Anti-tumor activity of an anti-endoglin monoclonal antibody is enhanced in immunocompetent mice

Authors


Abstract

In the present study, we investigated the mechanisms by which anti-endoglin (EDG; CD105) monoclonal antibodies (mAbs) suppress angiogenesis and tumor growth. Antihuman EDG mAb SN6j specifically bound to murine endothelial cells and was internalized into the cells in vitro. SN6j effectively suppressed angiogenesis in mice in the Matrigel plug assay. We found that SN6j is more effective for tumor suppression in immunocompetent mice than in SCID mice. We hypothesized that T cell immunity is important for effective antitumor efficacy of SN6j in vivo. To test this hypothesis, we investigated effects of CpG oligodeoxynucleotides (ODN) and depletion of CD4+ T cells and/or CD8+ T cells on antitumor efficacy of SN6j in mice. Systemic (i.v.) administration of a relatively small dose (0.6 μg/g body weight/dose) of SN6j suppressed growth of established s.c. tumors of colon-26 in BALB/c mice and improved survival of the tumor-bearing mice. Addition of CpG ODN to SN6j synergistically enhanced antitumor efficacy of SN6j. In contrast, such enhancing effects of CpG ODN were not detected in SCID mice. Antitumor efficacy of SN6j in BALB/c mice was abrogated when CD4+ T cells and/or CD8+ T cells were depleted; effect of CD8+ T cell depletion was stronger. Interestingly, CD4-depletion decreased tumor growth while CD8-depletion enhanced tumor growth in the absence of SN6j. SN6j induced apoptosis in human umbilical vein endothelial cells in a dose-dependent manner which indicates an additional mechanism of antiangiogenesis by SN6j. © 2008 Wiley-Liss, Inc.

Angiogenesis is essential for tumor growth, progression and metastasis and antiangiogenic therapy has become a new promising strategy in the treatment of cancer patients.1, 2 Antiangiogenic therapy may have advantages over conventional tumor cell-targeted therapies including chemotherapy and immunotherapy. For instance, tumor vessel-targeted therapies may be able to overcome three major problems associated with other anticancer therapies, i.e., the problems of drug resistance,3, 4 poor delivery5, 6 and tumor heterogeneity.7

EDG (CD105) is a homodimeric cell surface glycoprotein which is mainly expressed on immature B-lineage leukemia cells and endothelial cells.8–10 EDG is a proliferation-associated cell membrane antigen on endothelial cells11–14 and leukemia cells,15 and is strongly expressed on the tumor-associated angiogenic vascular endothelium.12, 13, 16 In addition, EDG is essential for angiogenesis17, 18 and a component of the transforming growth factor (TGF)-β receptor complex.19 Recently, several studies indicated that EDG represents a more specific and sensitive marker for tumor angiogenesis and/or tumor progression than the commonly used pan-endothelial markers such as CD34 and CD31 in various types of human malignancies.20–25 These data demonstrated that EDG is a useful prognostic marker of various solid cancers. In addition, they suggest that EDG may be a useful target for antiangiogenic therapy and vascular targeting therapy.

Previously, we showed that immunoconjugates and naked form of selected anti-EDG mAbs were effective for tumor suppression by targeting angiogenic vasculature in mice.7, 13, 26–28 However, the underlying mechanisms by which the naked mAbs exert antiangiogenic activities in vivo are poorly understood. In the present study, we tested our hypothesis that the immune status of the hosts will be an important factor for effective mAb-based EDG-targeted tumor therapy. This hypothesis was conceived based on our observation that anti-EDG mAb was more effective for tumor therapy in immunocompetent mice than in immunodeficient mice. The present results of T cell depletion experiments as well as combination therapy with CpG ODN and anti-EDG mAb support the hypothesis and indicate that T cell immunity, especially CD8+ T cells, play a pivotal role in mAb-based EDG-targeted therapy of tumors in vivo.

Abbreviations:

BW, body weight; CAM, camptothecin; EDG, endoglin; FITC, fluorescein isothiocyanate; HUVEC, human umbilical vein endothelial cell; mAb, monoclonal antibody; MSA, mouse serum albumin; MVD, microvessel density; ODN, oligodeoxynucleotides; p.t., peritumoral.

