Abstract

The prognosis and treatment of both major forms of advanced childhood B-NHL (BL and DLBCL) is similar with short and intensive multi-agent chemotherapy (Cairo/Patte et al., Blood, 2007 and Patte/Cairo et al., Blood, 2007). Despite both BL and DLBCL being germinal center derived, our recent cytogenetic results of BL vs DLBCL in the FAB LMB 96 study have demonstrated significant differences in secondary chromosomal aberrations in BL vs DLBCL and a differential prognosis based on secondary cytogenetic findings (Poirel/Cairo/Patte, Blood, 2003a). Thus, we sought to identify genes that could uniquely differentiate childhood BL vs DLBCL and discover potential genetic mechanisms of differential molecular pathogenesis and to determine the signal pathways that contribute to the genetic disparity between these two histological types of childhood B-NHL. Nine BL (7 patient samples and 2 cell lines, Raji and Ramos) and 3 DLBCL (1 patient sample and 2 cell lines, Pfeiffer and DB) were compared. Total RNA was isolated, reverse transcribed to cDNA biotinylated cRNA and hybridized to Affymetrix U133A_2 as we have previously described (Jiang/Cairo et al., Journal of Immunology, 2004). Data were analyzed using Agilent GeneSpring 7.3. Signal intensities were compared using one way ANOVA and Welch Test for statistical analysis. Two-fold changes between BL and DLBCL were considered as significant (p<0.05). KEGG Pathways were evaluated for the genes identified. There were 120 genes over-expressed and 217 genes under-expressed in BL vs DLBCL. BL expressed significantly higher level of Ki-67 (a measure of lymphoma-cell proliferation) than DLBCL (2.68F). BL also expressed higher level of the pro-apoptotic gene, p53 compared to DLBCL (1.46F). Over-expressed genes in BL vs DLBCL included TNFSF10 (11.87F), RHOQ (3.16F), PIP5K1B (5.22F) among many others. The genes significantly under-expressed in BL vs DLBCL included PIGL (0.45F), Inositol (myo)-1 (or 4)-monophosphatase 1 (IMPA1; 0.28F), cAMP-dependent regulatory type I, alpha protein kinase (PRKAR1A; 0.37F) among many others. TNFSF10 induces apoptosis in transformed and tumor cells and is known to participate in pathways including cytokine-cytokine receptor interaction and induction of apoptosis through DR3 and DR4/5 death receptors. PIP5K1B is involved in the Rho signaling pathway and PIGL catalyzes the second step of glycosylphosphatidylinositol (GPI) biosynthesis. Since activation of IL3R-mediated cAMP-dependent protein kinase leads to increased cell survival, we searched gene expression profiles in BL vs DLBCL that were involved in IL signaling pathways. The genes that were identified to be over-expressed in BL vs DLBCL included IL2RG (2.24F), IL8RB, IL18 receptor accessory protein (IL18RAP), IL18, IL18R1, and IL1R2 (natural log values of 11.11, 22.95, 2.16, 1.73 and 11.84, respectively in BL vs non-detectable values in DLBCL). Taken together, since IL1, IL2, IL8, and IL18 all belong to IL1 super family, these results suggest significant involvement of TNF (TRAIL) and IL1 super family via cytokine-cytokine receptor interaction and activation of the Rho signaling pathway in Burkitt vs DLBCL lymphomagenesis.

Author notes

Disclosure:Research Funding: The Pediatric Cancer Research foundation.