A Novel Targeting Therapy of Malignant Mesothelioma Using Anti-Podoplanin Antibody

Malignant pleural mesothelioma (MPM) is an aggressive malignancy arising from the mesothelial cells lining the pleura (1). Previously, MPM was considered to be rare; however, its incidence is increasing worldwide. According to various studies, the number of cases is estimated to reach a peak between 2010 and 2020 in Europe (2) and ∼2030–2040 in Asia (3). Despite the combined therapies including surgery, radiotherapy, and chemotherapy, the prognosis of MPM is very poor: median survival from the time of diagnosis is 4–12 mo, and the 5-y survival rate is <5% (4–6). Therefore, novel therapeutic strategies for MPM are urgently necessary for improving the prognosis.

On the basis of the current status of MPM, immunotherapy has also been considered as one of the novel therapeutic approaches. Recently, tumor-specific immunotherapy using vaccination with antigenic peptides of tumor-associated Ags (TAAs) with or without dendritic cells or Abs has been conducted because several TAAs specific for MPM were reported, although their number is insufficient (7–9). Among them, immunotherapy with specific Abs seems to be the most promising approach, whereas clinical trials with humanized Abs for mesothelin as well as CD26 are still in phase I (10).

Podoplanin (Aggrus), which is a type I transmembrane sialomucin-like glycoprotein, induces platelet aggregation (11–14). Originally, podoplanin was detected on the surface of podocytes (15). Although podoplanin is reported to be expressed in limited normal tissues such as endothelium of lymphatic vessels and in type I alveolar epithelium (16, 17), various tumors such as squamous cell carcinomas, testicular seminomas, malignant brain tumors, fibrosarcomas, and malignant mesotheliomas (MMs) including MPM are known to show its overexpression (18–22). Furthermore, it has been reported that podoplanin is associated with cell migration (23, 24), epithelial–mesenchymal transition (25), and tumor metastasis (26, 27). Moreover, podoplanin serves through extracellular matrix signaling pathways for cell cycle and cell proliferation (28), and increase of expression of podoplanin is also related to tumor malignancy and poor clinical outcome (22, 29, 30).

To establish targeted therapy to podoplanin, we previously generated a rat anti-human podoplanin mAb NZ-1 (31, 32). NZ-1 suppressed podoplanin-induced pulmonary metastasis by inhibiting tumor-induced platelet aggregation (27). Furthermore, we showed that NZ-1 has not only high specificity and sensitivity but also high binding affinity against podoplanin, leading to its status as a candidate for radioimmunotherapy or immunotoxin therapy (33). However, it has not been clarified whether NZ-1 has antitumor effector functions including Ab-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC), which is the most important mechanism contributing to the clinical efficacy of therapeutic Abs (34).

In the current study, we investigated whether NZ-1 can induce antitumor effects mediated by ADCC and CDC against MPM in vitro and in vivo. Furthermore, we also evaluated antitumor activity of NZ-8, a novel rat-human chimeric anti-human podoplanin Ab, using human mononuclear cells (MNC) as effector cells against MPM cells.

In this study, we used 15 human MPM cell lines. ACC-MESO-1, ACC-MESO-4, Y-MESO-8A, Y-MESO-9, Y-MESO-12, and Y-MESO-14 were established in Aichi Cancer Research Center Institute (35). EHMES-1 was provided by Dr. Hironobu Hamada (Ehime University, Matsuyama, Japan) (36). NCI-H290 and NCI-H513 were provided by Dr. Adi F. Gazdar (University of Texas Southwestern Medical Center, Dallas, TX). MSTO-211H, NCI-H28, NCI-H226, NCI-H2052, NCI-H2373, NCI-H2452, and Chinese hamster ovary (CHO) were purchased from American Type Culture Collection (Rockville, MD). MPM cell lines are divided into three histological origins, an epithelioid type (ACC-MESO-1, ACC-MESO-4, Y-MESO-9, Y-MESO-12, NCI-H290, NCI-H513, NCI-H226, and NCI-H2452), a sarcomatoid type (NCI-H28, NCI-H2052, and NCI-H2373), and a biphasic type (Y-MESO-8A, Y-MESO-14, EHMES-1, and MSTO-211H) (37–39). These cells were cultured in RPMI 1640 medium supplemented with 10% FBS (CRPMI 1640; Life Technologies, Grand Island, NY), 100 U/ml penicillin, and 100 μg/ml streptomycin (Meiji Seika Kaisha, Tokyo, Japan). Cell lines were authenticated by DNA fingerprinting. MSTO-211H cells were transfected with appropriate amounts of plasmids, pcDNA3/human podoplanin (PDPN) or pcDNA3/mock (11), using Metafectene (Nippon Genetics, Tokyo, Japan) according to the manufacturer’s instructions. Stable transfectants (MSTO-211H/PDPN and MSTO-211H/mock) were selected by cultivating the transfectants in medium containing 0.5 mg/ml Geneticin (Invitrogen, Carlsbad, CA).

