NCoA3 upregulation in breast cancer-associated adipocytes elicits an inflammatory profile

doi:10.21203/rs.3.rs-1730698/v2

Download PDF

Article

NCoA3 upregulation in breast cancer-associated adipocytes elicits an inflammatory profile

https://doi.org/10.21203/rs.3.rs-1730698/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted

You are reading this latest preprint version

Purpose

NCoA3 is a transcriptional coactivator of NF-κB and other factors, present in undetectable levels into normal cells and amplified or overexpressed in several types of cancer, including breast tumors. We have previously demonstrated that NCoA3 levels decrease during adipogenesis, but its role in tumor-surrounding adipose tissue is still unknown. Therefore, in this work, we studied the modulation of NCoA3 in breast cancer-associated adipocytes and evaluated its correlation with the expression of inflammatory markers.

Methods

We collected mammary adipose tissue (MAT) from female mice that were stimulated with conditioned media from breast cancer cell lines or MAT adjacent to tumors from breast cancer patients. Then, the expression levels of NCoA3 were evaluated by qPCR and immunohistochemistry. TNF and MCP1 levels were measured. Ex vivo models were validated using 3T3-L1 cell lines and bioinformatics analysis.

Results

We found that in MAT samples expressing high levels of NCoA3 were mainly associated to a pro-inflammatory profile. Even more, patients with a worse prognosis and advanced clinical stage of disease showed high levels of this coactivator. Furthermore, in 3T3-L1 cell line-derived adipocytes, NCoA3 silencing or NF-κB inhibition reverted the expression of inflammatory molecules.

Conclusion

The modulation of NCoA3 levels in synergy with NF-κB activity in MAT in tumoral context, could be determinant factors to establish breast cancer-associated inflammation. As adipocytes are involved in the development and progression of breast cancer, this signaling network deserves to be further investigated to improve future tumoral treatments.

Mammary adipose tissue

breast cancer

inflammation

TNF

NCoA3

The risk of endometrial and breast cancer development is known to be augmented in postmenopausal overweight women. Moreover, after menopause the composition of mammary adipose tissue (MAT) is modified and provides estrogens to the mammary gland [1].

Surrounding breast cancer cells, the most abundant tissue is adipose tissue (AT), and nowadays, it is recognized how important the role of the microenvironment is during tumorigenesis, for instance in proliferation, invasion and chemoresistance mechanisms [2]. Indeed, even though autocrine signals regulate the tumoral behavior, effects exerted by adjacent stromal cells also play a role in this signaling network. These signals could be soluble secreted factors, extracellular matrix components, and also cell–cell contact. This signaling involves a constant crosstalk between normal and transformed cells [3].

It has already been demonstrated that MAT adjacent to breast tumors serves as a source of pro-inflammatory cytokines and chemokines such as interleukin-6 (IL-6), tumor necrosis factor (TNF) and monocyte chemoattractant protein 1 (MCP1) [4]. One of the key transcriptional factors that lead to the expression of inflammation-associated molecules is NF-κB.

NCoA3/RAC3 (Nuclear coactivator 3/Receptor Associated-Coactivator 3) is a NF-κB coactivator, member of the p160 family of Steroid Receptor Coactivators. Previously, our group has demonstrated that NCoA3 expression levels are upregulated by TNF in response to inflammatory injury [5] and are downregulated during adipocyte differentiation [6]. It is highly expressed in several human cancers, including breast cancer [7, 8]. Up to now, we have found that NCoA3 overexpression has an anti-apoptotic, anti-autophagic and pro-proliferative role [9–11]. Moreover, NCoA3 expression levels correlate with higher chemoresistance [12]. However, its levels in the adipose tissue adjacent to breast cancer remain uncharacterized.

In this study we evidence that patients with more aggressive breast tumors or with worse prognosis tend to have higher expression of NCoA3 in the MAT adjacent to tumors and this also correlates with different markers of inflammation. Moreover, the silencing of NCoA3 or the inhibition of NF-κB reverts the pro-inflammatory phenotype of MAT in the tumoral context.

All these results suggest the importance of NCoA3 in the induction of molecular features of inflammation in MAT surrounding breast cancer.

Cell culture

The murine pre-adipocyte cell line, 3T3-L1 (ATCC® CL-173™), and the human breast cancer cell lines, T-47D (ATCC® HTB133™) and MDA-MB-231 (ATCC® CRMHTB26™), were purchased from American Type Culture Collection (ATCC® Manassas, VA, USA). The 3T3-L1 cell line was grown in Dulbecco’s modified Eagle’s medium (DMEM), MDA-MB-231 and T-47D cell lines in DMEM/F12 (1:1). Every medium was supplemented with 10% fetal bovine serum (FBS) (Gibco Laboratories, Grand Island, NY, USA), penicillin (100 U/ml) and streptomycin (100 mg/ml). Cells were maintained at 37°C in a humidified atmosphere with 5% CO₂.

Conditioned medium from breast cancer cell lines (BrCM) was collected after 24 h with free-serum medium and cell debris were pulled down by centrifugation. BrCM aliquots were conserved at -20ºC until adipocyte stimulation.

3T3-L1 cells were plated and differentiated by induction differentiation medium, described previously (1µg/ml insulin, 1µM dexamethasone, 0.5mM isobutylmethylxanthine and 0.1mM indomethacin) [6]. Insulin (1µg/ml) was added every 3 days until the week was completed.