Material and methods

Cells and animals

SVEC4-10 mouse endothelial cells were kindly provide by Dr. Kathryn O'Connell of Johns Hopkins University and cultured in Dulbecco's Modified Eagle's Medium (DMEM) media containing 10% fetal bovine serum (FBS), 100 unit/ml penicillin and 50 μg/ml streptomycin as described previously.13 Human umbilical vein endothelial cells (HUVECs) were obtained from Clonetics (San Diego, CA) and grown in endothelial growth medium containing supplements (Clonetics) and 5% FBS. Colon-26, a murine colon adenocarcinoma cell line, and 4T1, a murine mammary carcinoma cell line, were cultured in a monolayer in RPMI 1640 media containing 10% FBS, 100 unit/ml penicillin, 100 μg/ml streptomycin and 250 ng/ml amphotericin B. NIH-3T3, a murine fibroblast cell line, and human EDG-transfected NIH-3T329 were cultured in RPMI 1640 supplemented with 10% FBS. GK1.5 rat hybridoma for anti-CD4 mAb and 2.43 rat hybridoma for anti-CD8a mAb were obtained from American Type Culture Collection (ATCC, Manassas, VA), and cultured in DMEM containing 10% FBS, 4 mM L-glutamine, 4.5 g/l glucose and 1.5 g/l sodium bicarbonate. All cells were cultured in a humidified atmosphere with 5% CO2 at 37°C. Five to six-week-old BALB/c and SCID mice were obtained from National Cancer Institute (Bethesda, MD) and Roswell Park Cancer Institute (RPCI), respectively. They were supplied autoclaved food and water ad libitum. All handling of the mice was performed in a laminar flow hood.

Antibodies

Anti-EDG mAb SN6j that weakly cross-reacts with mouse endothelial cells was generated in our laboratory.26 An isotype-matched murine control IgG MOPC 195 variant (IgG1-κ) was prepared as described previously.8 The sterilized solutions of purified SN6j and control IgG were individually diluted with sterilized PBS containing mouse serum albumin (MSA, final concentration 0.05%). Anti-CD4 mAb and anti-CD8 mAb were purified from GK1.5 hybridoma ascites and 2.43 hybridoma ascites, respectively by affinity chromatography on protein A columns. Each mAb was individually sterilized by filtering through Millex-GV filter (0.22 μm; Millipore, Billerica, MA) in a laminar flow hood before used for injections into mice. Rat whole IgG (control) was purchased from SIGMA (St. Louis, MO). Rat anti-mouse CD31 (PECAM-1) mAb (rat IgG2a-κ), biotin-conjugated goat anti-rat immunoglobulin (Ig) specific polyclonal antibodies (pAb, multiple adsorption), fluorescein isothiocyanate (FITC)-conjugated rat anti-mouse CD4 mAb RM4-5, FITC-conjugated rat anti-mouse CD8a mAb 53-6.7 and FITC-conjugated rat IgG2a-κ (control) were purchased from BD Biosciences (San Jose, CA).

ODN

CpG ODN 1826 (sequence 5′-TCC ATG ACG TTC CTG ACG TT-3′) and non-CpG control ODN 2138 (sequence 5′-TCC ATG AGC TTC CTG AGC TT-3′) were obtained from Coley Pharmaceutical Canada (Ottawa, Canada).

Immunofluorescence staining of the cells

FITC-labeled SN6j and isotype-matched control IgG were individually incubated with SVEC4-10 proliferating mouse endothelial cells from subconfluent cultures for 4 hr at room temperature. After washes, the FITC-labeled cells were examined under a Reichert microscope (Cambridge Instruments, Buffalo, NY), and pictures were taken using ASA 160 Kodachrome color film (Eastman Kodak, Rochester, NY). In addition, colon-26, 4T1, HUVECs, NIH-3T3 and human EDG-transfected NIH-3T329 were examined for their reactivities with FITC-SN6j and FITC-labeled isotype-matched control IgG (MOPC 195 variant, IgG1-κ) after incubation of the reaction mixtures for 1 and/or 2 hr at 37°C. Details of the EDG transfection will be reported elsewhere.

In vivo Matrigel plug assay

Five-hundred microliter of Matrigel Matrix (BD Biosciences, San Diego, CA) were mixed with 1.25 × 105 colon-26 cells and implanted into mice. One day after the implantation, mice were divided into 2 groups (7 mice/group) according to plug size and body weight (BW). The mice were treated by injections of 1.8 μg/g BW of SN6j or isotype-matched control IgG via tail vein on day 1, 4 and 7. Ten days after implantation, Matrigel plugs were removed and fixed in zinc fixative (BD Biosciences) for 24 hr at room temperature, and stained with anti-mouse CD31 mAb using LSAB+ system-HRP (horse radish peroxidase) from Dako (Carpinteria, CA) according to manufacturer's instruction with minor modifications. Briefly, the sections were incubated with rat anti-mouse CD31 mAb (diluted 1:10) for 30 min at room temperature. An isotype-matched control IgG (rat IgG2a-κ) was used as a negative control. After washings, the sections were incubated with biotin-conjugated goat anti-rat Ig specific pAb (diluted 1:50) for 30 min at room temperature and followed by incubations with streptavidin peroxidase and substrate-chromogen solution according to the instruction. The sections were counterstained with hematoxylin. For the quantification of microvessel density (MVD), 12 hotspot fields (4 fields × 3 samples) of CD31 staining at 200× were captured from each group using Spot digital camera (Diagnostic Instruments, Sterling Heights, MI) mounted to Nikon ECLIPSE E600 (Kawasaki, Japan).7