NZ-1, a rat anti-human podoplanin mAb (IgG_2a), was developed as described previously (31). Rat IgG and human IgG were purchased from Beckman Coulter (Fullerton, CA). For the generation of rat-human chimera anti-human podoplanin (NZ-8), the appropriate V_H and V_L cDNAs of a rat NZ-1 Ab and C_H and C_L of human IgG₁ were subcloned into the pcDNA3.3/Neo or pcDNA3.1/Zeo vectors (Life Technologies), respectively. The cDNAs coding for the V_H and V_L (κ-chain) regions were constructed by a PCR-based method. Ab expression vectors were transfected into CHO cells using Lipofectamine 2000 reagent (Life Technologies). Stable transfectants of CHO/NZ-8 were selected by cultivating the transfectants in medium containing 1 mg/ml Geneticin and 0.5 mg/ml Zeocin (Life Technologies). CHO/NZ-8 cells were cultivated in CHO-S-SFM II medium (Life Technologies). The media containing NZ-8 were centrifuged, and the obtained supernatant was applied to a column of protein G-Sepharose (Thermo Scientific, Rockford, IL). After extensive washing with PBS, the fusion proteins were eluted using 0.1 M glycine and 0.15 M NaCl (pH 2.8) and then neutralized with 1 M Tris (pH 10). The Abs were dialyzed against PBS. Expression and purity of the proteins were confirmed by SDS-PAGE using 5–20% gradient gels (Wako Pure Chemical Industries, Osaka, Japan). F(ab′)₂ fragments of NZ-1 Ab were prepared using F(ab′)₂ Preparation Kits (Thermo Scientific) according to the manufacturer’s instructions. D2-40 (mouse anti-human podoplanin mAb) was purchased from DakoCytomation (Glostrup, Denmark).

Male SCID mice, 5 to 6 wk old, and male Wistar rats, 6 wk old, were obtained from CLEA Japan (Osaka, Japan) and maintained under specific pathogen-free conditions throughout this study. All animals were acclimatized for at least 1 wk before experiments. All experiments were performed in accordance with the guidelines of University of Tokushima, Committee on Animal Care and Use.

Expression of podoplanin was detected by flow cytometry. Cells (5 × 10⁵) were washed with cold PBS and stained on ice for 30 min with NZ-1 (1 μg/ml), NZ-1 F(ab′)₂ fragment (0.675 μg/ml), NZ-8 (1 μg/ml), rat IgG (1 μg/ml), or human IgG (1 μg/ml). After incubation with the primary mAbs, cells were washed with cold PBS and then incubated on ice for 30 min with FITC-conjugated goat F(ab′)₂ fragment anti-rat IgG (H+L) Ab or FITC-conjugated goat F(ab′)₂ Fragment anti-human IgG (Fcγ) Ab (Beckman Coulter). The cells were washed again and resuspended in cold PBS. The analysis was performed on an FACSCalibur flow cytometer with CellQuest software (BD Biosciences, Franklin Lakes, NJ). The mean specific fluorescence intensity (MSFI) was calculated as the ratio of the mean fluorescence intensity of NZ-1 to that of control mAb. Positive expression of podoplanin was defined as an MSFI of >2.0, and high-level expression of podoplanin was defined as an MSFI of >10.0.

Tissue arrays of human mesothelioma were obtained from Biomax (Rockville, MD). Staining was performed using R.T.U. VECTASTAIN Elite ABC Kit (Vector Laboratories, Burlingame, CA) (40). The tissue arrays were steamed in sodium citrate buffer for 10 min using a microwave to retrieve Ag and incubated in 3% H₂O₂ for 30 min to inhibit endogenous peroxidase. After incubation in blocking serum for 20 min, the slides were incubated overnight with NZ-1, NZ-8, or D2-40 at 4°C. The slides were washed and incubated in prediluted biotinylated secondary Ab for 30 min, followed by incubation in ready-to-use streptavidin/peroxidase complex reagent for 30 min. The tissue arrays were developed with a diaminobenzidine substrate kit (Vector Laboratories) and counterstained with Mayer’s hematoxylin (Muto Pure Chemicals, Tokyo, Japan).

