IL-1 induces p62/SQSTM1 and autophagy in ERα⁺/PR⁺ BCa cell lines concomitant with ERα and PR repression, conferring an ERα⁻/PR⁻ BCa-like phenotype

Abstract

Estrogen receptor alpha (ERα)^low/− tumors are associated with breast cancer (BCa) endocrine resistance, where ERα low tumors show a poor prognosis and a molecular profile similar to triple negative BCa tumors. Interleukin-1 (IL-1) downregulates ERα accumulation in BCa cell lines, yet the cells can remain viable. In kind, IL-1 and ERα show inverse accumulation in BCa patient tumors and IL-1 is implicated in BCa progression. IL-1 represses the androgen receptor (AR) hormone receptor in prostate cancer (PCa) cells concomitant with the upregulation of the prosurvival, autophagy-related protein, Sequestome-1 (p62/SQSTM1; hereinafter, p62); and given their similar etiology, we hypothesized that IL-1 also upregulates p62 in BCa cells concomitant with hormone receptor repression. To test our hypothesis, BCa cell lines were exposed to conditioned medium from IL-1-secreting bone marrow stromal cells (BMSCs), IL-1, or IL-1 receptor antagonist (IL-1RA). Cells were analyzed for the accumulation of ERα, progesterone receptor (PR), p62, or the autophagosome membrane protein, Microtubule-Associated Protein 1 Light Chain 3 (LC3), and for p62-LC3 interaction. We found that IL-1 is sufficient to mediate BMSC-induced ERα and PR repression, p62 and autophagy upregulation, and p62-LC3 interaction in ERα⁺/PR⁺ BCa cell lines. However, IL-1 does not significantly elevate the high basal p62 accumulation or high basal autophagy in the ERα⁻/PR⁻ BCa cell lines. Thus, our observations imply that IL-1 confers a prosurvival ERα⁻/PR⁻ molecular phenotype in ERα⁺/PR⁺ BCa cells that may be dependent on p62 function and autophagy and may underlie endocrine resistance.

1. INTRODUCTION

Estrogen receptor alpha (ERα) and progesterone receptor (PR) are nuclear hormone receptors that drive the proliferation and survival of breast cancer (BCa) cells, where PR is also an ERα target gene^1,2. Around 70% of BCa patients have tumors positive for ERα and PR (ERα⁺/PR⁺) at diagnosis; therefore ERα is a therapeutic target for BCa¹. ERα-targeting endocrine therapies include estrogen depravation (ovarian ablation or aromatase inhibitors), selective estrogen receptor modulators (SERMs; e.g., tamoxifen) and selective estrogen receptor downregulators (SERDs; e.g., fulvestrant) that block ERα activity and/or reduce ERα protein accumulation³. Unfortunately, reportedly 40–50% of BCa patients will eventually acquire resistance to ERα-targeting therapy¹. Studies suggest that endocrine therapy resistance is due to aberrant ERα activity as a result of ERα mutations, posttranslational modifications, altered chromatin binding, or estrogen-independent growth factor signaling¹. ERα reduction or loss is also a mechanism of endocrine resistance^1,4. ERα^low/- tumors are observed in 10–30% of endocrine resistant BCa patients^1,4 and ERα ^low/- tumors are predictive of endocrine resistance and poor prognosis as compared to BCa patients with higher ERα tumor levels^5–9.

Tumor inflammation positively correlates with ERα reduction or loss in BCa patient tumors and multiple different inflammatory cytokines have been shown to reduce ERα levels, thus, implicating inflammation in BCa endocrine resistance⁴. For example, the interleukin-1 (IL-1) inflammatory cytokine family has been shown to repress ERα levels^10–12 and IL-1 inversely correlates with ERα in BCa patient tumors^13–15, making IL-1 a likely culprit contributing to BCa endocrine resistance.

The major IL-1 family members are IL-1 alpha (IL-1α) and IL-1 beta (IL-1β), both of which bind to the cell surface IL-1 Receptor 1 (IL-1R1)¹⁶. IL-1 is elevated in BCa tumor tissue and patient serum and correlates with relapse and poor prognosis^17–20. Functionally, IL-1 signaling induces expression of pro-metastatic genes, angiogenic proteins, and growth factors that promote tumor progression²¹. Particularly in BCa, IL-1 can enhance progression through the induction of metastasis and cachexia²⁰. The IL-1 receptor antagonist (IL-1RA) blocks IL-1 signaling¹⁶ and IL-1RA has been shown to reduce breast cancer xenograft tumor volume and bone metastasis in vivo²². As such, IL-1RA is being investigated in clinical trials (ClinicalTrials.gov) as adjuvant therapy for multiple cancer types including BCa, to, in part, block IL-1 tumorigenic functions such as angiogenesis.

