Peptide analogues of a T-cell epitope of ricin toxin A-chain prevent agonist-mediated human T-cell response

The clinical efficacy of immunotoxins (IT) containing ricin toxin A-chain (RTA) can be drastically reduced by anti-toxin-neutralizing antibodies developed by patients. Strategies aimed at epitope-specific modulation of the immune response must be therefore set up to broaden the clinical applicability of RTA-based IT. Prevention or reduction of humoral immune responses against RTA could be achieved by peptide-based down-modulating strategies. Peptide analogues were investigated as candidate antagonist altered peptide ligands (APL) considering the sequence of a previously identified dominant T-cell epitope of RTA (i.e. I175-E185) presented in the context of the HLA-DRB1*03011 allele. Alanine-substituted peptides provided information on the role of individual residues of the wild-type peptide and allowed to identify one antagonist APL corresponding to the double-mutant peptide E177A/A178D. The analogue E177A/A178D not only prevented the agonist from stimulating anti-RTA human T-cell clones but also failed to induce down-regulation of surface-expressed TCR, thus suggesting its possible use for in vivo immune modulation of anti-RTA responses.

[Effect of modified ricin on the reduction of hepatotoxicity in mice]

OBJECTIVE

To observe the significance of ricin (RT) with chemical nidification to reduce the hepatotoxicity in mice and its anticancer effect.

METHODS

Mice were exposed to RT and RT-PDP [ricin chemically modified by N-succinimidyl 3-(2-pyridyldithio)-propionate, (SPDP)] respectively, and their serum activity of glutathione-S-transferase (GST) and liver glutathione (GSH) content were determined. The ultramicro-structure under electron microscope was also observed.

RESULTS

The GST activity increased with doses, and the increase in ricin group was higher than that in RT-PDP group; the activities of GST in RT group at 12.5, 15.0 micro g/kg [(93.65 +/- 12.30), (153.71 +/- 26.64) IU/L respectively] were higher than those in RT-PDP group [(62.97 +/- 11.22), (78.20 +/- 15.71) IU/L] (P < 0.05). The contents of GSH were decreased with doses; but the contents of GSH in RT-PDP group at 2.5, 5.0, 7.5, 10.0, 15.0 micro g/kg [(6.34 +/- 1.43), (4.14 +/- 1.82), (3.54 +/- 0.64), (2.73 +/- 1.82), (1.82 +/- 0.62) micro mol/L respectively] were still higher than those in RT group [(3.53 +/- 0.95), (2.12 +/- 0.54), (1.82 +/- 0.71), (1.52 +/- 0.34), (0.81 +/- 0.36) micro mol/L] (P < 0.01). Electron microscopic examination showed that the injury of liver cells in RT group was more severe than that in RT-PDP group.

CONCLUSION

The hepatotoxicity of ricin in mice may be reduced by chemical modification.

Immunotherapeutic strategies in neuroblastoma: antitumoral activity of deglycosylated Ricin A conjugated anti-GD2 antibodies and anti-CD3xanti-GD2 bispecific antibodies

BACKGROUND

The antigen GD2 is selectively expressed on the surface of neuroblastoma cells, and is detected by the monoclonal antibody BW704. In this study, we describe the antitumoral capacity of the immunotoxin BW704dgA (BW704 conjugated to deglycosylated ricin A), and of anti-CD3xanti-GD2 bispecific antibodies that are capable of redirecting cytotoxic T cells towards neuroblastoma cells. We further investigate the in vivo activity of BW704dgA immunotoxins in a human neuroblastoma model in SCID mice.

PROCEDURE

BW704dgA immunotoxins were injected i.p. as a single close (48 microg/mouse) on day 4 or divided into three doses on day 4, 5, and 6 after i.v. inoculation of the human neuroblastoma cell line IMR5-75.