After 3T3-L1 cell differentiation, media were replaced by T-47D BrCM or fresh free-serum medium for 24 h and then protein or RNA extraction were performed. To study the role of NF-κB activation, 3T3-L1-derived adipocytes were pre-incubated with NF-κB inhibitor Sulfasalazine (SSZ 0.2mM) for 30 min before stimulation with T-47D BrCM. 3T3-L1 adipocyte conditioned media were collected after being cultivated in fresh DMEM/F12 free-serum medium for 24h for dot blot assay.

Plasmid construction and transfection

The murine short hairpin RNA for NCoA3 (shNCoA3) and the scrambled sequence (control) were prepared using OriGene system following the manufacturer’s protocol and 3T3-L1 cells were transfected using Lipofectamine 2000 (Invitrogen Corp., USA). Three days after transfection, the cells were incubated in a selection medium containing 0.5µg/ml Puromycin (Invitrogen Corp., USA). After 14 days of selection, protein and mRNA expression were analyzed by quantitative real-time PCR (qPCR) and immunoblotting, respectively [6].

Patients and tissue specimens

MAT specimens (n = 36) were obtained from Institute of Medical Research Alfredo Lanari, School of Medicine, University of Buenos Aires. All samples correspond to MAT around breast tumors (n = 32) or non-malignant mammary lesions (n = 4) obtained during tumor extirpation or exploratory biopsies (less than 2 cm away from the invasive front or lesion), respectively. By pathological examination, breast tumors were determined as well as their molecular status (Estrogen Receptor (ER), Progesterone Receptor (PR) and Human Epidermal Growth Factor 2 (Her2)). Staging was performed according to the National Comprehensive Cancer Network (NCCN Guidelines Version 4.2017, Breast Cancer Staging)[13]. Meanwhile, patients with non-malignant lesions (mastopathy, mastitis or usual epithelial hyperplasia) were monitored for at least 6 years and no one has presented tumors up to date. Patients' ages were within the range of 53–94 years (median 75 years). Samples were classified according to NCoA3 mRNA expression levels as “low” or “high”. The cut-off was the average of NCoA3 levels obtained by qPCR in non-malignant mammary lesions.

The experimental protocol was approved by the Institutional Research Ethic Committee of Institute of Medical Research Alfredo Lanari and all procedures performed were in accordance with its ethical standards, National Laws, 1964 Helsinki declaration, and its later amendments. Informed consent was obtained from all patients.

For collection of CM, human explants of MAT were incubated in PBS plus 5mg/ml albumin (3 ml/g of tissue) for 30 min and centrifuged for 30 s at 400 g in order to reduce contamination with blood cells, soluble factors and tissue containing insufficient adipocytes to float [14]. Then, explants were cultivated in fresh DMEM/F12 free-serum medium for 24h and CM were collected for dot blot assay.

Animals

Animal manipulations and experimental protocols performed in this work were supervised and approved by the National University of Quilmes Institutional Animal Care and Use Committee, in accordance with the National Institutes of Health and, ARRIVE guidelines for the care and use of laboratory animals.

Adult (6–8 months, 25–30 gr) female C57-BL/6J mice were kept with ad libitum access to food and water, under 12 h:12 h light:dark conditions. To isolate the mammary glands, animals were sacrificed using 5% isoflurane inhalation (Piramal Healthcare, India), through an equipment of gas anesthesia (SurgiVet®, USA) and cervical dislocation. The mammary glands (#2–4) were surgically removed and explants were cultivated.

Murine explants of MAT were prepared as human explants. Then, explants were stimulated with BrCM or fresh free-serum medium for 24 h and processed for RNA extraction or immunohistochemistry. For collection of CM, explants were cultivated in fresh DMEM/F12 free-serum medium for 24h, and dot blot assay was performed.

Western blot analysis

Western blots were performed as previously described [12]. For experiments where NCoA3 expression was determined, 3T3-L1-derived adipocytes were stimulated with T-47D BrCM or not and then were harvested and treated with lysis buffer plus protease inhibitors. To validate the knockdown efficiency of shNCoA3, protein extracts were obtained from different passages after initial selection.

In both cases, samples were separated on 7% SDS-PAGE and membranes were incubated with anti-NCoA3 or anti-Tubulin antibodies overnight at 4ºC in T-TBS plus 0.5% bovine serum albumin. Subsequently, washed membranes were incubated for 1 h with HRP-conjugated secondary antibody (VECTOR). Proteins were detected using luminol reagent (Santa Cruz Biotechnology, USA). Densitometry units for each band were obtained by NIH-ImageJ software and normalized to anti-Tubulin densitometry values.

Immunohistochemical and Immunofluorescence staining

Paraffin-embedded tissue blocks from the Department of Pathological Anatomy were selected taking into account the proximity to the tumor front (less than 2 cm). Human and murine blocks were sectioned, deparaffinized in xylene and rehydrated for immunohistochemical (IHC) staining [12]. Antigen retrieval was performed using sodium citrate. The sections were then incubated in H₂O₂ (3%) for 10 min, blocked in horse serum for 60 min and incubated with anti-NCoA3 or anti-p-p65 (Ser536) antibodies (Santa Cruz Biotechnology, 1:100) at 4° C overnight. After incubation with a universal biotinylated secondary antibody for 60 min, the specimens were incubated with H₂O₂-diaminobenzidine until the desired staining intensity was observed. The sections were counterstained with hematoxylin, dehydrated, mounted and observed under Olympus BX51 microscope.

