Two human NSCLC cell lines, A549 [American Type Culture Collection (ATCC)^® CCL-185] and H460 (HTB-177^TM), and a human lung fibroblast cell line, LL97A (CCL-191^TM) were purchased from the ATCC (Manassas, VA, USA). The A549-Red-FLuc Bioware^® Brite cell line, which has been stably transduced with the red-shifted firefly luciferase gene from Luciola italica (Red-FLuc) was purchased from PerkinElmer Inc. (Waltham, WA, USA). Authentication of A549, H460, LL97A and A549-Red-Fluc cell lines was performed by IDEXX BioAnalytics to ensure that individual cell lines were correctly identified and free of contamination of non-human cells (Columbia, MO, USA). A549, A549-Red-Fluc, and LL97A cell lines were cultured in a mixture of Dulbecco’s Modified Eagle’s Medium (DMEM)/Ham’s F12 at a ratio of 1:1 (Corning Inc., Corning, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA, USA). The H460 cell line was cultured in RPMI-1640 medium (Corning Inc., Corning, NY, USA) supplemented with 10% heat-inactivated FBS. Penicillin (100 units/mL) and streptomycin (100 µg/mL) were added to the cell culture media to prevent contamination. Cells were maintained at 37℃ in a humidified atmosphere of 5% CO₂ in air and tested for mycoplasma using the MycoAlert Mycoplasma Detection Kit (catalog no.: LT07-118, Lonza, Allendale, NJ, USA) periodically throughout the studies to ensure that all cells used in the study were free from mycoplasma contamination.

Male athymic nude mice (Hsd: Athymic Nude-Foxn1nu; 5–6 weeks old) were purchased from Envigo (Indianapolis, IN, USA).

The therapeutic anti-human GPC1 mouse mAb was prepared by MegaNano Biotech Inc. (Tampa, FL, USA). In brief, human GPC1 gene and fragments (two overlapping fragments with one from AA1−280 and the other one from AA240−559. Sequence information was provided in Supplemental materials) were cloned into PGEX-6p-1 expression vector and expressed in BL21 bacteria. Fusion proteins were cleaved from glutathione Sepharose 4B beads with PreScission protease (Sigma-Aldrich Inc., St. Louis, Mo, USA). For immunization and hybridoma generation, mouse splenocytes were isolated and fused with Sp2/0 mouse myeloma cells. Positive clones selected with human recombinant protein were amplified in large quantity. Antibodies were lyophilized after being isolated with protein A beads.

The in vitro cytotoxic activity of anti-GPC1 mAb was determined in cultured A549 and H460 human NSCLC cells and LL97A human lung fibroblasts by an MTT assay as described previously with minor modifications [30]. A549 and H460 human NSCLC cells and LL97A lung fibroblasts were seeded in 96-well plates at a density of 4 × 10E3 cells/well and allowed to attach overnight. On the next day, culture media containing anti-GPC1 mAb (4.5–290 μg/mL) were added to appropriate wells. After the cells were treated for 72 h, an aliquot of 10 μL of 5 mg/mL MTT dissolved in phosphate-buffered saline (PBS) was added to each well, and cells were incubated at 37℃ for another 4 h. Finally, cells were lysed for 1 h with the addition of 100 μL of 100% DMSO. Plates were read at 570 nm on a SpectraMax 190 microplate reader equipped with SoftMax Pro software (Molecular Devices, Sunnyvale, CA, USA). The growth of treated cells was expressed as a percentage of vehicle control cultures. The concentration of anti-GPC1 mAb required for 50% inhibition of cell growth (i.e., IC₅₀) as compared with the control cells was calculated by nonlinear fitting of the experimental data obtained from multiple independent experiments performed in quadruplicates using the GraphPad Prism 5.0 program (GraphPad Software, Inc. La Jolla, CA, USA).

