The blood group-related Lewis Y antigen is expressed on the majority of human cancers of epithelial origin with only limited expression on normal tissue. Therefore, the Lewis Y antigen represents an interesting candidate for antibody-based treatment strategies. Previous experiments showed that the humanized Lewis Y-specific monoclonal antibody, IGN311, reduced ErbB-receptor-mediated stimulation of mitogen-activated protein kinase by altering receptor recycling. Here, we tested whether binding of IGN311 to growth factor receptors is relevant also to inhibition of tumor growth in vivo. Prolonged incubation with IGN311 of human tumor cell lines, which express high levels of ErbB1 (A431) or ErbB2 (SK-BR-3), resulted in down-regulation of the receptors and inhibition of cell proliferation. IGN311 inhibited the growth of tumors derived from A431 cells xenografted in nude mice. Treatment with IGN311 was associated with a down-regulation of ErbB1 in the excised tumor tissue. Importantly, these effects of IGN311 were also mimicked by the Fab fragment of IGN311. These data indicate that tumor cell growth inhibition by IGN311 cannot solely be accounted for by invoking cellular and humoral immunological mechanisms. A direct effect on signaling via binding to Lewis Y glycosylated growth factor receptors on tumor cells is also likely to contribute to the therapeutic effect of IGN311 in vivo.
The Lewis Y (LeY) antigen is a blood group-related difucosylated oligosaccharide with the chemical structure Fucα1 → 2Galβ1 → 4[Fucα1 → 3]GlcNAcβ1 → R. Whereas LeY is expressed at high levels during embryogenesis, its expression on adult tissue is restricted to epithelial surfaces and activated granulocytes. However, the vast majority of human carcinomas shows an overexpression of the LeY antigen (Hellstrom et al., 1990), either as a glycolipid in the cell membrane or linked to cell surface proteins (Basu et al., 1987).
Growth factor receptors have been selected as targets for anti-tumor therapy using monoclonal antibodies against their extracellular domains. The best-studied example is the ErbB-receptor family. The aberrant expression of ErbB receptors tyrosine kinases has been implicated in tumor growth and progression (Yarden and Sliwkowski, 2001). This family includes receptors of the epidermal growth factor (ErbB1), ErbB2, and the heregulin-receptors ErbB3 and ErbB4. Upon ligand binding, these receptors may homo- or heterodimerize, which leads to phosphorylation of intracellular tyrosine residues. These result in the recruitment of Src homology 2 domain containing adaptor proteins that transduce the signal to mitogen-activated protein kinases (MAPK) and phosphoinositide 3-kinase (Ullrich and Schlessinger, 1990). The use of trastuzumab, a humanized monoclonal antibody directed against ErbB2, represented a major break-through in the therapy of breast cancer. Nevertheless, targeting growth factor receptors can also lead to inhibition of physiological responses elicited by these receptors. In the case of trastuzumab, inhibition of cardiac ErbB2 may cause heart failure (Slamon et al., 2001).
The frequent overexpression of Lewis Y on cancer cells has stimulated the testing of monoclonal anti-LeY antibodies in animal cancer models (Clarke et al., 2000; Scott et al., 2000; Wahl et al., 2001). The observed growth inhibition has been attributed to recruitment of effector functions such as antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) (Co et al., 1996). Recently we have shown that the Lewis Y-specific humanized monoclonal antibody IGN311 inhibited EGF-mediated signaling in the epidermoid cancer cell line A431 and in the breast cancer cell line SK-BR-3 (Klinger et al., 2004). The effect of IGN311 (or its murine parent antibody ABL364) was not due to competition with binding of EGF to the receptor but rather to altered kinetics of EGF-receptor recycling. In the present work, we extended these in vitro investigations and tested whether this mechanism may also contribute to inhibition of tumor growth in vivo. The observations show that, upon prolonged exposure to IGN311, tumor cells down-regulate the expression of ErbB-receptors in vitro as well as in vivo in a xenograft tumor model. These data and the efficacy of Fab fragments of IGN311 suggest that inhibition of tumor growth is, at least in part, due to an effect of IGN311 on signaling via binding to Lewis Y glycosylated growth factor receptors present on cancer cells.
