1. Introduction
Lung cancer is a fast-progressing tumor with low prevalence of oncogene-addicted subsets, and with each subset requiring unique therapies. Recent therapeutic improvements in advanced non-small cell lung cancer (NSCLC) include, but are not limited to, targeted therapies and checkpoint immunotherapy that have changed clinical practice. However, the 5-year survival of NSCLC is still ~16% and there is an unmet need for highly effective treatments to better outcomes for this cancer type [ 1 ]. In advanced NSCLC, the current standards of care are chemotherapy, immunotherapy, and targeted therapy in specific oncogene-addicted tumors. Available targeted therapies are mostly directed towards oncogene aberrations: EGFR, ALK, ROS1, BRAF, RET, MET [ 2 ], NTRK [ 3 ], and KRAS [ 4 ]. Targeted inhibitors have achieved good anti-tumor effects in patients with target-specific genomic alterations.
Currently, several antibody-drug conjugates (ADCs) are under clinical investigation with promising activity in advanced cancers [ 5 ]. Approved ADCs include trastuzumab emtansine that is a conjugation between a cytotoxic payload and an anti-HER2 antibody, targeting HER2 overexpressing cells. Other ‘smart’ medications that direct cytotoxic therapy to cancer cells by using a known biomarker include deruxtecan, trop-2 ADC in triple negative breast cancer [ 6 ], and nectin-4 ADC in urothelial cancer [ 7 ]. Surgery with or without adjuvant chemotherapy is regarded as the standard-of-care treatment for early-stage NSCLC, although only 25–30% of tumors are suitable for potentially curative resection. Additional treatment options for NSCLC are urgently needed to improve patient outcomes and quality of life. Although mutation-specific therapeutic targets offer the exciting prospect of efficacious therapies, broad-spectrum novel approaches are in demand.
Tumor cell surface glycosylation acts as key regulatory mechanisms controlling several pathophysiological processes, such as cell signaling, cell dissociation, cell-matrix interaction, angiogenesis, immune modulation, invasion, and metastasis [ 8 ]. Changes in glycosylation patterns are associated with carcinogenesis [ 9 , 10 ]. Specifically, chondroitin sulfate (CS) alterations are found in most solid tumors [ 11 ]. Modification of proteins with CS form biologically active molecules called chondroitin sulfate proteoglycans (CSPG) [ 12 ].
CS and CSPGs play crucial roles in malaria pathogenesis [ 13 ], and syndecan-1 (SDC1) is the major CSPG involved in pregnancy-associated Plasmodium falciparum malaria [ 14 ]. The parasitized erythrocytes can sequester in the placenta via specific adhesion to a CS signature expressed exclusively in the placental syncytium and intervillous space, mediated by the malaria-encoded protein VAR2CSA [ 15 ]. Many solid tumors express placental-type CS as a secondary oncofetal CS modification on proteoglycans, and the recombinant VAR2CSA malaria protein (rVAR2) can, therefore, also bind to malignant cells [ 16 , 17 ]. Moreover, oncofetal CS is important for cancer cell migration and can regulate metastasis via integrin signaling pathways [ 18 ]. Previous studies have shown that oncofetal CS is highly expressed in solid tumors such as melanoma and muscle-invasive bladder cancer, and is associated with advanced tumor stage and poor patient outcome [ 16 , 19 ]. However, its expression in lung cancer, especially in early-stage tumorigenesis, is poorly understood. This study aimed to analyze the expression of oncofetal CS in NSCLC and determine the susceptibility of NSCLC cells to oncofetal CS-targeting therapeutics.
2. Material and Methods
2.1. Lung Cancer Cohorts
In this retrospective study, surgically resected and histologically confirmed NSCLC tumor tissues were obtained, with the approval of the local institutional review board, from patients diagnosed with NSCLC with early-stage disease: 318 cases of stage I and 187 cases of stage II (according to the contemporary guidelines and AJCC 7th edition) [ 20 ]. No data were available for 34 cases. Patients were diagnosed between the years 2001 and 2013 and received follow-up care according to the contemporary NCCN guidelines at Scottsdale Healthcare, Scottsdale, AZ (SHC); ProHealth Care, Waukesha, WI (PHC); National Koranyi Institute of Pulmonology, Budapest; or University of Tennessee, Knoxville, TN (UTenn). SHC and UTenn cohort data without detailed prognostic or OS data were published before [ 16 , 21 ]. Clinicopathological data included gender, age, stage, molecular characteristics (EGFR and KRAS where available), smoking, disease-fee survival (DFS), and overall survival (OS). We assessed the pathological tumor (pT) and lymph node (LN) status according to the Union for International Cancer Control (7th edition), and recorded age at the time of diagnosis. Patients who underwent lung resection surgery including lobectomy or pulmonectomy were included. OS was estimated from the time of diagnosis, until death, or last available follow-up. The study and all treatments were conducted based on the institutional guidelines.
2.2. Tissue Micro Arrays (TMA)
NSCLC resected tumors were formalin fixed and paraffin embedded (FFPE) after surgery. TMA construction from FFPE blocks was performed. Briefly, from each tissue block, 4-micron sections were cut using a microtome and placed on slides. Slides were stained for H&E. Board-certified pathologists marked the tumor area on H&E slides. Two 1-mm punches of tissue were taken from each primary tumor tissue block for TMA construction. TMAs from four separate patient cohorts, with two cores from each patient, representing a total of 539 clinically annotated stage I–II NSCLC cases were examined. The TMAs were built with Veridiam Semi-Automated Tissue Arrayer (VTA-100, Veridiam, San Diego, CA, USA).
2.3. Cell Lines
HCC827, H358, H1975, H2030, H460, H520, A427, H1993, H2073, H1792, H2122, H23, H1703, and A549 NSCLC cell lines (ATCC) were grown in appropriate cell culture medium.
2.4. Protein Constructs
All recombinant rVAR2 proteins were expressed in E. coli Shuffle cells (NEB) using a pET28 vector. The constructs of rVAR2 (ID1-ID2a) contained either a SpyTag [ 22 , 23 ] or an albumin-binding domain (ABD) [ 24 ] for extension of plasma half-life fused in the N-terminal, and in the C-terminal a V5-tag and 6×His-tag were added for detection and affinity purification. The control proteins, DBL4 and DBL5, both contained a V5-tag and His-tag in the C-terminal similar to rVAR2 and for DBL5 an ABD was fused in the N-terminal.
