A peptide-induced regression of low-grade malignant breast tumor via activation of autophagy in breast cancer cells

Abstract

Peptides have recently gained much attention for their therapeutic role in several diseases without having significant side effects. These are short chain amino acids and can be synthesized chemically with a great extent of purity, which renders them very effective, precise and safe for the body. So far, very few artificially synthesized peptides have been developed that have autophagy-inducing properties. Among them, a very small number of peptides have been demonstrated to have anti-tumorigenic effect by modulating autophagy with most of them lacking specificity for target cells. This study aims to investigate the regulation of autophagy through noncanonical activation of protease-activated receptor 1 (PAR1) by hemagglutinin protease (HAP). Unlike the canonical activation of PAR1 by thrombin, which enhances cell proliferation via mechanistic target of rapamycin (mTOR) signaling, HAP-induced noncanonical activation downregulates mTOR activation and triggers autophagy in both estrogen receptor positive/ progesterone receptor positive/ HER2 negative (ER+/PR+/HER2-) and triple negative (ER-/PR-/HER2-) breast cancer cells. This noncanonical activation generates a N-terminal sequence in PAR1. When this sequence is mimicked by a synthetic peptide, it induces autophagy independently of HAP in the breast cancer cells in vitro. Further in vivo investigation in BALB/c mouse model of low-grade malignant breast tumor reveals that the peptide-induced autophagy significantly inhibits tumor growth and delays tumor progression. Importantly, the lack of PAR1 expression in normal, healthy breast epithelial cells facilitates the peptide to selectively target breast cancer cells with relatively high PAR1 expression, highlighting its potential as a therapeutic agent against low-grade malignant breast tumor. These findings provide valuable insights into autophagy activation by a synthetic peptide that targets breast cancer cells with high PAR1 expression and for the first time shows peptide mediated targeted therapy of low-grade malignant breast tumor.

Graphical Abstract

Schematic representation of the plausible molecular mechanism underlying the induction of autophagy through HAP-mediated non-canonical activation of PAR1. HAP cleaves PAR1 at a distinct site, thereby generating a N-terminal sequence, PAFISED (in humans), which functions as a tethered ligand and subsequently activates PAR1. Following activation, non-canonically stimulated PAR1 engages downstream G protein-mediated signaling pathways, resulting in suppression of MAP kinase activation and consequent inhibition of mTOR signaling. Concurrent attenuation of mTOR activity, together with enhanced phosphorylation of AMPK, promotes activation of the AMPK-ULK1-BECN1 signaling axis, leading to the induction of autophagy

Supplementary Information

The online version contains supplementary material available at 10.1186/s10020-026-01514-4.

Introduction

Breast cancer, being the most frequently occurring cancer among women and second leading cause of death by cancer in women, imposes a high burden on the health care (Siegel et al. 2018). Although rapidly evolving treatment strategies and discovery of quite a number of drugs have unraveled a new horizon for breast cancer treatment, the major challenges lie in the non-specificity and cytotoxicity of these drugs and development of drug resistance in the breast cancer cells (Masoud and Pages 2017) along with the limited drug sensitivity of the triple negative breast cancer cells (Burguin et al. 2021). Despite the progress in breast cancer treatment modalities, it still stands crucial to find new therapeutic agent that shows precise selectivity in targeting breast cancer cells and effective against triple negative breast cancer cells along with other subtypes. Beside the conventional drugs, artificially synthesized peptides with anti-cancer property have recently emerged as suitable therapeutic agents for their purity, high stability and low cytotoxic effects against non-carcinogenic cells. Till now, five artificially synthesized peptides have been reported to exert anti-tumorigenic effect by inducing macroautophagy (hereafter referred to as autophagy). However, many of them lack the specificity for cancer cells. Hence, they are not suitable as drug candidates (Cai et al. 2025).

Beyond its cytoprotective functions, autophagy can lead to a specific form of cellular death known as type II programmed cell death. Recruiting this type of programmed cell death by inducing autophagy in malignant cells is a suitable treatment approach to kill cancer cells and inhibit cancer progression. Generally, cells recruit autophagy as a sophisticated recycling mechanism, in which dysfunctional or superfluous organelles, invading pathogens, or nonessential organelles under stressed condition are enclosed within a double-membrane vesicle termed phagophore, resulting in the formation of autophagosome (AP). Sequestosome 1 (SQSTM1/p62) protein interacts with LC3 protein and acts as adapter molecule for transporting selected ubiquitinated molecules to the phagophore (Ichimura et al. 2008; Pankiv et al. 2007; Kirkin et al. 2009). AP subsequently fuses with lysosome to form autolysosome (AL) that degrades sequestered components through proteolytic activity of lysosomal hydrolases and the building blocks are released to the cytosol for further utilization (Parzych and Klionsky 2014). In dysregulated autophagy, various stimuli initiate the process of phagophore formation and sequestration but fail to further proceed to the fusion of AP and lysosome, resulting into accumulation of cargo (Zhang et al. 2022), which is implicated in several neurodegenerative disorders including Alzheimer’s and Parkinson’s diseases, cancers, autoimmune diseases and metabolic syndromes including type 2 diabetes (Klionsky et al. 2021).

In cancer, autophagy plays a complex and dual role that varies with the different stages of cancer. In the early stages of cancer, autophagy acts as a tumor-suppressive mechanism by degrading misfolded proteins, potentially preventing further tumor development. Conversely, in later stages, cancer cells exploit autophagy for surviving and adapting to the hostile environment within the host (Yun and Lee 2018; Rosenfeldt and Ryan 2011; Gewirtz 2014; Russell and Guan 2022). Cancer cells utilize autophagy as a resistance mechanism to sustain the metabolic stress, disrupted protein metabolism, kinase inhibitors and chemotherapy. Considering this property of cancer cells, autophagy inhibiting agents, like hydroxychloroquine has been clinically evaluated in combination with other conventional chemotherapy or immunotherapy and showed promising result in cancer treatment (Amaravadi et al. 2011). However, considering the anti-tumorigenic role of autophagy in early-stage cancer or low-grade malignant tumor, several autophagy inducing agents being investigated for the treatment for early-stage cancer. Although, a basal level of autophagy is present in almost all living cells, an over-activation of autophagy can lead to type II programmed cell death, which is also utilized in cancer treatment (Fulda and Kogel 2015).

The coordinated interplay between proteasomal degradation and autophagy constitutes a dynamic compensatory network for cellular proteostasis, wherein impairment of one degradative pathway frequently elicits upregulation of the other. For instance, inhibition of 26S proteasome leads to the activation of AMP-activated protein kinase (AMPK) (Deshmukh and Dou 2015; Jiang et al. 2015; Min et al. 2014), a positive regulator of autophagy (Kim et al. 2011). In this context, it is imperative to acknowledge that a spectrum of proteolytic mediators, including some cathepsins and proteases like calpains, exert pleiotropic and context-dependent dual regulatory effect on both apoptotic and autophagic pathways (Tsapras and Nezis 2017; Yousefi et al. 2006). Proteases can trigger various signaling pathways by interacting with protease-activated receptors (PARs), a class of G protein-coupled receptors, which include four subtypes (PAR1–4) (Soh et al. 2010). Among them, the expression of PAR1, PAR2 and PAR4 is significantly higher in cancer cells than in normal cells, making them promising targets for cancer therapy (Flynn and Buret 2004). Thrombin, the conventional activator of PAR1, activates PAR1 by making a proteolytic cleavage at its N-terminal extracellular region and the N-terminal region acts as a tethered ligand, which binds to the second extracellular loop of the receptor. This thrombin-mediated activation of PAR1 upregulates the phosphorylation of mechanistic target of rapamycin (mTOR) (Parrales et al. 2013), a master regulator of autophagy (Saxton and Sabatini 2017; Rabanal-Ruiz et al. 2017). Therefore, the canonical activation of PAR1 by thrombin activates coagulation cascades and leads to several other cellular events, including induction of angiogenesis and promotion of growth and proliferation of metastatic cancer cells (Elste and Petersen 2010).

Hemagglutinin protease (HAP), a 35-kDa zinc-dependent secretory metalloprotease, isolated from Vibrio choleraeO1 strain (C6709), induces apoptosis in breast cancer cells (Ray et al. 2016). Among the four subtypes of PARs, HAP specifically engages with PAR1, facilitating its activation through proteolytic cleavage at the N-terminus. In Ehrlich ascites carcinoma (EAC) cells, a murine breast cancer cell line, this cleavage generates a new N-terminal extracellular segment, sequenced as PFISEDASGY. However, the precise molecular region of the newly generated N-terminus that acts as a tethered ligand to intramolecularly engage the receptor and initiate downstream signaling remains unresolved (Ray and Pal 2016). Subsequently, based on the sequence of the HAP-mediated cleavage site of PAR1 in EAC cells, a peptide (PFISED) has been designed, which induced apoptosis in both breast and colon cancer cell lines (Ray et al. 2018).

Evidently, autophagy being an early response of cells and apoptosis being a late one (Lockshin and Zakeri 2004), they share some common upstream signaling pathways, allowing a single stimulus to activate both the processes in many instances (Thorburn 2008; Crighton et al. 2006; Arico et al. 2001). However, the regulation of autophagy by the peptide designed from the sequence of N-terminal end of noncanonically activated PAR1 and its impact on breast cancer remains largely unexplored. Therefore, the current study explored the potential of the peptide as anti-tumorigenic agent in low-grade malignant breast tumor by inducing autophagy in a dose- and time-dependent manner in breast cancer cells.

Materials and methods

Bacterial strain growth conditions

The Vibrio cholerae O1 strain C6709 was selected for this study due to its harboring of a functional hapR gene, which encodes a regulatory protein that modulates several cellular processes in Vibrio cholerae, including the expression of hemagglutinin HAP (Jobling and Holmes 1997). Routine cultivation ofVibrio cholerae(C6709) was carried out in Luria broth (Himedia, M575) at 37℃ with shaking for 18 h (Saha et al. 2004). To preserve bacterial stocks, cells were stored at −70℃ in Luria broth supplemented with 20% sterile glycerol (Sigma-Aldrich, G5516) (Jobling and Holmes 1997).

Purification of the protease

HAP was isolated from Vibrio choleraeO1 strain C6709 and purified according to the protocol outlined by Ghosh et al. (2016). TheV. choleraeO1 strain C6709 was grown and harvested as previously described. The cell-free culture supernatant was sequentially filtered through 0.45 μm and 0.22 μm cellulose acetate membranes (Merck, HAWP04700 and Merck, GSWP04700) to eliminate residual cells. Outer membrane vesicles were subsequently removed from the filtrate by ultracentrifugation at 150,000 * g for 3 h at 4℃ using a T-865 rotor (Sorval Instruments, USA) (Lin 1996). The supernatant was carefully aspirated and subjected to salting out with 60% saturated ammonium sulfate (Sigma-Aldrich, A4418). Protein precipitation was achieved by centrifugation at 10,000 rotation per minute (RPM) for 20 min at 4℃. The resulting pellet was resuspended in 25 mM Tris (hydroxymethyl) aminomethane hydrochloride buffer (pH 7.4) (Sigma-Aldrich, 648317), dialyzed against the same buffer and concentrated using an Amicon® Ultra Centrifugal Filter (Millipore, UFC9050). The protein was then purified through a single-step ion-exchange chromatography column (2.5 * 20 cm) packed with Diethylaminoethyl cellulose-52 (DEAE-52) matrix (Whatman, 4057050) and pre-equilibrated with 25 mM Tris (hydroxymethyl) aminomethane hydrochloride buffer (pH 7.4). Proteins were eluted in the unbound fractions, which formed distinct peaks, pooled, dialyzed, concentrated and assayed for protease activity. Notably, the purification of HAP was performed without the use of Ethylenediamine tetraacetic acid (EDTA). Chromatographic separation was conducted using a Biologic Duo Flow Chromatographic System (Bio-Rad, USA). The purified HAP was stored in aliquots at −20℃.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE)

SDS-PAGE was performed using the method described by Laemmli (1970). Protein samples were denatured in sample buffer and separated on a 12.5% polyacrylamide gel using a discontinuous buffer system.