We have previously shown that IL-1β downregulates androgen receptor (AR) mRNA and protein in prostate cancer (PCa) cell lines²³. AR is a nuclear hormone receptor that promotes PCa growth and proliferation and, thus, is a PCa therapeutic target²⁴. However, as with ERα^low/- BCa tumors, PCa tumors enriched in AR ^low/- cells are associated with hormone therapy resistance^24,25. Furthermore, as with BCa, IL-1 also promotes PCa metastasis to bone²⁶ and IL-1 is implicated in PCa hormone therapy resistance²⁷. Therefore, given the etiological similarities between BCa and PCa, we hypothesize that the IL-1 mechanisms that promote resistance to hormone receptor-targeting therapies are similar in BCa and PCa.

We found that IL-1 signaling downregulates AR in PCa cell lines concomitant with the induction of several prosurvival, pro-growth pathways that may contribute to treatment resistance (Thomas-Jardin et al., in press), such as p62 signaling²⁸. p62 is a multi-domain scaffold protein that has been shown to be cytoprotective for multiple different cancer types^29–31 and p62 mediates various prosurvival signaling cascades, including autophagy³². Autophagy-dependent p62 function involves p62 sequestration of cytoplasmic proteins and organelles into the autophagosome for degradation and biomolecule recycling³². Cancer cells utilize autophagy to survive cellular stresses such as low nutrient availability, genotoxic stress, chemotherapy, and radiation³³. Not surprisingly, p62 overexpression^34–38 or elevated autophagy^39,40 in BCa patient tumors is associated with poor prognosis. Therefore, we investigated the effect of IL-1 on p62 and autophagy concomitant with ERα and PR loss in BCa cell lines.

Here, we characterize ERα⁺/PR⁺ BCa cell lines, MCF7 and T47D, and ERα^-/PR^- BCa cell lines, BT549 and MDA-MB-231. We find that ERα^-/PR^- BCa cell lines have high basal p62 accumulation and autophagy flux relative to ERα⁺/PR⁺ BCa cell lines. We show that IL-1 represses ERα and PR in hormone receptor positive BCa cell lines concomitant with p62 upregulation, autophagy induction, and p62 interaction with the autophagosome membrane protein, LC3. Thus IL-1-induced p62-autophagy signaling might be an underlying mechanism promoting BCa tumor cell survival when hormone receptor accumulation is repressed.

2. MATERIALS AND METHODS

2.1. Cell Culture:

Breast cancer cell lines (MCF7, T47D, MDA-MB-231), prostate cancer cell lines (LNCaP, C4–2, PC3, DU145), and bone marrow stromal cell lines (HS-5, HS-27a) were grown in a 37°C, 5.0% (v/v) CO₂ incubator in DMEM medium (Gibco; 1185–076) supplemented with 10% FB Essence (Seradigm; 3100–500), 0.4 mM L-glutamine (L-glut; Gibco/Invitrogen, 25030081), and 10 U/ml penicillin G sodium and 10 mg/ml streptomycin sulfate (pen-strep; Gibco/Invitrogen, 15140122). BT549 breast cancer cell line was grown in RPMI-1640 media (Hyclone; SH30027.01) supplemented with 10% FB essence, L-glut, Pen-strep.

2.2. Conditioned Media (CM) Treatment:

Culture medium (CM) was removed from BCa cell lines, HS-5 or HS-27a cells and replaced with fresh DMEM or RPMI supplemented with 10% FB Essence. After 3 days incubation, the conditioned DMEM or RPMI media was collected, filtered and stored at −80°C. For CM treatment experiments, cells were grown in BCa CM (control CM), HS-5 CM, or HS-27a CM for 3 days.

2.3. Cytokines and Drugs:

Human recombinant IL-1α, IL-1β and IL-1RA were purchased from R&D systems (200-LA/CF; 201-LB-005; 280-RA/CF) and resuspended in 0.1% bovine serum albumin (BSA) in 1X phosphate buffered saline (PBS). Chloroquine (CQ) diphosphate aqueous solution was purchased from Invitrogen (P-36239). Cells were treated with 25ng/mL IL-1 or vehicle control (0.1% BSA in 1X PBS) added to DMEM or RPMI growth media for 3 days. For the IL-1RA experiments, cells grown in DMEM were pre-treated with 200 ng/mL IL-1RA in for 1 day and then the medium replaced with fresh DMEM or HS-5 CM with or without IL-1RA for an additional 3 days.