Cells were rinsed with ice-cold PBS and lysed in M-PER reagent (Pierce, Rockford, IL) containing phosphatase and protease inhibitor cocktails (Roche, Indianapolis, IN). The concentrations of protein were determined using a Bio-Rad Protein Assay Kit (Bio-Rad, Hercules, CA). Cell lysate (15 μg/lane) was separated by SDS-PAGE. Immunoblotting was conducted as previously described (41).

Rat splenocytes were harvested from Wistar rat spleens. Spleens were minced, homogenized in RPMI 1640, and centrifuged after passing through the cell strainer (BD Biosciences) (40). The cell pellet was suspended in RBC lysis buffer (Sigma-Aldrich, St. Louis, MO) on ice for 10 min. After depletion of RBCs, splenocytes were resuspended in CRPMI 1640 and used as effector cells. The method by which SCID mouse splenocytes were harvested was the same as that of rat splenocytes. Human PBMCs (MNC) were obtained from leukocytes in lymphocyte separation medium (Litton Bionetics, Kensington, MD) as previously described (42). Leukocytes were separated from peripheral blood of healthy donors using an RS-6600 rotor of a Kubota KR-400 centrifuge (Kubota, Tokyo, Japan). The human study was approved by the ethics committee of University of Tokushima, and written informed consent was obtained from all of the subjects.

ADCC was evaluated using a [⁵¹Cr]release assay as described previously (40, 43). Target cells were targeted with 0.1 μCi [⁵¹Cr]sodium chromate at 37°C for 1 h. ⁵¹Cr-labeled target cells were placed in 96-well plates in triplicate. Effector cells and NZ-1, NZ-1 F(ab′)₂ fragment, NZ-8, or control IgG were added to the plates. After 6 h incubation, 100 μl supernatant was measured in a γ counter (PerkinElmer, Waltham, MA). Percent of cytotoxicity was calculated from the following formula: percent specific lysis = (E − S)/(M − S) × 100, where E is the release in the test sample, S is the spontaneous release, and M is the maximum release.

CDC was determined by [⁵¹Cr]release assay (40). Target cells were incubated with 0.1 μCi [⁵¹Cr]sodium chromate at 37°C for 1 h. After incubation, the cells were washed with CRPMI 1640 three times. ⁵¹Cr-labeled cells were added into 96-well plates and incubated with baby rabbit complement (Cedarlane Laboratories, Ontario, Canada) at a dilution of 1:4 and NZ-1 (1 μg/ml), NZ-8 (1 μg/ml), or control IgG (1 μg/ml) for 6 h. [⁵¹Cr]Release of the supernatant from each well (100 μl) was measured using a γ counter. Percent of cytotoxicity was calculated as above.

Depletion of NK cells from rat splenocytes was performed using the anti-asialo GM1 Ab as follows. Harvested rat splenocytes were treated with anti-asialo GM1 Ab (Wako, Osaka, Japan) at room temperature for 30 min. After treatment, splenocytes were incubated with baby rabbit complement at 37°C for 30 min, then washed with CRPMI 1640 three times, and the number of cells for use in experiments was counted. Rat NK cells were separated from Wistar rat splenocytes using a magnetic cell-sorting system. Splenocytes were incubated with FITC-conjugated anti-CD161a Ab (BD Biosciences) at 4°C for 30 min and then with anti-FITC mAb-coupled superparamagnetic microbeads (Miltenyi Biotec, Auburn, CA) at 4°C for 15 min. CD161a-positive or -negative selection was carried out using an autoMACS (Miltenyi Biotec) according to the manufacturer’s instructions (44). Isolated CD161a⁺ and CD161a⁻ cells that yielded purity ≥90 and ≥99.9%, respectively, as determined by flow cytometry were used in experiments. To purify CD14⁺ cells and CD56⁺ cells from human MNC, AutoMACS and CD14 or CD56 microbeads (Miltenyi Biotec) were used. Human MNC were treated with CD14 or CD56 microbeads at 4°C for 15 min, and then cells were separated by autoMACS. The purities of CD14⁺ and CD56⁺ cells were ≥90%.

