Search

We report the preparation and characterization of a novel, fully human antitumor immunoRNase (IR). The IR, a human RNase and fusion protein made up of a human single chain variable fragment (scFv), is directed to the ErbB-2 receptor and overexpressed in many carcinomas. The anti-ErbB-2 IR, named hERB-hRNase, retains the enzymatic activity of the wild-type enzyme (human pancreatic RNase) and specifically binds to ErbB-2-positive cells with the high affinity (K(d) = 4.5 nm) of the parental scFv. hERB-hRNase behaves as an immunoprotoxin and on internalization by target cells becomes selectively cytotoxic in a dose-dependent manner at nanomolar concentrations. Administered in five doses of 1.5 mg/kg to mice bearing an ErbB-2-positive tumor, hERB-hRNase induced a dramatic reduction in tumor volume. hERB-hRNase is the first fully human antitumor IR produced thus far, with a high potential as a poorly immunogenic human drug devoid of nonspecific toxicity, directed against ErbB-2-positive malignancies.

HuN901 is a humanized monoclonal antibody that binds with high affinity to CD56, the neuronal cell adhesion molecule. HuN901 conjugated with the maytansinoid N(2')-deacetyl-N(2')-(3-mercapto-1-oxopropyl)-maytansine (DM1), a potent antimicrotubular cytotoxic agent, may provide targeted delivery of the drug to CD56 expressing tumors. Based on gene expression profiles of primary multiple myeloma (MM) cells showing expression of CD56 in 10 out of 15 patients (66.6%) and flow cytometric profiles of MM (CD38(bright)CD45(lo)) cells showing CD56 expression in 22 out of 28 patients (79%), we assessed the efficacy of huN901-DM1 for the treatment of MM. We first examined the in vitro cytotoxicity and specificity of huN901-DM1 on a panel of CD56(+) and CD56(-) MM cell lines, as well as a CD56(-) Waldenstrom's macroglobulinemia cell line. HuN901-DM1 treatment selectively decreased survival of CD56(+) MM cell lines and depleted CD56(+) MM cells from mixed cultures with a CD56(-) cell line or adherent bone marrow stromal cells. In vivo antitumor activity of huN901-DM1 was then studied in a tumor xenograft model using a CD56(+) OPM2 human MM cell line in SCID mice. We observed inhibition of serum paraprotein secretion, inhibition of tumor growth, and increase in survival of mice treated with huN901-DM1. Our data therefore demonstrate that huN901-DM1 has significant in vitro and in vivo antimyeloma activity at doses that are well tolerated in a murine model. Taken together, these data provide the framework for clinical trials of this agent to improve patient outcome in MM.

We have used the Eos Hu03 GeneChip array, which represents over 92% of the transcribed human genome, to measure gene expression in a panel of normal and diseased human tissues. This analysis revealed that E-selectin mRNA is selectively overexpressed in prostate cancer epithelium, a finding that correlated strongly with E-selectin protein expression as assessed by immunohistochemistry. Antibodies against E-selectin that blocked function failed to impede cancer cell growth, suggesting that overexpression of E-selectin was not essential for cell growth. However, a novel auristatin E-based antibody drug conjugate (ADC), E-selectin antibody valine-citrulline monomethyl-auristatin E, was a potent and selective agent against E-selectin-expressing cancer cell lines in vitro, with the degree of cytotoxicity varying with surface antigen density. Interestingly, sensitivity to the ADC differed among cell lines from different tissues expressing similar amounts of E-selectin and was found to correlate with sensitivity to free auristatin E. Furthermore, E-selectin-expressing tumors grown as xenografts in severe combined immunodeficient mice were responsive to treatment with E-selectin antibody valine-citrulline monomethyl-auristatin E in vivo, with more than 85% inhibition of tumor growth observed in treated mice. These findings demonstrate that an E-selectin-targeting ADC has potential as a prostate cancer therapy and validates a genomics-based paradigm for the identification of cancer-specific antigens suitable for targeted therapy.

A number of different immunotoxins composed of cell-specific targeting structures coupled to plant or bacterial toxins have increasingly been evaluated for immunotherapy. Because these foreign proteins are highly immunogenic in humans, we have developed a new CD30 ligand-based fusion toxin (Ang-CD30L) using the human RNase angiogenin. The completely human fusion gene was inserted into a pET-based expression plasmid. Transformed Escherichia coli BL21(DE3) were grown under osmotic stress conditions in the presence of compatible solutes. After isopropyl beta-D-thiogalactoside induction, the M(r) 37,000 His(10)-tagged Ang-CD30L was directed into the periplasmic space and functionally purified by a combination of metal ion affinity followed by enterokinase cleavage of the His(10)-Tag and molecular size chromatography. The characteristics of the recombinant protein were assessed by ELISA, flow cytometry, and toxicity assays showing specific activity against CD30(+) Hodgkin-derived cells. Specific binding activity of Ang-CD30L was verified by competition with anti-CD30 monoclonal antibody Ki-4 and commercially available CD30L-CD8 chimeric protein. Ang-CD30L showed RNase activity in vitro. The human recombinant immunotoxin showed significant toxicity toward several CD30-positive cell lines (HDLM-2, L1236, KM-H2, and L540Cy) and exhibited highest cytotoxicity against L540 cells (IC(50) = 8 ng/ml) as determined by cell proliferation assays. CD30 specificity was confirmed by competitive toxicity assays. This is the first report on the specific cytotoxicity of a recombinant completely human fusion toxin with possibly largely reduced immunogenicity for the treatment of CD30-positive malignancies.

A theoretical drawback to alpha-particle therapy with 213Bi is the short range of the particle track coupled with the short half-life of the radionuclide, thereby potentially limiting effective cytotoxicity to rapidly accessible, disseminated individual tumor cells (e.g., as in leukemia). In this work, a prostate carcinoma spheroid model was used to evaluate the feasibility of targeting micrometastatic clusters of tumor cells using 213Bi-labeled anti-prostate-specific membrane antigen (PSMA) antibody, J591. In prostate cancer, vascular dissemination of tumor cells or tumor cell clusters to the marrow constitutes an important step in the progression of this disease to widespread skeletal involvement, an incurable state. Such prevascularized clusters are ideal targets for radiolabeled antibodies because the barriers to antibody penetration that are associated with the capillary basal lamina have not yet formed. Beta- and gamma-emitting radionuclides such as 131I, which are widely used in radioimmunotherapy, are not expected to be effective when targeting single cells or small cell clusters. This is because the range of the emissions is one to two orders of magnitude greater than the target size, and the energy deposited per traversal is insufficient to produce any significant radiobiological effect. Spheroids of the prostate cancer cell line, LNCaP-LN3, were used as a model of prevascularized micrometastases; their response to an anti-PSMA antibody, J591, radiolabeled with the alpha-particle emitter 213Bi (T(1/2), 45.6 min.) has been measured. The time course of spheroid volume reductions was found to be sensitive to the initial spheroid volume. J591 labeled with 0.9 MBq/ml 213Bi resulted in a 3-log reduction in spheroid volume on day 33, relative to control, for spheroids with an initial diameter of 130 microm; 1.8 MBq/ml were required to achieve a similar response for spheroids with an initial diameter of 180 microm. Equivalent spheroid responses were observed after 12 Gy of acute external beam photon irradiation. Monte Carlo-based microdosimetric analyses of the 213Bi decay distribution in individual spheroids of 130-microm diameter yielded an average alpha-particle dose of 3.7 Gy to the spheroids, resulting in a relative biological effectiveness factor of 3.2 over photon irradiation. The activity concentrations used in the experiments were clinically relevant, and this work supports the possibility of using 213Bi-labeled antibodies not only for disseminated single tumor cells, as found in patients with leukemia, but also for micrometastatic tumor deposits up to 180 microm in diameter (1200 cells).

A novel alpha-particle emitting monoclonal antibody construct targeting the external domain of prostate-specific membrane antigen (PSMA) was prepared and evaluated in vitro and in vivo. The chelating agent, N-[2-amino-3-(p-isothiocyanatophen-yl)propyl]-trans-cyclohexane-1, 2-diamine-N,N',N',N'',N''-pentaacetic acid, was appended to J591 monoclonal antibody to stably bind the 213Bi radiometal ion. Bismuth-213 is a short-lived (t 1/2 = 46 min) radionuclide that emits high energy alpha-particles with an effective range of 0.07-0.10 mm that are ideally suited to treating single-celled neoplasms and micrometastatic carcinomas. The LNCaP prostate cancer cell line had an estimated 180,000 molecules of PSMA per cell; J591 bound to PSMA with a 3-nM affinity. After binding, the radiolabeled construct-antigen complex was rapidly internalized into the cell, carrying the radiometal inside. [213Bi]J591 was specifically cytotoxic to LNCaP. The LD50 value of [213Bi]J591 was 220 nCi/ml at a specific activity of 6.4 Ci/g. The potency and specificity of [213Bi]J591 directed against LNCaP spheroids, an in vitro model for micrometastatic cancer, also was investigated. [213Bi]J591 effectively stopped growth of LNCaP spheroids relative to an equivalent dose of the irrelevant control [213Bi]HuM195 or unlabeled J591. Cytotoxicity experiments in vivo were carried out in an athymic nude mouse model with an i.m. xenograft of LNCaP cells. [213Bi]J591 was able to significantly improve (P < 0.0031) median tumor-free survival (54 days) in these experiments relative to treatment with irrelevant control [213Bi]HuM195 (33 days), or no treatment (31 days). Prostate-specific antigen (PSA) was also specifically reduced in treated animals. At day 51, mean PSA values were 104 ng/ml +/- 54 ng/ml (n = 4, untreated animals), 66 ng/ml +/- 16 ng/ml (n = 6, animals treated with [213Bi]HuM195), and 28 ng/ml +/- 22 ng/ml (n = 6, animals treated with [213Bi]J591). The reduction of PSA levels in mice treated with [213Bi]J591 relative to mice treated with [213Bi]HuM195 and untreated control animals was significant with P < 0.007 and P < 0.0136, respectively. In conclusion, a novel [213Bi]-radiolabeled J591 has been constructed that selectively delivers alpha-particles to prostate cancer cells for potent and specific killing in vitro and in vivo.