Materials and Methods
Materials. Fetal calf serum was purchased from PAA Laboratories (Linz, Austria), Dulbecco's modified Eagle medium, RPMI 1640 medium, nonessential amino acids, and β-mercaptoethanol were obtained from GIBCO-BRL (Grand Island, NY). A431 and SK-BR-3 cell lines were purchased from ATCC (Manassas, VA). SK-BR-5 cells were kindly provided by Novartis Institute for Biomedical Research (Vienna, Austria). The anti-EGF receptor antibody Ab-2 (clone 225) was from NeoMarkers (Fremont, CA). The antibodies recognizing ErbB1 and ErbB2 were from Cell Signaling Technologies Inc. (Beverly, MA), the antibody against cytokeratin 8 was from Abcam (M20 ab 9023-1; Cambridge, UK), and the antiserum recognizing the carboxyl terminus of erk1/erk2 was from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-Lewis murine monoclonal antibody ABL364 was provided by Novartis (Basel, Switzerland), and its humanized version, IGN311, was produced under GMP conditions by BioInvent (Lund, Sweden). Trastuzumab (Herceptin) was a gift from Roche (Basel, Switzerland).
Cell Culture. A431, SK-BR-3, and SK-BR-5 cells were propagated in Dulbecco's modified Eagle medium at 5% CO2 and 37°C supplemented with 10% fetal calf serum, 4 mM l-glutamine, 100 U/ml penicillin G, and 100 μg/ml streptomycin. For cell proliferation assays, A431, SK-BR-5, or SK-BR-3 cells (105 cells/dish) were plated into 3-cm culture dishes in the absence or presence of indicated concentrations of antibody. Media were changed daily with the addition of fresh antibodies. After 24, 48, and 72 h, the cells were detached with trypsin and counted.
Determination of ErbB Receptor Levels in Cell Lysates and in Tumor Tissues Samples. Cell culture conditions were as outlined above (105 cells/3-cm dish). Media were changed daily with fresh addition of antibodies. After washing with PBS, cells were lysed in lysis buffer (20 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM Na3VO4, 40 mM β-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 10 mM NaF, 1% Triton X-100, 250 U/ml aprotinin, and 40 μg/μl leupeptin; pH adjusted to 7.5 with HCl). The cellular debris was removed by centrifugation at 10,000g for 10 min, and the total protein content was measured photometrically using the bicinchoninic acid method (Micro-BCA kit; Pierce Chemical, Rockford, IL). Equal amounts of protein (20 μg) were resolved on SDS polyacrylamide gels. Subsequently, proteins were transferred to nitrocellulose membranes, which were probed with antibodies against ErbB1 and ErbB2; immunoreactive bands were visualized by chemoluminescence using a secondary antibody coupled to horseradish peroxidase. The loading control was done by staining for p42/p44 MAPK (erk1/erk2).
ErbB1 levels in the excised xenograft tumors (see below) were determined as described here. The frozen tissue samples (see below) were pounded with a mortar; the resulting powder was resuspended in two volumes of buffer (20 mM Tris-HCl, pH 7.4, 1% SDS); and insoluble material and debris were removed by centrifugation at 10,000g for 10 min. The amount of material applied onto SDS-polyacrylamide gels (∼5% of original tissue) was normalized by determining the content of cytokeratin-8 in each sample. Immunoblotting for cytokeratin-8 and ErbB1 was done as outlined above. Immunoreactive bands were quantitated by densitometry using the Scion Image software (Scion Corporation, Frederick, MD).
