MYCN amplification is a key bottleneck restricting the overall efficacy of NB and is one of the major causes of treatment failure. Disrupting MYCN expression impedes proliferation and metastasis and induces differentiation and death of MYCN-AM NB cells (33–35). Although MYCN deserves to be considered as an attractive candidate target for high-risk NB, direct targeting of MYCN proteins is fraught with challenges owing to the lack of surface structures for small molecules binding and the enzymatic activity that can be specifically inhibited. Given this, indirect targeting strategies through manipulating interactive molecules associated with MYCN have been rapidly discovered and developed over the years, which has become a research hotspot in the NB field. The alternative MYCN-targeting strategies mainly involve the transcription, translation, protein stability and target gene transcription of MYCN (7). For example, ALDH18A1 has been reported as a target gene of MYCN and its protein reciprocally participates in the regulation of MYCN transcriptional activation, thus forming a positive feedback loop (25). YG1702 was identified as an inhibitor of ALDH18A1 based on a computer-assisted virtual screen and displayed potent inhibition of MYCN-driven tumorigenicity in cellular and animal experiments. Aurora-A was reported to bind to MYCN protein through the same motif recognized by Fbxw7 ubiquitin ligase in MYCN Box I domain, resulting in an effective reduction of MYCN affinity for Fbxw7 (9, 19). Currently, MLN8237 (alisertib) has been investigated in clinical trials for the treatment of recurrent NB as an inhibitor capable of distorting the Aurora-A conformation (36, 37). Similarly, the polycomb repressive complex 2 component EZH2 was shown to counteract Fbxw7-mediated polyubiquitination and proteasomal degradation of MYCN by competing with Fbxw7 for MYCN binding in a methyltransferase-independent manner (38). And the EZH2 depleting agent DZNep but not the enzyme inhibitor was found to induce the degradation of MYCN and arrest the growth of tumor cells (38). Moreover, EZH2 is directly positively regulated by MYCN transcriptionally and is also involved in MYCN-mediated transcription of some target genes (10, 38, 39). These indicate that drug development based on the identified key molecules that regulate MYCN or are regulated by MYCN will definitely be a promising approach.
Studies have confirmed that there are significant differences in gene expression profiles between MYCN-AM and NA NB. In addition to MYCN itself, the growth of MYCN-AM NB cells is highly dependent on the certain genes downstream of MYCN, that are unnecessary for MYCN-NA NB cells (9, 12). Encouragingly, manipulation of these genes would selectively suppress the malignant biological phenotype of MYCN-AM NB, which exemplifies the promising potential of this strategy in drug development and clinical applications (32, 40, 41). Therefore, it is of profound relevance to search for potential regulatory factors and elaborate new strategies based on the specific activation pathways of MYCN amplification from the perspective of "selective inhibition". Here, we identified 22 shared differential genes based on transcriptomic differences between MYCN-AM and NA NB tissues and cells in multiple datasets (Fig. 1A-B), which also contain several potential targets for NB that have been and are being investigated in NB, such as ODC1, PHGDH and AURKB (42–44). Gene expression correlation analysis revealed a high positive correlation between CCNB1IP1 and MYCN expression in all the four datasets (Fig. 1C-G). Our results showed that CCNB1IP1 was overexpressed in both MYCN-AM NB patient specimens and cell lines, which is consistent with the database analysis (Fig. 2A-G). And its higher expression level is closely associated with lower OS, EFS of patients, which further implied the malignant function of CCNB1IP1 in NB. All these have aroused our curiosity about the role and function of CCNB1IP1 in MYCN-AM NB (Fig. 2H-K).
CCNB1IP1 was initially identified as an interacting protein of cyclin B1 for the regulation of cell cycle progression (13). To date, the biological function of CCNB1IP1 in a variety of tumors has been noted in several studies. Some of these studies stand on the side that CCNB1IP1 may produce a positive effect on tumorigenesis. Earlier, CCNB1IP1 was found to be a component of HMG1C gene translocation fusion in uterine leiomyoma, but its biological effect has hardly been discussed (14). Similarly, CCNB1IP1 was found aberrantly overexpressed in metastatic melanoma and hepatocellular carcinoma (15, 16). However, some results contradicting the above studies have also been reported. In an in-situ hybridization study based on patient-derived tumor tissue microarrays from multiple cancer types, CCNB1IP1 was observed to be under-expressed in colorectal, breast and non-small cell lung cancer (NSCLC), and its expression level was negatively correlated with survival time in adenocarcinoma, small cell squamous carcinoma and NSCLC (15). Furthermore, silencing of CCNB1IP1 in gastric cancer cells U2OS and breast cancer cells MCF-7 promoted cells metastasis and invasion, but suppressed cells proliferation and growth, suggesting that the extent or nature of CCNB1IP1 being required in different biological behaviors of tumor cells may not be consistent (17). Notably, a recent study based on transcriptome analysis of a MYCN-driven NB mouse model revealed that CCNB1IP1 was highly expressed in recurrent metastatic tissues (18). Although the role of CCNB1IP1 has not been further investigated, the results partially implicated an oncogenic function of CCNB1IP1 in NB. In this study, a significant oncogenic driving effect of CCNB1IP1 on MYCN-expressing or AM NB cells was found in cellular and xenograft tumor studies, as manifested by a greater proliferation and tumor formation, while having little effect on cells with low MYCN expression or NA. In contrast, even restoration of CCNB1IP1 expression failed to rescue the diminished proliferative capacity and tumor growth of NB cells caused by MYCN deficiency (Fig. 6). We hypothesized that the oncogenic effect of CCNB1IP1 probably exerts effectively in NB cells with abnormally high expression of MYCN, whereas in MYCN-NA cells, a gene network sufficient for CCNB1IP1-induced cell carcinogenesis seems to be infeasible. All these inspired us that CCNB1IP1 may cooperate with MYCN to drive the tumorigenic development of NB, and its specific biological role in NB needs to be further explored in a mouse model of NB primary that better resembles the characteristics of tumorigenesis and progression.