Adenovirus (Ad) have been used as vectors to deliver genes to a wide variety of tissues. Despite achieving high expression levels in vivo, Ad vectors display normal tissue toxicity, transient expression, and antivector immune responses that limit therapeutic potential. To circumvent these problems, several retargeting strategies to abrogate native tropism and redirect Ad uptake through defined receptors have been attempted. Despite success in cell culture, in vivo results have generally not shown sufficient selectivity for target tissues. We have previously identified (C. K. Goldman et al., Cancer Res., 57: 1447-1451, 1997) the fibroblast growth factor (FGF) ligand and receptor families as conferring sufficient specificity and binding affinity to be useful for targeting DNA in vivo. In the present studies, we retargeted Ad using basic FGF (FGF2) as a targeting ligand. Cellular uptake is redirected through high-affinity FGF receptors (FGFRs) and not the more ubiquitous lower-affinity Ad receptors. Initial in vitro experiments demonstrated a 10- to 100-fold increase in gene expression in numerous FGFR positive (FGFR+) cell lines using FGF2-Ad when compared with Ad. To determine whether increased selectivity could be detected in vivo, FGF2-Ad was administered i.v. to normal mice. FGF2-Ad demonstrates markedly decreased hepatic toxicity and liver transgene expression compared with Ad treatment. Importantly, FGF2-Ad encoding the herpes simplex virus thymidine kinase (TK) gene transduces Ad-resistant FGFR+ tumor cells both ex vivo and in vivo, which results in substantially enhanced survival (180-260%) when the prodrug ganciclovir is administered. Because FGFRs are up-regulated on many types of malignant or injured cells, this broadly useful method to redirect native Ad tropism and to increase the potency of gene expression may offer significant therapeutic advantages.

Tumor pretargeting with biotinylated antibodies and avidin, followed by a delayed delivery of radioactive-labeled biotin, is currently used for in vivo diagnosis and therapy in cancer patients. Herein, we describe the use of a three-step antibody/avidin targeting approach to increase the local concentration and the persistence of biotinylated human tumor necrosis factor alpha (bio-TNF) on a mouse tumor. Mouse RMA lymphoma cells were transfected with the Thy 1.1 allele (RMA-Thy 1.1) to generate a unique tumor-associated antigen. In vitro pretargeting of RMA-Thy 1.1 cells with the biotinylated anti-Thy 1.1 monoclonal antibody 19E12 (bio-19E12) and NeutrAvidin increased the amount of bio-TNF that bound to the cell (10-20 times in comparison with non-pretargeted cells), as well as its half-life on the surface (>30 times). Furthermore, cell pretargeting reduced by more than 2 orders of magnitude the LD50 of bio-TNF in a cytolytic assay with actinomycin D. Finally, RMA-Thy 1.1 cells, pretreated in vitro with bio-TNF according to the three-step procedure and injected into syngeneic C57/BL6 mice, were less tumorigenic than controls. These results indicate that the three-step targeting approach markedly increases the amount and the persistence of bio-TNF on the cell surface and that cell-bound bio-TNF can trigger cytolytic effects in vitro and antitumor effects in vivo. Tumor pretargeting with biotinylated antibodies and avidin could be a novel strategy for increasing the therapeutic index of TNF.

The use of tumor necrosis factor alpha (TNFalpha) in cancer therapy is limited by its short circulatory half-life and its severe systemic side effects. To overcome these limitations, we evaluated the capability of a bispecific antibody (BAb) directed against carcinoembryonic antigen (CEA) and human TNFalpha to target this cytokine in tumors. A BAb was constructed by coupling the Fab' fragments from an anti-CEA monoclonal antibody (MAb) to the Fab' fragments from an anti-TNFalpha MAb via a stable thioether linkage. The double specificity of the BAb for CEA and TNFalpha was demonstrated using a BIAcoreTM two-step analysis. The affinity constants of the BAb for CEA immobilized on a sensor chip and for soluble TNFalpha added to the CEA-BAb complex were as high as those of the parental MAbs (1.7 x 10(9) M-1 and 6.6 x 10(8) M-1, respectively). The radiolabeled 125I-labeled BAb retained high immunoreactivity with both CEA and TNFalpha immobilized on a solid phase. In nude mice xenografted with the human colorectal carcinoma T380, the 125I-labeled BAb showed a tumor localization and biodistribution comparable to that of 131I-labeled anti-CEA parental F(ab')2 with 25-30% of the injected dose (ID)/g tumor at 24 h and 20% ID/g tumor at 48 h. To target TNFalpha to the tumor, a two-step i.v. injection protocol was used first, in which a variable dose of 125I-labeled BAb was injected, followed 24 or 48 h later by a constant dose of 131I-labeled TNFalpha (1 microg). Mice pretreated with 3 microg of BAb and sacrificed 2, 4, 6, or 8 h after the injection of TNFalpha showed a 1.5- to 2-fold increased concentration of 131I-labeled TNFalpha in the tumor as compared to control mice, which received TNFalpha alone. With a higher dose of BAb (25 microg), mice showed a better targeting of TNFalpha with a 3.2-fold increased concentration of 131I-labeled TNFalpha in the tumor: 9.3% versus 2.9% ID/g in control mice 6 h after TNFa injection. In a one-step injection protocol using a premixed BAb-TNFalpha preparation, similar results were obtained 6 h postinjection (3.5-fold increased TNFalpha tumor concentration). A longer retention time of TNFalpha was observed leading to an 8.1-fold increased concentration of TNFalpha in the tumor 14 h postinjection (4.4 versus 0.5% ID/g tumor for BAb-treated and control mice, respectively). These results show that our BAb is able, first, to localize in a human colon carcinoma and, there, to immunoabsorb the i.v.-injected TNFalpha, leading to its increased concentration at the tumor site.

Previously, a quantitative analysis that related protein synthesis inhibition of transferrin-toxin conjugates to the cellular trafficking of transferrin was proposed (P.T. Yazdi and R. M. Murphy, Cancer Res., 54: 6387-6394, 1994). Here, this work is extended to evaluate cellular trafficking of anti-transferrin receptor antibodies and protein synthesis inhibition kinetics of immunotoxins constructed from the same antibodies and the toxin gelonin. Cellular trafficking models for two monoclonal anti-transferrin receptor antibodies (5E9 and OKT9) in HeLa cells were developed. The two mAbs had similar trafficking parameters, which differed significantly from those for transferrin. Protein synthesis inhibition kinetics for immunotoxins constructed from 5E9 or OKT9 and gelonin were measured. Analysis of the data using our previously proposed relationship between protein synthesis and cellular trafficking indicated that the relationship is also valid for these new systems. The protein synthesis inhibition constants for 5E9-gelonin and OKT9-gelonin conjugates were similar to those for the transferrin-gelonin conjugate. These results suggest that it may be possible to predict the efficacy of gelonin immunotoxins from knowledge of the trafficking of the corresponding targeting agent. A sensitivity analysis showed which cellular trafficking parameters have the greatest influence on immunotoxin efficacy and are, therefore, the most likely to be profitably manipulated.

The effect of using a "residualizing" iodine radiolabel, dilactitol-iodotyramine, for radioimmunolocalization of antibodies to tumors was investigated. This tracer is designed to be lysosomally trapped after catabolism of the labeled antibody. mAbs RS7 and RS11 were used for in vivo and in vitro studies on the uptake and retention of radioisotope into tumor cells. Both are murine IgG1 mAbs with pancarcinoma reactivity, which react with integral membrane glycoproteins. mAb RS7 has been shown to be relatively rapidly catabolized by the antigen-bearing cell line Calu-3, whereas RS11 is catabolized more slowly in the same cells. An 111In- or 88Y-p-isothiocyanatobenzyl-diethylenetriamine pentaacetic acid conjugate was also tested because these radiometals are known to be lysosomally trapped, and iodination via chloramine T was used to provide a baseline. In vitro, a substantial increase in retention of the label by cells was observed when the dilactitol-tyramine DLT- or 111In-labeled mAbs were used, and the improvement gained by the use of these residualizing labels was greater with the use of the rapidly catabolized mAb (RS7) than it was with the more slowly catabolized mAb (RS11). In biodistribution studies in nude mice bearing Calu-3 tumor xenografts, a dramatic improvement in the tumor accretion of the radiolabel was seen with the use of the 131I-labeled DLT- or 88Y-labeled mAbs. For example, at day 7 the percentage of injected dose/g in the tumor was 5.54 +/- 1.47% (SD), 38.06 +/- 8.04%, and 43.18 +/- 19.50% for the conventionally iodinated, DLT- and 88Y-labeled RS7, respectively. Dosimetry calculations performed on the biodistribution data predict increases of approximately 8- and 4-fold in the absorbed dose to tumor with the use of 131I-labeled DLT- and 90Y-labeled mAbs, respectively, compared to the conventional 131I. In contrast to in vitro findings, these results were similar for both RS7 and RS11, suggesting that the use of DLT may be advantageous for most of the mAbs binding to the cell surface, including antibodies that are catabolized relatively slowly. The advantage of 131I-labeled DLT over 90Y is due to the longer physical half-life of the 131I.

We have constructed anti-CD40 immunotoxins consisting of the single chain Fv (sFv) region of the anti-human CD40 mAb G28-5 fused to a truncated form of Pseudomonas exotoxin, PE40. CD40 is an integral membrane glycoprotein found on the surface of B-lineage cells, including lymphomas and leukemias, as well as certain carcinomas. Two forms of the immunotoxin were constructed, one with the light chain variable (VL) region of the sFv preceding the heavy chain variable region (VH) [G28-5 sFv(VL-VH)-PE40] and the second with the sFv in the opposite orientation [G28-5 sFv(VH-VL)-PE40]. Although both forms of G28-5 sFv-PE40 specifically bound to CD40 in ELISAs, the binding of G28-5 sFv(VL-VH)-PE40 was > 10-fold higher. A number of malignant B- and T-cell lines were screened for CD40 expression and susceptibility to G28-5 sFv(VL-VH)-PE40. All of the B-lineage cells tested were CD40 positive and sensitive to the anti-CD40 immunotoxin, with EC50s ranging from 2.5-70 ng/ml, whereas none of the T cell leukemias or lymphomas were antigen positive or were affected by the immunotoxins. Consistent with the antigen-binding results, the VL-VH immunotoxin was > 10-fold more cytotoxic than the VH-VL immunotoxin. The anti-CD40 single-chain immunotoxin fusion protein G28-5 sFv(VL-VH)-PE40 represents a potent and specific cytotoxic agent for the elimination of normal and transformed B-lineage cells expressing CD40.