Protease activity assay

In this study, azocasein (Sigma-Aldrich, A2765) was employed as a substrate to assess the proteolytic activity of HAP (Ghosh et al. 2006). The substrate-enzyme mixture (azocasein-HAP) was incubated at 37℃ for 1 h, followed by termination of the reaction by the addition of 10% (w/v) trichloroacetic acid (Sigma-Aldrich, T8657). The precipitated protein was eliminated by centrifugation at 12,000 RPM for 4 min and the supernatant was transferred to a clean tube containing 525 mM sodium hydroxide. Absorbance was measured at 440 nm using a Shimadzu spectrophotometer (Shimadzu, Japan). Negative controls included the substrate with buffer and the substrate with EDTA-inactivated enzyme.

Cell culture

Human epithelial breast cancer cells MCF7 (ATCC, HTB-22) and MDA-MB-231 (ATCC, HTB-26), murine epithelial breast cancer cells 4T1 (ATCC, CRL-2539), and human normal breast epithelial cells MCF10A (ATCC, CRL-10317) were cultured (passages 10–12) and maintained in Minimum Essential Medium (MEM) (Gibco, 61100061), DMEM- High Glucose (Gibco, 12100046), RPMI 1640 (Gibco, 23400021) and Mammary Epithelial Cell Basal Medium (MEBM) (Lonza, CC-3151) with Mammary Epithelial Cell Growth Medium (MEGM) SingleQuots supplements (Lonza, CC-4136) respectively. Each medium was supplemented with 10% Fetal bovine serum (FBS) (Gibco, 10270106), 100 U/ml penicillin G and 100 µg/ml streptomycin sulfate (Gibco, 15140122) and cells were incubated at 37℃ in a 5% CO2 atmosphere using a Heracell incubator (Thermo Fisher Scientific, USA, 51026331). The MCF7 and MCF10A cell lines were generously provided by Dr. Somsubhra Nath, Presidency University, Kolkata, while MDA-MB-231 and 4T1 cell lines were received as gift from Dr. Saptak Banerjee, CNCI, Kolkata.

Western blotting

Western blotting was conducted as previously described by Nag et al. (2024). In brief, cultured cells were lysed using Radioimmunoprecipitation assay (RIPA) buffer (Sigma-Aldrich, R0278) and protein concentrations were determined using the Bicinchonic acid (BCA) assay kit (Thermo Fisher Scientific, 23225). Equal amounts of protein were loaded onto SDS-PAGE for separation and subsequently transferred electrophoretically to a Polyvinylidene fluoride (PVDF) membrane (Merck, IPVH00010). Following transfer, the membrane was blocked with 3% Bovine serum albumin (BSA) (Sigma-Aldrich, A1470) in Tris-buffered saline (TBS) and incubated overnight at 4℃ with the primary antibody (diluted according to the manufacturer's instructions) against the target protein. The blot was then washed with Tris-buffered saline-Tween 20 (TBST) buffer and incubated with Horseradish peroxidase (HRP)-conjugated secondary antibody (Cell Signaling Technology, 7076; 1:1000 and Cell Signaling Technology, 7074; 1:1000) for 2 h at room temperature. Protein bands were visualized using a Bio-Rad gel documentation system with an Enhanced chemiluminescence (ECL) substrate (Thermo Fisher Scientific, 34580) specific for HRP-conjugated antibodies. Western blot quantification was performed using GelQuant (Thermo Fisher Scientific, USA) and ImageJ (NIH, Bethesda, MD) software. At least three independent experiments were conducted to validate the results and band intensities were normalized to loading controls: GAPDH for MCF7, MDA-MB-231, MCF10A cells, tumor tissue cell lysate and Cyclophilin B for 4T1 cells.

Analysis of lysosomal vacuoles by confocal microscopy

Variation in the number and in acidification status of lysosomal vacuoles was analyzed by staining with LysoTracker™ Green DND-26 (Thermo Fisher Scientific, L7526), following the method described by Klepka et al. (2022) with minor modifications tailored to the cell lines. Briefly, MCF7 and 4T1 cells were seeded onto cover slips in 6-well plates at a density of 0.3 × 10⁵ cells per well. After 24 h of incubation, cells were treated with HAP (0.5 μg/ml for MCF7 and 0.75 μg/ml for 4T1) for 30 min, with or without pre-treatment with 25 nM Bafilomycin A1, an inhibitor of lysosomal-acidification (Sigma-Aldrich, B1793) for 24 h. Untreated cells and those treated only with Bafilomycin A1 served as normal and negative controls, respectively. Following HAP treatment, cells were fixed with 4% paraformaldehyde for 10 min at room temperature. They were then permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, T8787) in 0.1% sodium citrate solution and blocked with 5% FBS prepared in TBST. The cells were incubated with LysoTracker™ Green DND-26 for 45 min at 37℃. Nuclei were stained with Hoechst 33342(Cell Signaling Technology, 4082; working concentration 1.0 μg/ml) for 10 min at 37℃. After washing twice, cells were mounted with 10% glycerol (Sigma-Aldrich, G5516). Images were acquired using a confocal microscope (Zeiss, Germany, LSM 710) with a Plan-Apochromat 63X/1.40 NA objective. Data were processed and analyzed using Zen 3.4 (blue edition) and ImageJ software.

Autophagy detection by confocal microscopy and flow cytometry

Autophagy was detected by confocal microscopy using the Autophagy Assay Kit (Abcam, ab139484) according to the manufacturer’s instructions. The kit contains a green fluorescent dye that specifically stains autophagosomal and autolysosomal vacuoles. Briefly, MCF7, 4T1 and MDA-MB-231 cells were seeded onto cover slips in 6-well plates at a density of 0.3 × 10⁵ cells per well. After 24 h of incubation, cells were treated with HAP (0.5 μg/ml for MCF7, 0.75 μg/ml for 4T1 and 0.5 μg/ml for MDA-MB-231) for 30 min and with peptides (50 μM dose of PAFISED for MCF7, 75 μM dose of PFISED for 4T1 and 25 μM dose of PAFISED for MDA-MB-231) for 4 h. Following treatment, cells were washed twice with 1X Assay Buffer (diluted from the 10X stock provided with the kit) supplemented with 5% FBS. Cells were then incubated at 37℃ for 30 min with 100 μl of staining solution, which contained 2 μl of the green fluorescent dye and 1 μl of nuclear stain (Hoechst), both provided in the kit, in 1 ml of 1X Assay Buffer, supplemented with 5% FBS. After incubation, the staining solution was removed and cells were washed twice with 1X Assay Buffer containing 5% FBS. Cells were then fixed with 4% paraformaldehyde for 15 min and washed thrice with 1X Assay Buffer containing 5% FBS. Finally, cells were mounted on glass slides with 10% glycerol. Green fluorescence was detected using the FITC filter and nuclear staining was visualized using the DAPI filter on a confocal microscope (Zeiss, Germany, LSM 710) with a Plan-Apochromat 63X/1.40 NA objective. Data were processed and analyzed using Zen 3.4 (blue edition) and ImageJ software.

The induction of autophagy was further analyzed by flow cytometry using the same Autophagy Assay Kit following the manufacturer’s protocol. In a brief, MCF7 and 4T1 cells were treated with HAP and the peptides as mentioned earlier and harvested by brief incubation with 0.25% trypsin-EDTA (Gibco, 25200072) at 37℃. Cells were washed with 1X Assay Buffer, supplemented with 5% FBS, as outlined earlier and stained with 250 µl of the staining solution, comprising of 1 μl of the green fluorescent dye (provided in the kit) in 1 ml of 1X Assay Buffer supplemented with 5% FBS. Flowcytometry data were acquired using BD FACSAria™ II (BD, USA). Analysis of the flow cytometry data was performed using FloJo™ 10 software.

Immunocytochemistry and confocal imaging

Immunocytochemistry and confocal imaging were conducted following the procedures outlined by Nag et al. (2024). In brief, MCF7, 4T1 and MDA-MB-231 cells were grown on coverslips in 6 well cell culture plates and treated with HAP (0.5 µg/ml for MCF7, 0.75 µg/ml for 4T1 and 0.5 µg/ml for MDA-MB-231) for 30 min. Following the treatment, cells were fixed with 4% paraformaldehyde for 10 min at room temperature. Cells were subsequently permeabilized using 0.1% Triton X-100 in 0.1% sodium citrate solution and blocked with TBST supplemented with 5% FBS for 2 h. They were incubated overnight at 4℃ in a moist chamber with primary antibody (diluted according to the manufacturer’s instructions) against the protein of interest. Following primary antibody incubation, cells were washed with TBS and incubated for 2 h at room temperature with Alexa 488 conjugated (Cell Signaling Technology, 4408; 1:200 and Cell Signaling Technology, 4412; 1:500) or Alexa 555 conjugated (Cell Signaling Technology, 4413; 1:500 and Cell Signaling Technology, 4409; 1:500) secondary antibodies. Nuclei were stained with Hoechst 33342 (Cell Signaling Technology, 4082; working concentration 1.0 μg/ml) for 10 min at 37℃. After washing the cells twice with TBS, they were mounted with 10% glycerol. Confocal images were captured using a confocal microscope (Zeiss, Germany, LSM 710) with a Plan-Apochromat 63X/1.40 NA objective. Image processing and analysis were performed using Zen 3.4 (blue edition) and ImageJ software.

Raising of antisera against purified HAP

Antiserum against HAP was generated by immunizing a New Zealand White rabbit, following the method described by Mondal et al. (2016). The rabbit was initially administered an intramuscular injection of 100 mg of purified HAP emulsified with an equal volume of Freund’s complete adjuvant (Sigma-Aldrich, F5881). This was followed by four booster injections of 100 mg of HAP emulsified with Freund’s incomplete adjuvant (Sigma-Aldrich, F5506) at 7 day intervals. Blood samples were collected from the rabbit on day 0 and 3 days after the final booster injection, allowing the blood to clot overnight at room temperature. The sera were then collected, centrifuged (1,000 RPM for 5 min), diluted in autoclaved glycerol and stored at −80℃ until use at a 1:1000 dilution.

Colocalization analysis by confocal imaging

Colocalization of HAP with PAR1 and of LAMP1 with LC3II was analyzed in HAP-treated (0.5 µg/ml for MCF7, 0.75 µg/ml for 4T1 and 0.5 µg/ml for MDA-MB-231 cells for 30 min) and peptide-treated (50 µM of PAFISED for MCF7, 75 µM of PFISED for 4T1 and 25 µM of PAFISED for MDA-MB-231 cells for 4 h) cells, respectively, by confocal microscopy, following the protocol described by Das et al. (2022). In brief, cells were grown on coverslips in 6 well cell culture plates. After the treatment, they were fixed with 4% paraformaldehyde for 10 min at room temperature and then permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate solution. The cells were blocked with 5% FBS in TBST for 2 h and incubated overnight at 4℃ in a moist chamber with both of the primary antibodies (diluted as per the manufacturer's instructions) against the respective proteins of interest. After washing twice with TBS, cells were incubated for 2 h at room temperature with Alexa 488 (Cell Signaling Technology, 4408; 1:500 and Cell Signaling Technology, 4412; 1:500) and Alexa 555 (Cell Signaling Technology, 4413; 1:500 and Cell Signaling Technology, 4409; 1:500) conjugated secondary antibodies. Nuclei were stained with Hoechst 33342 (Cell Signaling Technology, 4082; working concentration 1.0 μg/ml) for 10 min at 37℃. Following two washes with TBS, cells were mounted with 10% glycerol on a glass slide. Confocal images were acquired using a Zeiss LSM 710 microscope (Zeiss, Germany) with a Plan-Apochromat 63X/1.40 NA objective. Image processing and analysis were performed using Zen 3.4 (blue edition) and ImageJ software. Colocalization was quantified by calculating Pearson’s correlation coefficient.

Knock down of PAR1

MCF7 cells were first starved for 18 h in incomplete MEM before transfection. To transiently knock down PAR1, PAR1 siRNA (CGGUCUGUUAUGUGUCUAUdTdT) (Eurogentec, Seraing, Belgium) was used. Lipofectamine 3000 (Thermo Fisher Scientific, L3000001) served as the transfection reagent and the procedure was carried out according to the manufacturer's instructions. Transfection was performed when the cells reached approximately 70% confluency. The transfection reagents were diluted in Opti-MEM (Gibco, 31985062) and added onto the cells dropwise. After 4 h, the medium was replaced with fresh complete MEM (supplemented with 10% FBS). Cells were transfected for 72 h to achieve efficient knock down. Scramble siRNA (Santa Cruz Biotechnology, sc-37007) was used as a control in this knock down study. Knock down efficiency was confirmed by Western blotting.