A549 and H460 monoculture spheroids and A549/LL97A and H460/LL97A coculture spheroids (in a tumor cell to fibroblast ratio of 1:1) were generated by seeding a total of 4,000 cells per well in 200 μL of complete culture medium on a Spheron Nunclon 96-well plate (Thermo Fisher Scientific, Waltham, MA, USA) followed by centrifugation at 400 × g for 10 min. After the spheroids were cultured for 24 h, various amounts of anti-GPC1 mAb were added to achieve the final concentrations of 40 μg/mL, 80 μg/mL, 120 μg/mL, 160 μg/mL and 200 μg/mL and incubated for 72 h. After the 72-hour treatment with the vehicle control and anti-GPC1 mAb, spheroid viability was measured by reading luminescence (Synergy Neo2 Hybrid Multi-Mode Reader; Biotek Instruments, Inc. Winooski, VT, USA) after 20 min incubation with CellTiter-Glo 3D reagent (50 µL/well; Promega Co. Madison, WI, USA) according to the manufacturer’s protocol.

The inhibitory effect of anti-GPC1 mAb on the anchorage-independent growth of A549 and H460 tumor cells was evaluated using a modified two-layer soft agar culture system [31]. For the bottom layer of agar, 0.3 mL of complete culture medium containing 0.5% noble agar (Thermo Fisher Scientific, Waltham, MA, USA) was added to each well of a 24-well plate. After the bottom layer was solidified, 2 × 10E3 A549 or H460 cells were suspended in 0.2 mL of the same culture medium containing 0.3% noble agar and vehicle control, or 100 μg/mL mouse immunoglobulin G (IgG) isotype, or anti-GPC1 mAb at various concentrations and placed on the base layer. After the upper layer of agar was solidified, 200 μL of complete culture medium containing the same control or mAb at the same concentration as the upper layer of agar was added to prevent desiccation. Colony formation by A549 and H460 cells was evaluated after 10–12 days of culture for number and size of colonies. Twenty-four hours prior to the evaluation, 200 μL of 1 mg/mL of MTT was added to each well for the evaluation of metabolic viability of the colonies [32]. Colonies were photographed using the Bio-Rad Gel Doc XR+ gel documentation system (Bio-Rad Laboratories, Inc. Hercules, CA, USA). The number and size of colonies were quantified using ImageJ (https://imagej.nih.gov/ij/). The effect of the antibody treatment was expressed by relative number of colonies [33] and percent of colony area [34].

To generate 3D tumor cell monoculture and tumor cell/fibroblast coculture spheroids, A549 or H460 cells alone (3,000 cells per well), or A549 or H460 cells mixed with LL97A fibroblasts at a ratio of 1:1 (a total of 3,000 cells per well), were added to a Spheron Nunclon 96-well plate (Thermo Fisher Scientific) followed by centrifugation at 400 × g for 10 min. After 3 days of incubation under the standard culture condition, spheroids were suspended in 1.5 mg/mL CultrexTM rat collagen I (R&D Systems; Minneapolis, MN, USA) diluted with the complete DMEM culture medium, and then transferred to a flat-bottom 96-well plate. After collagen solidified, complete DMEM culture medium was added to the top of the collagen matrix to provide a chemogradient for the spheroids. To evaluate the effect of anti-GPC1 mAb on cell invasion, anti-GPC1 mAb (50 μg/mL and 100 μg/mL) or mouse IgG isotype (100 μg/mL) was added directly to the collagen matrix to achieve the desired final concentration before solidification and to the complete culture medium on top of the solidified collagen matrix. Spheroids were maintained under the standard culture condition for 4 days to allow for invasion. Bright-field microscopy images were acquired daily on the EVOS^TM XL imaging system (Thermo Fisher Scientific, Waltham, MA, USA). Cell-covered area was measured using ImageJ (https://imagej.nih.gov/ij/).

The indirect tumor cell-fibroblast co-culture was carried out using a transwell system with Corning^TM Transwell^TM 24 mm inserts and 0.4 μm pore polyester membranes (Corning Inc., Corning, NY, USA). A549 and H460 lung tumor cells were seeded in the lower chamber at the density of 5 × 10E4 cells/well, while LL97A lung fibroblasts were in the upper chamber at the density of 10 × 10E4 cells/insert. Co-cultured cells were maintained in DMEM/F12 media containing 10% FBS in the absence and presence of vehicle, mouse IgG isotype control or anti-GPC1 mAb.

