Inactivation of Wild-Type p53 by Asparagine Endopeptidase in Glioblastoma: An Opportunity to Target the “Undruggable”

Mutation of the TP53 gene is a frequent event in human cancer with p53 being clearly implicated as functioning as a tumor suppressor (1,2). The p53 protein is a transcription factor that binds in a sequence-specific manner and activates transcription of target genes (1). Many proteins encoded by activated oncogenes have been shown to be highly effective targets for therapeutic intervention, often through the use of selective inhibitors. However, targeting loss of tumor suppressors such as p53 remains a formidable challenge (2,3). The compelling problem is how to restore tumor suppressor activity in cancers in which this has been lost either genetically or functionally.

Although certain tumor types have high rates of missense mutation of the TP53 gene, in other cancers wild-type p53 predominates. One such example is glioblastoma where the majority of tumors do not show evidence of genetic alteration of the TP53 locus (Figure 1) (5–9). Mouse models that recapitulate human glioblastoma frequently involve p53 loss, arguing the importance of inactivating p53 in glioblastoma (10,11). It is, of course, formally possible that p53 plays no role in these tumors. However, the more likely explanation is to posit the existence of other means to inactivate p53 tumor suppressor function. Indeed, this is likely the case in glioblastoma. There are two negative regulators of p53: Mdm2 and the related Mdm4 (Figure 1) (12). An analysis of two published studies involving 372 glioblastomas shows 10.8% have upregulated Mdm2 and 7.3% for Mdm4, largely because of gene amplification (Figure 1). This is consistent with the idea that these p53 regulators are providing oncogenic function (5–8). Although such a mechanism appears to be relevant in some glioblastomas, the same analysis has 56.7% that do not show evidence of TP53, MDM2, or MDM4 alterations (Figure 1). It is reasonable to infer that in such tumors, other mechanisms exist to inactivate p53 besides direct genetic alteration. This notion has been confirmed in this issue of the Journal by Lin et al. (13).

Figure 1.

Although a subset of human glioblastoma show mutually exclusive alterations within the p53 pathway, the majority of tumors are genetically unaltered for p53 and its negative regulators Mdm2 and Mdm4. Of 372 glioblastoma tumors, 25.3% show TP53 alterations with no MDM2 gene amplification, whereas 10.8% show MDM2 gene amplification with no p53 alterations. A similar result is seen for MDM4 (7.3%). The majority of these tumors (57%) do not show any genetic changes in the TP53, MDM2, or MDM4 genes. Six patients (1.6%), indicated by ***, show both MDM4 and TP53 or MDM2 alterations in the same tumor. Analysis was performed at cBioPortal (www.cbioportal.org) using the OncoPrinter tool (4) and a combination of two published glioblastoma data sets consisting of 281 tumors (5) and 91 tumors (6).

Lin et al. (13) provide compelling evidence that upregulation of the asparagine endopeptidase (AEP) results in cleavage of wild-type p53 at a specific residue, thereby generating a truncated p53 protein that has lost tumor suppressor function. Manipulating AEP expression influences the malignant activity of glioblastoma cells both in vitro and in vivo. These effects can be overcome by the use of a p53 missense mutant protein (N311A), which is resistant to AEP-induced cleavage. Intriguingly, the authors go on to demonstrate that tumor cells release extracellular vesicles that contain AEP, which in turn is endocytosed by normal human glial or endothelial cells. This is shown to lead to p53 cleavage and altered cellular properties that are consistent with a tumor-promoting microenvironment. Most importantly, higher levels of AEP protein associate with decreased survival in human glioblastoma patients. Indeed, p53 cleavage can be seen in primary human tumor samples as well. Taken together, these findings make a strong case that upregulation of AEP represents a novel mechanism for wild-type p53 inactivation in a subset of human glioblastoma.

The compelling study by Lin. et al. (13) raise several important questions. Although clear data are presented for upregulation of AEP protein in human glioblastoma samples, it remains unclear what the molecular basis is for this. The authors suggest that Epidermal Growth Factor Receptor signaling may play a role, but more definitive experimental approaches are needed. Likewise, it is shown that mutant p53 in certain tumor cell lines is less sensitive to knockdown of AEP expression. Tumor-associated p53 mutations typically occur within the sequence-specific DNA binding domain, which is considered to be a highly structured region (1). It is unclear why disrupting this region might influence cleavage at asparagine 313, which is located in a more unstructured linker region between the DNA binding domain and that for another globular domain involved in tetramerization. One previous publication has suggested that AEP (also called legumain) may be a transcriptional target for p53 (14). Although the current study (13) was not able to confirm this latter observation, it is possible that the lack of cleavage of mutant p53 may be related to a failure to increase levels of AEP expression by a transcriptionally dead mutant p53 rather that the intrinsic sensitivity of the mutant protein itself to cleavage.

An additional intriguing finding of Lin et al. concerns the subcellular localization for the interaction between AEP and p53 (13). Immunostaining showed that the proteins colocalize with endosomal and lysosomal markers. Although AEP has been studied extensively for its role in these compartments (15), localization of p53 to these sites is a novel finding. More careful analyses are needed to validate that p53 is indeed recruited to such vesicles and what the underlying molecular basis is for this recruitment. This is a central question given that AEP localization may be a key determinant for its role in oncogenesis (16).

To date, it has been difficult to envision therapeutic approaches that will restore activity in a tumor in which the TP53 gene has been deleted or p53 function has been compromised. The demonstration that upregulation of AEP is a means to inactivate p53 in glioblastoma provides the attractive possibility that AEP may be a suitable target for inhibition under the appropriate conditions. Indeed, chemical inhibitors of AEP have been under development (17). Lin et al. generated and used one such reagent in their study. Targeting the Mdm2-P53 or Mdm4-p53 interaction has already proven to have possible efficacy (2,12,18,19). This further provides support to the idea that AEP inhibitors may be a new addition to the arsenal of therapies that can be used in tumors with impairment of the p53 pathway. Appropriate and effective treatments for glioblastoma are an unmet need that may very well be addressed by further study of AEP in this clinically challenging disease.

Funding

This work was supported by the National Cancer Institute at the National Institutes of Health (R01CA200256 and R01 CA196234).

Notes

The author has no conflicts of interest to disclose. The funder had no role in the writing of this editorial or decision to submit it for publication.