3.1. Functional classification of data set
First, the GO annotation was performed to determine the molecular function of 184 detected glycoproteins by Timpe et al. (7) in MCF-7 cell line. Only GO terms, including receptor proteins, proteins involved in endocytosis and proteins involved in metastasis, were selected in this step.
3.1.1. Receptor proteins
In the cancer cell, receptor proteins are the most important proteins associated with the cell growth and division by interaction with growth factors and mitogen. It is noteworthy that various receptor proteins in response to anticancer therapeutic regimens overexpress and show drug resistance properties so offer opportunities for cancer therapy (13). GO analysis of 184 DEGs based on molecular function released to four receptor proteins, which they have overexpressed in MCF-7 cells compared with normal cells. These proteins include as follows: CD239, CD55, CD47 and CD112. Details of receptor proteins are described in Table 1 and Fig. 1.
Table 1
Details of different receptor proteins expression released from GO analysis
No. | Protein name | Gen name | UniProt accession no. | Protein expression level |
MCF7 | MCF10A | MCF12A |
1 | Basal cell adhesion molecule | CD239 | P50895 | 7 | 0 | 0 |
2 | Complement decay-accelerating factor | CD55 | P08174 | 28 | 3.5 | 20 |
3 | Leukocyte surface antigen CD47 | CD47 | Q08722 | 29.5 | 6 | 15.5 |
4 | Nectin-2 | CD112 | Q92692 | 10.5 | 3 | 8.5 |
Among these receptor proteins, P50895 (CD239) is more important because it is expressed in MCF-7, while MCF10A and MCF12A standard cell lines do not express this receptor protein; and also, as shown in Fig. 1, CD239 is significantly upregulated in breast cancer tissues. CD239, or basal cell adhesion molecule (B-CAM) is a membrane-bound glycoprotein that belongs to the immunoglobulin superfamily. B-CAM acts as the receptor for laminin-10 (α5β1γ1), − 11 (α5β2γ1), and − 15 (α5β2γ3), which are the major components of the extracellular matrix proteins (14). B-CAM is induced in some epithelial malignancies and is known as a candidate molecule involved in cancer progression (15). So far, the modulation role of Lu/BCAM has been found in different types of cancers, such as ovarian, breast, brain, liver, skin, thyroid, colon, and bladder cancers (16). Proteomics investigation proposes that BCAM could use as a potential biomarker for pancreatic cancer (17). Lu/BCAM has been demonstrated that sustains cancer cell migration via regulating integrin-mediated cell adhesion to laminin 511 and mediates metastatic spreading of colorectal cancer (16). Kikkawa et al. reported that B-CAM is highly expressed in a number of breast cancer cells (18). They also generated a specific antibody against CD239 and demonstrated that this antibody has an antitumor function in CD239-highly positive SKBR3 cells but not in weakly positive cells. The importance of Lu/BCAM on tumorigenesis has been identified nevertheless, the downstream molecular signaling mechanisms of Lu/BCAM in tumorigenesis remain unclear, and further studies are needed on its function as a promising drug target.
CD55 or decay-acceleration factor, as a glycosylphosphatidylinositol-anchored protein is overexpressed in MCF7 cells and belongs to the receptors of the complement activation family, which are expressed in entire the body to inhibit over-regulation of the complement system (19). In breast cancer tissues, CD55 expression shows a notable downregulation, which contrasts with its overexpression in MCF7 cells (Fig. 1). Nevertheless, it has been shown that CD55 plays a critical role in the carcinogenesis of breast cancer, and the detection of a few cells with high CD55 expression would be enough to predict the poor prognosis of patients (20). Probably, overexpression of CD55 can reduce complement activation, prevent complement-mediated removal of cancer cells, and consequently enhance cancer progression. Moreover, CD55 can stimulate initiation and growth of tumor via prevention natural killer cells (21). A murine monoclonal antibody recognizing CD55 has been proposed for cancer treatment by inducing T-cell responses (22). Also, in a study, a short hairpin RNA (shRNA) reduced expression of CD55 in the SK-BR3 cells and improved trastuzumabinduced complementdependent cytotoxicity activation (23).
