Background
Desmogleins and desmocollins belong to “non-classical” desmosomal cadherin family members that constitute the adhesive interface at the core domain of desmosome junctions. Four desmoglein (Dsg1–4) and three desmocollin isoforms (Dsc1–3) have been found to be expressed in a tissue- and differentiation-specific manner in human [
1‐
3]. Their cytoplasmic domains bind to plakoglobin (γ-catenin), a close homolog of β-catenin, to assist in linking the desmosomes with the intermediate filament cytoskeleton [
4]. Dsg2 and Dsc2 are ubiquitous in desmosome-bearing tissues including simple epithelia, while the other isoforms are primarily expressed in stratified epithelia. Dsg2 and Dsc2 are regarded as partners since they act in the way of heterophilic interactions through opposite-charge attraction [
5,
6]. It was reported that deregulation of Dsg2 and its partner Dsc2 contributed to initiation and progression of cancers in a context-dependent manner. They have oncogenic function in some cancers, but in others they can function as suppressor genes. For instance, decreased Dsg2 expression indicated poor prognosis in gallbladder carcinoma, pancreatic cancer, melanoma, gastric cancer and anaplastic thyroid cancer [
7‐
11]. In contrast, higher Dsg2 expression was associated with progression in cancers of the lungs, the liver, the colon and the skin [
12‐
15]. Similarly, decreased expression of Dsc2 promoted tumorigenic behavior in colorectal carcinoma, pancreatic ductal carcinoma and esophageal squamous cell carcinoma [
16‐
18]. However, little is known about roles of Dsg2 and Dsc2 on tumorigenesis and progression in BC, which is the most common cancer among women.
Activated EGFR pathway contributes to onset and progression in many cancers including BC [
19,
20]. Previous studies identified Dsg2 loss induced a reduction of EGFR phosphorylation associated with activation in carcinoma cells of colon, gallbladder, skin and the lungs [
7,
12,
14,
21,
22]. But interestingly, there were different effects of Dsg2 loss on downstream pathways of EGFR activation, malignancy and resistance to EGFR-targeted therapy in these cancers.
BC is a complex and highly heterogeneous disease with several intrinsic molecular subtypes including luminal, triple-negative and some other subtypes according to ER, PR and HER2 status [
23,
24]. In normal mammary gland, Dsg2 and Dsc2 are expressed in both of luminal and myoepithelial cells, and they are required for luminal cell-cell adhesion and mammary epithelial morphogenesis [
25,
26]. This study aimed to investigate the effects of partner desmosomal cadherins Dsg2 and Dsc2 on the malignancy of BC and their underlying mechanisms, especially relating to EGFR and its main downstream pathways. Firstly, we found expression levels of Dsg2 and Dsc2 in BC tissues and cells were both decreased compared to normal counterparts. Next, we identified that shRNA-mediated loss of Dsg2 and Dsc2 could both promote malignant behaviors in tripe-negative MDA-MB-231 and luminal MCF-7 cells. Mechanistically, despite attenuating EGFR activation, Dsg2 or Dsc2 depletion can activate AKT and ERK pathways maybe through mediation of other tyrosine kinases in MDA-MB-231 cells, but can induce β-catenin accumulation and suppress AKT and ERK pathways in MCF-7 cells, respectively. These data demonstrate that Dsg2 and Dsc2 play important roles in regulating tumor-relevant signaling in a context-dependent manner beyond their function as cell adhesion molecules in BC.
Materials and methods
Human tissue specimens
A total of 30 human invasive ductal BC with adjacent normal tissues were collected from Pathology Department, Tianjin Medical University. The tissue collection and analysis conducted in this study was approved by the Ethical Committee of Tianjin Medical University, China. All 30 cases were randomly selected and were made anonymous to us. The pathologic diagnosis was counterchecked by two senior pathologists according to the 2003 World Health Organization histological classification of breast tumors. Immunohistochemistry was evaluated based on the percentage of positive staining (0: 0–10%, 1: 10–25%, 2: 26–50% and 3: >50%) and staining intensity (0: none, 1: weak, 2: intermediate, and 3: strong). The percentage and intensity scores were added to obtain an overall score. The protein expression with an overall score of 0–2 was considered “negative”, while that with an overall score of 3–6 was considered “positive”.