2.4. Western Blot and Antibodies:

Protein was isolated from cells using NP40 lysis buffer (0.5% NP-40 (US biological; N3500), 50mM Tris (pH 7.5), 150mM NaCl, 3mM MgCl2, 1× protease inhibitors (Roche; 05892953001). Protein concentration was measured using the Pierce BCA Protein Assay Kit (Thermo Scientific; 23227). For western blot analysis, equal protein concentrations were loaded onto and separated in 12% or 16% (w/v) sodium dodecyl sulfate polyacrylamide gel (40% acrylamide/bis-acrylamide solution; Bio-Rad; 161–0148). Proteins were transferred from the gel to 0.45 μm pore size nitrocellulose membrane (Maine manufacturing; 1215471) and total protein visualized using Ponceau S (Amresco; K793). The membrane was blocked with 2.5% (w/v) BSA (Fisher; BP 1600–1) in 1× TBST (20 mM Tris, pH 7.6, 150 mM NaCl, 0.05% Tween-20). Primary and secondary antibodies were diluted in 2.5% BSA in 1× TBST. Protein blot bands were visualized using Clarity Western ECL Substrate (BioRad; 1705061) and imaged using Amersham Imager 600 (GE). Primary antibodies: ERα (Cell signaling technology; 8644), LC3B (Novus Biologicals; NB600–1384), p62/SQSTM1 (Santa Cruz Biotechnology; sc-28359), SOD2 (Abgent; AP7579a) and β-actin (Abcam; ab8226). Secondary antibodies: sheep anti-mouse (Jackson ImmunoResearch Laboratories; 515–035-062), goat anti-rabbit (Sigma-Aldrich; A6154). Western blot densitometry was performed using Image J and graphed, where protein bands were first normalized to β-actin and then relative protein levels normalized to the lowest protein/β-actin ratio for a given protein.

2.5. RNA Extraction and Reverse Transcription-Quantitative Polymerase Chain Reaction (RT-QPCR):

RNA was extracted using GeneJET RNA Purification kit (Thermo-Fisher; K0732) as per manufacturer’s instructions. cDNA was synthesized using iScript Reverse Transcription Supermix (Biorad; 170–8841). RT-QPCR reactions were performed using the iTaq Universal SYBR Green Supermix (Biorad; 172–5125) as per manufacturer’s instructions using BioRad CFX Connect. ERα, PR, p62 or SOD2 cycle times were normalized to the β-actin. Relative mRNA levels were calculated using 2^−ΔΔCT method. 5’−3’ Primer sequences: ERα Forward, ACAAGGGAAGTATGGCTATGGA; ERα Reverse, GGTCTTTTCGTATCCCACCTTTC; PR Forward, ACATGGTAGCTGTGGGAAGG; PR Reverse, GCTAAGCCAGCAAGAAATGG; p62 Forward, AAATGGGTCCACCAGGAAACTGGA; p62 Reverse, TCAACTTCAATGCCCAGAGGGCTA; SOD2 Forward, GGCCTACGTGAACAACCTGA; SOD2 Reverse, GTTCTCCACCACCGTTAGGG; β-Actin Forward, GATGAGATTGGCATGGCTTT; β-Actin Reverse, CACCTTCACCGGTCCAGTTT

2.6. Immunoprecipitation:

Cells were lysed in NP40 lysis buffer and equal protein concentrations incubated with 2ug of LC3B antibody (Novus; NB100–2220) or rabbit IgG (Santa Cruz Biotechnology; sc-2027). Immunoprecipitation (IP) was performed using Pierce Magnetic co-IP kit (Thermo Scientific; 88804) per manufacturer’s instructions.

2.7. Immunofluorescence:

Cells were fixed and permeabilized with 100% methanol at −20°C for 30 minutes. Fixed cells were blocked with 2.5% BSA in 1× PBS at room temperature for at least 30 minutes. Antibodies were diluted in 2.5% BSA in 1× PBS. Cells were incubated in primary antibody overnight at 4°C, washed with 1× PBS, and then incubated with fluorescently labeled secondary antibody overnight at 4°C in the dark. Primary antibodies: LC3B (Novus Biologicals; NB600–1384), p62/SQSTM1 (Santa Cruz Biotechnology; sc-28359). Fluorescently labeled secondary antibodies: Alexafluor 488, goat anti-mouse (Invitrogen; A11001), Alexafluor 568, goat anti-rabbit (Invitrogen; A11061). Immunostained cells were imaged at 10x, 20x, and 100x magnification using a Nikon epifluorescence microscope. Images were analyzed using Nikon NIS Elements. Cell counts were performed for five representative microscopy fields at 10x magnification for each treatment. The cell count average and standard deviation of 3 biological replicates per treatment was calculated.

2.8. Statistical analysis:

Statistical significance between two groups was tested using Student’s t test. Error bars represent standard deviation (STDEV) of at least n = 3 biological replicates (Bio Rep) and asterisks denotes statistical significance (*, p<0.05; **, p<0.005; ***, p<0.0005).