A mathematical model was developed which relates the time and concentration dependence of protein synthesis inhibition of an immunotoxin to the properties of the targeting agent and the conjugated toxin. The role of the targeting agent and that of the toxin in determining the cytotoxicity were separated in this model by describing protein synthesis inhibition as a function of a cellular trafficking variable, which is calculated from the trafficking parameters of the targeting agent, and a protein synthesis inhibition constant, which is a property of the translocation and enzymatic rate constants of the toxin. Transferrin cellular trafficking parameters were determined experimentally for HeLa and SK-MEL-2 cells. Protein synthesis inhibition of transferrin-gelonin and transferrin-CRM107 conjugates in both cell lines was measured as a function of time and concentration. Analysis of the data showed that the model was a good representation of the experimental results, and correctly explained cell line differences in sensitivity to transferrin-toxin conjugates. The translocation rate constant for transferrin-CRM107 was approximately 3000 times greater than that for transferrin-gelonin. The model may be useful in understanding the factors that influence immunotoxin efficacy and in designing more lethal immunotoxins.

We analyzed the effect of isotype variation on effectiveness of B-cell CD19 immunotoxins (IT) by using class switch variants (CLB-B4-IgG1 and CLB-B4-IgG2a) conjugated to ricin A. Notably, IgG1-IT appeared to be approximately 100-fold more potent than IgG2a-IT toward B-cell lines Daudi and KM3. Binding and internalization studies with 125I-labeled monoclonal antibodies (mAbs) revealed a higher cellular uptake of IgG1 compared to IgG2a, despite similar binding affinities. Following removal of the Fc part, both mAbs internalized at the same rate as IgG2a, indicating that the Fc part of IgG1 is involved in enhanced cellular uptake. The involvement of Fc gamma RII (CD32) in this process was demonstrated by a decreased cytotoxicity of IgG1-IT (and not IgG2a-IT) in the presence of Fc gamma RII-blocking mAbs AT10 or IV.3. To identify the isoform responsible for this phenomenon, internalization of IgG1 and IgG2a in 11 B-cell lines and malignant B-cells of 8 patients was compared with expression of Fc gamma RII subclasses. In addition to four cell lines (Daudi, KM3, Nalm6, and Raji), the malignant B-cells of two patients showed enhanced uptake of IgG1 relative to IgG2a. Only the Fc gamma RIIa transcript was found in all B-cells. Furthermore, enhanced uptake of IgG1 correlated with rosetting of erythrocytes sensitized with anti-glycophorin A mAb of IgG1 isotype rather than with Fc gamma RIIa membrane expression levels. These data support the idea that functional Fc gamma RIIa is involved in the enhanced IgG1 uptake observed in a subset of B-cells. Our study, therefore, points to an important role for the Fc region of IT in the delivery of cytotoxic effects.

Recombinant immunotoxins were made by fusing the Fab domain of monoclonal antibody (MAb) B3 to PE38M, a truncated mutant form of Pseudomonas exotoxin (PE). The recombinant toxins were made in Escherichia coli by fusing genes encoding the antibody domains to a gene encoding the mutant form of PE. MAb B3 binds to a carbohydrate antigen found on many kinds of carcinomas. Immunotoxins in which MAb B3 has been chemically coupled to recombinant forms of PE have been shown to be very active cytotoxic agents. PE has also been targeted to tumor cells by replacing the cell-binding domain of PE (domain I) with a single-chain antibody to make a single-chain immunotoxin. In the current study, PE38M, a mutant form of PE, with a deletion of the cell-binding domain (amino acids 1-252) as well as mutations in domain III and some nonessential sequences in domain Ib (amino acids 365-380), was fused to the light chain of MAb B3. This protein was renatured in the presence of the Fd fragment of MAb B3 to produce a Fab-toxin fusion protein. Alternatively, the Fd fragment of MAb B3 was fused to PE38M and combined with the light chain. Both types of B3(Fab)-PE38M were just as active on target cells as previously described single-chain immunotoxins. Furthermore, the B3(Fab)-PE38M produced complete remissions of human tumor xenografts growing in nude mice. B3(Fab)-PE38M has two advantages over single-chain immunotoxins. One is that the yield of recombinant Fab-toxin is very high, with 10-22% of the starting protein recovered as cytotoxically active immunotoxin after chromatographic purification. The second is that the B3(Fab)-PE38M has a much longer survival in the circulation of mice with a t1/2 beta of approximately 5 h.

Analysis of human fibrosarcoma cells exposed to radiolabeled monoclonal antibody 19-24, which recognizes sarcoma-associated antigen p102, revealed that over 54% of the cell surface-bound radioactivity was internalized. No modulation of cell surface p102 antigen by monoclonal antibody 19-24 was observed in human fibrosarcoma cells. Monoclonal antibody 19-24 coupled to daunomycin via a dextran bridge was found to be most effective. In different preparations, the daunomycin:total protein molar ratio ranged from 1.9 to 6.1. In vitro cytotoxicity studies using human fibrosarcoma cells showed that, at 10 micrograms/ml concentration, this immunoconjugate was 79.4% as efficient as free daunomycin and, at 1 microgram/ml concentration, 36.8% as efficient. Control nonspecific murine monoclonal antibody P3 immunoconjugates were relatively ineffective. The distribution of 14C-Adriamycin, 125I-labeled monoclonal antibody 19-24, and 125I-labeled 19-24 immunoconjugate was also evaluated over a 24-h period in tumor and normal tissues of athymic mice bearing a human fibrosarcoma xenograft. Poor uptake of radiolabeled Adriamycin by the tumor tissue was observed. The level of 14C radioactivity in the tumor tissue never exceeded 1% of the total injected dose and was 24.8-fold lower than the radioactivity found in the spleen tissue. Tumor tissue uptake of radiolabeled monoclonal antibody 19-24 was characterized by the high tumor tissue:blood ratio of 1.62 +/- 0.28 (SD). However, for monoclonal antibody 19-24 immunoconjugates, this ratio decreased to 0.66 +/- 0.05, which was still higher than normal (liver, 0.48 +/- 0.02; lung, 0.48 +/- 0.07; spleen, 0.28 +/- 0.01) or nonspecific monoclonal antibody P3 immunoconjugates (0.22 +/- 0.03). Thus, it appears that, compared to free daunomycin, monoclonal antibody 19-24 immunoconjugates may be more efficient and less cytotoxic to normal tissues.

Modification of proteins with monomethoxy-polyethylene glycol (mPEG) has been shown to prolong circulation time and to reduce immunogenicity. To make a mPEG-modified recombinant toxin that retained cytotoxic activity but had a longer residence time in circulation, we have constructed an altered form of TGF alpha-PE40, a recombinant toxin composed of human transforming growth factor alpha (TGF alpha) fused to a fragment of Pseudomonas exotoxin (PE38) devoid of its cell-binding domain. In the newly designed protein, termed TGF alpha R29-L2-CH2-PE38QQ delta (TCP), there are no lysine residues in the TGF alpha and PE38 portions. Human IgG4 constant region CH2 and a tetradecapeptide linker; L2, are inserted between TGF alpha and PE38. Together, L2 and CH2 contain 13 lysine residues as potential modification sites for mPEG. mPEG conjugates of TCP (PEG-TCP) were generated and the products were resolved by ion exchange chromatography. Two PEG-TCP species termed B4 and B6 retained 15 and 4% of cytotoxicity, respectively, and 26% of their receptor binding activity compared with the unmodified TCP. Both B4 and B6 had prolonged circulation times in the blood and reduced toxicity in animals. The mean residence times of B4 and B6 were 37 and 68 min, respectively, compared to 7 min for TCP. When administered i.v. to tumor bearing mice, both B4 and B6 produced marked antitumor effects whereas the unmodified TCP had none. Also, the immunogenicity of PEG-TCP was 5-10 times less than that of TCP. We suggest that the prolonged circulating time and reduced toxicity of PEG-TCP compensate for a diminished cytotoxic activity and enlarge significantly the therapeutic window of this chimeric toxin.

A panel of bispecific F(ab')2 antibodies (BsAb) have been constructed for delivering the ribosome-inactivating protein saporin to human B cell lymphoma. Each derivative was prepared with specificity for saporin and CD19, CD22, CD37, or immunoglobulin. In vitro studies measuring inhibition of [3H]leucine uptake by cultured Daudi and Raji cells demonstrated that, despite all BsAb capturing saporin on the cell surface, BsAb targeting through CD22 were far more cytotoxic than those functioning via CD19, CD37, or surface immunoglobulin. This exceptional activity of the CD22-specific BsAb appears to derive from its ability to deliver and accumulate saporin inside the target cells. Further studies showed that four CD22-specific BsAb all performed with equal potency and were able to increase saporin toxicity (50% inhibitory concentration) up to 1000-fold, from 2 x 10(-7) M to 2 x 10(-10) M. Pairs of anti-CD22 BsAb which recognized different nonblocking epitopes on the saporin molecule were able to bind saporin more avidly to the target cell and, as a consequence, increased cytotoxicity by at least an additional 10-fold, resulting in 50% inhibitory concentration for protein synthesis of 2 x 10(-11) M. These results suggest that selected combinations of BsAb which bind cooperatively to a toxin and the cell surface may provide an efficient way of delivering toxins to unwanted cells in patients.

Local hyperthermia and the hypoxic cytotoxin SR 4233 were administered to nude mice with 693 +/- 47 mm3 (mean +/- SE) s.c. HCT-8 human colonic adenocarcinoma xenografts in an attempt to enhance the antitumor effects of radioimmunotherapy. Biodistribution studies revealed preferential binding of NR-Lu-10, a murine monoclonal antibody, to the tumors compared with an isotype-matched control antibody, CCOO16-3.A single injection of 25 microCi 90Y-NR-Lu-10 significantly inhibited tumor growth (control versus 90Y-NR-Lu-10: P = 0.048). The administration of hyperthermia at 41.5 degrees C for 1 h immediately following the injection of 111In-labeled NR-Lu-10 up-regulated tumor-associated antigen expression and increased antibody uptake in the tumors by 73% (P = 0.001) without significantly affecting antibody uptake in normal tissues. However, the heat treatment did not produce a more homogeneous distribution of the antibodies in the tumors and did not significantly enhance the tumor growth delay produced by 90Y-NR-Lu-10 (P = 0.07). The administration of local hyperthermia at 43.0 degrees C for 1 h, on the other hand, had direct cytotoxic effects (P = 0.03) and enhanced the tumor growth delay produced by 90Y-NR-Lu-10 (P = 0.01). SR 4233 also enhanced the tumor growth delay produced by 90Y-NR-Lu-10 (P = 0.03). The greatest antitumor effects were observed when both hyperthermia at 43.0 degrees C and SR 4233 were administered in combination with 90Y-NR-Lu-10 (P = 0.002). No toxicity was produced by the local hyperthermia, and the only toxicities produced by 90Y-NR-Lu-10 and SR 4233 were neutropenia and weight loss.

