TP53—a tumour suppressor with oncogenic potential?

With a prevalence of ~40%, TP53 encoding the stress-induced transcription factor p53 is the most frequently somatically altered gene in cancer [1]. TP53 is generally considered a tumour suppressor and ”guardian of the genome” due to its ability to induce transcriptional programmes leading to cell-cycle arrest and either DNA repair or apoptosis [2]. TP53’s tumour suppressive properties may best be illustrated by the fact that germline mutations of TP53 are associated with the cancer predisposition Li-Fraumeni syndrome that can give rise to a multitude of different malignancies, including osteosarcoma. However, discovered in 1979 in the context of virus-mediated malignant transformation [3], the first functional studies actually reported TP53 to be an immortalising oncogene [4]. Later studies suggested that the observed oncogenic effects of TP53 stemmed from the presence of non-synonymous point mutations [5]. This illustrates that the role of TP53 in cancer is more complex as well as context-dependent and that not all somatic TP53 alterations can be universally considered to inactivate a tumour suppressor. In particular, at least a subset of TP53 missense mutations have been associated with oncogenic functions in animal experiments, and this has led to a hypothesis that certain TP53 mutations might be considered ”separation-of-function” mutations altering the balance of pro- and antiproliferative effects (summarised in [6]).

TP53 promoter translocation as a separation-of-function paradigm in osteosarcoma

A recent genetic and transcriptomic study on 148 osteosarcoma patients published by Saba, Difilippo et al. [7] provides a novel paradigm in which TP53 rearrangements can simultaneously result in inactivation of p53 tumour suppressor functions and activation of oncogenic pathways by fusing the TP53 promoter region to new target genes (promoter swapping). In ~40% of analysed cases, evidence for TP53 promoter translocation was found, and in ~20% of cases, a putative fusion partner was readily identified. The authors further demonstrated that the resulting fusions were in-frame, and transcription of the TP53 fusion partner was increased. At the same time, expression of TP53 was lost, suggesting that promoter translocations of one TP53 allele co-occur with inactivating genetic aberrations of the other allele in osteosarcoma, and a selective advantage can be inferred. Functionally, the authors elegantly showed that DNA damage induced by cisplatin (a front-line drug to treat osteosarcoma) readily induced expression of the fusion partner in several different cell lines with different fusion partners.

Hence, these translocations do not only bring potential oncogenes under the control of the TP53 promoter, but they also disrupt the expression of a functional p53 protein. In essence, this results in both disruption of safeguarding TP53 responses upon, e.g., replication stress, reactive oxygen species and DNA damage and rewiring of the upstream stress response machinery to effector functions of a potentially oncogenic fusion partner (Fig. 1). In fact, a subset of the identified fusion partners had already been implicated in the pathobiology of osteosarcoma and other malignancies.

Fig. 1: Oncogenic effects TP53 promoter translocation.
figure 1

Upper panel: Cellular stress results in transactivation of the TP53 promoter, leading to transcription and eventually translation of TP53 protein whose transcriptional activity is responsible for physiological stress responses like cell cycle arrest, DNA repair and apoptosis. Lower panel: Translocation of the TP53 promoter brings an oncogene under the control of cellular stress, which can cause oncogenesis, treatment resistance and tumour survival.

Full size image

Of note, this separation-of-function paradigm (a term originally developed for specific TP53 missense mutations [6]) combines a canonical loss-of-function of the TP53 gene body with the overexpression of an oncogene through TP53 promoter hijacking as described in other tumours like lipoblastoma [8].

Osteosarcoma—a fusion-driven cancer?

Interestingly, TP53 hotspot mutations and TP53 promoter translocation were mutually exclusive even though the global gene expression patterns were similar. Furthermore, TP53 promoter translocations positively correlated with young age and a number of chromosomal rearrangements.

The high prevalence of these fusions in osteosarcoma, in particular in young patients, suggests a distinct role in the oncogenesis of osteosarcoma. This is of fundamental importance as osteosarcoma has hitherto not been considered a fusion-driven cancer, and fusions that arise by chance in the highly rearranged osteosarcoma genome have been reported to be out-of-frame and, thus, non-functional [9]. More experimental evidence, however, is required to answer the question of whether TP53 promoter translocations are indeed an early oncogenic event and whether this is causally linked to the subsequent accumulation of structural variants. Recently, single-cell analyses in a mouse model of pancreatic ductal adenocarcinoma showed that loss-of-function mutations in TP53 precede extensive copy number variations [10]. A similar approach is conceivable also for osteosarcoma research. Alternatively, spatiotemporal sampling and genetic analyses followed by clonal deconvolution could be considered in primary patient material as demonstrated for other diagnoses [11].

Conclusions

Identification of recurrent TP53 promoter translocations resulting in functional fusion genes is a major new paradigm for osteosarcoma in particular and for cancer research in general.

Future studies should be directed at exploring whether the subgroup of TP53-fusion-positive osteosarcoma comprises a distinct clinicobiological subset of osteosarcoma with respect to metastatic potential, therapeutic responses and immunological control.