Functional inhibition of F11 receptor (F11R/junctional adhesion molecule-A/JAM-A) activity by a F11R-derived peptide in breast cancer and its microenvironment

Female BALB/c mice aged 8–12 weeks from the Animal House of the Institute of Biology, University of Bialystok (Bialystok, Poland) were housed under standard conditions at the Animal House of the Medical University of Lodz (Lodz, Poland). The procedures were conducted as per the National Animal Care Committee regulations, and were approved by the Local Ethical Committee for Animal Research in Lodz (license number 9/LB51/2017). All efforts were made to minimize animal suffering.

Murine 4T1 cells (American Type Culture Collection, ATCC) were cultured in RPMI-1640 with 10% fetal bovine serum (FBS) in flasks to 80% confluence. The 4T1 cells were harvested, resuspended at a concentration of 10⁴ cells in 100 µl PBS per mouse, and administered subcutaneously in inguinal nipple area. Seven days after the injections, the mice were divided into three groups: the Tβ4 group, the TGF-β1 group, and the control group (PBS). Non-control groups were given intraperitoneally Tβ4 or TGF-β1 resuspended in PBS at a concentration of 15 mg/1 kg body weight or 5 mg/1 kg body weight, respectively. Each injection was followed by 3 days of an injection-free period. After 21 days of treatment, the mice were anesthetized intraperitoneally with the ketamine and xylazine solution (0.1 ml/20 g mouse). Ten mice were used for each experimental group. Blood samples were collected by cardiac blood drawn using a syringe with K2-EDTA solution. Samples were centrifuged (662×g, 10 min, 4 °C), and the plasma was aliquoted and stored at − 80 °C for further analysis.

MCF-10A (non-tumorigenic human breast epithelium), MCF-7 (human breast adenocarcinoma, Luminal A) [30], MDA-MB-231 (human breast adenocarcinoma, Claudin-low) [30], and HMEC-1 (human microvascular endothelial) cell lines were from ATCC. MCF-10A cells were grown in MEBM supplemented with MEGM Single Quots and cholera toxin (Lonza). MCF-7 and MDA-MB-231 cells were cultured in RPMI-1640 medium with 10% FBS. HMEC-1 cells were grown in MCDB 131 medium with 10 ng/mL epidermal growth factor (EGF), 1 µg/mL hydrocortisone, 10 mM glutamine, and 10% FBS. The cells were cultured in a humidified incubator (37 °C, 5% CO₂). Confluent cells were passaged using trypsin–EDTA. The culture media were changed each 2–3 days.

The cells were cultured overnight in a serum-free medium and subsequently incubated for 0, 2, 5, 12, or 24 h with Tβ4 (100 nM), and then for 24 h with TGF-β1 (10 ng/ml), tumor necrosis factor-α (TNF-α; 10 ng/ml), or with a mixture of proinflammatory cytokines (TNF/IFN)—TNF-α (10 ng/ml) and interferon-γ (IFN-γ; 20 ng/ml). For the inhibition of TJ formation between the cancer cells and endothelium, the cells were treated with the P4D peptide with the sequence of a 70–82 amino acids fragment of the F11R/JAM-A mature polypeptide chain that is responsible for F11R/JAM-A trans-homodimerization [31]. The peptide P4D is a d-amino acid analog of the native peptide that binds to the 70–82 amino acid fragments and blocks the trans-homodimerization of F11R/JAM-A molecules, thus inhibiting the TJs formation in the endothelial monolayer as well as between the cancer cells and endothelial cells [32]. The sequence of control peptide (Scr) corresponded to P4D peptide, but was scrambled by random insertion of amino acid residues during the synthesis process. The peptides were synthesized and purified by LifeTein, LLC and their sequences were as follows: NH₂-(dK)-SVT-(dR)-EDTGTYTC-CONH₂ for P4D and NH₂-S-(dK)-TVE-(dR)-TDTGTYC-OH for Scr. The cells were treated with P4D or Scr (500 µM) for 0, 1, 5, or 24 h. The cells incubated with forskolin (FSK; BioShop) for 24 h at a concentration of 10 µM served as the positive control of the junctions tightening for impedance-based measurements of endothelial permeability.