Animal Maintenance and Xenograft Inoculation. The experiments were approved by the Animal Welfare Committee of the Medical University of Vienna; animals were maintained according to the standards set forth by the Good Scientific Practice regulations of the Medical University of Vienna. Female nude mice (Balb/c nu/nu) were purchased from the Institut für Versuchstierzucht Himberg (Vienna, Austria); they were typically 5 to 6 weeks old. The animals were allowed to adapt for 1 week under sterile conditions (Seal-Safe-IVC-Cages; Techniplast, Munich, Germany). On the 1st experimental day, A431 cells (1 to 3 × 106 suspended in 0.2 ml phosphate-buffered saline) were injected subcutaneously. Before injection, the presence of Lewis Y-antigen on the cell surface was verified by fluorescence-activated cell sorting. On the next day, the appropriate antibody or vehicle was injected intraperitoneally: The injection of antibody was repeated every 72 h. The tumor size was measured, and the tumor volume was computed from the formula for an ellipsoid body (volume = length × breadth × height × π /6), where for small tumors, height (= thickness) was assumed to correspond to the smaller of the two diameters. At the end of the experiment, mice were sacrificed, and tumors were excised, shock-frozen in liquid N2, and stored at –80°C for further analysis.
Preparation of Fab. Antibody IGN311 (63 mg) was dialyzed against 20 mM sodium phosphate buffer (pH 7.0) supplemented with 10 mM EDTA. Aliquots of 0.5 ml (containing 5 mg of IGN311) were added to 0.5 ml of immobilized papain (Immuno Pure Fab-Kit; Pierce Biotechnology) equilibrated against digestion buffer cysteine-monohydrochloride/phosphate buffer, pH 7.0, for 24 h at 37°C. Fab fragment was purified by negative protein G affinity chromatography (HiTrap; Amersham Biosciences, Piscataway, NJ) after separation of immobilized Papain. Flow-through was concentrated to 3.6 mg/ml by Macrosep OMEGA 30K (Pall Life Sciences, Ann Arbor, MI) against PBS. Cleavage efficiency and fragment purity of >95% were confirmed by SDS-polyacrylamide gel electrophoresis and size-exclusion chromatography.
Statistical Analysis. Statistical comparisons were done by ANOVA followed by Tukey's multiple comparison test using the statistical package implemented in the GraphPad Prism software (GraphPad Software, Inc., San Diego, CA).
IGN311 Inhibits Cell Growth of A431 and SK-BR-3 Cells. The anti-Lewis Y antibodies IGN311 and ABL364 have been shown to block signaling via binding to Lewis Y-glycosylated ErbB receptors and other cell surface receptors (Klinger et al., 2004). In vivo, tumor cells may be stimulated by a cocktail of growth factors rather than by a single paracrine factor. We mimicked this situation by allowing the Lewis Y-positive cancer cell line A431 or SK-BR-3 to grow in the presence of serum and by testing the ability of IGN311 to block serum-driven growth. When added at 100 nM, IGN311 caused growth inhibition that was only observed after a delay (Fig. 1A). In subsequent experiments, the response to the antibodies was therefore assessed at ≥72 h. IGN311 and ABL364 exert their half-maximal inhibitory effect on EGF-dependent signaling at ∼10 nM (Co et al., 1996), which is consistent with the affinity of these antibodies for their respective epitope Lewis Y (Klinger et al., 2004). Accordingly, there was no appreciable inhibition of cell proliferation at 1 nM IGN311 (full triangles in Fig. 1C). In contrast, trastuzumab (Herceptin) caused ∼50% of its antiproliferative effect at 1 nM (empty triangles in Fig. 1C), a concentration that is in the range of its affinity for ErbB2. When employed at the saturating concentration of 100 nM, IGN311 and trastuzumab inhibited growth to a comparable extent (Fig. 1, C and D); there was not any additional inhibition if trastuzumab and IGN311 were combined (full squares in Fig. 1C). The data in Fig. 1C were obtained with SK-BR-3 cells; in contrast, trastuzumab did not inhibit the growth of A431 cells (Fig. 1D). A431 cells express large amounts of ErbB1 but only very small amounts of ErbB2; thus, in A431 cells, the efficacy of IGN311 was compared with that of the anti-erbB1-antibody 2C225 (the parent antibody of cetuximab). At 100 nM, IGN311 and 2C225 were equieffective.
