[CANCER RESEARCH 52, 4448-4452, August 15, 1992] Cytotoxicity of Streptavidin-blocked Biotinyl-Ricin Is Retrieved by in Vitro Immunotargeting via Biotinyl Monoclonal Antibody1

Bilha Schechter, Ruth Armin,2 and Meir Wilchek

Departments of Chemical Immunology [B. S., R. A.J and Membrane Research and Biophysics [M. W.], The Weidmann Institute of Science, Rehorot, 76100

ABSTRACT of the B chain (5, 8). Such A chain immunotoxins were shown to be specifically toxic to -reactive target cells. How The streptavidin- system has been used to immunotarget whole ricin to tumor cells in a system that overcomes ricin-nonspecific cyto- ever, various studies have shown that ricin A chain immuno toxicity. Biotin was linked to ricin via a disulfide-containing reagent, toxins may sometimes be inactive, or not as active as expected sulfosuccinimidyl-2-(biotinamido)ethyl-l,3'-dithiopropionate. The (9), and often require potentiating agents (10, 11). Addition of product, l>iotiiiyl-.Y,.S'-ricin(b-ricin), retained most of its in vitro cyto ricin B chain-antibody conjugate (12, 13) or free B chain toxic activity against human epidermoid carcinoma (KB) cells. Com- (14, 15) has shown to augment weak ricin A immunotoxin plexing b-ricin to streptavidin resulted in >99% loss of its cellular activity, demonstrating the potentiating effect of the B chain on toxicity which is associated with loss of cell-binding activity. The A chain function. The property of B chain in intact ricin im streptavidin-b-ricin complex could, however, be targeted to KB cells via munotoxin that facilitates internalization of the A chain to the the biotinylated monoclonal antibody 108 which is specific to the epi dermal growth factor overexpressed on KB cells. The complex cytosol appears to be, at least in part, independent of galactose did not regain its activity if the specific antibody was not biotinylated or recognition (16). if the biotinylated antibody was of a different specificity. Streptavidin is These observations have initiated new approaches to the use thus used to block b-ricin, presumably due to a storie restraint of the of whole ricin immunotoxin, mainly by introducing new con streptavidin on the ricin B-chain, and to bridge it to biotinyl antibody jugation methods (16-18) or blocking techniques (19) that ob recognizing the target cell. could not replace streptavidin in this struct the B chain galactose-binding sites. In the present study system since a complex between b-ricin and avidin retained a major part we have used streptavidin to block the cytotoxic activity of (60%) of ricin cytotoxic activity. This is attributed to the nonspecific b-ricin3 in which the biotin group is linked to ricin via a disul- binding of avidin to cells in vitro, including the KB cells. It is suggested that b-ricin is blocked by both streptavidin and avidin, but once the fide bond. Whole ricin can thus be specifically targeted to cells complex gains access to the cell surface, its cytotoxic activity is specif when incorporated in a cytotoxically blocked streptavidin-b- ically retrieved. ricin complex carrying free biotin binding sites, that can inter act with biotinyl groups on specific associated with INTRODUCTION the cell surface. It is assumed that ricin activity is regained following internalization and release of the ricin A chain. Ricin is a toxic plant glycoprotein composed of two polypep- tides, A chain and B chain, linked by a single interchain disul- fide bond (1). The A chain is a potent inhibitor of protein MATERIALS AND METHODS synthesis by inactivating 60S ribosomal subunits (2). The B Radiolabeling chain consists of two domains, each possessing a lactose bind ing site presumed to be associated with the strong binding af Ricin and b-ricin (0.2 mg/0.2 ml 0.05 Msodium phosphate buffer, pH finity of ricin to cells (3). The B chain binds to D-galactose or 8) were radiolabeled with 0.5 mCi I25I (Amersham International pic, /V-acetylgalactosamine-terminal residues of oligosaccharides on Amersham, United Kingdom) using the chloramine-T method (20). cell surface membranes and facilitates transfer of the toxin A After purification on Sephadex G-25 the specific activity was in the chain into the cytosol (4). Several approaches have been used to range of 0.5-1.0 ¿iCi/Mgprotein, target ricin to tumor cells. Linkage of a tumor-specific antibody to the whole