Glypican-1-targeted antibody–drug conjugate inhibits the growth of glypican-1-positive glioblastoma

Highlights

• •Humanized anti-glypican-1 antibody was conjugated with monomethyl auristatin E.
• •Elevated glypican-1 expression was shown in glioblastoma by immunohistochemistry.
• •GPC1-ADC bound to GPC1 was efficiently internalized in glioblastoma cell lines.
• •An orthotopic xenograft was established by intracranial implantation of KS-1-Luc.
• •Intravenous administration of GPC1-ADC showed potent intracranial activity.

Abstract

Glioblastoma is the deadliest form of brain tumor. The presence of the blood–brain barrier (BBB) significantly hinders chemotherapy, necessitating the development of innovative treatment options for this tumor. This report presents the in vitro and in vivo efficacy of an antibody–drug conjugate (ADC) that targets glypican-1 (GPC1) in glioblastoma. The GPC1-ADC was created by conjugating a humanized anti-GPC1 antibody (clone T2) with monomethyl auristatin E (MMAE) via maleimidocaproyl-valine-citrulline-p-aminobenzyloxycarbonyl linkers. Immunohistochemical staining analysis of a glioblastoma tissue microarray revealed that GPC1 expression was elevated in more than half of the cases. GPC1-ADC, when bound to GPC1, was efficiently and rapidly internalized in glioblastoma cell lines. It inhibited the growth of GPC1-positive glioma cell lines by inducing cell cycle arrest in the G2/M phase and triggering apoptosis in vitro. We established a heterotopic xenograft model by subcutaneously implanting KALS-1 and administered GPC1-ADC intravenously. GPC1-ADC significantly inhibited tumor growth and increased the number of mitotic cells. We also established an orthotopic xenograft model by intracranially implanting luciferase-transfected KS-1-Luc#19. After injecting Evans blue and resecting brain tissues, dye leakage was observed in the implantation area, confirming BBB disruption. We administered GPC1-ADC intravenously and measured the luciferase activity using an in vivo imaging system. GPC1-ADC significantly inhibited tumor growth and extended survival. In conclusion, GPC1-ADC demonstrated potent intracranial activity against GPC1-positive glioblastoma in an orthotopic xenograft model. These results indicate that GPC1-ADC could represent a groundbreaking new therapy for treating glioblastoma beyond the BBB.

Introduction

Glioblastoma is the most prevalent and lethal primary malignant brain tumor [1]. Despite first-line treatments, which include surgical resection followed by radiotherapy and chemotherapy, glioblastoma is associated with a median overall survival (OS) duration of only 1.5 years, indicating a persistently poor prognosis [2,3]. The blood–brain barrier (BBB) significantly restricts the delivery of substances to the brain to protect it. However, the BBB also poses a major barrier to chemotherapy for brain tumors. Many pharmacological interventions have failed to alter the course of glioblastoma [4]. Temozolomide, small enough to penetrate the BBB, remains a key drug, but it has not yielded satisfactory OS [2]. Consequently, there is an urgent need to develop novel treatment options for glioblastoma.

Antibody–drug conjugates (ADCs) are a type of targeted cancer therapy that uses immunoconjugates, in which anticancer drugs are chemically and enzymatically bound to monoclonal antibodies (mAbs). ADCs can deliver a highly cytotoxic payload to cancer cells expressing specific antigens, potentially achieving a broad therapeutic range while simultaneously reducing systemic adverse effects. ADCs have seen rapid development in recent decades, with the FDA approving brentuximab vedotin for the treatment of CD30-positive lymphoma in 2011 [5,6], followed by the approval of trastuzumab emtansine for HER2-positive breast cancer, marking the first approval for solid tumors in 2013 [7]. In the context of brain tumors, phase II trials demonstrated that trastuzumab deruxtecan was also effective for brain metastases from HER2-positive breast cancer in 2022 [8].