Apoptosis assay using cell death detection ELISA

HUVECs (5 × 104 cells/well) were placed into 6-well plates and cultured in endothelial growth medium overnight. Cells were incubated with SN6j (50 or 100 μg/ml), or an isotype matched control IgG (100 μg/ml) for 48 hr or with camptothecin (CAM; 4 μg/ml) for 4 hr. Nucleosome fragmentation was assessed using the Cell Death Detection ELISA (Roche, Indianapolis, IN) according to the manufacturer's instruction.

Tumor model

Cultured colon-26 cells were harvested using Hanks solution containing 3 mM EDTA and 25 mM HEPES, washed twice and then suspended in PBS, pH 7.2. A portion (0.1 ml) of the cell suspensions containing 1.25 × 105 cells was inoculated s.c. into the left flank of mice using a 30G1/2 needle (BD 30G1/2 PrecisionGlide Needle; Becton Dickinson, Franklin Lakes, NJ) to establish s.c. tumors. Recently, we have reported that two different types of tumors appear when tumor cells are injected to make s.c. tumors in mice; one is SS type which grows in the skin-side tissue (i.e., epidermis, corium or subcutis), and the other is MS type which grows in the muscle-side tissue (i.e., fascia, muscle or peritoneum/pleura).28 MS tumors grow faster and are less responsive to SN6j treatment than SS tumors, and the number of the mice with MS tumor is much less than that of the mice with SS tumor.28 Therefore, we chose only the mice bearing SS type tumors for animal experiments in this study.

Therapy of mice bearing tumors

Three to five days after tumor inoculation, mice with distinct SS type tumor were selected and divided into desired number of groups according to tumor volume and BW. SN6j or an isotype-matched control IgG was administered i.v. at a dose of 0.6 μg/g BW in 0.2 ml PBS containing 0.05% MSA via the tail vein. The treatment was initiated 3–5 days after the tumor inoculation, and was repeated every 3–4 days. CpG ODN and control ODN were administered peritumorally (p.t.) at a dose of 30 μg in 0.1 ml PBS/mouse. When CpG or control ODN were administered in combination with SN6j or control IgG, the treatment of ODN was started 1 day after the first mAb/control IgG treatment and was repeated every other day.

Pilot Experiments of in vivo depletion of CD4+ and CD8+ T cells

Pilot experiments were performed using purified anti-CD4 mAb GK1.5 and anti-CD8 mAb 2.43 in order to determine the effective doses. Mice received i.p. administration of varying doses (0.15, 0.3 or 0.6 mg/0.2 ml PBS/mouse) of anti-CD4 mAb or anti-CD8 mAb, or the maximum dose (0.6 mg/0.2 ml PBS/mouse) of control rat IgG for 3 consecutive days (day 0, 1 and 2). On day 5, flow cytometric analysis was performed to verify depletion of an appropriate subset of T cells.

In vivo Depletion of CD4+ and/or CD8+ T cells

Mice received i.p. administration of anti-CD4 mAb and/or anti-CD8 mAb, or rat control IgG at a dose determined in the titration experiment (see earlier) on day −1, 0, 1, 8, 15 and 22. Tumor challenge was performed on day 0.

Preparation of single-cell suspensions from spleens

Spleens from mice were aseptically removed and placed in RPMI 1640 containing 5% FBS. Several holes were made in the organs with a needle, and blood cells were retrieved by running the media through the holes of the organs. Single-cell suspensions were obtained by passing through a 70 μm sterile nylon mesh (BD Biosciences), and erythrocytes were lysed with ACK (ammonium chloride potassium) lysis buffer (containing 0.15 M NH4Cl, 10.0 mM KHCO3 and 0.1 mM Na2EDTA). The cells were washed twice with PBS, resuspended in PBS and then used for flow cytometric analysis.

Flow cytometric analysis

The efficacy of in vivo depletion of CD4+ and CD8+ T cells was determined by staining cells from spleens with FITC-conjugated anti-CD4 mAb RM4-5 and FITC-conjugated anti-CD8a mAb 53-6.7. RM4-5 and 53-6.7 do not compete for the epitope with GK1.5 and 2.43, respectively. FITC-conjugated rat IgG2a-κ was used as a control. Cells (1 × 106) in individual wells of a 96-well round bottomed plate were incubated with an appropriate antibody reagent in PBS-BSA (1% BSA in PBS) at 4°C for an hour with constant shaking. Cells were washed thrice and analyzed on a FACScan (BD Biosciences) using CellQuest Version 3.1 software (BD Biosciences).

