KR102523845B1 - Method for predicting drug resistance of cancer cells based on protein glycosylation - Google Patents

Abstract

The present invention relates to a method for predicting drug resistance of cancer cells based on protein glycosylation, and predicts the presence or absence of drug-resistant cells based on the degree of fucosylation of secretory proteins including PON1 in the serum of a patient of the present invention to determine the effect of a drug. Predictable. In addition, by measuring the fucosylation of proteins present in the patient's serum before and after administration of the anticancer drug, it can be used for selection of an anticancer drug and monitoring of treatment progress.

Description

Method for predicting drug resistance of cancer cells based on protein glycosylation}

The present invention relates to a method for predicting drug resistance of cancer cells based on protein glycosylation.

For the majority of cancer patients, a complete response to targeted therapy is rare. Therapeutic inhibition of tumor-inducing factors can stabilize disease in the long term, but inevitably results in resistance to targeted therapies. Inhibitors that target amplified or mutagenicly activated kinases (e.g., EGFR mutations or ALK translocations in lung adenocarcinoma, BRAF mutations in melanoma, or HER2 amplification in breast cancer) may respond to treatment; However, genetic and non-genetic mechanisms of resistance to these inhibitors exist. However, this resistance acquisition model excludes the importance of the tumor microenvironment (TME). For example, a complex network of signals secreted by drug-stressed tumors, called therapy-induced secretomes (TIS), can selectively increase the number of pre-existing small number of resistant clones. , which paradoxically explains the relapse of targeted therapy. A holistic understanding of drug resistance could lead to a paradigm shift in the management of clinical drug resistance in current cancer.

The cancer secretome is a collection of secreted proteins that naturally induce tumors. Many components of these secretory proteomes are biomarkers of disease and are major drug targets. The classical and non-classical pathways regulate the secretion of components including extracellular matrix proteins, exosomes, growth factors, cytokines, receptor shedding and proteases. When stress occurs, these secretory proteomic components are remodeled according to the tissue and cellular composition of the tumor microenvironment (TME), stress-induced stimuli, or conditions affecting liver homeostasis. Substantially secreted soluble proteins undergo post-translational modifications (PTMs) that result in trafficking, stability and folding prior to secretion. These post-translational modifications (PTMs) in the secretory pathway are constitutive for oncogenesis in chemotherapy, radiation therapy, targeted therapy or immunotherapy. Phosphorylation and glycosylation are the most common of these post-translational modifications (PTMs). Glycosylation (covalent association of sugar moieties with targeting scaffolds) is the most abundant PTM in secreted proteins because secreted mammalian proteins attach at least one glycan to a specific site. For example, treatment-induced apoptotic degradation of the Golgi apparatus has been associated with aberrant synthesis of certain glycan types. In some cases, direct glycosylation of apoptosis signals may inhibit or trigger cell death capacity during treatment. Moreover, therapies that act as endoplasmic reticulum (ER) stressors can inhibit protein glycosylation and reduce disulfide bonds that initiate the unfolded protein response (UPR). Although there is little evidence to suggest post-ER quality control operating in the Golgi after the UPR, stress-induced regulation of terminal glycosylation is a mechanism located in the Golgi that governs the assembly of newly synthesized secreted proteins ( It is a complementary mechanism of the Golgi-localized machinery).

Abnormal glycomes are a hallmark of cancer. Cancer-specific, the most frequent types of glycosylation are O- and N-linked glycosylation, which is the expression of genes encoding glycosyltransferases, enzymes that catalyze glycosidic linkages, and secretion pathways (Golgi and ER). It is regulated by my positioning. Malignant transformation is predominantly evident with N-glycomes. Through this process, intrinsic changes in glycan levels and composition, conjugation and connectivity are primarily reflected in cell surface, intracellular and extracellular scaffolds of lipids and proteins. Lewis antigen, a component of exocrine epithelial secretion, overexpresses the most frequently fucosylated epitope during carcinogenesis. It is mainly caused by fucosyltransferase (FUT). However, complex interactions between glycosyltransferases such as FUT and other factors regulating fucose metabolism in the Golgi apparatus/ER lead to incomplete synthesis (truncated glycosylation common in early carcinogenesis) or neosynthesis (atypical More subtle and complex dysregulation can arise from de novo generation of glycosylation patterns. As a result, various types of Lewis antigens, including sialylated isoforms, are currently being used as biomarkers for diagnosing cancer prognosis in clinical practice. Given that these glycan changes affect cancer secretory proteomes, treatment-induced remodeling of the tumor microenvironment (TME), particularly its secreted components, is accompanied by altered function at various stages of glycosylation.

Although the prior art related to the diagnostic use of small alveolar cancer for glycosylation of secretory proteins, particularly PON1, is disclosed in Korean Registered Patent No. 1013341230000, the method for predicting drug resistance of cancer cells based on protein glycosylation of the present invention is disclosed. does not exist.

Thus, the present inventors confirmed the core fucosylation of the secretory protein (TIS) induced by treatment. Specifically (i) differential induction of relatively smaller fucosylated proteins (<60 kDa), (ii) GDP- into N-glycan core structures by α1,6-fucosyltransferase (FUT8) in the Golgi. β-I-fucose (GDP-Fuc) delivery, (iii) expression of the fucose salvage gene and GDP-Fuc transporter SLC35C1, (iv) antioxidant paraoxonase 1 (PON1) It was confirmed to contain core fucosylation. In addition, deglycosylation of secreted proteins by glycosidases, inhibition of fucosylation by FUT8 or SLC35C1 RNA interference (RNAi), or site-specific blockade of PON1 core fucosylation can reduce secreted proteins induced by treatment of minority resistant clones ( TIS) was confirmed to dramatically inhibit growth. In addition, the present invention revealed that the effector of redox stress sensing and unfolded protein response (UPR) is a specific resistance modulator for secretory protein fucosylation through target screen and transcriptome-wide gene expression analysis. has been completed.

The present invention provides a method for predicting drug-resistant cancer cells by discovering biomarkers specific to drug resistance in cancer cells.

In order to achieve the above object, the present invention is 1) isolating the fucosylation (fucosylation) secretome (secretome) from the sample; and 2) measuring the degree of fucosylation of the 30 to 60 kDa secreted protein.

In the present invention, the sample may be selected from the group consisting of cells, tissues, blood, serum, plasma, saliva and urine of patients who have received drug treatment.

In the present invention, the fucosylation may be core fucosylation in which N-glycan is coupled at α-1,6. In order to achieve the above object, the present invention 1) fucosylation from a sample ) isolating the secretome; and 2) measuring the degree of fucosylation of the 30 to 60 kDa secreted protein.

In the present invention, the sample may be selected from the group consisting of cells, tissues, blood, serum, plasma, saliva and urine of patients who have received drug treatment.

In the present invention, the fucosylation may be core fucosylation in which N-glycan is coupled at α-1,6.

In the present invention, the size of the protein whose degree of fucosylation increases may be less than 100 kDa, preferably 30 to 60 kDa.

In the present invention, the secreted protein of 30 to 60 kDa is PRSS3 (Serine Protease 3), PON1 (Paraoxonase 1), HSPA7 (Heat shock 70 kDa protein 7), ACTG1 (Actin Gamma 1), CYR61 (Cysteine-rich angiogenic inducer 61) , LCAT (Lecithin-cholesterol acyltransferase), CLU (clusterin), PDIA3 (Protein disulfide-isomerase A3), P4HB (prolyl 4-hydroxylase subunit beta), STAM2 (Signal Transducing Adaptor Molecule 2) and AFP (alpha-fetoprotein) Characterized in that it is selected from the group.

In the present invention, the degree of fucosylation of the secreted protein of 30 to 60 kDa is increased in patients with drug resistance compared to normal subjects.

In the present invention, the patient is characterized in that he is selected from the group consisting of patients with liver cancer, stomach cancer, colon cancer, lung cancer, uterine cancer, breast cancer, prostate cancer, thyroid cancer and pancreatic cancer.

In addition, the present invention provides a composition for predicting drug resistance comprising an antibody specifically binding to a 30 to 60 kDa fucosylated secretome protein.

In addition, the present invention provides a kit for predicting drug resistance comprising an antibody that specifically binds to a 30 to 60 kDa fucosylated secretome protein.

In addition, the present invention provides a microarray for predicting drug resistance, characterized in that an antibody specific to a fucosylated secretome protein of 30 to 60 kDa is attached.

In the present invention, the fucosylated secretome proteins of 30 to 60 kDa are PRSS3 (Serine Protease 3), PON1 (Paraoxonase 1), HSPA7 (Heat shock 70kDa protein 7), ACTG1 (Actin Gamma 1) , Cysteine-rich angiogenic inducer 61 (CYR61), lecithin-cholesterol acyltransferase (LCAT), clusterin (CLU), protein disulfide-isomerase A3 (PDIA3), prolyl 4-hydroxylase subunit beta (P4HB), signal transducing adapter molecule 2 (STAM2) ) and at least one selected from the group consisting of alpha-fetoprotein (AFP).

The effect of the drug can be predicted by predicting the presence or absence of drug-resistant cells based on the degree of fucosylation of secreted proteins including PON1 in the serum of the patient of the present invention. In addition, by measuring the fucosylation of proteins present in the patient's serum before and after administration of the anticancer drug, it can be used for selection of an anticancer drug and monitoring of treatment progress.

