Targeted Delivery of Saporin Toxin by Monoclonal Antibody to the Transcobalamin Receptor, TCblR/CD320 Free

Cellular uptake of vitamin B12 (cobalamin; Cbl) is mediated by the transcobalamin receptor (TCblR/CD320) on the plasma membrane that specifically binds transcobalamin (TC), a plasma protein saturated with Cbl (1–3). The TCblR protein is structurally related to the LDLR family with two LDLR type A domains separated by a cysteine-rich CUB-like domain and contains a single transmembrane region followed by a short cytoplasmic tail (4). The expression of this receptor seems to be cell cycle–associated with highest expression in actively proliferating cells and is substantially downregulated in quiescent cells (5–7). This differential expression of the receptor serves to provide optimum delivery of the vitamin to cells during the early phase of DNA synthesis. This process ensures adequate functioning of Cbl-dependent enzymes, especially the methionine synthase that is essential for recycling of methyl folate to generate folates needed for purine and pyrimidine biosynthesis (8). The more proliferative a cell, the higher the need for folates and Cbl, and this need for Cbl is met by the increased expression of TCblR in cancer cells that may have inherently lost the ability to stop dividing and differentiate. Selective targeting of cancer cells for destruction by delivering drugs and toxins preferentially to these cells has been the ultimate objective of cancer therapy. The search for tumor-specific markers and the strategies to utilize these in cancer therapy have been pursued for decades with mixed results. This can be attributed to multiple factors that include the lack of specificity of the target antigen, cellular events that can alter the targeting, and the complex and diverse nature of cancer itself. Recently, potent toxins, such as ricin A chain, gelonin, saporin, cholera toxin, diphtheria toxin, and pseudomonas exotoxin, have been utilized to target these molecules to tumors as conjugates of antibodies or as chimeric proteins with varying success (9, 10). We have utilized the cell cycle–associated expression of TCblR to deliver saporin, an inhibitor of ribosomal assembly (11) to cancer cells with the use of monoclonal antibodies (mAb) to the extracellular domain of TCblR.

mAbs were generated to the recombinant extracellular domain of TCblR expressed in human embryonic kidney stem cells (HEK-293) and purified as described previously (4). Purified antibodies were used to study the delivery of saporin-conjugated goat anti-mouse IgG secondary antibody (Advanced Targeting Systems) to various cell lines maintained in culture.

We used K562 (ATCC CCL 243) human erythroleukemia cells and U266 (ATCC TIB 196) human myeloma that propagate as a suspension culture; SW48 (ATCC CCL-231) human colon adenocarcinoma cells and KB (ATCC CCL-17) human epidermoid carcinoma that propagate as adherent cells; and HEK-293 (ATCC CRL-1573). These cell lines were obtained from the American Type Culture Collection Center and have been identified by karyotyping. Second-passage and third-passage cells were frozen in aliquots and used for these studies. MCH 064, MCH 065, and RF peripheral skin fibroblast cultures in passages 9 to 12 were from The Repository for Mutant Human Cell Strains. Fresh human bone marrow mononuclear cells were obtained from Lonza. In addition, HEK-293 cells stably transfected with the cDNA for TCblR in pcDNA 3.1 to overexpress the receptor were used (HEK-293TR). These cells have ∼10-fold higher expression of TCblR constitutively driven by the cytomegalovirus (CMV) promoter, and therefore, TCblR expression is not cell cycle associated. All cell lines were maintained in DMEM supplemented with 10% fetal bovine serum and antibiotics. The primary antibody was incubated with the saporin-conjugated secondary antibody for 1 hour at room temperature to form the complex before use in the culture.

Binding and internalization of antibody directed to TCblR was determined in HEK-293 cells that were engineered to express a green fluorescent protein (GFP) tag in the cytoplasmic end of TCblR. For this determination, mAb1-25 was preincubated with quantum dot (qdot) 625–conjugated goat anti-mouse IgG secondary antibody (Invitrogen) for 60 minutes to form a complex and then incubated with cells in culture.