3. RESULTS

3.1. ERα^-/PR^- BCa cell lines have high basal p62 accumulation and autophagy flux

We previously discovered that AR^- PCa cell lines have high basal p62 and/or LC3-II relative to AR⁺ PCa cell lines²⁹. LC3-II is the membrane-conjugated form of Microtubule Associated Protein 1 Light Chain 3 (LC3) that comprises the autophagosome double membrane³². When autophagy is induced, free LC3 (LC3-I) is conjugated to phosphatidylethanolamine lipid to form LC3-II³². p62 binds LC3-II to sequester proteins and organelles into the autophagosome for subsequent degradation and recycling to maintain cellular homeostasis³². To determine if p62 and LC3-II accumulation might be stratified by hormone receptor status in BCa cells we compared p62 and LC3 mRNA and/or protein in ERα^-/PR^- versus ERα⁺/PR⁺ cell lines. RT-QPCR and western blot reveal that relative to the ERα⁺/PR⁺ BCa cell lines, MCF7 and T47D, the ERα^-/PR^- BCa cell lines, BT549 and MDA-MB-231, have high basal p62 mRNA (Fig. 1A) and protein (Fig. 1B) and high basal LC3-II protein (Fig. 1B). Thus, as we previously discovered for AR^- versus AR⁺ PCa cell lines²⁹, p62 and LC3-II are also basally high in ERα^-/PR^- BCa cell lines relative to ERα⁺/PR⁺ BCa cell lines.

Figure 1. Basal ERα, PR, p62 and LC3 levels in ERα^-/PR^- versus ERα⁺/PR⁺ BCa cell lines.

(A) RT-QPCR and (B) western blot analysis were performed for the ERα^-/PR^- BCa cell lines, BT549 and MDA-MB-231, and the ERα⁺/PR⁺ BCa cell lines, MCF7 and T47D, grown in BCa CM for 3 days. BT549 and MDA-MB-231 cells have little or no detectable ERα and PR mRNA (A) or protein (B) relative to MCF7 and T47D cells. BT549 and MDA-MB-231 cells have higher basal p62 or LC3-II mRNA (A) and/or protein (B) compared to MCF7 and T47D cells. Error bars = ± STDEV of 3 biological replicates; mRNA levels were normalized to T47D for ERα and p62 QPCR and normalized to MCF7 for PR QPCR. For western blot densitometry, MCF7 was set at 1. β-actin is the western blot loading control.

p62 binds to inner membrane LC3-II to sequester cargo into the autophagosome³², subjecting both p62 and LC3-II to autophagy-mediated degradation. Therefore, while increased LC3-II levels can indicate active autophagosome formation, elevated p62 and LC3-II protein levels could also be due to a block in autophagosome fusion with the lysosome or a block in autolysosomal degradation⁴¹. To determine if the high basal p62 and LC3-II in the ERα^-/PR^- BCa cell lines is due to a block in autophagy, we treated BT549 and MDA-MB-231 BCa cell lines with the lysotrophic drug, chloroquine (CQ). When autophagosomes fuse with lysosomes for form autolysosomes, the lysosomal hydrolases degrade the vesicle cargo. CQ increases lysosomal pH, thereby preventing degradation of autolysosomal cargo, including p62 and inner membrane LC3 (LC3-II), contained in autolysosomes. If autophagy flux is functional, then CQ will prevent autolysosomal turnover, causing a cellular accumulation of autophagosomes, autolysosomes and their cargo (e.g. p62 and LC3-II). Thus, cellular sensitivity to CQ indicates that autophagosome formation, lysosomal fusion, and autolysosomal degradation are functional⁴¹. Indeed, CQ increases p62 and LC3-II protein accumulation in the BT549 and MDA-MB-231 ERα^-/PR^- BCa cell lines (Fig. 4C), demonstrating that autophagy flux is functional in these BCa cell lines, where the high basal LC3-II (Fig. 1B) likely reflects enhanced autophagy flux and the high basal p62 is likely due to constitutive p62 gene expression (Fig. 1A). Taken together, we observe similar p62 and LC3-II expression and accumulation patterns in hormone receptor negative versus positive BCa cell lines, as we previously observed for hormone receptor negative versus positive PCa cell lines²⁹. Thus, BCa and PCa may have evolved a similar regulation and functional requirement for p62 and autophagy that is stratified by hormone receptor status.

Figure 4. IL-1α or IL-1β modulate ER, PR, p62 and LC3 in ERα⁺/PR⁺ BCa cells.