Our group recently identified glypican-1 (GPC1) as a novel cancer antigen in esophageal squamous cell carcinoma (ESCC) through a quantitative plasma membrane proteomic analysis [9]. We have since been developing an ADC that targets GPC1. GPC1 is a heparan sulfate proteoglycan that attaches to the plasma membrane via a glycosylphosphatidylinositol anchor [10,11]. It has been reported to promote tumor growth, metastasis, and invasion by acting as a coreceptor of heparin-binding growth factors, thereby enhancing various signaling pathways such as Wnt, Hedgehog [12], hepatocyte growth factor, and fibroblast growth factor-2 [13]. Elevated GPC1 expression has been observed in ESCC [9], pancreatic cancer [14], cholangiocarcinoma [15], uterine cervical cancer [16], and other cancers [13,17,18]. This elevated GPC1 expression has also been correlated with poor prognosis in several types of these cancers [9,14,15,18]. However, GPC1 expression in normal tissue is primarily restricted to the testis or ovary [9]. We developed GPC1-ADC by conjugating a humanized anti-GPC1 antibody (clone T2) with monomethyl auristatin E (MMAE) via maleimidocaproyl-valine-citrulline-p-aminobenzyloxycarbonyl (mc-vc-PABC) linkers. We have reported that GPC1-ADC exhibited potent tumor growth inhibition in GPC1-positive pancreatic cancer and ESCC patient-derived xenograft (PDX) models due to its bystander killing activity [19]. Recent studies have shown that GPC1 was highly expressed in half of glioblastoma patients [18] and GPC1 was identified as a promising therapeutic target for glioblastoma [20]. However, no study has examined the antitumor activity of ADCs targeting GPC1 against glioblastoma.

In this report, we present the first results of the antitumor efficacy of ADC targeting GPC1 for glioblastoma from in vitro and in vivo studies. The aim of the present study was to determine whether intravenous administration of GPC1-ADC inhibits the growth of GPC1-positive glioblastoma in an intracranial orthotopic xenograft model.

Materials and methods

Generation of anti-GPC1 monoclonal antibodies for immunohistochemistry

For immunohistochemical (IHC) studies, we generated a monoclonal antibody (mAb; PPY7462) against human GPC1 using conventional mouse hybridoma technology. This antibody was validated for its specificity to GPC1 on formalin-fixed, paraffin-embedded sections of xenograft tumor tissues, including BxPC3 (GPC1 positive) and BxPC3-GKO (GPC1 negative). Detailed methodologies are provided in the online Supplementary Material.

Tissue microarray and immunohistochemistry

A tissue microarray was procured from US Biomax (GL806g, U.S. Biomax human brain cancer tissue microarray). This array comprised formalin-fixed, paraffin-embedded tissue, including glioblastoma tissue from 35 patients and 5 samples of healthy brain tissue. As previously described [9], the tissue was deparaffinized with xylene and rehydrated through a series of graded alcohol solutions (70 %, 80 %, 90 %, and 100 %). IHC staining for GPC1 was conducted using a mouse anti-GPC1 monoclonal antibody (clone PPY7462, 0.08 μg/ml), and visualized using Envision ChemMate (Dako, Glostrup, Denmark) as per the manufacturer's instructions. Images of the sections were captured using a fluorescence microscope (BZ-X700, Keyence, Osaka, Japan).

Immunostaining was evaluated based on the intensity of the staining, with the following scoring system: 0 denoting no or weak staining 1 representing normal staining and 2 indicating strong staining. The staining density, termed the positivity score, was categorized as follows: 1 denoting less than 50 % positivity and 2 indicating more than 50 % positivity. The final IHC score was calculated by multiplying the intensity score by the positivity score, yielding a maximum possible score of 4. These results were designated as the GPC1 score. Scores ≥2 were classified as high-GPC1 expression, while scores <2 were considered low-GPC1 expression, in accordance with previous reports.

Cell lines and culture

The A172 human glioma cell line (RRID: CVCL_0131) was acquired from the American Type Culture Collection in Rockville, Maryland. Additionally, four human glioma cell lines, namely KNS42 (RRID: CVCL_0378), U-251-MG (RRID: CVCL_0021), KALS-1 (RRID: CVCL_1323), and KS-1 (RRID: CVCL_1343) were sourced from the Japanese Collection of Research Bioresources in Osaka, Japan. Furthermore, the human pancreatic cancer cell line BxPC-3 (RRID: CVCL_0186) was obtained from the European Collection of Authenticated Cell Cultures in Salisbury, UK. BxPC-3-GKO was established as described previously [19].

The A172, KNS42, and U-251-MG cell lines were cultured in Dulbecco's modified eagle medium (DMEM) with high glucose supplemented with 10 % fetal bovine serum (FBS; Thermo Fisher Scientific Inc., Waltham, MA, USA), 100 U/ml penicillin, and 100 μg/ml streptomycin. KALS-1 was cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10 % FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. KS-1 was cultured in DMED with low glucose supplemented with 10 % FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. BxPC-3 and BxPC-3-GKO were cultured in RPMI 1640 medium supplemented with 10 % FBS, 1 % GlutaMAX (Thermo Fisher Scientific), 100 U/ml penicillin, and 100 μg/ml streptomycin. All cultures were maintained at 37°C in a humidified atmosphere with 5 % CO₂. Cells lines were routinely tested for Mycoplasma using a MycoAlert Mycoplasma Detection Kit (Lonza) and were utilized within 3 months after resuscitation.

