) with more prolonged PFS in patients with the heterozygous TG genotype (60 mo PFS: GG 55.3%, TT 54.1%, and TG 81.7%). Finally, we found no difference in chromosome 5 deletion size or deletion location in del(5q) MDS according to SNP309 genotype.
MDM2 SNP309 may be linked to MDS susceptibility, as well as LEN responsiveness and PFS in del(5q) and non-del(5q) MDS, respectively. These data warrant validation in a larger patient cohort. Investigation of the interaction between MDM2 SNPs and the well described TP53 R72P SNP is underway.
Asterisk with author names denotes non-ASH members.