A novel affinity system for a two-phase delivery of radionuclides to tumor cells has been developed. In the first phase, a nontoxic bivalent monoclonal antibody conjugated to an enzyme is targeted to the tumor cells. In the second phase, a radionuclide-derivatized enzyme inhibitor, specific for the enzyme conjugated to the antibody, is administered. The model system selected for this study is the recombinant human enzyme dihydrofolate reductase (rhDHFR) and its high-affinity competitive inhibitor methotrexate (MTX). MTX was labeled with a radionuclide by covalent attachment of diethylenetriaminepentaacetic acid (DTPA) complexed with 111In. Using the gamma-carboxyl residue of MTX for the attachment of DTPA, binding of the inhibitor to rhDHFR was not affected. The inhibitory activities of nonderivatized MTX and DTPA-MTX were indistinguishable. Human K562 erythroleukemia cells were used to evaluate under in vitro conditions the DHFR-MTX affinity system for the delivery of 111In-labeled DTPA-MTX to pretargeted alpha-transferrin receptor antibody-rhDHFR conjugates (alpha-TFR-DHFR). The data demonstrate that the delivery of 111In is dose dependent and highly specific. Under saturating conditions, binding of 111In-DTPA-MTX to alpha-TFR-DHFR-treated cells was 14-fold higher than to cells treated with nonconjugated alpha-TFR antibody. Further experiments indicated that the low level of nonspecific binding of 111In-DTPA-MTX was comparable to that of 111In-DTPA, known for its complete extracellular distribution and rapid clearance through the kidneys. Based on the data of this study, antibody-conjugated rhDHFR and radionuclide-labeled DTPA-MTX complexes provide components for an alternative radioimmunotherapeutic approach that can be expected to result in improved tumor tissue ratios of both the targeting moiety and the radionuclide-labeled derivative as compared to current approaches.

The spatial distribution of i.v. administered immunotoxins in s.c. human rhabdomyosarcoma RD2 xenografts was studied. The toxin and immunotoxins were: (a) diphtheria toxin (DT); (b) a binding-deficient form of DT (CRM107) linked to a monoclonal IgG1 antibody (454A12) directed against the human transferrin receptor (454A12-107); (c) the binding-deficient form of DT linked to the Fab' fragment of 454A12 (Fab'-107); and (d) the binding-deficient form of DT coupled to MOPC21, a monoclonal IgG1 with no significant binding to RD2 cells. DT and the immunotoxins were radiolabeled with 125I and injected via the tail vein into tumor-bearing athymic mice (median tumor weight, 0.25 g). Tumors were removed 2, 6, and 24 h after injection of DT or immunotoxin. Film images of 20-microns frozen sections were digitized by video microscopy, and gray levels were converted to tissue concentrations based upon the film response to radioactivity standards and the specific activity of the radiolabeled toxins. Images of the tumors were characterized quantitatively by the kurtosis and the area above threshold; the kurtosis is a measure of the spatial heterogeneity of the radiolabeled immunotoxins, and the area above threshold is defined here as the fractional tumor area that reaches or exceeds 1.5% of the initial plasma concentration. The spatial distribution of DT in the tumors was extremely uniform, characterized by low kurtosis values. In contrast, the autoradiograms of 454A12-107 were punctate in appearance and were characterized by very high kurtosis values. Fab'-107, which has approximately one-half the molecular weight of the intact immunotoxin and binds only monovalently, also produced punctate images with kurtosis values similar to those for 454A12-107. The nonbinding immunotoxin distributed somewhat less uniformly than DT but much more homogeneously than either of the binding immunotoxins. DT, 454A12-107, and Fab'-107 have similar affinities for their respective receptors, but the concentration of binding sites for DT on RD2 cells (<3,000 receptors/cell) is much lower than the concentration of transferrin receptor (60,000 receptors/cell). Thus, the heterogeneous distribution of 454A12-107 and Fab'-107 probably reflects retarded penetration due to binding to the tumor cells. The area above threshold was greatest for DT and lowest for 454A12-107; the fragment and nonbinding immunotoxins had intermediate values. The lower area above threshold for the nonbinding immunotoxin as compared with DT may be due to the considerably large molecular weight and hence the lower capillary permeability and diffusion coefficient of the immunotoxin.(ABSTRACT TRUNCATED AT 400 WORDS)

The action of a number of toxins used in the formation of immunotoxins was studied in polarized cells. Diphtheria toxin inhibited protein synthesis most efficiently when added to the basolateral side of the kidney cells, MDCK-I, MDBK and Pt K2, and the colon carcinoma cell Caco-2. Similar findings were made with Pseudomonas aeruginosa exotoxin A in MDCK-I, Pt K2, and Caco-2 cells, and with modeccin and volkensin in MDCK-I cells. In accordance with the toxicity data, diphtheria toxin bound specifically to the basolateral side of MDCK-I cells but not to the apical side. On the other hand, in the trophoblastic BeWo cell line there was little or no difference in the toxic effect of P. aeruginosa exotoxin A and modeccin added to the two sides. The plant toxins ricin and abrin and the bacterial Shigella toxin inhibited protein synthesis approximately equally well in all cell lines tested whether they were added apically or basolaterally. The results indicate that protein toxins are able to enter cells from both the apical and basolateral sides provided receptors are present. The consequences for the preparation of immunotoxins are discussed.

We have produced immunotoxins using LL2, a monoclonal antibody which binds to human B-cell lymphomas and which, in a radioiodinated form, induced responses in lymphoma patients (D.M. Goldberg et al., J. Clin. Oncol., 9: 548-564, 1991). We have coupled LL2 to Lys-PE38KDEL, a derivative of Pseudomonas exotoxin (PE) which does not bind to the PE receptor. LL2-PE38KDEL was cytotoxic toward several Burkitt's lymphoma lines, with 50% inhibitory concentration values ranging from 2 to 6 ng/ml (10-30 pM). Another immunotoxin, LL2-Fab'-PE38KDEL, was produced by chemically coupling the Fab' fragment of LL2 to Lys-PE38KDEL. LL2-Fab'-PE38KDEL also was cytotoxic toward the Burkitt's cells, with a 50% inhibitory concentration of 1-2 ng/ml (13-24 PM). The antibody LL2 alone had no cytotoxicity toward the malignant cells, and excess LL2 prevented the cytotoxicity of LL2-PE38KDEL and LL2-Fab'-PE38KDEL. Control immunotoxins UPC-10-PE38KDEL and Mu-9-Fab'-PE38KDEL were not cytotoxic. LL2-PE38KDEL and LL2-Fab'-PE38KDEL bound to cells with 50% and 17% of the affinity of LL2, respectively. Both immunotoxins, but not UPC-10-PE38KDEL, prevented the development of CA-46 tumors in nude mice. LL2-PE38KDEL and LL2-Fab'-PE38KDEL, but not the control immunotoxins, led to complete regressions of measurable s.c. CA-46 tumors in nude mice, when given at 50% and 35% of the 50% lethal dose, respectively. LL2 alone significantly retarded the growth of CA-46 tumors but did not cause complete tumor regressions. Immunotoxins containing derivatives of Pseudomonas exotoxin can be targeted to human B-cell lymphoma and merit further study as potential therapeutic agents.

The moderately differentiated human colon tumor cell line, HT-29, constitutively expresses low levels of the high molecular weight mucin, tumor-associated glycoprotein 72 (TAG-72), and the M(r) 180,000 carcinoembryonic antigen (CEA) when grown as s.c. tumors in athymic mice. We report that the in vivo administration of gamma-interferon (IFN-gamma) resulted in a time- and dose-dependent increase in both TAG-72 and CEA expression in the HT-29 tumors. Immunohistochemical staining revealed a more homogeneous TAG-72-positive tumor cell population after IFN-gamma. Furthermore, both anti-TAG-72 and anti-CEA monoclonal antibodies (MAbs) showed enhanced localization to the HT-29 tumors in mice treated with IFN-gamma. Using that experimental model, subsequent studies presented evidence showing that the combination of IFN-gamma with 131I-CC49, an anti-TAG-72 MAb, resulted in a statistically significant improvement in therapeutic efficacy when compared with 131I-CC49 alone. For example, treatment with 300 microCi of 131I-CC49 initially suppressed HT-29 tumor growth; however, that reduction in tumor growth was transient as evidenced by the emergence of additional tumor growth at later time points. In contrast, an 8-day treatment with IFN-gamma in combination with 300 microCi 131I-CC49 resulted in sustained suppression of HT-29 tumor growth. Thus, IFN-gamma in vivo can substantially increase the TAG-72 expression in human colon tumor xenografts which leads to an increased tumor localization of anti-TAG-72 MAbs and seems to be responsible for the enhanced antitumor effects when IFN-gamma was combined with 131I-CC49. The results provide further evidence for including a biological response modifier, such as IFN-gamma, which can increase the expression of specific tumor antigens (i.e., TAG-72 and CEA) subsequently leading to a dramatic improvement in the antitumor efficacy of a radionuclide-conjugated MAb.