Establishment of the luciferase stably expressing cells

KS-1 cells, which stably express luciferase, were generated through a process of transfection using pcDNA3.1-Luc and selected with 700 μg/mL of geneticin (Invitrogen). The resulting clones were then maintained in DMEM with low glucose, supplemented with 10 % FBS and 250 μg/mL of geneticin to ensure stable expression. This process led to the establishment of luciferase-expressing cell lines, which were subsequently designated as KS-1-Luc#19 cells.

Generation of humanized anti-GPC1 antibodies and humanized GPC1-ADC for in vitro study

The production of the humanized anti-GPC1 antibody (clone T2) has been detailed in a prior publication [19]. Similarly, the synthesis of the humanized GPC1-ADC, intended for conjugation with the humanized anti-GPC1 antibodies (clone T2) and monomethyl auristatin E (MMAE), was also previously documented [19].

In vivo efficacy study in glioblastoma cell line subcutaneous heterotopic xenograft model

In this study, the assessment of the efficacy and safety of the drug in animal models followed the protocol outlined in a previous report [16]. All animal experiments were conducted in accordance with the institutional ethical guidelines for animal experimentation of Iwate Medical University (Approval number: 03-004). Healthy female CB17/severe combined immunodeficient (SCID) mice, aged 6 weeks, were procured from The Jackson Laboratory Japan Inc. (Yokohama, Japan). The mice were housed in a pathogen-free facility at the Center for In Vivo Science, Iwate Medical University (Yahaba, Japan). They were kept in a temperature-controlled room with 12-h light/12-h dark cycle and provided with free access to water. For heterotopic xenograft experiments, mice were anesthetized with 3 % isoflurane and subcutaneously injected in the flank with 5 × 10⁶ KALS-1 cells in 100 μL of a 1:1 (v/v) PBS: Matrigel solution (BD Biosciences, Franklin Lakes, NJ, USA). Once tumor volumes exceeded 75 mm³, mice were randomly divided into four groups (7 mice per group). Subsequent to randomization, PBS or GPC1-ADC (1 mg/kg, 3 mg/kg, 10 mg/kg) was administered via the tail veins twice a week, until a total of four doses had been administered. Tumor sizes were measured twice a week using a vernier caliper. Tumor volumes were calculated as W² × L/2, where W represents the width (smaller dimension) and L represents the length (larger dimension). The body weights of mice were also recorded twice a week. Mice were anesthetized with 3 % isoflurane and euthanized via cervical dislocation 42 days after the initiation of treatment. Subsequently, tumors were resected and weighed.

To examine the pharmacological effects of GPC1-ADC at the cellular level, animals bearing KALS-1 tumor xenografts were administered either PBS or GPC1-ADC (at doses of 1 mg/kg, 3 mg/kg, or 10 mg/kg). Tumors were harvested 24 h post-injection. These tumors were then fixed in formalin, embedded in paraffin, and sectioned into 4 μm slices. IHC was performed using an anti-phospho-histone H3 (Ser10) antibody (#9701, 1:400, Cell Signaling Technologies, Danvers, MA, USA). To verify the expression of GPC1 in tumors, a tumor was harvested following the subcutaneous transplantation of KALS-1, without any drug administration. IHC staining of GPC1 was carried out using a rabbit polyclonal anti-GPC1 antibody (1:2,000, catalog No. GTX104557; GeneTex, San Antonio, TX, USA).

In vivo intracranial activity study in glioblastoma cell line intracranial orthotopic xenograft model

Luciferase-transfected KS-1-Luc#19 glioblastoma cells were implanted into the right cerebral hemisphere of 6-week-old female NOD/Shi-scid, IL-2RgKO (NOG) mice. For the procedure [21,22], the mice were anesthetized with 3 % isoflurane. A 1 cm midline incision was made in the skin and a burr hole was created using a 25G needle. Subsequently, 5 × 10⁵ KS-1-Luc#19 cells in 5 μl PBS were injected into the right brain (2.0 mm lateral and 1.0 mm anterior to the bregma, 3.5 mm depth) using a 26G Hamilton syringe. The luciferase activity of each mouse was measured weekly using the IVIS Lumina (Caliper Life Sciences) and analyzed with the Living Image 4.2 software (Caliper Life Sciences), as previously described [23]. When the luciferase activity reached 1.5 × 10⁶ (photons/sec) after 8 weeks of implantation, the tumor-bearing mice were randomized into two groups (n = 10 per group) based on luciferase activity, and dosing was initiated (day 0). PBS or GPC1-ADC (10 mg/kg) was administered via the tail vein to the mice twice a week for a total of four times. After grouping, the luciferase activity continued to be measured once a week. When the mice lost 20 % of their body weight, some mice were anesthetized with 3 % isoflurane, and transcardial perfusion was performed, followed by brain resection. Other mice were euthanized via cervical dislocation after anesthesia.