Flow cytometric analysis of lysosomal vacuoles in PAR1 knocked down (KD) cells

Flow cytometry was employed to analyse the lysosomal activity in HAP-treated wild type (WT) and PAR1 knocked down (KD) MCF7 cells. Detection of these vacuoles was achieved by staining with LysoTracker™ Green DND-26, following the protocol outlined by Chikte et al. (2014) with cell-specific optimization of dye concentration and incubation time. WT and PAR1 KD MCF7 cells were seeded at a density of 0.3 × 10⁵ cells per well in 6-well tissue culture plates and were grown overnight. Both WT and PAR1 KD cells were treated with HAP at doses of 0.5 μg/ml and 0.75 μg/ml for 30 as well as 60 min. After treatment, cells were harvested by trypsinization (0.25% Trypsin-EDTA), washed twice with TBS and resuspended in 500 μl of TBS. Cells were incubated with 60 nM LysoTracker™ Green DND-26 for 45 min at 37℃ for staining followed by washing with TBS twice. Finally, cells were resuspended in 500 µl of TBS and Flow cytometric analysis was conducted on the single cell populations, gated by forward scatter (FSC) and side scatter (SSC). The FITC channel (excitation at 488 nm and detection range of 515-545 nm) was used to identify LysoTracker™ Green DND-26-positive cells. Data acquisition was performed using a BD FACSAria™ II (BD, USA) and analysis was carried out using FloJo™ 10 software.

in silico analysis

The crystal structure of human PAR1 bound to Vorapaxar (PDB ID: 3VW7) was retrieved from Protein Data Bank (https://www.rcsb.org) (Zhang et al. 2012). The physio-chemical properties, drug-likeness and pharmacokinetics of peptide were checked using swissADME (https://www.swissadme.ch) and toxicity was checked by ToxinPred (https://crdd.osdd.net/raghava/toxinpred) (Daina et al. 2017; Gupta et al. 2013). After checking the peptides properties, three servers were used one for peptide structure modelling and other two were used for molecular docking. Servers were PEP-FOLD4 (https://bioserv.rpbs.univ-paris-diderot.fr/services/PEP-FOLD4), Autodock Vina (https://www.swissdock.ch), CABS-dock (https://biocomp.chem.uw.edu.pl/CABSdock) (Rey et al. 2023; Eberhardt et al. 2021; Trott and Olson 2010; Kurcinski et al. 2019). PEP- FOLD4 was used for the peptide structure prediction; Autodock Vina and CABS-dock were used for the molecular docking in which Blind docking was performed by CABS-dock. The docking study was conducted in the following steps: 1. Re-docked Vorapaxar by Autodock Vina (Hidayat et al. 2015), 2. Peptide docking against PAR 1 by Autodock Vina 3. Blind docking by CABS-dock 4. Binding interactions and analysis by Biovia Discovery Studio visualizer 2024 (Dassault Systèmes BIOVIA, USA).

Transmission electron microscopy (TEM) imaging

For the TEM experiment, HAP and peptide treated and untreated MCF7 and 4T1 cells were washed three times in phosphate-buffered saline (PBS) and fixed with 2.5% glutaraldehyde in PBS (pH 7.4) overnight at 4℃. The samples were then dyed with 2% uranyl acetate for 30 min, washed three times with distilled water and dehydrated for 15 min. Subsequently, the samples were embedded in EPON 812 resin (Head Biotechnology, E8000) and cured for 24 h at 70℃. Ultrathin slices were procured using an ultramicrotome and then subjected to staining with a solution containing 2% uranyl acetate and lead citrate. The samples were imaged with a FEI Tecnai 12 BioTwin Transmission Electron Microscope operating at 100 kV.

Analysis of autophagic flux by pMXs GFP-LC3-RFP reporter construct

Autophagic flux in MCF7, 4T1 and MDA-MB-231 cells treated with HAP and peptides was assessed by transiently expressing the pMXs GFP-LC3-RFP plasmid in these cells. This plasmid was originally cloned in NEB® Stable Competent E. coli (New England Biolabs, USA) and obtained as an agarose stab culture as a generous gift from Noboru Mizushima Laboratory (Addgene, 117413) (Morita et al. 2018). The plasmid was isolated using the QIAprep Spin Miniprep Kit (Qiagen, 27104) according to the manufacturer’s protocol and its purity was confirmed by agarose gel electrophoresis.

Prior to transfection, MCF7, 4T1 and MDA-MB-231 cells were cultured on coverslips and subjected to starvation for 18 h in their respective media. Transfection was performed using Lipofectamine 3000 (Thermo Fisher Scientific, L3000001), following the manufacturer’s protocol. Cells were transfected at approximately 70% confluency and all the reagents, including P3000, were added in Opti-MEM (Gibco, 31985062). After 4 h, the media was replaced with fresh complete media (supplemented with 10% FBS) suitable for the respective cell lines. MCF7 cells were transfected for 72 h and 4T1 and MDA-MB-231 cells were transfected for 48 h, to ensure optimal transfection, which was confirmed by fluorescence microscopy. pAM_1C Empty Vector (Active Motiff, 53023) transfected cells were used as vehicle controls.

Following transfection, cells were treated with HAP (0.5 µg/ml for MCF7, 0.75 µg/ml for 4T1 and 0.5 µg/ml for MDA-MB-231) for 30 min or with peptides (50 µM dose of PAFISED for MCF7, 75 µM dose of PFISED for 4T1 and 25 µM dose of PAFISED for MDA-MB-231) for 4 h and then fixed with 4% paraformaldehyde. After fixation, cells were stained with Hoechst 33342 (Cell Signaling Technology, 4082; working concentration 1.0 μg/ml) for 10 min at 37℃ and mounted with 10% glycerol on glass slides. Images were acquired using a confocal microscope (Zeiss, Germany, LSM 710) with a Plan-Apochromat 63X/1.40 NA objective. Data processing was carried out using Zen 3.4 (blue edition) software and the number of red and green puncta was quantified and analyzed using ImageJ software.

Cell viability and cell proliferation assay

MTT assay was conducted according to the standard protocol described by Kar et al. (2014) to assess the effect of HAP and peptides (PAFISED for MCF7 and PFISED for 4T1 cells) on the cellular viability of breast cancer cells. Approximately 1 × 10⁴ viable cells per well were seeded in flat bottom 96-well cell culture plates, where viability was determined using 0.4% Trypan blue solution (Thermo Fisher Scientific, 15250061) before seeding. After 24 h of incubation, MCF7 and 4T1 cells were treated with the optimized concentrations of HAP (0.5 μg/ml for MCF7 and 0.75 μg/ml for 4T1) for 30 min and peptides (50 μM dose of PAFISED for MCF7 and 75 μM dose of PFISED for 4T1) for 4 h, with or without pre-treatment with 25 mM 3-MA, an autophagy inhibitor for 24 h. Following the treatment with HAP and the peptides, MTT solution (Thermo Fisher Scientific, M6494; 0.8 mg/ml in serum-free medium) was added and cells were incubated for 4 h in the dark at 37℃. The media was then removed and 100 µl Dimethyl sulfoxide (Sigma-Aldrich, D5879) was added in each well to solubilize the purple crystal formazan. After a 15-min incubation in the dark, absorbance was measured at 595 nm using a spectrophotomteric plate reader. The mean of three replicates was used to calculate the cell survival percentage.

To examine the impact of HAP- and peptide-induced autophagy on breast cancer cell proliferation, Ki-67 nuclear antigen immunofluorescence assay was performed in MCF7 and 4T1 cells following the standard protocol (Li et al. 2023) with slight modifications optimized for the current experimental condition. Cells were treated with HAP and the respective peptides (PAFISED for MCF7 and PFISED for 4T1) as described above. After treatment, cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate solution and blocked with 5% serum in TBST. Cells were incubated overnight at 4℃ with Ki-67 (D3B5) Rabbit mAb (Cell Signaling Technology, 9129; 1:400) in a moist chamber. Following washing with TBS, cells were incubated with Alexa 488-conjugated anti-rabbit IgG antibody (Cell Signaling Technology, 4412; 1:200). Nuclei were stained with Hoechst 33342 (Cell Signaling Technology, 4082; working concentration 1.0 μg/ml) for 10 min at 37℃. After two additional washes with TBS, cells were mounted on glass slides with 10% glycerol. Images were captured using a confocal microscope (Zeiss, Germany, LSM 710) with a Plan-Apochromat 63X/1.40 NA objective. Data were processed and analyzed using Zen 3.4 (blue edition) and ImageJ software.

Enzyme-linked immunosorbent assay (ELISA) for cytokine quantification

Cell culture media were collected from MCF7, MDA-MB-231 and 4T1 cells grown in 6-well cell culture plate following the treatment with peptides of previously optimized doses for induction of autophagy. Harvested media were centrifuged at 1000 * g for 10 min at 4 °C to remove cellular debris. The cell free supernatants were used for analysis of secreted cytokines by ELISA. The concentrations of the secreted interleukin-2 (IL-2), interferon-alpha (IFN-α), transforming growth factor-beta (TGF-β), granulocyte macrophage colony-stimulating factor (GM-CSF) and interleukin-10 (IL-10) in the cell supernatants were quantified using sandwich ELISA kits of R&D Systems (R&D Systems, S2050, 41100-1, DY240-05, SGM00, S1000B for MCF7 and MDA-MB-231 cells and R&D Systems, DY402-05, 42120-2, DY1679-05, SMGM00, SM1000B for 4T1 cells) according to the manufacturers’ protocols. For TGF-β measurement, samples were subjected to acid activation with HCl and subsequent neutralization with NaOH/HEPES prior to assay to convert latent TGF-β into its immunoreactive form, as recommended. Cytokine concentrations were calculated from standard curves generated using recombinant standards provided with the kits. Data were expressed as pg/ml. All samples were assayed in triplicate and independent experiments were performed at least three times to ensure reproducibility.

in vivo experiments

Animal model

In order to establish a low-grade malignant breast tumor mouse model for studying the effect of the peptide (PFISED) induced autophagy in early-stage breast cancer, first, malignant solid breast tumors were developed by implanting 4T1 cells in 4–6 weeks old female BALB/c mice following the protocol described by Pulaski et al. (2000). Briefly, malignant solid tumors were developed by implanting 4T1 cells (0.5*10⁷ cells/kg of body weight) subcutaneously at the abdominal mammary fat pad and allowed them to multiply. Cells were suspended in serum free RPMI 1640 media and each mouse was injected with 100 µl of the suspension. The mice were randomly distributed in four groups, each containing 5 mice (n = 5). The groups were stratified based on the time which the tumors were harvested at after the implantation of 4T1 cells. (i) Group I: harvested the tumor at day 7, (ii) group II: harvested the tumor at day 14, (ii) group III: harvested the tumor at day 21, (iv) group IV: harvested the tumor at day 28.

Tumors from all the groups were scored on the basis of histopathological analysis and grades were determined following the Nottingham Grading System. Using the Nottingham Grading System, we classified tumors based on the assigned scores considering three aspects of the tumor tissue, including ductal/tubular/glandular differentiation, nuclear pleomorphism & mitotic figure and correlated them with Ki-67 expression. A cumulative score between 3-5 refers to a low-grade/ grade I (well-differentiated) malignant breast tumor, whereas a cumulative score between 6-7 indicates an intermediate-grade/ grade II (moderately differentiated) malignant breast tumor. A cumulative score between 8-9 signifies a high-grade/ grade III (poorly differentiated) malignant breast tumor (Boiesen et al. 2000; van Dooijeweert et al. 2022) They have been further correlated with the size of the tumors.

Before evaluating the effect of the peptide PFISED on low-grade malignant breast tumor in BALB/c mice, the optimum dose of the peptide was identified based on the toxicity evaluation of the peptide following the protocol described by Lamichhane et al. (2019). Briefly peptide PFISED of six different doses (1, 2.5, 5, 7.5, 10 and 20 mg/kg body weight) was administered at an interval of 3 days starting from Day 0 (1st dose) to Day 12 (5th dose) and at day 15, blood was collected from retro-orbital sinus behind the eye. For the evaluation of hematological parameters, blood was collected in EDTA coated tubes. Hematological parameters including eosinophil, hemoglobin, lymphocyte, monocyte and neutrophil were counted using Medonic M32 Cell Counter (Boule, Sweden, M32B). Liver function was assessed by measuring ALT and AST in blood serum by ALT Activity Assay Kit (Sigma-Aldrich, MAK052) and Aspartate Transaminase (AST) Assay Kit (Sigma-Aldrich, MAK467) respectively and kidney function was evaluated by measuring creatinine and BUN (calculated from serum urea) by Creatinine Assay Kit (Sigma-Aldrich, MAK475) and Urea Assay Kit (Sigma-Aldrich, MAK471) respectively.