Follow-up of treatment efficacy

During the treatment, mice were monitored daily for morbidity. Mice were weighed every 2 or 3 days using an electronic balance (Scout Pro 202, Ohaus, NJ). Tumor size was measured every 2 or 3 days using an electronic digital caliper (Ultra-Cal IV, Ted Pella, CA) that was connected to a computer using GageWedge software (Fred V. Fowler Company, Newton, MA). The measured tumor diameters were converted to tumor volume using Excel 2003 and the following formula: V = length × width × height × π/6.

Statistical analysis

Statistical analysis of the data was performed using the Mann-Whitney U test and the Student's t test. Survival of mice was evaluated by Kaplan–Meier analysis and log-rank test (StatView software, version 5).

Results

Cross-reactivity of anti-EDG mAb SN6j with mouse endothelial cells

As shown in Figure 1b, FITC-SN6j reacted with SVEC4-10 murine endothelial cells and strong staining of the intracellular part of the cells is seen, whereas only very faint background staining of the cells is seen when the cells were allowed to react with an FITC-labeled isotype-matched control IgG (Fig. 1a). The results show that SN6j effectively bound to mouse endothelial cells and were internalized into the cells. In an additional study, reactivities of FITC-SN6j and FITC-control IgG with EDG-negative colon-26 and 4T1 were examined. In this test, EDG-positive HUVECs were included as a positive control. Recently we showed that neither colon-26 nor 4T1 reacts with SN6j in a cellular radioimmunoassay.28 FITC-SN6j showed only faint background staining of colon-26 (Fig. 1c) and 4T1 (data not shown), whereas it strongly stained the intracellular area of HUVECs (Fig. 1d). In a further study, NIH-3T3 and human EDG-transfected NIH-3T3 were compared for their reactivities with FITC-SN6j and FITC-control IgG. FITC-SN6j showed only faint background staining with NIH-3T3 (Fig. 1e), whereas it strongly stained the EDG-transfected NIH-3T3 (Fig. 1f). FITC-control IgG showed no significant staining of any of the tested cells.

Figure 1.

Reactivity of SN6j with SVEC4-10 murine endothelial cells, colon-26 murine carcinoma cells, HUVECs, NIH-3T3 murine fibroblast cells and human EDG-transfected NIH-3T3 cells. SN6j cross-reacted with viable SVEC4-10 cells and was internalized into the cells as shown by strong FITC-SN6j staining in the intracellular part of the cells (b). In contrast, only faint background staining is seen with the cells treated with the FITC-labeled isotype-matched control IgG (arrows in a). FITC-SN6j shows only faint background staining with colon-26 cells (c), whereas it strongly stains the intracellular area of HUVECs. Only faint background staining of FITC-SN6j is detected with NIH-3T3 cells (e), whereas it strongly stains the cell surface of the transfected cells (f). FITC-control IgG did not stain significantly any of the tested cells. Scale bar = 20 μm.

Antiangiogenic effect of SN6j on angiogenic murine blood vessels

Matrigel plug assay was performed to investigate the effect of SN6j on angiogenesis in mice. A mixture of Matrigel Matrix (500 μl) and colon-26 cells (1.25 × 105) was injected s.c. into the left flank of individual mice. Matrigel plugs were removed 10 days after the injection, and MVD was assessed by immunostaining with anti-mouse CD31 mAb. Representative images are shown in Figure 2a. MVD of Matrigel plugs of SN6j-treated mice was significantly lower than that of isotype-matched control IgG-treated mice (p < 0.05; Fig. 2b). Moreover, the vessels in plugs of SN6j-treated mice showed shrinkage and attenuation compared with those of control IgG-treated mice.

Figure 2.

Effect of SN6j on the Matrigel plug angiogenesis in mice. Matrigel Matrix mixed with 1.25 × 105 colon-26 cells was injected s.c. in the left frank of BALB/c mice on day 0. The mice were treated with 1.8 μg/g BW of SN6j or isotype-matched control IgG via tail vein on day 1, 4 and 7. The Matrigel plugs were fixed on day 10 for immunohistochemical staining. (a) examples of rat anti-mouse CD31 (PECAM-1) mAb staining of Matrigel plug sections from SN6j-treated and control IgG-treated mice. The sections were counterstained with hematoxylin. Arrowheads in the figure indicate representative CD31 positive vessels. (b) the CD31 positive vessels in hotspots were counted in four separate ×200 fields from 3 samples per group.