Expression and purification of the rVAR2 protein were performed, as described in a previous publication [ 16 ]. Briefly, the E. coli pellet harboring the recombinant protein was resuspended in lysis buffer with benzonase and homogenized. After centrifugation and sterile-filtration, the supernatant was loaded onto a HisTrap HP (Cytiva, Uppsala, Sweden) column and eluted using imidazole. The protein was further purified by HiTrap SP HP (Cytiva, Uppsala, Sweden), before elution an endotoxin reduction wash step was included where 0.1% Triton X114 in binding buffer was passed over the column. After removing the wash buffer, protein was eluted with a linear gradient from 0 to 1 M NaCl in 25 mM phosphate buffer pH 7.2. The eluted rVAR2 protein was pooled and analyzed by SDS-PAGE, aliquoted, flash-frozen, and stored at −80 °C until use.
2.5. Immunohistochemistry (IHC)
Freshly cut TMA sections were analyzed for oncofetal CS expression and IHC was performed using the Ventana Discovery platform, as described previously [ 16 ]. Interpretation of all immunostainings was blinded from clinicopathological parameters and patient outcome. The intensity and percent positivity were examined for both pericellular and extracellular immunostaining. While an overall score was calculated for each tissue core, an average score of 100 to 300 was considered as oncofetal CS high. The data were analyzed for correlation between low and high oncofetal CS expression of both tumor and stroma cells, with respect to EGFR and KRAS mutations, as well as clinical characteristics including DFS and OS.
2.6. Flow Cytometry
Cells were grown to 70–80% confluency in appropriate growth media and harvested in an EDTA detachment solution (Cellstripper ® , Mediatech, Inc., Manassas, VA, USA). Cells were incubated with rVAR2-V5-tagged protein (200 nM–25 nM) in PBS containing 2% FBS for 30 min at 4 °C and binding was analyzed in a FACSCalibur (BD FACSCanto II, Becton, Dickinson and Company, Franklin Lakes, NJ, USA) after secondary incubation with anti-V5-FITC antibody for 30 min at 4 °C (1:500, Invitrogen, Carlsbad, CA, USA; Cat#R963-25). For inhibition studies, the protein was co-incubated with 400 μg/mL CSA (Sigma, Saint Louis, MO, USA; Cat#27042). A recombinant V5-tagged non-CS binding domain of the VAR2CSA protein (DBL4) was used as a control, and results were analyzed using FlowJo software (version 10.1r7, Ashland, OR, USA). Mean fluorescence intensity (MFI) was plotted as a percentage of the highest MFI signal within the same cell line. Background was calculated from cells incubated with secondary anti-V5-FITC only and the signal was subtracted from each MFI.
2.7. Preparation of Drug Conjugate VDC-MMAE and Control-MMAE
We chemically conjugated a potent anti-mitotic agent, monomethyl auristatin E (MMAE), derived from peptides occurring in marine shell-less mollusk Dolabella auricularia to rVAR2 via a maleimide-activated cathepsin B valine-citrulline cleavable linker, utilizing free cysteines in the recombinant protein. Briefly, the rVAR2-drug conjugate (VDC) was obtained by mixing rVAR2 (ABD-ID1-ID2a) or control protein (ABD-DBL5) and the drug linker in a 1:2.8 molar excess and incubating for 60 min at room temperature. To concentrate and remove unreacted drug linker, the protein was then loaded on a HiTrap SP HP column equilibrated with binding buffer (25 mM sodium phosphate, pH 7.2) and eluted with 1 M NaCl in 25 mM sodium phosphate, pH 7.2. The eluted material was buffer exchanged into 20 mM Histidine, 200 mM NaCl, pH 6.0, flash-frozen, and stored at −80 °C prior to use. Purity and Drug-VAR2 Ratio (DVR) was determined by SDS-PAGE, HPLC-SEC, and UV-vis absorbance measurement. The rVAR2-MMAE drug conjugate (VDC-MMAE) had a DVR of 1.8, whereas the DBL5-MMAE (Control-MMAE) had a DVR of 2.3.
2.8. In Vitro Cytotoxicity Assay of VDC-MMAE of Human Lung Cancer Cell Lines
Cells were removed from their culture vessel using Corning™ CellStripper Dissociation Reagent (Mediatech, Inc., Manassas, VA, USA; #MT25056CI). Detached cells were diluted in respective growth medium (Thermo Fisher Scientific, Waltham, MA, USA; #11095-080) + 10% Fetal bovine serum (Corning; #35-015-CV) to 25,000 cells/mL such that 100 µL/well would dispense 2500 cells/well. Cells were seeded into black-walled, clear, flat-bottomed, 96-well plates (Corning™ Costar™, Fisher Scientific, Waltham, MA, USA; #3595). Cells were incubated for one night at 37 °C in a 5% CO 2 atmosphere to allow the cells to attach to the microtiter plate surface. VDC-MMAE was diluted directly in the appropriate cell growth medium and then titrated 1:2, over nine steps starting at 300 nM. A control with no VDC-MMAE present (growth medium alone) was included on each microtiter plate in triplicates. Then, 400 μg/mL CSA was used as a specificity control as well as a toxicity rescue assay in three of the 14 cell lines tested. Then, 50 µL/well of the prepared VDC-MMAE titrations was added in triplicate to each cell line assayed. The cells and titrations were incubated at 37 °C with 5% CO 2 for 48 hours. After the incubation, cell viability was measured by using a crystal violet proliferation assay. The collected relative light absorbance [LA] was converted to % cytotoxicity using the absorbance values measured from the growth medium alone control as follows: % Cytotoxicity = 1 − [Well LA/average medium alone control LA] (1)
Data (% Cytotoxicity vs. Concentration of VDC (log 10 [nM]) were plotted and were analyzed by non-linear regression methods using GraphPad Prism software version 8 (GraphPad Software, San Diego, CA, USA) to obtain IC50 estimates.
2.9. The rVAR2 Pulldown and Mass Spectrometry
Cells were homogenized and spun down, and membrane proteins were extracted by incubating the membrane pellet in EBC lysis buffer (20 mM NaCl, 50 mM Tris-HCl, 2.5 mM MgCl 2 , 1 mM EDTA) supplemented with a protease inhibitor cocktail (Roche Diagnostics GmbH, Mannheim, Germany). CS proteoglycan enrichment was done using an rVAR2 affinity chromatography workflow, as previously described [ 25 ]. Briefly, Spy-tagged rVAR2 protein was loaded onto a SpyCatcher modified HiTrap Sepharose column (GE Healthcare, Chicago, IL, USA), which allowed immobilization of rVAR2 through isopeptide formation between the tag-catcher system. The extracted samples were then loaded onto the column and washed extensively (0.25 M NaCl), and bound proteins were eluted using high-salt buffer (1.6 M NaCl). The samples were desalted using PD10 columns (GE Healthcare, Sigma) and reduced with DTT (5 mM), alkylated with iodoacetamide (15 mM), and digested with trypsin 1:40 w / w (Promega, Madison, WI, USA) overnight. Finally, detergent was removed from the digested samples by Pierce detergent removal spin columns (Thermo Scientific, Waltham, MA, USA) and the peptides were desalted using C18 spin columns (Thermo Scientific). Samples were dried and stored at −20 °C until LC-MS/MS analysis.