IGN311 Down-Regulates ErbB1 Receptors in A431 Cells. The mechanism of action that is generally considered relevant for clinical action of trastuzumab is down-regulation of ErbB2 (Baselga et al., 2001). We recently showed that IGN311 alters the kinetics of ErbB receptor recycling. This change in the cycle of endocytosis and externalization is causally related to suppression of EGF-dependent stimulation of MAPK (Klinger et al., 2004). However, this blockage of signal transduction occurs on a short time scale. In particular, it is not clear whether these early inhibitory actions can be directly linked to the growth inhibitory effect of the anti-LeY antibody, which required several days to become manifest (see Fig. 1A). Thus, we asked whether prolonged treatment with IGN311 altered the levels of ErbB receptors. A431 or SK-BR-3 cells, respectively, were treated for 72 h with 100 nM IGN311 (or its murine counterpart ABL364) as well as control antibodies (trastuzumab and 2C225). Cellular lysates were prepared, and ErbB receptors were detected by immunoblotting. Incubation in the presence of IGN311 and of ABL364 led to a substantial reduction of ErbB1 and ErbB2 protein levels in both A431 (Fig. 2A, top) and SK-BR-3 cell lines (Fig. 2B, top), respectively. The anti-epidermal growth factor receptor antibody (2C225) was more effective in reducing ErbB1 expression in A431 cells (Fig. 2A, top) but failed to alter the levels of ErbB2 in SK-BR-3 cells (Fig. 2B, top). In contrast and as expected, trastuzumab caused down-regulation of ErbB2 levels in SK-BR-3 cells (Fig. 2B, top). It is noteworthy that the down-regulation of ErbB2 induced by IGN311 or by ABL364 was comparable in magnitude to that induced by trastuzumab (Fig. 2B). Finally, the changes observed cannot be attributed to variations in the amount of protein that were loaded in individual lanes, because we detected comparable levels of MAPK in all samples; this determination was done with the lower part of the same gel (Fig. 2, A and B, bottom).
IGN311 Inhibits Growth of A431 Xenografts in Nude Mice. IGN311 inhibited serum-induced growth of cancer cells and caused down-regulation of ErbB-receptors under cell culture conditions. We tested whether both phenomena can be reproduced in vivo, which would further indicate a causal relation. When injected subcutaneously into nude mice, SK-BR-3 cells failed to grow and give rise to measurable tumors. Hence, the in vivo experiments were limited to A431 cells. In the first experiment, we injected a tumor cell load of 1 × 106 A431 cells/nude mouse. The experiment was terminated when the tumor volume was estimated to be larger than 1 ml. In untreated controls, this threshold was reached 25 days after injection. When administered intraperitoneally at 10 mg/kg and every 3rd day, IGN311 (open circle in Fig. 3A) and ABL364 (closed triangle in Fig. 3A) inhibited tumor growth to a similar extent. We verified that the antibodies were also efficacious at a higher tumor cell load by injecting 3 × 106 A431 cells (Fig. 3, B and C). In this case, the latency was shortened, and the tumor volume reached the threshold of 1 ml after 16 days in the control animals. Figure 3B also shows that IGN311 was more effective at the highest dose of 30 mg/kg compared with the lower dose of 10 mg/kg (also see Fig. 3D). In contrast, although ABL364 suppressed growth of the xenograft when applied at 10 mg/kg (open circle in Fig. 3C), growth was actually enhanced at the higher dose of 30 mg/kg; some of the animals developed large tumors on day 12. This is not readily evident from Fig. 3C (closed triangle), because all animals were sacrificed on this day and thus not subjected to further evaluation. We observed previously that, at high concentrations, ABL364 was capable per se of inducing MAPK stimulation; this effect was thought to result from the fact that ABL364 is an IgG3 isotype. Hence, it tends to aggregate, and this may support cross-linking and thus activation of growth factor receptors (Greenspan and Cooper, 1992; Klinger et al., 2004). This mechanism may also be responsible for the stimulation of xenograft growth that we observed at 30 mg/kg ABL364.