Murine MAT sections were stained with hematoxylin and Crown-like structures (CLS) were identified as adipocytes with at least two infiltrating immune cells (macrophage morphology) around the cell and the number of CLS per field was counted.

For immunofluorescence assays, 3T3-L1-derived adipocytes were seeded in 24-well plates on 12 mm glass coverslips. After 24 h, cells were stimulated with T-47D BrCM or serum-free medium for 1 h, then the cells were fixed with 3% formaldehyde and 0.02% glutaraldehyde for 15 min. Incubation with primary antibody against anti-p-p65 was performed 2 h at room temperature. Then, cells were washed and incubated with a TRITC-labeled secondary antibody (Jackson ImmunoResearch Laboratories Inc., 1:250) for 1 h, washed with PBS, counterstained with Hoechst 33342 (Santa Cruz Biotechnology) and mounted on glass slides with PBS/Glycerol 1:1 solution. Cells were analyzed with an Olympus BX51 fluorescence microscope. Images were taken with a digital camera and 100 cells per field were analyzed with NIH-ImageJ software.

Primers and qPCR

Total RNA was isolated using the TRIzol® protocol (Invitrogen) adapted for AT. Briefly, approximately 0.5 g of tissue was homogenized using a Tissumizer homogenizer (Tekmar Company, USA) with 1 ml TRIzol® reagent. The extracts were spun at 12000 g for 10 min at 4°C; the top layer of excess lipid was removed and solution containing RNA was transferred to a fresh tube [15]. For 3T3-L1 adipocytes standard TRIzol® protocol was used [6].

Reverse transcription was carried out using SuperScript II kit (Invitrogen) according to the manufacturer’s instructions. qPCR was performed using Homo sapiens (hs) and Mus musculus (mm) sequence-specific primers for:

− hsNCoA3: Fw 5’- AAGTGAAGAGGGATCTGGA − 3’

Rv 5’- CAGATGACTACCATTTGAGG − 3’

− hsCyclophilin A (hCyA): Fw 5’- ATGCTGGACCCAACACAAAT − 3’

Rv 5’- TCTTTCACTTTGCCAAACACC − 3’

− mmNCoA3: Fw 5’- ACATGGTGCATATGAACAGC − 3’

Rv 5’- GATGTCAGCAGTATTTCTGATCG − 3’

− mmTNF: Fw 5’- GACAGTGACCTGGACTGTGG − 3’

Rv 5’- GAGACAGAGGCAACCTGACC − 3’

-mmMCP1: Fw 5’-CGGCTGGAGCATCCACGT-3’

Rv 5’-TGGGGTCAGCACAGACCT-3’

− mmCyA: Fw 5’-CCACCGTGTTCTTCGACATC − 3’

Rv 5’-GCTCGAAAGTTTTCTGCTGT − 3’

For gene expression quantitative analysis, 2^− ΔΔCT method was used.

Dot blot assay

Conditioned medium was spotted onto nitrocellulose membrane with Bio-Dot Microfiltration Apparatus (Bio-Rad, USA). The membranes were blocked in 10% skim milk, incubated with anti-TNF (Santa Cruz Biotechnology, USA) or anti-MCP1 (ThermoFisher Scientific, Austria) and then incubated for 1 h with horseradish peroxidase-conjugated secondary antibody. Proteins were visualized by autoradiography using the luminol reagent (Santa Cruz Biotechnology, USA). The densitometry values for each band were determined by the NIH-ImageJ program and normalized to total RNA. All samples were performed in triplicate for mice explants and 3T3-L1 adipocytes.

Analysis of gene expression values from data repositories

In order to compare NCoA3 and different adipokines expression levels between MAT surrounding normal mammary gland or breast cancer, we analyzed values retrieved from two databases from two different public repositories: GSE66744 database, GPL13112 platform (Illumina HiSeq 2000 for mus musculus) from the NCBI’s databank, Gene Expression Omnibus (GEO), [16], and E-MTAB-8638 database, A-MEXP-2072 platform (Illumina HumanHT12 V4.0) from the EBI’s databank, ArrayExpress [17]. Reads per kilobase million (RPKM) and quantile normalized expression values were retrieved, respectively. Heatmap was performed by heatmap function, limma R package

Statistical analysis

For comparing two groups, the significance differences between means were determined using t test. For multiple comparisons, ANOVA test was used and Tukey post-test was performed for paired observation in case there was homogeneity of variance and normal distribution of data, otherwise the non-parametric Kruskal-Wallis test and Dunn’s multiple comparison post-test were assessed. Differences in a given categorical variable between groups were assessed using Fisher’s exact test for patients. For bioinformatics analysis, t test for paired samples and Chi square test for RPKM values were performed. Pearson’s r correlation coefficient was calculated for determining correlation. A two-tailed p value less than 0.05 was considered statistically significant.

NCoA3 expression levels in adipocytes are modulated in the breast cancer context

First, NCoA3 expression levels were evaluated in murine MAT in the presence or absence of CM from breast cancer cells and, in samples from patients with breast cancer.

We found that MAT from murine mammary gland showed a significant increase in the coactivator mRNA expression after stimulation with BrCM from T-47D (Fig. 1a). We also observed that NCoA3 mRNA levels correlate with its protein levels measured by IHC and the protein localization was mainly in the nuclei (Fig. 1b).