CD47, or integrin-associated protein is a receptor for the cell-binding domain of thrombospondin family members. CD47 has displayed a higher expression in breast cancer cells (24). CD47 acts as a counter-receptor for signal-regulatory proteinα (SIRPα), which is mostly expressed on macrophages. The binding of CD47 to SIRPα induces a signaling cascade that inhibits the phagocytic activity of macrophages and limits the trafficking of dendritic cells (25). Therefore, one of the strategies that cancer cells use to hide from the immune system is overexpression of the surface glycoprotein CD47, which functions as an anti-apoptotic protein or initiates the “don’t-eat-me signal” that intervenes with phagocytosis and tumor antigen cross-presentation (26). Targeting the phagocytosis checkpoint of the CD47-SIRPα axis has demonstrated significant therapeutic efficacy in pre-clinical and clinical cancer treatment studies (27). CRISPR-mediated CD47 and HER2 dual blockade prevented clonogenicity and increased macrophage-mediated attack in a mouse breast tumor (28). It's noteworthy to mention that while CD47 is overexpressed in MCF-7 cells, in breast cancer tissues obtained from the UALCAN database, there has been a reported decrease in CD47 expression levels. This complex interplay of CD47 expression within different contexts underscores the multifaceted role of CD47 in cancer and the potential for therapeutic interventions.
CD112, or Nectin-2 (poliovirus receptor-related 2 protein) is a Ca2+-independent immunoglobulin‐like cell adhesion molecule. Nectin-2 is an immune checkpoint that deregulates immune cell function (29). Previous studies have indicated the excessive expression and critical roles of Nectins in several cancer types, such as breast cancer, ovarian cancer, leukemic blasts, myeloma, adenosquamous carcinomas, colorectal carcinoma, adenocarcinoma of gallbladder, and pancreatic ductal adenocarcinomas (29). High expression of Nectin‐2 is related to high malignancy, progressive tumor stage, and poor outcome/prognosis (30). Simultaneously, the termination of proliferation of tumor cells in cell culture was demonstrated due to inhibition of Nectin-2. In mouse therapeutic models, a representative anti-Nectin-2 mAb demonstrated anti-tumor effects against MDA-MB-231 breast cancer cells (31).
3.1.2. Proteins involved in endocytosis
Endocytosis is a major regulator of cell cycle, mitosis, apoptosis and cell fate determination, as well as plays a critical role in the regulation of tumor metastasis. Several endocytic proteins are dysregulated in cancer cells and stimulate tumor migration and invasion. Target therapy in cancer using proteins involved in endocytosis can influence internal molecules to induce apoptosis (32), so proteins incorporated in endocytosis are important. Using GO analysis, we enriched eighteen glycoproteins involved in endocytosis, with significant differential expression in MCF-7 cell line. Most of the detected proteins involved in endocytosis are down-regulated in the MCF-7 compared to the standard cells, however the galectin-3-binding protein (Gal-3BP) and vitronectin have been overexpressed and can use to endocytosis the drugs. The details of glycoproteins incorporated in endocytosis are indicated in Table 2 and Fig. 2.
Table 2
Proteins incorporated in endocytosis released from GO analysis
No. | Protein name | Gen name | UniProt accession no. | Protein expression level | |
MCF7 | MCF10A | MCF12A | |
1 | Monocyte differentiation antigen CD14 | CD14 | P08571 | 0 | 18 | 12 | |
2 | Prolow-density lipoprotein receptor-related protein 1 | LRP1 | Q07954 | 1 | 65 | 82 | |
3 | NPC intracellular cholesterol transporter 1 | NPC1 | O15118 | 3.5 | 36.5 | 38 | |
4 | Amyloid-beta precursor protein | APP | P05067 | 0 | 3.5 | 2.5 | |
5 | Cadherin-13 | CDH13 | P55290 | 0 | 21.5 | 73 | |
6 | Calreticulin | CALR | P27797 | 3.5 | 260.5 | 99 | |
7 | Galectin-3-binding protein | LGALS3BP | Q08380 | 81.5 | 48.5 | 17.5 | |
8 | Endoplasmin | HSP90B1 | P14625 | 12.5 | 211 | 9 | |
9 | Hypoxia up-regulated protein 1 | HYOU1 | Q9Y4L1 | 43 | 206 | 39.5 | |
10 | Cation-independent mannose-6-phosphate receptor | IGF2R | P11717 | 11.5 | 88 | 204.5 | |
11 | Low-density lipoprotein receptor | LDLR | P01130 | 0 | 24.5 | 64 | |
12 | C-type mannose receptor 2 | MRC2 | Q9UBG0 | 0 | 4 | 34 | |
13 | Cation-dependent mannose-6-phosphate receptor | M6PR | P20645 | 56.5 | 52 | 134 | |
14 | Palmitoyl-protein thioesterase 1 | PPT1 | P50897 | 50.5 | 21.5 | 119 | |
15 | Sortilin | SORT1 | Q99523 | 0 | 6 | 0 | |
16 | Sushi domain-containing protein 2 | SUSD2 | Q9UGT4 | 3.5 | 6.5 | 100 | |
17 | Tetraspanin-1 | TSPAN1 | O60635 | 0 | 18 | 1 | |
18 | Vitronectin | VTN | P04004 | 21.5 | 0 | 0 | |
Galectin-3-binding protein (Gal-3BP) is a secreted multifunctional hyperglycosylated protein that boosts integrin-mediated cell-to-cell adhesion. It can also boost the cancer cells' survival in the bloodstream by inducing galectin-mediated tumor cell aggregation during the metastatic process (33). Galectin-3 mediates the endocytosis of β-1 integrins in a lactose-dependent manner in breast carcinoma cells (34). The Gal-3BP expression has been discovered to be associated with poor prognosis in breast cancer patients (35). A study indicated a powerful relationship between high Gal-3BP expression and reduced survival in node-negative breast cancer patients who after surgery didn't receive systemic therapy (36). These results were further confirmed by a cohort study indicating a significant relationship among Gal-3BP and the occurrence of metastasis in ER-negative breast cancer patients (37). Some investigations have been performed to inhibit Gal-3BP functions using monoclonal antibodies, particularly against various domains of the protein (38). For example, SP-2, an antibody detecting a conformational epitope along the lectin binding domain of Gal-3BP has exhibited beneficial therapeutic effects in multiple tumor xenografts (39).