Figure 1a is a diagram showing the relationship between FUT gene expression and drug resistance in several cancer types obtained from GDSC and CCLE.
Figure 1b is a diagram showing an experimental procedure for obtaining a sample with increased core fucosylation from patient serum.
Figure 1c is a diagram showing the AAL blot results for analyzing fucosylation from patient serum samples prepared in Figure 1b.
Figure 1d is a diagram showing the N-glycan release analysis from the patient serum sample prepared in Figure 1b.
Figure 1e is a diagram showing the preparation of cell-secreted protein and sandwich ELLA test method.
FIG. 1f is a diagram illustrating analysis of drug-sensitive cells by the method of FIG. 1e.
FIG. 1g is a diagram illustrating drug-resistant clones analyzed by the method of FIG. 1e.
1H is a diagram showing the results of confocal microscopy of drug-resistant clones.
FIG. 1i is a diagram showing the results of AAL blot analysis using drug-sensitive cells in the method of FIG. 1e.
Figure 1j is a diagram showing the results of AAL blot analysis using drug-resistant clones in the method of Figure 1e.
FIG. 1K is a view showing the results of sandwich ELLA analysis after filtering patient serum samples prepared by the method shown in FIG. 1B or H1993 cells prepared by the method shown in FIG. 1E with NMWL.
Figure 2a is a diagram showing a cell tracking analysis method using 'one pot' mixture culture.
Figure 2b is a view of observing the 3D mixture prepared by the method shown in Figure 2a with a confocal microscope.
Figure 2c is a diagram showing the results of Coomassie staining after separation by SDS-PAGE of the 3D mixture prepared by the method shown in Figure 2a.
Figure 2d is a diagram showing the result of sandwich ELLA analysis of the cell mixture in the condition of Figure 2c.
Figure 2e is a diagram showing the AAL blot, sandwich ELLA, N- glycan release analysis of the 2D mixture prepared by the method shown in Figure 2a.
Figure 2f is a diagram showing the results of confocal microscopy observation of GR clones of the 3D mixture prepared by the method shown in Figure 2a.
Figure 2g is a diagram showing the cell tracking results of the 2D mixture prepared by the method shown in Figure 2a.
Figure 2h is a diagram showing the results of a cell tracking experiment similar to Figure 2g, except that FUT8 or SLC35C1 siRNA was used.
Figure 2i is a diagram showing the results of sandwich ELLA analysis using the apoptotic body and secretory protein of the cell mixture of Figure 2g.
Figure 2j is a diagram showing the phosphorylation-RTK array results of the cell mixture under the same conditions as in Figure 2g.
Figure 2k is a diagram showing CM co-culture.
2l is a diagram showing colony formation of drug-resistant clones prepared as in FIG. 2k.
2m is a diagram showing ELISA sandwich-based measurement of drug-resistant clones prepared as in FIG. 2k.
Figure 2n is a diagram showing the q-PCR result of the 3D mixture prepared as in Figure 2a.
Figure 3a is a diagram showing label-free secreted protein analysis.
Figure 3b is a diagram showing the top 11 secreted proteins of 30 to 70 kDA predicted.
Figure 3c is a diagram showing the results of immunoblot and AAL blot showing the expression and fucosylation level of PON1.
Figure 3d is a diagram showing HLE analysis for measuring fucosylation of PON1.
Figure 3e is a diagram showing the results of HLE analysis using secreted proteomes of cells and drug-resistant clones.
3F is a diagram showing immunoblot results using patient serum.
Figure 3g is a diagram showing the results of AAL blotting using PON1 immunoprecipitates from patient serum.
Figure 3h is a diagram showing the results of HLE analysis using patient serum.
Figure 3i is a diagram showing the degree of fucosylation of PON1 in lung cancer patients by HLE analysis.
Figure 3j is a diagram showing the results of observing drug-resistant clones with a confocal microscope.
Figure 3k is a diagram showing the results of AAL analysis after fractionation and immunoprecipitation of H1993-GR.
3L is a diagram showing the results of HLE analysis of drug-resistant clones treated with SLC35C1 siRNA.
3m is a view showing the results of observation of SLC35C1 siRNA-treated H1993-GR with a confocal microscope.
3n is a diagram showing the results of confocal microscopy after staining drug-resistant clones with PON1, PON3, and DAPI.
Figure 3o is a diagram showing qPCR analysis for the expression of PON1 and PON3 and immunohistochemical analysis for the expression of PON3 in tissues of breast cancer patients who received multi-drug therapy.
Figure 3p is a diagram showing the degree of fucosylation and activation of PON1 after fractionation of H1993-GR treated with PON1, PON3 or SLC35C1 siRNA by HLE analysis.
Figure 3q is a diagram confirming GDP-Fuc activity using immunoprecipitates cross-linked from Golgi/ER-fractionated H1993-GR to FUT8 and PON1.
Figure 4a is a diagram showing the structure of N- glycan analyzed as a result of tandem MS / MS.
Figure 4b is a diagram showing the results of AAL blotting using PON1 immunoprecipitates of H1993-GR transduced with PON1, PON1-N253G or PON1-N324G.
Figure 4c is a diagram confirming GDP-Fuc activity using immunoprecipitates cross-linked to FUT8 and PON1 from H1993-GR transduced under the same conditions as in Figure 4b.
Figure 4d is a diagram showing the immunoblot results of H1993-GR transduced under the same conditions as in Figure 4b.
Figure 4e is a diagram showing the ELISA results of H1993-GR transduced under the same conditions as in Figure 4b.
Figure 4 is a diagram confirming the polypeptide synthesis of H1993-GR transduced under the same conditions as in Figure 4b.
FIG. 4g is a diagram showing the results of immunoblotting by CHX treatment of H1993-GR transduced under the same conditions as in FIG. 4b.
FIG. 4h is a diagram showing ELISA results of H1993-GR transduced under the same conditions as in FIG. 4b.
Figure 5a is a diagram showing the Cignal 45 pathway array of the indicated cell mixture.
Figure 5b is a diagram showing the results of observing H1993-GR under the indicated conditions with a confocal microscope.
FIG. 5c is a diagram showing the results of measuring ROS/RNS using the cell mixture shown in FIG. 5a.
Figure 5d is a diagram showing the results of measuring ROS/RNS after ATF6 siRNA-treated H1993-GR was grown in PON1-N253G-transduced H1993 cells in gefinitib-treated secretory protein.
5E is a diagram illustrating a 'sequential layer' method.
Figure 5f is a diagram showing the results of cell tracking analysis after ATF6 siRNA treatment on the 3D mixture prepared as in Figure 5e.
Figure 5g is a diagram illustrating the analysis of cytokine secretion from the secretory protein of the cell mixture prepared as in Figure 2b.
Figure 6a is a diagram of analyzing genes differentially expressed under the indicated conditions.
6B is a Venn diagram showing up-regulated and down-regulated genes under the indicated conditions.
Figure 6c is a diagram showing the top 20 genes shown in Figure 6d.
6D is a diagram analyzing first progression survival (FPS) or relapse-free survival (RFS) from a cohort of lung cancer patients.

Hereinafter, the present invention will be described in detail.

On the other hand, the embodiments of the present invention can be modified in various other forms, and the scope of the present invention is not limited to the embodiments described below. In addition, the embodiments of the present invention are provided to more completely explain the present invention to those skilled in the art.

The present invention comprises the steps of 1) separating a fucosylated secretome from a sample; and 2) measuring the degree of fucosylation of the 30 to 60 kDa secreted protein.

In the present invention, the sample is characterized in that it is any one selected from the group consisting of cells, tissues, blood, serum, plasma, saliva and urine of patients who have received drug treatment.

In the present invention, the fucosylation is characterized in that it is core fucosylation.

In the present invention, the secreted protein of 30 to 60 kDa is PRSS3 (Serine Protease 3), PON1 (Paraoxonase 1), HSPA7 (Heat shock 70 kDa protein 7), ACTG1 (Actin Gamma 1), CYR61 (Cysteine-rich angiogenic inducer 61) , LCAT (Lecithin-cholesterol acyltransferase), CLU (clusterin), PDIA3 (Protein disulfide-isomerase A3), P4HB (prolyl 4-hydroxylase subunit beta), STAM2 (Signal Transducing Adaptor Molecule 2) and AFP (alpha-fetoprotein) It is characterized in that at least one selected from the group.

In the present invention, the degree of fucosylation of the secreted protein of 30 to 60 kDa is characterized in that fucosylation is increased in drug-resistant patients compared to normal subjects.

In the present invention, the patient is characterized in that he is selected from the group consisting of patients with liver cancer, stomach cancer, colon cancer, lung cancer, uterine cancer, breast cancer, prostate cancer, thyroid cancer and pancreatic cancer.

Methods for measuring the degree of fucosylation include sandwich ELLA analysis, N-glycan release analysis, AAL blot (Aleuria aurantia lectin blotting), HLE (hybrid lectin ELISA) analysis, glycoprotein staining, lectin fluorescent staining, Conventional enzyme immunoassay (ELISA), radioimmunoassay (RIA), sandwich assay, western blotting, immunoprecipitation, immunohistochemical staining, flow cytometry, It can be performed using methods such as fluorescence-activated cell sorting (FACS), enzyme substrate staining, and antigen-antibody aggregation.

In addition, the present invention provides a composition for predicting drug resistance comprising an antibody specifically binding to a 30 to 60 kDa fucosylated secretome protein.

In the present invention, the antibody may be a whole antibody (hereinafter referred to as "whole antibody") or a functional fragment thereof. The whole antibody may be in the form of a monomer or a multimer in which two or more whole antibodies are bound.

The functional fragment of the antibody is an antibody having heavy and light chain variable regions of the entire antibody, and means to recognize substantially the same epitope as that of the entire antibody. Functional fragments of the antibody include single-chain variable region fragments (scFv), (scFv)2, Fab, Fab' and F(ab')2, etc., but are not limited thereto. The single-chain variable region (scFv) refers to an antibody fragment taking the form of a single-chain polypeptide in which a heavy chain variable region and a light chain variable region are connected through a linker peptide.

The antibody may be modified by binding to various molecules such as enzymes, fluorescent materials, radioactive materials, and proteins. Modified antibodies can be obtained by chemically modifying an antibody. This modification method is commonly used in the art. In addition, the antibody is obtained as a chimeric antibody in which a variable region derived from a non-human antibody and a constant region derived from a human antibody are combined, or a complementary antibody derived from a non-human antibody is obtained. It can be obtained as a humanized antibody in which a frame work region (FR) derived from a human antibody, including a crystal region, and a constant region are combined. Such antibodies can be prepared using methods known in the art.

The composition for prediction of the present invention may further include reagents known in the art used in immunological analysis in addition to protein-specific antibodies. In the above immunological analysis, any method capable of measuring the binding of an antigen and an antibody may be included.

In addition, the present invention provides a kit for predicting drug resistance comprising an antibody that specifically binds to a 30 to 60 kDa fucosylated secretome protein.

The prediction kit may be provided in the form of, for example, a lateral flow assay kit based on immunochromatography to detect a specific protein in a sample. The lateral flow assay kit usually consists of a sample pad to which a sample is applied, a releasing pad coated with an antibody for detection, and a membrane for development where the sample moves and separates and where the antigen-antibody reaction occurs (e.g. nitrocellulose) or strips, and an absorption pad.

In addition, the present invention provides a microarray for predicting drug resistance, characterized in that an antibody specific to a fucosylated secretome protein of 30 to 60 kDa is attached.

In the microarray, antibodies are generally attached to a slide glass surface treated with a specific reagent so that a protein specifically attached to the antibody can be detected through an antigen-antibody reaction.

In the present invention, the fucosylated secretome proteins of 30 to 60 kDa are PRSS3 (Serine Protease 3), PON1 (Paraoxonase 1), HSPA7 (Heat shock 70kDa protein 7), ACTG1 (Actin Gamma 1) , Cysteine-rich angiogenic inducer 61 (CYR61), lecithin-cholesterol acyltransferase (LCAT), clusterin (CLU), protein disulfide-isomerase A3 (PDIA3), prolyl 4-hydroxylase subunit beta (P4HB), signal transducing adapter molecule 2 (STAM2) ) and at least one selected from the group consisting of alpha-fetoprotein (AFP).

In addition, after separating and concentrating fucosylated glycoproteins from H1993 cells, H1993 cells resistant to gefitinib, and H1993-GR clones, they were separated on a 1D-SDS PAGE gel and subjected to in-gel trypsin digestion (In gel tryptic digestion). digestion) to pre-treat fucosylated proteins of 30 to 70 kDa into a peptide state, and then analyzed using a mass spectrometer. Three sample groups were compared and analyzed in drug-sensitive cells and drug-resistant clones compared to drug-sensitive cells. 11 secretory proteins with increased fucosylation [PRSS3 (Serine Protease 3), PON1 (Paraoxonase 1), HSPA7 (Heat shock 70kDa protein 7), ACTG1 (Actin Gamma 1), CYR61 (Cysteine-rich angiogenic inducer 61), LCAT (Lecithin-cholesterol acyltransferase), CLU (clusterin), PDIA3 (Protein disulfide-isomerase A3), P4HB (prolyl 4-hydroxylase subunit beta), STAM2 (Signal Transducing Adaptor Molecule 2) and AFP (alpha-fetoprotein)] were identified. (Fig. 3b).

Hereinafter, the present invention will be described in detail by the following Examples and Experimental Examples.

However, the following Examples and Experimental Examples are only to illustrate the present invention, and the content of the present invention is not limited by the following Examples and Experimental Examples.