RESULTS

The mean survival time (MST) of BW704dgA treated animals was significantly increased (MST 49 days) compared to the control animals treated with irrelevant immunotoxin, unconjugated BW704, or control buffer (MST 33 to 39 days, P < 0.0001), without differences in the application schedules. Anti-CD3xanti-NP antibodies and NP-conjugated GD2-antibodies (BW704-NP) were used in a cytotoxicity assay with cytotoxic T-cells as effectors, and tracer labeled neuroblastoma cell line IMR5 as target cells. Anti-CD3xanti-NP antibodies, together with BW704-NP, showed increased cytotoxic activity compared to the incubation with CD3xanti-NP antibodies alone or with unconjugated anti-GD2. Additionally, a dose-dependent effect of NP-conjugated anti-GD2-antibodies upon the lysis of the target cells could be demonstrated. In this report, we describe two immunotherapeutic approaches using GD2 binding BW704 antibodies, modified as immunotoxin and a bispecific antibody, for the targeting and elimination of neuroblastoma cells.

CONCLUSIONS

We envisage a combined immunotherapeutic regimen consisting of BW704dgA mediated stem cell purging, followed by a systemic treatment with anti-CD3xanti-GD2 bispecific antibodies in neuroblastoma.

Isolated ricin B-chain-mediated apoptosis in U937 cells

We have previously reported that ricin, a toxic lectin that inhibits protein synthesis induced apoptotic cell death. In this study, we have found that isolated ricin CM-B-chain, which has no effect on cellular protein synthesis, induced DNA fragmentation in U937 cells in a dose- and time-dependent manner, albeit it required a longer incubation time and higher concentration than those of holotoxin ricin. Z-Asp-CH2-DCB, a caspase family inhibitor and serine protease inhibitor, 3,4-dichloroisocoumarine (DCI) effectively inhibited the CM-B-chain-mediated DNA fragmentation as well as in ricin. Thus, like ricin, multiple proteases with different substrate specificity may also be involved in the CM-B-chain-mediated apoptotic pathway. Furthermore, BFA inhibited both ricin- and CM-B-chain-mediated DNA fragmentation, suggesting an intracellular vesicle transport system through the Golgi complex may be involved in the apoptotic induction by these proteins as a common feature. On the other hand, cycloheximide (CHA) strongly increased the CM-B-chain-mediated DNA fragmentation, but inhibited ricin-mediated DNA fragmentation. The opposite effects of CHA may reflect the difference in the apoptotic mechanism between ricin and CM-B-chain. In conclusion, our results suggest that ricin-B-chain can induce apoptosis through its lectin activity, but the underlying mechanism may be distinct from that of ricin in which the A-chain contributes profoundly to the apoptotic induction.

Phase II clinical trial of bolus infusion anti-B4 blocked ricin immunoconjugate in patients with relapsed B-cell non-Hodgkin's lymphoma

Immunotoxins, composed of a monoclonal antibody conjugated to a protein toxin, mediate cell death through novel cytotoxic mechanisms. Anti-B4-blocked ricin (anti-B4-bR) recognizes CD19-positive cells, which includes most B-cell non-Hodgkin's lymphomas (NHLs). Previous Phase I clinical studies of anti-B4-bR, using both bolus and continuous dosing regimens, demonstrated no safety or efficacy advantage to the continuous infusion regimen. This Phase II trial in 16 patients with relapsed CD19-positive NHL was conducted to evaluate the efficacy of anti-B4-bR when administered at the previously established maximum tolerated dose using a daily bolus for a 5 consecutive days schedule. Serum pharmacokinetics were measured in selected patients. Tissue samples of involved lymph nodes and bone marrow were also obtained from a portion of patients for determination of anti-B4-bR penetration into tissues. Toxicity was similar to what has been described previously for anti-B4-bR and consisted mainly of reversible elevations of hepatic transaminases and mild to moderate thrombocytopenia. No sustained clinical responses were documented. Pharmacokinetic measurements demonstrated that serum levels compatible with 3 logs of cell kill in vitro could be sustained for several hours in most patients. Immunohistochemical analysis of tissue samples provided some insight into the low efficacy. The immunotoxin could be detected in three of the four bone marrow aspirate samples but in only two of the seven lymph node specimens. Thus, anti-B4-bR, using a single daily bolus for a 5 consecutive day schedule, is not an active agent in relapsed NHL. Poor penetration into certain sites of disease may be one explanation for its lack of efficacy.

