Spermine Oxidase Serves as a Key Functional Node in Microbial Dysbiosis-Induced Breast Carcinogenesis

Abstract

Emerging evidence has highlighted the unequivocal importance of microbiota as a risk factor for breast carcinogenesis. Elucidating the underlying mechanisms is needed to identify key nodes that can be inhibited to abolish pathogenic microbe-mediated breast cancer growth and metastatic progression. Focusing on a pro-carcinogenic colon microbe, enterotoxigenic Bacteroides fragilis (ETBF), we uncovered the involvement of spermine oxidase (SMOX), a key enzyme of polyamine metabolism, in promoting breast tumorigenesis. Interestingly, exposure of breast cancer cells to multiple pathogenic microbes such as Fusobacterium nucleatum and pks⁺ Escherichia coli, and to bacterial toxins induced SMOX expression, while non-pathogenic bacteria exhibited no impact. Elevated levels of pro-inflammatory cytokines interleukin-6 (IL6) and tumor necrosis factor-alpha (TNFα) were observed in cells exposed to multiple pathogenic bacteria, and both of these cytokines were capable of upregulating SMOX expression and activity. ETBF and B. fragilis toxin (BFT) exposure led to a considerable rise in reactive oxygen species (ROS) activity, induction of γ-H2AX foci formation, and altered expression of major DNA damage response (DDR) proteins, which were abrogated with SMOX inhibition. Pharmacological inhibition of SMOX using MDL72527 and SXG-1 effectively impeded ETBF-induced breast carcinogenesis with long-lasting effects on tumor-dissociated cells. This work implicates the IL6/TNFα-SMOX-DDR axis as a functional mediator of the oncogenic effects of pathogenic bacteria and proposes SMOX inhibitors as an intervention strategy for treating breast cancer patients with microbial dysbiosis.

Introduction

Breast cancer persists as the most commonly diagnosed cancer globally and ranks second in cancer-related mortality among women in the United States (1). Identifying major risk factors and underlying mechanisms can help in prevention, early detection and more effective treatment. The microbiota and the host maintain a dynamic equilibrium, referred to as eubiosis, that positively influences many physiological processes. However, a state of disequilibrium or dysbiosis may evolve, contributing to various disease states (2,3). Recent studies have indicated the pivotal role of gut and/or breast microbial dysbiosis in breast cancer initiation, progression and therapeutic response (4–7). Nevertheless, there remain distinct gaps in our understanding regarding the molecular mechanisms underlying microbial dysbiosis-driven breast tumorigenesis, making this an imperative area of research. We previously discovered the presence of Bacteroides fragilis (B. fragilis) in breast cancer tissue and established its oncogenic role in breast cancer (8). Virulence of enterotoxigenic B. fragilis (ETBF) is credited to a secreted matrix-metalloprotease termed the B. fragilis toxin (BFT) (9). Owing to its unique virulence traits, ETBF has been proposed as an ‘alpha bug’ that can remodel the bacterial community to enhance its own induction and cause selective ‘crowding out’ or other impacts on protective symbionts, leading to dysbiosis (10). In recent years, additional pathogenic microbes have been connected with cancer initiation and progression. Fusobacterium nucleatum is associated with increased progression in colorectal cancer (11,12), tongue cancer (13), and mouse models of breast cancer (14). Pathogenic Escherichia coli has also been widely studied for its role in multiple cancers (15,16). We, therefore, sought to identify a key node whose inhibition may abolish pathogenic microbe-mediated breast cancer growth and metastatic progression.

Recent studies indicate a connection between oncogenic microbes and the polyamine pathway and, dysregulation of the polyamine pathway has been observed during breast carcinogenesis (17,18). Polyamines are polycationic alkylamines commonly found in all living cells and are indispensable for normal cell functioning and scavenging free radicals. However, their catabolism can be a source of toxic reactive oxygen species (ROS), indicating their potential to affect oxidative status (19). Unsurprisingly, cancer cells exhibit increased polyamine production to support their enhanced growth requirement, and the polyamine pathway is a downstream target for many oncogenes. Accumulating evidences suggest a strong interconnection between polyamines and breast cancer (20,21). Spermine oxidase (SMOX) is a key enzyme of the polyamine pathway that regulates intracellular polyamine pool and catalyzes spermine oxidation into spermidine while releasing hydrogen peroxide and 3-aminopropanal (22). Chronic inflammation or infection modulates SMOX-induced generation of ROS, resulting in DNA damage, potentially leading to cancer development (23). SMOX inhibition drastically impedes tumor incidence in animal models of colon and gastric cancer (24,25). A recent study showed that BFT-mediated genomic damage via SMOX modulation promotes colon cancer (24). These studies indicate SMOX as an important node connecting microbes and cancer progression. Ornithine decarboxylase (ODC), a rate-limiting enzyme in polyamine biosynthesis that decarboxylates ornithine to putrescine (26), has also been linked with cancer development. ODC levels are elevated in TNBC patient samples and treatment with difluoromethylornithine (DFMO), a specific inhibitor of ODC activity, sensitizes TNBC cells to chemotherapy (27). The possible contribution of polyamine biosynthesis to breast cancer is an interesting area of investigation and additional mechanistic understanding governing breast cancer development could assist in improved therapeutic strategies. Hence, we asked: does ETBF dysregulate the key enzymes of the polyamine pathway to induce breast cancer initiation and progression? If so, can we inhibit ETBF-mediated breast cancer by pharmacological targeting of polyamine pathway? Is polyamine pathway dysregulation limited exclusively to ETBF or is it a broader mechanism exploited by microbial dysbiosis during breast carcinogenesis?

Our current findings establish that infection with ETBF results in aberrant upregulation of ODC and SMOX, and downstream effects, such as induction of DNA damage and oxidative stress, in breast cancer cells. Pharmacological inhibition of SMOX inhibits ETBF-stimulated breast tumorigenesis in vitro and in vivo, thereby indicating SMOX inhibition as a critical intervention for microbial dysbiosis-mediated breast carcinogenesis. We also provide evidence to indicate that SMOX is a target of a diverse array of oncogenic microbes and microbial toxins, including but not limited to ETBF. Thus, SMOX acts as a potent functional node underlying microbial dysbiosis-regulated breast carcinoma and SMOX inhibition may prove to be a valuable intervention strategy for management of this disease.

Materials and Methods

Antibodies and Reagents

Antibodies used in this work are as follows: SMOX (PA5-100112 for IHC), MRE11 (PA517573), Rad50 (MA538619), p-NBS1 (MA532093), ATM (MA546912), p-ATM (14904682), p-BRCA1 (PA5104888) and DNA-PKc (MA532192) from Thermo Scientific; SMOX (15052-1-AP, Proteintech for immunoblotting); γ-H2AX (9818S), Ki-67 (9027S), E-cadherin (3195S), CD45 (70257S), CD3 (99940S), CD8 (98941S), CD11b (93169S), F4/80 (70076S) and Rad51 (65653) from CST; ODC (ab193338, Abcam); β-catenin (sc-7963), IL6 (sc-32296) and TNFα (sc-133192) from Santa Cruz Biotechnology. Horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgG were purchased from Sigma-Aldrich. Bacteroides fragilis toxin (BFT) was HPLC purified from ETBF culture supernatants as previously elaborated (8) and 5 nmol/L (100 ng/mL) concentration of purified BFT was used for experiments. E. coli LPS (L2630), Cholera toxin (C8052), Mitomycin C (M4287) and cytochalasin B (C6762) were purchased from Sigma-Aldrich. Two inhibitors of SMOX - MDL72527 (N¹,N⁴−bis(2,3-butadienyl)-1,4-butanediamine) (28) and SXG1 (compound “6” in (29)) were generously provided by Dr. Patrick Woster (Medical University of South Carolina).