ricin molecule generates an immunotoxin which Cells and Antibodies combines the tumor specificity of the antibody and the potent The KB human tumor cell line derived from oral epidermoid carci toxicity of ricin (5). However, the intact ricin immunotoxin noma was obtained from the American Type Tissue Culture Collection. cannot be used in vivo since it generates nonspecific effects due The cells were maintained in DMEM supplemented with 10% heat- to the B chain galactose-binding activity. However, whole ricin inactivated fetal calf serum, 2 HML-glutamine, penicillin, and strepto immunotoxins are still powerful reagents for the in vitro elim mycin (Biolab, Jerusalem, Israel) at 37°Cin 5% CO2 in air. The cells ination of T-cells (6) or leukemia cells (7) prior to transplanta were harvested by incubation with 0.1% trypsin and 1 HIMEDTA (Bi tion. Such a procedure can be carried out only in vitro since high olab) at 37°Cfor 5 min. The mAb 108 IgG2a hybridoma line, specific concentrations of galactose are needed to inhibit the B chain to the epidermal growth factor receptor overexpressed on KB cells, was binding, to allow conjugate reactivity only via its antibody rec generated by immunizing mice with the KB cells (21). The antibody ognition site. To overcome the in vivo effects of whole ricin fraction was isolated by ammonium sulfate precipitation at 45% satu immunotoxin, the antibody has been attached to the ricin A ration. Monoclonal antibody against CEA, which does not bind to KB chain, generating a conjugate in which the antibody carrying the cells, was used as control antibody. tumor cell specificity is used to replace the cell-binding function

Received 2/17/92; accepted 6/8/92. 3 The abbreviations used are: b-ricin, biotinyl-5,5-ricin; NHS-S.S-biotin, sulfo- The costs of publication of this article were defrayed in part by the payment of succinimidyl-2-(biotinamido)ethyl-l,3-dithiopropionate: mAb, monoclonal anti page charges. This article must therefore be hereby marked advertisement in accord body; b-mAb, biotinyl monoclonal antibody: HABA, 2(4-hydroxyazobenzene) ance with 18 U.S.C. Section 1734 solely to indicate this fact. benzoic acid; PBS, 0.15 MNaCI - 0.02 Mphosphate buffer, pH7.2; CEA, carcino- 1This research was supported by a grant from Makor Chemicals, Jerusalem, embryonic ; IC$o. concentration of ricin causing 50% inhibition of ['111 leucine incorporation by the cells; DMEM. Dulbecco's modified Eagle's medium; Israel. 2 To whom requests for reprints should be addressed. b-mAb, biotinylated monoclonal antibody. 4448

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research. IMMUNOTARGETING OF BIOTINYL-RICIN VIA STREPTAVIDIN

Conjugating Biotin to Ricin and Antibody- Binding of I25l-Streptavidin to b-mAb 108-coated KB Cells

Conjugating NHS-5,5-Biotin (Pierce, Rockford, IL) to Ricin. Ricin KB cell monolayers in 24-well plates were incubated for 40 min with (Sigma Chemical Co., St. Louis, MO; 0.93 mg) in 0.3 ml 0.05 Msodium 10 Mg/ml b-mAb 108, mAb-108, or b-mAb anti-CEA in 0.2 ml DMEM bicarbonate at pH 8.5 was mixed with NHS-S.S-biotin (0.15 or 0.4 mg) at room temperature. The monolayers were then washed twice with dissolved in 0.3 ml ice-cooled water. The reaction proceeded for 3 h at PBS and 0.2-ml samples containing 125I-streptavidin in a biotin-free 0°Cafter which the b-ricin was extensively dialyzed against PBS. medium (DMEM supplemented with dialyzed fetal calf serum) were added. After 30 min incubation at 37°Cthe cells were washed and Conjugating Biocytin Hydrazide to Modified Carbohydrates in An tibody. Monoclonal antibodies (mAb; 0.6 mg) were oxidized in 1 ml of dissolved in 0.5 M NaOH, and the cell-bound radioactivity was moni 0.1 M sodium acetate buffer, pH 5.5, containing 10 imi sodium meta- tored as described above. periodate for l h at 4°C.Theprotein was then dialyzed against the same buffer and biocytin hydrazide (Sigma; 0.6 mg) was added together with Immunotargeting in Vitro sodium cyanoborohydride (1.2 mg). The coupling reaction proceeded for l h at 4°Cafter which the b-mAb was dialyzed extensively against Experiments of immunotargeting streptavidin-b-ricin to KB cells us PBS. Retention of antibody binding activity of b-mAb 108 was deter ing b-mAb 108 were carried out in microtiter plates in a biotin-free mined by competing with 125I-mAb 108 for the binding to