In an additional study, MVD was compared between colon-26 tumors of SN6j-treated mice (n = 4) and those of isotype-matched control IgG-treated mice (n = 4). Although the MVD value of the SN6j-treated tumors was lower than that of the control IgG-treated tumors, the difference was statistically not significant (p = 0.14 and 0.39, respectively, for the CD105-stained vessels and CD31-stained vessels). The result supports our previous finding7 and supports the notion that Matrigel Matrix facilitates tumor angiogenesis.

SN6j-induced apoptosis in HUVECs

HUVECs were incubated with SN6j (50 and 100 μg), control IgG (100 μg) or CAM (positive control). Apoptosis was measured by ELISA. SN6j induced apoptosis in a dose-dependent manner in HUVECs from either a single-donor (Fig. 3a) or multi-donors (Fig. 3b).

Figure 3.

SN6j-induced apoptosis of HUVECs measured by ELISA. HUVECs (5 × 104 cells/well) were cultured in endothelial growth medium. Cells were incubated with SN6j (50 or 100 μg/ml), an isotype matched control IgG (100 μg/ml) or camptothecin (CAM; 4 μg/ml). Nucleosome fragmentation was assessed using ELISA. Differences in the value between SN6j (100 μg/ml) and control IgG (100 μg/ml) in each of the two tests are statistically significant (p < 0.05).

Antitumor effect of SN6j combined with CpG ODN

In a preliminary experiment, SN6j was more effective in BALB/c mice than in SCID mice for suppressing growth of colon-26 s.c. tumors. We hypothesized that the observed difference may be attributable to the difference in T cell immunity in view of the known deficiencies of T and B cells in SCID mice. To test this hypothesis, we investigated the in vivo antitumor effect of SN6j combined with CpG ODN which is known as a strong immune activator and induces Th1-responses in mice.30 First we examined the production of the Th1-promoting cytokines IL-12 and IFN-γ in response to p.t. injection of CpG ODN in BALB/c mice bearing colon-26 s.c. tumors. As expected, production of these cytokines significantly increased in mouse plasma (data not shown). Next, colon-26 cells were injected s.c. (day 0) and the mice with distinct palpable tumors were divided into 4 groups (n = 12 in each group) on day 4. The tumor-bearing mice were treated by i.v. administration of SN6j or an isotype-matched control IgG combined with p.t. injection of CpG ODN or control ODN. Treatment with SN6j/control IgG was initiated on day 4 after tumor inoculation and repeated every 4 days. Injections of ODN were initiated on day 5 and repeated every other day. As shown in Figure 4a, treatment with either SN6j or CpG ODN significantly suppressed the growth of colon-26 s.c. tumors (p < 0.05). Moreover, the antitumor efficacy of SN6j was significantly enhanced by CpG ODN (p < 0.05). The potential synergy in the tumor suppression between SN6j and CpG ODN was examined as described previously7, 31, 32 and the results are shown in Table I. The results demonstrate synergy between SN6j and CpG ODN in the suppression of tumor growth in BALB/c mice. Of 12 tumor-bearing mice, tumors completely disappeared in 9 of the mice by the combination therapy with SN6j and CpG ODN (Fig. 4a). Furthermore, the combined therapy improved the overall survival compared with single agent therapy with SN6j or CpG ODN (Fig. 4b). On the other hand, in SCID mice, SN6j did not show significant antitumor effect against fast-growing colon-26 s.c. tumors (Fig. 4c). In addition, CpG ODN did not enhance antitumor efficacy of SN6j although CpG ODN alone showed significant inhibitory effect on tumor growth compared with control ODN (p < 0.05) (Fig. 4c). These results suggest that T cell or B cell immunity may play an important role in SN6j-based tumor therapy in mice.

Figure 4.

Antitumor effect of SN6j combined with CpG ODN in BALB/c mice or SICD mice. Colon-26 cells (1.25 × 105) were injected s.c. on day 0 and the mice with distinct palpable tumors were divided into 4 groups (n = 12 in each group) and treated by i.v. administration of SN6j (0.6 μg/g BW) or an isotype-matched control IgG (0.6 μg/g BW) combined with p.t. injection of CpG ODN (30 μg/mouse) or control ODN (30 μg/mouse). □ (in a and c) and equation image (in b), control IgG plus control ODN; ○ and ▪ ▪ ▪, SN6j plus control ODN; ▵ and equation image, control IgG plus CpG ODN; • and equation image, SN6j plus CpG ODN. Arrows and arrowheads in the figure indicate administration of mAb/control IgG and ODN, respectively. (a) average size of tumors in BALB/c mice. (b) difference in survival of tumor-bearing BALB/c mice. (c) average size of tumors in SCID mice.