Analysis of peptides was carried out on an Orbitrap Fusion Lumos MS platform (Thermo Scientific). Samples were introduced using an Easy-nLC 1200 system with an ES801, 50 µm ID × 15 cm column (Thermo Scientific). Prior to each sample injection, the analytical column was equilibrated at 500 bar, for a total volume of 10 μL. After injection, sample loading was carried out, for a total volume of 6 μL at 500 bar. The injection volumes for all samples were 2 µL, and were run using a gradient of mobile phase A (water and 0.1% formic acid) and B (80% acetonitrile with 0.1% formic acid) at 0.3 µL/min, 2–28% B from 2–92 min followed by 28–40% B, 92–102 min; 40–95% B, 102–108 min; 95%, 108–120 min; 50 °C.
Data acquisition (control software version 3.1.2412.17, Thermo Scientific, Waltham, MA) was carried out using a data-dependent method with MS2 in the Orbitrap. The Lumos was operated with a positive ion spray voltage of 2000 and a transfer tube temperature of 325 °C. The default charge state was set as 2. Survey scans (MS1) were acquired in the Orbitrap at a resolution of 60 K, across a mass range of 375–1500 m / z , with RF lens setting of 30, an AGC target of 4 × 10 5 , and max injection time of 50 ms in profile mode. For MS2 scans, an intensity threshold of 5 × 10 4 , charge state filtering of 2–5, and dynamic exclusion for 15 seconds with 10 ppm tolerances were used with a 1.2- m / z window prior to HCD fragmentation of 33%. MS2 data acquisition, carried out in the Orbitrap, used a 15-K resolution, a fixed first mass of 110 m / z , an AGC target of 5 × 10 4 , and a max injection time of 100 ms in centroid mode with parallelizable time turned off.
2.10. Mass Spectrometry Data Analysis
All data files were processed with Protein Discoverer 2.2.0.388 (Thermo Scientific, Waltham, MA, USA). Spectrum files were recalibrated and features extracted with Minora. Searches were carried out with Sequest HT with SwissProt TaxID = 9606 (v2017-10-25, Swiss Institute of Bioinformatics, Lausanne, Switzerland) with precursor mass tolerance 10 ppm and fragment mass tolerance 0.01 Da with C carbamidomethyl as permitted and fixed, and M, P oxidation, as permitted dynamic peptide modifications and acetyl N-terminal protein modification. Decoy database strict and relaxed FDR targets were 0.01 and 0.05 based on q value. Precursor quantification was intensity based with unique and razor peptides used, normalizing on total peptide amount with scaling on all average, and protein lists exported to Excel for any further analysis. The CSPG hits were verified against control pulldown samples from cell media.
2.11. In Vivo Study
All animal work was approved by the animal ethics application (A18-0278) and biosafety application (B19-0104) Animal Care Committee of the University of British Columbia. Initially, 7-week-old female nude mice (ENVIGO, Indianapolis, IN, USA; Stock No. 069) were anesthetized with 3% isoflurane. Then, 100 μL of a cell suspension in PBS:Matrigel (Corning, Ref. No. 354234) in 1:1 ratio, containing 1.0 × 10 6 A549 cells, were injected into a total of 23 mice using a 27-G needle by subcutaneous injection. Starting on day 2, tumor volume was measured by caliper three times a week. Mice with growing tumor size between 120 mm 3 to 150 mm 3 were assigned to the following treatment groups: Vehicle (His-buffer; 20 mM His, 200 mM NaCl, pH 6.0) ( n = 9), VDC-MMAE ( n = 7) at 15 mg/kg, and Control-MMAE ( n = 7) at 8.9 mg/kg, using sequential randomization. The VDC-MMAE dose was selected based on a previous dose-escalation study in healthy female CD-1 mice, which was well tolerated by the animals, without signs of morbidity or physical distress [ 16 ]. Treatments were administered twice a week for two weeks by intravenous injection in the tail vein. Survival was determined by reaching either experimental (tumor volume between 800–1000 mm 3 ) or humane endpoints (such as weight loss >15%, tumor with a visible ulceration, which did not heal with metacam and polysporin treatments, irregular/labored respirations, ulcerated skin >1 cm patch, no response when stimulated, immobile, constantly shaking, vocalizations, and severe self-mutilation/trauma). The difference in tumor sizes between the two groups were analyzed using Student’s t -test.
2.12. Statistical Analyses
Statistical analyses were conducted using the PASW Statistics 22.0 package (SPSS Inc., Chicago, IL, USA). OS was calculated from the date of surgery to the date of death. Patients still alive were censored at the date of last follow-up. DFS was calculated from surgery until disease recurrence. To identify independent prognostic factors, Cox proportional hazards models were applied.
We used average H -score from TMA core A and B values, including both stroma and tumor compartments in the statistical analyses. Statistical analyses were performed only in patients with available clinicopathological data. Next, patients were divided into “high” (≥100) and “low” (<100) groups, based on the oncofetal CS expression averaged of both stroma and tumor tissue. According to low vs high oncofetal CS expression scores, clinicopathological characteristics of patients were analyzed by the Chi-square tests.
Kaplan–Meier curves and two-sided log-rank tests were used for univariate survival analysis. Two-sided p -values less than 0.05 were considered statistically significant. All variables with p -values less than 0.05 were included in the multivariate analysis. The Cox proportional hazards model was used for univariate and multivariate survival analyses to calculate the hazard ratios (HR) and corresponding 95% confidence intervals (CI). For multivariate survival analyses, the Cox regression model was adjusted for significant variables in the univariate analysis. Metric data are shown as median or mean and corresponding range or, in the case of OS, as median and corresponding 95% CI. We included gender, age, oncofetal CS expression, histology, and tumor stage in the Cox model.
3. Results
3.1. Baseline Clinical Characteristics and Patient Outcomes
We identified a total of 539 patients diagnosed with NSCLC who underwent surgical lung resection. The mean age at diagnosis was 65.6 ± 9.8 years. The median overall survival (OS) of the cohort was 66.7 ± 4.8 years. The clinicopathological characteristics of surgically resected NSCLC patients are outlined according to available low vs. high oncofetal CS expression in both tumor and stroma as average H -score in Table 1 . Figure S1A represents the study design flow chart, including case numbers for this retrospective analysis with inclusion and exclusion criteria.