Cell culture and transfection
The cell lines used in this study were HEK293T, the non-tumorigenic human breast epithelial cell line MCF10A and human BC cell lines MDA-MB-231 and MCF-7. All cell lines were obtained from the ATCC and then underwent verification using the short tandem repeat method. MCF10A was cultured in DMEM/F-12 (KeyGEN Bio TECH) with 5% FBS (NEWZERUM ), 10 µg/ml insulin, 20ng/ml EGF, 0.5 µg/ml hydrocortisone, 100ng/ml cholera toxin and 1% penicillin-streptomycin. The other three cell lines were cultured in DMEM (KeyGEN Bio TECH) supplemented with 10% FBS (NEWZERUM) and 1% penicillin-streptomycin. All cells were cultured at 37 °C with 5% CO2. Cells were incubated in serum-free DMEM for 24 h prior to treatment with EGF (10 ng/ml; Gibco) for another 24 h.
The plasmids were synthesized by Genecopoeia, including four short hairpin RNAs (shRNAs) targeting Dsg2, three shRNAs targeting Dsc2, and 2 non-targeting Control shRNAs. HEK293T cells were used for lentivirus packaging according to the manufacturer’s instructions (Lenti-PacTM HIV Expression Kit, Genecopoeia). The virus was added to BC cells along with 8 µg/ml polybrene. After 20 h, the medium was removed and replaced with fresh medium containing 2 µg/ml and 0.8 µg/ml of puromycin in MDA-MB-231 and MCF-7 cells, respectively. Puromycin-resistant clones which were selected by culturing for another 2 weeks in the presence of puromycin were regarded as stably transfected cells.
Western blotting
Cells were lysed in RIPA buffer supplemented with proteinase and phosphatase inhibitors. Protein concentration was determined by ultraviolet spectrophotometer. Then equal amounts of proteins were separated by SDS-PAGE and then electrotransferred to PVDF membrane. After blocking in TBST with 5% non-fat milk, the membranes were cut into strips and then incubated overnight with various primary antibodies at 4 °C, followed by incubation with a secondary antibody (1:3000 dilution) at room temperature for 2 h. The primary antibodies were against Dsg2 (1:1000, #21880, Proteintech), Dsc2 (1:1000, #13876, Proteintech), cyclin D1 (1:5000, #60186, Proteintech), N-cadherin (1:1000, #PTM-5221, PTMBIO), E-cadherin (1:500, #7870, Santa Cruz Biotechnology), CD133 (1:500, #ab226355, Abcam), β-catenin (1:5000, #ab32572, Abcam), EGFR (1:500, #373746, Santa Cruz Biotechnology), p-EGFR (Y845) (1:500, #ab97613, Abcam), p-EGFR (Y1092) (1:1000, #PTM-6870, PTMBIO), p-ERK1/2 (T202/Y204) (1:2000, #4370, Cell Signaling Technology), p-AKT (S473) (1:1000, #PTM-6649, PTMBIO), and GAPDH for internal reference (1:3000, #TA-08, ZS Bio).
5 × 102 cells were seeded into 6-well plates. Cells were maintained at 37 °C in an incubator containing 5% CO2. After incubation, the medium was renewed and the plates were incubated for 14 days under the same culture conditions. The inhibitors were added on the 7th day after cell seeding. The cells were washed twice with ice-cold phosphate-buffered saline (PBS) and fixed with methanol for 20 min in room temperature, followed by staining with 0.5% crystal violet.
MTT assay
Human BC cells (1 × 104/well) were placed in 96-well plates and continually cultured for different periods of time (1, 2, 3, 4 or 5 days). Subsequently, 50 µl of 0.5 mg/ml MTT (Catalog no. KGA9301, KeyGEN Bio TECH) was added to each well. The cells were incubated at 37 °C for another 4 h, the medium was removed, and the precipitated formazan was dissolved in 100 µl of DMSO. After the solution was shaken for 10 min using an Eppendorf Mix Mate (Eppendorf, GRE), the absorbance was detected at 490 nm (A490) on a Bio Tek ELx800 (Bio Tek, USA).