The specificity of mAb binding was tested in K562 cells that expressed the native TCblR. The binding and internalization of mAb1-19-qdot 625 complex was determined at 4°C and 37°C, and normal mouse IgG was used as a negative control.

To determine the optimum concentration of mAb for the in vitro cell-kill studies, mAb1-25 was preincubated with saporin-conjugated goat anti-mouse IgG secondary antibody at a 1:1 molar ratio for 60 minutes to form a complex. Various concentrations of this mAb/saporin antibody complex (0.1 pmol/L to 50 nmol/L) were incubated with 10,000 SW48 (colon carcinoma) or K562 (erythroleukemia) cells in 96-well culture plates for 72 hours, and viable cells were quantified by the MTS assay (Promega).

For using the anti-TCblR antibodies as carrier of saporin into cells through TCblR, the optimum ratio of primary mAb to saporin-conjugated secondary antibody was determined. SW48 colon carcinoma cells were seeded at a density of 10,000 cells per well in 96-well culture plates. In one set of experiments, the concentration of primary mAb was varied from 0.078 to 80 nmol/L, whereas the concentration of secondary antibody was kept constant at 10 nmol/L. In another set of experiments, the concentration of the primary mAb was kept constant at 2.5 nmol/L, and the concentration of saporin antibody was varied from 10 to 40 nmol/L. Cell viability was determined after 72 hours by the MTS assay.

Seeding density defines the proliferative phase of the culture, and therefore, cells seeded at lower density would replicate longer until the cell population reaches confluency. Because TCblR expression is highest in actively dividing cells, we tested cell lines at seeding densities varying from 1,000 to 10,000 cells per well with three different primary antibodies at 2.5 nmol/L mAb and 10 nmol/L saporin antibody concentration.

Because the toxic effect of saporin antibody was more pronounced in cell cultures seeded at lower density, the IC₅₀ determinations were done with cells seeded at 1,000 cells per well in 96-well plates, at a mAb–saporin antibody ratio of 1:4 and a primary antibody concentration range of 0.046 to 2.5 nmol/L. Viable cells were determined by the MTS assay after 96 hours in culture.

The specificity of the TCblR-mediated pathway for internalization of the mAb/saporin antibody toxin complex was determined by adding soluble receptor to the culture medium. The soluble receptor would compete with the cell surface receptor for the antibody, and this would reduce the antibody toxin available for cellular uptake, resulting in a decrease in percent inhibition. For this experiment, SW48 cells were seeded in 96-well plates at 2,000 cells per well, and the amount of mAb/saporin antinbody used was equivalent to the IC₅₀.

A 100-fold excess primary mAb or normal mouse IgG was added to the incubation medium containing a mAb/saporin antibody concentration of 2.5/10 nmol/L. A decrease in the anti-TCblR mAb/saporin antibody–induced inhibition of cell growth should be observed when excess primary antibody is present because the ratio of saporin antibody–labeled primary mAb to unlabelled primary mAb should be lower, and this increases the probability of unlabelled mAb binding to TCblR. The addition of normal mouse IgG should also result in a decrease in cell kill because the secondary saporin antibody would also bind to the normal mouse IgG, and this complex cannot bind to TCblR.

Cells stably transfected to express a chimeric TCblR with GFP tagged to the cytoplasmic end of the receptor show discrete membrane-associated fluorescence. As shown in Fig. 1A, binding of the mAb1-25-qdot 625red complex at 4°C was restricted to surface receptors, as indicated by mostly membrane-associated diffused fluorescence dispersed throughout the periphery of the cell. Binding and internalization of the mAb1-25-qdot 625red complex bound to TCblR occurred at 37°C as indicated by segregation of receptors to discrete regions in the membrane as well as the cytoplasm, indicated by colocalization of the red and green fluorescence (Fig. 1B). Similar binding and internalization was observed with mAb1-19-qdot 625red complex when incubated with K562 cells expressing normal levels of non-GFP native TCblR (Fig. 1D). The specificity of binding and internalization was confirmed by substituting normal mouse IgG for the primary antibody and incubating with K562 cells, which failed to show any binding and internalization of qdot 625red (Fig. 1C).