ERα⁺/PR⁺ BCa cell lines, MCF7 and T47D, and ERα^-/PR^- BCa cell lines, BT549 and MDA-MB-231, were treated for 3 days with vehicle control (0.1% BSA in PBS), IL-1α (25ng/mL) or IL-1β (25ng/mL) in the absence (A, B) or presence (C) of 20μM chloroquine (CQ). Cells were analyzed for mRNA by RT-QPCR (A) or protein by western blot (B, C). IL-1α or IL-1β downregulate ERα and PR mRNA (A) and protein (B) and induce p62 and LC3 mRNA (A) and/or protein (B) and in MCF7 and T47D cells, but does not modulate p62 or LC3 levels in BT549 or MDA-MB-231 cells. IL-1α- or IL-1β-induced SOD2 mRNA and protein accumulation serve as positive controls for treatment efficacy. (C) CQ blocks autophagy-mediated degradation of p62 and LC3-II in IL-1α- or IL-1β-treated cells. Error bars = ± STDEV of 4 biological replicates (Bio Rep); p-value = * ≤ 0.05, ** ≤ 0.005, *** ≤ 0.0005. mRNA fold change is normalized within each cell line to its control and protein fold change is normalized to the lowest protein/β-actin ratio for a given protein. β-actin is the western blot loading control.

3.2. Bone marrow stromal cell paracrine factors upregulate p62 and autophagy in ERα⁺/PR⁺ BCa cell lines

Bone-derived immune cells infiltrate tumors and secrete inflammatory cytokines that promote tumor progression both in the primary tumor and metastatic niche.⁴² Furthermore, BCa and PCa preferentially metastasize to bone. Thus, in investigating the paracrine effects of the HS-5 and HS-27a bone marrow stromal cell lines (BMSC) on PCa tumor cell biology, we discovered that HS-5 BMSCs, but not HS-27a BMSCs, secrete factors that induce p62 and autophagy in AR⁺ PCa cell lines^28,43,44. HS-5 BMSCs are secretory, while HS-27a BMSCs represent a structural component of the bone, where the HS-5-secreted milieu includes IL-1α and IL-1β⁴⁵. To determine if ERα⁺/PR⁺ BCa cell lines respond similarly to HS-5 BMSC paracrine factors, we treated MCF7 and T47D BCa cell lines with condition medium (CM) from the HS-5 or HS-27a BMSC cell lines. RT-QPCR and western blot revealed that HS-5 CM, but not HS-27a CM, induces p62 mRNA (Fig. 2A) and protein (Fig. 2B) and upregulates LC3-I and/or LC3-II accumulation (Fig. 2B) in MCF7 and T47D cell lines. Furthermore, p62 and LC3-II levels increase in the presence of CQ (Fig. 2C), indicating that autophagy is functional in the presence of HS-5 CM. Thus, HS-5 BMSCs secrete paracrine factors that have a similar effect on p62 and autophagy regulation in both hormone receptor positive BCa and PCa cell lines, suggesting p62 and autophagy are regulated through conserved mechanisms in both cancer types.

Figure 2. HS-5 BMSC CM modulates ERα, PR, p62 and LC3 levels in ERα⁺/PR⁺ BCa cell lines.

MCF7 and T47D BCa cell lines were grown in BCa CM (control CM), HS-5 CM, or HS27-a CM for 3 days in the absence (A, B) or presence (C) of 20μM chloroquine (CQ). Cells were analyzed for mRNA by QPCR (A) or protein by western blot (B, C). HS-5 CM downregulates ERα and PR mRNA (A) and protein (B) and induces p62 and LC3 mRNA (A) and/or protein (B) and in MCF7 and T47D cells. HS-5 CM-induced SOD2 mRNA and protein accumulation serves as a positive control for treatment efficacy. (C) CQ blocks autophagy-mediated degradation of p62 and LC3-II in control CM, HS-5 CM and HS-27a CM treated cells. Error bars = ± STDEV of 4 biological replicates (Bio Rep); p-value = * ≤ 0.05, ** ≤ 0.005, *** ≤ 0.0005. mRNA fold change is normalized within each cell line to its control and protein fold change is normalized to the lowest protein/β-actin ratio for a given protein. β-actin is the western blot loading control.

3.3. HS-5 BMSC paracrine factors repress hormone receptors in hormone receptor positive BCa cell lines

Given that high basal p62 and autophagy flux are associated with ERα^-/PR^- BCa cell lines (Fig. 1A), we predicted HS-5 BMSCs might induce ERα and PR repression concomitant with p62 and autophagy induction in ERα⁺/PR⁺ BCa cell lines. Our rationale is based on our previous observations that HS-5 CM represses AR concomitant with p62 and autophagy induction in AR⁺ PCa cell lines^23,44. Treatment with HS-5 CM represses ERα and PR mRNA and protein (Fig. 2A & B). HS-5 CM can also repress AR mRNA and protein in BCa cell lines (data not shown). Interestingly, we could not detect HS-5 CM-mediated repression of mRNA levels or protein accumulation for ER beta (ERβ) or glucocorticoid hormone receptors in BCa cell lines (data not shown). Thus, HS-5 BMSC paracrine regulation of AR, ERα, and PR hormone receptors may be through a conserved mechanism.