Table i. Combination Therapy with SN6j and CpG ODN in BALB/c Mice
Day2Fractional tumor volume (FTV)1 relative to untreated controls
SN6j3CpG ODN4Combination treatmentRatio of expected: observed FTV6
Expected5Observed
  • 1

    FTV, calculated as (mean tumor volume experimental)/(mean tumor volume control).

  • 2

    Day after tumor cell inoculation.

  • 3

    SN6j + control ODN.

  • 4

    Control IgG + CpG ODN.

  • 5

    Mean FTV of SN6j × mean FTV of CpG ODN.

  • 6

    Obtained by dividing the expected FTV by the observed FTV. A ratio of >1 indicates a synergistic effect, and a ratio of <1 indicates a less than additive effect.

40.991.081.080.991.09
90.930.840.780.621.26
140.780.600.470.281.65
180.640.390.250.073.66
230.650.220.150.034.31

Effect of CD4+ and/or CD8+ T cell depletion on the antitumor efficacy of SN6j

To examine the role of CD4+ T cells and CD8+ T cells in the antitumor effect of SN6j, we depleted BALB/c mice of CD4+ and/or CD8+ T cells by anti-mouse CD4 mAb and/or anti-mouse CD8 mAb in vivo. According to the dose titration experiment of anti-CD4 mAb or anti-CD8 mAb, more than 95% of targeted T cells were depleted by 3 consecutive-day i.p. injection of 0.15 mg mAb/mouse/day (data not shown). On the basis of this observation, mice received i.p. administration of 0.15 mg anti-CD4 mAb and/or anti-CD8 mAb, on day −1, 0, 1, 8, 15 and 22. Tumor challenge of colon-26 cells (1.25 × 105) was performed on day 0. Treatment with SN6j or control IgG was initiated on day 5 and repeated every 3 days.

First, we investigated effects of CD4+ T cells plus CD8+ T cells on antitumor efficacy of SN6j in mice (Figs. 5a and 5b). As shown in Figure 5a, SN6j showed significant antitumor efficacy against established tumors compared with an isotype-matched control IgG before depletion of CD4+ T cells and CD8+ T cells (termed CD4/CD8-depletion) (p < 0.05). However, this antitumor efficacy of SN6j was abrogated by CD4/CD8-depletion (Fig. 5a). Furthermore, SN6j's capacity to improve survival of the tumor-bearing mice was abrogated by CD4/CD8-depletion (Fig. 5b).

Figure 5.

Effect of CD4 and/or CD8-depletion on the SN6j induced-growth inhibition of colon-26 tumors in BALB/c mice. Mice to be depleted of CD4/CD8 T cells received i.p. administration of 0.15 mg/mouse anti-CD4 mAb and/or anti-CD8 mAb or control IgG on day −1, 0, 1, 8, 15 and 22. Tumor challenge of colon-26 cells (1.25 × 105/mouse) was performed on day 0. □ (in the left panels) and equation image (in the right panels), nondepleted mice treated with control IgG; ○ and ▪ ▪ ▪, nondepleted mice treated with SN6j; ▵ and, equation image depleted mice treated with control IgG; • and equation image, depleted mice treated with SN6j. Arrows in the figure indicate i.v. injections of SN6j or an isotype-matched control IgG. (a) average size of the tumors treated with SN6j or control IgG in mice with or without CD4/CD8-depletion. Each group consisted of 12 mice. (b) difference in survival of tumor-bearing mice treated with SN6j or control IgG with or without CD4/CD8-depletion. (c) average size of tumors treated with SN6j or control IgG in mice with or without depletion of CD4+ T cells. Nondepleted control group and CD4-depleted group consisted of 13 and 12 mice, respectively. (d) difference in survival of tumor-bearing mice treated with SN6j or control IgG with or without CD4-depletion. (e) average size of tumors treated with SN6j or control IgG in mice with or without depletion of CD8+ T cells. Each nondepleted control group and CD8-depleted group had 11 and 10 mice, respectively. (f) difference in survival of tumor-bearing mice treated with SN6j or control IgG with or without CD8-depletion.