F11R/JAM-A levels in murine plasma were assayed by RayBio^® Mouse JAMA ELISA Kit (RayBiotech) according to the manufacturer’s protocol and measured spectrophotometrically using the Wallac 1420 VICTOR2 Multilabel Counter (PerkinElmer).

Total RNA was isolated from the cells with the TriPure Reagent (Roche) due to the manufacturer’s protocol, dissolved in nuclease-free water, and analyzed spectrophotometrically for concentration and purity. The RNA concentrations in all samples were equalized and RNA was transcribed to cDNA using the iScript cDNA Synthesis Kit (BioRad) according to the manufacturer’s instructions. The cDNA samples were amplified by the Dream Taq Green PCR Master Mix (Thermo) and the Whatman Biometra T3 Thermocycler (Biometra) according to the provided instructions. The specific oligonucleotide primer pairs were synthesized (Genomed, Warsaw, Poland) and used in PCR reactions for the following transcripts: forward 5′-CGAGAGGAAACTGTTGTGCC-3′ and reverse 5′-AACGAGTCTGGTGGTGTCTC-3′ for F11R/JAM-A, forward 5′-GAGAGATGATGACCCTTTTGGC-3′ and reverse 5′-CCATCACCATCTTCCCAGGAGCG-3′ for glyceraldehyde 3-phosphate dehydrogenase (GAPDH, used as the loading control). The amplified products were separated in 7% polyacrylamide gels in Tris–acetate-EDTA buffer (TAE; tris(hydroxymethyl)aminomethane; ethylenediaminetetraacetic acid). Bands were visualized by ethidium bromide and UV light and documented using Gel Doc 2000 (BioRad).

The cells were lysed with lysis buffer (1% Triton X-100, 0.1% SDS, and protease inhibitors in PBS). The protein concentrations were determined with the BCA Protein Assay Kit (Thermo). Cell lysates (50 µg proteins per sample) were boiled in the loading buffer (50 mM Tris–HCl, pH 6.8, 8% glycerol, 2% SDS, 5% β-mercaptoethanol, 0.002% bromophenol blue), separated under reducing conditions on 4–20% gradient SDS-PAGE gels (BioRad), and electroblotted (120 min, 200 mA, 4 °C) onto polyvinylidene difluoride (PVDF) membranes (BioRad). The membranes were blocked for 2 h at room temperature with Tris-buffered saline (TBS) containing 5% non-fat dry milk and incubated overnight at 4 °C with a primary antibody diluted in TBS containing 0.05% Tween 20 (TBST). The following primary antibodies were used: monoclonal rabbit anti-JAM-A (Abcam) and monoclonal mouse anti-GAPDH (R&D Systems). The membranes were washed with TBST and incubated for 1 h at room temperature with the corresponding HRP-conjugated secondary antibodies (Santa Cruz) and processed for chemiluminescent detection using Westar ηC ECL Substrate (Cyanagen). The images were documented by ChemiDoc MP Imager (BioRad) and analyzed densitometrically with ImageJ software (National Institute of Health, USA). GAPDH bands served as loading controls.

The cells were labeled with the FITC anti-human CD321 (F11R/JAM-A) Mouse IgG₁ Antibody or with the relevant isotype control (BioLegend). Labeled cells were washed, harvested with ice cold EDTA/PBS, fixed with BD CellFix (Becton–Dickinson), and analyzed by FACS Canto II flow cytometer (Becton–Dickinson) upon the fluorescence excitation at 488 nm and the emission at 517 nm. Data were recorded with DIVA (Becton–Dickinson) and analyzed using FCSalyzer software (https://​sourceforge.​net/​projects/​fcsalyzer/​).