KB cells with medium. KB cell monolayers were preincubated with 0.1-ml aliquots of b-mAb 108, b-mAb anti-CEA or unbiotinylated mAb 108, at a concen reference to cold mAb 108, as described before (22). tration of 20 Mg/ml. After 30 min incubation at 4°Cthe antibody- The molar substitution ratio (mol biotin coupled/mol protein) was containing medium was replaced by fresh medium and the streptavidin- determined spectrophotometrically according to the method described b-ricin complex (at a 5:1 molar ratio) was added. Incubation proceeded by Green (23). The assay uses the change in absorbance at 500 nm for 18 h at 37°Cafter which the cells were pulsed and processed as which occurs when biotin-binding sites of avidin (S. C. Belovo, Bas- togne, Belgium) saturated with HABA (Sigma) interact with free or described above. bound biotin. Due to its higher binding affinity to avidin, biotin replaces HABA and the absorbance at 500 nm decreases. The average number of RESULTS biotin groups per protein molecule was calculated by titrating a solution of 0.3 mg/ml avidin and 0.025 mg/ml HABA with biotinyl-protein in Cytotoxicity of b-Ricin to KB Cells. Experiments were first comparison to a standard solution of biotin. Analysis of the prepara performed to assess the effect of on ricin activity tions showed that ricin reacted with 0.15 or 0.4 mg NHS-S.S-biotin and to define the conditions for obtaining cytotoxically active, gave rise to conjugates with molar substitution ratios of 2.6 and 4.7, fully biotinylated ricin preparations, that will be inhibited by respectively. The molar substitution ratio for antibody ranged between 3.2 and 4.7 in different b-mAb preparations. streptavidin to greater than 99%. Fig. 1 illustrates the proper ties of two b-ricin preparations differing in their level of biotin To estimate the efficiency of ricin biotinylation (nonbiotinylated/ biotinylated ricin), b-ricin preparations were chromatographed on avi- substitution. The two preparations were tested for their cyto din-containing affinity resin equilibrated with PBS. The effluents were toxic activity against human KB cells in comparison to native tested for biological activity in an in vitro cytotoxicity assay in compar ricin. The extent of their protein biotinylation, i.e., the occur ison to their unfractionated counterparts. rence of unbiotinylated species in each preparation, was evalu ated by comparing the two biotinylated ricins before and after In Vitro Cytotoxicity Assay on avidin-resin. As is shown in Fig. 1, ricin itself reduced the capacity of the cells to incorporate [3H]- Inhibition of protein synthesis was used as a measure for the in vitro cytotoxic activity exerted by the various ricin derivatives. KB cell mono- leucine by 50% (IC50) at a concentration of 0.005 nM. Conju gates of b-ricin with molar substitution ratios of 2.6 (low) and layers were prepared in wells of microtiter plates (Falcon; Becton Dick inson) by plating 2 x IO4 trypsinized cells in 0.1 ml supplemented 4.7 (high) were almost as active as ricin itself, with IC50 values DMEM medium. After 24 h incubation at 37°C,varyingconcentrations of 0.006 and 0.008 n\i, respectively. Testing the cytotoxic ac of the ricin derivatives were added to the cultures and incubation proc- tivity of avidin-resin effluents following chromatography of ceeded for 24 h. Control cells were incubated in medium alone. The equivalent doses of the two biotinyl-ricin preparations showed cells were pulsed with [3H]leucine (Nuclear Research Center, Negev, that more than 99.9% in each preparation was attached to the Israel) (1 MCi/well)during the last 6 h before termination of culture. The cells were then trypsinized and harvested onto glass fiber filters, and the incorporation of radioactivity was monitored using a Kontron liquid 100 - scintillation counter. The results were expressed as percentage of inhi bition of [3H]leucine incorporation relative to control cells. Preformed ¡fi « complexes between b-ricin and either avidin or streptavidin (Boehring- 80- er-Mannheim GmbH) were prepared by mixing the proteins at the desired molar ratios in PBS for 20-30 min at room temperature prior to the assay. The in vitro experiments described in this paper were 60- carried out in duplicates and standard deviation of the duplicate sam ples did not exceed 10% of the mean. 40 - Cell-binding Activity