SMOX and ODC enzymatic activities

SMOX enzymatic activity was determined as previously described (30). SMOX catalyzes the direct back-conversion of spermine to spermidine, 3-aminopropanal, and H₂O₂. This release of H₂O₂ by SMOX is coupled to chemiluminescence generated by the horseradish peroxidase (HRP)-catalyzed oxidation of luminol. Cells were harvested in glycine buffer and snap-frozen at −80°C. The assay reaction mix is composed of the cell lysate, HRP, luminol, pargyline (monoamine oxidase inhibitor), and aminoguanidine (diamine oxidase inhibitor), in a glycine buffer at pH 8.0. Following a short incubation at 37°C, spermine (substrate for SMOX) was added to the reaction mix and chemiluminescence was measured. The slightly basic pH facilitates HRP-mediated oxidation of luminol in a reaction coupled to H₂O₂ production by SMOX. The amount of chemiluminescence measured in a sample corresponds to the relative oxidation of spermine by SMOX in the cell lysate. The ODC enzymatic activity is evaluated by following the release of radiolabeled ¹⁴CO₂ from L-[1-¹⁴C] ornithine as previously published (31). The assay mixture contained L-ornithine, L-[1-¹⁴C] ornithine, pyridoxal phosphate, dithiothreitol, Tris-HCl at pH 7.5, and enzyme. After incubation at 37°C for 30 min in a shaking water bath, the reaction is stopped by addition of sulfuric acid. This releases ¹⁴CO₂ from the assay medium, and, an additional 15 mins incubation at 37°C ensures complete absorption of the ¹⁴CO₂ into sodium hydroxide placed in scintillation vials. A liquid scintillation spectrometer is used for measurement and one unit is defined as the amount of enzyme releasing 1 nmol of ¹⁴CO₂ per 30 mins at 37°C. Enzyme activities were presented relative to total cellular protein, as estimated using Bradford reagent (#5000205, BioRad) with a bovine serum albumin standard curve.

Cell Lines and Bacterial Strains

Human breast cancer cell lines MCF7 (RRID:CVCL_0031), T47D (RRID:CVCL_0553), HCC1806 (RRID:CVCL_1258), MDA-MB-231 (RRID:CVCL_0062) and normal breast epithelial cell line MCF10A (RRID:CVCL_0598) were procured from the ATCC and maintained at 37°C in 5% CO₂ and 95% humidity. MCF10A-KRas cells (RRID:CVCL_YI68) were a gift from Dr. Ben Park (Vanderbilt University Medical Center). BT474 (RRID:CVCL_0179) and 4T1-luc cells were a gift from Dr. Saraswati Sukumar (Johns Hopkins SOM). The experiments were conducted within 10 to 20 passages from thawing of the cell lines. Authentication for all cells was performed using short tandem repeat testing. The MycoAlert Detection Kit (LT07-218, Lonza) was regularly executed for mycoplasma detection. All bacterial strains used in the study and Streptococcus sp. VT_162 extract have been kindly gifted by Dr. Cynthia L. Sears (SOM, Johns Hopkins). Cultures of enterotoxigenic Bacteroides fragilis (ETBF) strain 86-5443-2-2 (BFT2-secreting strain) were maintained anaerobically at 37°C. Bacterial pellets from overnight cultures were washed and resuspended with 1X Dulbecco PBS (1X PBS free of calcium chloride and magnesium chloride) for mouse inoculums as described earlier (8) or co-cultured with cells for defined durations in serum-free media. Similar procedures were followed for co-culture of F. nucleatum and E. coli. MTB supernatant was a gift from Dr. William Bishai (Johns Hopkins University).

siRNA transfection

Cells were transiently transfected with SMOX-specific stealth siRNA (#HSS122962, Thermo Fisher Scientific), using Oligofectamine (#12252011, Thermo Fisher Scientific) transfection reagent, according to the manufacturer’s instructions. A scrambled siRNA served as a negative control.

Migration Assays

Scratch migration assay: Cells were trypsinized and seeded in 6-well plates and allowed to form monolayer. The monolayer of cells was washed with PBS to remove any cell debris or unattached cells, wound was created by scratching firmly with a 20 μl tip and fresh medium was added with appropriate treatments and the proliferation blocker, mitomycin C (10 μg/ml). Cells were photographed immediately, and migration of cells was followed for the defined time intervals. The width of the injury line remaining was measured against each time-point and graphically depicted. Transwell migration assay: Briefly, 1 × 10⁴ cells were seeded in serum-free media in the upper chamber of transwell inserts in 12-well plates. The lower chambers were filled with serum-supplemented media and incubated for 48h. Migrated cells were fixed in formalin and stained with 0.05% crystal violet. Excess cells were removed using a damp cotton swab and cells on the bottom surface of the inserts were imaged using a microscope, quantified using Leica ImageScope software and graphically presented. Spheroid migration assay: Migration of cells from 3D-tumor cell spheroid was examined using our previously published protocol (8). Cells (2 × 10⁴) were seeded in 1% agar-coated 96-well plates and cultured on an orbital shaker for 72h at 37°C in a humidified atmosphere of 5% CO₂ for the formation of tumor spheroids. Intact tumor spheroids were selected and transferred onto 12-well plates followed by incubation for a defined duration to allow migration of tumor cells from the spheroids. Fixed spheroids were stained with crystal violet, and the migrated cells were observed microscopically, quantified using Leica ImageScope software and graphically depicted.

Matrigel invasion assay

Invasion potential of breast cancer cells was tested using Matrigel invasion assay. Cells (2 × 10⁴) were seeded in the Matrigel invasion chamber from BD Biocoat Cellware as per manufacturer’s instructions. Cells that invaded through Matrigel were fixed, stained with 0.05% crystal violet, imaged using microscope, quantified using Leica ImageScope software and graphically represented.

Mammosphere assay

For liquid mammosphere assay, 5,000 cells were seeded in 2-3 mL of liquid mammosphere media in 6-well ultra-low attachment plates. Cells were treated accordingly and allowed to grow for 7-10 days. These primary mammospheres were trypsinized and cells were counted, seeded for secondary mammospheres and used for additional assays. Cultures were observed under microscope and spheres (>50 μm) were measured and graphically presented.

ALDEFLOUR assay

Cells dissociated from primary mammospheres were suspended in ALDEFLOUR assay buffer, incubated with activated ALDEFLOUR reagent for 30 min at 37°C in dark and ALDH enzymatic activity was measured using the ALDEFLOUR kit (# 01700, Stem Cell Technologies) following the manufacturer’s protocol. Labeled cells were acquired by BD Celesta and analyzed using FACSDiva 6.0 software.

Immunofluorescence, immunohistochemistry and imaging

Immunofluorescence: Cells (5 × 10⁴) were seeded in four-well chamber slides (Nunc, Rochester, NY). The following day, cells were subjected to appropriate treatments for a definite duration. Thereafter, cells were fixed in 4% paraformaldehyde prior to permeabilization using 0.1% Triton-X-100. Cells were then blocked with 3% BSA for 1 h at room temperature and incubated with primary antibody at 1:100 dilution in 3% BSA for overnight at 4ºC. After washing the cells in 1X PBS, they were subjected to incubation with Alexa Fluor 488 or Alexa Fluor 594-conjugated secondary antibody and mounted on coverslips. Images were captured using a Nikon spinning disc confocal microscope at the Johns Hopkins SOM Microscope facility at 40× magnification using oil immersion objective. The intensity of signal per cell was quantified using ImageJ software and graphically depicted (32,33). Immunohistochemistry: Tumor tissue sections were fixed in 10% formalin, paraffin-embedded and sectioned. IHC analyses of the tissue sections was carried out with appropriate antibodies, followed by incubation with HRP-conjugated secondary antibodies and developed using DAB peroxidase substrate kit (SK-4100, Vector Laboratories, CA). Images were captured with a Leica microscope at 20× magnification. Images were analyzed by Aperio ImageScope Software, Leica. Quantification of micrographs and IHC was done using Leica Aperio ImageScope, Leica Biosystems.

Cytokinesis-Block Micronucleus Assay

This is a widely used assay for assessing DNA damage-events are scored in once-divided binucleated cells micronuclei-ensured by addition of cytochalasin B. Cells were treated with this cytokinesis blocker (6 μg/ml) for 24h followed by co-culture with 10⁸ CFU of ETBF culture. Cells were, thereafter, washed with PBS, fixed in 4% paraformaldehyde prior to permeabilization using 0.1% Triton-X-100. Nuclei were stained with DAPI and mounted on a coverslip. Images were captured using a Nikon spinning disc confocal microscope at Johns Hopkins SOM Microscope facility at 40× magnification using oil immersion objective. We looked for biomarkers of DNA misrepair characterized by nuclear buds.

Acrolein Red assay

Acrolein, the simplest α,β-unsaturated aldehyde, is considered as an oxidative stress marker. Oxidation of polyamines represents one of the physiological sources of acrolein in cells. AcroleinRED, used for semi-quantitative measurement of cellular acrolein in live cells, reacts with intracellular acrolein generated via enzymatic pathway and labels the acrolein with TAMRA fluorophore (34). This was subsequently detected using an immunofluorescence microscope.

DCFDA assay

This cell-based ROS assay (ab113851, Abcam, MA, USA) uses the cell permeant reagent 2’,7’-dichlorofluorescin diacetate (DCFDA) to quantitatively assess reactive oxygen species in live cell samples. The DCFDA assay protocol is based on the diffusion of DCFDA into the cell, where it is then deacetylated by cellular esterases to a non-fluorescent compound, which is later oxidized by ROS into a highly fluorescent compound that can be detected by fluorescence spectroscopy with excitation/emission at 485 nm/535 nm. The protocol was performed according to the manufacturer’s instructions.

