2006 - Impact of nucleotide metabolism on CLL B-cell pathobiology and CLL disease progression

It has become increasingly clear that external signals from the leukemic microenvironment make pivotal contributions to disease progression in CLL. Multiple studies including ours have demonstrated that bone marrow stromal cells (BMSCs) promote CLL B-cell survival and protect these cells from the cytotoxic effects of therapy, but the complete nature of this interaction is still under studied.
To further uncover the mechanism of this interaction, we performed RNA-seq in paired CLL B-cells isolated from untreated blood and bone marrow (BM) (n=6) and in untreated CLL B-cells (n=4) cultured alone or co-cultured with primary BMSCs derived from normal individuals (n=2) or CLL patients (n=2). We found upregulation of 232 and 917 genes in CLL B-cells from BM and in co-cultured CLL B-cells compared to CLL B-cells from the same patient’s blood and CLL B-cells cultured alone respectively (p < 0.05, fold change >1.5). However, we found only 13 genes overlapped between BM CLL B-cells and co-cultured CLL B-cells (Figure 1). To explore the functional importance of these 13 genes, we analyzed the expression of these genes in CLL blood B-cells from an additional 162 untreated patients by RNA-seq and compared them with their clinical outcomes. We found a positive relationship between 4 of these 13 genes (PNP, C16orf54, MOB3A, CDK2AP2) with CLL patient’s clinical outcomes including overall survival (OS), progression-free survival (PFS), and time to first treatment (TTFT) (Figure 1) using Cox multivariable proportional hazard model (MCPHR p < 0.05). Out of these 4 genes, purine nucleotide phosphorylase (PNP) an enzyme in the purine salvage pathway, showed the most significant association for patient’s OS, PFS, and TTFT using MCPHR [hazard ratio (HR) {HR (95% CI)}: 2.2 (1.4-3.7), 2.2 (1.5-3.2), 2.5 (1.7-3.7) respectively, p < 0.005]. We also observed an increase in PNP protein levels in CLL B-cells co-cultured with BMSCs or from BM using Western blot (WB) analysis but only when baseline PNP protein is low. Additionally, untargeted metabolomic profiling (LC-MS+GC-MS) showed an increased level of purine salvage pathway metabolites adenosine, inosine, and hypoxanthine in CLL B-cells when co-cultured with BMSCs. When we treated high and low PNP expressing CLL B-cells from untreated patients with drugs known to be active in CLL, interestingly we found high PNP expressing CLL B-cells are sensitive to Bcl2 inhibitor venetoclax and AXL kinase inhibitor TP-0903 versus low PNP expressing CLL B-cells. These results suggest purine metabolism plays an important role in CLL B-cell survival and /or drug resistance. Consistent with this altered purine nucleotide metabolism data, we also found increased pyrimidine pathway enzyme thymidylate synthase (TYMS) at both RNA and protein levels in co-cultured CLL B-cells and in CLL B-cells isolated from BM. We further analyzed the RNA-seq data generated from the 162 CLL patient cohort and found a positive association of TYMS gene with their TTFT (Figure 1) using MCPHR [HR: 1.4, 95% CI (1-1.8), p < 0.05]. However treatment of CLL B-cells from untreated patients cultured alone or co-cultured with BMSCs with TYMS inhibitor raltitrexed did not show any killing indicating an alternate role of TYMS in BMSC cultured CLL B-cells. To that end we have detected increased expression of epithelial to mesenchymal (EMT) markers with increased TYMS expression in co-cultured CLL B-cells compared to CLL B-cells cultured alone by WB analysis.
In conclusion, these results indicate that active purine and pyrimidine metabolism in CLL B-cells augmented by the BM environment can support CLL B-cell survival and drug resistance and possible induction of EMT components. Future studies are underway to understand the role and regulation of PNP and TYMS in CLL disease progression and drug resistance for both untreated and treated CLL patients.