Vitronectin (VTN) is a multifunctional glycoprotein that can be found almost in all tissues and works as an adhesive agent and a link between cells and ECM (40). VTN encoding mRNA is mostly up-regulated in tissues with a stress-like proinflammatory response and gathers in the extravascular and interstitial spaces around the tumor (41). Integrins like αvβ1, αvβ3, αvβ5, αvβ6, αvβ8, urokinase plasminogen activator receptor (uPAR), and plasminogen activator inhibitor-1 (PAI-1) act as ligands for VTN (42). The αvβ5 integrin functions as an endocytic receptor for VTN. In multiple malignancies, VTN plays an important role in metastasis and tumor invasion by promoting degradation of cell adhesion and matrix through binding to uPAR, and PAI-1, and possibly works as structural support for tumor cell migration (40). Stuart et al. developed a VTN-binding aptamer conjugated with doxorubicin, demonstrating increased cytotoxicity to MDA-MB-231 breast cancer cells. They suggested that cells were taking up the drug complex due to nonspecific mechanisms such as endocytosis (43).
Calreticulin (CALR) is an endoplasmic reticulum-resident protein involved in a range of cellular processes, including calcium regulation, cell adhesion, protein folding, and cancer development. In breast cancer, CALR is found to be overexpressed compared to normal tissue and correlates with patient mortality and stemness indicators. Its expression is notably heightened in breast cancer stem cell-enriched populations. Interestingly, surface-exposed CALR appears to trigger immune-stimulating effects beyond mere phagocytosis. Dysfunctional CALR production causes tumorigenesis and disrupts the balance of healthy cells and immune surveillance. Therefore, CALR can serve as both a diagnostic marker and a potential target for cancer treatment (44).
Heat shock protein 90-alpha (HSP90AA1), a highly conserved molecular chaperone, is prominently expressed under conditions of trauma, infection, and in tumors. This protein can be secreted into the extracellular environment and enter the nucleus, influencing immune memory formation and participating in tumor development. HSP90AA1 has been implicated in various endocytic mechanisms and is essential for regulating DNA damage, cell cycle, and gene expression (45). In cancer, it activates various oncogenic client proteins, promoting cell survival, growth, and invasiveness, making it a potential cancer biomarker, and suggesting its potential as a therapeutic target (46, 47).
Hypoxia-upregulated 1 gene (Hyou1) is a member of the Hsp70 family. It plays a key role in protecting cells from endoplasmic reticulum (ER)-stress induced apoptosis and is induced by various stressors. HYOU1 is highly expressed in various cancers, including ovarian, breast, bladder, prostate, gastric, thyroid, lung and colorectal cancers, making it a critical survival chaperone in cancer cells, protecting them from ER stress-induced cell death. This highlights its importance as a biomarker or drug target in cancer therapy (48).
Cation-dependent mannose-6-phosphate receptor (M6PR) is a type I transmembrane receptor and its related pathways are vesicle-mediated transport and clathrin-mediated endocytosis. M6PR holds potential in lysosomal disease therapy and targeted delivery of apoptosis-inducing therapeutics. M6PR are widely expressed in various tissues and are upregulated in conditions like fibrosis, neurodegenerative diseases, and certain cancers, including breast, pancreatic, gastric, melanoma, and prostate cancers. They offer promise as cancer-specific drug delivery target due to its increased affinity in the acidic tumor microenvironment. However, M6PR can also function as tumor suppressor and may be downregulated or mutated in certain malignancies, such as hepatocarcinoma, adding complexity to its role in cancer treatment (49).