A toxin conjugate containing transforming growth factor-alpha and ricin A specifically inhibits growth of A431 human epidermoid cancer cells

The inhibitory effect of human epidermoid cancer cells A431 caused by conjugate toxin containing transforming growth factor (TGF-alpha) and ricin A was studied. TGF-alpha is a protein with 50 amino acids that specifically binds and stimulates phosphorylation of cell surface epidermal growth factor receptor (EGFR) and, subsequently, triggers cell proliferation. TGF-alpha as a ligand for EGFR is internalized upon binding and decomposed within lysosome. Lectin ricin is contained in the castor bean plant. The lectin consists of two subunits, ricin A and B. Toxic ricin A binds to ribosome and inhibits protein synthesis of target cells. In view of the abundance of EGFR in human cancer cells, the receptor-mediated endocytosis with the conjugate toxin composed of TGF-alpha and ricin A was synthesized, purified and tested for growth inhibition in both normal and tumor cells. The cytotoxicity of the conjugate was studied within the range of 10(-12) and 10(-8) M and IC50 was found to be 10(-10) M for human A431 epidermoid cells that over-express EGFR. Compared to A431 cells, the brain metastatic variant of human Non-Small Cell Lung Cancer (NSCLC) H226Br squamous cells showed a reduced inhibitory effect. On the other hand, no inhibitory effect was found with other NSCLC cells studied and normal human lung cells because of the fewer available EGF binding sites on the surface of the cells. These results indicated that the amount of the available EGFR contributes to the cytotoxic effect on human cancer cells, thereby demonstrating involvement of the receptor-mediated endocytosis of the conjugate. The result from 12labeled EGF-mediated competition assay further demonstrated the specific inhibition activity conferred by TGF-alpha and ricin A conjugation. Due to poor recovery of the chemical conjugation, modification in the form of a recombinant toxin is needed for further in-depth studies.

Autocrine growth factors and neuroendocrine markers in the development of small-cell lung cancer

Two different clinical trials using biological agents directed against an autocrine growth factor and a surface marker of neuroendocrine differentiation have been used for patients with relapsed small-cell lung cancer. In a phase II trial, an antibody (2A11) directed against the autocrine growth factor gastrin-releasing peptide has been used to treat patients with relapsed small-cell lung cancer. One of 12 evaluable patients treated with 2A11 250 mg/m2 three times weekly for 4 weeks achieved a complete response. An antibody directed against the neural cell adhesion molecule has been linked to a modified ricin molecule. This immunotoxin, N901-bR, has undergone phase I testing, and a recommended phase II dose of 30 micrograms/kg/day for 7 days by continuous infusion has been determined. In the phase I trial, one of 21 patients with relapsed or refractory small-cell lung cancer had a partial response to this treatment. Therefore, it appears that an antibody directed against an autocrine growth factor and an immunotoxin directed against a surface marker of neuroendocrine differentiation can inhibit the growth of small-cell lung cancer in vitro and in vivo; both produced some evidence of antitumor activity in patients. Further studies with agents directed against autocrine growth factors and surface markers of neuroendocrine differentiation appear warranted.

Immunotoxin therapy of small-cell lung cancer: a phase I study of N901-blocked ricin

PURPOSE

Immunotoxins could improve outcome in small-cell lung cancer (SCLC) by targeting tumor cells that are resistant to chemotherapy and radiation. N901 is a murine monoclonal antibody that binds to the CD56 (neural cell adhesion molecule [NCAM]) antigen found on cells of neuroendocrine origin, including SCLC. N901-bR is an immunoconjugate of N901 antibody with blocked ricin (bR) as the cytotoxic effector moiety. N901-bR has more than 700-fold greater selectivity in vitro for killing the CD56+ SCLC cell line SW-2 than for an antigen-negative lymphoma cell line. Preclinical studies suggested the potential for clinically significant cardiac and neurologic toxicity. We present a phase I study of N901-bR in relapsed SCLC.

PATIENTS AND METHODS

Twenty-one patients (18 relapsed, three primary refractory) with SCLC were entered onto this study. Successive cohorts of at least three patients were treated at doses from 5 to 40 microg/kg/d for 7 days. The initial three cohorts received the first day's dose (one seventh of planned dose) as a bolus infusion before they began the continuous infusion on the second day to observe acute toxicity and determine bolus pharmacokinetics. Toxicity assessment included nerve-conduction studies (NCS) and radionuclide assessment of left ventricular ejection fraction (LVEF) before and after N901-bR administration to fully assess potential neurologic and cardiac toxicity.