Two days after seeding, HMEC-1 cells have formed the confluent monolayers in the wells of 24-well plates and were subjected to a suitable treatment. Mammary epithelial cells (MECs) were starved overnight and on the assay day were labeled with CellTracker Green CMFDA Dye (Molecular Probes). Just before the assay, the HMEC-1 cells culture medium was substituted with the adhesion medium (MCDB 131 containing 1 mM CaCl₂, 1 mM MgCl₂, and 1% bovine serum albumin). Subsequently the aliquots of MECs (1 × 10⁵/well) were applied onto the endothelial monolayers and allowed to adhere for 1 h in a humidified incubator (37 °C, 5% CO₂). Non-adherent cells were washed away by PBS. The adherent cells were lysed by 2% SDS, and the fluorescence was measured using the Wallac 1420 VICTOR2 Multilabel Counter at 492 nm of excitation and 517 nm of emission wavelengths.

HMEC-1 cells were cultured in Transwell units with polycarbonate filter and 8 μm pores (Costar) in 24-well plates for 2 days at 37 °C to confluence. MECs were starved overnight and labeled with CellTracker Green CMFDA Dye. An aliquot (1 × 10⁵/well) of fluorescent MECs in serum-free medium was added in the top of the Transwell chamber, whereas the conditioned medium was used as a chemoattractant in the lower compartment. The wells containing the serum-free medium in the lower compartment were used to estimate the background migration. MECs were migrating across the endothelial monolayer for 6 h in a humidified incubator (temperature of 37 °C and 5% CO₂ atmosphere). At the end of the assay, non-migrated cells were wiped out from the inner surface of the membrane with a cotton swab. The assay was documented at 200 × magnification with a fluorescent inverted microscope and a digital camera (Zeiss). The migrated cells (five randomly chosen fields of each well) were counted using the ImageJ software.

Endothelial permeability was analyzed by macromolecular tracer assay. HMEC-1 cells were grown in Transwell inserts with polyethylene terephthalate filters and 0.4-μm pores in 24-well plates at 1.5 × 10⁵ cells/insert and immediately treated with P4D or Scr peptide at a concentration of 500 μM (the samples indicated as 0 h of incubation). Non-treated cells served as a control sample (Ctrl). Moreover, HMEC-1 cells were treated with the peptides at 1, 5, or 24 h after seeding. One hour after the last treatment FITC-dextran with a molecular weight of 40 kDa was added into the inserts and the culture medium from the wells was replaced with PBS. After 1 h of incubation in darkness in a humidified incubator (37 °C, 5% CO₂), the PBS from the wells (outside the inserts) was transferred into the wells of a black 96-well plate and the fluorescence was measured using the Wallac 1420 VICTOR2 Multilabel Counter at 492 nm of excitation and 517 nm of emission wavelengths.

Experiments based on impedance measurements were performed at Bionanopark, Ltd. (Lodz, Poland) with the ACEA xCELLigence^® Real-Time Cell Analysis (RTCA) DP system (Roche) according to the manufacturer’s instructions. Briefly, HMEC-1 cells were seeded at 40,000 cells per well and the plates were locked into the RTCA instrument for continuous recording of impedance changes expressed as Cell Index (CI) values [33]. During the phase of logarithmic cell growth HMEC-1 cells were pretreated with cytokines (TNF/IFN) and 24 h later they were treated with the tested compounds (Scr, P4D, or FSK). Alternatively, the cells were treated with the compounds 2 h after the cytokine treatment or the compound treatment was applied about 3 h after HMEC-1 seeding and about 67 h later was followed by the cytokine treatment. For the sample designated as ‘P4D & TNF/IFN’ after the pretreatment with P4D, the cells were subjected to the concurrent treatment with P4D and cytokines.

Experiments were performed in vivo in the murine 4T1 breast cancer model [34‐36]. Plasma samples obtained from the blood of 4T1-treated mice and untreated mice were analyzed by ELISA for soluble F11R/JAM-A (sJAM-A) quantitation. The injection of 4T1 cells did not change the level of sJAM-A in the murine plasma (Fig. 1). However, a significant decrease in the murine plasma level of sJAM-A was found upon the administration of either Tβ4 or TGF-β1 (Fig. 1 and Supplementary Table S1).

The expression of F11R/JAM-A was examined on non-tumorigenic human breast epithelial cells and on tumorigenic cells by flow cytometry (Fig. 2a). The lowest level of F11R/JAM-A/CD321 was detected on the surface of tumorigenic MDA-MB-231 cells (green). The tumorigenic MCF-7 cells and the non-tumorigenic MCF-10A cells exhibited similar F11R/JAM-A protein levels (Fig. 2a).