PPT1, a gene associated with synaptic vesicle exo- and endocytosis, plays a crucial role in regulating neurotransmission and synaptic integrity. Clinical data reveals elevated PPT1 mRNA expression in various cancers, including breast cancer, renal cell carcinoma, head and neck squamous cell carcinoma, thyroid cancer, and more, particularly in metastatic tumors. Increased PPT1 levels are associated with poor patient prognosis in liver, esophagus, kidney, and head and neck cancers. Mechanistically, PPT1 depalmitoylates the V0a1 subunit of v-ATPase, affecting lysosomal pH, autophagy, and mTORC1 signaling. Inhibiting PPT1's depalmitoylating activity impairs autophagy and mTORC1-regulated signal transduction. Additionally, PPT1 overexpression can confer resistance to apoptosis inducers by modulating Akt and Ras. This information highlights the diverse functions and clinical relevance of PPT1 in cancer and neurodegenerative diseases (50–52).
SUSD2 is a type I transmembrane protein predominantly located on the plasma membrane of breast cancer cells. Its overexpression does not impact cell migration but significantly enhances cell invasion. SUSD2 interacts with Gal-1, a member of the galectin family known for its involvement in tumor immune evasion and angiogenesis (53, 54). Recent evidence indicates that SUSD2 represents a highly promising therapeutic target for breast cancer.
Tetraspanin 1 (TSPAN1), a member of the TSPAN superfamily, facilitates interaction with cell surface signaling molecules and has gained attention for its influence on cancer cell malignancy. While TSPAN1 is notably overexpressed in various cancers, its role in breast cancer has remained relatively unexplored. TSPAN1 depletion has been shown to suppress breast cancer tumor growth in mouse models, further emphasizing its potential as a therapeutic target (55).
3.1.3. Proteins involved in metastasis
The metastatic process is associated with several processes in the cancer cells and extracellular matrix (56). Cell adhesion molecules, integrins, and proteases are important molecules in cellular migration (57, 58). GO analysis of 184 detected glycoproteins determined the seven proteins involved in cell adhesion with down-regulated expression in cancer cells. Details of these cell adhesion molecules described in Table 3 and Fig. 3. Also ITGB1 was determined as hub-bottleneck protein in PPI network. Although three of these proteins, ITGAV, ITGA5, and ITGB5, exhibited increased expression in cancer tissues. Loss of cell adhesion molecules allows malignant cells to detach from each other and the ECM and move from their site of origin, obtain an increased invasion phenotype, and eventually invade and metastasize (59). Different methods have been used for the repair of adhesion among tumor cells and the surrounding ECM at their early site for satisfactory prevention of the development of metastases. Some of these approaches are including the use of biochemical agents for inhibiting tumor cell migration (60), the use of heparins to manage metastasis through their anti-clotting features and their interactions with integrins and selectins (61), targeting cancer cells by liposome nanoparticles (62), and increasing adhesion of cancer cells with clay nanoparticles (60). So far, the research has not yielded acceptable results, and there is a need to study new methods at the molecular levels.
Table 3
Protein complexes involved in cell adhesion released from GO analysis
No. | Protein name | Gen name | UniProt accession no. | Protein expression level |
MCF7 | MCF10A | MCF12A |
1 | Integrin alpha-3 | ITGA3 | P26006 | 2 | 144 | 174.5 |
2 | Integrin alpha-6 | ITGA6 | P23229 | 0 | 90.5 | 10.5 |
3 | Integrin beta-1 | ITGB1 | P05556 | 85.5 | 307.5 | 472.5 |
4 | Integrin alpha-2 | ITGA2 | P17301 | 4 | 144 | 13.5 |
5 | Integrin alpha-V | ITGAV | P06756 | 40.5 | 105 | 142 |
6 | Integrin alpha-5 | ITGA5 | P08648 | 0 | 60.5 | 74 |
7 | Integrin beta-5 | ITGB5 | P18084 | 12.5 | 14.5 | 64.5 |
Proteolytic enzymes are other proteins that play a critical role in metastasis. The release of some proteases from metastatic tumors has an essential role in the degradation of various proteins and glycoproteins, which form a highly cross-linked structure in basement membranes, and so they caused metastatic growth (63). GO analysis based on molecular function determined cathepsin D and dipeptidyl peptidase II as proteases, which overexpressed in MCF-7 and have several roles in the cell surface.
