Background
With over 2 million new cases in 2020, breast cancer (BC) is the most common cancer occurring in women and the first most common type of tumor overall (source World Cancer Research Fund). BC represents a heterogeneous disease classified in several complex subsets on the basis of cellular compositions, molecular alterations, and clinical behavior.
Molecular subtyping of BC is now based on classical immunohistochemistry markers such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) that led to the distinction between luminal (A and B), basal, and HER2-positive classes [
1,
2]. Luminal A are the most prevalent type of BC (58.5%) and are ER
+/PR
+/HER2
− and Ki-67
low; Luminal B account for around 14% of BC and are ER
+/PR
+/HER2
− and Ki-67
high; HER2-positive BC represent 11.5% of BC and are ER
−/PR
−/HER2
+. Finally, the 15–20% of BCs are basal-like and are referred as triple negative breast cancers (TNBCs) due to the absence of classical molecular markers (ER
−/PR
−/HER2
−) [
3,
4]. Due to its aggressive biological behavior and the lack of potential markers and targets, TNBC represents the most dangerous BC subtype, with the poorest prognosis and outcome [
5].
The soluble pattern recognition receptor Long Pentraxin-3 (PTX3) is a member of the pentraxin family and is produced locally in response to inflammatory signals as a functional component of the innate immunity. PTX3 exerts non-redundant functions in various physio-pathological conditions and it has been described to be involved in tumor cell proliferation, angiogenesis, metastatic dissemination and cancer immune-modulation [
6‐
8]. As a secreted protein, PTX3 can be produced and released by both tumor and stroma cells, depending on tumor type [
9]. Different studies reported the role of PTX3 as an oncosuppressor acting through the modulation of tumor-associated inflammation [
10] and/or by blocking pro-tumor growth factors like various members of the FGF family [
9,
11]. Indeed, tumor and/or host PTX3 overexpression can inhibit FGF-driven epithelial-to-mesenchymal transition (EMT) and tumor/metastatic burden in melanoma models [
12] and hampers cancer growth in models of fibrosarcoma, prostate and bladder cancer [
6,
8,
13]. On the other hand, PTX3 has been shown to promote cell migration and invasion in some experimental tumor models, its expression levels being correlated with tumor progression in different human tumor types. For instance, high levels of PTX3 have been reported in all subtypes of human soft tissue liposarcoma [
14] as well as in pancreatic carcinoma cells and in advanced gastric cancer tissues where it promotes the migratory potential of tumor cells and macrophages recruitment [
15,
16]. Similarly, PTX3 levels in cervical cancers and gliomas appear to correlate with tumor grade and severity in vitro and in patients [
17,
18]. However, even though the mechanisms by which PTX3 exerts its anti-tumor activity are at least partially known, the mechanisms by which PTX3 exerts its tumorigenic activity have still to be revealed.
Recent studies suggest a possible tumorigenic role of PTX3 also in BC. Indeed, high expression levels of
PTX3 have been found to be associated with EMT in high-grade ductal infiltrating carcinomas [
19]. Elevated expression of
PTX3 has been observed in distant bone metastases of BC and correlated with osteoclast formation, suggesting that PTX3 might be involved also in the osteolytic bone metastatic process in BC [
20]. In addition,
PTX3 expression has been shown to be regulated by PI3K and to foster tumor stem-like features and bad prognosis in basal-like TNBC [
21,
22]. However, several issues still remain to be addressed in order to clarify (i) if PTX3 is differentially expressed by the different BC subtypes, (ii) which is the main source of PTX3 (tumor or stromal cells), and (iii) what are the biological effects and the molecular mechanism(s) exerted by PTX3 in BC.
In this study we show that, if compared to the other BC subtypes, high levels of PTX3 are mainly found in TNBC, where
PTX3 transcript is predominantly expressed by tumor cells in respect to cells associated with tumor stroma/microenvironment. Also, in vitro and in vivo data show that PTX3 confers more aggressive biological features to TNBC cells, resulting in augmented tumor cell proliferation and growth. Importantly, we demonstrate that the pro-tumor activity of PTX3 is exerted
via the activation of the TLR4 pathway which is known to play a relevant role in TNBC aggressiveness [
23,
24]. Indeed, our findings reveal for the first time that TNBC cell aggressiveness is fostered by a PTX3/TLR4 autocrine loop of stimulation, and that its inhibition may represent a promising therapeutic approach for the treatment of the most dangerous BC subtype.