of Streptavidin-complexed b-Ricin 20 - KB cell monolayers were prepared in 24-well plates (Falcon) by plating 2-3 x IO5 trypsinized cells in 0.4 ml supplemented DMEM. Following 24 h incubation at 37°C,the medium was replaced by fresh •¿â€¢*' medium (0.2 ml) and I25l-ricin, I25l-b-ricin or preformed complexes 0.001 0.01 0.1 10 between 125I-b-ricin and streptavidin at a molar ratio of 1:6 were added to the cultures in the absence or presence of o-galactose. Incubation icin (nM) proceeded for l h at 37°C,the cells were then washed 3 times with 1 ml Fig. 1. Effect of biotinylation on the cytotoxic activity of ricin. Inhibition of pHJIeucine incorporation into KB cells following 24 h incubation with ricin (O), PBS and dissolved in 0.5 MNaOH, and the cell-bound radioactivity was b-ricin-low (•),b-ricin-high (A) before ( ) or after ( ) chromatography on monitored in a Kontron gamma counting system. avidin-Sepharose. 4449

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research. IMMUNOTARGETING OF BIOTINYL-RICIN VIA STREPTAVIDIN avidin-resin. Native ricin similarly chromatographed was as ac 100- tive as unchromatographed ricin. These results indicate that under the conjugation conditions used in these experiments, attachment of NHS-S.S'-biotin to ricin gave rise to preparations 80- e retaining most of the cytotoxic activity of ricin. Less than >. 0.01% of the ricin in each preparation was unbiotinylated. 60 Cytotoxicity of b-Ricin Complexed to Streptavidin or Avidin. To test the effect of Streptavidin on b-ricin cytotoxicity, the two o b-ricin preparations, as well as unbiotinylated ricin, were incu 40 bated with Streptavidin at a molar ratio of 1:20, and the cyto toxic activity against KB cells was tested in comparison to their e 20- free counterparts. As shown in Fig. 2, the activity of unbioti o nylated ricin was not affected by Streptavidin, indicating that JC nonspecific blocking exerted by Streptavidin did not take place. C. The complexes formed between the two b-ricin preparations 0.1 1 10100 and Streptavidin lost a major part of their cytotoxic activity, but Streptavldln:b-rlcln molar ratio as is shown in Fig. 2, the degree of biotinylation had an impor Fig. 3. Titration of the streptavidin:b-ricin blocking activity. Inhibition of [3H]- leucine incorporation into KB cells following 24 h incubation with preformed tant effect on the activity of the complex: the complex formed complexes between b-ricin and Streptavidin (•)atdifferent molar ratios. (0.1 nM with the low b-ricin preparation was not fully blocked (still b-ricin). exerting 10% of its toxic activity); whereas the one formed with the high b-ricin was effectively blocked by Streptavidin, with a 300 -i residual activity of less than 0.5%. Fig. 2 also illustrates the effect of avidin on the cytotoxic activity of b-ricin (high). As is shown, avidin was unable to effectively block b-ricin, since a complex between b-ricin and Ricin avidin was almost as active as the unblocked b-ricin. -t-Gal Titration of the Streptavidin blocking activity was performed in order to determine the lowest Streptavidin to b-ricin molar ratio required to block the cytotoxic activity of b-ricin. Com plexes between a constant amount of b-ricin and Streptavidin at different molar ratios were prepared, and their cytotoxic activ ity was examined (Fig. 3). As shown, a complex between Streptavidin and b-ricin at a molar ratio of 3:1 was essentially inactive, indicating that this ratio was sufficient to block b-ricin Ricin b-Ricin Strept.--b-ricin at a concentration (0.1 nM) that by itself caused 98% toxicity to Fig. 4. Binding of ricin, b-ricin, and the streptavidin-b-ricin complex (molar the cells. ratio, 6:1; ricin concentration, 5 HM)to KB cell monolayers (3 x 10* cells) in the Binding of Streptavidin-b-Ricin Complexes to KB Cells in the absence or presence of o-galactose (Gal, 20 IHM)following l h incubation at 37'C. Absence and Presence of D-Galactose. To test whether the in hibition of b-ricin cytotoxicity by Streptavidin is attributed to complex (molar ratio 6:1; ricin concentration, 5 HM)and the diminished cellular binding, KB-cell monolayers were incu cell-bound radioactivity was monitored. As is shown in Fig. 4, bated with 125I-ricin, 125I-b-ricin and a Streptavidin-I25l-b-ricin the cellular binding of free b-ricin was similar to that of ricin but decreased