Assessment of BBB permeability

BBB disruption was evaluated using Evans blue. Evans blue (2 % in saline, Sigma-Aldrich) was administered at a dose of 4 mL/kg body weight through the tail vein, following established protocols [24]. After 30 min, the mice were subjected to transcardial perfusion with PBS followed by 10 % formalin, over a 30-minute period under anesthesia. Subsequently, the brains were isolated and fixed with 10 % formalin for a day. These fixed brains were then cut into 3-mm axial slices, further fixed with 10 % formalin for an additional day, and ultimately preserved in 70 % ethanol.

Immunohistochemistry in intracranial tumor

To verify the expression of GPC1 in brain-formed tumors, resected mouse brains were fixed with 10 % formalin for 2 days. Subsequently, they were preserved in 70 % ethanol for over a week, followed by a degreasing process. The brain samples were then embedded in paraffin and sectioned into 4 μm slices. IHC staining of GPC1 was performed in a manner similar to previous procedures.

Statistical analyses

Statistical analyses were conducted using GraphPad Prism 10 (GraphPad Software Inc., San Diego, CA, USA) or BellCurve for Excel 4.04 (Social Survey Research Information Co., Ltd.). Data from in vitro experiments are presented as mean ± SD, while data from in vivo experiments are shown as mean ± SEM. For comparisons between two groups, the Student's t-test was utilized. For comparisons among three or more groups, a one-way analysis of variance was performed, followed by either Tukey's honestly significant difference test or Dunnett's test. OS was evaluated using the Kaplan–Meier method and assessed by the generalized Wilcoxon test. A p-values less than 0.05 were considered statistically significant.

Other experiments are described in the Supporting Information Methods.

Results

Glioblastomas showed high expression of GPC1 in the tissue microarray

To evaluate the expression of GPC1 in glioblastoma, we conducted IHC staining for GPC1 using a glioblastoma tissue microarray. The 35 glioblastoma cases were categorized into two groups based on their GPC1-staining scores. Twenty-two cases (62.9 %) scored more than 2 points and were classified as our high-expression group (HG), as shown in Fig. 1A (right panel). The remaining 13 cases (37.1 %) scored less than 1 point and were classified as the low-expression group (LG), as depicted in Fig. 1A (middle panel). All five normal cerebrum samples exhibited low-expression, scoring less than 1 point (Fig. 1A (left panel)). The distribution of GPC1 expression is illustrated in Fig. 1B.

Fig. 1.

GPC1 expression in clinical glioblastomas and normal cerebrums in tissue microarray. (A) Representative images of immunohistochemistry (IHC) staining for GPC1 in a tissue microarray. It shows the expression of GPC1 in the membranes of glioblastomas and normal brain tissue. The left panel displays the normal brain tissue, the middle panel shows the glioblastoma tissue from the low-expression group (LG), and the right panel illustrates the glioblastoma tissue from the high-expression group (HG). The scale bar represents 100 μm. (B) The distribution of GPC1 scores in the LG and HG.

Human glioma cell lines expressed GPC1 protein

Flow cytometry was utilized to measure the presence of GPC1 protein on the surfaces of glioma cells, using a humanized anti-GPC1 antibody (clone T2). High GPC1 expression was observed in KNS42, U-251-MG, and KALS-1, while low expression was noted in A172 and KS-1. The GPC1 expression of the luciferase-transfected glioma cell line KS-1-Luc#19, which we established, was comparable to that of the parental KS-1 cell line (Fig. 2A). GPC1 expression on the plasma membrane was quantified using an indirect immunofluorescence assay. High levels of expression were observed in A172 (225,521 sites cell-1), KNS42 (132,787 sites cell-1), U-251-MG (223,176 sites cell-1), and KALS-1 (155,353 sites cell-1). In contrast, low levels of expression were detected in KS-1 (35,634 sites cell-1) and KS-1-Luc#19 (30,507 sites cell-1) (Table 1).

Fig. 2.

GPC-1 expression in glioma cell lines and characteristics of GPC-1 protein. (A) Flow cytometric analysis of GPC1 expression in five glioma cell lines and a luciferase-transfected glioma cell line. The red lines histogram profile represents the isotype control, while the blue lines histogram illustrates the staining results of the humanized anti-GPC1 antibody (clone T2). (B) Western blot analysis of GPC1 expression. BxPC3 serves as the positive control, and BxPC3-GKO is the negative control. The protein level of GPC1 was detected by western blot, with β-actin used as a loading control. (C) Time-course analysis of the internalization activity of the GPC1-ADC in KALS-1 and KS-1 cells.