<Example>

Experimental methods and materials

1. Data reporting and statistics

All quantitative data are expressed as mean ± SD unless otherwise indicated. Student's t, Mann-Whitney, Dunnett, Wilcoxon rank-sum, Mantel-Cox and chi-squared tests; ROC analysis was performed with GraphPad Prism 8.4. The number of samples or biological replicates (n) is indicated in each figure.

2. Human Cancer Patient Samples and Ethics Statement

All human blood and tissues from three groups of patients diagnosed with lung adenocarcinoma, squamous cell carcinoma, or breast cancer were collected and analyzed by approved protocols in accordance with the ethical requirements and regulations of the Institutional Ethics Review Committee of Seoul National University Hospital after informed written consent (IRB No. 1104-086-359 and B-1201/ 143-003). All samples were randomly selected and classified. Patients underwent surgical resection of primary or metastatic tumors at Seoul National University Hospital. Tissue and blood samples were obtained by core needle biopsy. Fifty-three pairs of lung cancer tumor tissues and adjacent normal tissues were obtained intraoperatively. All patients in this cohort did not receive preoperative chemotherapy. No postoperative or preoperative radiotherapy was performed in this cohort. Blood and tissue processing and histopathology data interpretation were supervised by expert pathologists (Ryu Han-Seok, Cho Seok-Ki, and Kim Tae-Min). The cohort's clinicopathological information was extracted from medical records and anonymized. Source DNA and RNA were extracted from formalin-fixed paraffin-embedded (FFPE) tumors and adjacent normal tissues. Lysates were obtained from frozen tumors. Frozen samples were snap-frozen in liquid nitrogen and stored at -80 °C. To obtain plasma, samples were centrifuged at 1,600g for 10 minutes and then further separated at 20,000g for 10 minutes within 1 hour after blood collection. To obtain serum, blood was allowed to clot at room temperature (RT) for 15–30 min before centrifugation. All aliquots were stored at -80 °C and thawed no more than twice before use. Multiple Affinity Removal System (MARS) HSA/IgG spin columns (Agilent) were used to remove albumin and IgG from blood samples. Samples were then run through Amicon Ultra-2 mL Centrifugal Filters [Merck; 3k molecular weight cut-off (MWCO)] and enriched according to the manufacturer's instructions.

3. Cultivation of cell lines

Human H292, H1993, H358, HCC4006, H460, H1299 and A549 cell lines [American Type Culture Collection (ATCC) nos. CRL-1848, CRL-5909, CRL-5807, CRL-2871, HTB-177, CRL-5803 and CCL-185; obtained from 2014 to 2016] were supplemented with 10% fetal bovine serum (FBS) alternative fetal bovine growth serum (RMBIO) or EqualFETAL bovine serum (Atlas bios), 2 mM L-glutamine and penicillin (100 U/ml)-streptomycin ( Cultured under standard conditions in RPMI 1640 (Welgene) supplemented with 100 μg/ml; Invitrogen).

PC9 and HCC827 [J. provided by K. Rho (Seoul Asan Medical Center, University of Ulsan)], EBC-1 [Japanese Collection of Research Bioresources (JCRB) Cell Bank no. JCRB0820], HCC78 [German Collection of Microorganisms and Cell Cultures (DSMZ) GmbH no. ACC563], H3122 [originally provided by PA Janne (Dana-Farber Cancer Institute, Boston, MA, USA)] and SKBR3 [provided by DM Helfman (KAIST, Daejeon, Korea)] are all cell lines obtained in 2017. It was cultured in RPMI 1640 under the same conditions.

Human H1975, H2009, A375 and HEK-293T cell lines [ATCC nos. CRL-5908, CRL-5911, CRL-1619 and CRL-3216; obtained in 2017] in Eagle's minimum essential medium (Merck), DMEM/F12 (Gibco) and DMEM (Welgene) containing no L-glutamine and additionally containing 1μg/mL amphotericin B Except for that, it was cultured under the same conditions as above. Mice LL/2 (LLC1; ATCC number CRL-1642; obtained in 2015) were cultured in BME with Earle's salts (Merck) using the same supplement as above. All cells were cultured in a humidified incubator at 37° C. with 5% CO ₂ and were regularly tested for mycoplasma infection. All cell lines used were negative for mycoplasma (Cosmogenetech mycoplasma test service).

4. Drug Resistant Clones

To generate drug-resistant clones, sensitive cell lines were seeded at low densities and continuously exposed to gradually increasing drug concentrations for 12 to 52 weeks. All clones were derived and expanded in colonies and maintained at specific drug concentrations. Clones were passaged every 2 or 3 days with the addition of new drug concentrations.

5. Preparation of secreted cell proteins

To prepare the secretome, 3x10 ⁶ sensitive cells and 7x10 ⁶ drug resistant clones were plated in 15 cm plates using standard media and allowed to attach overnight. The medium was then replaced with fresh medium supplemented with 2% dialyzed FBS and labeled drug for 48 hours. FBS was dialyzed against 0.15M NaCl at 4°C for 6 hours using 10k MWCO dialysis tubing (Fisher Scientific) until glucose reached below 5mg/dL. The secretome was centrifuged at 1,000 rpm. Vacuum filtered using 0.45 μm cellulose acetate membranes (Whatman) for 5 min and immediately placed on ice.

For 2D co-culture, secretomes were stored at 4°C, warmed before use, and used only within 48 hours.

For 3D co-culture, only freshly prepared secretomes were used and further concentrated using Amicon Ultra-15 mL Centrifugal Filters (3k MWCO).

For biochemical analysis, the secretome was further concentrated using Amicon Ultra-15 mL Centrifugal Filters (3k, 10k, 30k, 50k, 100k MWCO markings) and MARS HAS/IgG spin columns. was used to remove albumin and IgG. Aliquots were flash frozen in liquid nitrogen and stored at -80 °C until use, allowed to thaw only once.

6. Fucosylation Assay

(i) AAL-enrichment assay

To enrich core-fucasylated protein/lipid samples, AAL (Aleuria aurantia lectin) was used as a carbohydrate probe for core fucose, N-acetylglucosamine Scaffolds with fucose linked (α1,6) to N-acetylglucosamine or linked (α1,3) to N-acetyllactosamine-related structures were captured (Fig. 1b). In addition, AAL binds reversibly to fucose attached to nucleic acids. Bio-spin columns (Bio-Rad) are AAL (Vector Laboratories) coupled with 1.5 mL agarose beads. Agarose beads were maintained at 4° C. in inhibition solution [10 mM HEPES (pH 7.5), 0.15 M NaCl, 10 mM fucose, 0.04% NaN3]. The secretome (500 μL) or serum/plasma (40 μL) was thawed on ice at 4°C, diluted with 1.5 mL AAL adsorption buffer (AffiSpin-AAL kit; GALAB), and incubated on ice for 5 days. minutes, and incubated for at least 12 hours at 4° C. using the Biospin. Unbound proteins/lipids were passed through by gravity only and removed by washing with adsorption buffer and PBS. Fucosylated proteins were eluted twice for 1 hour each at 4° C. with 50 μL AAL elution buffer B1 or 40 μL of glycoprotein eluting solution for fucose-binding lectins (Vector Laboratories). . Remaining bound fucosylated proteins were forced to elute. Samples were scaled up to produce at least 80 μL of eluted protein. All reagents and columns were pre-cooled on ice before use. Eluted proteins were precipitated using the trichloroacetic acid (TCA)-sodium deoxycholate (DOC) method. Protein concentration was measured using Bradford reagent (Bio-Rad).

(ii) sandwich ELLA analysis

A sandwich ELLA assay was developed to quantify fucosylated proteins in AAL-rich samples. 96-well microtiter plates (Koma Biotech) _were plated with 0.4 _μg MAL II, SNA _, LCA, _BTL , PSA, It was coated with UEA1, ConA, or RCA1 lectins (Vector Laboratories) at 37° C. for 2 hours. Plates were then incubated with 0.1 mL oxidation buffer (20 mM NaIO4) per well. The lectin solution was removed by washing three times with PBS-Tween-20 (0.05%; PBST). Then, the plate was blocked for 1 hour at room temperature using PBST containing 3% bovine serum albumin (BSA). Concentrated secretome, lysate or serum/plasma was added to each well and incubated for 2 hours at room temperature. The plate was gently washed 3 times with PBST to remove unbound proteins. 100 μL of 4 μg/mL biotinylated AAL (Vector Laboratories) was added and incubated for 90 minutes at room temperature. The lectin solution was removed, HRP-conjugated streptavidin (Biolegend) was added, incubated at room temperature for 90 minutes, and then washed twice with PBST. For detection, 1-Step Turbo TMB-ELISA substrate solution (Thermo Scientific) was used. Absorbance was measured at 450 nm using a microplate reader (VersaMax, Molecular Devices).

(iii) N-glycan release assay

20 μL concentrated sample was mixed with 2.5 μL sodium phosphate or citrate buffer (500 mM, pH 7.5) and 10 μL 8U PNGase F or 10U Endo F/S, followed by 12 hours in a humidified chamber. Incubated at 37°C and heated at 95°C for 5 minutes. The reaction was then mixed with 20 μL 2.5 M TCA solution, vortexed for 5 minutes and centrifuged at 12,000 g for 30 minutes. 15 μL supernatant was mixed with 7.5 μL of 4M NaOH, 12.5 μL 1.7 mM WST-1 aqueous solution and incubated at 50° C. for 1 hour. For in-gel N-glycan release, proteins in the gel were digested with trypsin overnight. Samples containing released peptides were reduced in a SpeedVac until a minimum of 10 μL was reached. The reduced sample was mixed with 10 μL H ₂ O and the same protocol as mentioned above was used. Absorbance was measured at 584 nm. The released N-glycan was quantified using maltose (Sigma).

(iv) AAL blotting

AAL-enriched precipitated samples (10-15 μg concentrated cell secretion or 3-5 μg serum/plasma protein) were loaded on 12% SDS-PAGE. After transferring the gel to nitrocellulose membranes (Whatman), the membrane was blocked (4° C., <2 hours) with 1x Carbo-free blocking solution (Vector Laboratories). Then, it was incubated with 5 to 20 μg/mL of biotinylated AAL (room temperature, 1 hour), washed three times with PBST, and then incubated with HRP-conjugated streptavidin (room temperature, 1 hour). ) was done. Finally, after washing three times with PBST, it was developed using an ECL system (Amersham).

(v) SDS-PAGE gel glycoprotein staining

Glycoprotein staining of SDS-PAGE gels was performed using the GelCode glycoprotein staining kit (Pierce) according to the manufacturer's protocol. Stained glycol appears as a magenta/pink band.

(vi) HLE (hybrid lectin ELISA) analysis

An ELISA starter kit (Koma Biotech) was used for HLE (hybrid lectin ELISA) of target protein fucosylation. Specifically, 120 ng PON1 (18155-1-AP, Protein Tech), AFP (ab3980, Abcam), or A1AT (ab9399, Abcam) monoclonal antibody was placed in a 96-well microtiter plate and 100 Coating was performed at 37° C. for 3 hours using μL coating buffer. 100 μL of oxidation buffer was added per well for 30 minutes, and then blocked for 2 hours at room temperature using PBS supplemented with 3% BSA. Plates were washed 4 times with PBST. All AAL-enriched samples were diluted 10-fold in PBS, and 100 μL of each sample was added to each well and incubated for 2 hours at room temperature. After washing several times with PBST, 2 μg/mL biotinylated AAL was added and incubated for 90 min at room temperature. The lectin solution was removed, HRP-conjugated streptavidin (Biolegend) was added, incubated at room temperature for 90 minutes, and then washed twice with PBST. For detection, 1-Step Turbo TMB-ELISA substrate solution (Thermo Scientific) was used. Absorbance was measured at 450 nm using a microplate reader (VersaMax, Molecular Devices).