Based on the gradation of malignancy, out of the three groups with low-grade tumor, group II (carrying tumor of 14 days) was selected randomly for further analyses to evaluate the role of the peptide (PFISED) induced autophagy in the treatment of low-grade malignant murine breast tumor. Three groups of experimental mice were designed, each group containing 5 mice, carrying breast tumor of 14 days. (i) Group I: without any treatment, considered as control, (ii) group II: injected with 5 mg/kg of body weight dose (as optimized by toxicity assessment of the peptide) of the peptide PFISED, (ii) group III: injected with the aforementioned dose of peptide PFISED along with autophagy inhibitor, 3-MA (15 µg/gram of body weight). Mice of group II were injected with the peptide PFISED subcutaneously at the site of tumor at an interval of 3 days starting from the 14th day to 26th day after the implantation of 4T1 cells. In the mice of group III, 3-MA along with the peptide PFISED was administered according to the same experimental design of group II. The peptide PFISED and 3-MA were diluted in serum free RPMI 1640 and at one time each mouse was injected with 100 µl of the solution. Since, in silico analysis indicated that the peptide PFISED didn’t follow the Lipinski and Veber rules, the possibility of oral administration of the peptide had been excluded.

All the mice were procured from the animal breeding facility of ICMR-National Institute of Research in Bacterial Infections and were supplied with food pellets and autoclaved water ad libitum. All the mice were given numbers and randomly assigned to different groups by computer-based randomization using “RAND” command in a spreadsheet in Microsoft Excel. The inclusion criteria for selecting the experimental mice were i) female, ii) age between 4 to 6 weeks, iii) initial body weight between 15-24 g and the exclusion criteria were i) ongoing or previous pregnancy, ii) lactating mice. When required, all the mice were euthanized via asphyxiation in a CO₂ chamber and the tumors were surgically excised. All the animal experimental procedures were carried out following the ARRIVE (Animal Research: Reporting of in vivo Experiments) guidelines.

Histology and immunohistochemistry

Immunohistochemical and hematoxylin-eosin (H&E) staining of tumor tissues was done as per protocol described by Kim et al. (2019) with some minor modifications as required for the experiment. In a brief, tissues were fixed in 10% formaldehyde for at least 24 h and then processed by an automated tissue processor (Leica TP1020, Leica Microsystems, Germany). Processed tissues were embedded in paraffin and sectioned (5 µM thin) from the paraffin block by microtome, deparaffinized using xylene (Sigma-Aldrich, 6.08685), serially rehydrated and permeabilized with 0.5% Triton X-100 (Sigma-Aldrich, 93443) for 10 min. Tissues were then incubated with 3% Hydrogen peroxide (H₂O₂) to inactivate the endogenous peroxidase activity and boiled in 10 mM sodium citrate buffer pH 6.0 for 10 min antigen retrieval. Tissue sections were blocked by incubating with TBST containing 5% FBS for 1 h. They were stained with hematoxylin and eosin for histopathological analysis and for immunohistochemical analysis they were incubated with Ki-67 antibody (Cell Signaling Technology, 12202; 1:200) overnight in a humidified chamber at 4℃, followed by an incubation with Alexa 488 conjugated goat anti-rabbit IgG antibody (Cell Signaling Technology, 4412) for 2 h at room temperature. Nuclei were stained with Hoechst.

Hematoxylin and eosin (H&E)-stained tissue sections were viewed under 20X magnification of Zeiss Axiovert 40 C microscope (Zeiss, Germany). Immunohistochemical images of randomly selected fields were captured using confocal microscope (Zeiss, Germany, LSM 710); objective-Plan-Apochromat 40X/1.40 NA and quantified using Image J software.

Preparation of tumor tissue lysate for analysis of the expression of autophagy-associated markers and PAR1

Solid tumors were harvested from mice of three experimental groups, viz. tumor control, tumor treated with peptide and tumor treated with peptide and 3-MA at 15th day (one day after the first injection of peptide or peptide and 3-MA), 21st day (one day after the third injection of peptide or peptide and 3-MA) and 28th day (two days after the third injection of peptide or peptide and 3-MA). Tumor tissue lysate was prepared following the protocol described by Dignam (1990) with slight modification. Briefly, tumors were harvested from the experimental mice immediately after euthanizing them via asphyxiation in a CO₂ chamber. Tumors were frozen by flash freezing in liquid nitrogen and stored at −80℃. Before preparation of tumor tissue lysate, frozen tumor tissues were first broken and then cut into small pieces keeping on dry ice. The small pieces were homogenized by a tissue homogenizer after addition of T-PER™ Tissue protein extraction reagent (Thermo Fisher Scientific, 78510) along with protease inhibitor cocktail (Sigma-Aldrich, P8340) and 0.5 M EDTA (Invitrogen, 15575020). After sonicating the samples for 20 s keeping on ice, they were incubated on ice for 30 min. Supernatant was collected following the centrifugation of the samples at 13000 RPM for 15 min at 4℃. Protein concentration of the supernatants was estimated by BCA assay kit and further proceeded for immunoblot analysis of LAMP1, LC3 and PAR1.

Statistical analyses

All experiments were replicated at least three times (n ≥ 3). All animal experimental groups contained 5 animals each. The experimental results were presented as mean ± standard error of mean (SEM). All statistical analyses were done by applying ordinary one-way ANOVA or two-way ANOVA or Student’s t-test (Mann-Whitney test) or one sample t-test using GraphPad Prism 8. In all the graph panels, ns (non-significant) p ≥ 0.05, * (significant) 0.01 ≤ p < 0.05, ** (very significant) 0.001 ≤ p < 0.01, *** (extremely significant) p < 0.001.

List of key materials:

Reagent	Source	Identifier
Antibody
LC3B antibody (for western blot)	Cell Signaling Technology	Cat. No. 2775
LC3B (E5Q2K) Mouse mAb (for immunofluorescence)	Cell Signaling Technology	Cat. No. 83506
LAMP1 (C54H11) Rabbit mAb (for western blot)	Cell Signaling Technology	Cat. No. 3243
LAMP1 (D2D11) XP® Rabbit mAb (for immunofluorescence)	Cell Signaling Technology	Cat. No. 9091
LAMP1 (E5N9Z) Rabbit mAb	Cell Signaling Technology	Cat. No. 99437
GAPDH (14C10) Rabbit mAb	Cell Signaling Technology	Cat. No. 2118
Cyclophilin B (D1V5J) Rabbit mAb	Cell Signaling Technology	Cat. No. 43603
Cathepsin L (E3R3P) Rabbit mAb	Cell Signaling Technology	Cat. No. 55914
Cathepsin L Polyclonal Antibody	Invitrogen	Cat. No. PA5-100424
Anti-PAR-1, clone ATAP2 Antibody	Merck Millipore	Cat. No. MABF244
SQSTM1/p62 Antibody	Cell Signaling Technology	Cat. No. 5114
Phospho-mTOR (Ser2448)	Cell Signaling Technology	Cat. No. 2971
mTOR Antibody	Cell Signaling Technology	Cat. No. 2972
Phospho-AMPK alpha-1,2 (Thr172) Polyclonal Antibody	Invitrogen	Cat. No. PA5-114550
AMPKα Antibody	Cell Signaling Technology	Cat. No. 2532
Phospho-ULK1 (Ser757) Antibody	Cell Signaling Technology	Cat. No. 6888
Phospho-ULK1 (Ser317) Antibody	Cell Signaling Technology	Cat. No. 37762
ULK1 (D8H5) Rabbit mAb	Cell Signaling Technology	Cat. No. 8054
Beclin-1 Antibody	Cell Signaling Technology	Cat. No. 3738
P44/42 MAPK (ERK1/2) antibody	Cell Signaling Technology	Cat. No. 9102
Phospho-p44/42 MAPK (ERK1/2) (Thr202/Tyr204) antibody	Cell Signaling Technology	Cat. No. 9101
p38 MAPK antibody	Cell Signaling Technology	Cat. No. 9212
Phospho-p38 MAPK (Thr180/Tyr182) antibody	Cell Signaling Technology	Cat. No. 9211
SAPK/JNK antibody	Cell Signaling Technology	Cat. No. 9252
Phospho-SAPK/JNK (Thr183/Tyr185) antibody	Cell Signaling Technology	Cat. No. 9251
Ki-67 (D3B5) Rabbit mAb	Cell Signaling Technology	Cat. No. 9129
Ki-67 (D3B5) Rabbit mAb (IHC Formulated)	Cell Signaling Technology	Cat. No. 12202
Anti-rabbit IgG, HRP-linked Antibody	Cell Signaling Technology	Cat. No. 7074
Anti-mouse IgG, HRP-linked Antibody	Cell Signaling Technology	Cat. No. 7076
Anti-mouse IgG (H + L), F(ab')2 Fragment (Alexa Fluor® 488 Conjugate)	Cell Signaling Technology	Cat. No. 4408
Anti-rabbit IgG (H + L), F(ab')2 Fragment (Alexa Fluor® 488 Conjugate)	Cell Signaling Technology	Cat. No. 4412
Anti-rabbit IgG (H + L), F(ab')2 Fragment (Alexa Fluor® 555 Conjugate)	Cell Signaling Technology	Cat. No. 4413
Anti-mouse IgG (H + L), F(ab')2 Fragment (Alexa Fluor® 555 Conjugate)	Cell Signaling Technology	Cat. No. 4409
siRNA
PAR1 siRNA (CGGUCUGUUAUGUGUCUAUdTdT)	Eurogentec	N/A
Scramble siRNA	Santa Cruz Biotechnology	Cat. No. sc-37007
Peptide
PAFISED	Eurogentec	N/A
PFISED	Eurogentec	N/A
SFLLRN	Eurogentec	N/A
SFFLRN	Eurogentec	N/A
Scramble peptide (HTPPPPFISEDASGY)	Eurogentec	N/A
Plasmid
pMXs GFP-LC3-RFP	Addgene	Cat. No. 117413
pAM_1C Empty Vector	Active Motiff	Cat. No. 53023

Results

HAP-induced autophagy

Immunoblot analysis was conducted to assess the expression of autophagy-associated markers LC3II and Lysosome associated membrane protein (LAMP1) in human (MCF7, estrogen receptor positive and MDA-MB-231, triple negative) and murine (4T1, estrogen receptor negative) breast cancer cells treated with 0.25 to 1 µg/ml HAP for 30 and 60 min (Fig. 1Ai and S1Ai). Densitometric analysis revealed dose- and time-dependent increases in their expression. In MCF7 cells, optimal expression occurred with 0.5 µg/ml HAP treatment for 30 min, showing 3- and 22-fold increase for LC3II:LC3I ratio and LAMP1 expression, respectively. In 4T1 cells, LC3II:LC3I ratio and LAMP1 expression increased by 5- and 1.5-fold, respectively, with 0.75 µg/ml for 30 min as relevant dose of HAP treatment. 0.5 µg/ml dose of HAP for 30 min was optimized for MDA-MB-231 cells, which showed an increase of 2.5- and 5.6-fold in expression of LC3II:LC3I ratio and LAMP1. After 60 min, peak expression of LC3II:LC3I ratio and LAMP1 in 4T1 and MDA-MB-231 cells occurred with 0.5 and 0.25 µg/ml HAP respectively, whereas MCF7 cells showed peak LAMP1 expression with 0.25 µg/ml HAP and highest value for LC3II:LC3I ratio with 0.5 µg/ml HAP (Fig. 1Aii and S1Aii). Based on these findings, HAP doses were standardized to 0.5 µg/ml for MCF7 and MDA-MB-231 and 0.75 µg/ml for 4T1 with a uniform treatment period of 30 min for subsequent experiments.

Fig. 1.