Table 1.

Expression levels GPC1 and the IC₅₀ values of GPC1-ADC and MMAE in various glioma cell lines.

Cell Lines	GPC1 expression (ABC/cell)	GPC1-ADC (nM)	Control-ADC (nM)	MMAE (nM)
A172	225,521	0.992	N.D.	0.062
KNS42	132,787	2.989	N.D.	0.507
U-251-MG	223,176	0.200	N.D.	0.171
KALS-1	155,353	5.787	N.D.	0.081
KS-1	35,634	0.128	N.D.	0.032
KS-1-Luc#19	30,507	0.548	N.D.	0.052

ABC, antibody-binding capacity; ADC, antibody–drug conjugate; GPC1; glypican-1; MMAE, monomethyl auristatin E; N.D., none detected.

Western blot analysis was also conducted to confirm GPC1 protein expression. In addition to the six glioma cell lines, the GPC1-positive pancreatic cancer cell line BxPC3 was analyzed as a positive control, and the GPC1-knockout pancreatic cell line BxPC3-GKO served as a negative control. GPC1 expression was confirmed in all glioma cell lines (Fig. 2B).

GPC1-ADC bound to GPC1 was internalized in glioma cell lines

The binding capacity and internalization percentage of GPC1-ADC were evaluated using flow cytometry in KALS-1 and KS-1 cells. After exposure to GPC1-ADC, the amount of GPC1 remaining on the cell surface was measured using flow cytometry. This was achieved using a biotin-labeled anti-GPC1 mAb (clone 02b006), which recognizes an epitope distinct from that bound by clone T2. The internalization of GPC1-ADC occurred rapidly in both KALS-1 and KS-1 cells (Fig. 2C).

GPC1-ADC inhibited the proliferation of GPC1-expressing glioma cells

GPC1-ADC induced a dose-dependent decrease in cell viability in GPC1-positive glioma cell lines in vitro (Fig. 3A). No significant cytotoxic activity was observed in any of the cell lines treated with a control-ADC. The IC₅₀ values for GPC1-ADC in GPC1-positive cell lines ranged from 0.128 to 5.787 nM (Table 1). The IC₅₀ values for MMAE against the cell lines ranged from 0.032 to 0.507 nM (Table 1).

Fig. 3.

In vitro cell growth inhibitory activity of antibody–drug conjugates (ADC). (A) Cell viability following treatment with GPC1-ADC or IgG4-ADC (control-ADC [MMAE]). The cell lines A172, U-251-MG, KALS-1, KS-1, and KS-1-Luc#19 were measured 144 h post-drug administration. The KNS42 was measured 192 h post-drug administration. (B) Induction of G2/M phase cell cycle arrest in KNS42, U-251-MG, KALS-1, and KS-1 cells treated with GPC1-ADC. KNS42 was treated with either 32 nM control-ADC or 32 nM GPC1-ADC. U-251-MG, KALS-1, and KS-1 cells were treated with either 16 nM control-ADC or 16 nM GPC1-ADC. After 24 h, cell cycle analysis for KNS42, U-251-MG, and KS-1 was performed by flow cytometry with propidium iodide DNA staining. For KALS-1, the analysis was conducted 48 h later. Statistical significance is indicated by *p < 0.05 and **p < 0.01, as determined by one-way ANOVA, followed by Dunnett's post-hoc test. (C) Induction of apoptosis in A172, KNS42, U-251-MG, KALS-1, and KS-1 cells treated with control-ADC or GPC1-ADC. Caspase 3/7 activation was monitored using the Caspase Glo 3/7 assay system at 48 h after the addition of ADCs. Statistical significance is indicated by *p < 0.05 and **p < 0.01, as determined by Student's t test.

We also investigated the effects of GPC1-ADC on the cell cycle and apoptosis. GPC1-ADC significantly increased the proportion of cells in the G2/M phase, whereas the proportion of cells in each cycle remained unchanged with control-ADC (Fig. 3B). Additionally, GPC1-ADC triggered a dose-dependent increase in caspase 3/7 activity compared to control-ADC (Fig. 3C). These data suggest that GPC1-ADC induces cell cycle arrest in the G2/M phase and promotes apoptosis via a caspase3/7-dependent pathway.