RESULTS

The dose-limiting toxicity (DLT) of N901-bR given by 7-day continuous infusion is capillary leak syndrome, which occurred in two of three patients at the dose of 40 microg/kg (lean body weight [LBW])/d. Detectable serum drug levels equivalent to effective in vitro drug levels were achieved at the 20-, 30-, and 40-microg/kg(LBW)/d dose levels. Specific binding of the immunotoxin to tumor cells in bone marrow, liver, and lung was observed. Cardiac function remained normal in 15 of 16 patients. No patient developed clinically significant neuropathy. However, a trend was noted for amplitude decline in serial NCS of both sensory and motor neurons. One patient with refractory SCLC achieved a partial response.

CONCLUSION

N901-bR is an immunotoxin with potential clinical activity in SCLC. N901-bR is well tolerated when given by 7-day continuous infusion at the dose of 30 microg/kg(LBW)/d. Neurologic and cardiac toxicity were acceptable when given to patients with refractory SCLC. A second study to evaluate this agent after induction chemoradiotherapy in both limited- and extensive-stage disease was started following completion of this study.

Elimination of neuroblastoma and small-cell lung cancer cells with an anti-neural cell adhesion molecule immunotoxin

BACKGROUND

The development of immunotoxins has been hampered by difficulties, particularly in solid tumors, of finding appropriate target antigens and of linking sufficiently potent toxins.

PURPOSE

We evaluated the tissue specificity of an immunotoxin, N901-blocked ricin (N901-bR), and assessed its potential for eliminating neural cell adhesion molecule (NCAM)-positive tumor cells in conditions appropriate for in vitro purging, prior to autologous stem cell transplantation, and its potential for myelosuppression. N901-bR consists of a monoclonal antibody (MAb), N901, directed against CD56, an antigen of the family of NCAMs, covalently linked to blocked ricin as the cytotoxic effector moiety.

METHODS

The tissue specificity of the N901 MAb and the N901-bR immunotoxin was tested against a wide array of human tumor tissues and normal human tissues by immunohistochemical staining. The cytotoxic activity of N901-bR was tested against both small-cell lung cancer (SCLC) cells and neuroblastoma cells, either alone or among normal bone marrow mononuclear cells, and the efficacy of this treatment to specifically eliminate these cells was evaluated in a limiting dilution assay. In addition, normal bone marrow mononuclear cells were incubated with N901-bR, and the toxic effects of the immunotoxin on normal hematopoietic progenitors was evaluated.

RESULTS

N901 and N901-bR exhibited specificity for several neoplasms of neuroectodermal origin, including SCLC and neuroblastoma. Staining of normal tissues was essentially limited to various neuroendocrine cells, cardiac muscle cells, and cells in peripheral nerve tissue. We observed a time- and dose-dependent elimination of tumor cells in vitro, with three logs (i.e., > 99.9%) of malignant cells being killed following only 5 hours of exposure to 10 nM N901-bR. Unconjugated N901 MAb specifically blocked the elimination of NCAM-positive cells by N901-bR, whereas neither an isotype-matched control MAb nor galactose (the ligand of native ricin) had any effect on the activity of the immunotoxin, confirming the specificity of its cytotoxic activity. Importantly, N901-bR used under optimal conditions for in vitro tumor cell depletion was not toxic to hematopoietic precursors.

CONCLUSIONS

N901-bR has the properties required to target CD56, an antigen present not only on cells from a large number of cancers of neuroendocrine origin, but also on some important normal tissues. In addition, treatment with this immunotoxin results in the highly effective and specific elimination of neuroblastoma and SCLC cells and does not affect normal hematopoietic progenitors.

IMPLICATIONS

N901-bR may have clinical utility for purging of neuroblastoma cells and SCLC cells before autologous stem cell transplantation. Further toxicology studies are warranted to assess the potential of N901-bR for in vivo administration.