Materials and methods
Reagents and cell cultures
The TLR4 inhibitor TAK-242 was purchase from Selleckchem (Houston, TX, USA). Human MDA-MB-231, MDA-MB-468 cells were obtained from American Type Culture Collection (ATCC) and cultured in DMEM
plus 10% FBS; BT549 cells were obtained from American Type Culture Collection (ATCC) and cultured in RPMI
plus 10% FBS and 1 µg/mL of bovine insulin; murine E0771 cells, derived from a spontaneous mammary tumour in a C57BL/6 mouse were kindly provided by R. Giavazzi (Istituto M. Negri, Milan, Italy) and cultured in DMEM
plus 20% FBS [
25].
For overexpression, breast cancer cells were infected with a pLentiPGK-Puro (Addgene Plasmid #19,070) lentiviral vector harbouring or not the full length human PTX3 cDNA (GenBank accession n° X63613). For silencing, cells were infected with lentiviral vector containing short-hairpin RNA (shRNA) targeting human PTX3 (TRCN0000436981 or TRCN0000430959) or a non-targeting/control sequence (SHC002V, Merck Millipore, Burlington, MA, USA). Transduced cells were selected with 1 µg/ml puromycin. Cells were authenticated by microsatellite genotyping before the starting of the project and periodically all along the project, maintained at low passage, returning to original frozen stocks every 3 to 4 months, and tested regularly for Mycoplasma negativity by PCR and DAPI staining.
Analyses on human samples
Breast cancer samples used for Western blot, RNAscope analyses and single cell RNA-seq were from different source. Western blot samples were from the institutional biobank of the University Hospital Liege Belgium and clinical data available are reported in Table
S1. TNBC samples used for RNAscope (cases #A-D) were from the Unit of Pathology (Spedali civili di Brescia, Italy). TNBC cases analysed in single cell RNA-seq are derived from the publication [
26] (see details below).
Single cell RNA-seq analysis
To examine PTX3 expression in breast cancer we reanalysed previously published single cell RNA-seq analysis involving 5 TNBC patients [
26]. The selected cases were all TNBC tumors with no pre-treatment, fulfilling the following histological criteria: staining by immunohistochemistry for estrogen receptor (< 1%) and progesterone receptor (< 1%), and fluorescence in situ hybridization analysis of HER2 amplification using the CEP-17 centromere control probe (ratio of HER2/CEP-17 < 2.2). Processed 10X Genomics (Pleasanton, CA, USA) data were downloaded from GEO repository (GSE148673). The data were imported into R computational environment (4.0) and then analysed using
Seurat 3.1 package using default parameters [
27].
In vitro assays
Cell Proliferation. Cells were seeded (104) in 48-well culture plates in complete medium, detached at different time points and counted using the MACSQuant Analyzer (Miltenyi Biotec, Bergisch Gladbach, Germany).
Clonogenic Assay. Five hundred cells were seeded in 6-well culture plates and incubated in complete growth medium until visible colonies were formed. Then, the supernatant was removed and cells stained with 0.1% crystal violet/20% methanol. Plates were photographed to count formed colonies using the ImageJ software. Finally, crystal violet staining was solubilized with 1% SDS solution to measure absorbance at 595 nm.
Soft Agar Assay. Cells (5 × 10
4) were suspended in 3ml of complete growth medium containing 0.3% agar and poured on to 2ml pre-solidified 0.6% agar in a 6-well plate. After 3 weeks of incubation, colonies were observed under a phase contrast microscope, photographed, and their area was measured using the ImageJ Software and the SA_NJ algorithm [
8].
Wound-Healing assay. Confluent cells were scraped with a 200 µl tip to obtain a 2-mm-thick denuded area. After 24 h, wounded monolayers were photographed and the width of the wounds was measured in 3-independent sites per group.
qPCR analysis
Total RNA was extracted using QIAzol reagent, treated with DNAse and 2 µg of total RNA were retro transcribed with MMLV-RT using random hexaprimers, cDNA was analyzed by quantitative PCR using primers specific for human or murine PTX3 (hPTX3: Forward primer: 5’-CATCTCCTTGCGATTCTGTTTTG-3’; reverse primer: 5’-CCCATTCCGAGTGCTCCTGA-3’). Housekeeping gene human GAPDH was detected for normalization (hGAPDH: Forward primer: 5’-GAAGGTCGGAGTCAACGGATT-3’; reverse primer: 5’-TGACGGTGCCATGGAATTTG-3’).