Cathepsin D (CTSD) belongs to the lysosomal aspartic protease family, which is proteolytically active at low pH and has an essential role in many physiological and pathological processes, such as intracellular catabolism in lysosomal compartments, activation of inflammatory cells and regulation of cell death (64, 65). CTSD is well known for its roles in proliferation, metastasis, carcinogenesis, and angiogenesis in cancer (66). CTSD is identified in salivary adenoid cystic carcinoma, most metastatic breast cancer cell lines and ovarian cancers. CTSD is a poor prognosis marker in breast cancer. It up-regulates in the most malignancies breast cancers, and increases the dissemination and invasion of tumors through various mechanisms associated with its proteolytic activities (64, 67). Some results suggest that inhibition of CTSD increases drug-induced death of cancer cells (66). A study demonstrated that graphene oxide acted as an adsorbent for CTSD and cathepsin L and diminished their functions through binding to cationic and hydrophilic residues (68). One recent study demonstrated that CTSD is a potential tumor-specific extracellular target in triple-negative breast cancer for antibody-based therapy and inhibition tumor growth (69).
Dipeptidyl peptidase II (DPPII) is a serine protease that mostly appearances in the vesicular system and extracellular region. It is active in acidic pH and has an essential role in the degradation some oligopeptides. This protease has a role in the management cell differentiation and protection from the degradation of collagen fragments and cell death (70). These proteases, which overexpress in MCF7 and cause metastasis, can be used as potential targets for cancer therapy.
3.2. Some detected glycoproteins that are expressed in MCF-7 cell line with zero expression in normal cells
Based on the detected DEGs, we determined the proteins with zero expression in standard cells, while they expressed in MCF-7 cell line including: APOD, BCAM, CLU, IFI30, GRN, LYPD3, NAAA, PODXL, SEZ6L2 and VTN. Since these proteins are not expressed in standard cells, they can be used as drug targets for the treatment of cancer. Nevertheless, ulcan database analysis revealed reduced expression of APOD, CLU, PODXL, and VTN in cancer tissues. Additionally, the expression of the NAAA protein in breast tissue was not found to be significantly meaningful. Table 4 shows details of these glycoproteins in cell lines and Fig. 4 demonstrates expression levels of these genes in cancer tissues.
Table 4
Proteins that are expressed only in MCF7 cell line
No. | Protein name | Gen name | UniProt accession no. | Protein expression level |
MCF7 | MCF10A | MCF12A |
1 | Apolipoprotein D | APOD | P05090 | 191 | 0 | 0 |
2 | Basal cell adhesion molecule | BCAM | P50895 | 7 | 0 | 0 |
3 | Clusterin | CLU | P10909 | 4 | 0 | 0 |
4 | Gamma-interferon-inducible lysosomal thiol reductase | IFI30 | P13284 | 20 | 0 | 0 |
5 | Progranulin | GRN | P28799 | 26.5 | 0 | 7 |
6 | Ly6/PLAUR domain-containing protein 3 | LYPD3 | O95274 | 22 | 0 | 6.5 |
7 | N-acylethanolamine-hydrolyzing acid amidase | NAAA | Q02083 | 13.5 | 0 | 0 |
8 | Podocalyxin | PODXL | O00592 | 25 | 0 | 0 |
9 | Seizure 6-like protein 2 | SEZ6L2 | Q6UXD5 | 15.5 | 0 | 0 |
10 | Vitronectin | VTN | P04004 | 21.5 | 0 | 0 |
Apolipoprotein D (ApoD), is a member the lipocalin family of glycoproteins. ApoD has multiple intracellular mechanistic functions, mainly binding to various ligands such as arachidonic acid and progesterone for signaling or metabolism (71). Deregulated expression of ApoD has been occurred in several pathological conditions, such as breast carcinoma, prostate cancer, neurodegenerative diseases, etc (72). Up-regulated apoD is identified in breast nipple aspirate fluid in breast cancer patients (72). ApoD is overexpressed in cyst fluid (up to 1000-fold higher than plasma) from women with gross cystic disease of the breast, a benign condition related to an enhanced occurrence of breast cancer (73). Although the information from the UALCAN database demonstrates a significant decrease in its expression in breast cancer tissues (Fig. 4). It has been shown that estradiol barricades the expression of ApoD (74). In breast cancer patients, a subgroup with positive estrogen receptor-α and negative ApoD had a greater breast cancer survival rate than those with positive ApoD (75). Enhanced levels of tumor ApoD independently predict poor outcomes in patients with breast cancer (75). This protein is a potential drug target and is helpful in the diagnosis of cancer.