Estimation of tumor growth by determination of tumor volume may not always be completely accurate, because it assumes that the shape of all tumors are approximated by an ellipsoid body; tumors also may differ with respect to the depth at which they infiltrate the subcutaneous tissue, and the height of tumors is notoriously difficult to measure. Therefore, we also assessed the effect of IGN311 on tumor weight as a more objective parameter (Fig. 3D). The data were pooled from four independent experiments; in each of these, at least two different treatment groups were compared with a (vehicle-treated) control group. The number of injected tumor cells varied between 1 × 106 and 3 × 106 cells in these experiments. This resulted in variability of tumor weight. To account for this variability, the average control value was set 100% for each individual experiment, and the weight of each tumor was related to this reference value. It is evident from Fig. 3D that IGN311 caused a dose-dependent reduction in tumor weight; i.e., IGN311 did not cause any appreciable effect at 3 mg/kg, but an intermediate effect and maximal efficacy were seen at 10 and 30 mg/kg, respectively.
IGN311 Exerts Effects That Are Not Dependent on Its Effector Functions. When tested with human peripheral mononuclear effector cells, IGN311 supports a higher ADCC than the parental murine antibody ABL364 (Co et al., 1996). The efficacy of ADCC in a given model, however, eventually depends on the degree of fitting between the humanized or murine antibody Fc part with the respective Fcγ receptors on the effector cells of the host. There was no appreciable difference between the growth of the tumors in animals treated with 10 mg/kg ABL364 or with 10 mg/kg IGN311 (Fig. 3D). Furthermore, IGN311 inhibited proliferation of tumor cells under cell culture conditions in the absence of any Fcγ receptor-bearing effector cells or complement source (Fig. 1D), indicating that this effect cannot be explained by ADCC or CDC. If IGN311 exerted its effect, at least in part, in a manner independent on effector functions, i.e., ADCC or CDC, a Fab fragment should be also active in the tumor xenograft model. We verified that the Fab of IGN311 was also capable of reducing cell growth in vitro. Under conditions comparable with those outlined in Fig. 1, 300 nM were equieffective to 100 nM IGN311 (data not shown). Next, we compared the efficacies of the Fab fragment of IGN311 and uncleaved IGN311 in vivo by injecting nude mice with 3 × 106 A431 cells and treating them with either IGN311 (10 mg/kg) and Fab (20 mg/kg). As shown in Fig. 3D, 20 mg/kg of its Fab exerted a growth inhibitory effect that was similar in magnitude to that elicited by 10 mg/kg IGN311. Similarly, when monitored by tumor volume based on the measured dimensions, the time course of tumor growth was comparable in mice treated with IGN311 or the Fab fragment thereof (data not shown). Although this observation does not rule out effector functions as a mechanism of action, it clearly shows that there are effects other than antibody-dependent cellular cytotoxicity or complement-dependent cytotoxicity that contribute to the action of IGN311.
Prolonged treatment with trastuzumab predisposes patients to develop heart failure (Slamon et al., 2001); signaling via ErbB2 is also required to maintain normal cardiac performance, because tissue-specific deletion of ErbB2 in the adult mouse heart results in dilated cardiomyopathy (Ozcelik et al., 2002). We did not observe any macroscopic change in the heart and lung that was suggestive of dilated cardiomyopathy. Likewise, the ratio heart and lung weight to body weight was not affected by treatment with IGN311 (data not shown).
IGN311 Down-Regulates ErbB1 Receptor Expression in Vivo. In vitro, prolonged incubation of A431 and SK-BR-3 cells with IGN311 down-regulated the levels of ErbB1 and ErbB2, respectively (Fig. 2). If the action of IGN311 in vivo was also related in part to binding of LeY antigen carrying ErbB receptors, IGN311 should also down-regulate ErbB1 in A431 xenografts. This prediction was verified by comparing the levels of ErbB1 in xenograft tumors excised from control animals and from animals treated with IGN311 or its Fab fragment. Tumor lysates were prepared, and receptor expression was determined by immunoblotting. ErbB1 protein levels were markedly reduced in tumors from mice that had been treated with 10 and 30 mg/kg IGN311 or with 20 mg/kg Fab (Fig. 4, top blots). The tumors excised from treated and control animals may differ in their cellular composition, particularly the relative proportion of stromal cells. Hence, we used cytokeratin 8 as a loading control (Fig. 4, bottom blots). This protein is only expressed in cells of epithelial origin. Hence, in the subcutaneous tissue, the xenografted A431 cells can be considered the only candidate source of cytokeratin 8. The levels of cytokeratin 8 can therefore serve as a marker for the proportion of tumor cell proteins in the total lysate. As can be seen in Fig. 4, the levels of cytokeratin 8 were reasonably comparable in all samples. Thus, we rule out that the differences in ErbB1 levels are due to variation in the amount of tumor cell protein applied to individual lanes.