3.2. Oncofetal CS Expression Is an Independent Prognostic Classifier in Early-Stage NSCLC
We examined four cohorts of early-stage NSCLC samples as well as normal lung tissues for basal oncofetal CS expression using rVAR2 IHC. Oncofetal CS expression was observed in lung cancer tissue but was minimal or absent in normal lung ( Figure 1 A). There were 351 patients (71%) with low and 142 patients (29%) with high oncofetal CS expression. High oncofetal CS levels were associated with shorter DFS in all cases (39 vs. 67 months, p < 0.01) ( Figure 1 B) and smokers ( p < 0.05) ( Figure 1 C). A similar prognosis stratification was also seen in OS for all cases (51 vs. 69 months, p = 0.044) ( Figure 1 D) and smokers ( p = 0.028) ( Figure 1 E). Moreover, high oncofetal CS expression was associated with shorter DFS ( p = 0.009) and OS ( p < 0.01) in male patients ( Figure S1B,C ). There was no significant DFS or OS difference based on oncofetal CS expression when stratifying by EGFR or KRAS mutation status as well as female and non-smokers. Prevalence of high oncofetal CS level was highest in EGFR wild-type/KRAS wild-type cases followed by in EGFR wild-type/KRAS mutation patient samples ( Table S1 ). High oncofetal CS expression was seen more frequently in squamous cell carcinoma than adenocarcinoma (46% vs. 19%, Table 1 ) and associated with poor DFS in squamous cell carcinoma ( p = 0.012) ( Figure S1D ). In univariate survival analyses, we observed significant shorter median OS/DFS in male as well as in stage I and II groups with high oncofetal CS expression ( Table 2 ). In multivariate survival analyses, when Cox regression model was adjusted for oncofetal CS, histology (adenocarcinoma vs. squamous cell carcinoma), and stage (I vs. II), high oncofetal CS expression was significantly associated with shorter DFS (HR, 1.76; 95% CI, 1.32–2.48; p < 0.01) ( Table 3 ). Taken together, high oncofetal CS expression is present in around 30% of early-stage NSCLC patients and is an independent prognostic factor for poor DFS in NSCLC patients.
3.3. NSCLC Cell Lines Are Sensitive to VDC-MMAE
We next analyzed a panel of six NSCLC cell lines for oncofetal CS expression by flow cytometry. Using recombinant V5-tagged control protein (DBL4) as the control, we observed a concentration-dependent binding of the rVAR2 protein to oncofetal CS in all NSCLC cell lines tested, saturating at 100 nM rVAR2 ( Figure 2 A). The binding of rVAR2 to the NSCLC cells could be competed by purified CSA, demonstrating specificity of the rVAR2-oncofetal CS interaction ( Figure 2 A). Next, we tested the sensitivity of 14 NSCLC cell lines with and without KRAS mutations to the oncofetal CS targeting rVAR2-drug conjugate VDC-MMAE. All NSCLC cell lines were sensitive to VDC-MMAE treatment with IC50 values ranging from 1 to 6.5 nM ( Figure 2 B). Importantly, VDC-MMAE cytotoxicity was reduced by CSA competition with 7–9 times higher IC50 values, indicating drug-specificity towards oncofetal CS ( Figure 2 C). These data show that both KRAS wild-type and mutant NSCLC cells are positive for oncofetal CS and sensitive to VDC-MMAE in vitro.
3.4. VDC-MMAE Inhibits Growth of NSCLC Tumors In Vivo
We next investigated the sensitivity of established in vivo NSCLC xenograft tumors to VDC-MMAE. A549 is one of the best characterized NSCLC cell lines with a homozygous KRAS G12S mutation and extensively used for xenograft studies in mice [ 26 , 27 ]. We first confirmed the expression of known CSPGs in A549 cells by an affinity purification and mass spectrometry workflow ( Figure 3 A). The analysis identified multiple proteoglycans with oncofetal CS modifications including SDC1, SDC4, CD44, and Versican ( Figure 3 B). We next established A549 xenograft tumors in mice followed by treatment with 15 mg/kg of VDC-MMAE twice a week for two weeks. Contrary to vehicle and control-MMAE treatments, VDC-MMAE effectively inhibited tumor growth ( Figure 3 C) and prolonged survival of the mice ( Figure 3 D). These results show that oncofetal CS-positive NSCLC tumors can be targeted by VDC-MMAE in vivo independent of KRAS-mutation status.
4. Discussion
ADCs are approved for the treatment of several cancer types and represent a promising approach to improve the prognosis of advanced NSCLC. ADC targets that are present both on tumor cells and stromal compartments represent a more effective therapeutic opportunity.
Informative biomarkers are important for disease management and are most commonly assessed by IHC and/or genomics-based approaches. IHC biomarker stainings are affordable and widely available, such as PD-L1 stainings that offer fast laboratory turnaround time, which is essential for patients with highly progressing cancers. Gene sequencing facilities able to perform genomic interrogations are often located in centers and/or require shipment of biospecimens, which can be time-demanding and costly. These approaches are challenged by fast tumor progression in NSCLC that associate with pleural effusion, brain or liver metastases, or cachexia that limits the therapeutic window. Therefore, patients often receive no active therapy and supportive care that are under-represented in clinical trials. A large portion of NSCLCs are not linked to any oncogenic drivers. When an oncogene-based therapy is identified, a number of these subtypes have a prevalence of less than 5%. Additionally, a recent analysis showed that less than 50% of patients with metastatic NSCLC do not have their tumors tested broadly for genomic aberrations, missing treatment opportunities with small molecules [ 28 ].
We previously described that an evolutionarily refined parasite–host anchor protein, VAR2CSA, derived from P. falciparum malaria parasites, specifically binds to an oncofetal CS modification on a subset of cancer-associated proteoglycans with high affinity and specificity [ 16 , 17 , 18 , 19 ]. In the present study, we added that oncofetal CS was highly expressed in stage I and II tumors and associated with poor OS and DFS in NSCLC patients. We observed strong oncofetal CS expression in both pericellular and extracellular matrix compartments with no correlation to EGFR or KRAS mutation. Accordingly, there is a potential opportunity for broad therapeutic application towards NSCLC with no limitation to certain subsets. Notably, we frequently observed high oncofetal CS expression in NSCLC with squamous cell histology, suggesting that oncofetal CS-derived therapeutics can be a useful therapeutic option for lung squamous cell carcinomas where no targeted therapy is otherwise available. In agreement with this, we observed VDC-MMAE sensitivity in all advanced NSCLC cell lines tested regardless of gene mutation status, histology, or oncogene expression. Hence, oncofetal CS constitutes a biomarker applicable for both targeted therapy as well as prognostic stratification of NSCLC.