We have constructed a single-chain immunotoxin composed of the carcinoma-reactive antibody BR96 and a truncated form of Pseudomonas exotoxin. The chimeric molecule, BR96 sFv-PE40, was expressed in Escherichia coli and localized to the inclusion bodies. We purified and identified two species of BR96 sFv-PE40, monomers and aggregates. The monomeric form was able to bind well to the BR96 antigen, a Lewisy-related antigen, while the aggregate was not. The binding affinity of the monomeric recombinant immunotoxin was 5-fold less than intact BR96 IgG, and its specificity for the BR96 antigen was confirmed by competition analysis. Monomeric BR96 sFv-PE40 was found to be extremely cytotoxic against cancer cells displaying the BR96 antigen. The cytotoxicity of the fusion protein correlates directly with antigen density on the tumor cell lines tested. The breast carcinoma cell line MCF-7, which has the highest density of BR96 antigen, was the most sensitive to BR96 sFv-PE40, with a concentration producing 50% protein synthesis inhibition of 5 pM. BR96 sFv-PE40 was found to have a t1/2 in serum of 28.5 min in athymic mice, compared to that of the chemical conjugate, chiBR96-LysPE40, which was 54 min. These data indicate that the single-chain immunotoxin BR96 sFv-PE40 is a potent inhibitor of protein synthesis in target cell lines and may be an effective agent for the treatment of cancer.

This report investigates the application of monoclonal antibody A7 and its drug conjugate in locally controlling colorectal cancer. The experimental protocol consisted of local retention, lymphatic delivery, normal organ distribution, systemic toxicity, and tumoricidal effects. When 125I-labeled monoclonal antibody (Mab) A7 was injected into the pelvis and the thigh of Balb/c mice, a high local retention unrelated to antigen-antibody interaction was observed at the injected site for 24 h after injection. An analysis of local retension properties related to antigen-antibody interaction, conducted by intratumorally or peritumorally injecting 125I-Mab A7 into the tumor-bearing athymic nude mice, revealed a significantly higher tumor localization of Mab A7 in comparison to i.v. injection. 125I-Mab A7 accumulated to a great extent in the ipsilateral regional lymph node but not in the contralateral regional lymph node. Normal organ accumulation of Mab A7 was lower in the locally injected group than in the i.v. injected group. Intratumoral injection of Mab A7-neocarzinostatin (A7-NCS) led to the complete remission of established tumor in 5 of 6 antigen-positive xenograft-bearing mice but exhibited a complete remission in only 1 of 6 antigen-negative xenograft-bearing mice. A single local injection of A7-NCS inhibited tumor development in 12 of 16 and 5 of 15 antigen-positive tumor-bearing mice and antigen-negative tumor-bearing mice, respectively, whereas neither a systemic injection of A7-NCS and NCS nor a local injection of NCS and saline had a notable inhibitory effect on tumor development. Systemic toxicity of NCS was markedly reduced when it was locally administered in the antibody-conjugated form. These findings indicate that local injection of immunoconjugate is a promising new field for controlling the local recurrence of colorectal cancer.

The cytotoxicity of WT32 (CD3)-ricin A immunotoxin (IT) to the acute lymphoblastic leukemia T-cell line Jurkat was compared with the rate of internalization and intracellular degradation of WT32 and WT32-ricin A during continuous exposure. Moreover, the influence of NH4Cl and monensin on these processes was studied. Based on protein synthesis inhibition ([3H]leucine incorporation), it appeared that cytotoxicity was not fully expressed directly after exposure to IT due to a delay in either the internalization of membrane-bound IT or the action of intracellular ricin A. Varying the duration of incubation and postponing [3H]leucine addition for up to 24 h after initiation showed that cytotoxicity occurred in two phases, rapid internalization of initially bound IT followed by a continuous but slower uptake, possibly due to reexpression of the CD3 antigen. No differences were found in the rate of internalization and degradation of 125I-labeled WT32 and WT32-ricin A. Internalization started rapidly after binding at 37 degrees C, was fastest during the first 12 h (+/- 360,000 molecules/cell), and continued for at least 24 h (+/- 420,000 molecules/cell). Exocytosis of intracellularly degraded molecules became measurable after 1 to 2 h of incubation at 37 degrees C and increased to approximately 400,000 molecules/cell in 24 h. After 4 h of incubation at 37 degrees C the number of internalized molecules exceeded the amount of WT32 that could maximally bind to the cell membrane (+/- 150,000 molecules/cell), confirming reexpression of antigen. The addition of NH4Cl and monensin enhanced the cytotoxicity of WT32-ricin A, probably due to an increased intracellular amount of IT. These agents appeared to reduce strongly the degradation of internalized WT32, resulting in an accumulation of intracellular molecules. NH4Cl was most effective during the first 12 h of incubation, whereas monensin increased the amount of intracellular WT32 molecules after 2 to 24 h. Our observations suggest that incubation conditions for the optimal cytotoxicity of IT treatment can be predicted by studying the internalization and degradation of the IT or respective monoclonal antibody.

Recombinant truncated forms of Pseudomonas exotoxin A that lack the cell binding domain of Pseudomonas exotoxin A were coupled to an F(ab') fragment of a monoclonal antibody HB21 directed against the human transferrin receptor. One of these was NlysPE40. The other, NlysPE38QQR, has two amino groups on residues near the NH2-terminus and has no amino groups near the COOH-terminus. The proteins were linked by a stable thioether bond that connected the sulfhydryl group present in the hinge region of the antibody fragment to an amino group on the toxin. The F(ab')-PE40 immunotoxin, containing NlysPE40, exhibited potent cytotoxic activity on human carcinoma cell lines with a concentration of immunotoxin at which isotope incorporation falls by 50% when compared to nontreated cells (ID50) of 5.3 pM (0.5 ng/ml) on both the epidermoid carcinoma A431 and on the colon carcinoma Colo205. Immunotoxins made with whole antibody were considerably less active, with an ID50 of 15.9 pM (3.1 ng/ml) on these cell lines. F(ab')-PE38QQR, the immunotoxin containing NlysPE38QQR, was found to be the most active agent with an ID50 of 1.05 pM (0.1 ng/ml) on A431 cells. The greater cytotoxicity of immunotoxins containing fragmented antibody was probably due to the higher binding affinity of F(ab') conjugates in comparison to whole antibody conjugates to the transferrin receptor. The increase in cytotoxic activity of the immunotoxin made with NlysPE38QQR than that with NlysPE40 may reflect selective coupling of the toxin through NH2-terminal amino groups. The monovalent and divalent immunotoxins had dose-dependent antitumor effects on human epidermoid carcinoma xenografts in nude mice. A431 tumors completely regressed in all animals at a total dose of 105 pmol (10 micrograms) of F(ab')-PE38QQR and of 154 pmol (30 micrograms) of IgG-PE38QQR. Furthermore, the F(ab') immunotoxin was less toxic to mice than the conjugate containing IgG (840 pmol or 80 micrograms of total dose causing measurable adverse effects versus 208 pmol or 40 micrograms, respectively). Thus, a truncated Pseudomonas exotoxin A molecule coupled to the F(ab') fragment of an antibody is more active and less toxic in mice than an immunotoxin made with a whole antibody. Therefore, the therapeutic index for the monovalent immunotoxin is about four times better than that for the divalent immunotoxin.

A conjugate of 4-desacetylvinblastine-3-carboxyhydrazide (DAVLBHY) and the glioma-reactive monoclonal antibody (mAb) 9.2.27 induced long-term suppression of tumor growth in athymic nude mice engrafted with U87MG human glioma cells. In vitro, DAVLBHY had the strongest antiproliferative activity (inhibitory concentration at which incorporation of [3H]thymidine is at 50% of untreated control is 2.0 x 10(-9) M) of seven cytotoxic drugs tested and so was chosen for conjugation to mAb 9.2.27, which reacts specifically with the core protein of chondroitin sulfate proteoglycans found in human glioblastomas. After conjugation of DAVLBHY to the carbohydrate residues of mAb 9.2.27 it retained its full binding capacity. For in vivo studies, DAVLBHY and several conjugate derivatives were evaluated by using two dosages of i.v. injections, each starting 2 days after s.c. tumor inoculation. The control tumors reached a volume of nearly 3000 mm3 within 30 days. Tumor growth was delayed by about 20 days with four i.v. injections of 0.5 mg/kg 9.2.27-DAVLBHY, which was slightly superior to the unconjugated drug. Moreover, 9.2.27-DAVLBHY produced a highly significant suppression of growth so that the average tumor volume was only 3% of that observed in untreated controls after 28 days. Four injections of this conjugate at a larger dose, 2.0 mg/kg, prevented recurrence of the tumors for 130 days in all animals tested, thus demonstrating a significant increase in the therapeutic index, since the unconjugated drug provided limited inhibition of tumor growth for only 40 days. The specificity of the antitumor effect was demonstrated in a comparison with the control conjugate, KS1/4-DAVLBHY, which despite partial tumor suppression had only a transient effect. The specific antitumor effect of 9.2.27-DAVLBHY was unexpected, since the target antigen is expressed at a relatively low density (40,000 sites/cell) on U87MG glioma cells.

The endocytosis and degradation of 125I-labeled anti-mu monoclonal antibody DA4-4 by a Burkitt's lymphoma cell line was investigated using biochemical, chromatographic, electrophoretic, radioautographic, and electron microscopic techniques. 125I-DA4-4 was rapidly internalized by Ramos cells and routed from endosomes to lysosomes. Proteolysis of radiolabeled antibodies began in a late endosomal compartment, but lysosomes were primarily responsible for the terminal degradation of 125I-DA4-4. Catabolism of 125I-DA4-4 could be inhibited by 74-95% by blocking its delivery to late endosomes and lysosomes by incubation at 18 degrees C, by neutralizing the pH in intracellular organelles with monensin or ammonium chloride, or by inhibiting lysosomal enzymes with leupeptin. Radiolabeled antibodies synthesized using the chloramine T or Iodo-Gen techniques were degraded three times faster than conjugates made using a nonmetabolizable 125I-tyramine cellobiose adduct. Five major intermediate metabolites (Mr 48,000, 42,000, 25,000, 15,000, and 10,000) were generated during the intracellular catabolism of 125I-DA4-4, but 125I-tyrosine was responsible for 95% of the small-molecular-weight metabolites released by cells into the culture medium. We anticipate that a full comprehension of the catabolism of radiolabeled antibodies by tumor cells will make possible the development of clinical interventions which will enhance the retention of radioimmunoconjugates by hematologic malignancies and improve the efficacy of radioimmunotherapy.