Western blot analysis
Cells and fresh frozen tumor tissues were homogenized in NP-40 lysis buffer (1% NP-40, 20 mM Tris–HCl pH 8, 137 mM NaCl, 10% glycerol, 2 mM EDTA, 1 mM sodium orthovanadate, 10 µg/mL aprotinin, 10 µg/mL leupeptin). Protein concentrations were determined using the Bradford protein assay (Bio-Rad Laboratories, Milano, Italy). Then, 30 µg protein/sample were separated by SDS-PAGE and blotted on a PVDF membrane. The following antibodies were used: anti-PTX3 (from B. Bottazzi, Humanitas Clinical Institute, Rozzano, Italy), anti-TLR4 (Bio-Rad), anti-phospho IRAK1 (Sigma-Aldrich, MO, USA), anti-phospho AKTser473 (Cell Signaling Technology, MA, USA), anti-phospho p65 (Santa Cruz Biotechnology, CA, USA). To normalize the amount of loaded proteins, all blots were probed with anti-β-actin (Sigma-Aldrich), anti-α-tubulin (Sigma-Aldrich), anti-GAPDH (Santa Cruz Biotechnology) or anti-HSC70 (Santa Cruz Biotechnology) antibodies. All primary antibodies were diluted 1:1000 and the secondary HRP-conjugated antibodies 1:5000. Chemiluminescent signal was automatically acquired by ChemiDoc™ Imaging System (Bio-Rad) at a final resolution of 62.2 pixel/mm2.
Genome-wide expression profiling (GEP)
GEP was performed on shNT/shPTX3 MDA-MB-231 and mock/PTX3 MDA-MB-468 cells. Total RNA was extracted using TRIzol Reagent according to manufacturer’s instructions (Invitrogen, Waltham, MA, USA). RNA integrity and the purity of the treated cells were assessed using a Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Hybridization to HuGene-2_1-st-v1 array strips (ThermoFisher Scientific, Waltham, MA, USA) was performed. Normalized data were imported into Partek® Genomic Suite® 6.6 software (Partek, Chesterfield, MO, USA). After quality controls, Analysis of variance (ANOVA) test was performed to assess the effects of PTX3 modulation on gene expression, comparing MDA-MB-231 shNT vs. MDA-MB-231 shPTX3 and MDA-MB-468 mock vs. MDA-MB-468 PTX3. A cut-off of p-value < 0.01 (FDR corrected) and Log2 fold change ± 2 was applied to select differentially expressed genes. Specific cellular pathways and biological networks modulated by differentially expressed genes were identified through the Core Analysis function in Ingenuity Pathway Analysis (IPA) software (QIAGEN, Hilden, Germany). To identify significantly enriched or depleted groups of genes involved in the same biological pathways, Gene Set Enrichment Analysis (GSEA) on GEP data was performed (
http://software.broadinstitute.org/gsea/index.jsp).
Five thousand cells were resuspended in DMEM/F-12 medium (GIBCO) containing 10 ng/ml basic Fibroblast Growth Factor (bFGF), 10 ng/ml Epidermal Growth Factor (EGF) and 2% of B27 supplement (Sigma-Aldrich) and plated into each well of 24-well Ultra-Low Attachment Plates (Corning, NY, USA). After 7 days of incubation, tumor spheres were counted and assayed for ALDH activity using the Aldefluor kit (Stemcell technologies, Vancouver, Canada) according to manufacturer’s instructions. ALDH-positive cell was quantified by cytofluorimetric analysis (MACSQuant Analyzer). Samples treated with the specific ALDH inhibitor diethylaminobenzaldehyde (DEAB) were used as controls to set the gates defining the ALDH-negative and the ALDH-positive regions.
