Cell Death Regulatory Gene Expression Correlates with MLL Rearrangement Status and Prognostic Clinical Covariates in Acute Leukemia in Infants.

Infants with acute leukemia, especially neonates, comprise a unique high-risk population prone to chemotherapy resistance and treatment complications. Basal levels of expression of cell death and cell survival factors vary greatly in different types of cancer and are key determinants of response to treatment. Until now, expression of these factors had not been evaluated in detail in acute leukemia of infants. In this study the basal expression of pro-death and pro-survival genes in the intrinsic apoptosis pathway, apoptosis execution genes and mitochondrial maintenance genes was quantified in primary leukemia cells in a large cohort of infants, and relationships between gene expression and MLL status and clinical covariates were analyzed.

Methods: Bone marrow or peripheral blood samples from the time of diagnosis were examined in 103 infants (age 1–363 d, median 177 d) with acute leukemia (63 ALL, 40 AML). White blood cell (WBC) counts ranged from 4.4–1676×10³/μL (median 127.6×10³/μL). MLL status was characterized by standard cytogenetics, FISH, Southern blot analysis, reverse transcriptase PCR and/or panhandle PCR approaches. The ALL cases were classified as MLL-AF4⁺ (n= 31), other MLL rearrangement positive (other MLL⁺) (n= 22) or MLL rearrangement negative (MLL⁻) (n= 10). Due to heterogeneity in partner genes, the AML cases were classified as MLL⁺ (n= 22) or MLL⁻ (n= 18). Quantitative real time PCR analysis of total RNAs was performed using a custom TaqMan^® low-density array with one or more assays, each in quadruplicate, for 35 cell death regulatory genes and 3 housekeeping genes, and ΔC_T values (average cycle threshold values normalized to 18S rRNA) were computed. The Behrens-Fisher nonparametric multiple comparison test in the npmc package for R was used to determine the significance of the difference in median −ΔC_T values among the three ALL groups and between the two AML groups (p<0.001 considered as significant). Relationships were sought between −ΔC_T values and clinical covariates (age, sex, WBC count, CNS status, events) by computing Spearman correlation coefficients and their levels of significance.

Results: In ALL, expression of the following genes was greater in MLL-AF4⁺ than MLL⁻ cases: anti-apoptotic BCL-2 and MCL-1; pro-apoptotic BAX and BID; apoptosis execution APAF1, CASP3 and CASP9; and mitochondrial maintenance COX5A, CYC1, UQCR1 and UQCR. For some of these genes, expression was also greater in other MLL⁺ than MLL⁻ cases. In AML, expression of pro-apoptotic BAK and BIK was higher in MLL⁺ than MLL⁻ cases. For the sample set as a whole, expression of the following genes was positively correlated with WBC count (r>0.3; p<0.01): anti-apoptotic BCL-2 and BCL-W; pro-apoptotic NOXA; apoptosis execution CASP8; and mitochondrial maintenance HCCS. The correlation between WBC count and BCL-2 was the strongest (r=0.40;p=4E-5). Only the expression of pro-apoptotic BIM was correlated with age at diagnosis and the correlation was not particularly strong (r = −0.246;p=0.01).

Conclusions: These results suggest that infant ALL and AML differ with respect to the expression of genes that regulate cell death. Furthermore, cell death regulatory gene expression is affected more by MLL translocations in infant ALL than infant AML. This finding may explain the greater impact of MLL translocations on prognosis in infant ALL than AML. The abundance of specific pro- as well as anti-apoptotic BCL-2 family mRNAs in MLL⁺ infant ALL is also relevant to the identification of novel therapeutic targets in this pathway. The positive correlation between anti-apoptotic BCL-2 expression and WBC count is biologically plausible and consistent with the aggressive nature of acute leukemia in infants.