by 60% when complexed to Streptavidin. The obser vation that 40% of the cellular binding capacity of b-ricin was 100 -i still retained following its complexing to Streptavidin could be in agreement with the cytotoxicity studies, if ricin concentra 80- tion in the complexes is considered. As is shown in Figs. 1 and 2, 40-50% cytotoxic activity of streptavidin-b-ricin was ob served at a ricin concentration of 5 nM (binding at lower ricin 60- concentrations was restricted due to limited specific activity). Incubation in the presence of o-galactose diminished the cellu 40- lar binding of ricin, b-ricin, and the streptavidin-b-ricin com icL plex to similar low levels, indicating that nonspecific binding *o via Streptavidin, damaged radioiodinated protein, or aggrega 20- tion did not take place. Immunotargeting of Streptavidin-b-Ricin to KB Cells. We JD -C have previously demonstrated that biotinylation of mAb 108 c did not cause any significant damage to the antibody-combining 0.01 0.1 1 10 sites and most of the antibody-binding activity to KB cells was Ricin (nM) retained (22). The next experiment was performed to assess the Fig. 2. Blocking of b-ricin cytoioxic activity by Streptavidin or avidin. Inhibi capacity of b-mAb 108 coated KB cells to bind Streptavidin. As tion of [3H]leucine incorporation into KB cells following 24 h incubation with is shown in Fig. 5, preincubation of KB cell monolayers with ricin (O), b-ricin-low (•)orb-ricin-high (A) alone ( ). or following preincuba- b-mAb 108 followed by a short incubation with I25l-streptavi- tion wilh Streptavidin ( ) or avidin (with b-ricin-high, D) at a molar ratio of 1:20. din in a biotin-free environment resulted in specific binding of 4450

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research. IMMUNOTARGETING OF BIOTINYL-RICIN VIA STREPTAVIDIN

60- From observations on different b-ricin preparations and sev "5JD eral in vitro targeting experiments it appears that the rate of £. 50- biotin substitution and the formation of optimally structured, nonaggregated, streptavidin-b-ricin complexes are important •¿o 40 -i factors in the later regeneration of cytotoxic activity. The pos sible formation of aggregated complexes that might be cytotox- 30- ically inactive or unretrieved, could account for the incomplete regeneration of cytotoxic activity via avidin or biotinyl-anti- > 20- body, as compared to free ricin. In addition, the balance among b-ricin, streptavidin, and antibody as well as the potency of the o. •¿ 10- cell-antibody association seem to play important roles in the retrieval of cytotoxic activity in vitro. During the last decade, the streptavidin- or avidin-biotin sys ^ 0 5 10 15 tems have become extremely useful intermediates in a variety of 125l-streptavldln (nM) different applications (24). The two biotin-binding proteins are Fig. 5. Binding of I25I-streptavidin to KB cell monolayers (2 x IO5 cells) tetrameric, carrying one biotin-binding site per subunit, both preincubated with b-mAb 108. KB monolayers preincubated with b-mAb 108 in exhibiting a remarkable thermodynamic affinity to the biotin the absence (•)orpresence (•)ofbiotin, or with b-mAb anti-CEA (O), mAb 108 ligand(10'5 M~'). In principle, any biotinylated molecule can be (A) (10 Mg/rnl antibody) or medium alone (x) were washed twice with PBS and 125I-streptavidin was added in a biotin-free medium (DMEM + dialyzed PCS) for recognized by avidin or streptavidin, either as such, or when 40 min at 37'C. these two proteins are coupled to any one of a number of probes (e.g., fluorescent or radioactive labels, , or in immobi the I25l-streptavidin to the cells. Addition of biotin to the me lized matrices). Alternatively, avidin or streptavidin can be used dium abolished the binding of 125I-streptavidin to b-mAb 108- to bridge two types of biotinylated molecules. However, avidin coated cells. Replacing b-mAb 108 by unbiotinylated mAb 108 and streptavidin are entirely different in their amino acid com or by biotinylated antibody of a different specificity did not position, primary sequence (25), charge, and glycosylation. Avi result in any significant binding. din is a cationic glycoprotein which displays strong adherent The cytotoxic activity of ricin in streptavidin-b-ricin com properties to cells in vitro and rapid blood and tissue clearance plexes targeted to KB cells via