Clusterin is a secreted disulfide-linked heterodimeric glycoprotein (76). This glycoprotein is induced by chemotherapy and during apoptosis in hormone-dependent tissues, such as mammary and prostate glands, as well as in several other tissues in stressful condition (76). Expression of clusterin has been reportedly related to chemotherapy resistance in different tumor cell lines, and prevention of its expression improve the sensitivity of cancer cells to chemotherapy (77). However, data obtained from the UALCAN database shows a significant reduction in clusterin expression in breast cancer patients, which might be contradictory to the earlier findings. Recently, it was indicated that clusterin-overexpressed in breast cancer cells presumably suppresses apoptosis during cellular transformation and metastasis (78). HY et al. investigated the targeted delivery of anti-clusterin small-interfering RNAs (siRNAs) in breast cancer cells showing drug resistance (79). The siRNAs were delivered by lipid-polyethylenimine hybrid nanocarriers decorated with apolipoprotein E into chemo-treated cancer cells. Their data demonstrated that the clusterin-targeting strategy is a promising treatment for chemoresistant breast cancer, increased clusterin knockdown, and enhanced chemosensitization to further paclitaxel treatment.
Gamma-interferon-inducible lysosomal thiol reductase (GILT, IFI30) is localized in phagosomes and lysosomes and is optimally active at acidic pH. It catalyzes the reduction of protein disulfide bonds and has a critical role in major histocompatibility complex (MHC) class II-restricted antigen processing (80, 81). It has been demonstrated that IFI30 in breast cancer tissues is one of the most remarkably increased genes and is related to improved disease-free survival and breast cancer-specific survival of patients (82). Nevertheless, a number of studies reported controversial data. For example, a study reported significantly reduced GILT expression in primary and metastatic breast cancer cells compared to normal epithelial cells (81). Another study manifested that IFI30 is highly expressed in breast cancer tissues and induces the proliferation of cancer cells by regulating autophagy, and it is related to the poor outcome of patients (83). They showed that the knockdown of IFI30 prevented the proliferation, migration and invasion of cancerous cells and notably barricade the growth of breast tumors in vivo (83). Thus, IFI30 has the potential to be a novel prognostic biomarker in breast cancer, and additional research is required to manifest the mechanism of its action.
Progranulin (PGRN), composed of cysteine-rich tandem repeats identified as an autocrine growth factor, which has a complex role in some biological and pathological processes, including tissue repair, embryogenesis, inflammation, tumorigenesis, and neurodegeneration (84, 85). This protein has major biological impacts on various cancers. PGRN promotes tumorigenesis through stimulation of cell proliferation, migration, invasion, malignant transformation, angiogenesis, immune evasion and resistance to anticancer drugs (84). The expression level of PGRN in normal mammary tissues is weak or zero. However, PGRN receptors expression, such as sortilin, on the surface of breast cancer cells is much more than that in normal breast tissue (85). High tumor expression of PGRN is associated with enhanced recurrence and mortality. In breast cancer patients (stage I–IV), PGRN levels in serum are significantly higher, and it was suggested that PGRN can be used as a reliable biomarker for screening individuals at high risk or as an early sign of breast cancer. Treatment with PGRN inhibitors such as siRNA, shRNA or specific neutralizing antibodies has been shown in various studies to reduce the effect of PGRN on tumorigenesis. For instance, anti-PGRN neutralizing antibody treatment in MDA-MB-468 breast cancer cells reduces their proliferation and colony formation (86, 87). Moreover, co-expression of PGRN and its receptor sortilin is suggested as a novel combined prognostic biomarker indicating an uncontrollable malignant breast cancer and potentially can be targeted for cancer treatment (88). A recent study used an orally bioavailable small molecule (AF38469) for target sortilin in a breast cancer xenograft model; as a result, they observed, blocking lung metastases and inhibiting cancer cell infiltration of the skin (89).
Ly6/PLAUR domain-containing protein 3 is a high-glycosylated cell surface protein that in humans is encoded by the LYPD3 gene. It has been shown to be overexpressed in many human malignancies, such as breast cancer, colon cancer, pancreatic cancer, acute myelogenous leukemia, and ovarian carcinoma (90). LYPD3, via interactions with cell surface integrins and regulation of the focal adhesion pathway has been associated with enhanced invasion and metastasis (90). Studies have manifested the expression of LYPD3 is sustained in anti-estrogen therapy and enhanced in aromatase and tamoxifen resistance models. Therefore, it is used as a biomarker of poor prognosis in breast cancer patients (90). In addition, it was indicated that high expression of AGR2, ligand of LYPD3, is also a predictor of poor prognosis and reduced response to endocrine treatment in luminal breast cancer patients (91). Kimberly et al. demonstrated anti-tumor effects of humanized antibodies against LYDP3, or its ligand in endocrine treatment-resistant breast tumors in animal models (90).