Next, we examined effects of CD4+ T cells alone on tumor growth and antitumor activity of SN6j (Figs. 5c and 5d). CD4-depletion abrogated antitumor activity of SN6j and antitumor effects of SN6j on tumor growth and survival were not evident in CD4-depleted mice (Figs. 5c and 5d). In contrast, SN6j showed significant suppressive activity against tumor growth in BALB/c mice and improved survival of the tumor-bearing mice before CD4-depletion (p < 0.05; Figs. 5c and 5d). On the other hand, tumor growth was significantly slower in the CD4-depleted mice than in nondepleted mice in the control IgG-treated group (p < 0.05; Fig. 5c). The results suggest that CD4-depletion enhances antitumor immunity in the mice. This enhancement of antitumor immunity is probably attributable to a consequence of depletion of CD4+, CD25+, Fox3+ regulatory T cells.33

Lastly, we investigated effects of CD8+ T cells alone on the tumor growth and antitumor efficacy of SN6j. SN6j prolonged survival of tumor-bearing mice (Fig. 5f; p < 0.05) and appears to suppress tumor growth (Fig. 5e; p = 0.09). These antitumor effects of SN6j were abrogated in CD8-depleted mice (Figs. 5e and 5f). On the other hand, CD8-depletion resulted in a significant increase in the size of tumors compared with nondepleted mice in the control IgG-treated group (p < 0.05; Fig. 5e). In addition, CD8-depletion significantly shortened survival time of tumor-bearing mice in the control IgG-treated mice (p < 0.05; Fig. 5f). Neither suppressive activity against tumor growth nor improvement of survival was observed by injections of SN6j into CD8-depleted mice (Figs. 5e and 5f). The results indicate an important role of CD8+ T cells in mAb-based EDG-targeted tumor therapy. Taken together, our results indicate that T cell immunity, especially CD8+ T cells, play an important role in the effective antibody-based EDG-targeted tumor therapy in immunocompetent animals.

Discussion

In the present study, cross-reactivity of anti-human EDG mAb SN6j with murine endothelial cells was confirmed by two different assays. One is binding and internalization of FITC-labeled SN6j into murine endothelial cells and the other is inhibition of angiogenesis by SN6j in mice. In the first assay, FITC-labeled SN6j is visually observed in the murine endothelial cells after binding and internalization (Fig. 1). Previously we used flow cytometry (FACS) analysis to test reactivities of two cross-reactive anti-human EDG mAbs SN6j and SN6k with mouse endothelial cells.26 The results showed that these mAbs reacted with proliferating mouse endothelial cells more strongly than with resting mouse endothelial cells.26 Therefore, proliferating mouse endothelial cells are used in the present assay (Fig. 1). In the second assay, SN6j inhibited tumor angiogenesis in Matrigel plug assay in mice. The tumors treated with SN6j showed a significant decrease in MVD in tumors and the remaining tumor vessels showed shrinkage or attenuation (Fig. 2). These data indicate that SN6j effectively targets tumor angiogenesis in mice.

In our earlier studies, we reported that immunoconjugates (immunotoxins and/or radioimmunoconjugates) containing certain anti-EDG mAbs showed antiangiogenic activity by inhibiting tumor growth13, 27 and vascular targeting activity by inducing regression of established tumors.26 For the sake of simplicity, we use the term antiangiogenesis for both antiangiogenic activity and vascular targeting activity in this manuscript. Antibody-drug conjugates targeting EDG or immunoliposomes containing Fv fragment of an anti-EDG mAb may also be useful for targeting tumor vasculature.34, 35 In an additional study in our laboratory, selected naked (unconjugated) anti-EDG mAbs (i.e., SN6f, SN6j and SN6k) suppressed growth of established tumors of MCF-7 human breast cancer cells that were grown in human skin/SCID mouse chimeras.7 The tumor growth was supported by a mixture of murine blood vessels and human blood vessels. Antitumor effects of these mAbs on the MCF-7 tumors were primarily attributable to the effects on human vessels in the tumors.7

Despite the observed antiangiogenic activity of the naked anti-EDG mAbs, underlying mechanisms by which these mAbs exert antiangiogenic activity is poorly understood.

Several mechanisms may potentially contribute to antiangiogenic/antitumor activity of anti-EDG mAbs. These include (i) binding of a mAb to EDG of proliferating endothelial cells of angiogenic vessels in tumors and subsequent induction of down-stream signaling events that may lead to growth suppression of the endothelial cells, (ii) binding of an anti-EDG mAb to endothelial cells in the tumor vessels may induce apoptosis of the endothelial cells and consequently may kill the tumor cells, (iii) interaction of the Fc region of EDG-bound mAb with Fcγ receptors on effector cells, leading to antibody-dependent cell mediated cytotoxicity (ADCC), (iv) Fc-mediated complement activation leading to target cell lysis (complement-dependent cytotoxicity, CDC), and (v) target antigen-specific CD4+ and/or CD8+ T cell immunity induced by antigen presenting dendritic cells (cross-presentation). Recently we showed that certain anti-EDG mAbs termed SN6 series mAbs were able to suppress proliferation of human endothelial cells in vitro without any accessory cells.36 However, little is known about the mechanisms of this suppression. Our current in vitro study shows that SN6j induces apoptosis of HUVECs. This induction of apoptosis may be one of the underlying mechanisms by which SN6j suppresses growth of proliferating endothelial cells. The potential involvement of other mechanisms in the suppression of endothelial cells by anti-EDG mAbs is under study in our laboratory.