MAT specimens from patients with breast cancer were analyzed by qPCR and 59% of them showed high NCoA3 mRNA levels (Table 1, Fig. 1c). We also confirmed the correlation between this result and the protein levels by IHC and, as in mice, the subcellular location of NCoA3 was nuclear (Fig. 1d).

Table 1

Correlation between NCoA3 expression in mammary adipose tissue and clinical-pathological features in breast cancer
Variable	Number of patients	RAC3 expression (qPCR)		Fisher's P value
Variable	Number of patients	low	high	Fisher's P value
Sample type				0.3033
non-tumoral lesion	4	3 (75%)	1 (25%)
malignant breast cancer	32	13 (41%)	19 (59%)
Tumor size				1.000
T < 1	21	9 (43%)	12 (57%)
T > 1	10	4 (40%)	6 (60%)
Nodal status				0.2475
N0	14	7 (50%)	7(50%)
N1-3	12	3 (25%)	9 (75%)
Clinical stage				0.413
I	7	4 (57%)	3 (43%)
II-III	24	9 (37%)	15 (63%)
ER status				0.1469
positive	19	10 (53%)	9 (47%)
negative	13	3 (23%)	10 (77%)
PR status				0.0317
positive	18	10 (56%)	8 (44%)
negative	13	2 (15%)	11 (85%)
Her2 status				0.6236
positive	5	1 (20%)	4 (80%)
negative	26	11 (42%)	15 (58%)

Then, we investigated the association between NCoA3 expression levels in TA adjacent to breast tumors from patients and molecular markers or clinical stages. We observed that patients with molecular characteristics related to the worst prognosis (triple negative (TN) for ER, PR and Her2) or worse clinical stages showed high NCoA3 mRNA levels in adjacent MAT (Fig. 1e-f). In particular, MAT adjacent to tumors that are negative for PR expressed high levels of NCoA3. However, we did not detect this association with the expression of ER or Her2 (Table 1).

In summary, more than a half of the studied patients showed upregulation of NCoA3 expression in cancer-associated adipose tissue and the corresponding patients had a poorer prognosis than the rest, which was remarkable. Furthermore, ex vivo murine model showed the modulation of this coactivator because of tumoral factors and this modulation seems to depend on the phenotypic characteristics of the breast tumor.

NCoA3 expression is correlated with NF-κB activation and the expression of pro-inflammatory markers in adipocytes in a tumor context

NF-κB is a key transcription factor activated by a wide range of stimuli such as cytokines and mitogens, and regulates the expression of these molecules and others like adhesion molecules, metalloproteinases and apoptotic genes [18]. Therefore, its activation in cancer-associated adipocytes may exert several effects on the tumor microenvironment.

Since NCoA3 expression in tumor-surrounding human MAT and murine MAT stimulated with BrCM was mainly localized in the nucleus and we have previously demonstrated that NCoA3 is a NF-κB coactivator [19], we investigated NF-κB activation using immunodetection of the phosphorylated form of p65 subunit [20].

We observed that murine MAT stimulated with T-47D BrCM showed a qualitatively greater labeling of p-p65 by IHC, demonstrating a higher activation of this transcription factor induced by tumor-secreted molecules, than in their absence or with the MDA-MB-231 BrCM (Fig. 2a). Even more, we found that human MAT in the tumoral context expressing high levels of NCoA3 showed an increase in the nuclear staining of p-p65 (Fig. 2b).

The fact that adipocytes in tumor context expressed high levels of NCoA3 and had an increased degree of NF-κB activation suggests that gene expression in adipocytes changes to a pro-tumoral and inflammatory profile. Therefore, we then studied whether the increase in NF-κB activation in adipocytes was accompanied by the expression of its target genes, TNF and MCP1 [18, 21].

We found that BrCM increased the mRNA expression of the cytokine TNF, in murine explants (Fig. 2c). Remarkably, T-47D BrCM triggered higher TNF secretion than MDA-MB-231 BrCM (Fig. 2d). Furthermore, we observed in breast cancer patients that higher secretion of TNF was mainly associated with samples expressing high levels of NCoA3 (Fig. 2e).

We next studied MCP1 expression in the both models and we found that, as it happened with TNF, the BrCM from T-47D induced MCP1 expression in murine MAT (Fig. 2f). Moreover, in MAT from breast cancer patients, we evidenced a positive association between NCoA3 expression levels and the secretion of MCP1 (Fig. 2g). As MCP1 is a chemokine responsible for recruiting macrophages and it has been described that during local white adipose tissue inflammation, it can be observed the presence of dead adipocytes surrounded by macrophages which form CLS [22], we decided to analyze the number of CLS in Hematoxylin-stained sections. A major percentage of CLS was observed per field in MAT expressing high levels of NCoA3 (Fig. 2h-i).

Altogether, these results suggest that not only is there a positive association between the degree of NF-κB activation and NCoA3 expression levels but also this is correlated with the levels of TNF, MCP1 and the presence of CLS. Remarkably, all of these features are indicators of inflammation.