Protein isolation and western blotting

Whole-cell lysates were prepared using modified RIPA buffer, supplemented with 2 mM phenylmethanesulfonyl fluoride, 2 mM NaF, 1 mM Na₃VO₄ and protease inhibitor cocktail (Roche Applied Science, Mannheim, Germany). Protein concentration was estimated using Bradford’s reagent. Equal amount of protein lysates was subjected to SDS-PAGE followed by transfer onto polyvinylidene difluoride (PVDF) membrane (Millipore, Billerica, MA, USA). The membrane was blocked with 5% non-fat milk for 1 h, incubated with specific primary and horseradish peroxidase-conjugated secondary antibodies, and developed employing enhanced chemiluminescence (Radiance Plus Substrate, Azure Biosystems).

RT-PCR analysis

Total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. One microgram of total RNA was used to synthesize cDNA with an iScript cDNA Synthesis Kit (Bio-Rad, CA, USA). RT-PCR was performed and imaged using the Gel Doc image system (Bio-Rad, CA, USA). For quantitative Real-time PCR, the SYBR GREEN Master Mix (Applied Biosystems, Foster City, CA, USA) was employed as per the manufacturer’s protocol. List of primers is provided in the supplementary table 1.

ELISA

Antigen capture ELISA was performed to measure IL6 and TNFα levels. Herein, 100 μl of sample (media of cells following treatment) was used to coat the wells of a microtiter plate overnight at 4°C. Following day, antigen blocking was done using 5% BSA for 2h at RT. Incubation with primary antibody (human IL6 (sc-32296) and TNFα (sc-133192) diluted in 5% BSA) was carried out overnight at 4°C. Antibody was flicked and plates were washed with PBS four times. HRP-conjugated secondary antibody was added to the wells and allowed to incubate for 2h at RT. Following four washes with PBS, TMB liquid substrate was added and allowed to oxidize for about 20 mins. Absorbance for the blue oxidized chromogen was measured at 450 nm. The intensity of the color was proportional to the amount of antigen present in the sample. A standard was generated for each cytokine to calculate the absolute antigen concentration.

Mouse models

MMTV.f.HuHer2 mice model with oral infection of ETBF: Human HER2-overexpressing transgenic mice in FVB background (HuHER2) that spontaneously develop mammary tumors and lung metastases were generously gifted by Dr. Robert Ivkov, SKCCC, Johns Hopkins. MMTV-Hu-Her2 neu-LucG2 cells were a generous gift from Dr. George Sgouros, SKCCC, Johns Hopkins. The mice were subjected to antibiotic cocktail (clindamycin 0.1 g/L and streptomycin 5 g/L) in water bottles (Hospira and Amresco) for 7 days followed by their random segregation into ETBF and Sham control group. ETBF group received oral administration with ~10⁸ CFU of ETBF in 1X PBS. Sham control mice were administered with 1X PBS. Mammary gland colonization with ETBF: For mammary gland colonization, twice parous BALB/c mice (obtained from Charles River and maintained in house) were given antibiotic cocktail (clindamycin 0.1 g/L and streptomycin 5 g/L) in water bottles (Hospira and Amresco) for 7 days and discontinued. Mice were also injected with antibiotic cocktail intraductally to clear ductal microbiome at the same time. Following random segregation into groups, mice in ETBF group were injected with ∼10⁸ CFU of ETBF in 1X PBS via intraductal administration. For sham control, mice were intraductally injected with 1X PBS. Orthotopic Xenograft Model: NOD/SCID mice (female, 6–8 weeks old) were acquired from Sidney Kimmel Comprehensive Cancer Center (SKCCC) animal facility and maintained in house. Exponentially growing control MCF7 cells and cells treated with 5 nmol/L BFT for 72h (5 × 10⁷ cells in 100 μl matrigel), were implanted in the fourth mammary fat pad on either side. Tumor-dissociated cells from these mice were further implanted in a second set of NOD/SCID mice in a limiting dilution assay to form secondary tumors. These secondary tumors were used for RNA-seq analysis. For the SMOX inhibitor treatment experiment, NSG mice were given antibiotic cocktail (clindamycin 0.1 g/L and streptomycin 5 g/L) in water bottles (Hospira and Amresco) for 7 days and discontinued. They were then randomly segregated into 2 groups and ETBF group was orally gavaged with ∼10⁸ CFU of ETBF in 1X PBS. Sham control mice were administered with 1X PBS. This was followed by implantation of exponentially growing HCC1806 cells (5 × 10⁶ cells in 100 μL matrigel). Once tumors of measurable size were formed, the Sham control and ETBF groups were again divided into a total of 6 groups: Sham control, ETBF, MDL alone, SXG alone, ETBF+MDL and ETBF+SXG (n = 5/group). Accordingly, animals were subjected to intraperitoneal injections of either SXG1 (1 mg/kg body weight) or MDL72527 (20 mg/kg body weight) as indicated and followed. 4T1 Luc2 tumor model with enteric ETBF colonization: BALB/c mice were subjected to antibiotic cocktail (clindamycin 0.1 g/L and streptomycin 5 g/L) for 1 week prior to random segregation and oral administration of either PBS (Sham control) or oral infection with ∼10⁸ CFU of ETBF (n = 5/group). Mammary gland ducts of mice were then injected (via teats) with 4T1-Luc2 cells. For all the models, at the end of the experiment, mice were euthanized and whole mammary glands were excised and tissue sections were prepared for subsequent IHC staining or processed for further analysis as described earlier (8). All animal studies were approved by the Johns Hopkins University Animal Care and Use Committee. The data presented is representative of all the animals included in each study.

Gene expression analysis

ssGSEA analysis with ROC was performed on TCGA-TNBC and BFT-treated xenograft on the Genepattern server (https://cloud.genepattern.org/gp/pages/index.jsf). Further ssGSEA results were correlated with SMOX expression among the samples. We performed ssGSEA-ROC using oncogenic pathway signatures (MSigDB C6) in TCGA-TNBC (N=113) and BFT-treated tumor xenografts (N=7) through Genepattern. Important oncogene pathways were selected based on an AUC score ≥ 0.7 in both datasets (TCGA (44 pathways) and BFT-treated xenograft (57 pathways)). Further Pearson correlation analysis was performed between the SMOX expression and ssGSEA scores. The correction cutoff was calculated through 2/(√N).

Statistical analysis

All experiments were performed thrice in triplicates. Statistical analyses were done using Microsoft Excel, R and GraphPad Prism 5. Data was transformed when necessary to achieve normality. Statistical significance was evaluated by two-tailed student t test. Pairwise comparisons were done using Student’s t-test with Bonferroni correction. Results were expressed as mean ± standard deviations between triplicate experiments performed thrice. * signifies a p-value of < 0.05; ** signifies a p-value of < 0.01; *** signifies a p-value of < 0.001.

Ethics approval and consent to participate:

All animal studies were conducted in accordance with the guidelines of Johns Hopkins University Animal Care and Use Committee and are reviewed by Johns Hopkins ACUC. No Human data are reported here.

Results

ETBF colonization potentiates spontaneous mammary tumorigenesis and leads to increased SMOX expression.

Given the exceptional capacity of ETBF in promoting breast cancer growth in intraductal and mammary gland implantation xenograft models (8), we examined the effect of ETBF colonization (Figure 1) on the development of spontaneous mammary tumors in Her2-/neu transgenic mice, MMTV.f.HuHer2 (Fig. 1a). Mice bearing ETBF exhibited tumor incidence at a younger age and significantly higher tumor multiplicity compared to sham mice (Fig. 1b, c). While mice in the sham group developed palpable tumors after 37 weeks, ETBF-infected mice started developing multifocal tumors (Supplementary Fig. 1) at 29 weeks of age (Supplementary Fig. 1a). ETBF-group showed more aggressive tumor progression and succumbed to higher tumor load with poor survival compared to sham-control mice (Supplementary Fig. 1b, Fig 1e). Persistent colonization of ETBF in the gut was confirmed by stool culture on selective media and PCR for the bft gene at the end of the experiment (~1.5 years of age) (Supplementary Fig. 1c). Histological staining of the tumors revealed marked alterations in mammary tissue architecture of ETBF-harboring mice with noticeable tissue fibrosis and a higher Ki-67 staining, suggesting enhanced epithelial cell proliferation, relative to the sham-control (Fig. 1f). ETBF-harboring mice exhibited significantly higher lung metastasis and micro-metastasis in the liver compared to sham-control mice (Fig. 1d, Supplementary Fig. 1d). Metastasis was confirmed using human Her2-specific immunohistochemistry (Fig. 1g). Mammary tissue sections from ETBF-harboring mice demonstrated higher expression of CD8, CD45, CD3, F4/80 and CD11b, reinforcing that ETBF remodels the breast tumor microenvironment to favor cancer growth and metastasis (Supplementary Fig. 1g–i). For further validation, we used MMTV.f.HuHer2 mammary tumor-derived cells (MMTV-Hu-Her2 neu-LucG2) to perform in vitro assays. BFT-treatment stimulated migration and invasion in MMTV-Hu-Her2 neu-LucG2 cells (Supplementary Fig. 1e, f).