(vii) Lectin fluorescence staining

For lectin fluorescence staining of cells and paraffin sections, 15 μg/mL fluorescein-labeled AAL (Vector Laboratories) or 4 μg/mL FITC-conjugated UEA1 ( FITC-conjugated UEA1, Thermo Scientific) was used.

(viii) GDP-Fuc activity assay of FUT8

GDP-Fuc activity of FUT8 was assayed using the GDP-Glo glycosyltransferase assay kit (Promega) according to the manufacturer's protocol. Luminescence was read using a luminometer (POLARstar Omega).

7. Cell tracking experiments

To fabricate a 3D tumor spheroid array, a polydimethylsiloxane (PDMS)-based positive master mold with an array of 225 spherical microwells (15x15) was prepared. The mold was immersed in 70% (v / v) ethanol and sterilized under UV for 30 minutes before use. Agarose powder (LPS solution) was added to RPMI 1640 at a concentration of 3% (w/v) and heated for a short time to completely dissolve. Before gelation, the completely dissolved agarose solution was poured into a 35 mm cell culture dish (3 mL/dish, SPL), and a sterilized master mold was immediately inserted into the gel solution to create microwells. After solidifying the agarose at room temperature for 20 minutes, the master mold was gently removed. PBS (3mL/dish) was added to the agarose-based microwells to keep them hydrated prior to use. Cell mixtures seeded in these microwells immediately formed spheroids (Fig. 2B).

For tracking experiments in cell mixtures, sensitive cells are labeled with CellTracker-Green (CMFDA) and drug-resistant clones are labeled with CellTracker-Red (CMTPX) or CellTracker-Deep Red (Thermo Scientific) or pCAG-LifeAct according to the manufacturer's protocol. -RFP (Ibidi). Clones labeled with RFP were stably selected using geneticin (Thermo Scientific) according to the manufacturer's instructions. The labeled cells and clones were filtered using a cell strainer (Merck) and seeded in 'one pot' or 'sequential layer' 2D and 3D mixtures (Fig. 2a). 3D mixtures were imaged live and 2D mixtures were fixed with 3.7% formaldehyde in PBS for 5 minutes and washed with PBS before imaging. Live imaging of 3D tumor spheroids was performed using a fluorescence inverted microscope (Nikon Eclipse Ti) equipped with a CFI Apochromat TIRF objective. Time-lapse images were obtained at 15 min frame intervals to minimize photobleaching and phototoxicity due to high illumination and analyzed as 3D reconstructions of stacked axes. Imaging of the 2D mixture was performed using a fluorescence inverted microscope (Leica DMI3000 B). Cell tracking and fluorescence analysis were performed using the Fiji/ImageJ plug-in TrackMate.

8. Plasmids, RNAi and Transfection

60 nM to 120 nM target-specific smart pools of short interfering RNA (siRNA) or non-targeting scrambled siRNA/siLuciferase (IDT Korea) were prepared by Lipofectamine 3000 (Life Technologies) or DharmaFECT (Dharmacon). Targeting siRNAs were purchased from IDT Korea, Life Technologies or Bioneer. Unless otherwise specified, assays were performed 48 hours after transfection. The pCMV6-AC-Myc-DDK and pCMV6-FUT8-Myc-DDK (Origene) expression plasmids were delivered using Lipofectamine 3000 or FuGene 6 (Promega) according to the manufacturer's instructions. Transfection was performed for 48 hours.

To establish PON1 and SLC35C1 knockout (KO) cells, pLKO.1-puro or pLKO.1 plasmids (TRC shRNA library, selected from Broad, purchased from Origene) encoding target shRNA constructs were cloned. . The sequence of the construct was verified by DNA sequencing (Origene). Scrambled shRNA (Addgene) was used as shControl. Lentiviral co-transfection of HEK293T cells with 8 μg of cloned transgene plasmid, 1 μg pMD2.G (packaging plasmid; Addgene), and 3 μg psPAX2 (packaging plasmid; Addgene) using iN-fect (Intron Biotechnology) according to the manufacturer's protocol. and transduction was performed in the indicated cell lines using standard procedures. Lentivirus titer was determined using the Lenti-X p24 rapid titer kit (Takara Bio). To select stable cell lines, 2 to 8 μg/mL puromycin was gradually added for 2 weeks. Stably selected KO cells were maintained in complete medium containing 0.1 μg/mL puromycin.

To establish PON1 overexpressing cells, bicistronic pLVX-EF1α-IRES-puro (Takara Bio) encoding the CDS of human PON1 single mRNA transcript was cloned. An empty vector was used as a control (-CC). Co-transfection of plasmids, transduction and selection were performed as above except that infected cells were selected in 1 μg/mL puromycin.

9. Site-directed mutagenesis

To create the mutant PON1 construct, PCR-based Q5 site-directed mutagenesis kit (NEB) was used according to the manufacturer's instructions. The PON1 cDNA template was cloned into pcDNA3.1 (Genscript). Mutagenesis primers were designed using Primer X Tool (http://bioinformatics.org/primerx/). full-length full-length; FL) or mutant PON1 constructs were transfected using Xfect transfection reagent (Clontech) according to the manufacturer's protocol.

10. Gene expression analysis

Total RNA (1-3 μg total per 10 μL volume) was isolated using the RNAeasy mini kit (QIAGEN) or TRIzol (Life Technologies) according to the manufacturer's protocol. Tumor tissues were homogenized in a hand-held homogenizer using RLT-ME buffer (Qiagen). Complementary DNA (cDNA) was made using the Transcriptor First Strand cDNA synthesis kit (Roche). RNA was treated with deoxyribonuclease I (DNase I; Takara) and reverse transcribed using RevertAid reverse transcriptase (Fermentas). cDNA was amplified using SYBR Green PCR master mix (Applied Biosystems). Differential RNA levels were analyzed using Taqman Gene Expression Analysis (Life technologies). Quantitative polymerase chain reaction (qPCR) was performed using the SureCycler 8800 (Agilent) and AriaMx Real-Time PCR System (Agilent). Relative gene expression was normalized to internal control genes such as GAPDH or ACTB. Experiments were performed according to the manufacturer's protocol using the FFPE All-Prep kit (QIAGEN) for nucleic acid extraction (total RNA and genomic DNA) from FFPE tumor samples. Small sections of specimens were prepared from approximately 80 μm slices of FFPE tumor blocks and then dewaxed using Deparaffinization Solution (QIAGEN). Purified RNA was used for reverse transcription PCR (RT-PCR) and qPCR as above.

11. Immunoblotting and immunoprecipitation

Golgi and ER were enriched using the Minute Golgi/ER enrichment kit (Invent Biotech), and nuclear and cytoplasmic extracts were isolated using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Pierce) according to the manufacturer's instructions.

After surgery, xenograft tumors were flash frozen in liquid nitrogen. Portions of frozen tumors excised from mice were thawed on ice and homogenized in Complete Lysis Buffer (Active Motif) for total lysate extraction. Protein concentration was determined using Bradford reagent. Samples were boiled in Laemmli buffer for 5 minutes. Equal amounts of protein (30-50 μg) were separated by SDS-PAGE (typically 7.5, 10 and 12% gels).

For immunoprecipitation, PON1 monoclonal antibody was bound to protein G-Sepharose 4B beads (GE Healthcare) and eluted. Proteins were transferred to Immobilon PVDF membranes (Millipore) using a semidry transfer system (Amersham). Detection was performed using Clarity Max Western ECL Substrate (Bio-Rad) and Western Lightning Plus-ECL (PerkinElmer). As the secondary antibody, a goat antibody to mouse immunoglobulin G-horseradish peroxidase (IgG-HRP; DACO) mouse IgGκ (Santa Cruz Biotechnology) or a donkey antibody to rabbit IgG-HRP (GE Healthcare) was used. For cross-linking, cells were starved in medium containing 2% dialyzed FBS before cross-linking (using 1 mM EGS for 45 min at 4°C). Specifically, lysates were diluted 2-fold in assay buffer and incubated overnight with capture beads for FUT8 (protein G-agarose beads; Abcam). The lysate was clarified by centrifugation at 16, 400 g for 15 min and then pre-clarified with agarose resin for 30 min. Lysates were then incubated overnight at 4°C with protein A/G agarose and PON1 antibody. The next day, the resin was washed 6 times with lysis buffer and then incubated with 2M hydroxylamine HCL in phosphate buffered saline (pH 8.5) at 37°C for 6 hours. The resin was then removed and the supernatant used for the indicated assay. Primary antibodies used for immunoblotting were PON1 (ab24261, Abcam), RCAS1 (12290, CST) and GAPDH (6C5, Santa Cruz Biotechnology). Antibodies used for immunoprecipitation were PON1 (17A12, Santa Cruz Biotechnology) and FUT8 (ab191571, Abcam).

12.ELISA

A Sandwich-based ELISA kit was used to detect PON1 (RayBiotech), FUT8 (LSBio) and ATF6 (Novus Biologicals) according to the manufacturer's protocol. For ELISA detection of SLC35C1, 96 well microtiter plates were pre-coated with an HLE-like SLC35C1 antibody (CSB-PA839285LA01HU, Cusabio). Absorbance was measured at 450 nm.

13. Polypeptide synthesis assay

To detect early protein synthesis, EZClick global protein synthesis kit (Biovision) was used according to the manufacturer's protocol. This assay is based on the alkyne analog of puromycin, O-Propargyl-puromycin (OP-puro). OP-puro stops translation by forming a covalent link with the nascent polypeptide chain. The cleaved polypeptide is rapidly converted by the proteasome and can be detected through a click reaction with fluorescent azide. Fluorescence was measured by flow cytometry using an LSR Fortessa (BD Biosciences). Excitation and emission wavelengths were set to 440 and 530 nm, respectively. Analysis was performed using BD FACSDiva software.

14. Trypsin Sensitivity and CHX Trace Assay

Lysates were trypsinized to assess the folding state of PON1. Specifically, the lysate was clarified by centrifugation at 17,800 g at 4° C. for 10 minutes. Clear lysates of 1 mg/mL 50 μL aliquots were treated with 2 μL of the indicated trypsin concentration (Promega), at 4° C. for 15 minutes. 50 μL stop buffer (1x SDS sample buffer, 100 mM dithiothreitol, 10x protease inhibitor cocktail) was added to the samples and incubated at 100° C. for 5 minutes. 30 μg of each sample was separated by SDS-PAGE and subjected to immunoblotting.

To evaluate PON1 stability, cells cultured in 6-well plates were incubated with 25 μg/mL CHX (Sigma) at indicated times and used for immunoblotting or other analyses.

15. Phospho-RTK Array and Kinase Phosphorylation Assay

Phosphorylated RTK was measured using the PathScan human RTK signaling antibody array kit (R & D Systems) according to the manufacturer's instructions. Tyrosine 1068 phosphorylation of EGFR, pan-tyrosine phosphorylation of MET and HER3/ErbB3, and tyrosine 1150/1151 phosphorylation of insulin receptor β were evaluated by solid-phase sandwich ELISA (CST PathScan kits) according to the manufacturer's protocol. This assay quantitatively detects endogenous levels of the indicated target. Absorbance was measured at 450 nm.