High oncofetal CS expression is associated with poor DFS in NSCLC patients [ 16 ]. This is in line with other solid tumor malignancies such as melanoma and muscle-invasive bladder cancer [ 16 , 19 ]. Interestingly, elevated CHST11 expression, the enzyme required for CSA 4- O -sulfation, is also associated with poor DFS in three independent lung cancer cohorts [ 16 ]. A higher proportion of highly-sulfated CS was observed in the highly-metastatic LM660H11 lung cancer cell line, as compared to the P29 cell line with low metastatic potential [ 29 ]. Notably, preincubating Lewis lung cancer cells with rVAR2 lectins before intra-cardiac injection, strongly impeded the metastatic seeding in distant organs as compared to control groups [ 18 ]. Additionally, rVAR2-coated beads can capture circulating lung cancer cells with high specificity and efficiency from patient blood samples, as well as A549 cells spiked into healthy donor blood, even after transforming them into mesenchymal phenotype to resemble a step within metastatic cascade [ 30 ]. These studies indicate that in addition to early-stage NSCLC, oncofetal CS modifications are also expressed in later stages of lung cancer.
Curative intent lung resection surgery is the only effective therapy that results in long-term survival, but 55% of NSCLC patients experience recurrence [ 31 ]. For advanced disease, the current treatment paradigm includes immunotherapy, chemotherapy, and oncogenic driver-based targeted therapy. Recently, ADCs are being developed for NSCLC. Although tremendous resources are put into the search for new, targeted therapeutic options, no clear improvement in OS has been seen in combination with anti-angiogenic therapy in NSCLC [ 32 ] and there are no consistent results from tyrosine kinase inhibitor adjuvant studies [ 33 ]. Cancers with a high mutation burden, such as smoking-related lung cancer, require high sequencing capacity to reach sufficient sensitivity for the identification of low-frequency driver genes [ 34 ]. Targeting broadly expressed cancer-specific post-translational modifications such as oncofetal CS is an attractive strategy as these modifications are often expressed redundantly on different proteoglycans and do not depend on single gene alterations [ 35 , 36 ], making treatment resistance less likely to occur.
Aberrant glycosylation and expression of CSPGs are common in tumor initiation, progression, and prognosis [ 8 , 16 , 37 ]. Our oncofetal CS pulldown experiment on the A549 cell line identified a limited repertoire of oncofetal CS-modified CSPGs ( Figure 3 B). Among them, SDC1 expression is reportedly associated with NSCLC patient survival [ 38 ], independent of EGFR expression [ 39 ]. Additionally, high serum levels of SDC1, measured by ELISA, is an independent, poor-prognostic classifier in lung cancer patients [ 40 ]. Additionally, pre-treatment serum SDC1 levels can predict outcome in small cell lung cancer patients treated with platinum-based chemotherapy [ 41 ]. Recently, we validated the novel CS-glycosylation site on human SDC4 protein, which can be modified by CS chains at the attachment sites Ser39, Ser61, and Ser63 [ 25 ]. CD44 (also known as CSPG8) is another major CSPG identified in our analysis that is involved in tumorigenesis. CD44 inhibition attenuates EGFR signaling and enhances cisplatin sensitivity in EGFR wild-type NSCLC [ 42 ]. Additionally, CD44 promotes PD-L1 expression and its tumor-intrinsic function in both breast and lung malignancies [ 43 ]. High stromal expression of Versican correlates with poor tumor differentiation, disease recurrence, advanced tumor stage, and lymph node metastases [ 44 ]. Neuropilin 1 (NRP1) modulates TGF-β1-induced epithelial-mesenchymal transition in NSCLC [ 45 ], and dual-targeting of EGFR and NRP1 attenuates resistance to EGFR-targeted antibody therapy in KRAS-mutant NSCLC [ 46 ]. NRP1 expression correlates with radio-resistance [ 47 ], and NRP1 antagonism in human cancer cells inhibits migration and enhances chemosensitivity [ 48 , 49 ]. In our study, all NSCLC cells were effectively killed by VDC-MMAE in low-nM concentration. The A549 in vivo data confirmed that VDC-MMAE can effectively inhibit growth of oncofetal CS-positive NSCLC tumors and extend survival. We did not observe immune-related side effects or organ toxicity in this study or in our previous studies [ 16 , 18 , 19 ]. Based on dihydrodiol dehydrogenase (DDH) enzyme expression, Chen et al. showed that A549 is one of the most cisplatin-insensitive cell lines [ 50 ]. Indeed, VDCs have also shown promising results in cisplatin-resistant bladder tumors [ 19 ]. These observations support the idea that patients progressing after platinum-based therapy might still be eligible for oncofetal CS targeting regimens.
Limitations to our study include that our cohort represents a large, early-stage cancer cohort that provides a reasonable basis with both stroma and tumor compartment detailed assessments for the expression landscape of glycosaminoglycan. However, advanced NSCLC cases might represent a slightly different pattern. Thus, additional independent clinical prevalence studies in advanced NSCLC are needed to validate our findings in these patient groups.
5. Conclusions
In conclusion, our findings identify oncofetal CS as an independent prognosticator in early-stage NSCLC and a potential actionable therapeutic target in advanced NSCLC. Accordingly, our findings support a rationale for exploring oncofetal CS targeting opportunities in NSCLC.
Acknowledgments
In memory of friend, husband, and father Kristian Nim, who lost the battle to NSCLC in 2015.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Materials
The following are available online at https://www.mdpi.com/article/10.3390/cancers13174489/s1 . Figure S1: Study design and patient outcome of surgically resected NSCLC cohort. Table S1: Relationship between oncofetal CS expression and EGFR/KRAS mutation status.
Click here for additional data file.
Author Contributions
Conceptualization, H.Z.O., Z.L., G.J.W. and M.D.; methodology, H.Z.O., Z.L., N.K., J.L., G.K., J.P., H.A., H.B., B.Z., R.D., S.C., T.G., T.M.C., J.D.E., J.W.A., M.A.T., N.L.T., J.M., B.D., A.S. and N.A.-N.; software, H.A. and N.N.; formal analysis, J.P., H.A. and N.N.; investigation, Z.L.; resources, J.P. and H.A.; data curation, H.Z.O., Z.L., N.K., J.L., G.K., J.P., H.A., N.N., H.B., B.Z., R.D., S.C., T.G., T.M.C., J.D.E., J.W.A., M.A.T., N.L.T., J.M., B.D., A.S., N.A.-N. and G.J.W.; writing—original draft preparation, H.Z.O.; writing—review and editing, H.Z.O., Z.L., N.K., J.L., G.K., J.P., H.A., N.N., H.B., B.Z., R.D., S.C., T.G., T.M.C., J.D.E., J.W.A., M.A.T., N.L.T., J.M., B.D., A.S., N.A.-N., G.J.W. and M.D.; supervision, G.J.W. and M.D.; project administration, H.Z.O., N.A.-N. and G.J.W.; funding acquisition, M.D. All authors have read and agreed to the published version of the manuscript.