The administration of Tβ4 or TGF-β1 significantly dropped the F11R/JAM-A levels in the murine plasma (Fig. 1). Consequently, we examined whether these molecules could affect the expression of F11R/JAM-A on the protein and/or message level. The presence of either Tβ4 or TGF-β1 did not alter F11R/JAM-A message in treated cells (Fig. 2b, d). However, Tβ4 significantly decreased the level of the F11R/JAM-A protein in HMEC-1 cells after a 24-hour incubation period similar to TGF-β1 that was used as the positive control for demonstrating the F11R/JAM-A downregulation (Fig. 2c) [20]. The incubation of cells with Tβ4 for up to a 12-hr period of time resulted in significant time-dependent changes in the F11R/JAM-A protein level (Fig. 2e).

Any changes in the F11R/JAM-A/CD321 protein expression on HMEC-1 cells treated with Tβ4, TGF-β1, or TNF-α were analyzed by flow cytometry. The presence of Tβ4 (red) and TGF-β1 (green) decreased the amount of the F11R/JAM-A protein (Fig. 3). However, TNF-α slightly increased the F11R/JAM-A/CD321 protein level (blue).

The soluble form of F11R/JAM-A was reported to inhibit both the human platelet aggregation induced by a stimulatory monoclonal antibody M.Ab.F11 and the platelet adhesion to the inflamed endothelium [5, 7, 10, 31]. The F11R/JAM-A antagonistic peptide 4D (P4D) also blocked the M.Ab.F11-induced aggregation and the adhesion of platelets to activated endothelial cells. Consequently, we speculated whether the decrease in the plasma level of sJAM-A produced by Tβ4/TGF-β1 could interfere with adhesive processes and thereby promote the transendothelial migration (TEM) of breast cancer cells. Thus, we examined the interactions between mammary epithelial cells (MECs) and endothelial cells in the presence of the P4D peptide (Fig. 4).

The cytokines TNF-α and IFN-γ (TNF/IFN) enhanced the adhesion of MECs to the endothelium (Fig. 4a). The addition of P4D significantly inhibited the adhesion of MECs to the inflamed endothelium and this effect was the most pronounced for MDA-MB-231 cells (Fig. 4a). Likewise, we observed that the TEM of breast cancer cells in the presence of TNF/IFN and Tβ4 was blocked by P4D (Fig. 4b and Supplementary Fig. S1). Tβ4 activates the cellular motility and inhibits the adhesion [37], thus we did not utilize this peptide for adhesion assays.

Subsequently, we examined whether P4D affects the endothelial barrier permeability (Fig. 5). Transendothelial flux of 40-kDa FITC-dextran was elevated only when P4D was applied immediately (at 0 h) after the HMEC-1 cells were seeded (Fig. 5a). We conclude that P4D prevented de novo formation of TJs without disrupting the pre-existing ones.

The effect of P4D on endothelial barrier was monitored by the Real-Time Cell Analysis (RTCA). The P4D did not affect the endothelial permeability when applied ~ 24 h after the cell seeding or 24 h after the cytokine treatment (Fig. 5b, c). Therefore, we treated the HMEC-1 monolayer with P4D, Scr, or FSK 2.5 h after plating when the cells adhered to the bottom of the wells without forming a confluent monolayer, as demonstrated by the low CI levels (Fig. 5d, e). Presumably, at this time period, the TJs have not fully developed and only a minor fraction of F11R/JAM-A molecules have formed trans-homophilic interactions. The endothelium was treated with inflammatory cytokines ~ 60 h later, during the logarithmic growth phase (Fig. 5d, f). The cells from ‘P4D & TNF/IFN’ sample were retreated with P4D concurrently with TNF/IFN treatment. The barrier-protecting effect of P4D was more evident than that of FSK, when P4D pretreatment was followed by the second supplementary dose applied simultaneously with the cytokines (Fig. 5d, f). Otherwise, P4D did not reveal any barrier-protecting properties without the supplementary dose (Fig. 5f).