Silencing of NCoA3 and NF-kB inhibition in 3T3-L1 adipocytes reverse adipokines expression

The results presented heretofore show a relation between the expression levels of NCoA3 and inflammation markers in MAT in tumoral context. In order to deeply study this association, we used adipocytes derived from the 3T3-L1 cell line. As we did not find changes in the target genes expression analyzed with MDA-MB-231 BrCM, we only evaluated the effects of T-47D cells BrCM stimulation. We detected an increase in the NCoA3 expression, both at the mRNA and protein levels (Fig. 3a-b and Supplementary Fig. S1 online). Moreover, T-47D BrCM induced phosphorylation and nuclear localization of the p65 subunit of NF-κB in 3T3-L1 adipocytes, demonstrating increased activation of this transcription factor. This effect was inhibited in the presence of SSZ (Fig. 3c-e).

So as to explore the potential role of NCoA3, we used 3T3-L1 cells transfected with a shNCoA3 or control plasmid (control). qPCR and Western blot analysis were performed to validate the knockdown efficiency (Fig. 3f-g and Supplementary Fig. S1 online).

First, we observed that 3T3-L1 stimulated with T-47D BrCM secreted more TNF and expressed more MCP1 than the basal condition, mimicking the results obtained in murine explants (Fig. 3h-i). Second, to confirm the role of NF-κB activation in the regulation of TNF and MCP1 induced by T-47D BrCM, 3T3-L1 adipocytes were pre-incubated with SSZ 0.2mM before stimulation with T-47D BrCM. We found that both the levels of TNF and MCP1 were diminished in the presence of SSZ, confirming that the expression of these genes is regulated by NF-κB. Furthermore, results showed that NCoA3 silencing reverted the effects exerted by the BrCM from T-47D cells on TNF and MCP1 expression (Fig. 3h-i).

Hence, we confirmed that the signals released by tumor cells are capable of regulating the expression of NCoA3 in adipocytes and this increased expression was accompanied by the activation of the transcriptional factor NF-κB. Together, these molecules would be responsible for the modulation of inflammatory mediators, such as TNF, MCP1, among others.

MAT adjacent to tumors expressing high NCoA3 showed an inflammatory profile

We sought to validate the results obtained in our cohort of breast cancer patients, so that we analyzed the records from the E-MTAB-8638 dataset [17]. In accordance with our data, we observed an increase of NCoA3 mRNA expression in MAT adjacent to human breast tumors (adMAT) with respect to the distant MAT (dMAT) (Fig 4a). Even more, high expression of NCoA3 was associated with increased expression of several inflammatory molecules in adMAT (Fig 4b). There was also a positive correlation between NCoA3 expression and CD68, a macrophage marker (Fig. 4c). Additionally, we analyzed the GSE66744 dataset [16] and the MAT adjacent to tumor in mice injected with a human breast cancer cell line, showed increased expression levels of Ncoa3, Ccl2 (encoding MCP1), Tnf and Rela (encoding p65) with respect to normal MAT (Fig. 4d). Furthermore, Il1 and Il6, other pro-inflammatory cytokines activated by NF-kB were also increased in these mice.

All these results not only do confirm the modulation of NCoA3 in the adipose tissue adjacent to breast tumors, but also these increased levels correlate with different inflammatory markers that favor tumor progression.

Several works have shown that stromal cells and their secreted factors can exert different and diverse effects on tumor development, influencing proliferation, invasion, and metastasis processes [23]. Experiments performed with conditioned media from adipocytes evidenced that secreted proteins have effects more pronounced on breast cancer proliferation than factors released by other cell types [24]. However, it is unknown which transcriptional factors and cofactors regulate adipokine expression in tumor microenvironment.

In this work we have demonstrated for the first time that NCoA3 coactivator expression in breast tumor-associated adipocytes is upregulated with respect to normal fat cells, according to the data obtained in all the used models: MAT adjacent to human breast tumors, murine MAT and cell line-derived adipocytes stimulated with BrCM and also, bioinformatics analysis of public datasets.

Previous studies have described a role for NCoA3 in adipose tissue at normal physiological conditions [25, 26] and we have recently demonstrated that NCoA3 levels diminish during adipogenesis [6], revealing the importance of this coactivator in adipocyte behavior.

In breast cancer, it was reported that the NCoA3 gene is amplified in 5–10% of human tumors and its protein is overexpressed in 60% of tumors [7, 27]. In the breast cancer context, our results showed that 59% of MAT adjacent to tumors overexpressed NCoA3, suggesting that the expression of this coactivator gives a growth advantage to mammary tumors and the surrounding adipose tissue itself.

Furthermore, the correlation of tumoral NCoA3 expression levels and breast cancer molecular markers is contradictory, however, breast cancer patients expressing high levels of tumoral NCoA3 showed poor prognosis and chemoresistance [28–31]. Despite our cohort size, we did not find an association between adipose NCoA3 levels and breast cancer molecular markers in accordance to literature [28, 30].

In our ex vivo model, we observed T-47D BrCM upregulates NCoA3 expression in adipocytes more than MDA-MB-231. Interestingly, NCoA3 expression levels in the breast cancer cell line T-47D are greater than MDA-MB-231 cells [32, 33]. Therefore, an association between NCoA3 levels in breast cancer cell lines and adipocytes is suggested but further studies are needed.

The differential expression of NCoA3 between T-47D and MDA-MB-231 cell lines could lead them to different secretomes. In fact, it has been evidenced these mammary tumor cell lines secrete distinct cytokine patterns [34]. For instance, T-47D cells release more TNF than MDA-MB-231 and, indeed, we have previously demonstrated that this cytokine stimulates T-47D proliferation [11]. Moreover, we have formerly reported TNF is able to induce NCoA3 expression via NF-κB activation [5] and this could explain high expression of this coactivator in adipocytes stimulated by T-47D BrCM. However, due to the heterogeneity of BrCM, other molecules should not be excluded.

