Linking autophagy with inflammation through IRF1 signaling in ER+ breast cancer

Abstract

Resistance to antiestrogen therapy remains a critical determinant of mortality in patients affected by ER+ breast cancer. Our previous work identified autophagy and interferon regulatory factor 1 (IRF1) signaling as key regulators of this process. We have recently demonstrated a novel reciprocal interaction between IRF1 and ATG7, linking inflammation and autophagy.

Introduction

Approximately 1 in 8 American women will develop breast cancer during their lifetime.¹ Of these cases, 70% will have breast tumors that express estrogen receptor-α (ESR1; ERα) and are candidates for treatment with endocrine-targeted therapies such as the antiestrogen tamoxifen or the aromatase inhibitor letrozole. While often initially effective, the development of resistance limits the clinical efficacy of these agents. More women die from ER+ disease than from the other molecular subtypes of human epidermal growth factor receptor-2 (ERBB2; HER2) positive or triple negative (ERα, progesterone receptor, and HER2 negative) breast cancer. Thus, understanding the nature of endocrine resistance remains a central goal in breast cancer research.²

IRF1 signaling in breast cancer

It was previously reported that approximately 30% of neoplastic breast lesions have lost interferon regulatory factor-1 (IRF1) expression when compared with the surrounding normal breast tissue.³ Moreover, there is an increased loss of heterozygosity (approximately 50%) of IRF1 in breast cancer patients with mutations in breast cancer 1, early onset (BRCA1), demonstrating the importance of IRF1 as a tumor suppressor in breast cancer.³ IRF1 expression was shown to induce apoptosis and prevent the formation of mammary tumors.⁴ IRF1 has also been linked with endocrine therapy resistance.⁵ The decreased IRF1 expression in antiestrogen-resistant breast cancer cell lines implies a key role for IRF1 in mediating therapeutic responsiveness.

Autophagy as a driver of resistance

Autophagy is a cellular process of “self-eating” whereby damaged organelles or misfolded proteins are encapsulated by a double phospholipid membrane and then digested with lysosomal enzymes. The products of this digestion can feed into intermediate metabolism to support cell survival and proliferation.⁶ Antiestrogen-resistant breast cancer cells express elevated levels of lipid-conjugated microtubule-associated protein light chain 3 (MAP1LC3A; LC3-II), which is an essential component of the autophagosomal membrane.⁷ Inhibiting autophagy by using RNAi to knock down expression of autophagy related gene −5 (ATG5) or −7 (ATG7) resensitizes antiestrogen-resistant breast cancer cells to endocrine therapy and illustrates the importance of this pathway in mediating drug resistance.⁷ Moreover, using the chemical autophagy inhibitor chloroquine (CQ), a FDA-approved drug for the treatment of malaria, multiple myeloma, and rheumatoid arthritis, we recently showed that inhibiting autophagy in vivo restores tamoxifen sensitivity to resistant ER+ breast tumors.⁸ These data highlight the clinical relevance of targeting autophagy for the treatment of ER+ breast cancer.

Linking IRF1 and autophagy

One of the key pieces of information lacking in our understanding of resistance is how sensitive and resistant cells differentially regulate the balance between apoptosis and prosurvival autophagy. We explored this critical regulation in our recent publication in Cancer Research, “Interferon Regulatory Factor-1 signaling regulates the switch between autophagy and apoptosis to determine breast cancer cell fate.” In this study, we establish the novel reciprocal relationship between the putative breast cancer tumor suppressor protein IRF1 and 2 central regulators of autophagy, beclin-1 (BECN1) and autophagy protein ATG7.⁹ Using mice with a mutated ATG7 allele (ATG7^+/−) we showed elevated IRF1 protein in mammary tumors, kidneys, and spleens compared with wild-type mice. Moreover, we showed that human breast tumors that exhibit nuclear (active) IRF1 staining had reduced ATG7 staining, whereas tumors that were negative for nuclear IRF1 staining expressed high levels of ATG7. Thus, we established a reciprocal correlation between IRF1 and ATG7 that implied functional signaling interactions that could affect breast cancer cell fate decisions.⁹

Using molecular techniques and endocrine therapy-resistant human breast cancer cell lines, we studied the mechanistic relevance of these associations by showing that knockdown of IRF1 by RNAi increased ATG7 protein expression and autophagosome formation.⁹ Importantly, inhibition of ATG7 and BECN1 increased nuclear IRF1 (increased tumor suppressor signaling) while concurrently preventing nuclear ERα localization (thus inhibiting ERα-driven prosurvival signaling). The ability of nuclear IRF-1 to inhibit nuclear ERα localization may be a key component of the novel IRF1/ATG7 pathway that exists only in the ER+ breast cancer subtype; IRF1/ATG7 signaling was not observed in the triple negative MDA-MB-231 breast cancer cell line.⁹ Moreover, cytosolic accumulation of ER was previously shown to induce the endoplasmic reticulum stress pathway, the unfolded protein response (UPR) that might link IRF1/ERα/UPR signaling.¹⁰

A pathway schematic illustrating the IRF1/ERα/ATG7 signaling axis is shown in Figure 1. Interestingly, neither the chemical inhibitors HCQ and 3-methyladenine (3-MA) nor treatment with ATG5 siRNA resulted in elevated IRF1 levels in the ER+ breast cancer cell lines, suggesting that this relationship is directly mediated by ATG7 or BECN1 protein signaling and not by autophagic flux.⁹ We further show that signal transducer and activator of transcription (STAT1) inhibition and reactive oxygen species (ROS) blockade had no effect on ATG7 and BECN1 siRNA-mediated IRF1 induction, suggesting that the inverse relationship observed between IRF1 and ATG7/BECN1 is a STAT1-independent and ROS-independent process.⁹ The mechanism by which this is regulated is currently under investigation by our group.

Figure 1.

Reciprocal interactions between IRF1 and ATG7 signaling. Silencing of autophagy-related gene 7 (ATG7) or beclin-1 (BECN1) leads to inhibition of autophagosome formation and concurrent stimulation of interferon regulatory factor-1 (IRF1). Increased IRF1 inhibits B-cell lymphoma-2 (BCL2) signaling and results in the shuttling of estrogen receptor-α (ERα) out of the nucleus. Increased cytosolic ERα may result in stimulation of the unfolded protein response (UPR).

It is common practice to end such studies here, with the assumption that the cellular and molecular analyses have adequately addressed the central hypothesis and that no further signaling is likely to be relevant. However, we chose to go further and use mathematical modeling of the experimental data to explore the possibility that other activities might also be important in driving the effects of IRF1 on cell fate. The model was built using an ordinary differential equation formalism and calibrated to cell proliferation data using a least squares method. The resulting model supports the hypothesis that ATG7 knockdown can resensitize cells through both IRF1-dependent and -independent pathways. An experimental test of model predictions further supported this hypothesis and experiments are ongoing to identify the other pathway.⁹

Conclusions

Our data suggest that inducing nuclear IRF1 may be a critical component of inhibiting breast tumorigenesis and preventing endocrine therapy resistance in ER+ breast cancer. We are also the first to show the reciprocal relationship between IRF1 and ATG7, suggesting integration of inflammation and autophagy signaling in the regulation of endocrine therapy responsiveness. Nonetheless, the results of mathematical modeling and additional experimentation suggest that other signaling, integrated with that shown in our study, may also contribute to the cell fate decisions in response to endocrine therapies. In addition to showing the importance of the IRF1/ERα/ATG7 signaling axis in endocrine resistance, the study also exemplifies the utility of mathematical modeling in directing experimental design and interpreting results.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.