b-mAb 108 is demonstrated in in vivo, whereas streptavidin is a nonglycosylated neutral pro Fig. 6. KB cells were preincubated with b-mAb 108 and follow tein displaying minimal adherence properties to cells and its in ing removal of excess antibody the streptavidin-b-ricin complex vivo clearance is slower (26). In the in vitro system described in was added. As illustrated, the streptavidin-b-ricin complex was this study, avidin could not replace streptavidin since it did not exhibit a meaningful blocking effect on b-ricin toxicity. This is not cytotoxic to untreated KB cells. However, when reacted with KB cells previously treated with b-mAb 108 the streptavi attributed to the natural adherence of avidin to the surface of din-b-ricin exerted a significant level of cytotoxic activity, al cells in general (26) includind the KB cell. Accordingly, the activity of b-ricin in the avidin-b-ricin complex is blocked in the though the activity of the complex was lower than that of ricin itself (IC50 of 0.33 and 0.19 HM,respectively). Replacing b-mAb same way as in the streptavidin-b-ricin complex (as evidenced 108 by b-mAb of a different specificity (anti-CEA), or by unbi by the method used to determine biotinyl substitution in the b-ricin preparations, which is dependent upon interaction with otinylated mAb 108, did not lead to subsequent cytotoxic effect avidin; see "Materials and Methods"). However, in avidin-b- upon addition of streptavidin-b-ricin. ricin complexes, the cell-binding function of ricin B chain has

DISCUSSION 100-] The present work describes a new approach for immunotar- geting whole ricin using the streptavidin-biotin system. Accord ing to this approach, the cytotoxic activity of whole ricin, bi « 80- otinylated via a disulfide bond, is effectively inhibited when interacted with streptavidin. However, ricin in the streptavidin- 60- b-ricin complex specifically regained its cytotoxic action when I unoccupied biotin-binding sites of streptavidin in the complex interacted with biotin groups on the antibody. Streptavidin is S. 40- thus used to block b-ricin and to bridge it to biotinyl-antibody recognizing the target cell. Experiments were performed to es tablish the conditions for maximal ricin biotinylation yielding 20- effective blocking of the toxin, yet ensuring later retrieval of cytotoxic activity. The efficiency of the streptavidin blocking activity was correlated to the degree of ricin biotinylation and, 0.001 0 01 0 .1 1 as evidenced from experiments with several b-ricin prepara tions, an optimal range of 3-5 biotin groups attached per ricin R I c l n (nM) were required for inhibition greater than 99% and later retrieval Fig. 6. Streptavidin-b-ricin cytotoxicity to KB cells mediated via b-mAb 108. Inhibition of pHjleucine incorporation into KB cells following 24 h incubation of cytotoxic activity. This may suggest that effective inhibition with ricin (•),b-ricin (O), and streptavidin-b-ricin complex (molar ratio. 5:1) is achieved when more than 3 biotin groups/ricin are allowed to added to cells that were preincubated for 30 min with medium (x), mAb 108 (•), b-mAb-anti-CEA (D), or b-mAb 108 (A) (20 Mg/ml antibody). Following prein- interact with biotin binding sites of the same and/or different cubation, the antibody-containing medium or medium alone was removed and the streptavidin molecules. various ricin derivatives were added in fresh medium. 4451

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research. IMMUNOTARGETING OF BIOTINYL-RICIN VIA STREPTAVIDIN now been replaced by that of avidin. This observation suggests 2. Endo, Y., Mitsui, K., Motizuki, M., and Tsurugi, K. The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. J. Biol. that avidin facilitates the transfer of ricin into the cytosol and Chem., 262: 5908-5913, 1987. the eventual release of ricin A chain. It also demonstrates the 3. Villafranca, J. E., and Robertus J. D. Ricin B-chain is a product of gene retrieval of ricin cytotoxic activity when incorporated into such duplication. J. Biol. Chem., 25«:554-556, 1981. a complex. Thus, b-ricin may be targeted to cells via any ligand, 4. Eiklid, K., Olsnes, S., and Phil, A. Entry of lethal doses of abrin, ricin and moddecin into the cytosol of HeLa cells. Exp. Cell Res., 126: 321-326,1980. be it a biotin-binding molecule with natural cell adherence 5. Thorpe, P. E., and Ross, W. C. J. The preparation and cytotoxic properties properties such as avidin, or a biotinylated antibody specific to of antibody-toxin conjugates. Inumi noi. Rev., 62: 119-158, 1982. 