Age-related changes in the BACH2 and PRDM1 genes in lymphocytes from healthy donors and chronic lymphocytic leukemia patients.

BACKGROUND: Age-related genetic changes in lymphocyte subsets are not currently well documented. BACH2 is a transcription factor that plays an important role in immune-mediated homeostasis by tightly regulating PRDM1 expression in both B-cells and T-cells. BACH2 gene expression is highly sensitive to DNA damage in aged mice. This concept led us to investigate the variation in BACH2 and also PRDM1 expression in major lymphocyte subsets with age. METHODS: Lymphocyte subsets from 60 healthy donors, aged from 20 to 90 years, and 41 untreated chronic lymphocytic leukemia patients were studied. BACH2 and PRDM1 gene expression was analyzed by real-time quantitative PCR. BACH2 gene expression was correlated with its protein expression. Lymphocyte apoptosis was evaluated after intracellular oxidative stress-inducing etoposide treatment of T and B cells. RESULTS: Our analysis shows BACH2 mRNA downregulation with age in healthy donor CD4+, CD8+ T-cells and CD19+ B-cells. Decreased BACH2 expression was also correlated with an age-related reduction in CD8 + CD28+ T-cells. We found a strong correlation between age-related BACH2 downregulation and decreased CD4+ T-cell and CD19+ B-cell apoptosis. PRDM1, as expected, was significantly upregulated in CD4+ T-cells, CD8+ T-cells and CD19+ B-cells, and inversely correlated with BACH2. A comparison of untreated chronic lymphocytic leukemia patients with age-matched healthy donors reveals that BACH2 mRNA expression was further reduced in CD4+ T-cells, CD8+ T-cells and leukemic-B cells. PRDM1 gene expression was consequently significantly upregulated in CD4+ and CD8+ T-cells in chronic lymphocytic leukemia patients but not in their leukemic B-cells. CONCLUSION: Overall, our data suggest that BACH2 and PRDM1 genes are significantly correlated with age in human immune cells and may be involved in immunosenescence.

Comparative pharmacogenetic analysis of risk polymorphisms in Caucasian and Vietnamese children with acute lymphoblastic leukemia: prediction of therapeutic outcome?

AIMS: Acute lymphoblastic leukemia (ALL) is the most common of all paediatric cancers. Aside from predisposing to ALL, polymorphisms could also be associated with poor outcome. Indeed, genetic variations involved in drug metabolism could, at least partially, be responsible for heterogeneous responses to standardized leukemia treatments, hence requiring more personalized therapy. The aims of this study were to (a) to determine the prevalence of seven common genetic polymorphisms including those that affect the folate and/or thiopurine metabolic pathways, i.e. cyclin D1 (CCND1-G870A), γ-glutamyl hydrolase (GGH-C452T), methylenetetrahydrofolate reductase (MTHFR-C677T and MTHFR-A1298C), thymidylate synthase promoter (TYMS-TSER), thiopurine methyltransferase (TPMT*3A and TPMT*3C) and inosine triphosphate pyrophosphatase (ITPA-C94A), in Caucasian (n = 94, age < 20) and Vietnamese (n = 141, age < 16 years) childhood ALL and (b) to assess the impact of a multilocus genetic risk score (MGRS) on relapse-free survival (RFS) using a Cox proportional-hazards regression model. RESULTS: The prevalence of MTHFR-677TT genotype was significantly higher in Caucasians (P = 0.008), in contrast to the prevalence of TYMS-TSER*3R/3R and ITPA-94AA/AC genotypes which were significantly higher in Vietnamese (P < 0.001 and P = 0.02, respectively). Compared with children with a low MGRS (≤ 3), those with a high MGRS (≥ 4) were 2.06 (95% CI = 1.01, 4.22; P = 0.04) times more likely to relapse. Adding MGRS into a multivariate Cox regression model with race/ethnicity and four clinical variables improved the predictive accuracy of the model (AUC from 0.682 to 0.709 at 24 months). CONCLUSION: Including MGRS into a clinical model improved the predictive accuracy of short and medium term prognosis, hence confirming the association between well determined pharmacogenotypes and outcome of paediatric ALL. Whether variants on other genes associated with folate metabolism can substantially improve the predictive value of current MGRS is not known but deserves further evaluation.