The orthotopic lung tumor model was established by intrapulmonary injection of A549-Red-Fluc cells in athymic nude mice using the method described elsewhere with slight modifications [35]. In brief, A549-Red-Fluc cells harvested during the exponential growth in vitro were suspended in PBS containing 10% Matrigel^® growth factor reduced (GFR) basement membrane matrix (Corning Inc., Corning, NY, USA) prior to the tumor cell inoculation procedure. Athymic nude mice were anesthetized under isoflurane and positioned in the lateral decubitus position with the left chest facing up and a small (about 5 mm) incision was made in the skin just below the scapula. Next, the chest wall soft tissue was gently spread to expose the thoracic ribs and intercostal space to allow visualization of the left lobe of the lung. At this stage, 50 μL of tumor cell suspension (1 × 10E6 A549-Red-Fluc cells in 50 μL PBS containing 10% Matrigel^® GFR) was drawn into a 3/10-cc insulin syringe equipped with a 31-G hypodermic needle (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). Then the needle was inserted between the 6th and 7th ribs and the tumor cell suspension was injected 5 mm deep into the left lung parenchyma over a 1-minute period. After injection of the cells, the syringe was withdrawn from the tissue and the initial surgery incision made in the skin was closed with non-absorbable Ethilon^® 6-0 nylon sutures. After the orthotopic inoculation of tumor cells, weekly in vivo bioluminescence imaging (BLI) was performed throughout the experiment using the IVIS^® Spectrum in vivo imaging system (Perkin Elmer Inc., Waltham, MA, USA) to monitor changes in tumor burden over time in the same animals. To perform the in vivo BLI, tumor bearing mice received intraperitoneal (i.p.) injection of 150 mg/kg d-luciferin (Gold Biotechnology Inc., Olivette, MO, USA) 8 min before imaging and were anesthetized with isoflurane during imaging. Bioluminescence emitted from each tumor bearing animal was acquired using the Living Image Software (Perkin Elmer Inc., Waltham, MA, USA). Region of interest (ROI) measurements on the images were used to convert surface radiance to total flux of photons expressed in photons/second and used as readout for the orthotopic lung tumor size. For longitudinal comparison of bioluminescence, ROI was manually selected and kept consistent across all experiments.

Intravenous (i.v.) dosing regimen: to evaluate the effect of anti-GPC1 mAb on orthotopic lung tumor growth, athymic nude mice were randomly divided into (1) vehicle control, (2) mouse IgG isotype 10 mg/kg and (3) anti-GPC1 mAb 10 mg/kg groups once the orthotopic lung tumor lesions were detected by BLI in those animals, which was at 4–6 weeks after intrapulmonary tumor cell inoculation. Tumor-bearing animals were given i.v. injection of saline, 10 mg/kg of mouse IgG isotype and 10 mg/kg of anti-GPC1 mAb through tail vein once every 72 h for a total of 8 doses. All animals were euthanized 24 h after the last dose. Plasma, lung tissue and tumor tissue samples collected from individual animals were stored at −80℃ before subjected to further analyses.

I.p. dosing regimen: to evaluate the efficacy of anti-GPC1 mAb treatment on orthotopic lung tumor growth, athymic nude mice were randomly divided into four groups: (1) vehicle control, (2) anti-GPC1 mAb 1 mg/kg, (3) anti-GPC1 mAb 10 mg/kg and anti-GPC1 mAb 50 mg/kg groups. Two weeks after the orthotopic tumor cell inoculation, individual animals received i.p. injection of saline or anti-GPC1 mAb at different dose levels once a week for 3 doses (i.e., on Day 0, 7 and 14) and then once every 10 days for 2 more doses (i.e., on Day 24 and 34). Animals were euthanized 1 week after the last dose (i.e., on Day 41). Superficial and internal macroscopic tumors in the lungs were assessed at necropsy. Those that failed to develop lung tumors were excluded from further analyses. Plasma, lung tissue and tumor tissue samples collected from individual tumor-bearing animals were stored at −80℃ before subjected to further analyses.

The method of euthanasia for athymic nude mice used in this study was exsanguination under isoflurane anesthesia and the death was confirmed by cervical dislocation of the neck prior to tumor tissue resection.