Interestingly, CCNB1IP1 acts as the target gene of MYCN that reciprocally maintained the protein stability of MYCN (Fig. 5). Interfering with the post-transcriptional stability or activity of MYCN is relatively straightforward and is receiving increasing attention to develop more effective and feasible options to attack MYCN. According to our data, the regulation of MYCN protein ability by CCNB1IP1 is mainly associated with E3 ubiquitin ligase Fbxw7 and is not dependent on other partner molecules interacting with MYCN (Trim32, Huwe1, USP3 and USP5) (Fig. 7 and Fig. S2) (9, 28–31). In truncated mutation experiments, we found that CCNB1IP1 bound tightly to the reciprocal region of Fbxw7 on MYCN AA 48–89 in a C-terminal domain-dependent manner (Fig. 8). Interestingly, MYCN with deletion mutations in this region lost its ability to bind Fbxw7, and binding to CCNB1IP1 was also significantly abolished (Fig. 8). This indicates that the Fbxw7 binding segment is contained at least within the region where MYCN interacts with CCNB1IP1. In contrast, the MYCN AA 48–89 region is not be required for the binding of other MYCN cofactors (Trim32, Huwe1, USP3 and USP5). Therefore, CCNB1IP1 would not disturb their binding to MYCN or further affect the stability of MYCN protein, which presumably corroborated with the results in Fig. 7 and Fig. S2.
Currently, numerous studies suggest that destabilization of MYCN has great potential as an emerging therapeutic strategy, although it also faces various obstacles, such as specificity and clinical translation issues. For example, AURKA kinase is currently being studied as a well-established therapeutic target for MYCN-AM NB. Based on the interaction of AURKA with MYCN, several small molecules have been identified which can bind to AURKA and alter its conformation to disrupt the Fbxw7-MYCN complex and cause rapid MYCN degradation (9, 45). Among them, inhibitor alisertib was assumed to be effective in uncoupling the direct interaction of the catalytic domain of AURKA with MYCN and displayed good activity against pediatric solid tumors in preclinical trials (46). Moreover, the inhibitory effect of AURKA on tumors is not absolutely dependent on the state of MYCN amplification, as evidenced by the fact that a few Myc-driven tumors are also sensitive to AURKA suppression (47, 48). Whereas in clinical trials including NB, inhibition of AURKA resulted in significant unintended toxicity and largely disappointing responses (36, 37). Encouragingly, good responses were achieved in phase I and II clinical trials of alisertib in combination with irinotecan and temozolomide for the treatment of relapsed or refractory neuroblastoma, conferring a potentially more favorable application of this inhibitor in combination with chemotherapeutic drug therapies (49, 50). As targets for MYCN-AM NB are still limited or unspecific, the discovery and development of new targets and therapeutic applications remains urgent. Here, our results and those of others adequately indicate that destabilization of MYCN is going to be an overwhelming breakthrough in attacking MYCN NA NB. we demonstrated that CCNB1IP1 supported MYCN-AM NB cells proliferation and tumor growth by repressing Fbxw7-mediated ubiquitination degradation of MYCN through gain- and loss-of-function experiments. Moreover, CCNB1IP1 antagonized the interaction of Fbxw7 with MYCN in a competitive binding manner, during which the C-terminal domain of CCNB1IP1 is required (Fig. 10).
Here, we identified CCNB1IP1 as a novel cofactor that stabilizes MYCN and acts synergistically with MYCN to enhance the proliferation and tumorigenicity of NB cells. Mechanistically, CCNB1IP1 blocked MYCN degradation through competitive binding of MYCN to Fbxw7, which further diversified the regulatory mechanism of MYCN proteostasis. Our data supported the potential of CCNB1IP1 as a MYCN-specific intervention target. However, there remain certain limitations and pending issues to be addressed in our study. Although phenotypic effects were observed, further investigation may be required to clarify how the synergistic effect of CCNB1IP1 with MYCN modulates the tumorigenicity of NB cells. Given the cell cycle-dependent expression and nucleus co-localization of both CCNB1IP1 and MYCN, it is necessary to further explore the complexity and spatiotemporal-dependent alterations of their interactions. Therefore, multifaceted mechanistic studies need to be integrated to more accurately assess the extent to which CCNB1IP1 contributes to the post-transcriptional regulation of MYCN expression. Although no definitive inhibitor of CCNB1IP1 has been developed, our study indicates that targeting CCNB1IP1 may be a new option for the future treatment of MYCN-AM NB. Screening and validation of small molecule compounds or agents with known clinical indications that specifically disrupt the interaction of CCNB1IP1 with MYCN as candidate inhibitors would be expected to be available for preclinical testing and subsequent clinical applications of MYCN AM NB. Although targeting CCNB1IP1 is still far from clinical application, it remains undeniable that NB cell growth is impaired by disrupting the CCNB1IP1-MYCN interaction. Ideally, more effective drug combination regimens could be developed based on this idea to provide more effective and specific treatments for improving prognosis and survival of high-risk NB patients.