The intratumoral and whole-body distributions of 90Y-labeled C110 anticarcinoembryonic antigen immunotoxin after i.p. and i.v. injection were compared by quantitative autoradiography. During in vitro incubation of spherical tumor nodules of LS174T human colon cancer (about 5 mm in diameter) in a medium containing C110 radioimmunotoxin (RIT), the direct penetration of the immunotoxin increased with time but was limited to the outer 300 microns of the tumor nodule after 12 h of incubation. In vivo experiments were performed in nude mice bearing LS174T xenografts as i.p. tumor nodules. Injection of C110 RIT i.p. resulted in a ring-like distribution, i.e., high uptake at the tumor periphery and considerably lower uptake at the tumor center (ratio of peripheral to central concentration, 7:1 at 1 day and 2:1 at 5 days). In contrast, i.v. injection provided a much smaller gradient in C110 RIT distribution from peripheral to central regions (ratio of peripheral to central concentration, 3:1 at 1 day and 1:1 at 5 days). Estimates of total tumor uptake of C110 RIT by quantitative autoradiography demonstrated almost equivalent tumor uptake after either i.p. or i.v. injection, while i.v. injection was associated with increased C110 RIT uptake in various normal organs, especially in the liver, as compared to i.p. injection. The results in this study suggest that (a) i.v. injection may produce more homogeneous distribution of C110 RIT in i.p. tumor nodules of LS174T but may also result in increased liver toxicity, and (b) i.p. injection may decrease C110 RIT exposure of normal tissues, which can reduce systemic toxicity, but may also produce more restricted intratumoral distribution of C110 RIT. In addition, current methods using a nude mouse model of i.p. tumor nodules and quantitative autoradiography allow us to assess intratumoral and whole-body distributions of radiolabeled immunoconjugates from various administration routes.

With regard to increasing the clinical potential of ricin A-chain immunotoxins (RTA-ITs), a novel class of calcium channel blockers, indolizines SR33557 [2-isopropyl-1-[4-(3-N-methyl-N-(3,4-dimethoxy-beta- phenethyl)amino)propyloxy)benzenesulfonyl))indolizine] and SR33287 [isopropyl-2-((1-butylamino-3-propyl)oxy-4-benzoyl)-3-indolizine], were evaluated for their ability to enhance RTA-IT activity in vitro and in vivo. Five microM SR33287 and 5 microM SR33557 were potent enhancers of both anti-Thy 1.2 AT15E RTA-IT (84- and 64-fold, respectively) on T2 cells and anti-CD5 T101 (622- and 538-fold) and T101 F(ab')2 RTA-IT (34- and 28-fold) on CEM III cells. This was superior to the effect achieved by both 10 microM verapamil and 10 mM NH4Cl, albeit slightly inferior to that of 50 nM monensin and 5 microM perhexiline. Murine T2 lymphoma cells bearing the Thy 1.2 antigen were injected i.v. in Thy 1.2 (-) BL. 1.1 mice (median survival time, 17.7 days). Intravenous treatment with 10 micrograms of AT15E RTA-IT prolonged the survival of mice (median survival time, 26.8 days). When 400 micrograms of SR33287 were coinjected i.v. with 10 micrograms of AT15E RTA-IT, mouse survival was further increased, with 5 of 6 mice surviving, disease free, over 42 days. SR33287 had a significant impact on the intracellular routing of 125I-AT15E RTA-IT, which induced a greater than 2-fold increase in intracellular intact AT15E RTA-IT at 90 min. This effect on RTA-IT half-life was distinctly different from that observed with either NH4Cl or monensin and may be linked to the inhibition of acid lysosomal sphingomyelinase by SR33287, leading to cellular lipidosis. In conclusion, indolizines appear to be promising agents not only for immunotoxin enhancement but also for increasing the activity of any number of targeted therapeutic agents where modifying either the intracellular routing or increasing the intracellular half-life of the ligand would be beneficial to its cytotoxic activity.

The potential of immunoconjugates of cytotoxic drugs for the treatment of cancer has not yet been realized owing to the difficulty of delivering therapeutic concentrations of these drugs to the target cells. In an effort to overcome this problem we have synthesized maytansinoids that have 100- to 1000-fold higher cytotoxic potency than clinically used anticancer drugs. These maytansinoids are linked to antibodies via disulfide bonds, which ensures the release of fully active drug inside the cells. The conjugates show high antigen-specific cytotoxicity for cultured human cancer cells (50% inhibiting concentration, 10 to 40 pM), low systemic toxicity in mice, and good pharmacokinetic behavior.

The conjugation of antineoplastic drugs to monoclonal antibodies reactive with tumor associated antigens conveys selective cytotoxicity, overcoming the systemic toxicities caused by drugs during standard chemotherapy. 2'-Deoxy-5-fluorouridine, a more potent derivative of 5-fluorouracil, is an antimetabolite which exerts its cytotoxic action by inhibiting the enzyme thymidylate synthetase and as a result inhibits DNA synthesis. 2'-Deoxy-5-fluorouridine was successfully conjugated to anti-Ly-2.1 reactive with the murine thymoma ITT(1)75NS E3, I-1, and 250-30.6 reactive with human colon cancer cells using the active ester of 2'-deoxy-5-fluoro-3'-O-succinoyluridine (5FdUrdsucc). In vitro, 5Fd-Urdsucc-anti-Ly-2.1 was selectively cytotoxic against ITT(1)75NS E3 murine thymoma cells at nanomolar concentrations. The human colon carcinoma cell LIM1899 was inhibited by 5FdUrdsucc-I-1 conjugates in the range of 10(-7)-10(-8) M, as were Colo 205 cells by 5FdUrdsucc-250-30.6 conjugates. In vivo, 5FdUrdsucc conjugates were more effective than equivalent amounts of free 5FdUrdsucc. Against the ITT(1)75NS E3 murine thymoma, a single dose of 100 micrograms (5FdUrdsucc equivalents) 5FdUrdsucc-anti-Ly-2.1 resulted in 85% tumor inhibition compared to mean tumor size of control mice. Irrelevant 5FdUrdsucc conjugates failed to inhibit tumor growth. Multiple doses of 5FdUrdsucc-I-1 conjugate produced 50% growth inhibition of the moderately differentiated tumor LIM1899. In contrast, the human colon carcinoma Colo 205 was relatively resistant to free 5FdUrdsucc and 5FdUrdsucc-250-30.6 conjugates.

Ricin, the cytotoxic protein isolated from castor beans, is composed of two subunits, A-chain and B-chain. Ricin intoxicates cells by binding through its B-chain to galactose-terminated oligosaccharides found on the surface of all eukaryotic cells and then transferring its A-chain to the cytosol where it disrupts protein synthesis by inactivating ribosomes. In addition to binding, the B-chain plays an important, but not yet understood, role in the translocation of the A-chain through a cellular membrane to the cytosol. Blocking the two galactose-binding sites of native ricin by chemical modification with affinity ligands created an altered toxin, called blocked ricin, that has at least a 3500-fold lower binding affinity and is more than 1000-fold less cytotoxic than native ricin for Namalwa cells (a Burkitt's lymphoma line) but that has maintained the translocation function of the B-chain and the catalytic activity of the A-chain. Conjugation of blocked ricin to monoclonal antibodies that bind to cell surface antigens creates new cytotoxins that approach the potency of native ricin. These cytotoxins incorporate the three essential functions of natural toxins, i.e., binding to cells, transport through a membrane, and catalytic inactivation of an essential cellular process; but in addition they possess a defined cellular target specificity. Such potent immunotoxins may play an important therapeutic role in cancer treatment. Clinical trials with an anti-CD19-blocked ricin and an anti-CD33-blocked ricin conjugate against B-cell cancers and acute myeloblastic leukemia have begun.

In this study, the breast carcinoma-reactive monoclonal antibody 15A8 and a site-specific immunoconjugate of the antibody, 15A8-glycyl-tyrosyl-(N-epsilon-diethylenetriamine pentaacetic acid)-lysine (15A8-GYK-DTPA), were characterized by immunohistological methods for reactivity with normal and neoplastic human tissues and normal cynomolgus monkey tissues. In addition, 15A8-GYK-DTPA labeled with 111In was assessed by in vivo imaging and pharmacokinetic studies for localization to human tumor xenografts in nude mice. The native antibody and the site-specific immunoconjugate exhibited similar limited reactivity with normal human tissues. Specifically, epithelial structures, including normal breast epithelium, lung alveoli, bronchial epithelium and glands, liver bile ducts, pancreatic ducts, kidney distal and collecting tubules, epidermal and esophageal epithelium, endometrial glands, and thymic Hassall's corpuscles, were reactive. Normal monkey tissues stained with 15A8 exhibited a similar pattern of reactivities. Antibody 15A8 reacted broadly with epithelium-derived tumors; more than 60% of the cells in all of the breast, colon, non-small cell lung, ovarian, prostate, bladder, and renal carcinomas tested expressed the antigen. In contrast, a variety of nonepithelial neoplasms, including lymphomas, melanomas, sarcomas, and small cell lung carcinomas, were nonreactive. 15A8-GYK-DTPA-111In administered i.v. rapidly localized to and imaged both MX-1 and MCF-7 human breast carcinoma xenografts in nude mice, reaching maximal levels of about 20% of injected dose/g of tumor within 4 days. No unusual localization to any nontumor tissue or organ was seen; the level of radioactivity in the normal tissues and organs was at or below that seen in the blood. Furthermore, the immunoconjugate did not accumulate in xenografts of the antigen-negative breast carcinoma ZR-75-1, which indicates that tumor localization was antigen specific. Pharmacokinetic studies in cynomolgus monkeys suggested that significant amounts of 15A8-GYK-DTPA-111In did not localize to normal epithelia and demonstrated that the immunoconjugate was not toxic. These findings suggest that antibody 15A8 may be useful in the diagnosis and therapy of breast cancer and possibly other carcinomas.

Photodynamic therapy is a relatively new and potentially selective experimental approach to the treatment of malignant neoplasms. Its inherent dual selectivity is reinforced by the use of photosensitizer-monoclonal antibody conjugates. The goal of this study was to evaluate the phototoxicity and selectivity of an immunoconjugate (IC) synthesized from a chlorin derivative chlorin e6-monoethylenediamine monoamide (CMA) as the photosensitizer and an anti-ovarian carcinoma monoclonal antibody OC125. Binding efficiency and specificity of the IC were determined by enzyme-linked immunosorbent assay, and specific covalent linkage of the monoclonal antibody to the photosensitizer was demonstrated by fluorescence and electrophoresis. Phototoxicity was tested against ascites or pleural fluid cells from 15 patients with ovarian and nonovarian cancers. Tumor cells from the fluid were treated with the IC at 3 microM equivalent CMA concentration and irradiated at 654 nm (lambda max CMA in IC) at 25 J/cm2 from an argon ion-pumped dye laser. Phototoxic efficacy was assayed by [3H]thymidine incorporation. Ovarian cancer cells exhibited high cytotoxicity with [3H]thymidine incorporation of 2.4 +/- 2.2%, while nonovarian cancer cells under identical conditions exhibited none to reduced cytotoxicity with [3H]thymidine incorporation of 70 +/- 54%. Using a Wilcoxon test, there was a statistically significant difference between these two groups (P less than 0.001). Dose-response curves revealed reciprocity in photosensitizer concentration and fluence. These results demonstrate that photoimmunoconjugates retain significant antigen binding specificity and affinity, are effective in the selective photochemical eradication of target cells, and merit further evaluation as photochemotherapeutic agents.