3.4. IL-1 signaling mediates HS5 BMSC paracrine regulation of p62, autophagy, and hormone receptor accumulation in ERα⁺/PR⁺ BCa cell lines.

We found that IL-1 alone²³ or in the HS-5 CM milieu⁴⁶ is sufficient to repress AR and induce p62 mRNA and protein accumulation in PCa cell lines. To determine if IL-1 signaling is sufficient to mediate HS-5 BMSC paracrine modulation of ERα, PR, p62, and/or autophagy in ERα⁺/PR⁺ BCa cell lines, we blocked IL-1 signaling in MCF7 and T47D cells with the IL-1 receptor antagonist (IL-1Ra). Our IL-1Ra treatment does not restore ERα mRNA levels (Fig. 3A), but does attenuate HS-5-induced repression of ERα protein accumulation (Fig. 3B). Thus, other factors in HS-5 CM may regulate ERα mRNA levels and IL-1 may also regulate ERα protein stability. IL-1Ra does, however, restore both PR mRNA and protein in HS-5 CM-treated cells (Fig. 3A & B), which may be due to IL-1Ra-restored ERα protein that can subsequently transactivate PR expression. Finally, while IL-1Ra attenuates HS-5-induced LC3-II accumulation in both MCF7 and T47D cells (Fig. 3B), only T47D cells show IL-1Ra attenuation of HS-5-induced p62 mRNA expression and protein accumulation. Hence, MCF7 and T47D have differential sensitivity to the HS-5 CM milieu and additional HS-5 paracrine factors likely regulate p62 levels. Taken together, our data suggests that IL-1 signaling is sufficient, but not strictly necessary, to mediate HS-5 BMSC paracrine repression of hormone receptors and induction of p62 and autophagy.

Figure 3. IL-1Ra attenuates HS-5 BMSC modulation of ERα, PR, p62 and LC3 in ERα⁺/PR⁺ BCa cells.

Hormone receptor positive BCa cell lines were pretreated with 200 ng/mL Interleukin-1 receptor antagonist (IL-1Ra) for 1 day followed by treatment with HS-5 CM ± 200 ng/mL IL-1RA for an additional 3 days. DMEM (control CM) served as control for HS-5 CM. Treated cells were analyzed for mRNA (A) or protein accumulation (B) by QPCR or western blot, respectively. IL-1RA can attenuate HS-5 CM modulation of ERα, PR, p62 or LC3 mRNA (A) and/or protein (B) in MCF7 and T47D cells. The attenuation of HS-5 CM-induced SOD2 protein is a control for IL-1Ra efficacy. Error bars = ± STDEV of 4 biological replicates (Bio Rep); p-value = * ≤ 0.05, ** ≤ 0.005, *** ≤ 0.0005. NS = not statistically significant. mRNA fold change is normalized within each cell line to its control and protein fold change is normalized to the lowest protein/β-actin ratio for a given protein. β-actin is western blot loading control.

3.5. IL-1α and IL-1β are sufficient to repress nuclear receptors and induce p62 and autophagy in nuclear hormone receptor positive BCa cell lines

IL-1α and IL-1β are the major IL-1 family members and signal through the IL-1R1 receptor¹⁶. IL-1α and IL-1β show similar biological activity, but do have some distinct functions¹⁶. Therefore, to determine if each cytokine can individually repress hormone receptors or induce p62 or autophagy in BCa cells, we treated MCF7 and T47D BCa cell lines with IL-1α and IL-1β. Both IL-1α and IL-1β can repress ERα and PR mRNA and protein, induce p62 mRNA and protein, and induce LC3-I and/or LC3-II accumulation in MCF7 and T47D cells, while ERα^-/PR^- BT549 and MDA-MB-231 BCa cells lines show only a modest or no response to IL-1α or IL-1β (Fig. 4A & B). To determine if BT549 and MDA-MB-231 cells can respond to IL-1 and to ensure IL-1 treatment efficacy, we assayed the induction of the IL-1 target gene, mitochondrial Superoxide Dismutase 2 (SOD2). IL-1α and IL-1β induce SOD2 mRNA and protein in both hormone receptor negative and positive BCa cell lines (Fig. 4A & B), indicating that ERα^-/PR^- BCa cells are indeed responsive to IL-1 signaling, while IL-1-mediated regulation of p62 and autophagy is specific to ERα⁺/PR⁺ BCa cells. It is also possible that the high basal levels of p62 and LC3 in ERα^-/PR^- BCa cell lines masks any IL-1 effect on p62 induction or autophagy flux. Finally, p62 and LC3-II levels increase in the presence of CQ, indicating that autophagy is functional in the presence of IL-1 in ERα⁺/PR⁺ MCF7 and T47D cell lines and ERα^-/PR^- BT549 and MDA-MB-231 cell lines (Fig. 4C, Fig. 6C, data not shown). Taken together, IL-1α and IL-1β are sufficient to regulate hormone receptor accumulation, p62 accumulation, and autophagy induction in ERα⁺/PR⁺ BCa cells.