A 96-well Immulon^® Microtiter^TM plate (Thermo Fisher Scientific, Waltham, MA, USA) was coated with 50 μL of recombinant human GPC1 protein (2.5 μg/mL in PBS) for 1 h at 37℃. After the coated plate was washed twice with PBS containing 0.1% Tween 20 (PBST) and blocked with 100 μL/well of 1.5% BSA in PBST at 37℃ for 30 min, an aliquot of 100 μL of diluted plasma sample (1:200 dilution in PBST containing 1.5% BSA) was added to each well and incubated at 37℃ for 1 h. After the plasma samples were removed, the plate was washed four times with PBST and then incubated with 100 μL/well of diluted goat anti-mouse-horseradish peroxidase (HRP) antibody (1:4,000 dilution in PBST containing 1.5% BSA; Southern Biotech, Birmingham, AL, USA) at 37℃ for 1 h. For colorimetric enzyme-linked immunosorbent assay (ELISA) detection, 100 μL of 3,3’,5,5’-tetramethylbenzidine (TMB; Surmodics IVD, Inc., Eden Prairie, MN) was added to each well. After the plate was incubated at room temperature for 10–15 min, the enzymatic reaction was stopped by adding 100 μL of 0.4 mol/L of H₂SO₄ to each well. The plate was read at 450 nm on a SpectraMax 190 microplate reader equipped with SoftMax Pro software (Molecular Devices, Sunnyvale, CA, USA). The optical density was used to estimate plasma levels of anti-GPC1 mAb.

Monocultured and co-cultured A549, H460 and LL97A cells, and normal lung and A549 lung tumor tissue samples collected from the in vivo studies were subjected to semi-quantitative Western blot analysis to determine the phosphorylation of downstream effectors of FGFR signaling, including the MEK/ERK/RSK, Akt/GSK3/mTOR and Src/focal adhesion kinase (FAK) pathways, as well as the protein expression of GPC1 and selected epithelial-mesenchymal transition (EMT) markers. A complete list of primary antibodies used in the Western blot analysis is shown in Table S1.

Statistical analyses were performed using Number Cruncher Statistical Systems 2007 (Keysville, UT, USA). Data are presented as the mean ± standard deviation (SD) unless otherwise indicated. The two-sample t-test was used to determine if there was a statistically significant difference between the means of two independent groups. Comparison of means between more than two independent groups was made using the one-way ANOVA followed by Tukey-Kramer post hoc multiple comparison test. To compare the tumor growth rate among three study groups, tumor sizes were transformed with square root and linear regressions were applied to fit the tumor growth curves for individual mice. When three or more independent sample data did not follow a normal distribution, a nonparametric Kruskal-Wallis test was used to compare the independent groups instead of one-way ANOVA. Pearson’s correlation coefficient was used to describe the strength of linear correlation between two variables. A two-sided P-value of less than 0.05 was considered statistically significant.

Evaluation of the in vitro cytotoxic effect of anti-GPC1 mAb in 2D A549, H460 and LL97A monocultures was carried out using the MTT assay after the cells were treated with anti-GPC1 mAb or mouse IgG isotype at concentrations ranging from 4.5 μg/mL to 290 μg/mL for 72 h. The effect of anti-GPC1 mAb on the viability of 3D A549 and H460 monoculture spheroids and A549/LL97A and H460/LL97A coculture spheroids was determined using the CellTiter-Glo 3D reagent that generates the stable ATP-dependent luminescent signal. Figure 1 shows the concentration-effect curves for anti-GPC1 mAb and mouse IgG isotype treatments in different cell lines. IC₅₀ values were calculated by fitting the log(inhibitor) vs. normalized response model to the data using non-linear regression in GraphPad Prism 8.0.1 (GraphPad Software. Boston, MA, USA). The difference in the IC₅₀ values was then compared between the two treatment groups in individual cell lines. Results of the MTT assay showed that the IC₅₀ values of anti-GPC1 mAb for A549 NSCLC cells and LL97A lung fibroblasts were significantly lower than that for H460 cells (P < 0.01 for both; Figure 1A–C). Among the three cell lines tested, LL97A lung fibroblasts were shown with the lowest IC₅₀ value, suggesting that the lung fibroblasts are more sensitive to anti-GPC1 mAb treatment than the NSCLC cells. Treatment with the mouse IgG isotype for 72 h had no effect on the viability of all three cell lines tested (Figure 1A–C). Results of the 3D cell viability assay indicated that the IC₅₀ values of anti-GPC1 mAb in A549/LL97A and H460/LL97A coculture spheroids were 15% and 44% lower than those in A549 and H460 monoculture spheroids, respectively, (P < 0.01 for both; Figure 1D–E), suggesting tumor cell-fibroblast cocultures are more sensitive to the cytotoxic effect of anti-GPC1 mAb than tumor cell monocultures.