A dose-dependent induction of autophagy by HAP in breast cancer cells. Ai Representative immunoblots of autophagy-associated markers, viz. LAMP1 and LC3 in whole cell lysate of MCF7 and 4T1 treated with different doses of HAP (0.25, 0.5, 0.75 and 1 μg/ml) for two different time points (30 min and 60 min). Aii Quantification of the protein level of LAMP1 normalized against GAPDH (in MCF7) and Cyclophilin B (in 4T1) and LC3II:LC3I ratio by densitometric analysis and comparison with untreated control. Bi Representative confocal microscopy images of immunofluorescence of LysoTracker™ in MCF7 and 4T1 cells treated with the optimized dose of HAP (0.5 μg/ml for 30 min in MCF7 and and 0.75 μg/ml for 30 min in 4T1) along with or without the pretreatment of the autophagy inhibitor 3MA. Bii Quantification of the mean fluorescence intensity of LysoTracker™ per cell and comparison with the untreated controls was presented in graphical forms. Ci Representative fluorescent images of MCF7 and 4T1 cells treated with the previously optimized doses of HAP stained using Autophagy assay kit of Abcam (ab139484) and captured using the FITC channel of a confocal microscope. Cii Quantification of the mean fluorescence intensity of FITC per cell and comparison with the control. Di Representative fluorescent images of MCF7 and 4T1 cells treated with the previously optimized doses of HAP (0.5 µg/ml for MCF7 and 0.25 µg/ml for 4T1) for 30 min and incubated with anti-cathepsin L antibody. Further, cells were incubated with Alexa 488-tagged secondary antibody and images were captured using FITC channel of a confocal microscope. Cells pretreated with Rapamycin (500 nM) for 24 h were considered as positive control. Dii Quantification of the mean fluorescence intensity of FITC per cell and compared with control. For all the graphs in this figure, significance was determined by applying one-way ANOVA

As an increase in Lamp1 expression (a marker of lysosome) was observed, we further examined lysosomal activity. Confocal fluorescence microscopy confirmed an increased lysosomal activity over HAP-treatment (Fig. 1Bi). Quantitative analysis of the mean fluorescence intensity revealed a 9.8- and 4.7-fold increase of lysosomal vacuoles in HAP-treated MCF7 and 4T1 cells respectively. However, this increase sharply reduced to just 2.6- and 1.3-fold respectively in MCF7 and 4T1, when cells pre-treated with Bafilomycin A1, an inhibitor of lysomal-acidification, treated with HAP (Fig. 1Bii). Beside lysosomal activity, an elevation in the number of Autophagosomes and Autolysosomes was also observed in confocal microscopy images (Fig. 1Ci and S4Ci), as evident from the quantification of mean fluorescence intensity (MFI), indicating 13-, 3.6- and around 9-fold increases in MCF7, 4T1 and MDA-MB-231 cells, respectively (Fig. 1Cii and S4Cii). Cells pretreated with rapamycin (500 nM) and chloroquine (60 µM) simultaneously for 24 h were considered as positive controls for autophagy detection. Flow cytometric analysis of the HAP-treated cells for autophagy detection corroborated these findings (Fig. S2Ai and Aii) revealing a significant increase in MFI (Fig. S2Aiii). However, upregulation of lysosomal activity or increase of autophagosomal vacuoles alone do not confirm the completion of the process of autophagy or an enhanced autophagic flux. Rather, these findings establish the ability of HAP to induce autophagy in breast cancer cells.

Cathepsin L is known to induce autophagy and considered to be a mediator of autophagy-induced cell death (Hsu et al. 2009; Punnonen et al. 1992; Uchiyama 2001). Immunofluorescence studies for cathepsin L expression confirmed the ablation of autolysosomal membrane in HAP-treated MCF7 and 4T1 cells at the specified doses and time points (Fig. 1Di). Quantification of fluorescence revealed 2.7- and 6.1-fold higher expression of cathepsin L in the cytosol of HAP-treated MCF7 and 4T1 cells, respectively, than in their respective controls (Fig. 1Dii).

PAR1 mediates HAP-induced autophagy

PAR1 is a G protein coupled transmembrane receptor. It is widely expressed on the membrane of most of the breast cancer cells and its expression is important for cancer cell survival, metastasis and tumorigenesis (Yang et al. 2009; Boire et al. 2005). This ubiquitous expression of PAR1 in most of the breast cancer cells renders it an ideal target for targeted therapy of breast cancer. We have earlier reported that HAP targets PAR1 in breast cancer cells and activates it non-canonically, which results into induction of apoptosis when the cells are treated with a high dose (1 µg/ml) of HAP for 2 h (Ray et al. 2016; Ray and Pal 2016). In MCF7 cells, PAR1 expression increased by 0.5 and 0.75 μg/ml HAP treatment for 30 min and by 0.25 and 0.5 μg/ml HAP treatment for 60 min. In 4T1 cells, the highest upregulation occurred by 0.75 μg/ml HAP treatment for 30 min and by 0.25 and 0.5 μg/ml HAP treatment for 60 min (Fig. 2Ai and Aii). These doses aligned with those optimized for induction of autophagy, suggesting a plausible role of PAR1 in HAP-induced autophagy.

Fig. 2.

Involvement of PAR1 in HAP induced autophagy. Ai Representative immunoblots of PAR1 using whole cell lysate of MCF7 and 4T1 treated with different doses of HAP (0.25, 0.5, 0.75 and 1 μg/μl) for two different time points (30 min and 60 min). Aii Quantification of the protein level of PAR1 normalized against GAPDH (in MCF7) and Cyclophilin B (in 4T1) and the comparison of expression with untreated control was presented graphically. Statistical significance was calculated using one-way ANOVA. Bi Confocal microscopic immunofluorescent images of MCF7 and 4T1 cells demonstrating colocalization of HAP and PAR1.HAP was detected with antisera raised against HAP and PAR1 was detected with anti-PAR1 antibody further incubated with secondary antobodies tagged with Alexa Fluor 555 and Alexa Fluor 488 respectively. Bii Quantification of colocalization of HAP and PAR1 in term of Pearson’s correlation coefficient and the comparison with the untreated control was represented in graphical form. Statistical significance was determined by applying Student’s t-test (Mann-Whitney test). Ci Representative immunoblots of PAR1 using whole cell lysate of WT MCF7 cells and siRNA mediated PAR1 KD MCF7 cells. Cii Quantification of the protein level of PAR1 normalized against GAPDH by densitometry. Di Representative immunoblots of SQSTM1/p62 and LC3 using whole cell lysate of WT and PAR1 KD MCF7 cells, both treated with two different doses of HAP for two different time points (0.5 μg/ml for 30 min and 0.25 μg/ml for 60 min). Dii Quantification of protein level of SQSTM1/p62 normalized against GAPDH and LC3II:LC3I ratio by densitometry and presented in graphical form. One-way ANOVA was applied to determine the statistical significance. Ei Representative flow cytometric images of WT and PAR1 KD MCF7 cells, both treated with different doses of HAP for two different time points (0.5 and 0.75 μg/ml for 30 min and 0.25 and 0.5 μg/ml for 60 min) and stained with LysoTracker™. Eii Quantification of the percent positive cells for LysoTracker™ staining and comparison with their respective untreated controls was presented graphically. Statistical significance was calculated by applying one-way ANOVA

To investigate the interaction between HAP and PAR1, we used confocal microscopy to assess co-localization in cells treated with optimized HAP doses for 30 min (Fig. 2Bi and S1Bi). Significant co-localization of HAP and PAR1 was noticed, with Pearson's correlation coefficients indicating approximately 30- and 18-fold increase in MCF7 and 4T1 cells respectively and 12.7-fold in MDA-MB-231 cells compared to those of their respective controls (Fig. 2Bii and S2Bii).

To assess the role of PAR1 in autophagy, we transiently knocked down PAR1 in MCF7 cells (Fig. 2Ci and Cii). HAP treatment (0.5 and 0.25 µg/ml for 30 and 60 min, respectively) of WT cells led to significant changes in autophagy-associated markers, with 3.7- and 6.1-fold decrease in SQSTM1/p62 expression and 2.1- and 1.75-fold increase in LC3II:LC3I expression ratio, respectively compared to the untreated WT cells; however, compared to the untreated PAR1 KD cells, minimal or no significant change in the expression of SQSTM1/p62 and LC3II:LC3I ratio was observed in PAR1 KD cells, treated with similar doses of HAP for similar duration (Fig. 2Di and Dii). Flow cytometry analysis indicated that PAR1 knock down reduced the rate of increase in lysosomal activity upon HAP-treatment compared to WT MCF7 cells (Fig. 2Ei). When WT cells were treated with 0.5 and 0.75 μg/ml HAP for 30 min, they exhibited 10.2- and 9.8-fold increase respectively, in percentage of LysoTracker™-positive cells, whereas 9.4- and 8.4-fold increase in percentage of LysoTracker™-positive cells was observed when the cells were treated with 0.25 and 0.5 μg/ml HAP, respectively, for 60 min. However, change in percentage of LysoTracker™-positive cells was statistically non-significant in HAP-treated PAR1 KD cells compared to the untreated PAR1 KD cells, except when they were treated with 0.75 μg/ml HAP for 30 min, which resulted in a 3.2-fold increase in percentage of LysoTracker™-positive cells (Fig. 2Eii).

in silico analysis of the interaction between PAR1 and the peptides

We have previously reported the interaction between PAR-1 and its cleaved N-terminal sequence, PFISED (generated by HAP-mediated cleavage of PAR1) a tethered ligand in murine breast cancer cells, as the underlying mechanism of HAP-induced apoptosis (Ray et al. 2018). However, the interaction between the human homolog of PAR1 and its N-terminal amino acid sequence at the similar position (PAFISED) remains unexplored.

The physicochemical properties of the peptides, such as drug likeness, pharmacokinetics and toxicity, were analyzed before docking (Table S1). Three-dimensional structures of the peptides were predicted and the top pose was used as a ligand in Autodock Vina (Fig. S3A). Interactions with the native ligand vorapaxar suggested that Val257, Tyr337, Leu258 and Ala349 were important residues for the protein-ligand complex (Fig. 3A). The parameters after re-docking vorapaxar were used to perform docking in Autodock Vina and CABS-dock was used for blind docking (Fig. 3B). The binding affinities of PAFISED and PFISED with PAR1 were examined. Re-docked Vorapaxar showed a binding affinity of −13.136 kcal/mol with hydrogen atoms of Tyr337 and Val257 of PAR1, while PAFISED and PFISED showed −6.721 and −5.160 kcal/mol binding affinities, respectively. PAFISED interacted with hydrogen atoms of Tyr337, Leu263 and Thr261, whereas PFISED interacted with those of Tyr337, Tyr95, Tyr350 and Tyr187 in top most docking pose (Fig. 3Ci and Cii). Both peptides showed considerable interaction with PAR-1 in blind docking. The cluster densities of PAFISED and PFISED were 16.34 and 29.33; average root mean square deviations (RMSDs) were 7.40 and 2.86; and maximum RMSDs were 32.88 and 28.90 with the number of elements in selected cluster being 121 and 84, respectively (Fig. S3C and D). The hydrogen bond interactions (Table S2) and hydrophobicity (Fig. S3B) of the top docked poses by CABS dock were also analyzed by Biovia Discovery Studio visualizer.

Fig. 3.

Docking analysis of peptide against the PAR1 receptor. A Crystal structure of PAR1 (3VW7) with Vorapaxar with their binding interaction. B Result of re-docked Vorapaxar with binding interaction and superimposed pose of re-docked Vorapaxar (in yellow) with binding parameters that used in Autodock Vina. Docking Score = −13.136 kcal/mol (C) 2-D structure of peptide (Ci) PAFISED and (Cii) PFISED and hydrogen bond donor interaction of docked peptide with SAS view and their binding interaction of PAFISED AND PFISED respectively

Noncanonical activation of PAR1 by HAP regulates the induction of autophagy

Thrombin activates PAR1 in a canonical way by cleaving at a particular site of PAR1 and leaving a new N-terminal end (Vu et al. 1991; Nieman and Schmaier 2007). To compare between thrombin-mediated canonical PAR1 activation and noncanonical activation of PAR1 by HAP regarding their potential of inducing autophagy, we analyzed LC3II:LC3I ratio and LAMP1 expression in MCF7 and 4T1 cells. No significant upregulation of these markers was observed in thrombin-treated cells compared to that in the control. Conversely, HAP-treated MCF7 cells showed 16.8- and 6.6-fold increases in LC3II:LC3I ratio and LAMP1 expression, respectively, while 4T1 cells exhibited 8.65- and 3-fold increase in LC3II:LC3I ratio and LAMP1 expression, respectively. Moreover, we checked the upstream regulator of autophagy i.e. mTOR. HAP treatment downregulated phosphorylated mTOR (p-mTOR) expression by 4.2- and 6.8-fold in MCF7 and 4T1 cells, respectively, consistent with the notion that mTOR activation negatively regulates autophagy. However, the increase in mTOR phosphorylation in either of the thrombin-treated cells was non-significant (Fig. 4Ai and Aii).

Fig. 4.