Figure 1. ETBF colonization increases spontaneous mammary tumorigenesis and elevates SMOX expression.

(a) Schematic of experimental plan. (b) Age of MMTV.f.HuHer2 mice in weeks at tumor incidence, and (c) tumor multiplicity from control and ETBF group. (d) Lung metastasis. (e) Median survival plot. (f) Representative trichrome, H&E and IHC images of Ki-67-stained tumor sections from mammary tumors. Scale bar, 100 μm. (g) Representative IHC images of Her2-stained lungs and liver sections. Scale bar, 100 μm. (h) Correlation analysis between SMOX and oncogenic pathways in RNA-seq data from control and BFT-treated MCF7-based tumors of NOD/SCID mice. (i) Correlation analysis between SMOX and oncogenic pathways in human breast tumors from TCGA data. (j) qRT-PCR analysis of SMOX gene expression in mammary tumors in MMTV.f.HuHer2 mice. Graph represents the mean of three independent experiments ± SD. * p value < 0.05 (k) Representative IHC images of SMOX-stained mammary tumor sections from MMTV.f.HuHer2 mice. Scale bar, 100 μm. (l) Schematic of experimental plan. Representative IHC images of SMOX-stained tumor sections from 4T1-luc-derived mammary tumors formed in BALB/c mice bearing enteric ETBF infection. Scale bar, 100 μm. (m) Schematic of experimental plan. Representative IHC images of SMOX-stained tumor sections from mammary tumors formed in BALB/c mice intraductally infected with ETBF. Scale bar, 100 μm. (n) Schematic of experimental plan. Representative IHC images of SMOX-stained sections from tumors formed in mammary fat pads of NOD/SCID mice implanted with control and BFT-exposed (72h) MCF7 cells. Scale bar, 100 μm. Graphs represent the corresponding quantification of SMOX intensity as the mean of three independent experiments ± SD. Statistical significance was calculated using student’s two-tailed t-test. * p-value < 0.05, *** p-value < 0.001 compared to control.

Intrigued by a connection between oncogenic microbes and the polyamine pathway via SMOX modulation in recent studies (35,36), we evaluated SMOX expression in breast cancer patient samples. Comparison of TCGA samples demonstrated a considerable increase in SMOX expression in primary breast tumors, across the subtypes and stages of breast cancer relative to normal breast tissues (Supplementary Fig. 2a–c). Furthermore, Kaplan-Meier analysis indicated that a high expression of SMOX is associated with poor overall survival and recurrence-free survival in breast cancer patients (Supplementary Fig. 2d, e). Our earlier study showed that BFT exposure of MCF7 cells led to multifocal tumors with higher stemness potential as evident in the secondary tumors formed in in-vivo limiting dilution (8). We further analyzed the RNA-seq data obtained from these secondary tumors formed in the BFT and control group (RNA-seq data reported in (8)). Increased SMOX expression was noted in BFT-treated MCF7 cells-derived tumors compared to control (Supplementary Fig. 2f). Correlation analysis revealed a significant correlation between multiple oncogenic pathway signatures (13 pathways; ≥ 0.75 Pearson cutoff; p-value ≤ 0.05) and SMOX expression in BFT-treated MCF7 cells-derived tumors (Fig. 1h). A significant correlation between SMOX expression and multiple oncogenic pathways (25 pathways; ≥ 0.06 Pearson cutoff; p-value ≤ 0.05) was also observed in primary breast tumors from TCGA (Fig. 1i). These observations strongly support the clinical relevance of SMOX in breast cancer and indicate a relevance to ETBF-mediated breast cancer.

Next, we examined whether intraductal or enteric ETBF infection or exposure to BFT affects SMOX expression in mammary tumors. Mammary tumors from ETBF-infected MMTV.f.HuHer2 mice demonstrated a major increase in SMOX expression compared to sham-control mice (Fig. 1j, k, Supplementary Fig. 3a, b). BFT treatment of MMTV-Hu-Her2 neu-LucG2 cells in vitro also led to a time-dependent increase in SMOX expression (Supplementary Fig. 3a). Higher SMOX expression was also observed in 4T1-luc cells-derived intraductal tumors developed in BALB/c mice bearing enteric ETBF colonization (Fig. 1l, Supplementary Fig. 3c). When ETBF was intraductally administered in the mammary ducts of BALB/c mice, we found SMOX elevation in the mammary tissues of mice harboring ETBF compared to control animals (Fig. 1m, Supplementary Fig. 3d). Tumors derived from BFT-pretreated MCF7 cells implanted in mammary fat pads of NOD/SCID mice also exhibited increased SMOX expression (Fig. 1n, Supplementary Fig. 3e). These results unequivocally point to a novel connection between ETBF-mediated breast carcinogenesis and an elevated SMOX level.

B. fragilis toxin (BFT) exposure induces increased expression and activity of SMOX and ODC in breast cancer cells.

BFT treatment caused a time-dependent increase in the level of SMOX (Fig. 2, Supplementary Fig. 4), both at the transcript (Fig. 2a, Supplementary Fig. 4a) and protein level (Fig. 2c) in MCF7 and HCC1806 cells. Her2-overexpressing BT474 cells also showed elevated SMOX expression upon BFT exposure (Supplementary Fig. 4b). In agreement, following co-culture of the cells with ETBF, we observed a marked rise in SMOX expression (Fig. 2b, d). Exploring the potential relationship between ODC and ETBF, our immunoblotting analysis unveiled an induction in the expression of ODC in cells exposed to BFT (Fig. 2c). Immunofluorescence analysis further confirmed a substantial increase in the expression of SMOX and ODC following BFT exposure in multiple breast cancer cells (Fig. 2e–g). To investigate whether BFT exerts a similar effect on these polyamine pathway enzymes in normal breast epithelial cells, MCF10A cells were exposed to BFT for defined durations. BFT exposure resulted in prominent elevation in the expression of SMOX and ODC (Supplementary Fig. 4b, d, e). Earlier, we showed that BFT-pretreated MCF10A cells harboring a KRAS mutation (MCF10A-KRas) formed more aggressive tumors compared to control cells and tumor-dissociated cells derived from these tumors were more migratory and invasive in nature (8). Upon querying the involvement of the polyamine pathway in MCF10A-KRas cells, we observed elevated expression of SMOX and ODC following BFT exposure (Supplementary Fig. 4c, f, g). These findings are in strong agreement with our observations in the in vivo models of ETBF infection, thereby strengthening the potential association between SMOX overexpression and the ETBF-BFT axis in breast cancer. Since SMOX and ODC are key enzymes of polyamine pathway, we investigated whether BFT modulates their enzymatic activity. Multiple breast cancer cells were treated with BFT for specific durations and monitored for SMOX-dependent generation of hydrogen peroxide and ODC-mediated release of radiolabeled ¹⁴CO₂. BFT treatment significantly induced the enzymatic activities of both SMOX and ODC (Fig. 2h, i). Polyamine catabolism also leads to the generation of multiple reactive aldehydes. Acrolein is one such metabolite produced during polyamine oxidation and is an important downstream effector of SMOX (37). AcroleinRED-based assay was used to detect acrolein formation as a downstream effect of SMOX activity (schematic in Supplementary Fig. 5a). We observed a notable increase in the expression of AcroleinRED with increasing duration of treatment with BFT in both MCF7 and HCC1806 cells (Supplementary Fig. 5a). Together, our results showed rapid induction of SMOX and ODC expression and activity in response to ETBF infection and BFT treatment.

Figure 2. ETBF/ BFT increase SMOX expression and activity in breast cancer cells.