In vivo efficacy study showed that GPC1-ADC inhibited tumor growth in KALS-1 subcutaneous heterotopic xenograft model

To evaluate the antitumor efficacy of GPC1-ADC, we established a xenograft mouse model by subcutaneously implanting the KALS-1 glioblastoma cell line. Over a period of 8 to 13 weeks post-implantation, tumor formation was observed in SCID mice. The expression of GPC1 in the KALS-1 xenograft tumor was confirmed via IHC (Fig. 4A). Mice implanted with KALS-1 were segregated into four groups, each consisting of seven mice. Mice in each group were intravenously administered with either PBS or GPC1-ADC at doses of 1 mg/kg, 3 mg/kg, or 10 mg/kg twice a week, for a total of four administrations. Compared to PBS, GPC1-ADC significantly inhibited tumor growth (p < 0.01; Fig. 4A). No significant weight loss was observed across any of the groups (Fig. 4B). These findings suggest that GPC1-ADC effectively inhibits tumor growth in GPC1-positive glioblastoma, assuming the BBB is not a factor.

Fig. 4.

In vivo antitumor efficacy of GPC1-ADC in KALS-1 subcutaneous heterotopic xenograft model. (A) Antitumor efficacy of GPC1-ADC in the KALS-1 subcutaneous heterotopic xenograft model (n = 7 per group). Each point on the graph represents the average tumor volume. The tumor volume in each group was analyzed 42 days post-drug administration. Representative images of Immunohistochemistry (IHC) staining for GPC1 in subcutaneous heterotopic xenograft tumor tissues derived from KALS-1. The scale bar: represents 40 μm. (B) Changes in relative body weight. (C) Induction of mitotic arrest caused by GPC1-ADC in vivo. Animals bearing the KALS-1 tumor xenografts were administered a single dose of either PBS or GPC1-ADC (1 mg/kg, 3 mg/kg, 10 mg/kg). After 24 h, the tumors were harvested and stained with an anti-phospho-histone H3 (Ser10) antibody to detect mitotic cells. The scale bar represents 40 μm. (D) Phospho-histone H3 (Ser10) staining as the ratio of mitotic cells to the total number of tumor cells in four groups (magnification, × 200). The statistical analysis was performed using One-way ANOVA, followed by Tukey's test.

To further investigate the in vivo pharmacological action of GPC1-ADC, IHC staining of KALS-1 tumors was performed using an anti-phosphorylated histone H3 (Ser10) antibody, a known marker of mitosis. Treatment with GPC1-ADC resulted in an increase in mitotic cells (Fig. 4C). The percentage of mitotic cells significantly increased in a dose-dependent manner in the group administered with GPC1-ADC compared to the control group (p < 0.01; Fig. 4D). These results suggest that the tubulin polymerization inhibitor, MMAE, was effectively delivered to KALS-1 tumor cells by the anti-GPC1 mAb, resulting in mitotic arrest.

In vivo intracranial activity study showed that GPC1-ADC inhibited tumor growth in KS-1-Luc#19 intracranial orthotopic xenograft model

In order to evaluate the intracranial antitumor activity of GPC1-ADC, and to account for the impact of the BBB on drug delivery, we established a xenograft mouse model. This was achieved by intracranially implanting glioblastoma cells from the luciferase-transfected KS-1-Luc#19 cell line. Following the intracranial implantation of KS-1-Luc#19 cells, the formation of a mouse brain tumor expressing GPC1 was confirmed IHC staining when the mouse brain tissue was resected (Fig. 5A). Furthermore, when the mouse brain was resected after being intravenously injected with the vascular permeability marker, Evans blue, leakage of the Evans blue dye was observed in the region where the KS-1-Luc#19 cells were implanted (Fig. 5B). This observation suggests that the BBB within the tumor is compromised, leading to an increase in vascular permeability.

Fig. 5.

In vivo antitumor activity of GPC1-ADC in KS-1-Luc#19 intracranial orthotopic xenograft model. (A) Images of immunohistochemistry (IHC) staining for GPC1 in a non-implantation mouse brain and in brain tissue 130 days post-intracranial implantation of KS-1-Luc#19. The scale bar represents 1 mm. (B) Macroscopic imaging of Evans blue staining in a non-implantation mouse brain and in brain tissue intracranially implanted with KS-1-Luc#19. The arrow indicates staining of the dural mater, while the arrowhead indicates light staining of the brain tumor. The scale bar represents 1 mm. (C) Representative images showing the time course of luciferase activity. It compares the luciferase activity in the PBS and GPC1-ADC (10 mg/kg) groups before treatment and 28 days post-treatment. (D) Antitumor intracranial activity of GPC1-ADC in the KS-1-Luc#19 orthotopic xenograft model (n = 10 per group). Each point on the graph represents the average luciferase activity. The luciferase activity in the two groups was analyzed 28 days post-drug administration. The error bars denote the standard error of mean (SEM). (E) Changes in relative body weight. (F) Kaplan–Meier survival analysis of the KS-1-Luc#19 intracranial orthotopic xenograft model study. The generalized Wilcoxon test results indicate a significant difference (p < 0.01) between the GPC1-ADC and PBS groups.