It is widely believed that antigen heterogeneity and noninternalization of antigen-antibody complexes will severely limit the antitumor activity of monoclonal antibody-drug conjugates. The B72.3 monoclonal antibody binds to a tumor-associated antigen which is heterogeneously expressed in human carcinomas (J. Schlom, Cancer Res., 46: 3225-3238, 1986). We therefore performed studies to assess the degree of internalization of B72.3 antibody-antigen complexes and the level of in vivo antitumor activity that could be achieved with B72.3 conjugated to 4-desacetyl vinblastine-3-carboxhydrazide. Internalization studies were performed on LS174T colorectal carcinoma and OVCAR-3 ovarian carcinoma cells using iodinated B72.3 as well as an iodinated antibody that binds to the human transferrin receptor, IIB21. These data indicated that, in contrast to HB-21, the B72.3 antigen-antibody complex was not internalized. The B72.3-Vinca alkaloid immunoconjugate demonstrated significant antitumor activity against LS174T xenografts, although complete regressions of established tumors were not achieved. Immunohistochemical analyses indicated that the B72.3 antigen was heterogeneously expressed in the LS174T xenografts and that tumor cells which were not killed by high doses of B72.3-Vinca also expressed the B72.3 antigen. These studies indicated that significant antitumor activity may be achieved by monoclonal antibody-drug conjugates even when antigen heterogeneity and noninternalization of antigen-antibody complexes are encountered. The data also suggested that the formulation of antibody-drug conjugate cocktails to counteract antigen heterogeneity may not be sufficient to eradicate all malignant cells within a solid tumor mass.

We report that selected combinations of two or more monoclonal bispecific F(ab')2 antibodies (BsAbs) far outperform single derivatives in the delivery of the ribosome-inactivating protein, saporin, to guinea pig L2C leukemic cells. Throughout the work, BsAbs were constructed by thioether-linking the hinges of two Fab'gamma, one from monoclonal anti-L2C-idiotype antibody and the other from anti-saporin antibody. The latter was either affinity-purified rabbit polyclonal or one of a panel of five mouse monoclonal antibodies. In vitro cytotoxicity studies showed that, though all derivatives were effective, the BsAb made with the polyclonal antibody was always 10 to 20 times more potent than those made with a monoclonal antibody in yielding 50% inhibition of [3H]leucine uptake. This superior activity could be matched by selective mixtures of two or more of the monoclonal derivatives. Furthermore, in immunotherapeutic delivery of saporin to tumor, a pair of BsAbs performed significantly better than did either individually. Binding and uptake studies with radiolabeled saporin demonstrated a 20-fold increase in functional affinity when saporin was held at the cell surface by an appropriate BsAb mixture rather than by a single BsAb. In contrast, only small differences were recorded in the rate at which saporin was internalized as a result of the same maneuver. We conclude that the improved performance of combinations of BsAbs arises from their ability to provide multiple linkages between saporin molecules and cell surfaces, significantly increasing the functional affinity with which saporin is tethered to the cell, but, in this system at least, having only a minor effect on the rate at which it is internalized. Cocktails of two or more BsAbs, selected to bind to multiple epitopes on ribosome-inactivating proteins and perhaps also on unwanted cells, could provide an important new strategy in immunotherapy.

Intracellular ricin and immunotoxin trafficking has been difficult to study as only one to two cytosolic ricin A chain (RTA) molecules are sufficient to cause cell death. Previous studies (R.J. Youle and M. Colombatti, J. Biol. Chem., 262: 4676-4882, 1987) using anti-ricin hybridomas identified the secretory pre-Golgi as a critical site for RTA intoxication. We used ricin and RTA immunotoxins constructed with transferrin (TF) or anti-murine TF receptor antibody (RI7/217) to compare patterns of cytotoxicity and intracellular trafficking in anti-ricin hybridomas. Anti-RTA and anti-ricin B chain (RTB) hybridomas bound similar amounts of ricin and secreted comparable amounts of anti-ricin immunoglobulin. Anti-RTA hybridomas were 50- to 500-fold more resistant to ricin than nonsecretory and anti-RTB hybridomas, defining a ricin-resistant phenotype. All hybridomas expressed similar levels of surface TF receptors. RTA immunotoxins were constructed using human TF or RI7/217 and a disulfide linker. In protein synthesis inhibition assays, ricin-resistant hybridomas were manyfold more resistant to RI7/217-RTA than were ricin-sensitive hybridomas. In contrast, all hybridomas were equally sensitive to TF-RTA. Monensin increased ricin cytotoxicity minimally against all hybridomas, but dramatically increased RI7/217-RTA cytotoxicity in ricin-resistant and ricin-sensitive hybridomas in a way that abrogated the ricin-resistant phenotype. In contrast, monensin increased TF-RTA cytotoxicity equally in all hybridomas. Ammonium chloride had little effect on ricin or RI7/217-RTA cytotoxicity, but increased TF-RTA cytotoxicity against all hybridomas. Taken together, these results suggest that RTA molecules mediating cytotoxicity pass through an anti-RTA antibody-containing pre-Golgi compartment when bound to RTB or RI7/217, but not when bound to TF. Monensin abrogates the ricin-resistant phenotype when RTA is linked to RI7/217, but not RTB. This suggests that monensin alters RI7/217-RTA processing proximal to the pre-Golgi and that passage through the pre-Golgi may not be necessary for translocation of RTA to the cytoplasm. Ammonium chloride alters toxin cytotoxicity only when RTA is linked to TF, suggesting that only TF trafficks RTA through an acid-sensitive compartment prior to cytoplasmic translocation. With the addition of potentiating agents, each toxin studied showed a unique cytotoxicity profile against the anti-ricin hybridomas, demonstrating a dominant role of the cell binding ligand in intracellular toxin trafficking.

In this report we demonstrated that human alpha 2-macroglobulin (alpha 2M) reacts with deglycosylated ricin A chain (dgA) and its immunotoxins to form high molecular weight complexes (molecular mass approximately 800 kDa). This interaction has a t1/2 at 37 degrees C of 5 h and reaches completion at 24 h. Complexes of alpha 2M-dgA cannot be dissociated by guanidine, sodium dodecyl sulfate, or low pH, but can be partially dissociated by reducing agents, such as 2-mercaptoethanol in the presence of sodium dodecyl sulfate. This indicates that dgA or dgA-containing immunotoxins are bound to alpha 2M by disulfide bonds. The dgA-binding site on alpha 2M and the mechanism underlying its interaction with dgA are different from those described for proteases or methylamine. alpha 2M complexes do not bind to Blue-Sepharose 4B or anti-A chain-Sepharose, suggesting that the sites on dgA which bind Cibacron Blue or polyclonal anti-A chain antibodies are sterically blocked or modified by interaction with alpha 2M. The interaction of alpha 2M with dgA or its immunotoxins results in a 2- to 3-fold decrease in the activity of the dgA in both cell-free assays and cytotoxic assays. However 12 h after injection into mice, only 11% of immunotoxin was bound to alpha 2M because of the slow kinetics of the interaction versus the more rapid t1/2 of the immunotoxin in the circulation.

An immunotoxin consisting of the enzymatically active A-chain of mistletoe lectin I and a monoclonal antibody against a surface protein on mouse leukemia L1210V cells was found to inhibit protein synthesis in these cells as efficiently as the native mistletoe toxin. The immunotoxin was somewhat more slowly endocytosed than the native toxin, but in both cases the endocytic uptake continued under conditions in which uptake from clathrin-coated pits was inhibited by mild acidification of the cytosol. This indicates that the toxin and the immunotoxin were at least partially internalized by a non-clathrin-dependent uptake mechanism and that uptake by this pathway is responsible for most of the toxic effect on the cells. The results indicate that efficient immunotoxins can be made with antibodies against cell surface epitopes that are endocytosed by a mechanism not involving clathrin-coated pits.

The anti-Tac antibody is known to bind to the p55 chain of the human interleukin 2 receptor. An immunotoxin was produced by genetically linking Clostridium perfringens phospholipase C (PLC) to the Fab domain of anti-Tac. For this purpose, the PLC gene, with its own promoter and signal sequence, was fused to the 5' end of the VHCH1 segment of the anti-Tac heavy chain gene. The anti-Tac light chain gene, with an attached bacterial signal sequence, was made part of the same transcriptional unit. Escherichia coli transformed with the construct secreted a recombinant immunotoxin, anti-Tac(Fab)-PLC, in an active form. Anti-Tac(Fab)-PLC bound to cells expressing the interleukin 2 receptor and inhibited protein synthesis, with a 50% inhibitory concentration of 0.02 nM (1.8 ng/ml).

Antibody conjugates were prepared by coupling F(ab')2 or Fab' fragments of an antibody specific for the human high molecular weight-melanoma associated antigen to Fab' fragments of an antibody specific for indium-diethylenetriaminepentaacetate complexes. Monovalent and bivalent haptens were synthesized by reacting the dipeptide tyrosyl-lysine with diethylenetriaminepentaacetic cyclic anhydride. In vitro, the antibody conjugate mediated binding of the 111In-labeled haptens to melanoma cells. In vivo, it allowed specific localization of the haptens in A375 tumors. The bivalent hapten exhibited much higher efficiency at targeting 111In onto cells, both in vitro and in vivo. Antibody conjugate and hapten doses (2 micrograms and 1 pmol, respectively) and the delay between antibody conjugate and tracer injections (24 h) were adjusted to maximize tumor uptake (4% injected dose/g) and tumor to normal tissue contrast (greater than 3) obtained 3 h after injection of the 111In-labeled bivalent hapten. This two-step technique, when compared to direct targeting of 111In-labeled F(ab')2 fragments, provided lower localization of injected activity into the tumor (x 0.25), but higher tumor/tissue ratios, especially with respect to liver (x 7), spleen (x 8), and kidneys (x 10). In addition, high contrast images were obtained within 3 hours, instead of days. Thus, antibody conjugate-mediated targeting of small bivalent haptens, labeled with short half-life isotopes, is proposed as a general method for improving tumor radioimmunolocalization.