16. Cignal 45 Pathway Reporter Array

The activities of 45 transcription factors/signal transduction pathways were measured simultaneously using Cignal 45-pathway reporter arrays (QIAGEN) according to the manufacturer's protocol. Specifically, cell mixtures grown for 5 days in various conditions were transferred to Cignal Finder 96-well plates (minimum 30,000 cells/well). A reporter construct residing in each well was introduced into the cells by reverse transfection. Cell mixtures were cultured for 48 hours in Opti-MEM (Gibco) supplemented with 5% FBS and 0.1 mM MEM nonessential amino acids (NEAA; Gibco). Then, the cell mixture was lysed and the luciferase activity was measured using the dual-emission optics of a plate reader (POLARstar Omega).

17. Proliferation, survival, cell cycle, apoptosis and senescence assays

Cell proliferation and survival were assessed by sulforhodamine B (SRB) and colony formation assays. Cell sorting, cell cycle analysis by DNA content quantification, and cell death detection at the sub-G1 peak were performed by flow cytometry using a FACSCalibur (BD Biosciences). Analysis was performed using BD CellQuest Pro software. A minimum of 20,000 cells were used for each assay. Changes in cell distribution proportions at each stage of the cell cycle were determined. To isolate apoptotic bodies, cells grown under the specified conditions were transferred to serum-free medium containing 0.35% BSA, and cell debris was collected after 24 hours. Cells were centrifuged at 300 g for 10 minutes, and remaining cell debris was removed to obtain soluble secreted protein. The mixture was centrifuged at 16,500g for 20 minutes using an ultra high speed vacuum centrifuge (Vision Scientific).

To detect aging, SA-β activity was measured using a senescence βgalactosidase staining kit (CST) according to the manufacturer's protocol. SA-β-positive cells were quantified based on three independent images of different staining areas analyzed with a digital inverted light microscope (40x phase contrast, Leica DMi1). Gene expressions of p16, LNMB1, IL-1α, IL-6, MMP-3, MMP-9, CXCL-1, CXCL-10 and CCL20 were measured by qPCR to evaluate the aging gene signature and SASP activity.

18. Caspase activity and intracellular ATP assay

Caspase 3/7 and 9 activities were evaluated using the fluorescence-based Apo-ONE homogenous caspase 3/7 assay kit (Promega) and the luminescence-based caspase-glo 9 assay system (Promega), respectively, according to the manufacturer's protocol. Excitation and emission wavelengths were set to 560 and 590 nm, respectively. Luminescence was read on a luminometer.

For ATP measurements, cells were placed in 96 well plates and subjected to the indicated treatment/culture conditions in nutrient-limited medium (10% dialysis FBS). ATP levels were measured using a luminescence-based ATPLite system (Perkin-Elmer) according to the manufacturer's instructions.

19. Enzyme activity assay

Paraoxonase activity was evaluated based on 4-nitrophenol formation. Paraoxon (O,O-Diethyl O-(4-nitrophenyl) phosphate; Sigma) was used as a substrate. Absorbance was measured at 412 nm. One unit of paraoxonase activity was defined as 1 nM 4-nitrophenol formed per minute. Arylesterase activity was evaluated based on phenol formation. Phenylacetate (Sigma) was used as a substrate and absorbance was measured at 217 nm. One unit of arylesterase activity is equal to 1 mM of phenylacetate hydrolyzed per minute.

20. ROS/RNS and Cytokine Measurements

Free radical ROS/RNS was measured by OxiSelect in vitro ROS/RNS assay kit (Cell

Biolabs) was measured according to the manufacturer's protocol. This assay used a DCFH probe and oxidation reactions were measured against H ₂ O ₂ or DCF standards. Excitation and emission wavelengths were set to 480 and 530 nm, respectively.

IL-6, TFN-α and GM-CSF levels were quantified using ELISA kits pre-coated with the specified capture antibodies according to the manufacturer's instructions (Sigma). IL-6 levels were previously detected using the Q-Plex Human cytokine screen (16-plex; Quansys Biosciences). Absorbance was measured at 450 nm.

21. Immunofluorescence

Cells were plated in 0.1% gelatin coated glass bottom 30 mm dishes (except for 3D tumor spheroids) and cultured overnight unless otherwise specified. Cells were fixed with 4% paraformaldehyde in PBS at room temperature for 8 minutes, quenched in 10 mM Tris in PBS at room temperature for 1 minute, and permeabilized in 0.1% Triton X-100 in PBS. Cells were blocked with 2% bovine serum albumin (BSA) (PBS containing 0.01% Triton X-100) for 30 minutes at room temperature and incubated for 2 hours with primary antibody diluted in 2% BSA. . Primary antibodies were labeled using Alexa Fluor or fluorescein isothiocyanate-conjugated secondary antibodies. Nuclei were stained using DAPI (4', 6-diamidino-2-phenylindole; 0.35 μg/ml). Cells were embedded using VECTASHIELD Mounting Medium (Vector Laboratories). Confocal microscopy was performed using a ZEISS LSM 780 ApoTome microscope (Carl Zeiss) using a C-Apochromat 40x lens with a numerical aperture of 1.20. Primary antibodies used for immunofluorescence were PON1 (ab24261, Abcam), PON3 (ab42322, Abcam), ATF6 (PA5-20215, Invitrogen), RCAS1 (12290, CST) and XBP1 (ab37152, Abcam).

22. Immunohistochemistry

Human or mouse FFPE tumor tissue sections were deparaffinized in xylene substitute (Histo-Clear, EMS; 3 × 5 min) and plated in an EtOH/H ₂ O gradient series (100%, 95%, 70%, 40%, 5 min each). minutes) and rehydrated. The rehydrated part was washed in TBS for 10 minutes. The epitope was revealed using preheated Tris (10 mM Tris base, 0.05% Tween 20, pH 8.0) and citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0). Sections were pressure conditioned in Tris buffer for 12 minutes, transferred to citrate, heated for 12 minutes, cooled at room temperature for 40 minutes, and washed in TBS containing 0.1% Twee-20 for 10 minutes. Sections were permeabilized with 0.3% Triton X-100 in TBS for 45 minutes and washed in TBS (2 x 5 minutes). Endogenous peroxidase activity was quenched in peroxidase solution (0.3% H ₂ O _2/ TBS) and sections were embedded in 1% BSA/TBS. Blocking was performed for 2 hours using 10% normal donkey serum or carbon-free blocking solution (for fucosylation detection). After blotting the slides to remove serum, primary antibodies were added at a predetermined concentration (1:400 or 1:800). The slides were incubated overnight in a humidified chamber at 4°C and washed with TBS (3 x 5 min). Biotinylated LINK and HRP-conjugated secondary antibodies were applied to the sections and incubated for 2 hours in a dark, humidified chamber at room temperature, followed by washing. Replica slides were also stained with H&E (Vector Labs) according to the manufacturer's protocol. A VECTASHIELD hard set mounting medium (Vector Labs) was used to mount the slides. Positive staining density was measured using a Leica CCD camera coupled to a Leica DMi1 microscope. Fucosylation was detected using biotinylated AAL (20 μg/ml). Primary antibodies used were PON1 (18155-1-AP, Protein tech) and PON3 (OTI1A5, Thermo Scientific).

23. Induction of xenograft and lung metastasis

Experiments were performed at the College of Veterinary Medicine, Seoul National University, according to guidelines approved by the Institutional Animal Care and Use Committee. Male mice, 5 to 7 weeks old, C57BL/6 background, were purchased from Orient Bio Inc. and provided with normal diet (PMI LabDiet) and water.

For xenograft, LLC-CC and LLC-PON1 cell suspensions (1.2x10 ⁷ cells) in 200 μL of culture medium/growth factor reduced Matrigel (BD Biosciences) at a 1:1 ratio were added to the right flank of each mouse. was injected subcutaneously.

For metastasis, a suspension of LLC-CC and LLC-PON1 cells (2.5 x 10 ⁶ cells) in 150 μL culture medium was injected intravenously through the tail vein. Mice were sacrificed at the indicated times after injection. Lungs, livers, spleens and other sites where primary or metastatic tumors developed were stored in ice-cold PBS for further testing. All mice were maintained under continuous sedation by administering 2 to 4% isoflurane (Sigma) through an anesthesia mask during surgery and before euthanasia. Tumor volume was determined using digital caliper measurements and calculated using the formula: Tumor volume = (D × d2)/2, where D and d represent the long and short tumor diameters, respectively. The body weight of each mouse was also monitored.

To sample the serum, the mice were isolated from the cage for several minutes under an infrared lamp to increase blood flow. The tail was cleaned and blood was drawn. Blood samples were transferred to tubes, incubated at 4°C for a minimum of 4 hours, and centrifuged at 4°C at 10,000 g for 10 minutes. After serum was collected, it was centrifuged again and snap frozen in liquid nitrogen until further use.

24. Transwell invasion and gap-closure assay

A chemotactic invasion assay was performed in Boyden chamber wells (24 wells, 8 μm pore size, Corning) with matrigel/ECM based membranes. Matrigel matrix (Corning) was diluted to 1 mg/mL with serum-free Opti-MEM and applied to the insert in the upper chamber. For chemotaxis induction of cells, 800 μL culture medium supplemented with 8% FBS or indicated mouse serum was added to the lower chamber. After incubation for 24 or 48 hours under standard culture conditions, the membrane insert was removed and the non-enzymatically invading cells were removed from the upper surface of the membrane and non-enzymatically invading cells using 1x cell dissociation solution (Corning) following the manufacturer's protocol. was removed. Invading cells were stained with Mayer's modified hematoxylin (Abcam) for 20 minutes and washed with water.

For the gap closing assay, cells were seeded and grown until confluent. Gaps were made using P10 tips. Cells were washed and secreted protein was added. Images were acquired over time using a digital inverted light microscope to monitor gap closure.

25. Label-free proteomics

The precipitated AAL-rich secretory protein (45 μg) was separated on a 1 mm thick 10% SDS-PAGE gel and stained with CBB G-250 staining solution (Bio-Rad) for 1.5 hours at room temperature. Sections between 30 and 70 kDa were cut into 2 x 2 mm cubes and transferred to Protein Lo-Bind tubes (Eppendorf). The cut gel was divided into tubes and destained several times in 75 mM ammonium bicarbonate (Sigma) and 40% EtOH (1:1) on a shaking rack. The destained gel pieces were washed with 100 mM ammonium bicarbonate and acetonitrile (1:1), vortexed, and incubated at room temperature for 15 minutes. The gel piece was diluted with 100 mM ammonium bicarbonate and reduced with 10 mM dithiothreitol at 51 °C for 1 hour. The gel pieces were cooled to room temperature for 30 minutes and then alkylated with 20 mM iodoacetamide a at room temperature for 45 minutes in the dark. Gel pieces were dehydrated in 100% acetonitrile and dried in a SpeedVac. In-gel proteins were digested with trypsin at a protein:enzyme ratio of 20:1 at 37°C for 12 hours in a shaking incubator. Peptides were extracted in 100 μL extraction buffer (5% acetic acid/acetonitrile; 1:2) and incubated at 37°C for 15 min in a shaking incubator. The tryptic peptide mixture was eluted from the gel with 0.1% acetic acid.

For mass spectrometry, nanospray liquid chromatography tandem mass spectrometry (LC-MS / MS) was used. LC-MS/MS was performed with an LTQ-Orbitrap mass spectrometer (Thermo Electron) coupled to an Agilent 1200 series G1312B binary pump SL and a NanoLC AS-2 autosampler (Eksigent Technologies). Peptide mixture (sample 2 μL per) was loaded.