SCID mice were injected s.c. into the right flank with ACC-MESO-4 (4.0 × 10⁶ cells) or MSTO-211H/PDPN (1.0 × 10⁶ cells). Tumor size is expressed as tumor area (mm²): the longest diameter × the shortest diameter. The tumor-bearing mice were divided into several experimental groups before treatment with Abs was initiated. An i.p. injection of NZ-1 (100 μg), NZ-1 F(ab′)₂ fragment (67.5 μg), NZ-8 (100 μg), control rat IgG (100 μg), or control human IgG (100 μg) was started, on average, 14 d after tumor inoculation, when the tumor size had reached 15 (± 3 SD) mm², and continued twice a week for 2 wk. Rat CD161a⁺ cells (1.0 × 10⁵ cells), human CD56⁺ cells (1.0 × 10⁵ cells), or control normal saline injection (s.c.) around the tumors continued weekly for 2 wk.

In vivo fluorescence imaging was performed using IVIS Spectrum (Caliper Life Sciences, Hopkinton, MA) (45). ACC-MESO-4 (4.0 × 10⁶ cells) was injected into the right flank of SCID mice (s.c.). When the tumor grew to a sufficient size (25 mm²), 100 μg NZ-1 or rat IgG labeled with XenoLight CF770 (Caliper Life Sciences) was injected (i.p.). The fluorescence in each mouse was determined using IVIS Spectrum (Caliper Life Sciences), with an excitation wavelength of 730–760 nm and an emission wavelength of 830–850 nm, every day after the labeled Ab injection. Mice were anesthetized with isoflurane (Abbott Laboratories, Chicago, IL). Images were acquired and analyzed using Living Image 3.1 software (Caliper Life Sciences).

The statistical significance of differences in in vitro and in vivo data were analyzed by Student t test and one-way ANOVA. In this study, p values <0.05 were considered significant in all experiments.

First, we evaluated the expression of podoplanin on the surface of MPM cells using NZ-1, which is highly reactive for podoplanin. Fig. 1A shows that MPM cells expressed various levels of podoplanin as determined by flow cytometry. Expression of podoplanin was detected on 73% (11 out of 15) of MPM cells, and sarcomatoid type showed a higher expression rate of 100% (3 out of 3) as compared with 75% of epithelioid type (6 out of 8) and 50% of biphasic type (2 out of 4), although the number of cell lines is not sufficient to determine the expression rate. High level of expression of podoplanin, which is defined as a ratio of the mean fluorescence intensity of NZ-1 to that of control IgG of >10, was found in 33% (5 out of 15) of all cells, of which four cell lines are in an epithelioid type.

We next examined podoplanin expression in tissues of mesothelioma by immunohistochemistry with NZ-1. As shown in Fig. 1B, strong intensity of podoplanin staining was observed in tissues of MM, except benign mesothelioma. Podoplanin expression was found in 92% (33 out of 36) of all MM tissues.

Next, we attempted to clarify whether NZ-1 has the potential to induce ADCC-targeted MPM cells expressing podoplanin. As shown in Fig. 2A and 2B, using human MNC or SCID mouse splenocytes as effector cells, we could not detect ADCC induced by NZ-1. By contrast, significant ADCC against only podoplanin-positive, but not podoplanin-negative, MPM cells was induced by NZ-1 when splenocytes of Wistar rat were used as effector cells (Fig. 2C). The level of ADCC was strongly associated with the level of podoplanin expression (Figs. 1A, 2C). In contrast, CDC induced by NZ-1 was detected only against ACC-MESO-4 cells (Fig. 2D). Additionally, treatment with 1 μg/ml NZ-1 alone did not inhibit proliferation of these MPM cell lines, indicating that NZ-1 has no direct growth-inhibitory activity on MPM cells in our experimental conditions (data not shown).

To evaluate the kinetics of ADCC activity mediated by NZ-1, we analyzed the E:T ratio and time course of ADCC and dose dependence of NZ-1. As shown in Fig. 2E, the level of ADCC was dependent on E:T ratio. In addition, NZ-1 also effectively induced ADCC against podoplanin-expressing cells in a dose- and time-dependent manner (Fig. 2F, 2G). The optimal concentration of NZ-1 to induce ADCC seems to be >0.1 μg/ml. These results suggest that rat anti-podoplanin Ab NZ-1 exhibits potent ADCC against MPM cells.