(a) RT-PCR analysis of SMOX expression in MCF7 and HCC1806 cells treated with 5nM BFT and (b) MCF7 cells co-cultured with ETBF. (c) Immunoblotting of SMOX and ODC in MCF7 and HCC1806 cells treated with BFT. (d) Immunoblotting of SMOX in MCF7 cells co-cultured with ETBF. β-actin (labeled Actin) served as a control for all the western blots and RT-PCRs. (e-f) Immunofluorescence analyses of SMOX in MCF7 and HCC1806 cells treated with BFT. Scale bar, 20 μm. (g) Immunofluorescence analyses of ODC in T47D and HCC1806 cells treated with BFT. Nuclei are stained with DAPI. Scale bar, 20 μm. (h) Bar graphs represent fold-change in SMOX and (i) ODC enzymatic activity in MCF7, T47D and HCC1806 cells following BFT treatment for indicated timepoints. Graphs represent the mean of three independent experiments ± SD. Statistical significance was calculated using student’s two-tailed t-test. * p-value < 0.05, ** p-value < 0.01 compared to control.

Breast cancer cells exposed to B. fragilis toxin exhibit increased oxidative stress and involvement of γ-H2AX, an early marker of DNA double-strand break (DSB) damage.

An upregulation of SMOX, in response to various inflammatory stimuli, induces oxidative stress and oxidative DNA damage (19). Similar to other enteric bacteria, ETBF is well-established to stimulate ROS generation and, consequently, oxidative DNA damage in colorectal cancer models (38). We found a distinct increase in ROS activity in breast cancer cells exposed to BFT (Fig. 3a, Supplementary Fig. 5b) or co-cultured with ETBF (Fig. 3b, Supplementary Fig. 5c). Utilizing mammary tumor samples from mouse models harboring intraductal or enteric ETBF colonization or bearing BFT-exposed breast cancer cell-derived tumors, we interrogated the involvement of γ-H2AX. Indeed, elevated γ-H2AX level was observed in 4T1-luc cells-derived intraductal tumors developed in BALB/c mice bearing enteric colonization with ETBF (Fig.3c, Supplementary Fig. 6a). Ductal ETBF colonization in BALB/c mice leads to hyperplasia in normal mammary tissue, which also showed increased levels of γ-H2AX compared to control-treated animals (Fig.3d, Supplementary Fig. 6b). Increased γ-H2AX level was also observed in tumors derived from BFT-pretreated MCF7 cells implanted in mammary fat pads of NOD/SCID mice (Fig.3e, Supplementary Fig. 6c). Exhibiting the direct effect of BFT on breast cancer cells, MCF7, HCC1806 and BT474 cells showed increased levels of γ-H2AX upon BFT exposure (Fig. 3f, Supplementary Fig. 6d). A similar BFT-induced γ-H2AX elevation was noted in normal mammary epithelial cells and MCF10A-KRas cells (Supplementary Fig. 6e, f). Also, ETBF co-culture led to a marked rise in γ-H2AX levels (Fig. 3f), and this increase was independent of ETBF-mediated effects on cell proliferation (Fig. 3g, h). Immunofluorescence analysis of breast cancer cells exposed to BFT for various time-intervals further confirmed a substantial increase in the expression of γ-H2AX (Supplementary Fig. 6g, h). This sustained increased level of γ-H2AX indicates a potential role of DNA damage pertaining to ETBF-BFT exposure in breast cancer.

Figure 3. SMOX silencing inhibits BFT-induced ROS and sustained DNA damage.

Representative graphs showing ROS activity in MCF7 and HCC1806 cells, following (a) treatment with BFT and (b) co-culture with ETBF. Hydrogen peroxide (H₂O₂) served as a positive control. (c-e) Schematic of experimental plan and representative IHC images with corresponding quantification of γ-H2AX-stained tumor sections from (c) 4T1-luc-derived mammary tumors formed in BALB/c mice bearing enteric ETBF infection, (d) mammary tumors formed in BALB/c mice intraductally infected with ETBF, (e) tumors formed in mammary fat pads of NOD/SCID mice implanted with control and BFT-exposed (72h) MCF7 cells. Scale bar, 100 μm. (f) Immunoblotting of γ-H2AX in MCF7 and HCC1806 cells treated with BFT and MCF7 cells co-cultured with ETBF. β-actin (labeled Actin) served as a control. (g, h) Representative immunofluorescence analyses of γ-H2AX in MCF7 and HCC1806 cells co-cultured with ETBF, in media containing mitomycin C (10 μg/ml). Nuclei are stained with DAPI. Scale bar, 20 μm. (i) RT-PCR analysis of SMOX expression in MCF7 and HCC1806 cells transfected with scrambled si-RNA (control) or si-SMOX. GAPDH served as loading control. (j, k) Graphical presentation of immunofluorescence analyses of γ-H2AX foci in MCF7 and HCC1806 cells transfected with si-SMOX and treated with BFT as indicated. Graphs represent the mean of three independent experiments ± SD. Statistical significance was calculated using student’s two-tailed t-test. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001.

Involvement of SMOX in B. fragilis toxin-mediated γ-H2AX induction and modulation of DNA damage response (DDR) related key proteins.

We next asked whether SMOX is directly involved in ETBF-BFT-mediated DNA damage and DDR response. Towards this goal, we transiently transfected MCF7 and HCC1806 cells with SMOX-siRNA (Fig. 3i). As evident in the immunofluorescence analysis, SMOX-silenced breast cancer cells failed to induce γ-H2AX foci formation upon BFT exposure (Fig. 3j, k, Supplementary Fig. 6i, j). As iterated above, SMOX upregulation is closely related to increased ROS-mediated DNA damage and our results substantiated that microbial dysbiosis-induced elevation in SMOX level triggered higher γ-H2AX foci formation. A double-stranded break (DSB) in the genome of mammalian cells activates the DSB sensors, thereby initiating the DNA damage response (DDR) pathway. To understand the involvement of DDR, we further analyzed our RNA-seq data. Differential expression analysis of the global differences in RNA transcript levels induced in BFT-pretreated vs. control group revealed higher expression of DDR pathway. Genes associated with DDR pathway are presented in circular heatmap (Fig. 4a) and unsupervised hierarchical clustering heatmap (Supplementary Fig. 7a), and associated key nodes are validated (Supplementary Fig. 7b–j). Formation of γ-H2AX foci serves as a substrate for these DSB sensors, and, facilitates the recruitment of MRN complex (MRE11, Rad50 and NBS) to the sites of DNA damage (39). This leads to phosphorylation and activation of kinases, such as, ATM, Chk2, BRCA1 and Rad51, in consecutive steps, ultimately, resulting in homologous recombination (HR) and repair of the damaged DNA (40). An aberrant DNA repair mechanism is believed to support cancer progression. Independent reports support the close connections between members of the HR pathway with various oncogenic microbes (41–43). This drove us to explore whether SMOX dysregulation, as an aftermath of microbial dysbiosis, sabotages components of the host DDR pathway to promote accumulation of DNA damage. To this end, we subjected breast cancer cells to transient knockdown of Smox utilizing SMOX-siRNA, prior to treatment with BFT (Fig. 4b, c, Supplementary Fig. 8a, b). BFT-exposed cells demonstrated an increased expression of Rad50 and NBS while MRE11 expression remained unaffected (Fig. 4b, c, Supplementary Fig. 8b). A concomitant reduction in the levels of phospho-ATM and phospho-BRCA1 was also observed in BFT-treated cells compared to control (Fig. 4d, e, Supplementary Fig. 7b–j, 8c), indicating that BFT suppressed the activation of these DSB repair proteins, suggestive of defective DNA repair. BFT treatment also led to reduced level of Rad51 foci in the nuclei of the cells despite high expression of γ-H2AX (Fig. 4f, g, Supplementary Fig. 8d). Despite appreciable DNA damage, reflected by elevated nuclear γ-H2AX, BFT perturbed Rad51 recruitment to these sites of DNA injury. In sharp contrast, this BFT-mediated disruption to the HR pathway was severely impaired in cells with SMOX knockdown (Fig. 4b–g, Supplementary Fig. 7b–j, 8b–d). In agreement, we observed a significant increase in DNA-PKc expression accompanied by enhanced micronuclei formation following co-culture of breast cancer cells with ETBF, suggesting the lingering persistence of DNA damage (Fig. 4h–j). These evidences suggest that SMOX is essential for mediating BFT-induced disturbances to the host HR pathway, which may support accumulation of DNA damage in breast cancer cells.

Figure 4. Breast cancer cells exposed to BFT exhibit altered DNA damage response (DDR) and SMOX silencing modulates BFT-altered key DDR-pathway proteins.