The growth of tumors in intracranially implanted KS-1-Luc#19 cells was quantified based on luciferase activity. A total of thirty NOG mice were intracranially implanted with KS-1-Luc#19 cells, out of which 29 exhibited intracranial tumor formation over a period of 4 to 20 weeks. Out of these, twenty mice that showed tumor formation between 8 to 13 weeks were randomly divided into two groups, each consisting of ten mice. Mice in each group were intravenously administered either PBS or GPC1-ADC at a dose of 10 mg/kg twice a week, for a total of four administrations. The administration of GPC1-ADC resulted in a significant suppression of tumor growth compared to PBS (Fig. 5C–D). In contrast to GPC1-ADC, the administration of PBS led to significant weight loss (Fig. 5E). Moreover, the treatment with GPC1-ADC significantly extended the median survival of mice with intracranial KS-1-Luc#19 xenografts compared to those treated with PBS (from 39 to 53 days; p < 0.05, Fig. 5F). These findings suggest that GPC1-ADC demonstrates potent inhibitory effects on tumor growth in GPC1-positive glioblastoma, considering the influence of the BBB.

Discussion

Our study demonstrates that the intravenous administration of GPC1-ADC exhibits potent antitumor activity in an intracranial orthotopic xenograft model. These findings suggest that GPC1-ADC could potentially serve as a novel therapeutic approach for glioblastoma, beyond the disrupted BBB.

In the context of glioblastoma treatment, ADCs are anticipated to outperform conventional therapies. The BBB, which is composed of endothelial cells interconnected by tight junctions to pericytes and astrocytes, typically acts as a barrier against toxins and pathogens. However, this barrier also prevents many anticancer drugs, including numerous antibody drugs, from penetrating it. Increasing the dosage of anticancer drugs to overcome the BBB can lead to severe systemic side effects. While some drugs are administered directly into the tumor via methods such as convection-enhanced delivery, there are concerns that some drugs may decrease tolerability and increase the risk of infection. In contrast, ADCs, when administered intravenously and targeted at appropriate cancer antigens with conjugated suitable anticancer drugs, are expected to be effective against brain tumors. This is supported by data from phase II trials, which showed that HER-2 targeted ADCs, such as trastuzumab deruxtecan, elicited a high intracranial response in patients with brain metastases originating from breast cancer [8]. Given the effectiveness of ADCs against metastatic brain tumors beyond the BBB, it is anticipated that ADCs may also prove effective in treating glioblastoma.

In this study, we conducted a preclinical investigation of GPC1-ADC, which targets GPC1-positive glioblastoma. Our tissue microarray analysis revealed that approximately 60 % of glioblastomas exhibited high GPC1 expression (Fig. 1), a finding that aligns with the results obtained from patient specimens by Saito et al. [18]. GPC1-positive glioblastoma cells demonstrated a higher binding capacity and internalization efficiency for GPC1-ADC (Fig. 2C), suggesting that GPC1 could be a promising therapeutic target for glioblastoma. Our in vitro examination of GPC1-ADC's efficacy revealed its potent cell-killing ability against GPC1-expressing glioblastoma cells (Fig. 3A). KS-1 cells exhibited high sensitivity to GPC1-ADC, although their GPC1 expression levels were low (Table 1). This is likely attributed to their high sensitivity to MMAE and a rapid growth rate. The mechanism of action of GPC1-ADC involves cell cycle arrest in the G2/M phase and apoptosis via the caspase3/7-dependent pathway (Fig. 3B, C). These findings are consistent with the strong microtubule inhibiting properties of MMAE, the payload of GPC1-ADC, and suggest that GPC1-ADC selectively inhibits mitotic cells. In our in vivo study, which used a KALS-1 cell line subcutaneous heterotopic xenograft model to exclude the effects of the BBB, GPC1-ADC strongly inhibited tumor growth, with a mode of action similar to that observed in vitro (Fig. 4). These results indicate that GPC1-ADC exhibits potent antitumor efficacy when it reaches GPC1-positive glioblastoma. Interestingly, in our KS-1-Luc#19 intracranial orthotopic xenograft model study, GPC1-ADC significantly inhibited tumor growth and extended survival. The mechanism of intracranial activity is speculated to involve drug passage due to disrupted BBB (Fig. 5). This result suggests that a portion of the administered GPC1-ADC can penetrate the disrupted BBB and exhibit antitumor intracranial activity against glioblastoma.