To confirm that ADCC of NZ-1 depends on the expression of podoplanin on target cells, we used MSTO-211H/PDPN as target cells. MSTO-211H/PDPN was generated by transfecting pcDNA3/podoplanin into MSTO-211H, one of the podoplanin-negative MPM cell lines (Fig. 3A). Although ADCC of NZ-1 was not detected against MSTO-211H/Mock, NZ-1 exhibited significant ADCC activity against MSTO-211H/PDPN (Fig. 3B). E:T ratio–, dose-, and time-dependent ADCC activities against MSTO-211H/PDPN were also induced by NZ-1 (Fig. 3C).

For further investigation of the type of effector cells in rat splenocytes, we depleted NK cells using anti-asialo GM1 Ab and baby rabbit complement. As shown in Fig. 4A, ADCC was not induced by NK cell–depleted splenocytes. Moreover, we isolated NK (CD161a⁺) cells and examined ADCC using NZ-1 against ACC-MESO-4. Fig. 4B shows that the purified NK (CD161a⁺) cells from rat splenocytes showed a significant level of ADCC. In contrast, CD161a⁻ splenocytes did not induce ADCC by NZ-1.

To evaluate the antitumor activity of NZ-1 in vivo, we used s.c. xenograft of human MPM (ACC-MESO-4) in SCID mice. NZ-1 or control rat IgG was injected i.p. twice a week, and rat NK (CD161a⁺) cells or normal saline was injected s.c. weekly for 2 wk because ADCC of NZ-1 was mediated only by rat NK cells but not mouse splenocytes. Treatment with NZ-1 was started when tumors were established and formed a visible tumor mass (>15 mm²) owing to the evaluation of therapeutic antitumor ability of NZ-1. The antitumor activity of NZ-1 against ACC-MESO-4 in vivo is shown in Fig. 5A. Administration of NZ-1 with rat NK (CD161a⁺) cells significantly inhibited the growth of ACC-MESO-4 cells. By contrast, the injection of neither NZ-1 nor NK (CD161a⁺) cells alone showed antitumor effects in our experimental setting. Using an MSTO-211H/PDPN xenograft model, tumor growth was also significantly inhibited by treatment with NZ-1 and NK (CD161a⁺) cells (Fig. 5B).

It is known that ADCC activity is mediated by binding of the Fc portion of Abs to the FcR on NK cells. Therefore, we examined the antitumor activity of NZ-1 F(ab′)₂ fragment. Although the NZ-1 F(ab′)₂ fragment recognized podoplanin on ACC-MESO-4 in flow cytometric analysis to a similar extent as intact NZ-1, the F(ab′)₂ fragment did not induce ADCC against ACC-MESO-4 (Fig. 6A, 6B). Moreover, i.p. injection of NZ-1 F(ab′)₂ fragment unalterably showed no effect on tumor growth of ACC-MESO-4 compared with intact NZ-1, even when rat NK cells were simultaneously administered (Fig. 6C).

Next, we investigated the accumulation of injected NZ-1 on s.c. podoplanin-positive MPM tumor in SCID mouse model. As shown in Fig. 7A, Xenolight CF770-labeled NZ-1 accumulated on s.c. ACC-MESO-4 after single i.p. injection. Quantification of accumulation to the tumor of NZ-1 or control rat IgG showed a peak of the accumulation of NZ-1 on day 5, and the accumulation could be detected over 3 wk (Fig. 7B).

Because it was demonstrated that NZ-1 showed antitumor activity through ADCC mediated by rat NK cells, to apply targeted therapy to podoplanin with Ab in a clinic, we generated the chimeric Ab by fusing the V_H and V_L regions of rat Ab (NZ-1) with the C_H and C_L regions of human IgG₁, respectively (Fig. 8A). As shown in Fig. 8B and 8C, NZ-8, a novel rat-human chimeric anti-human podoplanin Ab, could recognize the podoplanin expressing in MPM like NZ-1 by flow cytometry as well as immunohistochemistry. Moreover, to confirm the recognition pattern of NZ-8 to podoplanin, the lysate of MPM cells were subjected to Western blotting. As shown in Fig. 8D, NZ-8 could detect the ladder bands of podoplanin. However, the recognizing pattern of NZ-8 was different from that of D2-40. Thinking of clinical application of NZ-8, the recognition of podoplanin expressing in normal tissue by NZ-8 is quite important. Therefore, we examined the staining pattern of podoplanin expressing in normal tissue such as lymphatic vessels, alveolar epithelial cells, and podocytes in the kidney by immunohistochemistry with NZ-8 because these tissues have been reported to express podoplanin. As shown in Fig. 8E, NZ-8 preferentially recognized the podoplanin expressing in MPM, but not in normal tissues including lymphatic vessels, alveolar epithelial cells, and podocytes. However, D2-40 strongly detected the podoplanin expressing in lymphatic vessels as compared with MPM tissues.