Intranasal stimulation of long-lasting immunity against aerosol ricin challenge with ricin toxoid vaccine encapsulated in polymeric microspheres

Intranasal (i.n.) immunization with ricin toxoid (RT) vaccine encapsulated in poly (lactide-co-glycolide) microspheres (RT-PLG-Ms) and poly (L-lactide) microspheres (RT-PLA-Ms) stimulated systemic and mucosal immune responses and protected mice from aerosolized ricin intoxication. High titers of anti-ricin IgG2a were stimulated in the serum of mice with one or two doses of RT-Ms 6 weeks postimmunization. However, in the lungs, no IgG2a or total IgG was elicited either with RT-Ms or with aqueous RT. At 6 weeks postimmunization, a single dose of the RT-Ms stimulated secretory IgA (sIgA) in the lungs of four of six mice, but a second immunizing dose did not enhance the stimulation. A single dose of aqueous RT vaccine failed to stimulate sIgA in the lungs, while, a second dose induced sIgA in 50% of the mice. One or two i.n. doses of RT-Ms protected most of the mice against lethal aerosol-delivered ricin toxin 6 weeks postimmunization. In contrast, protection was absent or marginal after one or two doses of aqueous RT vaccine. In both studies, the protection against lethal aerosol challenge was significantly better with one dose of RT-Ms than with two doses of aqueous vaccine, which may be attributed to the induction of sIgA in the lungs and the serum. Duration of the IgG2a and IgA in the serum, particularly that of IgG2a was much longer after the administration of RT-Ms than after the aqueous vaccine. The geometric mean IgG2a titers stimulated with two doses of RT-Ms remained high during 40 weeks postimmunization and were up to 25 times higher than the titers induced with aqueous RT vaccine. After 6 weeks, the IgG2a induced by two doses of aqueous vaccine was no longer detectable. Persistence of antibody response was predictive of efficacy. At 1 year postimmunization with two doses of RT-Ms, 100% of mice were protected against lethal ricin challenge. However, at the same time no protection was afforded by two doses of aqueous RT. The results of the present study consistently demonstrated the advantages of microencapsulated RT vaccine to stimulate effective and long-lasting protection by i.n. administration.

Immunotoxin therapy of small-cell lung cancer. N901-blocked ricin for relapsed small-cell lung cancer

Despite its initial chemosensitivity, small-cell lung cancer (SCLC) is rarely cured with chemotherapy alone, and fewer than 5% of patients are alive at 5 years. Immunotoxin therapy appears to offer promise in treating the minimal residual disease that remains after induction chemotherapy. We have studied N901-bR in patients with relapsed SCLS. N901-bR consists of the N901 monoclonal antibody (MoAb) and blocked ricin, an altered ricin molecule in which the galactose binding sites of the ricin B-chain which mediate nonspecific binding of the toxin are blocked through the covalent binding of ligands. N901 is an anti-NCAM (CD56) MoAb which binds to SCLC tumors and cell lines, cardiac muscle, natural killer (NK) cells, and peripheral nerve. N901-bR showed a 2.7 log greater in vitro cytotoxicity to the CD56-positive cell line SE-2 than to the antigen-negative Namalwa cell line. Nineteen patients with relapsed antigen-negative Namalwa cell line. Nineteen patients with relapsed and/or refractory SCLC have been entered into a phase I study at doses ranging from 5 to 40 micrograms/kg/day given as a 7-day continuous infusion. The dose-limiting toxicity is capillary leak syndrome observed in two thirds of the patients treated at 40 micrograms/kg/day. One patient at the maximum tolerated dose, 30 micrograms/kg/day x 7 days, has achieved a partial response to N901-bR. No patient has developed clinically significant peripheral or central neuropathy. We plan to begin a phase II study of N901-bR following induction chemotherapy in patients with SCLC.

Selective killing of laryngeal carcinoma cells by a monoclonal immunotoxin

To overcome the lack of selectivity of present anticancer drugs, an alternative approach is described for laryngeal carcinoma cells. An immunotoxin is synthesized containing the toxic chain from ricin coupled to a monoclonal antibody against a laryngeal carcinoma cell line. The newly formed monoclonal immunotoxin selectively kills cells from laryngeal carcinoma cell lines, whereas various normal living cells are not affected. Control experiments rule out the possibility that toxicity can be attributed to contaminating traces of whole ricin toxin. The results indicate that minute amounts of the immunotoxin are selectively cytotoxic for living laryngeal carcinoma cells in vitro.