(a) Circular heatmap showing the DDR pathway-related DEGs specifically downregulated/upregulated in the MCF7-tumor-derived RNA-seq data. (b, c) Immunoblotting of SMOX, Rad50, MRE11 and p-NBS1 in MCF7 and HCC1806 cells treated as indicated. β-actin (labeled Actin) served as a loading control. (d, e) Graphical depiction of immunofluorescence analyses of ATM, p-ATM and p-BRCA1 expression in MCF7 and HCC1806 cells treated as indicated. Graphs represent the mean of three independent experiments ± SD. Statistical significance was calculated using student’s two-tailed t-test. ** p-value < 0.01, *** p-value < 0.001. (f, g) Representative images of immunofluorescence analyses showing co-localization of γ-H2AX and RAD51 in MCF7 and HCC1806 cells with indicated treatments. Nuclei are stained with DAPI. Scale bar, 20 μm. (h) Immunoblotting for DNA-PKc in MCF7 and HCC1806 cells co-cultured with ETBF. Actin served as a loading control. (i, j) Representative images of micronuclei staining of MCF7 and HCC1806 cells co-cultured with ETBF following 24h treatment with cytochalasin B (6 μg/ml). Arrows point to the micronuclei. Scale bar, 10 μm.

Pharmacological inhibition of SMOX impedes BFT-induced pro-carcinogenic attributes of breast cancer cells.

Our earlier study showed that BFT exposure induces a network of genes associated with cell migration and cell movement, and increases the migration of breast cancer cells (8). Here, we investigated whether SMOX silencing interrupts BFT-induced migration of breast cancer cells. The results of transwell migration assay showed considerably decreased cellular migration in the presence of BFT after silencing SMOX (Supplementary Fig. 9a, b). Next, the effects of SMOX silencing on BFT-mediated spheroid-migration was determined. The results indicated that, contrary to the visibly evident increase in spheroid migration in BFT-exposed cells, SMOX silencing inhibited spheroid migration despite treatment with BFT (Supplementary Fig. 9c). Our results revealed that silencing SMOX markedly repressed BFT-facilitated migration potential in breast cancer cells.

Further, to examine the potential of pharmacological inhibitors of SMOX on functional aspects of breast cancer cells, BFT-exposed MCF7 and HCC1806 cells were treated with either of two potent inhibitors of SMOX, SXG-1 (29) and MDL72527 (44). SMOX enzymatic assay revealed a significant increase in SMOX activity in BFT-treated cells compared to control cells (~1.8-fold in MCF7, p < 0.01, and ~2.2-fold in HCC1806 cells, p < 0.05). This rise was considerably mitigated following treatment of the BFT-exposed cells with either MDL72527 (~1.25-fold in MCF7, p < 0.01, and ~1-fold in HCC1806 cells, p < 0.05) or SXG-1 (~1.1-fold in MCF7, p < 0.05, and ~1.1-fold in HCC1806 cells, p < 0.05) (Fig. 5a). These results were further corroborated with the AcroleinRED assay (Supplementary Fig. 10a, b) and γ-H2AX foci formation (Supplementary Fig. 10c–f). These findings clearly demonstrate that the SMOX inhibitors successfully blocked BFT-mediated amplification of SMOX activity in breast cancer cells.

Figure 5. Pharmacological inhibition of SMOX inhibits BFT-induced SMOX activity and oncogenic hallmarks in breast cancer cells.

Breast cancer cells were treated with SMOX inhibitors, SXG-1 (0.544μM) or MDL72527 (250μM), for 16h prior to BFT exposure for 6h. (a) Bar graphs show SMOX enzymatic activity in treated MCF7 and HCC1806 cells. (b, c) Representative images of treated MCF7 and HCC1806 cells undergoing transwell migration assay for 48h (Scale bar, 120 μm) with the corresponding quantification. (d, e) Representative images of primary and secondary mammospheres formed by treated MCF7 and HCC1806 cells (Scale bar, 100 μm) with the corresponding quantification. (f) Flow cytometry analysis for ALDH-positive population in cells from MCF7 primary mammospheres. (g) Representative qRT-PCR analysis for expression of stemness-enriched genes, c-Myc, Sox2, Plau and Nanog, in cells from HCC1806 secondary mammospheres. Graphical data represent mean ± SD from three independent experiments. Statistical significance was calculated using student’s two-tailed t-test. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001.

Next, we tested the effect of SXG-1 and MDL72527 on pro-carcinogenic attributes of BFT-treated breast cancer cells. While BFT exposure led to a considerable increase in cellular migratory potential, co-treatment with either of the SMOX inhibitors significantly repressed this effect (Fig. 5b, c). These observations were verified using a scratch-migration assay, wherein SMOX inhibitors inhibited BFT-induced increased migration of breast cancer cells (Supplementary Fig. 11a–d). Additionally, treatment with the SMOX inhibitors abrogated BFT-associated stemness potential of breast cancer cells, as depicted by our primary and secondary mammosphere assay. SMOX inhibition notably reduced both the number and size of primary and secondary mammospheres formed in BFT-exposed cells (Fig. 5d–e). Moreover, flow cytometry of cells from the MCF7 primary mammospheres corroborated a significant increase in the percentage population of ALDH-positive cells in response to BFT exposure- an effect that was mitigated with SMOX inhibition (Fig. 5f). Additionally, qRT-PCR analysis revealed a substantial increase in the expression of known stemness-associated genes, c-Myc, Sox2, Plau and Nanog, in HCC1806 secondary mammospheres in the BFT group which was inhibited with MDL/SXG treatment (Fig. 5g). Immunofluorescence analysis across different breast cancer cells further indicated a considerable repression in the expression of γ-H2AX foci in BFT-exposed cells upon treatment with either MDL72527 or SXG-1 (Supplementary Fig. 10c–f). These results indicated that SMOX inhibition using pharmacological agents drastically restricts BFT-mediated pro-oncogenic characteristics of breast cancer cells.

Treatment with SMOX inhibitors inhibits ETBF-accelerated breast cancer development and progression in vivo.

Our in vitro findings intrigued us to explore the possibility of using SMOX inhibitors in animal models as tumor suppressive agents against ETBF-induced breast carcinogenesis. The overall schematic of our experimental plan is presented in Fig. 6a. Mice harboring ETBF infection showed a more rapid tumor progression and an increased tumor weight (~1.7-fold) compared to control group (Fig. 6a, b). In contrast, mice bearing ETBF colonization, when subjected to either of the SMOX inhibitors, had a visible decline in their tumor volume in comparison to the ETBF-infected group (Fig. 6a, b, Supplementary Fig. 12a). IHC staining of mammary tumor sections (Fig. 6c–e, Supplementary fig. 12b–g) from ETBF-bearing mice revealed denser staining of SMOX and higher number of Ki-67-positive cells, signifying higher proliferation, compared to control (Fig. 6c, e; Supplementary Fig. 12b, e, d, g). Furthermore, these ETBF-exposed tumors exhibited sustained DNA damage, as evaluated by enhanced γ-H2AX staining (Fig. 6d; Supplementary Fig. 12c, f). These effects were remarkably inhibited in tumor sections from ETBF-exposed animals, who received treatment with either SXG-1 or MDL72527 (Fig. 6c–e; Supplementary Fig. 12b–g). Next, tumors were dissociated into single cells and the ex vivo cells were evaluated for their oncogenic characteristics. Our AcroleinRED assay verified the suppression of SMOX activity in the tumor-dissociated cells from ETBF-infected group treated with either SXG-1 or MDL72527 (Fig. 6f). For additional validation, tumor-dissociated cells from each experimental group were tested for their pro-cancer potential (Fig. 6g–l). Tumor cells from the ETBF group demonstrated augmented cellular migratory index and enhanced spheroid migration relative to control and this effect was mitigated following SMOX inhibition (Fig. 6g, h, k, l; Supplementary Fig. 13a–d). We also observed a distinct elevation in the mammosphere-forming potential of cells from the ETBF-infected mice compared to control, a feature that was significantly suppressed in the presence of the SMOX inhibitors (Fig. 6i, j). Our collated data exhibit that pharmacologically inhibiting SMOX may significantly ameliorate the oncogenic alterations pertaining to enhanced tumor growth and elevated migratory and stemness-rich phenotype brought upon by ETBF infection.

Figure 6. Pharmacological inhibition of SMOX blocks BFT-induced breast tumor growth in mice.