We further assessed whether NZ-8 can induce antitumor effect against MPM cells mediated by human MNC as effector cells. As shown in Fig. 9A, ADCC against MPM cell lines was not exhibited by NZ-1 with human MNC, whereas NZ-8 showed the induction of a significant level of ADCC mediated by human MNCs against podoplanin-positive MPM cells but not podoplanin-negative MPM cells. CDC was also induced by NZ-8 against ACC-MESO-4 and NCI-H226, which was significantly higher than that of NZ-1 (Fig. 9B). Further effector cell analysis of ADCC clearly demonstrated that human NK (CD56⁺) cells, but not human monocytes (CD14⁺), could induce ADCC by NZ-8 (Fig. 9C). Using the xenograft model of ACC-MESO-4 in SCID mice, the tumor growth was also significantly inhibited by treatment with NZ-8 and human NK (CD56⁺) cells (Fig. 9D). By contrast, no antitumor effects were shown by the injection of NZ-8 or human NK (CD56⁺) cells alone. Furthermore, administration of NZ-1 did not show any antitumor effects when human NK (CD56⁺) cells were used as an effector cells (Fig. 9E).

In the current study, we have demonstrated that a rat anti-human podoplanin mAb, NZ-1, possesses potent ADCC mediated by rat NK cells and CDC activity and mediates in vivo therapeutic antitumor activity against MPM in SCID mice. Moreover, we have also shown that a novel rat-human chimeric anti-human podoplanin Ab, NZ-8, effectively induces ADCC mediated by human NK cells and CDC and mediates in vivo antitumor activity against MPM cells. Furthermore, NZ-8 preferentially recognized the podoplanin expressing in MPM than those in normal tissues. These results strongly suggest the possible clinical application of anti-human podoplanin Ab against MPM.

In this study, we found that podoplanin was expressed on 73% of MPM cell lines including 33% of high-expression and 92% of MM tissues, which is consistent with previous studies (46–48). Furthermore, NZ-8 could recognize the expression of podoplanin in MPM like NZ-1. Compared with other mesothelioma-associated Ags such as calretinin, thrombomodulin, cytokeratin 5, WT1, and mesothelin in previous studies (47–49), the frequency of expression of podoplanin is the same or higher than that of others, suggesting that podoplanin could be a novel immunological target for MPM therapy. In contrast, although it was reported that podoplanin was expressed in lymphatic vessels and type I alveolar epithelium and podocyte (15–17), NZ-8 could not detect the expression of podoplanin in lymphatic vessels when it was used at the lower concentration of 0.1 μg/ml, which could still recognize the podoplanin in MPM tissues and induce the ADCC and CDC reactions. This staining pattern of NZ-8 is quite different from that of D2-40, which has been frequently used as an anti-podoplanin Ab in many studies (46, 47, 50). The reason why NZ-8 preferentially recognizes the podoplanin expressing in MPM, but not in normal tissues, is still unclear. However, it might be due to the difference in the epitopes recognized by each Ab in addition to the difference in the expression level of podoplanin between MPM and normal tissues, because we have shown that the epitope recognized by NZ-1 is platelet aggregation-stimulating domain-2/3, and the epitope of D2-40 is platelet aggregation-stimulating 1/2 (32). In the current study, it is considered that the recognition pattern of NZ-8, which possesses V_H and V_L of NZ-1, was also different from that of D2-40.