Peptides were separated by reverse phase liquid chromatography using a mobile phase. Serial mass spectra were processed using Sorcerer 3.4 beta2 (Sorcerer web interface). All MS/MS samples were analyzed using SEQUEST Cluster (Thermo Scientific) and Mascot generic format (MGF) files were set to query the human IPI v3.68 database. The search was performed with and without oxidation of methionine and carbamido methyl modifications of cysteine as variable modifications. In order to reduce the noise effect, the false positive and false positive rates were corrected through the decoy option during data retrieval in Sorcerer. Scaffold v4.0.5 (Proteome software) was used to verify MS/MS-based peptide and protein identification. PeptideProphet was used to verify peptide and protein assignments to MS/MS spectra (>95% probability). Subtractive proteomic analysis was performed on each data set, normalized to the total ion current (TIC, normalized to the average of all TIC values in the spectrum assigned to the protein).

26.RNA-seq

Low density H1993-GR was grown for 48 hours under the indicated conditions and total RNA was extracted using the RNAeasy mini kit. A 2x101 paired-end RNA-seq library was constructed using the TruSeq stranded total RNA H/M/R prep kit and sequenced using the Novaseq6000 system (Illumina). Raw paired-end sequencing reads were mapped to the human genome (build hg38). HISAT2 v2.1.0 used default parameters except for "--dta" and "--dta-cufflinks" options. Stringtie v.2.0.6 quantified the expression of genes and transcripts using the transcript information of GENCODE v27. The Ballgown package was used to perform differential gene expression analysis generating FPKMs for each gene. A differentially expressed gene was defined as a false discovery rate (FDR) <0.05 (a gene with a fold change of 2-fold or more or less than 0.7-fold) and an average read count of 10 or more.

Gene ontology analysis (GO) was performed using DAVID 6.8. GOs of biological processes or molecular functions were explored and summarized. GOs with P<0.01 were selected as significant. GO semantic-similarity network visualization was performed using REVIGO (www.revigo.irb.hr). Hierarchical clustering was performed using the pheatmap library in R. The FPKM values filtered by row-value were analyzed using the default option.

27. Bioinformatics

Drug response (IC50 per drug and cell line) and gene expression data (log2 transformed RNA values normalized to basal expression or RNA-seq TPM per cell line) were obtained from GDSC (v2, www.cancerRxgene.org) and CCLE (v2, www .depmap.org). All IC50s are expressed in μM. Cell lines were categorized by cancer type and marked with R. Cell lines lacking the drug or expression measurement of interest were not included in the analyses. Correlation analysis between drug response and gene expression by cancer type was performed by quantitatively matching the preprocessed values. Spearman's correlation coefficient, relative quantification and plotting were performed in Python. Only correlations with p < 0.05 are shown. Mutational datasets were obtained from CCLE (v2, www.cbioportal.org) and FUT domain information was retrieved from the Pfam database (www.pfam.xfam.org). The CTRP data set was analyzed using the CARE algorithm (www.care.dfci.harvard.edu).

The CCS gene set (n = 1,810) was obtained from UniProt. Glycosylation (N-/O-glycosylation) gene sets were obtained by performing gene set analysis using the GOs “Glycosylation”, “Protein N-linked glycosylation” and “Protein O-linked glycosylation” in MSigDB. Total and overlapping glycosylation genes (n = 264 or 19 for total glycosylation, n = 81 or 1 for N-linked, and n = 193 or 18 for O-linked) along with the CCS gene set were included in the analysis . For CCS, missing values of FC_2 were removed and 0 was considered as missing value. P-values were calculated with a two-sided Student's t-test and adjusted to control for FDR using the Benjamini-Hochberg procedure (Padj).

Methylation data analysis was performed using the preprocessed reduced representation bisulfite sequencing (RRBS) data set of CCLE v2. Drug sensitivity data were obtained from GDSC.

PON1 co-expressed genes were obtained CCLE v2 based on RNA-seq RPKM mRNA abundance data. Interaction rankings were based on Spieman's correlations and p-values. Quantitative analysis and plotting were performed in Python. The location of the protein encoded by each gene was searched in the Human Protein Atlas (www.proteinatlas.org).

N-glycosylation sites in the PON1 protein sequence were predicted using NetNGlyc (www.cbs.dtu.dk/services/NetNGlyc). Fold and charge regions within PON1 were visualized using FoldIndex (www.fold.weizmann.ac.il/fldbin/findex) and EMBOSS charge prediction tool (www.bioinformatics.nl/cgi-bin/emboss/charge). Stability and folding effects of proteins due to specific amino acid substitutions were evaluated using MutPred v2 (www.mutpred.mutdb.org) and I-Mutant v3 (www.gpcr2.biocomp.unibo.it/cgi/predictors/I-Mutant3.0/I -Mutant3.0.cgi) was used to predict.

Data on cancer dependence were obtained from the DepMap portal (www.depmap.org/portal) RNAi screen dataset (CRISPR Avana Public 20Q2). Dependence scores for all cancer types were grouped by phylogenetic type as predefined by DepMap and subsequently used for correlation analysis with drug response to indicated targeted therapies obtained from GDSC v2 (AUC values). Spieman's correlation coefficients and linear regression-derived p-values were obtained from pre-calculated associations in the DepMap portal. Lines with fewer than 4 cell lines for specific genetic investigations were removed from the data set. Raw essentiality scores were derived from the Profiling Relative Inhibition Simultaneously in Mixtures (PRISM) Drug Screening and Project Achilles Genetic Dependence Screening from the Broad Institute.

Data were obtained from KM plotter (www.kmplot.com/analysis) for lung cancer or pan-cancer for relapse-free survival (RFS) analysis.

Data from the breast cancer METABRIC cohort (cBioPortal) were used for co-occurrence gene analysis. Co-occurring genes in patients with fucosylation gene copy number amplification, deletion, mRNA upregulation or mRNA downregulation were stratified.

<Experimental Example 1> Confirmation of core fucosylation of cancer secretome induced by treatment

To investigate whether fucosylation correlates with drug sensitivity, gene expressions of fucosyltransferases (FUT) and protein O-fucosyltransferases (POFUT) were examined.

Specifically, the pharmacogenomic databases Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE) were used. After clustering cell line-derived data into 30 cancer types, correlations between gene expression and summary drug response measures were determined (based on IC50).

As a result, as shown in FIG. 1A, Spearman's correlation coefficient confirmed that there was a variable but broad correlation between fucosylation gene expression and drug resistance in the two data sets. In particular, FUT8, the only enzyme known to directly mediate core fucosylation through N-linkage, was consistently highly expressed in various cancer cells in both data sets.

To characterize fucosylation in patient serum, fucosylated samples were prepared from lung cancer patient (LC) serum and AAL blotting was performed, as shown in FIG. 1B.

As a result, as shown in FIG. 1c, compared to untreated patients, 30% in lung cancer (LC) patients who received multiple doses of osimertinib, a third-generation EGFR-tyrosine kinase inhibitor (TKI), Core fucosylarion of serum proteins between 2 and 60 kDa was clearly observed.

In order to quantitatively verify the results, N-glycan release analysis was performed.

Specifically, an N-glycan release assay was performed using the indicated glycosidase in crude patient serum prepared as in FIG. 1B. Prior to this analysis, samples were separated on SDS-PAGE and subjected to Coomassie staining. The 30 to 60 kDa in-gel proteins were excised and deglycosidized using glycosidase (total 8U PNGase F or total 10U Endo S/F). Glycan cleavage sites are indicated for each enzyme.

As a result of analyzing 30 to 60 kDa serum proteins in the gel, as shown in Figure 1d, the N-glycan containing the core fucose (α1,6 site) released by PNGase F had no treatment experience. Patients treated with osimertinib showed higher values than patients without. This difference was reduced in N-glycans released by Endo S/F, because the N-glycan cleavage sites (β1,4) of Endo S/F are different and do not contain core fucose. looks like

The above results suggest that the cancer secretory protein (TIS) induced by treatment contains many core fucosylarion proteins of less than 60 kDa.

In order to confirm the core fucosylation of cell-derived secreted proteins, a sandwich enzyme-linked lectin assay with various binding abilities to AAL-captured fucosylated proteins ;ELLA) was performed.

Specifically, as shown in FIG. 1e, after obtaining a secreted protein from cells or drug-resistant clones treated or not treated with the indicated drug for 48 hours, the fucosylated secreted protein was captured and eluted using ALL, Core fucosylation of cell-derived secretory proteins was analyzed using Ulex europaeus agglutinin I (UEA1)-AAL sandwich ELLA.

As a result, as shown in FIG. 1f, a therapy-induced secretome (TIS) derived from cancer cells treated with inhibitors of EGFR (gefitinib), BRAF (vemurafenib), and HER2 (lapatinib) signaling It was confirmed that fucosylation of was increased.

In addition, as shown in FIG. 1g, it was confirmed that the secretome derived from all drug-resistant clones had increased fucosylation compared to the control group.

In order to confirm whether core fucosylation occurred in the Golgi, immunofluorescence analysis was performed. Stained for RCAS1 (Golgi marker; green), fluorescein-conjugated AAL (core fucosylation; red) and DAPI (nuclear; white) and DR clones were observed by confocal microscopy.

As a result, as shown in FIG. 1h, it was confirmed that core fucosylation was immediately increased in the Golgi (16 hours after treatment) and that there was sustained activation in each drug-resistant clone.

The above results suggest that core fucosylation of the secretory protein proceeds in the Golgi before secretion of the secreted protein.

To confirm the core fucosylation of the secretome, fucosylation was characterized in the secretome derived from cancer cells.

Specifically, as shown in FIG. 1e, after obtaining a secreted protein from cells or drug-resistant clones treated or not treated with the indicated drug for 48 hours, the fucosylated secreted protein was captured and eluted using ALL, AAL blotting (ALL-blotting) was performed.

As a result, as shown in FIG. 1i, in the case of secreted proteomes induced by treatment derived from cancer cells, target kinase inhibition by EGFR (gefitinib), HER2 (lapatinib) or BRAFi (vemurafenib) inhibitors is 60 kDa Hereafter, fucosylation of the secreted protein was induced. Similarly, it was confirmed that fucosylation of proteins of 60 kDa or less was induced in the secretome derived from the drug-resistant clone (FIG. 1j).

The above results show the same results as the core fucosylation induced by osimertinib in the serum of lung cancer (LC) patients.

In order to examine the above results in detail, molecular weight cut-off filtration was performed, followed by sandwich enzyme-linked lectin assay (ELLA).

Specifically, the secreted protein was prepared as in Figure 1i with the secreted protein of sensitive or resistant H1993 cells treated with or without 1 μM gefitinib for 48 hours, and serum was treated with or without osimertinib. Serum of an untreated lung cancer (LC) patient was prepared as in 1F, and after filtering according to the indicated nominal molecular weight limit (NMWL), sandwich ELLA was performed.

As a result, as shown in FIG. 1K, secreted proteins of 30 kDa or more were concentrated from the serum of H1933 cells treated with the EGFR inhibitor gefitinib, gefitinib-resistant clones (GR), and patients treated with the EGFR inhibitor (ocimertinib). Although distinctly abundant core fucosylation was observed in , it was confirmed that there was little difference in secreted proteins of 100 kDa or more.

The above results suggest that targeted therapy induces core fucosylation of various cancer secretory proteins. These treatment-induced changes are probably evolutionary mechanisms for establishing resistance.

<Experimental Example 2> Confirmation of treatment resistance through drug-induced fucosylation of secreted protein

Based on the results of treatment-induced secretome core fucosylation in many cancers, we hypothesized that treatment-induced secretome deglycosylation inhibits the growth of surviving drug-resistant tumor cells.