N-acylethanolamine-hydrolyzing acid amidase (NAAA), is a cysteine hydrolase enzyme that is highly expressed in immune cells (92). NAAA mainly localizes in lysosomes and is highly active at acidic pH. NAAA is associated with many physiological processes, such as pain and inflammation (93). Its expression and activity were also shown in human prostate cancer cell lines (94) and very aggressive ovarian cancer in mice (95). In addition, a recent study showed the important role of NAAA in colorectal cancer growth (96). They suggested that this enzyme is a valid drug target for the inhibition of colorectal cancer progression. The knowledge of NAAA is fragmentary, with an unknown mechanism in cancer. One study demonstrated that inhibition NAAA with inhibitors, including amide derivatives of hexadecylamine, reduced the proliferation and migration of bladder cancer cell lines (97). Another study using combined data from different databases revealed abnormal expression of NAAA in most malignant tumors and its association with the poor prognosis of cancer patients (98), which can be used as a prognostic biomarker and a promising target for cancer immunotherapy.
Podocalyxin (PODXL) is a highly O-glycosylated and sialylated type I transmembrane protein that is expressed by hematopoietic progenitors, vascular endothelia, platelets, kidney podocytes, as well as a subset of neurons (99). It has two diverse roles: functions as an anti-adhesion in podocytes of the renal glomeruli, regulates adhesion, cell morphology, and development of the organ; Acts as a pro-adhesive molecule, enhancing the rate of migration and cell-cell contacts by an integrin-dependent process. Overexpression of PODXL was found in several cancers, particularly those that originated from the epithelium. Results of studies propose that overexpression of PODXL is associated with more progression of the aggressive tumor, unfavorable treatment outcomes, poor prognosis, and chemoresistance (99). The mean survival time for patients with highly PODXL-positive tumors was 6 year less than the other patients (100). A study showed that silencing of PODXL significantly diminished primary tumor progression and metastasis to the bone marrow, lung, and liver in a xenograft model (101). Moreover, they used a monoclonal antibody against PODXL in mouse models, which prevented breast tumor growth and metastatic progression. However, this protein showed decreased expression in breast cancer tissue.
Seizure 6-like protein 2 (SEZ6L2), is a type 1 transmembrane protein with an extracellular region that has specialized endoplasmic reticulum function in neurons (102). Ishikawa et al. demonstrated that SEZ6L2 protein was expressed on the surface of lung cancer cells (102). They showed patients whose tumors exhibited higher expression of SEZ6L2 suffered less tumor-specific survival in contrast to those with no SEZ6L2 expression (102). Another study revealed the upregulation of SEZ6L2 in drug-resistant lung tumor spheroid cells (103). They indicated that anti-SEZ6L2 antibody therapy decreased drug resistance and tumor spheroid formation. The expression of SEZ6L2 was also notably increased in the malignant tissues of colorectal cancer patients (104). SEZ6L2 was found to be highly expressed in breast cancer and significantly is associated with the tumor, node and metastasis (TNM) stage, HER-2 status and lymph node metastasis. Its knockdown remarkably inhibited proliferation, migration and invasion of breast cancer cells (105).
3.3. Results of PPIN
The PPIN of DEGs in the study was constructed using the STRING server. Visualization and analysis of PPIN performed using Cytoscape v3.2.0 software. The drawn PPI network consisted of 177 nodes and 1181 edges, which are indicated in Fig. 5. In the network, each node and edge represents protein and physical/functional interaction between two proteins, respectively. The network analyzer tool investigates the topological parameters of PPIN, which are described in Table 5.
Table 5
Topological parameters of protein-protein interaction network of DEGs
Topological parameter | values |
Number of nodes Number of edges Clustering coefficient Network centralization Network density Network diameter | 177 1181 0.423 0.257 0.076 6 |
The hubs, bottlenecks, and hub-bottlenecks, which identified using cytohuba, are represented in Table 6. The six hub-bottlneck proteins were identified based on the greatest degree and betweenness algorithms, including as follows: FN1, ITGB1, CDH1, TFRC, LAMP1 and APP. Fibronectin 1 was determined as top scorer hub protein. Gene expression levels of top hub proteins in breast cancer tissues are also obtained from the UALCAN database (Fig. 6).
Table 6
The top 10 hub proteins were identified in the PPI network using the ctytohubba plugin based on degree, bottleneck and betweenness centrality
Top 10 hub based on degree | | Top 10 hub based on betweenness | | Top 10 hub based on bottleneck |
No | Name | Score | | No | Name | Score | | No | Name | Score |
1 2 3 4 5 6 7 8 9 10 | FN1 ITGB1 CD44 CDH1 CD9 TFRC ICAM1 CTSD LAMP1 APP | 58 49 47 41 40 39 39 38 37 35 | | 1 2 3 4 5 6 7 8 9 10 | FN1 CALR CTSD CDH1 APP LAMP1 HSP90B1 BSG TFRC ITGB1 | 2925.972 1918.457 1794.801 1634.028 1553.161 1545.324 1495.659 1472.137 1460.095 1419.841 | | 1 2 3 4 5 6 7 8 9 10 | TFRC ITGB1 CD9 HSP90B1 BSG FN1 CDH1 LAMP1 APP SLC2A1 | 11 10 9 9 9 8 7 6 6 6 |
More analysis of network by MCODE generated four significant subnetworks with 21, 11, 27 and 18 nodes in each of them. MCODE identified seed nodes in each cluster, including ITGA6, THY1, PLAUR, SLC2A1 and TMED10 which are demonstrated in Fig. 7.