In soft agar colony formation assay, tumor cells were deprived of their anchorage by being suspended in 0.3% noble agar as single cells (Figure 2A). The relative number of colonies [colony area (%)] and percent of well area covered by colonies [colony area (%)] were determined to quantify the inhibitory effect of anti-GPC1 mAb on tumor cell survival and proliferation under the anchorage-independent condition. For A549 cells, treatment with 40 μg/mL and 60 μg/mL of anti-GPC1 mAb resulted in a significantly less number of colonies as compared with the number of colonies formed in the vehicle control and 80 μg/mL mouse IgG isotype groups (P < 0.01 for all; Figure 2B). Treatment with 20 μg/mL, 40 μg/mL and 60 μg/mL of anti-GPC1 mAb also significantly inhibited the anchorage-independent growth of A549 cells, which was reflected by the significant decrease in the percent colony areas in the anti-GPC1 mAb treated A549 cells as compared with those in A549 cells treated with the vehicle control or 80 μg/mL mouse IgG isotype. For H460 cells, treatment with 10 μg/mL and 20 μg/mL of anti-GPC1 mAb resulted in a significant decrease in the relative number of colonies formed as compared with the vehicle treatment (P < 0.01 for all; Figure 2B). Treatment with 5 μg/mL, 10 μg/mL and 20 μg/mL of anti-GPC1 mAb significantly decreased the colony areas (%) as compared with treatment with vehicle or 40 μg/mL of mouse IgG isotype. The inhibitory effect of anti-GPC1 mAb on the colony formation was concentration dependent irrespective of the cell line used (P < 0.01 for all; Pearson’s correlation). It was noted that the estimated concentrations at which 50% of colony formation was inhibited were significantly lower than the IC₅₀ values obtained from the MTT assay performed in 2D cell culture (41.2 μg/mL ± 0.93 μg/mL vs. 169 μg/mL ± 19.2 μg/mL for A549 cells; 13.6 μg/mL ± 0.59 μg/mL vs. 646 μg/mL ± 103 μg/mL for H460 cells. P < 0.01 for both; Figure 1A, 1B and 2B). This finding suggests that the anchorage-independent tumor cell growth is more sensitive to anti-GPC1 mAb treatment.

The 3D spheroid invasion assay was conducted to evaluate the effect of anti-GPC1 mAb on the invasion potential of A549 and H460 tumor cells that were either monocultured or cocultured with LL97A fibroblasts (Figure 3). The final concentrations of anti-GPC1 mAb used in monoculture and coculture spheroids were 50 μg/mL and 100 μg/mL. As indicated by the fold change in cell-covered area relative to Day 0, treatment with anti-GPC1 mAb at 50 μg/mL and 100 μg/mL had little effect on the invasion potential of A549 and H460 monoculture spheroids as compared with the vehicle control or 100 μg/mL mouse IgG isotype treatment (Figure 3A and 3C). GPC1 is known to function as a co-receptor/modulator of various growth factors to enhance the binding to their respective receptors [4–11], and cancer-associated fibroblasts (CAFs) are an important contributor of some of those growth factors in tumors, including FGF2 and TGFβ [36, 37]. In this regard, the tumor cell/fibroblast coculture spheroids were used to evaluate the anti-invasion potential of anti-GPC1 mAb. A549/LL97A coculture spheroids treated with 100 μg/mL of anti-GPC1 mAb exhibited significantly reduced cell-covered areas compared with those treated with vehicle control or 100 μg/mL of mouse IgG isotype from Day 2 through 4 (P < 0.05 for all; Figure 3B). In H460/LL97A coculture spheroids, treatment with anti-GPC1 mAb at 50 and 100 μg/mL resulted in significantly less cell-covered areas than treatment with 100 μg/mL of mouse IgG isotype on Day 2 (P < 0.01 for both), Day 3 (P < 0.01 for both) and Day 4 (P < 0.05 for both; Figure 3D). Overall, these results suggest that anti-GPC1 mAb treatment inhibits fibroblast-led collective tumor cell invasion.