Funding
This work was supported by a grant from the Canadian Institutes of Health Research to M.D. (CIHR #153092). Z.L. was supported by the 2015 IASLC WCLC international mentorship program and the 2018 LCFA-BMS/IASLC Young Investigator Scholarship Award. Z.L. acknowledges funding from the Hungarian National Research, Development and Innovation Office (OTKA #124652 and OTKA #129664, Z). J.M. acknowledges funding from the Hungarian National Research, Development and Innovation Office (NAP2-2017-1.2.1- NKP-0002 and K129065). B.D. acknowledges funding from Austrian Science Fund (FWF I3977, I3522 and I4677). G.J.W. received funding from TGen Foundation, SHC Foundations and Flinn Foundation.
Institutional Review Board Statement
The study was conducted according to the guidelines of the Declaration of Helsinki. For patient samples, Scottsdale Healthcare IRB #2008-024 and Western IRB #20100600 approval under Exemption 4 of Title 45 Code of Federal Regulations (CFR) concerning retrospective study of existing data, qualifying lung tumor tissues were collected. Animal work was approved by the Animal Care Committee of the University of British Columbia (A18-0278, B19-0104).
Informed Consent Statement
Patient consent was waived due to the IRB approval under Exemption 4, Title 45 CFR, Part 46. Under Exemption 4, patient consent is not required and was not obtained for this study, as Title 45 CFR Part 46 does not apply. Specimens were de-identified and the clinical information associated with these specimens is not individually identifiable such that subjects cannot be identified either directly or indirectly through identifiers linked to the subjects.
Data Availability Statement
No new data were created or analyzed in this study. Data sharing is not applicable to this article.
Conflicts of Interest
M.D., A.S. and T.M.C. are co-founders of and shareholders in VAR2 Pharmaceuticals ( www.var2pharma.com ) and VarCT Diagnostics ( https://varctdiagnostics.com ). N.A.-N. is a consultant for VAR2 Pharmaceuticals. VAR2 Pharmaceuticals and VarCT Diagnostics are biotechnology companies that specialize in therapeutic and diagnostic development of the VAR2CSA technology. G.J.W. is an employee of SOTIO, LCC, outside of this submitted work, and a former employee of Unum Therapeutics outside of this submitted work; reports personal fees from Spring Bank Pharmaceuticals, Imaging Endpoints II, MiRanostics Consulting, Gossamer Bio, Paradigm, International Genomics Consortium, Angiex, IBEX Medical Analytics, GLG Council, Guidepoint Global, Circulogene, and Genomic Health, all outside this submitted work; has ownership interest in MiRanostics Consulting, Unum Therapeutics (now Cogent Biosciences), Exact Sciences, Moderna, and Circulogene, outside the submitted work; and has issued patents PCT/US2008/072787, PCT/US2010/043777, PCT/US2011/020612, and PCT/US20211037616, all outside the submitted work. M.A.T. is an advisor or a consultant for the following institutions: Adaptive, Advisory Board; BMS (Celgene)—Connect MDS/AML/MF Registry: Sci Steering Committee; Doximity—Medical Advisory Board; Elsevier Clinical Path (prior named: VIA Oncology) (1) Co-Chair Medical Oncology Myeloma Committee, (2) Co-Chair Medical Oncology Lymphoma Committee, and (3) Co-Chair Medical Hematology ITP Committee; Epizyme—Advisor Board SETD2 inhibitor; Janssen—Advisory Board—Daratumumab; Syapse Precision Medicine Council; Takeda—Insight—Multiple Myeloma Registry; and UpToDate, Peer Review for Plasma Cell Dyscrasias. The rest of the authors declare no potential conflicts of interest.
References
1. Gettinger S. Horn L. Jackman D. Spigel D. Antonia S. Hellmann M. Powderly J. Heist R. Sequist L.V. Smith D.C. Five-Year Follow-Up of Nivolumab in Previously Treated Advanced Non-Small-Cell Lung Cancer: Results From the CA209-003 Study J. Clin. Oncol. 2018 36 1675 1684 10.1200/JCO.2017.77.0412 29570421
2. Reungwetwattana T. Liang Y. Zhu V. Ou S.I. The race to target MET exon 14 skipping alterations in non-small cell lung cancer: The Why, the How, the Who, the Unknown, and the Inevitable Lung Cancer 2017 103 27 37 10.1016/j.lungcan.2016.11.011 28024693
3. Ekman S. How selecting best therapy for metastatic NTRK fusion-positive non-small cell lung cancer? Transl. Lung Cancer Res. 2020 9 2535 2544 10.21037/tlcr-20-434 33489816 PMC7815373
4. Hong D.S. Fakih M.G. Strickler J.H. Desai J. Durm G.A. Shapiro G.I. Falchook G.S. Price T.J. Sacher A. Denlinger C.S. KRAS(G12C) Inhibition with Sotorasib in Advanced Solid Tumors N. Engl. J. Med. 2020 383 1207 1217 10.1056/NEJMoa1917239 32955176 PMC7571518
5. Rinnerthaler G. Gampenrieder S.P. Greil R. HER2 Directed Antibody-Drug-Conjugates beyond T-DM1 in Breast Cancer Int. J. Mol. Sci. 2019 20 1115 10.3390/ijms20051115 PMC6429068 30841523
6. Bardia A. Mayer I.A. Diamond J.R. Moroose R.L. Isakoff S.J. Starodub A.N. Shah N.C. O’Shaughnessy J. Kalinsky K. Guarino M. Efficacy and Safety of Anti-Trop-2 Antibody Drug Conjugate Sacituzumab Govitecan (IMMU-132) in Heavily Pretreated Patients With Metastatic Triple-Negative Breast Cancer J. Clin. Oncol. 2017 35 2141 2148 10.1200/JCO.2016.70.8297 28291390 PMC5559902
7. Powles T. Rosenberg J.E. Sonpavde G.P. Loriot Y. Duran I. Lee J.L. Matsubara N. Vulsteke C. Castellano D. Wu C. Enfortumab Vedotin in Previously Treated Advanced Urothelial Carcinoma N. Engl. J. Med. 2021 384 1125 1135 10.1056/NEJMoa2035807 33577729 PMC8450892