Although in our model we used BrCM from human cell lines to stimulate murine adipocytes and the cross-reaction should not be guaranteed, the capability of human TNF to bind and stimulate murine p55 TNF receptor has been already validated [35].

Remarkably, although we observed that one of the non-tumoral samples expressed high NCoA3 levels, it was from a patient diagnosed with mastitis, characterized by localized inflammation. That suggests the relevance of the inflammatory context to NCoA3 expression.

NCoA3 was first identified as a coactivator for nuclear receptors [36], and later for other transcription factors, such as NF-κB [19]. In breast cancer, constitutive NF-κB activation is associated with poor prognosis and chemoresistance, as well as with hormone-independent breast cancer progression [37]. Furthermore, in a hormone-dependent breast cancer model, we have previously demonstrated that NF-κB, ERα and NCoA3 are part of the same protein complex and this induces cyclin D1 expression [11].

Considering NF-κB has an important role in breast cancer development and its activity is enhanced by NCoA3, we studied the implication of this transcriptional factor in our models. We found that adipocytes in tumoral context had greater NF-κB activation, which is consistent with the upregulation of NCoA3 expression levels. Even more, T-47D-associated adipocytes exhibited an increase in the expression of target genes of NF-κB, encoding inflammatory molecules.

Previous works have shown that adipocytes can produce cytokines such as TNF and MCP1, in the inflammatory context [4, 38]. Moreover, we found that fat cells in tumoral context not only did express more NCoA3 but also they were capable of secreting high TNF levels. This cytokine could act on adipocytes itself, in an autocrine way, establishing a positive feedback loop in addition to tumor-released TNF.

The dysregulated secretion of pro-inflammatory cytokines and chemokines, such as TNF and MCP1, by adipocytes can lead to recruitment of macrophages and an increase in CLS [4, 22], as we observed in our results. CLS is a hallmark of chronic adipose tissue inflammation, and increased numbers of CLS in the breast have been associated with increased risk and poor prognosis of breast cancer patients in both obese individuals and those with normal body mass index [22]. CLS formation has been shown to enhance the expression of inflammatory cytokines/chemokines, further activating macrophages for clearing apoptotic adipocytes, generating positive feedback.

Of interest, in our results we observed that NCoA3 silencing or NF-κB inhibition reverted in a similar manner the cytokine expression, suggesting that these molecules are co-dependent in this process. These results are relevant considering that the evaluated cytokines favor tumor progression [11, 39] and contribute to the comprehension of the interaction between fat cells and breast cancer.

In conclusion, this is the first time that the expression of NCoA3 is reported to be increased in MAT adjacent to breast cancer. In Fig. 5, we propose that breast cancer-secreted factors, such as TNF, induce activation of NF-κB and upregulation of NCoA3 expression, which in turn power up the transcription of TNF and MCP1. Once these inflammatory cytokines are released, they could have effects on fat cells themself, breast cancer cells and also macrophages, propitiating an inflammatory context that promotes tumor progression and worse prognosis.

As adipocytes are involved in the development and progression of breast cancer, the modulation of NCoA3 levels in MAT and its possible role in inflammatory processes related to cancer deserve to be further investigated to improve future tumoral treatments.

Ethical Approval and Consent to participate

All procedures performed in studies involving human participants were in accordance with the ethical standards of the Institutional Research Ethic Committee of Institute of Medical Research Alfredo Lanari, National Law No 25326 “Protección de datos personales”, Res. 1480/11, Art 58 and 59 from Commercial and civil code from Argentine and, the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent was obtained from all individual participants included in the study.

Consent for publication

The authors affirm that human research participants provided informed consent for publication.

Data availability

The datasets generated during the current study are available from the corresponding author on reasonable request. The analyzed datasets retrieved from public repositories are openly available at the GEO (www.ncbi.nlm.nih.gov/geo/, accession number GSE66744) and Array Express (www.ebi.ac.uk/arrayexpress/, accession number E-MTAB-8638) websites.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Funding

This work has been supported by grants from CONICET, and National Agency for Scientific and Technological Promotion, Argentina (ANPCyT)

Author contributions

María Cecilia Lira, Francisco Rosa, Ignacio Aiello, Alejandra Palma and Mileni Soares Machado: Investigation, methodology, data curation, formal analysis. María Cecilia Salazar Güemes, Leonardo Paz, Silvia Burlando, Natalia Paladino and Pablo J Azurmendi: Data curation and validation. Mónica A. Costas: Funding acquisition, supervision and writing-review and editing. María Fernanda Rubio: Conceptualization, project administration, supervision and writing-original draft. All authors read and approved the final version of the manuscript.