Molecular signaling mechanism of the peptide induced autophagy. Ai Immunoblots of LC3, LAMP1, p-mTOR and mTOR in whole cell lysates of MCF7 and 4T1 treated with HAP (0.5 and 0.75 μg/ml in MCF7 and 4T1 respectively) for 30 min and thrombin (2 nmol/L for 2 h). Aii Densitometric analysis of the protein expression of LC3II:LC3I, p-mTOR/mTOR and LAMP1 normalized against GAPDH in MCF7 and against Cyclophilin B in 4T1 and the comparison with the untreated control was graphically presented. Bi Immunoblots of SQSTM1/p62 in the whole cell lysates of MCF7 and 4T1, where MCF7 cells were treated with the synthetic peptides PAFISED (50 μM), SFLLRN (50 μM) and scramble peptide HTPPPPFISEDASGY (50 μM) and 4T1 cells with synthetic peptides PFISED (75 μM), SFFLRN (75 μM) and scramble peptide HTPPPPFISEDASGY (75 μM) for 4 h. Bii Quantitative analysis of the expression of SQSTM1/p62 normalized against GAPDH (in MCF7) and Cyclophilin B (in 4T1) and comparison with the respective untreated control was presented in graph form. Ci Immunoblots of LC3, LAMP1 and SQSTM1/p62 in whole cell lysates of MCF7 and 4T1 treated with different doses (5, 25, 50, 75 and 100 μM) of both the peptides PAFISED and PFISED for 4 h. Cii Densitometric analysis of the expression of LC3II:LC3I and LAMP1 & SQSTM1/p62 normalized against GAPDH (in MCF7) and Cyclophilin B (in 4T1) and comparison with the respective untreated control was presented in graph form. Di Representative immunofluorescence images of MCF7 and 4T1 cells treated with the peptide PAFISED (50 μM) and PFISED (75 μM) respectively for 4 h and stained using dyes included in the Autophagy assay kit (Abcam, ab139484) and observed in a confocal microscope under FITC channel. Dii Quantification of the Mean fluorescence intensity of green fluorescence (FITC) per cell and graphical presentation of the comparison with the untreated control. Ei Immunoblots of different signaling molecules of autophagy including p-mTOR, mTOR, p-AMPK, AMPK, p-ULK1 (Ser757), p-ULK1 (Ser317), ULK1 and Beclin1 in the whole cell lysate of MCF7 treated with different doses of the peptide PAFISED, ranging from 5 μM to 100 μM. Eii Quantification of p-mTOR/mTOR, p-AMPK/AMPK, p-ULK1 (Ser757)/ULK1, p-ULK1 (Ser317)/ULK1 and Beclin1 normalized against GAPDH by densitometric analysis and graphical presentation of the comparison with the untreated control. Fi Immunoblots of the three MAP Kinase molecules: ERK1/2, p38 and JNK and their phosphorylated forms in the whole cell lysate of MCF7 treated with different doses of the peptide PAFISED, ranging from 5 μM to 100 μM. Fii Quantification of the expression of p-ERK1/2, p-p38 and p-JNK normalized with ERK1/2, p38 and JNK respectively and graphical presentation of the comparative analysis with the untreated control. One-way ANOVA was applied to determine the statistical significance for all the graphs in this figure

Potential of the synthetically designed peptides in inducing autophagy

Earlier, we have demonstrated that HAP-mediated cleavage of PAR1 exposes the PFISED sequence, which further binds to PAR1 as a tethered ligand (Ray et al. 2018). The human PAR1 homolog (UniProt ID: P25116) possesses a similar sequence, PAFISED, with an additional Ala residue between Pro and Phe at the N-terminal end. in silico analysis has shown that PAFISED can bind to PAR1. The synthetic hexapeptide SFLLRN mimics the effect of thrombin on PAR1 without requiring proteolytic cleavage (Chen et al. 1994). To explore the autophagic potential of PAFISED and PFISED, we assessed their effects on the expression of LC3, LAMP1 and SQSTM1/p62. SQSTM1/p62 acts as an autophagy adapter protein, which carries the substrates of autophagy to the autolysomal vacuoles and subsequently gets degraded (Bjorkoy et al. 2005). Optimal induction of autophagy was observed with 50 μM PAFISED in MCF7 cells, 75 μM PFISED in 4T1 cells and 25 μM PAFISED in MDA-MB-231 cells. (Fig. 4Ci and S4Ai). Densitometric analysis revealed a 2.47- and 2.52-fold increase in the LC3II:LC3I ratio and a 3.35- and 3.23-fold increase in LAMP1 expression in MCF7 and 4T1 cells respectively. SQSTM1/p62 expression decreased by 4.4- and 2.76-fold in MCF7 and 4T1 cells, respectively (Fig. 4Cii). In MDA-MB-231 cells, LC3II:LC3I ratio and LAMP1 expression increased by 4- and 1.76-fold respectively (Fig. S4Aii). Comparisons of PAFISED and PFISED with the thrombin-mimicking peptides SFLLRN (human) and SFFLRN (mouse) respectively and with a scramble peptide showed a slightly more than 4-fold reduction in SQSTM1/p62 expression in both MCF7 and 4T1 cells, treated with PAFISED and PFISED respectively and 1.76-fold reduction in MDA-MB-231 cells, treated with PAFISED compared to their respective controls. No significant change was noted in cells treated with SFLLRN, SFFLRN, or scramble peptide (Fig. 4Bi, Bii, S4Bi and Bii). Induction of autophagy was further confirmed by confocal microscopy (Fig. 4Di, S4Ci), showing significant increases in AP and AL vacuoles in peptide-treated cells accounting for 11.7-, 13.3- and 15.4-fold increase in MFI per cell in MCF7, 4T1 and MDA-MB-231 cells, respectively (Fig. 4Dii, S4Cii). Flow cytometric analysis showed similar results (Fig. S2Ai and Aii) indicating significant increase of MFI in peptide-treated cells (Fig. S2Aiii).

Additionally, no significant induction of autophagy was observed in normal breast epithelial cells (MCF10A), which showed reduced PAR1 expression compared to that in MCF7 cells and LC3 and LAMP1 expression was almost negligible (Fig. S5C).

Predictive signaling pathway of peptide-induced autophagy

Further we checked the upstream regulators of autophagy. Expression of various markers of the AMPK-ULK1-BECN1 pathway of autophagy was assessed by immunoblotting. Immunoblotting revealed that treatment of MCF7 cells with 50 µM PAFISED significantly reduced p-mTOR levels by 16.2-fold, while phosphorylated AMPK levels increased by 6.5-fold compared to that of the control. ULK1 phosphorylation showed a 1.5-fold decrease in p-ULK1^Ser757 and a 2.9-fold increase in p-ULK1^Ser317(active form), suggesting AMPK-mediated ULK1 activation, outperforming mTOR-mediated phosphorylation of ULK1 at Ser757 (Kim et al. 2011). Notably, Beclin 1 (BECN1) expression increased by 3.5-fold, suggesting the possibility of the recruitment of BECN1 to form BECN1-VPS34-VPS15 core complex that is instrumental for the nucleation phase of autophagy (Fig. 4Ei and Eii).

The three MAP kinases, ERK1/2, JNK and p38, function as critical upstream regulators of mTOR signaling. These kinases are, in turn, modulated by G protein-mediated signaling cascades initiated upon activation of PAR1 receptor (Rohani et al. 2010). So, they serve as key molecular intermediaries linking non-canonical PAR1 activation to the observed attenuation of mTOR phosphorylation. Comprehensive analysis of these MAP kinase components within the PAR1 signaling axis demonstrated a pronounced suppression of their kinase activity, evidenced by greater than 4-fold reductions in the phosphorylation levels of ERK1/2 and JNK, alongside a 1.7-fold decrease in p38 phosphorylation (Fig. 4Fi and Fii). These findings delineate a plausible mechanistic framework underlying PAR1-mediated downregulation of mTOR activation by PAFISED.

However, a complete mechanistic resolution of this signaling cascade will require more detailed study of the molecular determinants involved in PAR1-mediated modulation of mTOR activity.

Potential of HAP and the peptides in forming AL

To investigate the final stages of autophagy i.e. AP-lysosome fusion, we used pMXs GFP-LC3-RFP plasmid as an autophagic flux reporter. Following transfection, endogenous ATG4 proteins cleaved the plasmid, producing GFP-LC3 and red fluorescent protein (RFP) in equimolar amounts. Upon AP-lysosome fusion, GFP-LC3 is quenched in the acidic environment, whereas RFP fluorescence remains intact. An increased RFP:GFP ratio indicates autophagic flux (Morita et al. 2018). In MCF7, 4T1 and MDA-MB-231 cells treated with HAP, confocal microscopy revealed 3.3-, 5.6- and around 2-fold increase in the RFP:GFP ratio, respectively, while PAFISED treatment showed 8.7- and 2.8-fold increase in MCF7 and MDA-MB-231 cells respectively and PFISED treatment resulted in 2.9-fold increase in 4T1 cells (Fig. 5Ai, Aii and S4Di, Dii).

Fig. 5.

Formation of autolysosome and analysis of autophagic flux. Ai Representative confocal microscopic immunofluorescent images of MCF7 and 4T1 cells transfected with the plasmid construct pMXs GFP-LC3-RFP and treated with HAP (0.5 and 0.75 μg/ml for MCF7 and 4T1 respectively) for 30 min and with the peptides (50 µΜ of PAFISED for MCF7 and 75 μM of PFISED for 4T1) for 4 h. pAM_1C empty vector transfected cells were kept as vehicle control. Aii Ratio of the mean of numbers of red puncta (RFP) and green puncta (GFP in non-acidic enviroment) per cell was measured and the comparison with the vehicle control was graphically presented. Bi Immunofluorescent images of MCF7 and 4T1 treated with peptides (PAFISED of 50 μM and PFISED of 75 μM for MCF7 and 4T1 respectively) and stained against LAMP1 and LC3 with anti-LAMP1 and anti-LC3 antibodies and incubated them with their respective secondary antibodies tagged with AlexaFluor 555 and AlexaFluor 488 respectively. Bii Colocalization of AlexaFluor 555 tagged LAMP1 and AlexaFluor 488 tagged LC3 was measured in term of Pearson’s correlation coefficient and the comparison with the untreated control was graphically presented. For all the graphs in this figure, one-way ANOVA was used to determine the statistical significance. C Representative transmission electron microscopic (TEM) images of MCF7 and 4T1 treated with HAP (0.5 μg/ml for MCF7 and 0.75 μg/ml for 4T1) for 30 min and with peptides (50 μΜ of PAFISED for MCF7 and 75 μM of PFISED for 4T1) for 4 h. Autophagosomal structures (AP) and autolysosomal structures (AL) are marked with arrows

To validate our findings, we assessed the colocalization of LC3 and LAMP1 in peptide-treated MCF7, 4T1 and MDA-MB-231 cells using confocal microscopy (Fig. 5Bi and S4Ei). The Pearson’s correlation coefficient for colocalization increased by 15.5-, 4.7- and 3.2-fold in MCF7, 4T1 and MDA-MB-231 cells, respectively (Fig. 5Bii and S4Eii), indicating the fusion of AP with lysosome. Transmission electron microscopy further confirmed AL formation, characterized by multiple dark patches indicative of degraded cargo inside the autolysosomal vacuoles in HAP- and peptide-treated MCF7 and 4T1 cells (Fig. 5C). However, we also observed some AP structures containing intact cellular organelles in HAP- and peptide-treated 4T1 cells.

Determination of cancer stage of malignant breast tumors developed using 4T1 cells in BALB/c mice

The effect of peptide-induced autophagy on breast cancer cells was validated in vivo using a BALB/c mouse model of low-grade malignant breast tumor developed by injecting 4T1 cells. Tumor progression, histologically assessed weekly over 4 weeks, revealed tumors from weeks 1-3 as Grade I (highly differentiated) and from week 4 as Grade II (moderately differentiated) (Fig. S6D and Table S3). Tumor grading, based on the Nottingham system (van Dooijeweert et al. 2022), was correlated with Ki-67 expression. Immunohistochemical analysis revealed a substantial increase in Ki-67 expression in tumors from week 4 compared to that in tumors from weeks 1-3 (Fig. S6Ei and Eii), which supports the tumor grading based on histopathological analysis. Tumor volume highly increased in weeks 3 and 4 compared to that in weeks 1 and 2 (Fig. S6Ci and Cii). Mice showed a slight and non-significant decrease in weight until week 3, followed by a slight increase in weight in week 4 (Fig. S6B).

