The role of PTEN, PI3K-Akt-mTOR signaling and pseudogene PTENP1 in breast cancer

Background: Breast cancer is the most common type of cancer in women, and a major cause of premature death still today. The PI3K-Akt-mTOR signaling (in short: PI3K signaling) pathway is a growth and pro-survival pathway, and is frequently deregulated in breast cancer due to amplifications, mutations and deletions of the genes encoding these important signaling hubs. PTEN inhibits the activation of Akt, and acts as an endogenous tumor suppressor by negatively regulating PI3K signaling. Loss of PTEN protein expression is frequently observed in breast cancer, which could contribute to hyperactivation of the PI3K pathway. Whereas aberrations in the PI3K pathway are associated with resistance towards endocrine therapy, less is known of the role of this signaling pathway in resistance to chemotherapy. Epigenetic alterations like DNA methylation, histone modifications and non-coding RNAs can also influence cancer progression and resistance to therapy. Non-coding RNAs are functional RNA transcripts, but are not translated into proteins. MicroRNAs (miRNAs) regulate protein expression post-transcriptionally by binding to complementary sequences on mRNA transcripts, which is then degraded. PTEN and the PTEN pseudogene PTENP1 share several miRNA binding sites, and PTENP1 acts as a competing endogenous RNA (ceRNA) by adsorbing miRNAs which would otherwise degrade various protein-coding mRNA transcripts. In prostate cancer it has been showed that PTENP1 expression protects PTEN mRNA from degradation and increases PTEN protein expression through ceRNA interaction. The role of the noncoding RNA PTENP1 in breast cancer has not been addressed previously. Main objectives: The first two papers presented in this thesis focused on unraveling the potential relationship between PI3K signaling and PTEN status, and mechanisms of resistance to chemotherapy in breast cancer. In the third study we examined the potential role of PTENP1 pseudogen in breast cancer on tumor growth. Results: In Paper 1, we explored the relationship between PI3K signaling and clinical outcome using data from three clinical studies. By examining breast cancer biopsies using IHC staining for PTEN, phosphorylated Akt (pAkt), pS6K and p4EPB expression we found that neither of the protein staining’s were associated with survival outcome, or predictive of response to chemotherapy. However, we found that patients with a high intratumoral PTEN mRNA expression had a worse prognosis than patients with a low PTEN mRNA expression. Interestingly, this was only significant for patients with TP53 wild type tumors. PTENP1 was expressed in most of the breast cancer biopsies examined and correlated to PTEN expression. However, there was no prognostic impact of PTENP1 gene expression, and no predictive value was observed, with respect to response to chemotherapy. In Paper 2, we created anthracycline-resistant breast cancer cell lines to compare how PI3K signaling was affected by an acute dose of anthracycline in anthracyclineresistant vs. naïve cells. After 24h anthracycline exposure we observed an increase in phosphorylated (activated) Akt in ER positive cell lines, while no change was seen in ER negative cell lines. Also, the resistant cell lines showed a higher baseline expression of phosphorylated Akt. Accordingly, we examined the effect of Akt inhibition with and without doxorubicin in concert both in vitro and in vivo. A combination treatment with Akt inhibitor and doxorubicin reduced tumor size more than Akt inhibitor or chemotherapy alone in the ER positive MCF7. In contrast, in the ER negative MDA-MB-231 Akt inhibition had no additive benefit to the anthracycline. To examine the relevance of Akt and anthracycline resistance in a clinical setting, we examined the mRNA level of AKT1 expression in breast cancer biopsies from two clinical studies. We found that an initial increase in AKT1 24 hours after the first treatment characterized ER positive tumors that subsequently responded to doxorubicin treatment, although this was not observed after 16 weeks of treatment. Thus, analysis of biopsies after 24 hours, to dissect the acute response to anthracyclines, seems to be of value to predict whether the patient will subsequently respond to treatment. The third paper addressed the question of whether the non-coding RNA PTENP1 has a functional role in breast cancer through ceRNA interactions. We found a diverging role for PTENP1 that was dependent on the tumor´s ER status. PTENP1 overexpression increased both the PTEN transcript and PTEN protein in ER negative cell lines, while ER positive cell lines had decreased PTEN transcript and unchanged PTEN expression subsequent to pseudogene transduction. By implanting PTENP1- overexpressing breast cancer cells in mice, PTENP1 reduced tumor growth in ER negative breast cancer, while it increased tumor growth in ER positive breast cancer. This effect seemed unrelated to PI3K signaling, but rather related to AP2γ and ERsignaling. PTENP1 overexpression increased expression of hsa-miR-26a and profoundly decreased the hsa-miR-26a target ESR1 and thereby ERα expression in the ER positive MCF7. There was also a clinical correlate to this finding, where data from the TCGA database demonstrated that breast tumors with upregulated PTENP1 expression exhibited lower ERα expression than breast tumors with normal or no PTENP1 expression. Conclusion and implications: We found a high PTEN gene expression to be associated with worse prognosis for patients with breast cancers harboring preserved p53 function. The finding that a high level of PTEN expression is a marker for bad prognosis in breast cancer is surprising, but indicates that PTEN transcripts may not only influence PTEN protein levels and PI3K signaling, but also interact with other biological mechanisms to promote tumor progression. The interaction between PTEN and TP53 seems essential in this context. Also, we demonstrate that ceRNA interactions are implicated in PTEN regulation, via its pseudogene PTENP1, and this interaction is fundamentally dependent on the tumor´s estrogen receptor status. Furthermore, in estrogen receptor positive breast cancers, upregulated PI3K signaling, and in particular Akt activation, seems to play a key role in resistance to anthracycline treatment, and potentially Akt inhibitors could be introduced in this setting to target chemoresistance