Targeted quantitative analysis of secreted cytokines by Bio-Plex assay
The targeted quantitative analysis of secreted cytokines and chemokines in culture media was performed by using the Bio-Plex multiplex system (Bio-Rad) based on xMAP technology [
28]. Magnetic beads labeled with red and infrared fluorophores are coated with specific antibodies, thus allowing the simultaneous detection of multiple target analytes within one sample. Following reaction of beads with target analytes, detection is performed with a biotinylated antibody and phycoerythrin conjugated streptavidin. All steps were performed according to manufacturer’s instructions. Data were acquired using a Bio-Plex MAGPIX Multiplex Reader system (Bio-Rad).
In vivo studies
Animal Experiments were performed according to the Italian laws (D.L. 116/92 and following additions) that enforce the EU 86/109 Directive and were approved by the local animal ethics committee (OPBA, Organismo Preposto al Benessere degli Animali, Università degli Studi di Brescia, Italy).
Seven-week-old NOD/Scid female mice were injected orthotopically into the mammary fat pad with 4 × 106 MDA-MB-231 (shNT or shPTX3) and 8 × 106 MDA-MB-468 (mock or PTX3), while seven-week-old syngeneic C57BL/6 females were injected orthotopically with 5 × 105 E0771 (mock or PTX3) cells.
TAK-242 treatment (3 mg/Kg) was performed IP every other day when tumors were palpable. Tumors were measured with callipers and the volume was calculated according to the formula V = (D × d2)/2, where D and d are the major and minor perpendicular tumor diameters, respectively. At the end of the experimental procedure, tumors were surgically removed, weighed and paraffin embedded for immunohistochemical analysis.
Immunohistochemical and RNAscope analyses
For IHC on tumor xenograft samples, formalin-fixed, paraffin-embedded samples were sectioned at a thickness of 3 μm, dewaxed, hydrated, and stained with hematoxylin and eosin (H&E) or processed for immunohistochemistry with rabbit anti-human PTX3 (from B. Bottazzi, Humanitas Clinical Institute, Rozzano, Italy), rabbit anti-human phospho-Histone H3 (Merck Millipore), rabbit anti-CD44 (ThermoFisher Scientific), rabbit anti-phospho IRAK1 (Sigma-Aldrich) or rabbit anti-phospho p65 (Santa Cruz Biotechnology) antibodies. Positive signal was revealed by 3,3’-diaminibenzidine (Roche) stainings. Sections were finally counterstained with Carazzi’s hematoxylin before analysis by light microscopy. Images were acquired with the automatic high-resolution scanner Aperio System (Leica Biosystems, Wetzlar, Germany, EU) and image analysis was carried out using the open-source ImageJ software.
For RNAscope on TNBC patients’ samples, in situ hybridization was performed on FFPE TNBC biopsies using RNAscope® 2.5 HD Reagent Kit (RED 322,360, Advanced Cell Diagnostics (ACD), Hayward, CA). Sections were heated at 60 °C for 1 h and deparaffinized in fresh xylene. After dehydration in 100% ethanol, sections were incubated with the H2O2 for 10 min, target retrieval reagent for 15 min, and protease for 30 min (Pretreatment kit 322,330, ACD). The sections were then covered with a probe Hs-PTX3 (ref. 517,611) in the HybEZ oven (ACD) at 40 °C for 2 h. The Hs-PPIB probe was used as a control to ensure RNA quality. After probes’ hybridizations, sections were subjected to signal amplification using the HD 2.5 detection Kit, and hybridization signal was detected using a Fast- RED solution. Breast cancer biopsies were obtained from the institutional biobank of the University Hospital Liege Belgium, following the approval of the institutional ethical committee (reference number 2009/69).
Statistical analyses
Statistical analyses were performed using Prism 8 (GraphPad Software). Student’s t test for unpaired data (2-tailed) was used to test the probability of significant differences between two groups of samples. For more than two groups of samples, data were analyzed with a 1-way analysis of variance and corrected by the Bonferroni multiple comparison test. Tumor volume data were analyzed with a 2-way analysis of variance and corrected by the Bonferroni test. Differences were considered significant when p < 0.05 unless otherwise specified.
Discussion
In the last decades, considerable progresses have been made in BC treatment, especially through the introduction of targeted therapies against the signaling pathways governing cancer onset and progression [
33‐
35]. For instance, ER, PR and HER2 play key roles in the evolution of the majority of BCs, and selective targeting of these proteins has enabled the inhibition of their associated pathways, leading to a better prognosis for tumors that are positive for these receptors [
36‐
38]. At variance, TNBC, that accounts for 10–15% of all BCs, lacks of effective specific targeted therapies due to its aggressive clinical behavior and displays a risk death of 70% in the five years following diagnosis [
39‐
41]. For these reasons, the characterization of new molecular pathways and/or alternative druggable targets is of great interest for TNBC [
42].