8. Pinho S.S. Reis C.A. Glycosylation in cancer: Mechanisms and clinical implications Nat. Rev. Cancer 2015 15 540 555 10.1038/nrc3982 26289314
9. Ladenson R.P. Schwartz S.O. Ivy A.C. Incidence of the blood groups and the secretor factor in patients with pernicious anemia and stomach carcinoma Am. J. Med. Sci. 1949 217 194 197 10.1097/00000441-194902000-00011 18109280
10. Hakomori S.I. Murakami W.T. Glycolipids of hamster fibroblasts and derived malignant-transformed cell lines Proc. Natl. Acad. Sci. USA 1968 59 254 261 10.1073/pnas.59.1.254 4298334 PMC286029
11. Khazamipour N. Al-Nakouzi N. Oo H.Z. Orum-Madsen M. Steino A. Sorensen P.H. Daugaard M. Oncofetal Chondroitin Sulfate: A Putative Therapeutic Target in Adult and Pediatric Solid Tumors Cells 2020 9 818 10.3390/cells9040818 PMC7226838 32231047
12. Ajit Varki E.E. Cummings R.D. Esko J.D. Stanley P. Hart G.W. Aebi M. Darvill A.G. Kinoshita T. Packer N.H. Prestegard J.H. Essentials of Glycobiology [Internet] 3rd ed. Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY, USA 2015–2017 27010055
13. Rogerson S.J. Chaiyaroj S.C. Ng K. Reeder J.C. Brown G.V. Chondroitin sulfate A is a cell surface receptor for Plasmodium falciparum-infected erythrocytes J. Exp. Med. 1995 182 15 20 10.1084/jem.182.1.15 7790815 PMC2192085
14. Ayres Pereira M. Clausen T.M. Pehrson C. Mao Y. Resende M. Daugaard M. Kristensen A.R. Spliid C. Mathiesen L. Placental Sequestration of Plasmodium falciparum Malaria Parasites Is Mediated by the Interaction Between VAR2CSA and Chondroitin Sulfate A on Syndecan-1 PLoS Pathog. 2016 12 e1005831 10.1371/journal.ppat.1005831 27556547 PMC4996535
15. Salanti A. Dahlback M. Turner L. Nielsen M.A. Barfod L. Magistrado P. Jensen A.T. Lavstsen T. Ofori M.F. Marsh K. Evidence for the involvement of VAR2CSA in pregnancy-associated malaria J. Exp. Med. 2004 200 1197 1203 10.1084/jem.20041579 15520249 PMC2211857
16. Salanti A. Clausen T.M. Agerbaek M.O. al Nakouzi N. Dahlback M. Oo H.Z. Lee S. Gustavsson T. Rich J.R. Hedberg B.J. Targeting Human Cancer by a Glycosaminoglycan Binding Malaria Protein Cancer Cell 2015 28 500 514 10.1016/j.ccell.2015.09.003 26461094 PMC4790448
17. Clausen T.M. Pereira M.A. Oo H.Z. Resende M. Gustavson T. Mao Y. Sugiura N. Liew J. Fazli L. Theander T.G. Real-time and label free determination of ligand binding-kinetics to primary cancer tissue specimens; a novel tool for the assessment of biomarker targeting Sens. Biosensing Res. 2016 9 23 30 10.1016/j.sbsr.2016.05.003 27441183 PMC4942562
18. Clausen T.M. Pereira M.A. al Nakouzi N. Oo H.Z. Agerbaek M.O. Lee S. Orum-Madsen M.S. Kristensen A.R. El-Naggar A. Grandgenett P.M. Oncofetal Chondroitin Sulfate Glycosaminoglycans Are Key Players in Integrin Signaling and Tumor Cell Motility Mol. Cancer Res. 2016 14 1288 1299 10.1158/1541-7786.MCR-16-0103 27655130 PMC5136311
19. Seiler R. Oo H.Z. Tortora D. Clausen T.M. Wang C.K. Kumar G. Pereira M.A. Orum-Madsen M.S. Agerbaek M.O. Gustavsson T. An Oncofetal Glycosaminoglycan Modification Provides Therapeutic Access to Cisplatin-resistant Bladder Cancer Eur. Urol. 2017 72 142 150 10.1016/j.eururo.2017.03.021 28408175
20. Edge S.B. Compton C.C. The American Joint Committee on Cancer: The 7th edition of the AJCC cancer staging manual and the future of TNM Ann. Surg. Oncol. 2010 17 1471 1474 10.1245/s10434-010-0985-4 20180029
21. Raz G. Allen K.E. Kingsley C. Cherni I. Arora S. Watanabe A. Lorenzo C.D. Edwards V.D. Sridhar S. Hostetter G. Hedgehog signaling pathway molecules and ALDH1A1 expression in early-stage non-small cell lung cancer Lung Cancer 2012 76 191 196 10.1016/j.lungcan.2011.10.015 22115706
22. Yumura K. Akiba H. Nagatoishi S. Kusano-Arai O. Iwanari H. Hamakubo T. Tsumoto K. Use of SpyTag/SpyCatcher to construct bispecific antibodies that target two epitopes of a single antigen J. Biochem. 2017 162 203 210 10.1093/jb/mvx023 28637250
23. Reddington S.C. Howarth M. Secrets of a covalent interaction for biomaterials and biotechnology: SpyTag and SpyCatcher Curr. Opin. Chem. Biol. 2015 29 94 99 10.1016/j.cbpa.2015.10.002 26517567
24. Jonsson A. Dogan J. Herne N. Abrahmsen L. Nygren P.A. Engineering of a femtomolar affinity binding protein to human serum albumin Protein Eng. Des. Sel. 2008 21 515 527 10.1093/protein/gzn028 18499681
25. Toledo A.G. Pihl J. Spliid C.B. Persson A. Nilsson J. Pereira M.A. Gustavsson T. Choudhary S. Oo H.Z. Black P.C. An affinity chromatography and glycoproteomics workflow to profile the chondroitin sulfate proteoglycans that interact with malarial VAR2CSA in the placenta and in cancer Glycobiology 2020 30 989 1002 10.1093/glycob/cwaa039 32337544 PMC7724742
26. Nakano T. Kanai Y. Amano Y. Yoshimoto T. Matsubara D. Shibano T. Tamura T. Oguni S. Katashiba S. Ito T. Establishment of highly metastatic KRAS mutant lung cancer cell sublines in long-term three-dimensional low attachment cultures PLoS ONE 2017 12 e0181342 10.1371/journal.pone.0181342 28786996 PMC5546599
27. Moran D.M. Trusk P.B. Pry K. Paz K. Sidransky D. Bacus S.S. KRAS mutation status is associated with enhanced dependency on folate metabolism pathways in non-small cell lung cancer cells Mol. Cancer Ther. 2014 13 1611 1624 10.1158/1535-7163.MCT-13-0649 24688052
28. Robert N.J. Espirito J.L. Chen L. Karhade M. Evangelist M.C. Spira A.I. Neubauer M.A. Bullock S.A. Coleman R.L. Biomarker tissue journey among patients (pts) with untreated metastatic non-small cell lung cancer (mNSCLC) in the U.S. Oncology Network community practices J. Clin. Oncol. 2021 39 S9004 10.1200/JCO.2021.39.15_suppl.9004