With regard to antiangiogenic activity of SN6j in vivo, we hypothesized that T cell immunity may be important for antibody-based antiangiogenic therapy. This hypothesis was conceived based on our observation that anti-EDG mAb was more effective for tumor suppression in BALB/c mice than in SCID mice. Induction of such T cell immunity may be achieved in BALB/c mice but not in SCID mice. To test this hypothesis, we investigated the in vivo antitumor effect of SN6j combined with CpG ODN which is known as a strong immune activator and to induce Th1-responses in immunocompetent mice.30 An important immunostimulatory property of CpG ODN will be to activate antigen-presenting cells (APCs), including macrophages, monocytes and dendritic cells (DCs), to produce Th1-promoting cytokines, especially IL-12 and IFN-γ, which are hallmarks of an immune response to CpG-ODN. IL-12 can promote Th1 cell immune response and stimulate T cells and natural killer (NK) cells to secrete IFN-γ.37 The activated NK cells also release IL-12, which together with IFN-γ, in turn further activate NK cells, APCs and T cells and enhance their immune activity.38 Although CpG ODN have been reported to be effective for tumor immunotherapy either alone or in combination with other therapies, including tumor vaccines, antitumor antibodies, chemotherapy and other immunotherapies,30, 39 little is known about whether CpG can be an effective adjuvant for the antibody-based antiangiogenic therapy of tumors. In this study, we show that CpG ODN synergistically enhances antitumor efficacy of SN6j against established tumors in immunocompetent mice. However, CpG did not augment the antitumor efficacy of SN6j in SCID mice although SCID mice possess effector cells including NK cells, macrophages and monocytes. These results suggest that activation of T cell immunity by cross-presentation of mAb-coated antigen through DCs may be more important than other mechanisms (e.g., ADCC, or CDC) in the CpG-mediated antiangiogenic therapy of tumors. It may be possible that enhancement of the immunity with CpG is required because cross-reactivity of SN6j with mouse EDG is weak. In this regard, we are in the process of generating transgenic mice that express human EDG. Our recent study using slow growing s.c. tumors of MCF-7 (human breast carcinoma cells) in SCID mice showed that SN6j significantly suppressed tumor growth (unpublished data). In addition, recently we showed that efficacy of antiangiogenic therapy of tumors was influenced by tumor growth sites.28 Recently, we found that two different types of tumors form in mice when tumor cells are injected to make s.c. tumors, i.e., SS (skin side) type and MS (muscle side) type tumors (see Ref.28 and above Material and methods section). These tumors show different growth rates and different responses to therapy. We selected the SS type tumor in this study. Thus, antitumor efficacy of SN6j in mice will be significantly influenced by immune status of the tumor-bearing host, aggressiveness of the tumors, tumor growth sites and probably other factors. Understanding and analyses of these factors will be important for successful application of SN6j and other antiangiogenic agents to therapy of cancer.

In the study of BALB/c mice bearing colon-26 s.c. tumors, antitumor effects of SN6j on growth of established tumors and improvement of survival were abrogated when CD4+ and/or CD8+ T cells were depleted. Taken together, T cell immunity, especially CD8+ T cells, plays an important role in antibody-based EDG-targeted tumor therapy.

Recently, it was reported by Dhodapkar et al.40, 41 and others42 that cross-presentation of mAb coated-antigens by DCs to T cells may contribute to effective antibody-based tumor therapy. In this study, we show importance of T cell immunity for the mAb-based antiangiogenic therapy of tumors. Interestingly, our results showed that the growth rate of tumors in CD4-depleted mice was lower than that in nondepleted mice (Fig. 5c), whereas tumors grew faster in CD8-depleted mice than in non-depleted mice (Fig. 5e). In other animal models, CD25-depletion shortly before tumor challenge increased antitumor immunity.43, 44 Therefore, CD4+ T cell depletion in the present study was probably accompanied by depletion of CD4+ CD25+ Fox3+ regulatory T cells that may have induced stronger antitumor immunity in the CD4-depleted mice. When CD4-depletion was combined with CD8-depletion, either effect of CD4-depletion or CD8-depletion on tumor growth was not evident (Fig. 5a) which suggest that the two opposite effects neutralized each other.

Acknowledgements

The authors thank Dr. Xinwei She for help in the apoptosis assay and Dr. Maurice Barcos for expert advice concerning tissue histology.

Ancillary