Although growth factor receptors have been proposed as targets for tumor therapy more than two decades ago (Masui et al., 1984), breakthroughs have been only achieved in recent years, and the concept has been vindicated by the success of trastuzumab (Herceptin) (Slamon et al., 2001). We have developed a Lewis Y-specific humanized monoclonal antibody IGN311, which recently has been tested in a Phase I dose escalation study in biopsy-proven Lewis Y-positive carcinoma patients (D. V. Oruzio, G. Waxenecker, C. Aulmann, M. Sandherr, G. C. Mudde, R. Kircheis, H. Loibner, and G. Schlimok, submitted for publication). There are several reasons for choosing the LeY antigen as a target for therapeutic antibodies. Lewis Y has been found to be expressed on the majority of human cancers of epithelial origin (Blaszczyk-Thurin et al., 1987). Expression on normal tissue is limited to epithelial surfaces and activated granulocytes (Dettke et al., 2001). In contrast, LeY was found to be present on up to 70 to 90% adenocarcinomas of the lung, breast, colorectal, gastric, pancreatic, prostate, and ovarian cancers (Sakamoto et al., 1986; Inagaki et al., 1990; Steplewski et al., 1990; Murata et al., 1992). Furthermore, the levels of Lewis Y expression have been shown to correlate with survival in patients with lung carcinoma (Miyake et al., 1992).
There are several lines of experimental evidence for proof of principle for Lewis Y-targeted antibody therapy. Anti-LeY antibodies have been used for the delivery of chemotherapeutic drugs (e.g., doxorubicin; Wahl et al., 2001) or used in radioactive form (131I-labeled anti-LeY antibodies to enhance the effect of Taxol tumor chemotherapy; Clarke et al., 2000). The humanized anti-LeY antibody 3S193 was also shown to exert growth inhibitory effects on MCF7 cells xenografted into nude mice (Scott et al., 2000). Antibody-dependent cellular cytotoxicity and complement-dependent cytotoxicity have been proposed as the mechanism of action (Co et al., 1996; Klinger et al., 2004). Activation of CDC and ADCC has been demonstrated ex vivo in patients' serum after intravenous infusion of IGN311 and was found to correlate with the kinetics of IGN311 levels in the serum (D. V. Oruzio, G. Waxenecker, C. Aulmann, M. Sandherr, G. C. Mudde, R. Kircheis, H. Loibner, and G. Schlimok, submitted for publication). Here we provide evidence for an additional mechanism of action, namely blockage of growth factor receptor-dependent signaling. This interpretation is based on the following observations. (i) Growth factor receptors, including ErbB1 and ErbB2 are Lewis Y-glycosylated. (ii) In vitro, IGN311 caused growth inhibition with a delayed onset, which is inconsistent with cell death due to any residual complement-dependent toxicity. Furthermore, Fab prepared from IGN311 also inhibited cell growth in vitro, indicating that the Fc-portion was dispensable for growth inhibition under cell culture conditions. (iii) The addition of IGN311 (and its murine counterpart) to tumor cells caused loss of ErbB receptors. This reduction in ErbB1 was also seen in xenograft tumors of recipient mice that had been treated with IGN311 (or alternatively its Fab fragment). Thus, long-term exposure to IGN311 can cause down-regulation of Lewis Y antigen-modified cell surface receptors. Loss of growth factor receptors is likely to contribute to an antiproliferative effect as is the previously documented blockage by IGN311 of growth factor-induced signaling (Co et al., 1996). (iv) Nude mice, while lacking T-cell mediated immunity, still have NK cells, monocytes, and complement, which are candidate effectors by which IGN311 may blunt xenograft growth in nude mice. However, the fact that the Fab fragment of IGN311 also inhibited xenograft growth is difficult to reconcile with ADCC. Therefore, we conclude that a direct action on growth factor receptors contributes to the antitumor effect of IGN311 in vivo.