29. Li F. Dam G.B.T. Murugan S. Yamada S. Hashiguchi T. Mizumoto S. Oguri K. Okayama M. van Kuppevelt T.H. Sugahara K. Involvement of highly sulfated chondroitin sulfate in the metastasis of the Lewis lung carcinoma cells J. Biol. Chem. 2008 283 34294 34304 10.1074/jbc.M806015200 18930920 PMC2662238
30. Agerbaek M.O. Bang-Christensen S.R. Yang M.H. Clausen T.M. Pereira M.A. Sharma S. Ditlev S.B. Nielsen M.A. Choudhary S. Gustavsson T. The VAR2CSA malaria protein efficiently retrieves circulating tumor cells in an EpCAM-independent manner Nat. Commun. 2018 9 3279 10.1038/s41467-018-05793-2 30115931 PMC6095877
31. Uramoto H. Tanaka F. Recurrence after surgery in patients with NSCLC Transl. Lung Cancer Res. 2014 3 242 249 25806307 10.3978/j.issn.2218-6751.2013.12.05 PMC4367696
32. Alshangiti A. Chandhoke G. Ellis P.M. Antiangiogenic therapies in non-small-cell lung cancer Curr. Oncol. 2018 25 Suppl. 1 S45 S58 10.3747/co.25.3747 29910647 PMC6001774
33. Sandler J.E. D’Aiello A. Halmos B. Changes in store for early-stage non-small cell lung cancer J. Thorac. Dis. 2019 11 2117 2125 10.21037/jtd.2019.05.34 31285907 PMC6588773
34. Lawrence M.S. Stojanov P. Mermel C.H. Robinson J.T. Garraway L.A. Golub T.R. Meyerson M. Gabriel S.B. Lander E.S. Getz G. Discovery and saturation analysis of cancer genes across 21 tumour types Nature 2014 505 495 501 10.1038/nature12912 24390350 PMC4048962
35. Sinchaikul S. Hongsachart P. Sriyam S. Tantipaiboonwong P. Phutrakul S. Chen S.T. Current proteomic analysis and post-translational modifications of biomarkers in human lung cancer materials Chang. Gung Med. J. 2008 31 417 430 19097588
36. Oo H.Z. Seiler R. Black P.C. Daugaard M. Post-translational modifications in bladder cancer: Expanding the tumor target repertoire Urol. Oncol. 2020 38 858 866 10.1016/j.urolonc.2018.09.001 30342880
37. Iida J. Meijne A.M. Knutson J.R. Furcht L.T. McCarthy J.B. Cell surface chondroitin sulfate proteoglycans in tumor cell adhesion, motility and invasion Semin. Cancer Biol. 1996 7 155 162 10.1006/scbi.1996.0021 8773301
38. Shah L. Walter K.L. Borczuk A.C. Kawut S.M. Sonett J.R. Gorenstein L.A. Ginsburg M.E. Steinglass K.M. Powell C.A. Expression of syndecan-1 and expression of epidermal growth factor receptor are associated with survival in patients with nonsmall cell lung carcinoma Cancer 2004 101 1632 1638 10.1002/cncr.20542 15378500
39. Toyoshima E. Ohsaki Y. Nishigaki Y. Fujimoto Y. Kohgo Y. Kikuchi K. Expression of syndecan-1 is common in human lung cancers independent of expression of epidermal growth factor receptor Lung Cancer 2001 31 193 202 10.1016/S0169-5002(00)00184-7 11165398
40. Joensuu H. Anttonen A. Eriksson M. Makitaro R. Alfthan H. Kinnula V. Leppa S. Soluble syndecan-1 and serum basic fibroblast growth factor are new prognostic factors in lung cancer Cancer Res. 2002 62 5210 5217 12234986
41. Anttonen A. Leppa S. Ruotsalainen T. Alfthan H. Mattson K. Joensuu H. Pretreatment serum syndecan-1 levels and outcome in small cell lung cancer patients treated with platinum-based chemotherapy Lung Cancer 2003 41 171 177 10.1016/S0169-5002(03)00196-X 12871780
42. Yin J. Zhang H. Wu X. Zhang Y. Li J. Shen J. Zhao Y. Xiao Z. Lu L. Huang C. CD44 inhibition attenuates EGFR signaling and enhances cisplatin sensitivity in human EGFR wildtype nonsmallcell lung cancer cells Int. J. Mol. Med. 2020 45 1783 1792 32236608 10.3892/ijmm.2020.4562 PMC7169661
43. Kong T. Ahn R. Yang K. Zhu X. Fu Z. Morin G. Bramley R. Cliffe N.C. Xue Y. Kuasne H. CD44 Promotes PD-L1 Expression and Its Tumor-Intrinsic Function in Breast and Lung Cancers Cancer Res. 2020 80 444 457 10.1158/0008-5472.CAN-19-1108 31722999
44. Pirinen R. Leinonen T. Bohm J. Johansson R. Ropponen K. Kumpulainen E. Kosma V.M. Versican in nonsmall cell lung cancer: Relation to hyaluronan, clinicopathologic factors, and prognosis Hum. Pathol. 2005 36 44 50 10.1016/j.humpath.2004.10.010 15712181
45. Ding Z. Du W. Lei Z. Zhang Y. Zhu J. Zeng Y. Wang S. Zheng Y. Liu Z. Huang J.A. Neuropilin 1 modulates TGFbeta1induced epithelialmesenchymal transition in nonsmall cell lung cancer Int. J. Oncol. 2020 56 531 543 31894269 10.3892/ijo.2019.4938 PMC6959462
46. Kim Y.J. Baek D.S. Lee S. Park D. Kang H.N. Cho B.C. Kim Y.S. Dual-targeting of EGFR and Neuropilin-1 attenuates resistance to EGFR-targeted antibody therapy in KRAS-mutant non-small cell lung cancer Cancer Lett. 2019 466 23 34 10.1016/j.canlet.2019.09.005 31521695
47. Dong J.C. Gao H. Zuo S.Y. Zhang H.Q. Zhao G. Sun S.L. Han H.L. Jin L.L. Shao L.H. Wei W. Neuropilin 1 expression correlates with the Radio-resistance of human non-small-cell lung cancer cells J. Cell Mol. Med. 2015 19 2286 2295 10.1111/jcmm.12623 26147006 PMC4568932
48. Jia H. Cheng L. Tickner M. Bagherzadeh A. Selwood D. Zachary I. Neuropilin-1 antagonism in human carcinoma cells inhibits migration and enhances chemosensitivity Br. J. Cancer 2010 102 541 552 10.1038/sj.bjc.6605539 20087344 PMC2822953
49. Hong T.M. Chen Y.L. Wu Y.Y. Yuan A. Chao Y.C. Chung Y.C. Wu M.H. Yang S.C. Pan S.H. Shih J.Y. Targeting neuropilin 1 as an antitumor strategy in lung cancer Clin. Cancer Res. 2007 13 4759 4768 10.1158/1078-0432.CCR-07-0001 17699853
50. Chen J. Emara N. Solomides C. Parekh H. Simpkins H. Resistance to platinum-based chemotherapy in lung cancer cell lines Cancer Chemother. Pharmacol. 2010 66 1103 1111 10.1007/s00280-010-1268-2 20953859 PMC2957658