Initial titration of the mAb/saporin antibody complex at 1:1 molar ratio incubated with SW48 or K562 cells for 72 hours indicated that a primary antibody concentration of 2 to 5 nmol/L was adequate for testing the ability of these antibodies to deliver saporin toxin to cancer cells.

To determine the optimum ratio of saporin-conjugated secondary antibody to the anti-TCblR primary mAb, in the first set of experiments the saporin antibody concentration was kept constant at 10 nmol/L, and the concentration of primary mAb was varied from 0.078 to 80 nmol/L. In the second set of experiments, the concentration of the primary mAb was kept constant at 2.5 nmol/L and the concentration of saporin-antibody was varied from 10 to 40 nmol/L. All three anti-TCblR mAbs tested yielded similar results, with 2.5 nmol/L concentration of the primary mAb as most effective in inhibiting cell growth (Fig. 2A). The addition of 3.7 nmol/L holo-TC to the culture did not have a significant effect on the outcome although minor differences were observed due to TC-Cbl competing with the antibody for uptake. The concentration of TC-Cbl used was in excess of the normal 0.5 to 1.5 nmol/L in plasma. Increasing the concentration of saporin antibody did not produce any increase in cell death (Fig. 2B). Thus, a mAb concentration of 2.5 nmol/L and a saporin antibody concentration of 10 nmol/L (i.e., a mAb–saporin antibody ratio of 1:4) seemed optimum for delivering antibody toxin into cells through the TCblR pathway.

HEK-293 cells stably transfected to overexpress TCblR were highly sensitive to mAb/saporin antibody, resulting in >90% inhibition of cell growth. The expression of TCblR in these cells is driven by the CMV promoter and is not dependent on the cell cycle or proliferative state of the cell. The relationship between TCblR expression and the effect of the mAb-saporin conjugate was evident when normal HEK-293 cells and HEK-293TR cells overexpressing TCblR were seeded at varying densities and exposed to mAb-saporin. Whereas the effect of the toxin decreased with increasing cell density for normal HEK-293 cells, >80% of the HEK-293TR cells were killed at all cell densities (Fig. 3A). Thus, the effectiveness of the toxin was directly related to the level of receptor expression. For all cell lines tested in which the receptor expression is related to the proliferative state of the cells in culture, most effects were observed when seeded at lower density, as shown for U266 cells in Fig. 3B. All three mAbs tested yielded similar results. For comparing the effective delivery of toxin, IC₅₀ determinations were done and are shown in Fig. 4 for two suspension cultures and two adherent cell lines. For most cell lines, the IC₅₀ was in the 0.625 to 2.5 nmol/L range for the primary mAb concentration.

The specificity of the TCblR pathway was evident from the decreased effect of the toxin when soluble recombinant TCblR was added to the culture medium. Soluble receptor in the culture medium would compete with the cell surface receptor for the antibody and reduce the antibody toxin available for cellular uptake, resulting in a decrease in percent inhibition as shown in Fig. 5A. The specificity of mouse anti-TCblR mAb for delivering the saporin antibody toxin was determined by adding either a 100-fold excess primary mAb or normal mouse IgG to the incubation medium. The inhibition of cell growth seen with anti–TCblR mAb1-10/saporin antibody was decreased when either excess primary antibody or normal mouse IgG was added. Primary mAb or normal mouse IgG in the absence of saporin antibody had no effect on cell growth (Fig. 5B). The growth inhibitory effect of TCblR mAb/saporin antibody complex seemed to be specific for tumor cell lines because three different normal human fibroblast cell lines were not inhibited under the identical culture conditions (Fig. 5C). The effect on the bone marrow cells seemed to be considerably less when maintained in DMEM compared with MarrowMAX medium (Gibco) but far less than the inhibition observed with many tumor cell lines (Fig. 5D).