Hub-bottleneck proteins play a vital role in the protein network structure and may be considered drug targets or biomarkers. Based on our results, fibronectin 1 (FN1) is a highly connected protein in PPI network and can be regard a hub protein. FN1 is an extracellular matrix glycoprotein which acts as a key protein in the communication between the intracellular and extracellular environment by binding to cell surface integrin receptors, thus controls cell behavior (106). Geng et al. reported that FN1 expression levels correlate with THCA prognosis and levels of immune infiltration. Their bioinformatics analysis suggested that FN1 could be used as an immune-related biomarker and therapeutic target in THCA (107). The analysis of FN1 mRNA and protein expression data of breast cancer patients was reported that FN1 was high expressed in breast cancer tissues compared with normal tissues. High FN1 mRNA expression was associated with poor clinical outcomes and performed well in predicting patients' survival status. FN1 may be a new prognostic biomarker and a new therapeutic target for breast cancer (108). Our findings showed that, this protein is a hub-bottleneck in the network. So, it could be a good candidate to be introduced as a marker or therapeutic target for future studies.
Integrin beta 1 (ITGB1) or CD29 is another hub-bottleneck protein in the obtained network with a high connection that is a member of the integrin family. Previous studies have suggested that ITGB1 is the most dominant integrin in normal and tumor cells. This protein controls various processes such as angiogenesis, tumor progression and metastasis (109, 110). The bioinformatics study showed the association of ITGB1 expression with chemoresistance in breast cancer. Their findings showed that ITGB1 activates the Focal Adhesion Kinase (FAK) pathway and promotes invasion, migration, and chemoresistance in breast cancer by upregulating Erk phosphorylation (111).
CDH1 is the next hub-bottleneck protein in the PPI network. E-cadherin is a membrane protein which belongs to the family of classical cadherins. This protein is involved in cell-cell adhesion mechanism, mobility and epithelial cell proliferation. Loss of E-cadherin protein is a strong sign of invasive lobular breast cancer (ILBC), and methylation of the CDH1 gene is noticed as a potential mechanism for loss of this protein expression in human breast cancer (112). According to the detected glycoproteins in the study of Timpe et al. (7), CDH1 was not expressed in the MCF-7 cell line.
PPI network analysis also identified Transferrin receptor 1 protein (TFRC) as a hub-bottleneck protein. TFRC is a protein which needed for iron transport to cells. The expression of TFRC in breast cancer tissues was identified in 87 clinical samples (113). Although, its expression in MCF-7 cell line was poorly in the data used in the present study.
Lysosome-associated membrane glycoprotein 1 (LAMP1) is a major protein in the lysosomal membrane and is expressed in the plasma membranes of malignant cells. It has been reported that lysosomes containing LAMP1 are mainly accumulated in the cytoplasm around the nucleus. The spatial distribution of lysosomes plays a role in cancer cell metastasis and drug resistance (114). In this study, LAMP1 expression has been detected in both MCF-7 and normal cell lines and detected as a hub-bottleneck protein in the PPI network.
Amyloid precursor protein (APP) is another hub-bottleneck protein detected in PPI network analysis. In detected glycoproteins in MCF-7, the APP expression is zero compared with standard cell lines. APP is a type 1 transmembrane glycoprotein, and its homologs (APLP1, APLP2) are highly conserved in mammals (115).
Conclusion
Targeted therapy is an effective type of cancer treatment, which has some advantages over chemotherapy. In this research, we did the GO analysis of 184 differentially expressed glycoproteins in the MCF7 cell line detected by Timpe et al. (7) for identifying new biomarkers and drug targets. Using GO analysis based on molecular function, we selected protein categories, including receptor proteins, proteases, and proteins involved in endocytosis and metastasis. Also, some glycoproteins are determined that expressed in the MCF7 cell line but not in standard cell lines. Also, we used the UALCAN database to analyze the expression levels of the identified genes in breast cancer tissue. Previous studies on some identified proteins introduce them as new biomarkers for targeted breast cancer therapy. We suggest that further studies should be conducted on the proteins enriched in this study to discover new diagnostic and prognostic biomarkers and drug targets for breast cancer treatment.