Substantial heterogeneity has been observed in the expression of individual antigens within tumor cell populations. Immunotoxins which bind to different cell surface antigens might exert additive or synergistic cytotoxicity when used in combination to eliminate all clonogenic cells within a tumor. Immunotoxins have been prepared by conjugating recombinantly derived toxin A chain to different monoclonal reagents which recognize cell surface determinants of Mr 42,000 (317G5), 55,000 (260F9), and 200,000 (741F8). Each immunotoxin was evaluated for binding, internalization, and cytotoxicity with four breast cancer cell lines. Each of the three immunotoxins bound to the SKBr3 cell line and exerted antitumor activity in a limiting dilution clonogenic assay. Simultaneous treatment with two immunotoxins produced additive antitumor activity with each of the possible combinations. Additive binding could be demonstrated by immunofluorescent techniques, however, with only one of three combinations. With two of the three combinations, subpopulations of tumor cells could be identified which lacked one or the other antigenic determinant but not both. Consequently, log-additive antitumor activity was produced by immunotoxins in combination, and heterogeneity of antigenic targets may have contributed to the combined cytotoxicity in some but not all cases.

The monoclonal antibody L6 recognizes a determinant that is expressed on lung, breast, colon, and ovarian carcinomas and is present only at trace levels in normal tissues. L6 was covalently linked to intact ricin by a thioether bond to produce an immunotoxin (IT). Gel analysis revealed that this IT was heterogeneous, but mostly one monoclonal antibody molecule linked to one ricin molecule. The L6-ricin IT selectively bound and was selectively toxic to L6-positive H2981-T3 adenocarcinoma cells in protein synthesis inhibition assays in which lactose was added to block the native ricin binding site. Clonogenic studies showed that 1 microgram/ml L6-ricin could inhibit about 99.99% of H2981-T3 growth in a limiting dilution assay, even in the presence of a 20-fold excess of human bone marrow cells. Treatment of bone marrow cells with the same dose of L6-ricin resulted in the growth of ample numbers of bone marrow progenitor cells (colony-forming units-mixed, colony-forming units granulocyte/macrophage, and blast-forming units erythroid) after 14 days. We also evaluated the antitumor effect of L6-ricin administered intratumorally with lactose against established H2981-T3 tumors in a nude mouse model. Thirty % of the tumor-bearing animals responded completely to single-dose treatment, while 60% gave partial responses. The in vivo effects were not absolutely specific, since irrelevant anti-CD5 IT also induced tumor regression in this model (10% responded completely, while 30% gave partial responses). However, irrelevant IT gave higher systemic toxicity (50% mortality) than L6-ricin (23% mortality). The nonspecific activity of IT was possibly due to Fc binding, which was demonstrated in vitro, or due to ricin B-chain binding. Ricin alone was too toxic for sustained tumor protection. Unconjugated L6 had no antitumor effect. The data suggest that L6-ricin may be useful for in vitro purging of autologous bone marrow from patients with solid tumors and marrow involvement and for in vivo regional therapy of L6-positive carcinomas.

A monoclonal anti-Thy-1.1 antibody (OX7) was coupled to either native or chemically deglycosylated ricin A-chain (dgA) using one of two different cross-linking agents. One cross-linker, N-succinimidyloxycarbonyl-alpha-methyl-alpha-(2-pyridyldithio)tolu ene (SMPT), generates a sterically hindered disulfide bond which is relatively resistant to reduction, whereas the other, 2-iminothiolane hydrochloride, generates an unhindered disulfide bond with greater lability. A two-compartment pharmacokinetic model was used to analyze the blood levels of each immunotoxin and its breakdown product (free antibody) after i.v. injection into mice. Immunotoxins prepared with SMPT broke down in vivo 6.3-fold more slowly than those prepared with 2-iminothiolane hydrochloride, and immunotoxins containing native A-chain were cleared 2- to 3-fold more rapidly from the bloodstream than those containing dgA. As a result, 24 h after injection, 16% of the OX7-SMPT-dgA remained in the blood as compared with 0.4 to 2.5% of the other immunotoxins. Immunotoxins prepared with dgA were about 3-fold more toxic to mice than those prepared with native A-chain, whereas immunotoxins prepared with SMPT were only slightly more toxic than those prepared with 2-iminothiolane hydrochloride. When equivalent toxic doses of the immunotoxins were administered i.v. to mice which had been given injections of Thy-1.1+ AKR-A/2 lymphoma cells, the OX7-SMPT-dgA gave the best antitumor effect. A dose equivalent to one-seventh of the median lethal dose extended the survival time of the animals by the extent expected if 99.999% of the tumor cells had been eradicated. Furthermore, the tumors that did develop in the mice treated with OX7-SMPT-dgA were mutants which were resistant to all the immunotoxins. Some of the mutants were deficient in Thy-1.1 whereas others were not. In conclusion, both the use of the SMPT cross-linker and deglycosylation of the A-chain significantly improve the therapeutic index of the immunotoxins in AKR-A/2 tumor-bearing mice.

Two new coupling agents were synthesized for making immunotoxins containing disulfide bonds with improved stability in vivo: sodium S-4-succinimidyloxycarbonyl-alpha-methyl benzyl thiosulfate (SMBT) and 4-succinimidyloxycarbonyl-alpha-methyl-alpha(2-pyridyldithio)tolue ne (SMPT). Both reagents generate the same hindered disulfide linkage in which a methyl group and a benzene ring are attached to the carbon atom adjacent to the disulfide bond and protect it from attack by thiolate anions. An immunotoxin consisting of monoclonal anti-Thy-1.1 antibody (OX7) linked by means of the SMPT reagent to chemically deglycosylated ricin A-chain had better stability in vivo than an immunotoxin prepared with 2-iminothiolane hydrochloride (2IT) which generates an unhindered disulfide linkage. About 48 h after i.v. injection into mice, one-half of the SMPT-linked immunotoxin present in the blood was in intact form and one-half as released free antibody, whereas equivalent breakdown of the 2IT-linked immunotoxin was seen at about 8 h after injection. Consequently, the blood levels of the SMPT-linked immunotoxin remained higher than those of the 2IT-linked immunotoxin despite loss of immunotoxin from the blood by other mechanisms. Forty-eight h after injection, 10% of the injected dose of the SMPT-linked immunotoxin remained in the bloodstream as compared with only 1.5% of the 2IT-linked immunotoxin. The ability of immunotoxins prepared with the new reagents to inhibit protein synthesis by Thy-1.1-expressing AKR-A/2 lymphoma cells in vitro was identical to that of immunotoxins prepared with 2IT or N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP). Clonogenic assays showed that fewer than 0.01% of AKR-A/2 cells survived exposure to high concentrations of OX7-abrin A-chain immunotoxins prepared with SMBT, 2IT, or SPDP. Twelve clones of cells which had survived treatment with the SMBT-linked immunotoxin were isolated. None of the clones was selectively resistant to the SMBT-linked immunotoxin when retested in cytotoxicity assays. In conclusion, immunotoxins prepared with the new coupling agents should have improved antitumor activity in vivo because they are longer lived and do not break down so readily to release free antibody which could compete for the target antigens.

The binding of methotrexate (MTX) to IgG in conjugates was examined by studies on a direct MTX conjugate with a monoclonal antibody (aMM46) to mouse mammary tumor MM46 cells and corresponding irrelevant antibody and normal gamma-globulin conjugates, all prepared by the active ester method with MTX N-succinimidyl ester (MTX-OSu). The binding was examined in terms of effects on the potency and selectivity of the cytotoxic activity of the aMM46 conjugate. The results obtained supported the following conclusions: (a) MTX-OSu reacts not only with the amino group of IgG to give an amide bond, but also with another group(s) to give a less stable bond(s) such as an ester bond; (b) in contrast to the amide bond-linked MTX, which is taken up by the cells by endocytic internalization, a substantial portion of the MTX linked by an ester or other less stable bond(s) is released from the conjugates extracellularly and enters the cells by the MTX active transport system, as revealed by the inhibitory effect of thiamine pyrophosphate on the cytotoxicity; (c) this MTX linked by a less stable bond(s) that causes nonspecific cytotoxicity can be removed by treatment with hydroxylamine; (d) the overall cytotoxicity of aMM46-MTX decreased on removal of this less stably linked MTX, suggesting that the lysosomal degradation of the conjugate carrying amide bond-linked MTX to liberate MTX derivatives of low molecular weight is insufficient; (e) the liberation of low-molecular-weight substances in the lysosomes is probably more important for efficient entry of active substances into the cytosol, than for inhibition of the activity of dihydrofolate reductase, because after hydroxylamine treatment, the amide bond-linked MTX showed greater decrease in drug cytotoxicity than in inhibitory activity against dihydrofolate reductase; (f) in combination with hydroxylamine treatment, insertion between MTX and IgG of a linkage capable of ready cleavage in lysosomes deserves exploitation as a method for making potent conjugates with less nonspecific activity.

A number of compounds that interfere with glycoprotein synthesis and transport have been tested for their ability to sensitize cells to cancerostatic protein toxins. Tunicamycin, swainsonine, cycloheximide, and puromycin sensitized Vero cells and HeLa cells to abrin and ricin, as we have found previously with monensin (K. Sandvig and S. Olsnes, J. Biol. Chem., 257: 7504-7513, 1982). Cycloheximide, but not swainsonine, sensitized Vero cells to Pseudomonas exotoxin A and Shigella toxin. The ability of ricin to intoxicate cells was much lower at 19 degrees C than at 37 degrees C and there was almost no sensitizing effect of cycloheximide and monensin at 19 degrees C. Studies by electron microscopy showed that ricin conjugated to horseradish peroxidase appeared in trans Golgi elements in Vero cells. Possibly, transport of ricin into the cytosol requires passage through the Golgi apparatus. The possibility that the sensitizing agents here described may be valuable in enhancing the action of immunotoxins is discussed.