Elimination of trinitrophenol-specific antibody response by antigen-toxin conjugates

The aim of the experiments presented here is the selective inhibition of an antigen-specific immune response. Antigen receptors are used as targets for conjugates of antigens and toxin to eliminate antigen-reactive cells. The trinitrophenol (TNP)-specific immune response can be specifically abrogated by incubating the TNP-keyhole limpet haemocyanin (KLH)-primed spleen cells with TNP-ricin or TNP-chicken IgG-ricin conjugates before in vitro stimulation with TNP-KLH. The rate of elimination is dose-dependent and related to the degree of TNP moieties bound to the toxin molecule. The specificity of the toxin conjugates is demonstrated by treating sheep erythrocyte-primed spleen cells with TNP-IgG-ricin conjugates. These results may have therapeutic relevance for treating autoimmune diseases.

Immunotoxins containing ricin A or B chains with modified carbohydrate residues act synergistically in killing neoplastic B cells in vitro

Partially deglycosylated (dg) ricin A and B chains were coupled to rabbit anti-human Ig (IT-dg-A) and goat anti-rabbit Ig (IT-dg-B), respectively. The deglycosylation was accomplished by treating intact ricin (before separation into its constituent chains) with a mixture of sodium metaperiodate and sodium cyanoborohydride. The two immunotoxins (ITs) acted synergistically in vitro to kill Daudi cells, albeit to a four-to-eight-fold lesser degree than ITs prepared with native A and B chains.

EGF-ricin A conjugates: kinetic profiles of cytotoxic effects and resistant cell variants

We have conjugated epidermal growth factor (EGF) with the A subunit of ricin (RICa) via a dithiopropionyl linkage. The EGF-RICa conjugate was competitive with [125I]EGF for binding to cell surface EGF receptors. Entry of the conjugate into cells was seen within 10 min at 37 degrees C and inhibition of protein synthesis was seen within 90 min. The blockage of protein synthesis continued for more than 20 h in sensitive cells. Protein synthesis in EGF receptor-deficient cells was not affected. The conjugate killed human epidermoid carcinoma (A431) cells, mouse 3T3 fibroblasts and Chinese hamster lung (CHL) cells with identical efficiencies (ED50 = 2 X 10(-9) M). The EGF-RICa conjugate was used as a selection agent and several resistant variants were isolated from CHL cells. These variants showed various degrees of [125I]EGF binding capacity. Some other characteristics of the variants are also described.

Preparation and properties of a hybrid toxin of modeccin A-chain and ricin B-chain

Hybrid molecules were prepared from the A- and B-chains of the two toxic lectins ricin and modeccin by dialyzing mixtures of isolated chains to allow a disulfide bridge to be formed between them. Whereas the hybrid consisting of ricin A-chain and modeccin B-chain was non-toxic, the converse hybrid, modeccin A-chain/ricin B-chain, was even more toxic to Vero cells than were the parent toxins, native ricin and modeccin. A number of drugs (NH4Cl, monensin, trifluoperazine, verapamil, ionophore A23187) which protect cells against modeccin, but not against ricin, protected to some extent against the toxic hybrid, but less so than against native modeccin. The possibility is discussed that the modeccin A-chain of the hybrid may enter the cytosol by two routes, one which is highly efficient and identical to that used by native modeccin and another less efficient one which cannot be used by native modeccin.

Studies on the galactose-binding site of ricin and the hybrid toxin Man6P-ricin

N-acetylimidazole (NAI) was used to O-acetylate the plant seed toxin ricin. O-acetylation of one to two tyrosine residues per molecule of ricin inhibited ricin binding to Sepharose 4B and decreased toxicity by 90% in a protein synthesis inhibition assay in HeLa cells. Lactose, known to block the binding site on the ricin B subunit, protected ricin from NAI modification of binding or toxicity. Thus NAI, under these conditions, can be a lactose site-specific inhibitor. The lactose site-specific modification of the hybrid toxin, Man6P-ricin, performed under the same conditions, exhibited the same 90% inhibition of Man6P receptor-mediated toxicity as the galactose-containing receptor-mediated toxicity of either Man6P-ricin or ricin. Thus the ricin B chain lactose-binding site appears to be essential for the high potency of Man6P-ricin via the new cell type-specific Man6P receptor. Treatment of fibroblasts with neuraminidase exposes galactose residues, thus increasing the sensitivity to ricin eight fold. The Man6P receptor-mediated toxicity of Man6P-ricin is not affected by this treatment, although the galactose-inhibited route is potentiated eight fold. The Man6P-ricin hybrid appears to require the ricin B chain galactose-binding site to enter the cytosol after initially binding to the Man6P receptor. These data provide some insights into the proper design of hybrid toxins. We discuss a number of possible models for hybrid toxin entry.