Acknowledgments

Not applicable

D.P. Rose and L. Vona-Davis, Maturitas 66, 33-38 (2010) doi: S0378-5122(10)00046-0 [pii] 10.1016/j.maturitas.2010.01.019
L. Bochet, A. Meulle, S. Imbert, B. Salles, P. Valet and C. Muller, Biochem Biophys Res Commun 411, 102-106 (2011) doi: S0006-291X(11)01081-3 [pii] 10.1016/j.bbrc.2011.06.101
D. Hanahan and R.A. Weinberg, Cell 144, 646-674 (2011) doi: S0092-8674(11)00127-9 [pii]10.1016/j.cell.2011.02.013
L.M. Berstein, A.Y. Kovalevskij, T.E. Poroshina, A.V. Kotov, I.G. Kovalenko, E.V. Tsyrlina, E.E. Leenman, S.Y. Revskoy, V.F. Semiglazov and K.M. Pozharisski, Int J Cancer 121, 514-519 (2007) doi: 10.1002/ijc.22552
C.V. Alvarado, M.F. Rubio, P.N. Fernandez Larrosa, L.C. Panelo, P.J. Azurmendi, M. Ruiz Grecco, G.A. Martinez-Noel and M.A. Costas, FEBS Open Bio 4, 450-457 (2014) doi: 10.1016/j.fob.2014.04.009S2211-5463(14)00045-X [pii]
M.C. Lira, F.D. Rosa, L.C. Panelo, M.A. Costas and M.F. Rubio, Cell Death Discov 5, 20 (2018) doi: 10.1038/s41420-018-0085-y85 [pii]
S.L. Anzick, J. Kononen, R.L. Walker, D.O. Azorsa, M.M. Tanner, X.Y. Guan, G. Sauter, O.P. Kallioniemi, J.M. Trent and P.S. Meltzer, Science 277, 965-968 (1997)
L. Liao, S.Q. Kuang, Y. Yuan, S.M. Gonzalez, B.W. O'Malley and J. Xu, J Steroid Biochem Mol Biol 83, 3-14 (2002) doi: S0960076002002546 [pii]
G.P. Colo, M.F. Rubio, I.M. Nojek, S.E. Werbajh, P.C. Echeverria, C.V. Alvarado, V.E. Nahmod, M.D. Galigniana and M.A. Costas, Oncogene 27, 2430-2444 (2008) doi: 1210900 [pii] 10.1038/sj.onc.1210900
P.N. Fernandez Larrosa, C.V. Alvarado, M.F. Rubio, M. Ruiz Grecco, S. Micenmacher, G.A. Martinez-Noel, L. Panelo and M.A. Costas, Cancer Sci 103, 2064-2071 (2012) doi: 10.1111/cas.12019
M.F. Rubio, S. Werbajh, E.G. Cafferata, A. Quaglino, G.P. Colo, I.M. Nojek, E.C. Kordon, V.E. Nahmod and M.A. Costas, Oncogene 25, 1367-1377 (2006) doi: 1209176 [pii] 10.1038/sj.onc.1209176
M.F. Rubio, M.C. Lira, F.D. Rosa, A.D. Sambresqui, M.C. Salazar Guemes and M.A. Costas, Cancer Cell Int 17, 111 (2017) doi: 10.1186/s12935-017-0483-x483 [pii]
W.J. Gradishar, B.O. Anderson, R. Balassanian, S.L. Blair, H.J. Burstein, A. Cyr, A.D. Elias, W.B. Farrar, A. Forero, S.H. Giordano, M.P. Goetz, L.J. Goldstein, S.J. Isakoff, J. Lyons, P.K. Marcom, I.A. Mayer, B. McCormick, M.S. Moran, R.M. O'Regan, S.A. Patel, L.J. Pierce, E.C. Reed, K.E. Salerno, L.S. Schwartzberg, A. Sitapati, K.L. Smith, M.L. Smith, H. Soliman, G. Somlo, M.L. Telli, J.H. Ward, R. Kumar and D.A. Shead, J Natl Compr Canc Netw 16, 310-320 (2018) doi: 10.6004/jnccn.2018.0012
J.N. Fain, B.M. Tagele, P. Cheema, A.K. Madan and D.S. Tichansky, Obesity (Silver Spring) 18, 890-896 (2009) doi: oby2009335 [pii]10.1038/oby.2009.335
J.N. Fain, B. Buehrer, S.W. Bahouth, D.S. Tichansky and A.K. Madan, Metabolism 57, 1005-1015 (2008) doi: S0026-0495(08)00090-5 [pii]10.1016/j.metabol.2008.02.019
I.V. Chivukula, D. Ramskold, H. Storvall, C. Anderberg, S. Jin, V. Mamaeva, C. Sahlgren, K. Pietras, R. Sandberg and U. Lendahl, Breast Cancer Res 17, 109 (2015) doi: 10.1186/s13058-015-0616-x 10.1186/s13058-015-0616-x [pii]
I. Miran, D. Scherer, P. Ostyn, C. Mazouni, F. Drusch, M. Bernard, E. Louvet, J. Adam, M.C. Mathieu, M. Haffa, J.P. Antignac, B. Le Bizec, P. Vielh, P. Dessen, H. Perdry, S. Delaloge and J. Feunteun, Front Oncol 10, 1506 (2020) doi: 10.3389/fonc.2020.01506
S. Ghosh and M. Karin, Cell 109 Suppl, S81-96 (2002) doi: S0092867402007031 [pii]
S. Werbajh, I. Nojek, R. Lanz and M.A. Costas, FEBS Lett 485, 195-199 (2000) doi: S0014-5793(00)02223-7 [pii]
C.Y. Sasaki, T.J. Barberi, P. Ghosh and D.L. Longo, J Biol Chem 280, 34538-34547 (2005) doi: M504943200 [pii]10.1074/jbc.M504943200