Determination of peptide dose in BALB/c mice

Peptide dose was optimized by assessing peptide-induced toxicity for 2 weeks at an interval of 3 days (Fig. S7A). Toxicity was evaluated by measuring different body parameters of mice, including differential blood counts, hemoglobin level, alanine aminotransferase (ALT) and aspartate transaminase (AST) activities, blood urea nitrogen (BUN) and creatinine levels (Fig. S7B). In a preliminary experiment, the maximum tolerated dose was identified as 20 mg/kg body weight. Beyond this dose, one out of the five mice in a group died and three started showing visible symptoms of discomfort and itching within 7 days from the time of first administration of peptide. In subsequent experiments, mice administered with doses up to 5 mg/kg body weight showed no notable hepatic toxicity such as increase in ALT and AST activities. Similar results were observed for nephrotoxicity as the increase in creatinine and BUN levels were non-significant up to the dose of 5 mg/kg body weight. However, hemoglobin levels were significantly altered starting from the dose of 2.5 mg/kg body weight and the lymphocyte count was altered from the dose of 1 mg/kg body weight, when compared to those of the control. Eosinophil, monocyte and neutrophil counts exhibited no significant alteration up to the dose of 5 mg/kg body weight, consistent with the results of hepatotoxicity and nephrotoxicity (Fig. S7B). Therefore, 5 mg/kg body weight dose of peptide PFISED was optimized for further experiments.

Peptide-induced autophagy improves survival rate of BALB/c mice carrying low-grade malignant solid breast tumors developed by 4T1 cells

Mice with 2-week-old tumors were used as an early-stage breast cancer model. Untreated mice survived for 45 days on average, whereas PFISED treatment increased the average survival to 96 days. Notably, co-treatment of PFISED and autophagy inhibitor 3-MA reduced the average survival to 54 days. Tumor-free mice lived for approximately 172 days (Fig. 6B).

Fig. 6.

in vivo analysis of the impact of peptide induced autophagy on low-grade malignant breast tumor. A Schematic representation of the in vivo experimental design. B Comparative survivability percentage of four sets of BALB/c mice, viz. (a) control: mice without tumor, (b) tumor control: mice bearing 4T1 induced tumor but without any treatment, (c) tumor + peptide: tumor-bearing mice treated with the peptide PFISED (5 mg/kg body weight) periodically at an interval of 3 days from day 14 to day 26 after the administration of 4T1 cells, (d) tumor + 3-MA + peptide: tumor-bearing mice treated with the peptide PFISED as described before with a simultaneous treatment of autophagy inhibitor 3-MA. Ci Images of the solid tumors from the aforementioned three groups of mice, viz. tumor control, tumor + peptide and tumor + 3-MA + peptide. Cii A comparative dynamic analysis of the volume of the tumors from these three groups of mice. Linear regression was used to determine the pattern of change in tumor volume at different time intervals. (D) Histological images of tumors from these three groups of mice. Images were captured using inverted light microscope with 20X (first image of every row) and 40X magnification (second image of every row). Glandular/tubular differentiations are marked with purple arrow in the first image of every row and nuclear pleomorphisms are marked with green arrow in the second image of every row. Ei Immunofluorescence images of the expression of the proliferation marker Ki-67 in the tumors of aforementioned three groups of mice. Tumor samples were immunohistochemically analyzed with anti-Ki-67 antibody, further incubated with AlexaFluor 488 tagged secondary antibody and images were captured under FITC channel of a confocal microscope. Nuclei were stained with Hoechst 33342. Eii Graphical presentation of comparative analysis of the expression of Ki-67 in all three tumor sets. One-way ANOVA was applied to calculate the statistical significance. Fi Schematic representation of the study design to analyse and compare LC3II:LC3I ratio, expression of LAMP1 and PAR1 in tumor tissue from different experimental groups. Fii Immunoblots of LC3, LAMP1 and PAR1 in tissue lysate of tumors, harvested from mice of all the three tumor-bearing experimental groups at different times (15th day, 21st day and 28th day). Fiii Quantification of LC3II:LC3I ratio, the expression of LAMP1 and PAR1 normalized against GAPDH by densitometric analysis and graphical presentation of the comparison with the untreated tumor control. Statistical significance was evaluated by applying one-way ANOVA

Inhibition of tumor growth, progression and proliferation of tumor cells by peptide-induced autophagy at early stages of breast cancer in BALB/c mice

PFISED treatment in tumor-bearing mice resulted in a 4.4-fold reduction of tumor volume in 4 weeks, compared to that of the tumor control. Conversely, co-treatment of PFISED and 3-MA resulted in only 2.2-fold reduction in tumor volume compared to that of the tumor control. Dynamic monitoring of the tumor volume revealed that tumor volume remained stable after first 3 doses of peptide in the peptide-treated mice, while in tumor control mice and mice co-treated with peptide and 3-MA, the tumor volume kept increasing upto 28th day. (Fig. 6Ci and Cii). Histopathological grading revealed intermediate-grade (II) malignancy for the tumor control and of tumor-bearing mice co-treated with PFISED and 3-MA, whereas in tumor-bearing mice treated with PFISED, low-grade (I) malignancy was noticed (Fig. 6D and Table 1). Ki-67 expression decreased by 10-fold in the PFISED-treated group, whereas it showed only 1.26-fold reduction by co-treatment of PFISED and 3-MA, compared to that of the tumor control group (Fig. 6Ei and Eii).

Table 1.

Analysis of the grade of tumors of different experimental groups

Tumor sample	Ductal/tubular/glandular differentiation	Nuclear pleomorphism	Mitotic figure	Cumulative score	Grade of tumor
Control	2	3	1	6	II
Tumor treated with 3-MA and peptide PFISED	2	2	1	5	II
Tumor treated with peptide PFISED	1	1	1	3	I

Scores of the experimental sets of tumors, assigned upon histopathological analysis of the three parameters, viz. Ductal/tubular/glandular differentiation, Nuclear pleomorphism and Mitotic figure and gradation of tumors based on their cumulative scores following Nottingham Grading System

Time-lapsed expression pattern of autophagy-associated markers and PAR1 in tumor tissues of different experimental groups

Expression of autophagy-associated markers in tumor tissue alters variably with time in different experimental groups. Analysis of two such markers, LAMP1 and LC3 in the tumor tissue at 15th day (one day after the first dose of peptide) revealed an increase of expression of LAMP1 and LC3II:LC3I ratio by 1.5- and 1.2-fold respectively compared to the untreated tumor. Eventually at 21st day (one day after the third dose of peptide), their respective expression increased by 2.1- and 1.5-fold. Interestingly, at 28th day (two days after the fifth dose of peptide), expression of LAMP1 decreased by 1.3-fold in the peptide-treated tumor tissue compared to the untreated one. However, no statistically significant change was observed in the expression of LC3II:LC3I ratio. In the tumor obtained from the mice, treated with peptide along with 3-MA, no statistically significant change in the expression of LAMP1 and LC3II:LC3I ratio was observed at 15th day. But, expression of both the markers declined in tumor tissue obtained from this group at 21 st and 28th day. While expression of LAMP1 reduced by 1.5- and 3-fold, LC3II:LC3I ratio showed a decline of 1.9- and 2.4-fold at 21 st and 28th day respectively. No significant alteration was found in PAR1 expression among tumor tissues obtained from all the three groups at 15th day. Significantly, at 21 st day, PAR1 expression increased by 1.5-fold in the peptide-treated tumor tissue compared to the untreated one. However, the tumor tissue obtained from mice, treated with peptide along with 3-MA, represented no significant alteration in the expression of PAR1 at 21 st day as well. At 28th day, the expression of PAR1 reduced by 1.6- and 1.4-fold respectively in the peptide-treated tumor and in the tumor, treated with peptide along with 3-MA (Fig. 6Fii and Fiii).

Autophagy-driven cell death by HAP and the peptides

A reduced Ki-67 expression indicated a decreased proliferation of MCF7 and 4T1 cells after HAP and peptide treatments. However, prior to HAP and peptide treatment, treatment with 3-MA (a broad inhibitor of autophagy) restored the HAP- and peptide-induced decline in Ki-67 expression (Fig. S5Ai and Aii).

In MCF7 cells, HAP and PAFISED treatments reduced survival by approximately 60% and 50%, respectively (Fig. S5Bi), while the survival of 4T1 cells decreased by approximately 46% and 55%, respectively, by HAP and PFISED treatment (Fig. S5Bii). Autophagy inhibition by 3-MA treatment prior to HAP or, peptide treatment increased cell survival.

Autophagy-driven Immunomodulation by the peptides

Treatment of MCF7, 4T1 and MDA-MB-231 cells with the optimized peptide dose required for induction of autophagy elicited a pronounced immunomodulatory response, characterized by enhanced immunostimulation alongside attenuation of immunosuppressive signaling components. Quantitative ELISA analysis of the supernatant collected from the peptide-treated MCF7 and 4T1 and MDA-MB-231 cells revealed a substantial alteration of key cytokines. Specifically, Interleukin-2 (IL-2) levels increased by 3.8-fold, 4.1-fold and approximately 5-fold, Interferon alpha (IFN-α) levels rose by 2.9-fold, 2.2-fold and 1.5-fold and TGF-β levels increased by approximately 6-fold, 3-fold and 2.9-fold in MCF7, 4T1 and MDA-MB-231 cells, respectively. In contrast, a marked reduction in Granulocyte-macrophage colony-stimulating factor (GM-CSF) secretion and Interleukin-10 (IL-10) secretion was observed, with decreases of 2.7-, 3.5- and 3.9-fold in GM-CSF and 3-, 4.5- and 4.7-fold in IL-10 secretion across the respective cell lines (Fig. S5D).

Collectively, these findings indicate that peptide-induced autophagy promotes a shift toward an immunostimulatory and anti-immunosuppressive tumor microenvironment. This immunological reprogramming may contribute to tumor regression and potentiate the anti-proliferative effects associated with peptide-mediated induction of autophagy in breast cancer cells.

Discussion

Autophagy, long regarded as a fundamental cellular self-survival mechanism, has recently reported to facilitate cell death as well as termed as type II programmed cell death (Liu et al. 2023; Green 2019; Hou et al. 2016). This new approach has led to the exploration of autophagy as a therapeutic target for various diseases, including cancer and neurodegenerative conditions, which requires selective elimination of specific cell populations as a treatment approach. Autophagy can be induced by different stimuli, indicating the presence of multiple upstream signaling pathways that may govern autophagy independently or in an intertwined manner (Yin et al. 2016). In this study, we investigated the potential of two peptides, designed by mimicking the N-terminal end of human and murine homolog of PAR1, noncanonically activated by a secreted metalloprotease, HAP, to induce autophagy selectively in breast cancer cells, alongside the therapeutic exploitation of the peptide, designed from murine homolog of PAR1 in low-grade malignant breast tumor in BALB/c mice. Although substantial progress has been achieved in the therapeutic management of ER+ breast cancer, the majority of available agents fail to elicit a robust antitumorigenic response in ER− counterparts and most importantly in triple negative (ER−/PR−/HER2−) breast cancer cells. To address this therapeutic gap, we sought to delineate the autophagy-inducing potential of the protease and peptide in triple negative (4T1 and MDA-MB-231) as well as ER+/PR+/HER2− (MCF7) breast cancer cells.

HAP, owing to its proteolytic activity, engages with PARs on the membranes of various mammalian cancer cells. HAP mediates the alternative or noncanonical activation of PAR1, a key mechanism driving HAP-induced apoptosis. Among the four known subtypes of PARs, HAP specifically binds to PAR1 and activates it (Ray and Pal 2016). In the present study, HAP was shown to induce autophagy in breast cancer cells at lower concentrations than those previously optimized for apoptosis induction (Ray et al. 2016). A single stimulus can simultaneously regulate apoptotic and autophagic pathways, with these processes often acting as mutually exclusive events (Lockshin and Zakeri 2004; Thorburn 2008). We identified that noncanonical activation of PAR1 by HAP was pivotal for the induction of autophagy. When apoptosis and autophagy are triggered by a common stimulus, autophagy is typically initiated as an early response, with apoptosis being recruited at later time points. In some instances, autophagy may even lead to apoptosis (Lockshin and Zakeri 2004; Nikoletopoulou et al. 2013). While apoptosis has long been known to be exploited to treat breast cancer, inhibiting autophagy has also become an effective mode of treatment in high-grade malignant breast tumor (late-stage breast cancer). Since autophagy plays a bipartite role in cancer, inducing autophagy and autophagy mediated programmed cell death has also emerged as a viable treatment approach to treat low-grade malignant breast tumor (early-stage breast cancer) (Jain et al. 2013).