Since the identification of cancer-specific antigens and mAbs, the goal of therapeutic developments has been to target the cancer cell for destruction without affecting the normal cellular components. This simple concept has proven to be difficult to execute because of the complex and diverse nature of cancer and the biological adaptations that escape treatment modalities. It is now recognized that multipronged approaches tailored to specific cancers and tissue types may be needed to succeed in devising treatments that are cancer-specific and less toxic. The therapeutic potential of mAbs as immune modulators or as carriers of drugs or toxins is being explored as engineered antibodies that are well tolerated are being developed (12, 13). The use of potent toxins, such as ricin, cholera toxin, gelonin, and saporin, if targeted specifically to cancer cells, could be highly effective in destroying these cells (9, 10, 14, 15). This requires a tumor cell–specific target antigen that the mAb can recognize and that can deliver the toxin to the cell. Numerous receptors have been explored as target antigens with some degree of success (16, 17). However, these antigens are also expressed on normal cells, and therefore, selective targeting of cancer cells is dictated primarily by differential expression of the target receptor. The generalized toxicity of the conjugate will depend on the number of receptors expressed on each cell type; therefore, receptors that are highly expressed in normal cells, although overexpressed in cancer cells, are not likely to be selective and effective targets because normal cells would also internalize sufficient toxin for destruction. The ideal target antigen would have to be tumor-specific or be expressed at such low levels in normal cells that it would render the dose ineffective. This same receptor, if overexpressed in cancer cells, would provide a target antigen that could be selective for the tumor cells. The TCblR expression is fairly low with only a few thousand receptors expressed in actively replicating cells during the S phase of the cell cycle (5–7). Due to the proliferative nature of neoplastic cells, this expression is 5-fold to 10-fold upregulated and sustained in certain cancer cells. This differential expression seems to be sufficient to provide the selectivity needed to target the cancer cell for destruction while sparing the normal cell population. This conclusion is based on the decreased effect on cells seeded at higher density, which do not divide at the same rate as cells seeded at lower density, and the virtual sparing of normal cells, which did divide but failed to internalize sufficient toxin. Cbl is essential for the recycling of methylfolate, the major form of folate in the body. Thus, Cbl deficiency produces metabolic folate deficiency that results in defective DNA synthesis and megaloblastic changes in the bone marrow (8). Blocking of Cbl uptake into cells by mAbs to TC has been used to show the utility of this strategy in cancer therapy (18, 19). An antibody to TCblR that blocks the binding of TC-Cbl can also prevent cellular uptake of TC-Cbl and could be used to block Cbl uptake into cancer cells. These approaches, although feasible with virtually no direct systemic toxicity, are not very practical because of the long treatment schedule required to deplete intracellular Cbl. The use of the TC-TCblR pathway to deliver toxins and drugs provides the dose needed to rapidly destroy highly proliferative tumor cells. The cell cycle–associated TCblR expression and the high level of expression in certain cancer cells versus normal cells provide the selectivity and specificity demanded of this targeting strategy. In utilizing these antibodies in vivo in humans, bone marrow toxicity is a potential concern. The low toxicity observed for these cells in DMEM and the slightly higher effect seen in MarrowMAX medium is related to the proliferation and differentiation of marrow cells in the latter. On average, TCblR expression in embryonic stem cells is similar to that observed in skin fibroblasts (4). Thus, the proliferative state of the cells and TCblR expression would dictate marrow toxicity. Although the effect on the marrow in vitro seems low, far greater toxicity is likely in vivo. Decreasing cell surface receptors by predosing with B₁₂ or with antibody without toxin are strategies likely to decrease cell surface receptors to minimize toxicity.

Patent application WO2007/117657 A2.

Grant Support: NIH grant R01DK064732 and KYTO Biopharma.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.