(a) Diagrammatic depiction of experimental design. Bar graph shows tumor weight in six experimental groups as indicated. (n = 5) * p-value < 0.05, ** p-value < 0.01. (b) Representative line graph shows tumor progression for each experimental group. Statistical significance was calculated using student’s two-tailed t-test. p-value are tabulated in supplementary figure 12a. (c-e) Representative IHC images of SMOX, γ-H2AX and Ki-67–stained mammary tumor sections formed in mammary fat pads of the aforementioned NSG mice (n = 5). Scale bar, 100 μm. (f-l) Resected tumors from all mice from the six experimental groups were dissociated and the isolated tumor cells were subjected to different in vitro functional assays. (f) Representative images of AcroleinRED assay from each group. Nuclei are stained with DAPI. Scale bar, 20 μm. (g-l) Representative images of the ex vivo HCC1806 tumor cells undergoing (g, h) transwell migration assay for 48h (Scale bar, 120 μm) with the corresponding graphical quantification. (i, j) mammosphere assay (Scale bar, 100 μm) with the corresponding graphical depiction (k, l) spheroid migration assay for 48h . Scale bar, 100 μm. Graphical data represent mean ± SD from three independent experiments. Statistical significance was calculated using student’s two-tailed t-test. * p < 0.05, ** p < 0.01, *** p < 0.001.

SMOX inhibitors abrogate the effect of BFT on E-cadherin-β-catenin localization.

Aberrant morphogenetic alterations occur as a result of a disrupted E-cadherin/β-catenin complex (45). SMOX deletion in H. pylori-infected gastric cancer models leads to the membrane translocation of β-catenin, suggestive of β-catenin inactivation (35). Also, ODC1 silencing is associated with reduced expression and decreased nuclear localization of β-catenin (46). Of importance, heightened levels of cytoplasmic and nuclear β-catenin are common in 40% of primary breast cancers and connected to poor patient outcome (47). Immunostaining revealed that BFT led to a distinct nuclear translocation of β-catenin (Supplementary Fig. 14) in MCF7, T47D and BT474 breast cancer cells due to disruption of the membrane-bound β-catenin/E-cadherin complex (Supplementary Fig. 14a–c). BFT-stimulated nuclear accumulation of cytoplasmic β-catenin was found in HCC1806 and MDA-MB-231 cells (Supplementary Fig. 14d–e). This nuclear localization associated with the activation of β-catenin-responsive genes, such as Axin2, Survivin, CyclinD1 and c-Myc, in MCF7 and HCC1806 cells following co-culture with ETBF (Supplementary Fig. 14g). Importantly, BFT-mediated nuclear localization of β-catenin was reversed in the presence of the ODC inhibitor, DFMO, validating the functional connection between β-catenin and ODC (Supplementary Fig. 14f). Furthermore, treatment with SMOX inhibitor reversed the BFT-induced translocation of β-catenin suggesting inactivation of the β-catenin pathway (Supplementary Fig. 15a–c). These results support the hypothesis that SMOX inhibition interrupts BFT-associated aberrant activation of the β-catenin signaling pathway (Supplementary Fig. 14–15), which aids in suppressing downstream oncogenic effects, such as breast cancer cell migration and stemness.

Multiple microbes and microbial toxins show functional convergence on the modulation of inflammatory cytokines- IL6 and TNFα levels, and SMOX upregulation.

We found that gut colonization with ETBF facilitates a systemic immune response, with enhanced levels of various pro-inflammatory cytokines, such as, IL6 and TNFα, rewiring the premetastatic niches and promoting an immune-suppressive and pro-metastatic environment (48). Also, our RNA-seq data showed a significant upregulation of the IL6 and TNFα pathways in the BFT-treated group (Supplementary Fig. 16a, b). Breast cancer cells exhibited a significant increase in the levels of both IL6 and TNFα following exposure to various microbial toxins, including E. coli LPS, Cholera toxin and BFT or co-culture with either ETBF or F. nucleatum (Fig. 7a, b). However, a non-pathogenic pks-negative strain of E. coli failed to significantly affect IL6/TNFα levels (Fig. 7a, b). Importantly, SMOX has been associated with immunomodulatory effects in various models (49,50). The collated reports suggest that the pro-inflammatory cytokines, such as IL6 and TNFα, may mediate, at least in part, their pro-tumorigenic effects, such as enhanced ROS generation and genomic injury, through perturbed expression of SMOX. To explore this possibility, we treated breast cancer cells with IL6 or TNFα and found a sharp increase in both the level (Fig. 7c, d) and enzymatic activity of SMOX (Fig. 7e, f) in response to these pro-inflammatory cytokines. Together, these data indicate the influence of microbial dysbiosis on the host inflammatory response and denote the subsequent upregulation of SMOX as a potent mechanism for triggering breast tumorigenesis.

Figure 7. Multiple pathogenic bacteria or bacterial toxins exhibit elevated SMOX expression and activity via stimulation of IL6 and TNFα.

(a, b) ELISA assay for the levels of cytokines IL6 and TNFα in the media of MCF7 and HCC1806 cells exposed to various microbial toxins (50 μg/ml of LPS, 100 ng/ml of Cholera toxin, 5 nM of BFT) or co-cultured with bacteria as indicated. Graphical data represent mean ± SD from two independent experiments. Statistical significance was calculated using student’s two-tailed t-test. * p-value < 0.05, ** p-value < 0.01. (c, d) Immunoblotting of SMOX in MCF7 and HCC1806 cells treated with 10 ng/ml of human IL6 or human TNFα. β-actin served as a loading control. (e, f) Representative images of AcroleinRED assay in MCF7 and HCC1806 cells following treatment with 10 ng/ml of human IL6 or human TNFα. Nuclei are stained with DAPI. Scale bar, 20 μm.

Our next goal was to explore the possibility that potentiation of SMOX may be an integral mechanism deployed by a variety of microbes for promoting breast carcinogenesis. To test this proposition, we exposed breast cancer cells to various bacteria and bacterial metabolites that are vital regulators of host cell functions. Of note, co-culture of MCF7 cells with either E. coli (pks⁺) or F. nucleatum resulted in a pronounced induction of both SMOX and γ-H2AX, as evident from our western blotting and immunostaining analysis (Fig. 8a, b). Furthermore, Mycobacterium tuberculosis culture supernatant (designated as MTB) led to a similar elevation in the expression of SMOX and γ-H2AX (Fig. 8c, d). A similar trend was found for LPS from E. coli or Cholera toxin, (Supplementary Fig. 17 a–g). Co-culture with the pks-E. coli, albeit a modest increase in the level of γ-H2AX, did not elicit SMOX upregulation in MCF7 and HCC1806 cells (Supplementary Fig. 18a–f). Exposure to culture extract of another oral opportunistic anaerobe, Streptococcus sp. VT_162, did not considerably affect SMOX expression though it induces DNA damage (Supplementary Fig. 18g, h), implying that our observed effect is specific to certain microbes. Our data, thus, showed that SMOX dysregulation as a molecular mechanism involved in breast tumorigenesis is not limited to ETBF-BFT but is employed by other putatively oncogenic microbes (pks+ E. coli, F. nucleatum) and toxins.

Figure 8. Exposure to multiple pathogenic bacteria results in modulation of SMOX and γ-H2AX in breast cancer cells.

Immunoblotting of SMOX and γ-H2AX in MCF7 cells co-cultured with (a) pks-positive E. coli, and (b) F. nucleatum. Lower panels show representative images of immunofluorescence analyses of γ-H2AX in these cells. Nuclei are stained with DAPI. Scale bar, 20 μm. (c) Immunoblotting of SMOX and γ-H2AX and, (d) representative images of immunofluorescence analyses of γ-H2AX in MCF7 and HCC1806 cells treated with MTB. Nuclei are stained with DAPI. Scale bar, 20 μm. β-actin (labeled Actin) served as a control for all the western blots. (e) Representative images of transwell migration assay and mammosphere-formation assay in MCF7 cells, co-cultured with F. nucleatum. (f) Representative images of transwell migration assay and (g) mammosphere-formation assay in MCF7 and HCC1806 cells treated with MTB supernatant. Scale bar, 100 μm.

We, then, verified the oncogenic impact of these pathogenic microbes and microbial toxins pertaining to changes in the migratory index and mammosphere-forming potential of breast cancer cells. As expected, there was a sharp rise in the number of migratory cells following exposure of breast cancer cells to the microbial toxins and upon co-culture with MTB or F. nucleatum (Supplementary Fig. 17e, g, Fig. 8e, f). Furthermore, we found a drastic increase in the mammosphere-forming potential of MCF7 and HCC1806 cells following treatment with either LPS or MTB (Supplementary Fig. 17f, Fig. 8g). A considerable augmentation in the number of mammospheres was observed for MCF7 cells co-cultured with F. nucleatum (Figure 8e). These results indicate a mechanism involving an aberrant upregulation of SMOX by microbial dysbiosis during breast oncogenesis and suggest that targeting SMOX may interfere with the oncogenic potential of several microbes.