The precise mechanism by which long-term exposure to IGN311 causes down-regulation of ErbB receptors is not known. However, our earlier work demonstrated that short-term exposure of cells to IGN311 (and its murine counterpart ABL364) altered the kinetics of ErbB1 receptor recycling and affected the distribution of the internalized receptors (Klinger et al., 2004). It is likely that these short-term effects, particularly the altered intracellular compartmentalization of the receptors, are also causally related to long-term down-regulation. Trastuzumab (Herceptin) is also thought to elicit its therapeutic effect, at least in part, by down-regulating ErbB2 (Baselga et al., 2001); the underlying mechanism still remains a matter of controversy (Austin et al., 2004).
The concentration of IGN311 at its site of action is not known; however, it can be roughly estimated based on the following considerations: 10 mg/kg corresponds to 0.2 mg/20-g mouse or 1.3 nmol/mouse; the systemic bioavailability of intraperitoneally administered antibody is ∼0.3 (Flessner, 2001). The intravascular space in a 20-g mouse is ∼1 ml. The half-life of a humanized IgG1 in mice is in the range of 3 to 7 days (Bazin et al., 1994; Vieira and Rajewsky, 1998). If, for the sake of simplicity, the half-life of IGN311 is assumed to be equal to the dosing interval (that is 3 days), the steady-state plasma concentration is ∼0.3 μM. The permeability of the vasculature within the tumor is not known; if it corresponds to that of normal subcutaneous tissue (O'Connor and Bale, 1984), the concentration of IGN311 in the extracellular space is ∼10% of that in the plasma, i.e., ∼30 and ∼100 nM upon repeated administration of 10 and 30 mg/kg IGN311. It is obvious that this is a rough estimate because of several uncertainties (in particular, the permeability of the capillaries in the tumor). Nevertheless, the concentration is within the range in which the antibody was active in inhibiting cell growth in vitro. We note that the effect of 20 mg/kg IGN311 Fab fragments was comparable in magnitude with that of 10 mg/kg IGN311. In other words, on a molar basis, the IGN311 was more potent than the Fab fragments thereof. However, this does not necessarily argue against our interpretation for the following reasons. First, in vitro, the potency of IGN311 also exceeded that of the Fab fragment. More importantly, the pharmacokinetics of IgG1 and of Fab fragments prepared thereof have been reported to differ substantially, with considerably longer half-life of IgG1 compared with its Fab fragments (Covell et al., 1986). However, Fab fragments have a larger volume of distribution; i.e., they permeate better into the interstitial fluid and do so more rapidly. Thus, it is not surprising that, on a molar basis, higher amounts of Fab fragments have to be administered to elicit the same effect as that of IGN311. Finally, the quantitative differences seen between whole IGN311 and its Fab when calculated at the molar basis may reflect the contribution of the effector functions, i.e., ADCC and CDC, in the overall antitumor efficacy of IGN311 in this model.
In conclusion, the presented data indicate that interfering with cell signaling via binding to Lewis Y-glycosylated growth factor receptors present on cancer cells may be one of the mechanisms for the antitumor activity of the Lewis Y-specific humanized monoclonal antibody IGN311 in vivo.
We thank Novartis and Roche for generous gifts of antibodies.
This work was supported by a grant from Igeneon GmbH (to M.F.).
H.F. and C.S. contributed equally to this work.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: LeY, Lewis Y antigen; ADCC, antibody-dependent cellular cytotoxicity; CDC, complement-dependent cytotoxicity; MAPK, mitogen-activated protein kinase; PBS, phosphate-buffered saline.
- Received May 5, 2006.
- Accepted September 6, 2006.
- The American Society for Pharmacology and Experimental Therapeutics