J.N. Fain and A.K. Madan, Int J Obes (Lond) 29, 1299-1307 (2005) doi: 10.1038/sj.ijo.0803032
N.M. Iyengar, P.G. Morris, X.K. Zhou, A. Gucalp, D. Giri, M.D. Harbus, D.J. Falcone, M.D. Krasne, L.T. Vahdat, K. Subbaramaiah, M. Morrow, C.A. Hudis and A.J. Dannenberg, Cancer Prev Res (Phila) 8, 349-358 (2015) doi: 10.1158/1940-6207.CAPR-14-0243
B. Dirat, L. Bochet, M. Dabek, D. Daviaud, S. Dauvillier, B. Majed, Y.Y. Wang, A. Meulle, B. Salles, S. Le Gonidec, I. Garrido, G. Escourrou, P. Valet and C. Muller, Cancer Res 71, 2455-2465 (2011) doi: 71/7/2455 [pii]10.1158/0008-5472.CAN-10-3323
P. Iyengar, T.P. Combs, S.J. Shah, V. Gouon-Evans, J.W. Pollard, C. Albanese, L. Flanagan, M.P. Tenniswood, C. Guha, M.P. Lisanti, R.G. Pestell and P.E. Scherer, Oncogene 22, 6408-6423 (2003) doi: 10.1038/sj.onc.12067371206737 [pii]
J.F. Louet, A. Coste, L. Amazit, M. Tannour-Louet, R.C. Wu, S.Y. Tsai, M.J. Tsai, J. Auwerx and B.W. O'Malley, Proc Natl Acad Sci U S A 103, 17868-17873 (2006) doi: 0608711103 [pii]10.1073/pnas.0608711103
Z. Wang, C. Qi, A. Krones, P. Woodring, X. Zhu, J.K. Reddy, R.M. Evans, M.G. Rosenfeld and T. Hunter, Cell Metab 3, 111-122 (2006) doi: S1550-4131(06)00026-X [pii]10.1016/j.cmet.2006.01.002
M.I. Torres-Arzayus, J. Font de Mora, J. Yuan, F. Vazquez, R. Bronson, M. Rue, W.R. Sellers and M. Brown, Cancer Cell 6, 263-274 (2004) doi: 10.1016/j.ccr.2004.06.027S1535610804002193 [pii]
E. Burandt, G. Jens, F. Holst, F. Janicke, V. Muller, A. Quaas, M. Choschzick, W. Wilczak, L. Terracciano, R. Simon, G. Sauter and A. Lebeau, Breast Cancer Res Treat 137, 745-753 (2013) doi: 10.1007/s10549-013-2406-4
H. Iwase, Y. Omoto, T. Toyama, H. Yamashita, Y. Hara, H. Sugiura and Z. Zhang, Breast Cancer Res Treat 80, 339-345 (2003) doi: 10.1023/A:1024916126532
K. Lee, A. Lee, B.J. Song and C.S. Kang, World J Surg Oncol 9, 139 (2011) doi: 10.1186/1477-7819-9-139
M. Weiner, L. Skoog, T. Fornander, B. Nordenskjold, D.C. Sgroi and O. Stal, Ann Oncol 24, 1994-1999 (2013) doi: 10.1093/annonc/mdt159
J. Kao, K. Salari, M. Bocanegra, Y.L. Choi, L. Girard, J. Gandhi, K.A. Kwei, T. Hernandez-Boussard, P. Wang, A.F. Gazdar, J.D. Minna and J.R. Pollack, PLoS One 4, e6146 (2009) doi: 10.1371/journal.pone.0006146
L.J. Wong, P. Dai, J.F. Lu, M.A. Lou, R. Clarke and V. Nazarov, BMC Cancer 6, 111 (2006) doi: 10.1186/1471-2407-6-111
K. Chen, L. Satlof, G. Stoffels, U. Kothapalli, N. Ziluck, M. Lema, L. Poretsky and D. Avtanski, Data Brief 28, 104798 (2020) doi: 10.1016/j.dib.2019.104798
R.A. Smith, M. Kirstein, W. Fiers and C. Baglioni, J Biol Chem 261, 14871-14874 (1986)
H. Li, P.J. Gomes and J.D. Chen, Proc Natl Acad Sci U S A 94, 8479-8484 (1997)
D.K. Biswas, Q. Shi, S. Baily, I. Strickland, S. Ghosh, A.B. Pardee and J.D. Iglehart, Proc Natl Acad Sci U S A 101, 10137-10142 (2004) doi: 10.1073/pnas.04036211010403621101 [pii]
H. Hauner, Proc Nutr Soc 64, 163-169 (2005) doi: S0029665105000200 [pii]
R. Divella, R. De Luca, I. Abbate, E. Naglieri and A. Daniele, J Cancer 7, 2346-2359 (2016) doi: 10.7150/jca.16884jcav07p2346 [pii]

No competing interests reported.

SupplementaryFigS1.pdf

Download PDF

Version 2

posted

You are reading this latest preprint version

NCoA3 upregulation in breast cancer-associated adipocytes elicits an inflammatory profile

Status:

Version 2

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Materials And Methods

Cell culture

Plasmid construction and transfection

Patients and tissue specimens

Animals

Western blot analysis

Immunohistochemical and Immunofluorescence staining

Primers and qPCR

Dot blot assay

Analysis of gene expression values from data repositories

Statistical analysis

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 2