Discussion

That microbial dysbiosis is a risk factor for breast cancer initiation and progression is now well-established and well-substantiated through studies with clinical samples and preclinical models (2–4,8,48). Microbes can directly or indirectly exert tumor-centric as well as tumor-microenvironment (TME)-centric effects leading to the initiation and progression of cancer. Our previous studies demonstrated that gut or mammary ductal colonization with ETBF, an “alpha bug”, considerably stimulated growth, stemness, migration, and invasion potential of breast cancer cells resulting in increased breast cancer progression and metastasis (8). The results presented here are the first to demonstrate a direct involvement of SMOX in mediating the effect of pathogenic microbes in breast cancer. We found the association of SMOX with oncogenic pathways in ETBF-BFT-exposed breast tumors as well as in clinical samples from TCGA. Validation studies involved multiple breast cancer-bearing mouse models harboring gut or ductal exposures with ETBF, all showing elevated levels of SMOX in their mammary tumors. We also observed an increase in ODC levels in breast cancer cells upon ETBF-BFT exposure. SMOX and ODC are the key enzymes of the polyamine pathway, and their interplay regulates the intracellular polyamine pool (17,19). Neoplastic tissues exhibit dysregulation of polyamine metabolism resulting in elevated concentrations of polyamines, in comparison to normal tissues. Since polyamines are required for many physiological processes and are strongly associated with cancer, reducing their level can impact tumor progression (51). In colon cancer, purified BFT was shown to induce SMOX, resulting in SMOX-dependent oxidative damage (24), potentiating ETBF-related colon tumorigenesis (24). Notably, our results clearly show that SMOX inhibition blocks ETBF-BFT induced pro-oncogenic characteristics of breast cancer cells and significantly suppresses ETBF-induced breast tumor growth, directly connecting polyamine metabolism to ETBF-induced breast tumor.

Enzymatic activity of SMOX involves the oxidation of spermine leading to spermidine production, as well as the aldehyde 3-amniopropanal, and hydrogen peroxide (52). Generation of ROS causes DNA damage, which if left untreated, leads to accumulation of mutations in driver genes contributing to carcinogenesis. Hence, increased levels of SMOX may promote the accumulation of genetic and epigenetic alterations, one of the hallmarks of cancer (53). H. pylori colonization of the stomach mucosa causes SMOX upregulation-mediated increase in H₂O₂ production and DNA damage without causing apoptosis or cell death (36). SMOX-mediated ROS induction and the ensuing DNA damage without associated cell death has also been implicated in colon cancer exposed to ETBF (24). In our study, a marked rise in ROS induction and γ-H2AX foci in response to ETBF or BFT exposure suggests a critical role of SMOX-related oxidative stress and DNA damage in ETBF-induced breast cancer. Furthermore, transient knockdown of SMOX was associated with a significant suppression in BFT-mediated increased γ-H2AX foci formation in breast cancer cells, supporting our notion that SMOX is an important mediator of BFT-facilitated DNA-damage in breast cancer. In gastric carcinoma, H. pylori is known to hinder the phosphorylation and nuclear translocation of BRCA1, thereby generating DNA-DSBs while disabling error-free HR-induced DNA repair (43). Infection with H. pylori also activated the DNA damage pathway in concert with Rad51 suppression in models of gastric carcinogenesis (54). We showed that BFT-mediated SMOX upregulation results in substantial generation of DSBs, which leads to enhanced recruitment of γ-H2AX and the MRN complex to the DSB sites. However, the BFT-SMOX axis impairs the downstream DNA repair process by preventing the normal activation of ATM, BRCA1 and Rad51, thus, interfering with regular homologous recombination process and potentially causing accumulation of genetic defects. We found higher expression of DNA-PKc and micronuclei formation in ETBF-co-cultured breast cancer cells. DNA-PKc overexpression is associated with higher tumor grade, chemoresistance and poor survival in breast cancer patients (55). Additional mechanistic insights can be procured by identifying the key factors of the ROS and DNA damage pathway that are dysregulated by the SMOX-ETBF axis in breast carcinoma.

Moreover, we observed the involvement of SMOX and ODC in BFT-mediated nuclear translocation and activation of the β-catenin pathway. Upregulation of β-catenin resonates with higher tumor stage and lower survival in breast cancer subjects (56). We found considerable enrichment of β-catenin-responsive genes that may contribute to BFT-associated enhanced stemness of breast tumors. Our results clearly showed that aberrant activation of the β-catenin pathway in breast tumor cells can be reversed to a significant extent following pharmacological inhibition of either SMOX or ODC. Research pertaining to numerous inhibitors of the Wnt-β-catenin pathway are currently under investigation (57). Our data hint at the potential of combining intervention strategies targeting the polyamine pathway enzymes with known β-catenin inhibitors for mitigating the oncogenic functions of ETBF in breast carcinoma.

Our results show that SMOX inhibition can suppress breast tumor growth and the highly migratory and stemness-rich phenotype that is observed with ETBF exposure. ETBF was examined in patients harboring gut-related ailments and healthy controls. While higher number of patients with diarrhea or colitis showed ETBF gut colonization, ETBF was found to asymptomatically colonize the gut of 5% to 35% of healthy controls (58). ETBF is a potent pathogen (10), therefore, SMOX inhibition is expected to exert a significant impact on breast cancer patients harboring ETBF infection. Given that ETBF is a common colonic colonizer, it has been widely explored in the context of colon tumorigenesis while there is a dearth of research related to its involvement in breast cancer. Our lab was the first to establish the oncogenic prowess of the common colonic colonizer, ETBF in mammary tumorigenesis (8,48). While ETBF may not initiate breast cancer in all persons harboring ETBF colonization, it may increase breast cancer growth and metastatic progression. It is also plausible that ETBF may only impact people at higher risk of developing breast cancer owing to other modifiable or non-modifiable risk factors. Clinical studies are needed to further examine the impact of ETBF on breast cancer initiation. The current study provides additional insight into the host molecular pathways that are hijacked and modulated by ETBF during its carcinogenic manifestations in the breast. Since few other microbes and microbial toxins have been recently implicated in breast cancer, we asked an important question if SMOX is a key node that is utilized by multiple microbes and microbial toxins. Interestingly, we found that pathogenic F. nucleatum, E. coli, MTB as well as LPS and cholera toxin induced SMOX level and DNA damage in multiple breast cancer cells. Functional impact of these microbes and microbial toxin is also clearly visible on breast cancer cells with respect to their invasion and migration potential. Interestingly, we also uncovered that these pathogenic microbes and toxins induce inflammatory cytokines, TNFα and IL6, which are capable of upregulating SMOX level and activity in breast cancer cells. Both TNFα and IL6 play an important role in activating various cellular functions, including cell proliferation, movement and stemness, as well as impacting various cell types in the tumor microenvironment. Co-expression of high TNFα and IL6 correlates with aggressive clinical parameters and significantly reduced survival of breast cancer patients (59,60). Our results also showed that ETBF remodels the breast immune microenvironment in the spontaneous mammary tumor murine model, characterized by higher T cell infiltration (increased CD3, CD8 and CD45 staining), tumor-associated macrophages (higher F4/80) and a pro-metastatic niche (higher CD11b). This further validates our previously published observations in a different breast cancer mice model (48). Given the ETBF/BFT-IL6/TNFα/SMOX axis, it will be worthwhile to interrogate how SMOX inhibition can affect the breast tumor microenvironment and overcome ETBF-associated immune remodeling. Since targeting multiple microbes individually is extremely difficult, it would be preferable to understand mechanistic underpinnings that mediate their biological impact, find common nodes and develop therapeutic strategies. Our results present SMOX as an actionable node whose inhibition can potentially mitigate the impact of various pathogenic microbes.

Collectively, these results provide new insight into the relationship between pathogenic microbes, inflammatory cytokines and polyamine pathway in breast cancer, and establish SMOX as a quintessential node of functional convergence for microbial dysbiosis-associated breast carcinogenesis. Future investigations should focus on further developing SMOX inhibition strategies in combination with standard therapeutical approaches, and developing tools to select breast cancer patients who harbor pathogenic microbes and design focused clinical studies.

Statement of Significance:

Overabundance of opportunistic pathogens elevates SMOX activity via pro-inflammatory cytokines to accelerate breast cancer progression, which can be targeted with pharmacological inhibitors of SMOX to significantly inhibit microbiota-associated carcinogenesis.

Availability of data and materials:

Data generated in this study are available within the article and its supplementary files. Raw data for RNA seq for this study were generated at The Single Cell & Transcriptomics Core, JHU. Derived data supporting the findings of this study are available from the corresponding author upon request. Also, the data analyzed in this study were obtained from publicly available database, The Cancer Genome Atlas (TCGA). All other raw data generated in this study are available upon request from the corresponding author.