Since GPC1 is believed to facilitate the activation of signaling pathways downstream of the receptors of HB growth factor family of proteins, it is anticipated that anti-GPC1 mAb blocks the interaction between GPC1 and those growth factors, and subsequently hinders the activation of the cognate receptors and downstream pathways. In this study, A549, H460 and LL97A cells that were cultured as monocultures or co-cultured in an indirect transwell coculture system were exposed to vehicle control, 100 μg/mL of mouse IgG isotype or 100 μg/mL of anti-GPC1 mAb for 24 h. The collected whole cell lysate samples were subjected to Western blotting analysis to examine the effect of in vitro anti-GPC1 mAb treatment on common downstream growth factor receptor signaling pathways, including the MEK/ERK/RSK, Akt/GSK3/mTOR and Src/FAK pathways, as well as the protein expression of GPC1 and selected EMT markers, including β-catenin and vimentin. As shown in Figures 4 and 5, GPC1 protein expression was detected in all three cell lines by Western blotting, indicating that the target antigen of anti-GPC1 mAb was present in all three cell lines tested. In monocultures, treatment with anti-GPC1 mAb had no effect on the levels of phosphorylated signaling effectors and the protein expression of GPC1, β-catenin and vimentin proteins in A549 and H460 NSCLC cell lines (Figure 4). However, in LL97A lung fibroblast monocultures, as compared with both of the vehicle control and mouse IgG isotype control treatment, anti-GPC1 mAb treatment significantly decreased the FAK phosphorylation at Tyr397 (P < 0.01 for both), RSK phosphorylation at Ser380 (P < 0.05 for both), Akt phosphorylation at Ser473 (P < 0.01 for both), and vimentin protein expression (P < 0.05 compared with the vehicle control and P < 0.01 compared with the mouse IgG isotype; Figure 6A). Moreover, compared with the vehicle control treatment, the anti-GPC1 mAb treatment significantly decreased the expression of phospho-mTOR at Ser2448 in LL97A fibroblasts (P < 0.05; Figure 6A). These data indicated that the anti-GPC1 mAb treatment significantly decreased the activities of FAK, Akt, mTOR and RSK in LL97A lung fibroblasts, suggesting that the antiproliferative effect of anti-GPC1 mAb on LL97A lung fibroblasts is attributable to its inhibitory effect on the FAK/Src, Akt/mTOR, MEK/ERK/RSK pathways (Figure 1C and Figure 6A). The significant decrease in vimentin protein expression in anti-GPC1 treated LL97A cells may be associated with the suppression of fibroblasts proliferation and activity in vitro (Figure 6A). When LL97A lung fibroblasts were co-cultured indirectly with H460 lung tumor cells, anti-GPC1 mAb treatment significantly decreased the phosphorylation of Src at Tyr416 and β-catenin protein expression as compared with both of the vehicle control and mouse IgG isotype control treatment (P < 0.01 for all; Figure 5 and Figure 6B). Moreover, the expression of phospho-Akt at Ser473 in anti-GPC1 mAb treated LL97A fibroblasts was significantly lower than that in the vehicle control-treated LL97A fibroblasts (P < 0.05), while the expression of phospho-MEK at Ser217/221 in anti-GPC1 mAb treated LL97A fibroblasts was significantly lower than that in the mouse IgG isotype-treated LL97A fibroblasts (P < 0.05; Figure 6B). When LL97A lung fibroblasts were co-cultured indirectly with A549 cells, the anti-GPC1 mAb treatment significantly decreased the expression of phospho-Src (Tyr416; P < 0.01 for both), phospho-Akt (Ser473; P < 0.05 for both) and β-catenin protein (P < 0.01 for both) as compared with both of the vehicle control and mouse IgG isotype treatment (Figure 5 and Figure 6C). In addition, in the co-cultured A549 cells, the anti-GPC1 mAb treatment significantly decreased the RSK phosphorylation at Ser380 (P < 0.05 compared with both of the vehicle and mouse IgG isotype controls) and MEK phosphorylation at Ser217/221 (P < 0.05 compared with the vehicle control; Figure 6D). Taken together, results of the Western blot analysis of the in vitro monoculture and co-culture samples suggest that lung fibroblasts are more sensitive to anti-GPC1 mAb than lung tumor cells and the interaction between NSCLC cells and lung fibroblasts influences the molecular response to anti-GPC1 mAb treatment.