Figure 6. p62 and LC3 co-localization in control and IL-1 treated MCF7 cells.

Vehicle control or IL-1 (25 ng/mL) treated MCF7 cells grown in the absence or presence of CQ (20 μM) for 3 days were co-immunostained for p62 (FITC; green) and LC3 (Texas Red, red). Nuclei are stained with DAPI (blue). (A) p62 and LC3 puncta co-localize under each growth condition. (B, C) Images show merged p62 (FITC), LC3 (Texas Red), and DAPI fluorescence. IL-1 increases p62-LC3 co-localization in the total cell population (images and graph); and in the presence of CQ, 100% of the cell population accumulates co-localized p62 and LC3, indicating that autophagy flux is functional under both basal and IL-1 treatment conditions. Cells were imaged at 100x (A; scale bar = 10 μM) and 20x (C; scale bar = 100 μM) magnification. Error bars = ± STDEV of 3 biological replicates; 12,000 < n < 22,000 total cells counted; p-value = * ≤ 0.05, ** ≤ 0.005, *** ≤ 0.0005.

3.6. p62 and LC3 interact in hormone receptor negative BCa cell lines and in response to IL-1 in hormone receptor positive BCa cell lines

p62 interacts with LC3-II to sequester protein and organelles into the autophagosome for degradation³². Given that hormone receptor negative BCa cell lines have high basal p62 and LC3 (Fig. 1) and IL-1 induces p62 and LC3-II accumulation in hormone receptor positive BCa cell lines (Fig. 4), we performed immunoprecipitation to determine if p62 and LC3 are interacting in these contexts. p62 and LC3 form a complex in ERα^-/PR^- BT459 and MDA-MB-231 BCa cell lines under basal growth conditions (Fig. 5A), while IL-1 promotes p62-LC3 interaction in HS-5 CM-treated (Fig. 5B) and IL-1-treated (Fig. 5C) ERα⁺/PR⁺ MCF7 and T47D BCa cell lines.

Figure 5. p62-LC3 interaction in hormone receptor negative or positive BCa cell lines.

LC3 was immunoprecipitated from untreated ERα^-/PR^- BCa cell lines, BT549 and MDA-MB-231, AR^- PCa cell lines, PC3 and DU145 (A), treated ERα⁺/PR⁺ BCa cell lines, MCF7 and T47D (B, C), or treated AR⁺ PCa cell lines, LNCaP and C4–2 (C). Western blot analysis was performed on the immunoprecipitation elution to detect LC3 and p62. (A) BT549, MDA-MB-231, and PC3 cell lines show LC3-p62 interaction under basal growth conditions. LC3-p62 interaction is not detected in DU145 cells, which do not accumulate LC3-II. HS-5 CM (B) or IL-1α or IL-1β (C) induce LC3-p62 interaction in MCF7, T47D LNCaP, and C4–2 cells. (B) IL-1RA attenuates HS-5 CM induced LC3-p62 interaction in MCF7 and T47D cells. IgG serves as control for LC3 antibody. * = Non-specific band.

We also co-immunostained IL-1-treated cells for p62 and LC3, and in support of our immunoprecipitation results, we see that p62 and LC3 co-localize as distinct puncta in MCF7 cells (Fig. 6A) and p62-LC3 colocalization is enriched in IL-1-treated MCF7 cell populations (Fig. 6B & C). Unlike for MCF7 cells, we did not detect p62-LC3 puncta fluorescence in T47D cells above background signal, likely because T47D cells accumulate less total LC3 than MCF7 cells (Fig. 1B) and MCF7 cells show are more robust response to IL-1 for p62 and LC3 induction (Fig. 4A & B). Nevertheless, MCF7 co-immunostaining suggests that IL-1 promotes p62-LC3 interaction.

Given the hormone receptor-stratified similarities between BCa and PCa cell lines, we also investigated p62-LC3 interaction in under basal growth conditions in the AR^- PCa cell lines, PC3 and DU145, and in response to IL-1β treatment in AR⁺ PCa cell lines, LNCaP and C4–2. As observed for ERα^-/PR^- BT459 and MDA-MB-231 BCa cell lines, p62 and LC3 interact under basal growth conditions in AR^- PC3 cells (Fig. 5A). DU145 cells do not perform canonical autophagy and, therefore, do not form LC3-II⁴⁷; therefore, as expected, we do not detect an interaction between p62 and LC3 in DU145 cells (Fig. 5A). As observed for ERα⁺/PR⁺ MCF7 and T47D BCa cell lines, p62 and LC3 interact in response to IL-1β in AR⁺ LNCaP and C4–2 cells (Fig. 5C). Of note, because our IL-1 treatment does not appear to induce autophagy flux in AR⁺ PCa cell lines (data not shown), as a positive control we also treated the cells with CQ to prevent basal p62 and LC3 turnover and maintain p62 and LC3 interaction.