While our data suggest promising preclinical efficacy of GPC1-ADC for glioblastoma, previous trials involving ADCs have demonstrated limited clinical efficacy against this condition [25]. For instance, despite extensive investigation of EGFR as a target molecule for ADCs against glioblastoma, EGFR-ADCs have proven ineffective in patients due to factors such as tumor size, payload-sensitizing mutations [26], and severe infusion-related reactions [27]. In contrast, the antitumor activity of GPC1-ADC in an intracranial orthotopic xenograft model can be attributed to two key factors. First, GPC1 appears to be a suitable therapeutic target for glioblastoma. Prior studies confirm that the BBB in glioblastoma is heterogeneously disrupted [28], allowing some ADCs to reach the glioblastoma [29]. The challenge lies in efficiently delivering the correct dosage of drugs. Our study demonstrated that GPC1-ADC was efficiently internalized into GPC1-positive cells (Fig. 2C), suggesting that GPC1 on tumor cells can internalize a small amount of GPC1-ADC that passes the BBB. The second factor is the bystander effect of the payload MMAE. Following the cleavage of the linker within the first internalized tumor cell, the released MMAE can enter nearby tumor cells [30], a phenomenon known as the “bystander killing effect.” This effect is particularly significant given the heterogeneity of BBB disruption [28] and the variability in drug concentrations within the tumor due to its size. In fact, an inverse correlation between tumor size and ADC accumulation has been reported [31]. We previously demonstrated that GPC1-ADC inhibited tumor growth in a heterogeneous GPC1-expressing pancreatic tumor model due to the bystander effect [19]. Therefore, the payload MMAE may diffuse within the intracranial tumor due to this bystander effect. In conclusion, our findings suggest that GPC1-ADC exhibits potent tumor growth inhibition due to the combined advantages of both the target molecule GPC1 and the payload MMAE.

Considering the attributes of both ADC and the cancer antigen GPC1, GPC1-ADC appears to be safe. GPC1-ADC is only taken up by cells that express the cancer-specific antigen GPC1. In this study, GPC1 was found to be minimally expressed in healthy human brain tissue (Fig. 1A). Our group has also previously demonstrated that a radioactively labeled humanized anti-GPC1 antibody (clone T2) rarely transfers to a normal mouse brain [32]. Furthermore, GPC1 was also infrequently expressed in normal tissue [33]. Even if GPC1-ADC does slightly transfer to normal tissue, it is only effective in mitotic cells. Therefore, GPC1-ADC will have minimal impact on normal tissue, including the brain. It is advisable to conduct safety studies for GPC1-ADC using animals, and toxicity assessments should be undertaken with meticulous care in future clinical studies.

This study, however, has several limitations. First, in the intracranial orthotopic xenograft model, only one cell line, KS-1-Luc#19, was examined. The effectiveness of GPC1-ADC might be due to this cell line's specific tendency to disrupt the BBB. It is necessary to investigate whether the same effect is observed in multiple glioblastoma cells, including patient-derived cells. Second, comparing antitumor efficacy with or without the BBB is challenging because different cell lines were used in the subcutaneous and intracranial models. The KALS-1 cell line was viable subcutaneously but not intracranially. Conversely, the KS-1 cell line was viable intracranially but not subcutaneously, indicating that KS-1 cells were proliferating intracranially, not extracranially. Tumor regrowth was observed in the orthotopic model with the same drug administration protocol as the heterotopic model. However, this result might reflect differences in the characteristics of cell lines. Simply increasing the number of drug administrations may inhibit the regrowth of the orthotopic model. Finally, a common question in orthotopic xenograft models is whether tumor formation by intracranial transplantation mimics spontaneous tumorigenesis, especially BBB invasion and destruction. Resolving this issue could reduce the discrepancy between the results of preclinical and clinical trials.

In conclusion, our results showed that intravenously administered GPC1-ADC exhibited potent intracranial activity against GPC1-positive glioblastoma in an orthotopic xenograft model. This finding suggests that GPC1-ADC could be a breakthrough novel therapy for treating glioblastoma beyond the BBB. Further studies are required to confirm whether GPC1-ADC is effective in PDX model experiments and clinical trials.

Consent for publication

Not applicable.

Data availability

All data generated and analyzed during this study are included in this published article and its supplementary information files.

CRediT authorship contribution statement

Shun Uchida: Methodology, Formal analysis, Investigation, Writing – original draft, Visualization. Satoshi Serada: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – review & editing, Visualization, Supervision, Project administration, Funding acquisition. Yuji Suzuki: Methodology, Investigation. Eiji Funajima: Investigation. Kei Kitakami: Investigation. Kazumasa Dobashi: Investigation. Satomi Tamatani: Investigation. Yuichi Sato: Investigation. Takaaki Beppu: Investigation. Kuniaki Ogasawara: Investigation. Testuji Naka: Conceptualization, Writing – review & editing, Supervision, Project administration, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.