Given the potential impact of anti-GPC1 mAb on the interaction between tumor cells and tumor associated fibroblasts observed in the in vitro model, the anti-tumor activity of anti-GPC1 mAb was evaluated and associated mechanisms were explored using an orthotopic A549 tumor xenograft model. In the initial study, the athymic nude mice bearing established orthotopic A549 lung tumors were given tail-vein injections of vehicle (saline) control, or 10 mg/kg mouse IgG isotype, or 10 mg/kg of anti-GPC1 mAb once every 72 h for a total of 8 doses. The treatment was started once the progression of orthotopic lung tumor growth was confirmed by the in vivo BLI, which was 4–6 weeks after the intrapulmonary inoculation of luciferase expressing A549 tumor cells. BLI was conducted once a week over the 24 days of treatment period (Figure 7A). The relative rate of lung tumor growth in the anti-GPC1 mAb group, which was calculated as the slope of the curve on the semi-logarithmic plot of BLI total signal intensity over time, was lower than those in the vehicle control and mouse IgG isotype groups. Since the lung tumor growth rate data did not follow a normal distribution, a nonparametric Kruskal-Wallis test was used to test the difference between study groups. The result showed that the difference was not statistically significant (P > 0.05; Figure 7B). Likewise, the difference in total lung weights that reflect the lung tumor burden was not statistically significant among the three study groups (P > 0.05; Figure 7C). In the second study, athymic nude mice were given the i.p. injections of anti-GPC1 mAb at 1 mg/kg, 10 mg/kg and 50 mg/kg once a week for 3 doses and then once every 10 days for 2 more doses. The treatment was initiated 2 weeks after the orthotopic inoculation of A549 cells for early blockage of GPC1 activity. The in vivo BLI performed a week after the last dose showed that bioluminescence signals were detected in 6 out of 7 (6/7), 6/8, 4/8, and 6/7 animals in the vehicle control, 1 mg/kg, 10 mg/kg and 50 mg/kg anti-GPC1mAb groups, respectively (Figure 8A). Tumor nodules were found in the lungs of two more animals in the 10 mg/kg anti-GPC1 mAb group during the macroscopic inspection. Animals that failed to develop macroscopic tumors in the lungs were excluded from further analyses. Results of the ELISA determination of the anti-GPC1 mAb levels in plasma on the final day of the experiment showed that the anti-GPC1 mAb plasma levels were strongly correlated with the administered dose (Pearson’s correlation coefficient = 0.9986, P < 0.01; Figure 8B). Tumor-bearing mice treated with 50 mg/kg anti-GPC1 mAb had 43-, 17- and 2-fold increase in the mean anti-GPC1 mAb plasma levels as compared with the control, 1 mg/kg and 10 mg/kg anti-GPC1 mAb groups, respectively (P < 0.01 for all; Figure 8B). The mean anti-GPC1 mAb level in the 10 mg/kg group was significantly higher than that in the control group (11-fold increase, P < 0.05; Figure 8B). These data indicated that i.p. administration of 5 doses of anti-GPC1 mAb at 10 mg/kg and 50 mg/kg over 41 days elicited significantly higher mAb plasma levels as compared with the vehicle control. With regard to the effect of anti-GPC1 mAb treatment on tumor burden, although the mean total lung weights in the 10 mg/kg and 50 mg/kg anti-GPC1 mAb groups were decreased by 28% and 14% as compared with those of the control group, the differences were not statistically significant (P > 0.05. Figure 8C).

The effect of in vivo anti-GPC1 mAb treatment on signaling pathways associated with aberrant activation of growth factor receptors, including the MEK/ERK/RSK, Akt/GSK3/mTOR and Src/FAK pathways, as well as the protein expression of selected EMT markers, was evaluated in A549 lung tumors and adjacent noncancerous lung tissues using Western blot analysis (Figure 9, Figure 10, Figure 11, and Figure 12). The i.v. administration of 10 mg/kg anti-GPC1 mAb once every 72 h for a total of 8 doses did not have any significant effect on the phosphorylation of MEK, ERK, RSK, Akt, GSK3α/β, mTOR, Src and FAK, nor did it change significantly the expression of E-cadherin and β-catenin protein in orthotopic A549 lung tumors and adjacent normal lung tissues (Figure 9 and Figure 13).

