Cereblon Deficiency Leads to Alteration in c-Myc Regulated Genes, Increased Glucose Uptake and Resistance to JQ1 Bromodomain Inhibition

Background: Immunomodulatory drugs (IMiDs), including thalidomide and its analogs, are highly effective for the treatment of both non-del5q and del5q Myelodysplastic Syndrome (MDS) and other hematological cancers. IMiDs have also been shown to play an additional role as a potent T-cell stimulant for cancer therapy. The first identified target of IMiDs is cereblon, an E3 ubiquitin ligase substrate receptor. From crystal structure analysis of human and murine crbn/ddb1 bound to lenalidomide or thalidomide, IMiDs bind to a conserved hydrophobic pocket of crbn called the thalidomide-binding domain. To explore possible mechanisms of crbn regulation of immune response, we studied immune regulation in cereblondeficient mice (crbn-/-), which have increased T-cell proliferation and cytokine production.

Results: We have previously shown that purified T-cells from crbn deficient mice have greater immune activation with increased proliferation and pro-inflammatory cytokines, including IL-2, IFNγ, and TNFα, when compared to wild-type T-cells. Additionally crbn-/- T-cells mount a greater alloimmune response in vivo. To determine the molecular effects of crbn deficiency on T-cell response following activation, we analyzed gene expression by Affymetrix microarray in wild-type and crbn-/- purified T-cells before and 12 hours following activation with anti-CD3 and anti-CD28 antibodies. Based on hierarchal clustering, the majority of gene expression changes were driven by T-cell activation. To explore the effect of crbn deficiency on T-cell activation, we compared genes from wild type and crbn-/- T-cells that showed a 2-fold (FDR<0.05) expression change following activation. Wild-type and crbn-/- T-cells shared an overlapping set of 3795 probesets. However, 798 and 3226 probesets were upregulated 2-fold only in wild-type and crbn-/- T-cells respectively following activation. Gene ontology analysis of the crbn-/- only subset showed enrichment for metabolic processes. Utilizing GSEA analysis, we identified c-Myc as an enriched transcription factor that regulates genes found in the crbn-/- only gene subset. Although no differences were seen in c-Myc expression by qRT-PCR, the c-Myc protein was increased in crbn-/- T-cells at earlier times of activation and prolonged through 48hrs post-activation compared to wild-type T-cells that showed a shorter duration of Myc activation. Indicative of differences in c-Myc activation, the c-Myc responsive gene known as SLC2A1, encoding glucose transporter 1 (or GLUT1), was elevated in crbn-/- T-cells compared to WT T-cells. Consistent with the role of GLUT1 in the facilitation of glucose transport across plasma membranes, crbn-/- T-cells showed an increased uptake of 2-NBDG, a fluorescent glucose analog. Treatment of the cells with JQ-1, an inhibitor of BET bromodomains and a suppressor of c-Myc expression, resulted in a decrease in wild-type T-cell proliferation while crbn-/- T-cells showed less sensitivity to JQ1 inhibition of proliferation.

Conclusions: Our results indicate that cereblon deficiency results in an increased immune response in T-cells. Analyzing gene expression pattern differences between wild-type and crbn-/- T-cells, we identified c-Myc as a possible driver of this phenotype. Interestingly, c-Myc regulates the metabolic switch from oxidative phosphorylation to glycolysis and expression levels of c-Myc can effect proliferation and cytokine production, both of which are upregulated in crbn-/- T-cells. The prolonged expression of c-Myc protein could be responsible for the increase in Glut1 expression, glucose uptake, and resistance to JQ1. Overall, cereblon may negatively regulate c-Myc and the metabolic switch to glycolysis, resulting in crbn-/- T-cells having an increased glycolytic phenotype following activation.