Wound-healing assay
Cells of each groups were incubated in 6-well plates. The cell monolayer was scraped in a straight line with 100 µl pipette tip and incubated with serum free medium. Photographs of scratches were monitored by an invert microscope at different time points. Cell motility was assessed by measuring the migration of cells into a scrape. The speed of wound closure was monitored after 48 and 72 h by measuring the ratio of the area size of the wound relative to that at hour 0. Each experiment was performed in triplicate.
Transwell migration and invasion assays
Transwell migration and invasion assays were performed with an 8.0 μm pore filter chamber (Invitrogen) without or with Matrigel inserted in 24-well plates. The BC cells (1 × 105 cells) in 100 µl of DMEM without FBS were seeded into the upper wells, and DMEM with 10% FBS were added to the bottom chamber. Transfected MDA-MB-231 cells were incubated and allowed to migrate for 24 h or invade through the Matrigel for 48 h. Transfected MCF-7 cells were incubated for an additional 24 h. After fixing with methanol, the non-invading cells were removed from the upper surface. The invaded cells adhering to the bottom surface of the membrane were stained with 0.5% crystal violet. Using an inverted light microscope (Nikon), we counted the number of invading cells. All experiments were repeated independently at least three times.
Immunofluorescence staining
The cells were tiled on coverslips, incubated at 37 °C overnight, fixed with ethanol and blocked with 5% FBS. Then, the cells were incubated with primary antibodies against Dsg2 (#21880; Proteintech; 1:50) and Dsc2 (#13876; Proteintech; 1:50). After incubation with fluorophore-conjugated secondary antibodies, the nuclei were counterstained with 4’,6-diamidino-2-phenylindole(DAPI) (Sigma). Images were acquired by fluorescence microscopy (Nikon, Japan).
Statistical analysis
For all analysis, mean ± SD of the measurements was calculated and illustrated in the histograms or broken line graphs. Student’s t-test and chi-square analysis were used to determine whether there was a significant difference between two means, and nonparametric ANOVA was performed for comparison of multiple means. The data were analysed with the GraphPad Prism 8.0 (GraphPad Software). P values (two-sided) less than 0.05 were considered statistically significant.
Discussion
Cancer initiation and progression are complex progresses in which many signaling pathways are involved. Crosstalk between different pathways allows the integration of the great diversity of stimuli that a cell receives [
30‐
33]. As cytomembrane proteins, desmosomal cadherins and EGFR, a common receptor tyrosine kinase (RTK), have been reported to interact in several cancers previously [
21,
22,
34]. EGFR pathway plays important roles in cancers including BC [
20,
35]. In this study, we simultaneously investigated the roles of two common partner desmosomal cadherins Dsg2 and Dsc2 played in BC as well as their effect on EGFR pathway. We found shRNA mediated-knockdown of Dsg2 and Dsc2 could both enhance proliferation, migration and invasion in triple-negative MDA-MB-231 and luminal MCF-7 cells. However, the phosphorylation levels of EGFR on Y845 and Y1092, two representatives of EGFR activation, were reduced upon loss of Dsg2 or Dsc2 whatever in triple-negative MDA-MB-231 cells expressing higher EGFR or in luminal MCF-7 cells with lower EGFR level [
36]. EGFR phosphorylation on Y845 is usually catalyzed by Src and Y1092 belongs to auto-phosphorylation sites. Y845 and Y1092 phosphorylation-mediated signaling is linked to higher cancer malignancy due to enhanced cell transformation, proliferation, motility, and invasion [
37,
38]. It is obvious that decreased phosphorylation of EGFR on Y845 and Y1092 could not account for increased malignancy in our study. We speculate that there must be other alternative signaling pathways enhancing malignant behaviors. However, our data revealed the alternative pathways were obviously distinct in MDA-MB-231 and MCF-7 cells partly due to their different cellular genetic background [
39].