To model regressing tumors in vitro, a multicolor isomorphic 'one-pot' was created by mixing 1% red tracer-labeled drug-resistant clones with 99% green tracer-labeled drug-sensitive cells in both 2D and 3D cultures. A mixture assay (multicolor homotypic 'one-pot' admixture assay) was performed. Then, the mixture was subjected to targeted treatment, and PNGase F was added to deglycosylate proteins of the secretome, and then rebound of drug-resistant clones and degeneration of sensitive cells were tracked (FIG. 2a).

First, the indicated 3D tumor spheroid mixtures were prepared as in FIG. 2a and observed under confocal microscopy with or without treatment with 5 μM gefitinib for 24 hours.

As a result, as shown in FIG. 2B, after the formation of 3D tumor spheroids, the population of gefitinib-resistant (GR) clones gradually increased (increased after 1 day and maintained until day 5) On the other hand, the sensitive cell population was greatly reduced.

Then, PNGase F was added to deglycosylate the secretome protein, and the AAL-enriched secretome was separated by SDS gel and subjected to coomassie staining (Fig. 2c) or sandwich ELLA experiment. (FIG. 2d) was performed.

Specifically, the 3D cell mixture of H1933 prepared as in Figure 2a was treated with 2 μM gefitinib or the 3D cell mixture of PC3 with 0.1 μM erlotinib for 1 or 5 days and treated with 10 μg/mL recombinant PNGase F or did not process

As a result of Coomassie staining and sandwich ELLA experiments, as shown in FIGS. 2c and 2d, it was confirmed that significant protein deglycosylation occurred when PNGase F was added to the secretome of the 3D cell mixture.

As shown in FIG. 2a, 2d cell mixtures were prepared and AAL blotting, sandwich ELLA, and N-glycation release assays were performed.

Specifically, H1993 cells were treated with 0 or 1 µM gefitinib, and PC9 cells were treated with 0 or 0.1 µM erlotinib, followed by 10 µg/mL recombinant PNGase F for 5 days. cultured. The secretome was filtered using a 30 kDa NMWL filter.

As a result, as shown in FIG. 2e, the regression of most sensitive cells induced by the treatment and the increase of a small number of gefitinib-resistant clones and erlotinib-resistant clones are secreted proteome ( It can be confirmed that it is closely related to the increased core fucosylation of the secretome.

The 3D cell mixture prepared as in FIG. 2a was treated with 2 μM gefitinib and cultured for 24 hours or 48 hours with or without treatment with 10 μg/mL recombinant PNGase F.

Fluorescently tagged 2D cell mixtures prepared as in FIG. 2a were treated with or without 1 μM gefitinib and cultured with or without 10 μg/mL recombinant PNGase F for the indicated times. On day 5, the indicated cell mixtures were followed under the same conditions, and the cell cycle status of adherent cells and apoptosis of floating cells were confirmed.

As a result of examining confocal images of fluorescent tag-conjugated gefitinib-resistant (GR) clones, as shown in FIG. It was confirmed that the accelerated growth of the promoted drug-resistant clone was blocked. In addition, as shown in Figure 2g, the 2D mixing analysis also showed similar results that deglycosylation of secreted proteome inhibited the growth of drug-resistant clones, delayed S phase of the residual cell population and promoted apoptosis. Confirmed.

In order to examine the correlation of the fucose pathway with FUT8 or SLC35C1 genes, RNAi was used and cell tracking experiments were performed.

Specifically, the H1993 mixture was treated or not with 1 μM gefitinib, the PC9 mixture was treated with or without 0.1 μM erlotinib, and the A375 mixture was treated with or without 0.1 μM vemurafenib. Sensitive cells were treated with FUT8 or SLC35C1 RNAi for 48 hours prior to mixing, and a similar tracking experiment as in Figure 2g was performed.

As a result, as shown in FIG. 2h, FUT8 RNAi or SLC35C1 RNAi treatment effectively blocked the expansion of drug-resistant clone populations.

The above results suggest that protein deglycosylation, a similar effect provided by PNGase F, is by FUT8 or SLC35C1.

In order to further investigate the characteristics of fucosylation, sandwich ELLA and Phospho-RTK arrays were performed using the same cell mixture as in FIG. 2g.

As a result, as shown in Figures 2i and 2j, it was confirmed that fucosylation was reduced by PNGase F treatment in both apoptotic bodies and soluble protein secretion bodies, and phosphorylation of proteins by kinases depletion was observed.

As shown in Figure 2k, conditioned media (CM) assays were performed.

First, as shown in FIG. 2k , colony formation analysis and phosphorylation analysis of the prepared drug-resistant clones were performed.

As a result, as shown in FIGS. 2L and 2M, drug-induced deglycosylation of secretory protein (TIS) prevents drug-resistant clones from forming colonies, and gefitinib-resistant (GR) and erlotinib-resistant The phosphorylation activity of EGFR, MET and ErbB3 kinase, respectively, was reduced in (ER) clones.

qPCR analysis was performed to investigate gene expression in the 3D cell mixture prepared as in FIG. 2a. Specifically, the cells were cultured for 5 days with or without treatment with 2 μM gefitinib and with or without 10 μg/mL recombinant PNGase F. Values are relative to DMSO and normalized to GAPDH levels.

As a result, as shown in Figure 1n, treatment-induced deglycosylation of secretory protein (TIS) triggered the senescence-associated secretory phenotype (SASP), and in the regressing tumor microenvironment (TME). Expression of genes that promote resistant cell growth was disrupted.

These results suggest that drug-resistant clonal populations are affected by treatment-induced secretory protein (TIS) fucosylation in a tumor regression model.

<Experimental Example 3> Confirmation of fucosylation of PON1 in cancer secretory protein induced by treatment

To identify components related to the treatment-induced secretory protein (TIS) and the drug-resistant clone-specific N-glycome, label-free secretome analysis (FIG. 3a) was performed.

After selecting overlapping fucosylated secretory proteins between gefitinib-treated H1993 cells and gefitinib-resistant (GR) clones, two different quantitative approaches [label-free quantification (LFQ) and intensity-based absolute quantification (iBAQ)] were used. 11 top proteins (PRSS3, PON1, HSPA7, ACTG1, CYR61, LCAT, CLU, PDIA3, P4HB, STAM2 and AFP) of 30 or more and less than 70 kDa were identified (Fig. 3b).

Among the identified fucosylated proteins, focusing on PON1, an antioxidant enzyme, immunoblot and AAL blotting analyzes were performed to confirm fucosylation of PON1.

Specifically, 1 μM gefitinib-treated H1993 cell-secreted protein was treated with 8U PNGase F, and immunoprecipitates of PON1 were analyzed.

As a result, as shown in FIG. 3c, significant fucosylation of PON1 was confirmed, and it was found to have two isoforms with a molecular mass of about 55 kDa or about 45 kDa. In cells, PON1 has both nuclear and cytoplasmic isoforms, where the approximately 40 kDa cytoplasmic isoform is known to be selectively enriched in tissues and cell lines of lung cancer (LC) patients.

To quantitatively verify PON1 fucosylation in drug-stressed cancer cells and drug-resistant clones, PON1 fucosylation-specific hybrid lectin ELISA (HLE; Fig. 3d) was performed.

As a result, as shown in FIG. 3e , despite treatment with different cell lineages, different tumor-inducing factors, and different drugs, it was confirmed that a wide range of fucosylated PON1 was increased in several cancer cells during targeted therapy.

To confirm fucosylation of PON1 in the serum of lung cancer (LC) patients, immunoblot was performed using patient serum samples treated with or without 8U PNGase F, or serum samples from patients treated with or without osimertinib. was captured with AAL, followed by PON1 immunoprecipitation and AAL blotting was performed.

As a result, as shown in FIGS. 3f and 3g , it was confirmed that the increased fucosylation of PON1 in patient serum was deglycosylated by PNGase F and significantly increased in osimertinib-treated lung cancer (LC) patient serum.

In order to examine whether fucosylation of PON1 can distinguish between untreated and osimertinib-treated lung cancer (LC) patient sera, HLE analysis of PON1 fucosylation and paraoxonase activity from serum of lung cancer patients Analysis was performed and expressed using a receiver operating characteristic (ROC) curve.

As a result, as shown in Fig. 3h, fucosylation of PON1 was distinguished with high sensitivity and specificity with an area under the curve (AUC) of 0.901 based on HLE analysis. In addition, the paraoxonase activity in the serum of patients with lung cancer (LC) not treated with osimertinib and patients treated with osimertinib was significantly different.

In order to confirm the correlation between cancer recurrence and fucosylation of PON1 in lung cancer patients, HLE analysis was performed.

As a result, as shown in FIG. 3i , it was confirmed that PON1 fucosylation increased in tissues of lung cancer (LC) patients whose cancer relapsed.

The above results suggest that PON1 fucosylation is associated with relapse and is distinct from non-recurrent cancer.

Next, intracellular PON1 fucosylation was characterized in drug-resistant clones.

The H1993 clone treated with gefitinib and the A357 clone treated with vemurafenib were stained with RCAS1 (Golgi marker; green), PON1 (red), and DAPI (nuclear; white) and observed under a confocal microscope.

As a result, as shown in Figure 3j, it was confirmed that PON1 is mainly located in the Golgi.

In addition, GR-treated H1993 cells were fractionated, PON1 was immunoprecipitated, and AAL blotting or glycoprotein staining was performed.

As a result, as shown in 3k, active fucosylation was confirmed in the Golgi/ER fraction of the gefitinib-resistant (GR) clone.

Lysates of two drug-resistant clones (H1993-GR and A375-VR) treated with SLC35C1 RNAi for 48 hours were fractionated and subjected to HLE analysis. In addition, a drug-resistant clone (H1993-GR) treated with SLC35C1 RNAi for 48 hours was stained with SLC35C1 (white) and DAPI (blue), and observed under a confocal microscope.

As a result, as shown in Figures 3l and 3m, SLC35C1 RNAi significantly reduced fucosylation of PON1 in the Golgi and increased fucosylation in the nucleus of the two drug-resistant clones.

The above results suggest that there is a functional defect in the transport of GDP-Fuc along the secretory pathway.

To confirm that PON3 is a direct modulator of PON1 fucosylation, two drug-resistant clones (H1993-GR and A375-VR) were stained with PON1 (red), PON3 (green) and DAPI (nucleus; white) and analyzed by confocal microscopy. Observed.

As a result, as shown in FIG. 3n, it was confirmed that there was active co-localization between PON1 and PON3 in both GR and VR clones.

In order to confirm the relationship between the expression of PON1 or PON3 and cancer recurrence, qPCR experiments and immunohistochemical analysis were performed using tissues of breast cancer patients who received multiple drug treatments.

As a result, as shown in FIG. 3O , it was confirmed that the expression of PON3 was increased in breast cancer patients whose cancer recurred.

The above results indicate that PON1 expression does not discriminate between relapsed and non-relapsed breast cancer patient tissues, but PON3 overexpression is associated with recurrence.

In addition, HLE analysis was performed using the Golgi/ER fraction of the lysate of the drug-resistant clone (H1993-GR) treated with SLC35C1, PON1 or PON3 RNAi for 48 hours.