In contrast, the i.p. administration of anti-GPC1 mAb at 10 and 50 mg/kg once a week for 3 doses and then once every 10 days for 2 more doses appeared to attenuate the FGF signal transduction pathway in the orthotopic A549 NSCLC xenografts (Figure 10, Figure 11 and Figure 12). Treatment with 10 and 50 mg/kg anti-GPC1 mAb significantly inhibited the FGFR1 phosphorylation at Tyr766 (P < 0.05 for both treatment groups; Figure 11 and Figure 14A), as well as the downstream ERK phosphorylation at Thr202/Tyr204 and RSK phosphorylation at Ser380 in tumors, as compared with the control group (P < 0.01 for all; Figure 10 and Figure 14B). The relative expression levels of phospho-GSK3α (Ser21; P < 0.05 for 1 mg/kg, and P < 0.01 for 10 mg/kg and 50 mg/kg groups), phospho-GSK3β (Ser9; P < 0.01 for all) and phospho-Src (Tyr416; P < 0.05 for 1 mg/kg, and P < 0.01 for 10 mg/kg and 50 mg/kg groups) were significantly decreased in all anti-GPC1 mAb treated tumors as compared with those in the control group (Figure 10 and Figure 14B). Moreover, comparison among the treatment groups showed that the expression levels of phospho-Src (Tyr416) in A549 tumors treated with 10 mg/kg and 50 mg/kg anti-GPC1 mAb were significantly lower than those treated with 1 mg/kg anti-GPC1 mAb (P < 0.05 for both 10 mg/kg and 50 mg/kg groups; Figure 14B). The phospho-ERK (Thr202/Tyr204) expression in the 10 mg/kg group was significantly lower than that in the 1 mg/kg group (P < 0.05), while the phospho-MEK (Ser380) expression in the 10 mg/kg group was significantly lower than those in the 1 mg/kg (P < 0.01) and 50 mg/kg (P < 0.05) groups (Figure 14B). It was noted that treatment with anti-GPC1 mAb at 10 mg/kg and 50 mg/kg resulted in a significant decrease in the intratumoral E-cadherin protein expression as compared with the vehicle treatment (Figure 14C), raising the concern that anti-GPC1 mAb treatment might elicit EMT in tumors.

In adjacent noncancerous lung tissues, anti-GPC1 mAb had no effect on the phospho-FGFR1 (Tyr766) expression (P > 0.05; Figure 12 and Figure 14D). Anti-GPC1 mAb treatment at 10 mg/kg and 50 mg/kg significantly decreased the phospho-Src (Tyr416) expression as compared with the vehicle control (P < 0.05 and P < 0.01 for 10 mg/kg and 50 mg/kg anti-GPC1 mAb treatment, respectively) and the 1 mg/kg anti-GPC1 mAb treatment group (P < 0.05 and P < 0.01 for 10 mg/kg and 50 mg/kg anti-GPC1 mAb groups, respectively; Figure 10 and Figure 14E). Moreover, the phospho-p90RSK (Ser380) expression levels in the 10 mg/kg and 50 mg/kg anti-GPC1 mAb treatment groups were significantly lower than that in the 1 mg/kg group (P < 0.01 for both; Figure 14E). A significant difference in the phospho-p90RSK (Ser380) expression level in the noncancerous lung tissues was found between control and 10 mg/kg anti-GPC1 mAb groups (P < 0.05), while a significant difference in the phospho-FAK (Tyr397) expression levels was found between control and 10 mg/kg anti-GPC1 mAb treatment groups (P < 0.05; Figure 14E). It was noted that the protein expression of GPC1 was detectable using the anti-GPC1 mAb, which is developed to target human GPC1, suggesting that the anti-GPC1 mAb is cross-reactive with the mouse GPC1. Treatment with the anti-GPC1 mAb at all three dose levels significantly decreased the vimentin protein expression levels in the adjacent noncancerous lung tissues (P < 0.05 for all; Figure 14F), implicating that certain types of cells in normal lung tissue, such as lung fibroblasts, may be sensitive to anti-GPC1 mAb treatment.