In 2012 (Xie L. et al., Blood, 2012) we paid attention on the low expression of FOXO1 in classical Hodgkin lymphoma in comparison with normal B cells and non-Hodgkin B cell lymphomas (cHL). cHL is a B-cell lymphoma which mainly lost its B-cell program. We have shown that overexpression of FOXO1 induced apoptosis in cHL cell lines and concluded that FOXO1 is a tumor suppressor in cHL. Latter we interrogated the molecular effects of FOXO1 overexpression in cHL with help of gene expression profiling. We identified the master regulator of plasma cell differentiation, PRDM1, as transcriptional target of FOXO1 which mediates its tumor suppressor function. Further, we demonstrated that the tumor-suppressor effect of PRDM1 depends on downregulation of the proto-oncogene MYC (Fig. 1). Moreover, we supposed that low FOXO1 expression is responsible for the characteristic “abortive plasma cell differentiation” phenotype of Hodgkin and Reed-Sternberg cells, the malignant component of cHL.

Continuing to investigate the role of FOXO1 in B cell lymphomas we paid attention on the reciprocal character of FOXO1 and FOXO3A mRNA expression in the process of mature B cell differentiation from germinal center dark zone (DZ) B cells to the plasma cells (PCs). The expression levels of FOXO1 decreased in the course of terminal differentiation from DZ B cell to PCs, whereas the FOXO1 mRNA levels increased. We noticed that FOXO3A mRNA levels in cHL were higher than in DZ B cells but lower than in PCs. We have further shown that this intermediate FOXO3A expression levels are essential for the survival of cHL, as either increase or decrease of the FOXO3A dosage was inappropriate (Fig. 2).

We concluded that cHL maintains FOXO3A/FOXO1 expression levels which reflect the differentiation status of the normal precursor cell of cHL, and that maintenance of the proliferation and survival program of this normal precursor is essential for the survival of cHL. In this way we postulated a Goldilocks principle for the FOXO effects in B cell lymphoma.

Searching for support of our Goldilocks hypothesis we found that B-ALL, which is highly sensitive to FOXO-activating treatments (e.g. inhibitors of AKT or overexpression of FOXO1) maintain high FOXO1 expression in the nuclei, indicating a role of FOXO1 in B-ALL. We have shown that genetic depletion of FOXO1 induces growth arrest and apoptosis in B-ALL cell lines (Fig. 3). Most importantly, pharmacological inhibition of FOXO1 showed an anti-leukemia activity on several primary patient-derived pediatric ALL xenografts with effective leukemia reduction in the hematopoietic, lymphoid, and central nervous system organ compartments, ultimately leading to prolonged survival without leukemia reoccurrence in a preclinical in vivo model of BCP-ALL. These results suggested that repression of FOXO1 might be a feasible approach for the treatment of BCP-ALL.

Re-analyzing our old data on FOXO1 expression in B cell non-Hodgkin lymphoma subtypes we identified high FOXO1 expression in nuclei of Burkitt lymphoma (BL). With help of loss-of-function experiments we identified an anti-proliferative effect of FOXO1 knockdown in BL cell lines, which was associated with the repression of the DZ B cell program including expression of MYB, CCND3, RAG2, BACH2, and CXCR4. In addition, the induction of signaling pathways of the light zone program like NF-κB and PI3K-AKT was observed. Using a rescue experiment we identified downregulation of the proto-oncogene MYB as a critical factor contributing to the antiproliferative effect of FOXO1 knockdown. In an attempt to estimate the feasibility of pharmacological FOXO1 repression, we found that the small molecular weight FOXO1 inhibitor AS1842856 induced cell death and growth arrest in BL cell lines at low concentrations. As in case of cHL and B-ALL, we found that overactivation of FOXO1 also induced growth inhibition in BL cell lines, indicating the importance of a tight regulation of the FOXO1 activity in BL (Fig. 4).

Investigating the role of FOXO1 in the survival and proliferation program of BL, we also addressed the apparent paradox of the preferably nuclear FOXO1 expression and allegedly high activity of FOXO-inactivating PI3K-PDPK1-AKT pathway in BL. We found that the PI3K-PDPK1-AKT activity does not exceed physiological levels typical for germinal center B cells from which BL derives. Both overexpression of the constitutively active version of AKT (myrAKT) and knockdown of the natural PI3K-AKT signaling inhibitor PTEN resulted in FOXO1 inactivation and repressed the DZ proliferation and survival program and consequently the growth of BL (Fig. 5).

Previously we reported that inactivation of FOXO1 confers growth arrest and apoptosis in B-ALL, partially mediated by subsequent depletion of CCND3 (Wang F. et al., 2018). Given that previously the canonical MYC target CCND2 has been considered to play a major role in B-ALL proliferation, further investigation of the link between FOXO1 and CCND3 in B-ALL was warranted. In this study, we demonstrated that CCND3 is essential for the proliferation and survival of B-ALL, independent of the mutational background. Respectively, its expression at mRNA level exceeded that of CCND1 and CCND2. Furthermore, we identified FOXO1 as a CCND3-activating transcription factor in B-ALL. By comparing the effects of CCND3 depletion and CDK4/6 inhibition by palbociclib on B-ALL cells harboring different driver mutations, we found that the anti-apoptotic effect of CCND3 is independent of the kinase activity of the CCND3-CDK4/6 complex. Moreover, we found that CCND3 contributes to CDK8 transcription, which in part might explain the anti-apoptotic effect of CCND3. Finally, we found that increased CCND3 expression is associated with the development of resistance to palbociclib. We conclude that CCND3 plays an essential role in the maintenance of B-ALL, regardless of the underlying driver mutation. Moreover, downregulation of CCND3 expression might be superior to inhibition of CDK4/6 kinase activity in terms of B-ALL treatment.