PAR1 is a transmembrane G protein-coupled receptor, which, upon activation, interacts with various G protein subtypes, including G12/13, Gi and Gq, depending on the cellular context and modulates different signaling pathways (Vouret-Craviari et al. 1998; Taylor et al. 1991; Offermanns et al. 1994; Klages et al. 1999; Hung et al. 1992). Although, PAR1 expression is minimal in most noncancerous mammalian cells (with the exception in platelets), it is abundantly expressed in breast cancer cells, making it an ideal target for therapeutic intervention (Boire et al. 2005). Thrombin, a serine protease, is a known activator of PAR1. It interacts with the transmembrane receptor, PAR1 through its N-terminal exodomain, contains two thrombin-binding sites (Nieman and Schmaier 2007). Thrombin activates PAR1 by cleaving the Arg 41/Ser 42 site of its exodomain (Vu et al. 1991; Chen et al. 1994). Upon activation, G protein-coupled receptors activate MAP kinases by signaling through G proteins and facilitate proliferation by mTOR activation (Lopez-Ilasaca 1998). This canonical pathway of PAR1 activation by thrombin also escalates proliferation via MAPK activation and leads to activation of mTOR (Parrales et al. 2013; Wang et al. 2002),which is considered as a negative regulator of autophagy (Jung et al. 2010). Our findings re-established the notion that mTOR activation was consistent with thrombin-induced canonical activation of PAR1. However, we observed that HAP mediated noncanonical activation of PAR1 downregulated mTOR activation, in contrast to the effects of thrombin mediated canonical activation. Moreover, our study found that thrombin did not induce significant autophagic responses, whereas HAP demonstrated a marked ability of inducing autophagy. To the best of our knowledge, this is the first report that noncanonical PAR1 activation by HAP induces autophagy, as opposed to its canonical thrombin-mediated activation, which primarily upregulates mTOR activation that leads to cell proliferation and inhibits autophagy. Whereas our study established PAR1 as a key mediator of HAP-induced autophagy, from our existing data, we cannot rule out the existence of any other upstream stress or co-signaling pathways that may also contribute toward induction of autophagy.

Thrombin activates the human PAR1 homolog by proteolytically cleaving its N-terminal extracellular domain, which exposes a sequence (SFLLRN) that intramolecularly binds PAR1 as a tethered ligand, initiating downstream signaling in platelets and vascular endothelial cells (Vu et al. 1991). The murine homolog of PAR1 contains a similar sequence (SFFLRN) with the alteration of one amino acid (Phe44 instead of Leu44) at the same position. HAP cleaves the murine homolog of PAR1 at the Pro90/Pro91 site, thereby generating a new N-terminal end with the sequence PFISED (Ray and Pal 2016). This sequence is analogous to PAFISED in the human PAR1 homolog, located between Leu84 and Ala92. Both PFISED and PAFISED, designed on the basis of last seven and six amino acids of the N-terminal cleaved ends of human and murine homologs of PAR1, respectively, possess high affinity for their respective PAR1 homologs, as shown by molecular docking studies. The docking result and free energy values suggest that these peptides are stable and can optimally bind to PAR1. The molecular docking analysis also suggested that hydrogen bonds contribute to the binding specificity and structural stability in protein-ligand binding. However, exact binding affinity cannot be conclusively determined form molecular docking alone. Previous studies have demonstrated that ligand binding is controlled by combination of non-covalent interactions, electrostatic interactions, hydrogen bonds and hydrophobic contacts (Pantsar and Poso 2018; Chen et al. 2020). Therefore, the interactions observed in this study also required further validation through experimental approaches.

The peptides PAFISED and PFISED efficiently induced autophagy in human and murine breast cancer cells respectively, in a dose-dependent manner, whereas SFLLRN and SFFLRN failed to do so in breast cancer cells of human and murine origin respectively. A dose of 50 µM PAFISED in MCF7 cells, a dose of 75 µM PFISED in 4T1 cells and a dose of 25 µM PAFISED in MDA-MB-231 cells for a treatment time of 4 h were optimized for induction of autophagy. Further investigation into the plausible signaling pathways revealed a significant downregulation of mTOR activation and upregulation of AMPK activation as evidenced by respectively a decrease and an increase in the expression of their respective phosphorylated forms. Whereas mTOR is known as a negative regulator of autophagy, AMPK, upon activation, initiates the process of autophagy by phosphorylating ULK1 at Ser317 and Ser777 residues, which are pivotal for the formation of initiation complex of autophagy. On contrary, mTOR, when activated, inhibits the formation of initiation complex by phosphorylating ULK1 at Ser757 residue. A substantial increase in p-ULK1^Ser317 expression and decrease in p-ULK1^Ser757expression and a subsequent increase in BECN1 expression in the peptide treated MCF7 cells speculated a plausible existence of AMPK-ULK1-BECN1 axis behind the peptide mediated induction of autophagy. In some cancer cells, a decrease in mTOR signaling can lead to the activation of autophagy, a process that breaks down cellular components and can lead to cell death (Kim et al. 2011). While hyperactivation of mTOR is often linked to drug resistance (Dong et al. 2021), some evidences suggest that mTOR signaling can regulate the expression of matrix metalloproteinases (MMPs) which are involved in metastasis. Inhibiting mTOR can reduce the expression of MMP9, leading to the suppression of cancer metastasis (Jiang et al. 2015).

However, further detailed study is required to establish the specific molecular mechanism of this peptide-mediated induction of autophagy in human breast cancer cells. As the autophagy-inducing potential of these peptides that resembled the cleaved N-terminal end of noncanonically activated PAR1 was established, it was proven that the induction of autophagy was not HAP-specific, rather it depended on the site of cleavage on the PAR1 exodomain.

A major limitation of many autophagy-inducing agents is their inability to form ALs or effectively degrade cargo, an event necessary for autophagy-mediated cell death. Our study confirmed that treatments with both HAP and peptides resulted in the formation of ALs in breast cancer cells, which was evidenced from a significant increase in RFP:GFP ratio in HAP- and peptide-treated cells and an elevation in LAMP1-LC3 colocalization in peptide-treated cells. Formation of AL in both HAP- and peptide-treated cells was further confirmed by transmission electron microscopic images. Degradation of cargoes is marked by the rupture of autolysosomal structure, which was evidenced by the elevated expression of Cathepsin L in cytosol of the HAP-treated MCF7 and 4T1 cells. A downregulation of SQSTM1/p62 expression in both HAP- and peptide-treated cells also confirmed autolysosomal degradation of cargoes.

In the context of cancer therapy, it is essential to confirm that induction of autophagy leads to cessation of cellular proliferation and a subsequent reduction in cancer cell survival. Our data showed a significant decline in both proliferation and survival of breast cancer cells upon induction of autophagy by HAP and the peptides. Proliferation and survival were restored when autophagy was inhibited by 3-MA. Although, a basal level of autophagy is always present in cancer cells, HAP- and peptide-driven cell death was plausibly due to over-activation of autophagy. Notably, breast cancer cells, which exhibit high expression of PAR1, were selectively targeted by HAP and the peptide for induction of autophagy, while noncancerous human breast epithelial cells, MCF10A, that lacks any significant PAR1 expression, remained unaffected in term of autophagy by either of HAP and the peptide PAFISED.

We also assessed the in vivo effect of peptide-induced autophagy in low-grade malignant breast tumor. Using the Nottingham Grading System, we classified tumors, based on the scores attributed to them, considering three aspects of the tumor tissue, viz. ductal/tubular/glandular differentiation, nuclear pleomorphism and mitotic figure and correlated them with Ki-67 expression. A cumulative score between 3-5 refers to a low-grade/ grade I (well-differentiated) breast tumor, whereas a cumulative score between 6-7 indicates an intermediate-grade/ grade II (moderately differentiated) breast tumor and a cumulative score between 8-9 signifies a high-grade/ grade III (poorly differentiated) breast tumor in term of malignancy (van Dooijeweert et al. 2022). Our findings revealed that malignant breast tumors developed in BALB/c mice by 4T1 cells remained well-differentiated (Grade I) up to 3 weeks. Following PFISED treatment, we observed significant inhibition of tumor growth and progression along with a partial cessation of increasing proliferation rate in this low-grade malignant breast tumor model. Additionally, the peptide was safe to be administered up to the dose of 5 mg/kg body weight. However, further studies on the physicochemical properties of the peptide and large-scale in vivo experiments in higher order animals including primates are required before advancing to clinical trials.

In spite of the fact that use of 3-MA as inhibitor of autophagy is well established and supported by several studies (Petiot et al. 2000; Caro et al. 1988; Blommaart et al. 1997), it’s use as inhibitor of autophagy in various experiments of the current study, cannot rule out the possibility of affecting other signaling pathways dependent on Phosphoinositide 3-kinase (PI3K). Since, 3-MA is a broad range inhibitor of PI3K, along with class III PI3K, that facilitates the nucleation step of autophagy, it also inhibits other classes of PI3K that may modulate other related signaling pathways. Notably, the in vivo use of 3-MA, as inhibitor of autophagy, also leaves the space of other PI3K-dependent signaling pathways being modulated. So further study, involving genetic inhibition of autophagy by modulating key autophagy-associated gene, needs to be carried on to selectively inhibit autophagy.

Although our study demonstrated that noncanonical activation of PAR1 inhibits tumor growth, a significant limitation lies in the lack of investigation into its effects on platelet activation. Considering the role of canonical PAR1 activation by thrombin in platelet activation, assessment of the impact of this noncanonical activation of PAR1 on platelets is imperative to rule out any potential thrombotic complication prior to any consideration of clinical translation. As this aspect was beyond the purview of the current study, we intend to address it in future research.

Conclusion

In conclusion, we report for the first time that contrary to the canonical activation of PAR1 by thrombin, a noncanonical activation of PAR1, irrespective of the activating agent, downregulates mTOR activation and subsequently induces autophagy in breast cancer cells to inhibit tumorigenesis in low-grade malignant tumors. In addition, this study opens new avenues for a peptide-mediated therapeutic strategy aimed at leveraging autophagy as a treatment modality for breast cancer, specifically at the early stage.

Abbreviations

3-MA

3-Methyladenine

4T1

410.4 Subpopulation Tumor 1

AL

Autolysosome

ALT

Alanine transaminase

AP

Autophagosome

AST

Aspartate aminotransferase

BUN

Blood urea nitrogen

EAC

Ehrlich ascites carcinoma

EDTA

Ethylene diamine tetraacetic acid

ER+ 

Estrogen receptor positive

ER−

Estrogen receptor negative

GM-CSF

Granulocyte-macrophage colony-stimulating factor

HAP

Hemagglutinin protease

KD

Knocked down

MCF10A

Michigan cancer foundation-10A

MCF7

Michigan cancer foundation-7

MEM

Minimum essential medium

PAR1

Protease activated receptor 1

PARs

Protease activated receptors

PI3K

Phosphoinositide 3-kinase

RPMI 1640

Roswell park memorial institute 1640

TEM

Transmission electron microscope

TGF-β

Transforming growth factor-beta

WT

Wild type

Authors’ contributions

S.S. conceptualized and designed the study, performed experiments, analyzed and validated the data, wrote the original draft and reviewed. R.K. performed the in silico experiments and analyzed the data. S.S. analyzed the histopathological data. A.G. provided technical assistance in confocal microscopy experiments. A.P. helped in flow cytometric data acquisition and formal analysis. R.T. did formal analysis and assisted in microbiological experiments. N.N. did formal analysis. S.B. reviewed and edited the manuscript. A.P. conceptualized the study, supervised the research work, helped in funding acquisition, reviewed and approved the final manuscript.

Funding

The study was supported by the institutional funding of ICMR-National Institute for Research in Bacterial Infections.

Data availability

The data presented in this study are available in the article and in the supplementary information. Additional datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All the animal experiments and animal study designs were approved by the institutional animal ethics committee (IAEC) of ICMR-National Institute for Research in Bacterial Infections (formerly ICMR-NICED) (License No: PRO/169/Jan 2022). The maximum tumor volume approved by the institutional animal ethics committee (IAEC) of ICMR-NIRBI in BALB/c mouse is 3000 mm³. Tumors in all the experimental animals remained well under this limit.

Competing interests

The authors declare no competing interests.