Is exclusive Skp2 targeting always beneficial in cancer therapy?

To the editor:

We read with great interest the work published in Blood by Chen et al concerning the therapeutic restoration of p27^KIP1 protein levels in cancer after applying the specific Skp2 inhibitor CpdA.¹  Skp2 is an E3-ubiquitin ligase that mediates degradation of several cell-cycle regulators.²  Among its targets are negative cell-cycle regulators, including the kinase inhibitor protein (KIP) family member p27^KIP1, and positive ones, such as cyclin E.²  In addition, p27^KIP1 also inhibits cyclin E.²  Thus, Skp2, p27^KIP1, and cyclin E form a tight network controlling S-phase entry.² 

In various tumors Skp2 is frequently overexpressed and represents a major cause of p27^KIP1 protein down-regulation.²  However, p27^KIP1 gene alterations and transcriptional silencing, due to microRNA-dependent repression, promoter methylation, and transcriptional suppressors, is not a rare event, as the authors mention,¹  but a significant source of p27^KIP1 inactivation in several malignancies (Table S1, available on the Blood website; see the Supplemental Materials link at the top of the article). Consequently, we cannot exclude the possibility that Skp2 overexpression coexists with transcriptionally silenced and/or mutant p27^KIP1. We have observed such a condition with the other KIP member and Skp2 target, p57^KIP2.³ 

Based on the above, Skp2 inhibition, in cases with transcriptionally silenced and/or mutant p27^KIP1, could result in increased expression of cyclin E with potential deleterious effects. Cyclin E provokes genomic instability, when overexpressed, by producing either DNA damage and/or centrosome amplification.^4,5  Up-regulation of cyclin E is frequently observed in cancer, and is associated in various malignancies with poor survival.⁶ 

To test this hypothesis we mimicked the above scenario by silencing Skp2 alone or in combination with p27^KIP1 in A549 cancer cells, which express high Skp2 levels.³  The experiment showed that sole Skp2 silencing resulted in elevation of p27^KIP1, reduction of cyclin E expression (Figure 1A,B), and a decrease in growth,³  while Skp2/p27^KIP1 double knockout, recapitulating the proposed scenario, led to increased levels of cyclin E (Figure 1A,B), centrosome amplification (Figure 1B), abnormal mitoses, and pronounced nuclear atypia characterized by micronuclei, lobulated nuclei, and nucleoplasmic bridges, features that are indicative of chromosomal instability (Figure 1C).⁷  In addition, marked p53 Ser-15 phosphorylation (Figure 1A), indicating a prominent DNA damage response, provides a mechanistic explanation for the observed genomic instability.^8,9  Furthermore, the elevated levels of cyclin E comprised not only the full-length but also the low-molecular-weight isoforms.^5,6  These isoforms are present in both cytoplasm and nucleus, and have increased affinity for cdk2.^5,6  They have been associated with genomic instability, resistance to CIP/KIP inhibition, and poor outcome of patients with various malignancies.⁶ 

Similarly, is exclusive Skp2 targeting efficient in patients with p27^KIP1 haploinsufficiency?¹⁰  In such cases, SKP2 blocking is effective when the levels of the remaining p27^KIP1 allele are up-regulated above a threshold, able to exert its negative effect on cell-cycle progression. Below this threshold the CIP/KIP molecules will be sequestered by the cyclin D/cdk complexes, further promoting cyclin E/cdk2 activity.¹¹ 

In conclusion, although the findings of Chen et al are significant and important, our results present an additional point of view, which stresses the impact of defining the transcriptional and/or mutational status of p27^KIP1 before applying a therapeutic approach based exclusively on Skp2 inhibition.