Imprinted Non-Coding RNA Genes, H19 and ZNF215, Are Potential Markers For Two-Year Survival For Acute Myeloid Leukemia Patients With Normal and Abnormal Karyotypes

Studies in large-scale genome sequencing have shown that only 2% of the mammalian genome encodes mRNAs, but the most part is transcribed as long and short non-coding RNAs (ncRNAs). The ncRNAs with gene regulatory functions are starting to be seen as a common feature of mammalian gene regulation. Genomic imprinting is a form of epigenetic regulation and imprinted genes are silenced in a parental-specific manner. Although the exact mechanism how imprinted ncRNA regulates gene expression remains largely unknown, it is general accepted that imprinted ncRNAs binds to chromatin modifying complexes, such as PRC2, TRX, and G9a, and generates specific silencing of genomic loci both in cis and trans. Imprinting is associated with many human diseases or syndromes (e.g. Prader-Willi, Angelman, Beckwith-Wiedemann, Retts, and Silver-Russell syndromes) and various cancers (e.g. breast, prostate, and colorectal cancers), but its role in leukemogenesis remain elusive. In this present study, the expression of a panel of 24 human imprinted ncRNA genes (AMPD3, C15orf2, COPG2, CPA4, GABRB3, H19, IGF2, IMPACT, INPP5F, L3MBTL, NR3251, NR3252, PEG3-AS, PPP1R9A, PRIM2, RASGRF1, RTL1, SFMBT2, SLC22A3, SNURF, TCEB3C, TSPAN32, ZNF215, ZNF264) and a panel of 66 human histone modifying enzymes (HME) genes was investigated in 68 newly-diagnosed acute myeloid leukemia patients with chromosome normal (AML-CN), 115 AML patients with chromosome abnormal (AML-CA), and 85 healthy individuals using real-time quantitative RT-PCR. Altered expression of 9 imprinted ncRNA genes (C15orf2, COPG2, H19, IGF2, IMPACT, PEG3-AS, PRIM2, SLC22A3, ZNF215) and 16 HME genes were observed. In AML-CN, patients’ survival days are correlated with the expression levels of H19 (p < 0.01), IMPACT (p < 0.05), DNMT3L (p < 0.05) and AURORA (p < 0.01). In AML-CA, patients’ survival days are correlated with the expression levels of PGE3-AS (p < 0.01), PRIM2 (p < 0.01), SLC22A3 (p < 0.05), and ZNF215 (p < 0.01). Multiple linear regression analysis further revealed the expression level of H19 and ZNF215 can be used as predictors for 2-year survival for AML-CN patients (p = 0.002) and AML-CA patients (p = 0.040), respectively. Cox proportional hazard model was used to analyze the hazard ratio (HR) for H19 (HR=0.868, 95.0% Confident Interval: 0.797-0.945, p = 0.001) and ZNF215 (HR=0.904, 95.0% Confident Interval: 0.821-0.995, p =0.040). In addition to survival, analysis has also been performed to correlate patients’ clinical parameters and expression levels of these altered genes and to correlate the expression levels between imprinted ncRNA genes and HME genes (results will be presented at the meeting). From our preliminary results, it is reasonable to hypothesize that loss imprinting of imprinted ncRNA is critical for the leukemogenesis of AML and under CN or CA conditions different ncRNAs are activated and affect different imprinted gene expression and thus leading to different clinical outcomes. Based on our findings, we will further perform in vitro functional analysis to elucidate the functions and mechanisms of these imprinted ncRNAs in AML tumorigenesis. Updated results of these analyzes will be presented at the meeting.