Originally, NZ-1 showed high binding affinity against podoplanin (K_D = 1.2 × 10⁻¹⁰ M) (33). In addition, it has been demonstrated that NZ-1 suppresses podoplanin-induced pulmonary metastasis through inhibiting platelet aggregation (27). In the current study, NZ-1 showed other effector functions, for ADCC and CDC. Recently, therapeutic Abs have been classified into three classes (34), in which class I Abs are defined to possess ADCC and CDC and ranked in the highly potential category, especially for cancer immunotherapy. On the basis of this concept, NZ-1 seems to be ranked as a highly potent therapeutic Ab because it has effector functions such as ADCC and CDC in addition to preventing metastasis by inhibiting platelet aggregation.

ADCC activity of NZ-1 has not been reported because NZ-1 did not show any ADCC activity mediated by murine splenocytes or human MNCs. However, when ADCC of NZ-1 was examined using rat splenocytes, significant ADCC activity was detected. To use rat splenocytes as effector cells is unusual to evaluate the ADCC activity of Ab; however, other studies have also reported that, in special cases, rat splenocytes are more useful than murine ones for testing the ADCC activity of rat mAb (51, 52). If the Ab has potency to induce the ADCC mediated by effector cells of several species except for humans effectively, it would be possible to prepare chimeric or humanized Ab that induces ADCC mediated by human effector cells. Therefore, from our study, it is also noted that rat splenocytes have the same importance to screen the ADCC activity of Ab as human MNC or mouse splenocytes.

In vivo experiments have shown that NZ-1 reduces the tumor growth of s.c. inoculated cell lines of both ACC-MESO-4 and MSTO-211H/PDPN, suggesting that NZ-1 possesses antitumor effects against MPM as well. In addition, it is very important that antitumor effects of NZ-1 were observed against established tumors, indicating that treatment with NZ-1 was apparently therapeutic, not preventive. Previously, we reported that NZ-1 directly inhibits podoplanin-dependent tumor metastasis in vivo (27). By contrast, in this study, administration of NZ-1 alone had no effect to reduce the podoplanin-positive MPM cell growth in SCID mice. The suppressive effects on tumor growth of NZ-1 are required for the administration with rat NK cells, probably as immune effector cells. In addition, NZ-1 F(ab′)₂ fragment induced neither ADCC in vitro nor antitumor effect in vivo, which supports the notion that interaction between Fc portion of the Ab and FcR of the effector cell leads to the antitumor effect of NZ-1. Moreover, in vivo fluorescence imaging results have demonstrated that injected NZ-1 specifically accumulates to s.c. ACC-MESO-4. Recent study has shown that tumor accumulation of trastuzumab, which is an anti-Her2 Ab widely used against Her2-positive breast cancer in a clinical context, peaked at 24 h and decreased at the same rate as the baseline after 24 h in SCID mice model (53). By contrast, in this study, specific accumulation of NZ-1 peaked at day 5 and could continue over 3 wk after a single injection. Recently, we have also shown that NZ-1 was selectively accumulated to podoplanin-positive glioblastoma but not normal tissues (33). Therefore, it was estimated that NZ-1 has strong binding affinity and specific accumulation potential against target Ags. These results indicated that NZ-1 might have strong antitumor effects against MPM in vivo by ADCC.

Finally, we have shown that NZ-8, a novel rat-human chimeric anti-human podoplanin Ab, generated from NZ-1 has the potency to induce ADCC against MPM cells mediated by human NK cells as effector cells. Interestingly, the level of ADCC and CDC of NZ-8 was significantly higher than that of NZ-1. Moreover, NZ-8 has the promising antitumor effects against MPM cells mediated by human NK cells in vivo, in contrast to administration of NZ-1 with human NK cells, which did not induce any antitumor effects. In addition, NZ-8 preferentially detected the podoplanin expressing in MPM, but not in normal tissues.

In conclusion, we found that both NZ-1 and NZ-8 possess potent and therapeutic antitumor effects based on ADCC against MPM in an SCID mouse model. Importantly, NZ-8 could induce higher ADCC mediated by human NK cells in addition to CDC as compared with NZ-1. Based on the ability of NZ-8, which preferentially recognizes the podoplanin expressing in MPM, but not in normal tissues, the current study suggests that targeting therapy to podoplanin using NZ-8 might be useful for patients with MPM as a novel immunotherapy.

We thank Tomoko Oka, Imi Saito, Ryo Yanagiya, Shunpei Morita, Kana Nasu, Hiroko Sasaki, Sachiko Suzuki, and Junko Aita for excellent technical assistance.

The authors have no financial conflicts of interest.