Neddylation Pathway Regulates Treg Differentiation and T Cell Function in Chronic Lymphocytic Leukemia (CLL) Ex Vivo and Murine In Vivo Studies

Introduction: Inhibitors of B-cell receptor associated kinases (e.g., Bruton tyrosine kinase and phoshpoinotiside-3 kinase) have led to significant improvement in outcomes of patients with CLL. Furthermore, such therapies have immunomodulatory effects. This is particularly relevant given that T cells from patients with CLL are functionally compromised and demonstrate impaired immune synapse formation, adhesion, migration and cytotoxicity. It is critical to improve understanding of the immunomodulatory properties of the novel agents as this will help understand their effect on the immune system in CLL, mitigate toxicities as well as inform future drug development.

Pevonedistat, (MLN4924) forms an irreversible covalent adduct with NEDD8, a ubiquitin-like modifier, thereby inhibiting the NEDD8-activating enzyme (NAE). This leads to decreased neddylation and reduced activity of cullin-RING E3 ubiquitin ligases (CRLs). Ultimately, a decrease in CRL activity leads to reduced ubiquitination and proteasomal degradation of CRL substrates, extending the half-life of these proteins, including inhibitor of NFκB (IκB). We have shown that pevonedistat abrogates NFκB activation in CLL cells. Importantly, NFκB activation is indispensable in T-cell activation. However, there is paucity of data regarding the immune effects of targeting neddylation. Here we address this knowledge gap and demonstrate that NAE inhibition may have favorable immunomodulatory effects in CLL.

Methods: Peripheral blood mononuclear cells were isolated from patients with CLL and T cells were purified using Dynabeads. Pevonedistat was provided by Millennium Pharmaceuticals, Inc. (Cambridge, MA). For gene expression analysis, FACS-sorted naïve CD4+ T cells were pre-treated with pevonedistat for 1 hour prior to T-cell receptor (TCR; αCD3/CD28) stimulation; RNA was harvested 3 or 24 hours after stimulation and analyzed on a Clariom S microarray chip. For polarization assays, FACS-sorted naïve CD4+ T cells were TCR-activated for 7 days under Th1/2/17/Treg-differentiation conditions. BALB/c mice were administered 60 mg/kg pevonedistat SC twice weekly for 3 weeks and T cell populations were analyzed by flow cytometry.

Results: In vitro treatment of T cells with pevonedistat led to rapid reduction in neddylated cullins and stabilization of pIκBα. NAE inhibition did not impede proximal TCR signaling following TCR stimulation (pZAP70, pERK). GSEA demonstrated downmodulation of NFκB and IL-2 signaling pathways in pevonedistat-treated cells by 3 h. Despite this, CD4/CD8+ T cells exhibited normal induction of early activation markers (CD40L, CD69). By contrast, we observed reduced expression of CD38, HLA-DR and PD-1 and diminished CD25 following continuous treatment with pevonedistat for 72 h. This was accompanied by dose-dependent decrease in IL-2 secretion and reduced proliferation of the CD4/CD8+ T cell subsets (CFSE), but no apoptosis.

Sorted naive T cells treated with pevonedistat in Th/Treg polarizing conditions exhibited an increase in IFNγ secretion and a decrease in IL-4, suggesting a shift toward Th1 phenotype. Furthermore, we observed a robust decrease of the inducible Treg (iTreg) FoxP3+ population. Loss of iTregs was accompanied by ablated IL-2/STAT5 signaling. Concurrently, we observed a modest increase in Th17 subpopulation following NAE inhibition. We found increased expression of HIF-1α, a CRL target, in pevonedistat-treated T cells, which may have contributed to this phenomenon.

To mimic the clinical pharmacokinetics of pevonedistat, we performed 2 h pulse treatment with pevonedistat prior to TCR stimulation. Under these conditions NFκB activity fully recovered by 24 h. Importantly, allogeneic (OCI-LY19 cells) and autologous (CD40L-stimulated CLL cells) T-cell cytotoxicity, perforin and granzyme B production were not disrupted by NAE inhibition.

In vivo administration of pevonedistat in immunocompetent BALB/c mice resulted in a decrease of Treg population, confirming in vitro data.

Conclusions: Our data suggest that targeting neddylation may help rebalance T cells towards healthy immune subsets in CLL via the reduction of the Treg/Th2 phenotypes. Combined with our earlier reports that targeting NAE kills CLL cells under lymph node-mimicking conditions, these data provide a strong rationale for continued investigation of pevonedistat in CLL and lymphoid malignancies.