PTX3 may exert anti-tumor or pro-tumor effects depending on tumor type and context [
43,
44]. In this frame, limited experimental evidence suggests that PTX3 may be endowed with a tumor-promoting activity in TNBC. Indeed, PTX3 has been shown to be a marker of poor prognosis in TNBC patients [
21]. However, the cellular source of PTX3, its biological effects and the mechanism(s) by which PTX3 exerts its pro-tumorigenic activity in TNBC have not been investigated so far.
Here we show that (i) among all subtypes of BC, PTX3 is highly expressed in the most aggressive TNBC subtype, (ii) the main source of PTX3 in TNBC patient-derived samples is represented by tumor cells rather than the stromal/immune component, and (iii) PTX3 expression by tumor cells fosters the tumorigenic potential of TNBC by activating a PTX3/TLR4 autocrine loop.
These findings indicate that PTX3 produced and secreted by tumor cells may act as an autocrine factor able to condition TNBC cell behavior. Relevant to this point, our data show for the first time that PTX3 exerts its pro-tumor activity in TNBC by activating TLR4/IRAK1/NF-kB signaling in tumor cells. Indeed, GSEA, Western blot and immunohistochemical analyses demonstrate that the TLR4 pathway is activated when PTX3 is expressed and inactivated when PTX3 is silenced in both in vitro and in vivo TNBC models. The strict correlation between PTX3 expression and TLR4 activation was confirmed by the fact that TLR4 blockade impairs PTX3-mediated tumorigenic activity in vitro and in vivo. Also, exogenous PTX3 is able to restore the activation of TLR4 pathway in PTX3 silenced TNBC cells.
PTX3 has been shown to exert a protective antifungal activity by directly activating TLR4 through the binding to myeloid differentiation protein 2 (MD-2), an accessory protein of TLR4 [
45]. In cancer, a PTX3/TLR4 interaction has been recently reported only for invasive melanoma [
46]. Our data extend these observations and strongly indicate that the PTX3/TLR4 system may play a non-redundant role in TNBC aggressiveness. Accordingly, TLR4 has been shown to be upregulated in human BC tissues [
47,
48] and constitutive activation of IRAK1 and NF-kB, key downstream effectors of TLR4 signaling, has been frequently reported in TNBC [
49‐
51]. The activation of this key pathway leads to the expression of pro-inflammatory cytokines and anti-apoptotic genes that foster aggressive growth, stemness and chemoresistance in TNBC cells. Indeed, pharmacological inhibition of TLR4 or IRAK1 has been reported to abolish the growth and metastatic progression of TNBC [
48,
49]. In this frame, our data indicate that PTX3 secreted by tumor cells promotes the activation of the TLR4/IRAK1 pathway in TNBC cells, and that the expression of PTX3 itself may determine the antitumor responses to TLR4 inhibition. In fact, TLR4 inhibition by TAK-242 treatment significantly impaired the proliferation and the clonogenic capacity in vitro and tumorigenic activity in vivo of
PTX3-expressing TNBC cells but did not affect the tumorigenic potential of
PTX3-silenced cells.
In a therapeutic perspective, our data indicate that the PTX3/TLR4 autocrine loop may represent a novel therapeutic target for TNBC. So far, several TLR antagonists/inhibitors have been investigated in clinical trials for the therapy of inflammatory diseases and disorders of the vascular system [
52]. In this frame, our observations suggest that the direct targeting of PTX3 or TLR4 may represent a promising therapeutic approach for the treatment of TNBC where TLR4 signaling activation strictly depends on PTX3 expression. This implies that TLR4 inhibition may affect only those TNBC lesions that express high levels of PTX3, a criterium to be taken into account for the selection of patients undergoing future anti-TLR4 therapies in TNBC. On the other side, our findings reinforce the therapeutic significance of recent approaches under phase II/III clinical evaluation [
42,
53] based on targeting TLR4 downstream effectors, such as Akt and NF-kB, in TNBC patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.