As a result, as shown in FIG. 3p, it was confirmed that both PON3 and SLC35C1 RNAi, but not PON1 RNAi, inhibit PON1 fucosylation in the Golgi/ER of H1993-GR. This is a result demonstrating that PON3 fucosylate PON1 before secretion. Moreover, only PON1 RNAi, but not PON3 or SLC35C1 RNAi, was found to interfere with Golgi/ER paraoxonase activity, a result reflecting known differences between the two PON proteins in paraoxon hydrolysis.

The P-Fuc activity of FUT8 was assayed in cross-linked FUT8 and PON1 co-immunoprecipitates from Golgi/ER fractionated H1993-GR lysates.

As a result, as shown in Figure 3q, it was confirmed that PON3 or SLC35C1 RNAi can remove the transfer of the fucose moiety of FUT8 (movement of PON1 from GDP-Fuc to the N-glycan GlcNAc residue).

These results suggest that core fucosylated PON1 is a major component of the constitutive N-glycome of secretory proteins (TIS) induced by cancer treatment and is a hallmark of target therapy resistance.

<Experimental Example 4> Confirmation of the effect of core fucosylation of PON1 on PON1 stability and secretion To structurally show N-glycans bound to PON1, analyze tandem mass spectrometry (MS/MS) data and FUT8 substrate specificity was analyzed through CID data.

As a result, as shown in FIG. 4a, 6 abnormally fucosylated glycans released from the immunoprecipitated PON1 were determined, which generally showed a GlcNAc ₂ Man ₃ + HexNAc ₂ Fuc ₁ glycan structure. In addition, peaks 1 and 2, the most abundant glycans, were found to have high FUT8 substrate specificity, whereas peaks 3 to 6 had low specificity or were not identified yet.

In order to determine the site of glycosylated PON1, single-point mutations were introduced into two PON1 sequences [N253 (WT) and N324 (GT)], and full-length AAL blot analysis of PON1 immunoprecipitates from drug-resistant clones (H1993-GR) transfected with (full-length) PON1, PON1-N253 or PON1-N324G constructs was performed, or AAL-enriched PON1 immunoprecipitates (AAL- HLE analysis of secretory protein PON1 fucosylation and N-glycan release analysis were performed in enriched PON1 immunoprecipitates).

As a result, as shown in Figure 4b, it was confirmed that PON1 core fucosylation was reduced in both the PON1 mutants of N253G (②) and N324G (③).

In addition, after obtaining cross-linked FUT8 and PON1 co-immunoprecipitates of the transfected drug-resistant clone (H1993-GR) as described above, GDP-Fuc activity of FUT8 was analyzed.

As a result, as shown in Figure 4c, it was shown that efficient GDP-Fuc delivery was prevented in both PON1 mutants of 253G (②) and N324G (③), and N253G (②) was more potent than N324G (③). showed effect.

To verify the effect of the N253G and N324G mutations on PON1 stability, PON1 folding and synthesis were analyzed for protein cleavage by trypsin or de novo protein synthesis inhibition by cycloheximide (CHX) treatment.

As a result of immunoblotting of the lysate of the gefitinib-resistant clone (H1993-GR), as shown in Fig. 4d, in the case of the N253G (②) mutant, PON1 was degraded by trypsin, resulting in PON1 folding errors (misfolding). ), and it was confirmed that N324G induces PON1 degradation only at higher trypsin concentrations.

In addition, similar results were confirmed in the results of experiments using ELISA in the Golgi / ER of the gefitinib-resistant clone (H1993-GR) (FIG. 4e).

Drug-resistant clones (H1993-GR) transfected with full-length PON1, PON1-N253 or PON1-N324G constructs for 36 hours were treated with or without 25 μg/mL CHX for the indicated times, followed by EZClick labeling analysis. or immunoblot was performed.

As a result of confirming protein synthesis by performing EZClick labeling analysis, as shown in FIG. 4f, it was confirmed that N253G did not significantly affect total protein synthesis in the gefitinib-resistant clone.

In addition, as a result of examining the expression of PON1 through immunoblot, as shown in FIG. 4g, N253G in the gefinitib-resistant clone accelerates the degradation of the small-sized isoform of PON1 in the fucosylated form. appear.

An ELISA assay was performed to confirm the expression of the secretory protein PON1 in drug-resistant clones (H1993-GR) transfected with full-length PON1, PON1-N253 or PON1-N324G constructs for 36 hours.

As a result, as shown in FIG. 4h, it was confirmed that N253G significantly eliminated PON1 secretion, but N324G showed a moderate effect.

The above results suggest that core fucosylation promotes PON1 stability before secretion in drug-resistant clones and PON1 overexpressing cells.

<Experimental Example 5> Confirmation of the effect of core fucosylation of secreted protein on treatment resistance through UPR effector and pro-inflammatory

In order to confirm changes in intracellular signal transduction pathways, a Cignal 45-pathway reporter array was performed.

Specifically, the cell mixture was treated with 1 μM gefitinib, and 10 μg/mL recombinant PNGase F was not treated (control) or treated for 5 days to confirm reporter transcriptional activity.

As a result, as shown in Fig. 5a, ER stress (ATF6), amino acid response (AAR elements; ATF2, ATF3, ATF4), androgen receptor (AR) pathway, and E2F transcription were markedly upregulated, and stem cell factors (SOX2, NANOG, OCT4), interferon-stimulated responses (ISR elements; STAT1, STAT2), STAT3, and hypoxia (HIF) signaling were selectively inhibited.

These results indicate that treatment-induced deglycosylation of secreted protein (TIS) is a mediator that inhibits the growth of drug-resistant clones, resulting in translational changes by ER stress and UPR.

To verify this process, experiments were conducted focusing on ATF6, a UPR-specific stress sensor located in the ER.

Secretory proteomes of H1993 cells treated with a total of 8U PNGase F or transfected with full-length PON1 or PON1-N253G constructs for 36 h and treated with 1 μM gefitinib and H1993-GR were co-cultured for 5 days, followed by confocal microscopy. Observed. Gefitinib resistant (GR) clones were stained for fluorescein-conjugated AAL (core fucosylation; green), ATF6 (red) and DAPI (nuclear; blue). In addition, after co-cultivation under the same conditions as above, Golgi/ER fractionation and ELISA analysis of ATF6 expression were performed.

As a result, as shown in Fig. 5b, treatment-induced secretome deglycosylation or fucosylated PON1 inhibition both inhibited ATF6 located in the Golgi/ER and fucosylated scaffold residues in drug-resistant clones. It was confirmed that it was actively induced to the same location. It seems that ATF6 is translocated/sorted from the ER to the Golgi apparatus by ER stress.

Next, in order to determine whether the stress-induced response due to inhibition of secretory protein (TIS)-specific PON1 fucosylation induced by treatment with the PON1-N253G mutation is coordinated with oxidative stress, 1 μM Zepty was added to the cell mixture. The nip was treated, and 10 μg/mL recombinant PNGase F was not treated (control) or treated for 5 days to confirm generation of active oxygen and nitrogen species (ROS/RNS). In addition, the production of reactive oxygen species and nitrogen species was confirmed using a cell mixture transfected with the PON1-N253G construct and treated with 1 μM gefitinib.

As a result, as shown in FIG. 5c, inhibition of treatment-induced deglycosylation of TIS and fucosylated PON1 in the degenerating cell mixture actually inhibits reactive oxygen species (ROS) and nitrogen species (ROS), which are characteristic of redox imbalance. / RNS) was confirmed to significantly stimulate the production.

In addition, generation of reactive oxygen species and nitrogen species was confirmed in H1993-GR against ATF6 RNAi cultured in the secretory proteome of PON1-N253G transfected H1993 cells treated with or without 1 μM gefitinib for 72 hours.

As a result, as shown in Fig. 5d, blocking the protein synthesis of ATF6 in drug-resistant clones co-cultured in TIS induced by modified treatment with little PON1 fucosylation resulted in intracellular reactive oxygen species and nitrogen species. It was confirmed that the production of was inhibited.

In 3D cell mixtures treated with gefitinib, sensitive cells harboring PON1-N253G inhibited growth of gefitinib-resistant clones, whereas ATF6 RNAi treatment did not (Figures 5e and 5f). .

In order to confirm the effect on the intracellular inflammatory response, the secretory protein of H1993 cells transfected with 1 μM gefitinib transfected with full-length PON1 or PON1-N253G construct for 36 hours and H1993-GR were co-cultured for 5 days, Sandwich ELISA was performed to analyze IL-6, TNF-α and GM-CSF cytokines.

As a result, as shown in FIG. 5g, PON1-N253G promotes a pro-inflammatory environment by increasing IL-6, TNF-α, and GM-CSF cytokines in the secretory protein, while ATF6 This phenomenon was not observed when RNAi was treated.

Through the above results, it was confirmed that inhibition of PON1 deglycosylation or PON1 fucosylation increases the production of reactive oxygen species and nitrogen species and induces pro-inflammatory secreted proteins.

<Experimental Example 6> Identification of resistance-related genes mediated by core fucosylation of secreted protein PON1

To identify the molecular causes of the response of drug-resistant clones to inhibition of PON1 fucosylation, data of differentially expressed genes under PNGase F treatment and PON1-N253G transfection conditions were integrated.

As a result, as shown in Figure 6a, deglycosylation of the secretory protein led to 135 down-regulated and 65 up-regulated genes, and inhibition of fucosylation of the secretory protein PON1 resulted in 150 down-regulated and 83 up-regulated genes. were generated, and all showed statistically significant p-values. Commonly in both conditions, 21 genes were upregulated and 90 genes were downregulated (Fig. 6b). Specifically, we predicted C19orf25, RPS27L, CLDN2, PAQR3, and SOX4 as negative regulators of drug-resistant clone growth by PON1 fucosylation, and THBS1, F3, TAGLN, ANKRD1, and DKK1 as positive regulators (Fig. 6c).

Patient survival data from a cohort of lung cancer patients were stratified by indicated gene expression (low or high) in primary tumors based on microarray (first progression survival; FPS) or RNA-seq (relapse-free survival; RFS) data.

As a result, as shown in Fig. 6d, high expression of upregulated RPS27L in a large lung cancer patient population correlates with increased first progression survival (FPS) or relapse-free survival (RFS). On the other hand, it was confirmed that high expression of downregulated DKK1 was associated with low survival.

The above results show that the gene controlling the response of drug-resistant clones to inhibition of PON1 fucosylation is functionally related to drug sensitivity to targeted therapy, and the gene is a potential therapeutic target that restricts the growth of drug-resistant clones. suggests that

Claims (12)

1) separating the fucosylated PON1 (Paraoxonase 1) protein from the sample; and
2) measuring the degree of fucosylation of the PON1 protein
Characterized in that it comprises a, anti-cancer drug resistance prediction method.
The method of claim 1, wherein the sample in step 1) is any one selected from the group consisting of cells, tissues, blood, serum, plasma, saliva and urine of patients who have received drug treatment. .
The method of claim 1, wherein the fucosylation of step 1) is core fucosylation.
delete The method of claim 1, wherein the degree of fucosylation of the PON1 protein in step 2) is increased in patients with anticancer drug resistance compared to normal subjects.
The method of claim 5, wherein the patient is selected from the group consisting of patients with liver cancer, stomach cancer, colon cancer, lung cancer, uterine cancer, breast cancer, prostate cancer, thyroid cancer and pancreatic cancer.
A composition for predicting anticancer drug resistance, comprising an antibody that specifically binds to fucosylated PON1 protein.
delete A kit for predicting anticancer drug resistance, comprising an antibody that specifically binds to fucosylated PON1 protein.
delete A microarray for prediction of anticancer drug resistance, characterized in that an antibody specific to fucosylated PON1 protein is attached. delete

Images