Taken together, our data suggest an autophagy-dependent function for p62 in ERα^-/PR^- BCa cell lines that is induced by IL-1 in ERα⁺/PR⁺ BCa cell lines, and this function is likely conserved in PCa cells. Studies are underway to definitively determine the mechanistic function(s) of p62 in hormone receptor negative versus positive BCa and PCa cell lines and in response to IL-1.

4. DISCUSSION

ERα^- BCa tumors, such as triple negative tumors, are innately resistant to ERα-targeting endocrine therapy and a subpopulation of endocrine resistant ERα⁺ BCa patients have or evolve tumors with low or no ERα accumulation^1,4. Undoubtedly, tumor cells with low or no ERα elicit compensatory survival and growth pathways to compensate for ERα loss and to evade endocrine therapy, but the underlying molecular survival mechanisms are still being explored.

The Kabos lab showed that HS-5 BMSCs repress ERα expression and, when co-cultured with MCF7 cells, reduce tumor sensitivity to tamoxifen in vivo, possibly through compensatory EGFR or HER2 signaling⁴⁸. We show that IL-1 signaling mediates HS-5 BMSC repression of ERα and PR levels concomitant with p62 upregulation and autophagy induction in ERα⁺/PR⁺ BCa cell lines (Fig. 2 & 3); and we contend that IL-1-induced p62 upregulation and autophagy induction also contribute to BCa cell survival, growth, and endocrine resistance when hormone receptors are repressed. In support of this notion, knockout or silencing of p62 or autophagy have been shown to prevent breast cancer initiation, progression, or metastasis in mouse models^37,38,49 and in vitro studies demonstrate that autophagy is cytoprotective against endocrine therapy^50–52.

IL-1 inversely correlates with ERα in BCa patient tumors^13–15, is elevated in BCa tumor tissue and patient serum, and correlates with relapse and poor prognosis^17–20. However, it will be important to determine if IL-1 levels can be used to identify patients that will and/or have acquired endocrine resistance due to low or no ERα accumulation. These patients could benefit from IL-1Ra adjuvant therapy to (re)sensitize their tumors to endocrine therapy. Furthermore, while p62 overexpression^34–38 or elevated autophagy^39,40 are known to be associated with poor prognosis in BCa, a correlation between IL-1, p62, and autophagy in BCa endocrine resistance has yet to be determined. If such a correlation exists, these patients would also benefit from p62- and autophagy-targeting therapies.

Based on our data showing that ERα^-/PR^- BCa cell lines are insensitive to IL-1-induced p62 accumulation or autophagy flux (Fig. 4), their high basal levels (Fig. 1) suggest that ERα^-/PR^- BCa cells have evolved a requirement for p62 and autophagy that could be therapeutically exploited. Indeed, p62³⁴ and autophagy⁵³ are elevated in triple negative BCa patients and autophagy is cytoprotective for triple negative BCa cells^53–55.

Finally, IL-1 induces p62-LC3 interaction (Fig. 5 & 6), but the autophagy-dependent p62 function in BCa progression or endocrine resistance is unknown. As such, diverse roles for p62 in BCa initiation and progression are constantly being discovered. For example, p62 interaction with LC3 and KEAP1 may support BCa mammosphere formation through NRF2 signaling⁵⁶, p62 interaction with VANGL2 may promote BCa tumor growth through non-canonical WNT signaling⁵⁷, and p62 interaction with VIMENTIN may mediate BCa cell metastasis³⁸. Thus, given the multifunctional nature of p62, it is likely that IL-1 promotes both autophagy-dependent and –independent p62 functions in BCa cells. Experiments are underway to delineate IL-1-induced p62 functions in BCa that may contribute to endocrine resistance.

5. CONCLUSION

Our current and previous^28,43,44 published reports suggest that p62 and autophagy regulation and function are conserved and stratified by receptor status in BCa and PCa cells. Given the similarities in BCa and PCa etiology, our previous and current data reveals an important, seemingly conserved, role for IL-1 in treatment resistance and aggressive disease – the repression of hormone receptors concomitant with the upregulation of prosurvival proteins. IL-1, p62, and autophagy signaling could each be exploited to prevent and treat progressive disease for both BCa and PCa; and the FDA approved and experimental therapies in cancer clinical trials targeting IL-1⁵⁸, p62⁵⁹, and autophagy^60–62 emphasize the importance of these pathways in driving cancer progression.