In triple-negative MDA-MB-231 cells, the p-AKT and p-ERK, two of the main downstream molecules of EGFR pathway were increased upon loss of Dsg2 or Dsc2. These results are similar to that in both gallbladder carcinoma and anaplastic thyroid cancer reported by Jeong-Ki Min previously [
7,
11]. Jeong-Ki Min et al. found loss of Dsg2 promoted gallbladder carcinoma progression and resistance to EGFR-targeted therapy through Src kinase activation, a non-receptor tyrosine kinase (nRTK), by which increased p-AKT and p-ERK were involved, indicating PI3K and MAPK pathways were activated. In addition, in anaplastic thyroid cancer Dsg2 depletion significantly increased metastatic potential through activation of c-Met which belonged to receptor tyrosine kinase (RTK) and also transmitted higher downstream p-AKT activity. Besides, it is worth noting that competition exists among different RTKs for the same downstream signaling molecules [
40‐
43]. Based on these above, we proposed that loss of Dsg2 or Dsc2 that inhibited EGFR activity may cause competitive activation of some other tyrosine kinases including nRTKs and RTKs, and then transmit positive signals to downstream AKT and ERK pathways in MDA-MB-231 cells which overexpressed and relied overly on tyrosine kinases [
44]. It needs to explore which tyrosine kinases are activated by loss of Dsg2 or Dsc2 and how they are activated next. Futhermore, in our study ERK inhibition exhibited much weaker anti-tumor effect than AKT inhibition in general, which could be attributed to feedback activation of p-AKT induced by ERK inhibitor PD98059 in shDsg2 and shDsc2 MDA-MB-231 cells. Feedback activation is commonly observed in cancers and it contributes to resistance to the targeted therapy of cancer [
45‐
47]. Therefore, inhibitors of AKT, not of ERK, can be considered for triple-negative BC patients with low Dsg2 or Dsc2 expression.
In luminal MCF-7 cells which expressed low tyrosine kinases but high E-cadherin, Dsg2 or Dsc2 depletion attenuated EGFR activity and the downstream PI3K and MAPK pathways [
44]. Consistent with the result, Bing-Xia Zhou et al. showed that in cervical cancer cells transfected with si-Dsg2, p-ERK was significantly decreased [
48]. Likewise, another study reported that shDsg2 colon cancer cells had decreased p-ERK and p-Src levels [
14]. However, shDsg2 or shDsc2 MCF-7 cells exhibited enhanced malignancy in our study while Dsg2-deficient cervical cancer and colon cancer cells showed inhibited progression. Moreover, we found that increased β-catenin contributed to enhanced tumor malignant behaviors in MCF-7 shDsg2 or shDsc2 cells, which was consistent with that in E-cadherin-expressing oesophageal squamous cell carcinoma (ESCC) cells reported by Wang-Kai Fang et al. They found Dsc2 loss can release more free γ-catenin which may compete with β-catenin, thus displacing the latter from E-cadherin and increasing β-catenin- dependent transcriptional activity [
49]. Besides, Kolegraff et al. reported that loss of Dsc2 contributed to the growth of colorectal cancer cells through the activation of AKT/β-catenin signaling [
16]. MCF-7 cell, a less aggressive BC cell line, overexpresses E-cadherin, a classic cadherin. Γ-catenin and β-catenin are highly homologous. Γ-catenin can interact with E-cadherin as β-catenin but, in addition, binds to the desmosomal cadherins desmocollin and desmogelin in desmosomes. In our study, loss of Dsg2 and Dsc2 in MCF-7 cells could release more free γ-catenin to compete with its close relative, β-catenin to bind to E-caherin. In addition, loss of Dsg2 and Dsc2 might activate Wnt signaling to reduce degradation of free β-catenin. Therefore, more free β-catenin was accumulated in the cytoplasm and the nucleus. Increasing accumulation in the nucleus promoted β-catenin-dependent transcriptional activity. All these above involving MCF-7 cells are our speculation, but how Dsg2 or Dsg2 loss induced β-catenin accumulation in MCF-7 cells specifically need much in-depth study in future.
In conclusion, this study determined that both of Dsg2 and Dsc2 were down regulated in human BC tissues and cells. In vitro experiments confirmed that in spite of reducing EGFR activity, single loss of Dsg2 or Dsc2 could activate AKT and ERK pathways in MDA-MB-231 cells and facilitate β-catenin accumulation in MCF-7 cells to promote malignancy, respectively. This study suggests that AKT inhibitors can be chosen for triple-negative BC patients with low Dsg2 or Dsc2 expression while luminal BC patients with low Dsg2 or Dsc2 may benefit from inhibitors targeting β-catenin. More specific mechanisms are needed to be clarified in future.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.