Ricin linked to monophosphopentamannose binds to fibroblast lysosomal hydrolase receptors, resulting in a cell-type-specific toxin

The receptor specificity of the plant seed toxin ricin, which ordinarily binds to galactose-containing receptors, has been altered by coupling monophosphopentamannose residues to ricin by reductive amination and by reversibly binding lactose to the modified ricin. The added monophosphopentamannose residues provide ricin with the recognition factor common to fibroblast lysosomal hydrolases and enable the modified ricin (Man6P-ricin) to bind to the fibroblast Man6P receptor and inhibit protein synthesis in the cells via this receptor. The addition of lactose to Man6P-ricin saturates the galactose site on Man6P-ricin and prevents the binding of Man6P-ricin to cells via galactose-containing ricin receptors. The Man6P receptor-mediated toxicity of Man6P-ricin, identified in human fibroblasts by competition by Man6P and blockade by alkaline phosphatase treatment, was not detected in HeLa cells or human amnion cells. Consequently, in the presence of lactose, the fibroblasts were 8 and 13 times more sensitive than amnion and HeLa cells, respectively. These results show that highly toxic cell-type-specific reagents can be made by the proper alteration of toxin receptor specificities. An attempt to construct a highly toxic altered toxin by adding Man6P residues to diphtheria toxin fragment A was unsuccessful. A possible explanation is that in Man6P-ricin the ricin B chain performs some entry function, even though the initial binding step occurs at the Man6P receptor.

Preparation of concanavalin A-ricin A-chain conjugate and its biologic activity against various cultured cells

For the development of therapeutic agents that possess tissue-specific carriers, a method was devised to synthesize an artificial protein hybrid conjugate containing a moiety which binds to a cell membrane receptor and an active fragment of a toxic protein. By the introduction of an activated sulfhydryl group into concanavalin A (Con A), a conjugate of Con A and the ricin A-chain cross-linked with a disulfide linkage was synthesized. The purified conjugate was studied with regard to its inhibitory activity against protein synthesis in cell-free and cultured cell systems. The Con A-rich A-chain conjugate retained about one-third the inhibitory activity of ricin in a cell-free protein synthesis system. It also was highly toxic to cultured normal cells. These results indicate that the conjugate is a structural and functional analog of ricin and that the original membrane-binding chain (B-chain of ricin) could be replaced by Con A. Transformed cells were insensitive to this conjugate and required a longer preincubation time. The sensitivity of the normal cells was reduced in the presence of local anesthetics.

Binding of ricin A chain to rat liver ribosomes: relationship to ribosome inactivation

Ricin A chain was radioactively labeled using reductive alkylation, lactoperoxidase catalyzed iodination, and reaction with iodoacetamide or N-ethylmaleimide (NEM). The inhibition of cell-free rat liver protein synthesis by the modified A chains and the ribosome binding characteristics of each of the labeled derivatives was examined. [3H] NEW was found to quantitatively react with the A chain sulfhydryl group normally involved in a disulfide bond with the B chain in intact ricin. Labeling the protein with [3H] NEM had no effect on the in vitro inhibition of protein synthesis by the A chain. [3H] NEM-labeled A chain binds to rat liver ribosomes in a manner which is dependent on the concentrations of NaCl and Mg2+. At optimal Mg2+ concentration (5.5 mM), A chain binding to ribosomes is saturable and fully reversible either by dilution of the reaction mixture or by addition of unlabeled A chain. At 5.5 mM Mg2+, A chain was found to bind to a single site on rat liver ribosomes with a dissociation constant of 6.2 x 10(-8) M. [3H] NEM-labeled A chain did not bind to isolated 40S ribosomal subunits and bound to 60S ribosomal subunits with a 1 : 1 molar stoichiometry and a dissociation constant of 2.2 x 10(-7) M. The relationship between ribosome binding and A chain inhibition of eucaryotic protein synthesis is discussed.