6. Filipovich, A. H., Vallera, D. A., Youle, R. J., et al. Ex-vivo treatment of the target cell that requires the participation of an intermediary donor bone marrow with anti-T cell immunotoxins for the prevention of party, i.e., streptavidin. It is not clear whether antibody inter- graft-vwius-host disease. Lancet, /: 469-472, 1984. nalization carrying with it the streptavidin-b-ricin complex is 7. Thorpe, P. E., Mason D. W., Brown, A. N. F., et al. Selective killing of malignant cells in leukemic rat bone marrow using an antibody-ricin conju required or whether the antibody or avidin serves merely to gate. Nature (Lond.), 297: 594-596, 1982. bring the b-ricin in close proximity to the plasma membrane of 8. Blakey, D. C., and Thorpe, P. E. An overview of therapy with immunotoxins the cell. containing ricin or its A chain. Antibody Immunoconjugates Radiopharma- ceuticals, /: 1-16, 1988. The rationale behind the present work is based on a previous 9. Vallera, D. A., Quiñones,R. R., Azemove, S. M., and Soderling, C. C. B. observation described by Bayer et al. (27), showing that affinity Monoclonal antibody-toxin conjugates reactive against human T lympho cleavage and targeted catalysis of proteins can be achieved using cytes: a comparison of antibody linked to intact ricin with antibody linked to ricin A chain. Transplantation (Baltimore), 37: 387-392, 1984. the streptavidin- or avidin-biotin complex. In that system, 10. Pirker, R. P., FitzGerald, D. J. P., Willingham, M. C., and Pastan. I. En streptavidin or avidin inhibits the action of biotinylated pro hancement of the activity of immunotoxins made with either ricin A chain or Pseudomonas exotoxin in human ovarian and epidermoid carcinoma cell teases on native proteins, but once these proteins are biotiny lines. Cancer Res., 48: 3919-3923, 1988. lated, a selective enzymatic activity towards these proteins is 11. Griffin, T. W., Richardson, C., Houston L. L., LePage, D., Bogden, A., and being conferred. Also, an unbiotinylated target protein could be Raso, V. Antitumor activity of intraperitoneal immunotoxins in a nude specifically cleaved via interaction with biotinyl-(anti-target)- mouse model of human malignant mesothelioma. Cancer Res., 47: 4266- 4270, 1987. -antibody which was avidin bridged to the biotinyl . The 12. Thiesen, H. J., Juhl, H., and Arndt, R. Selective killing of human bladder cancer cells by combined treatment with A and B chain ricin antibody con selective catalysis described in that system do not require the jugates. Cancer Res., 47:419-423, 1987. release of the effector molecule, i.e., the enzyme, and is 13. Vitteta, E. S., Fulton, R. J., and Uhr, J. W. Cytotoxicity of cell-reactive achieved, perhaps, through a proximity effect. immunotoxin containing ricin A chain is potentiated by an anti-immunotoxin containinig ricin-B chain. J. Exp. Med., 160: 341-346. 1984. The results of the present study describe an in vitro model 14. Mclntosh, D. P., Edwards, D. C., Cumber, A. J., Parnell, G. D., Dean, C. J., system, which is designed to be implemented in vivo. Com Ross, W. C. J., and Forrester, J. A. Ricin B chain converts a non-cytotoxic plexes between streptavidin-b-ricin and biotinyl-antibody may antibody-ricin A chain conjugate into a potent and specific cytotoxic agent. be administered directly, or else they can be used in a two-step FEBS Lett., 164: 17-20, 1983. 15. Eccles, S. A., Mclntosh, D. P., Purvies, H. P., Cumber, A. J., Parnell, G. D., system in which the biotinyl-antibody is administered first, al Forrester, A., Styles, J. M., and Dean, C. J. An ineffective monoclonal lowing accumulation at the tumor site and clearance of most antibody-ricin A chain conjugate is converted to a tumoricidal ¿gentin vivoby subsequent systemic administration of ricin B chain. Cancer Immunol. Im irrelevant antibody from the circulation and tissues, followed by munother., 24:37-41, 1987. the administration of the streptavidin-b-ricin complex. Our pre 16. Thorpe, P. E., Ross, W. C. J., Brown. A. N. F., Myers, C. D., Cumber, A. J., liminary results show that this complex is 20-30-fold less toxic Foxwell, B. M. J., and Forrester, J. T. Blockade of the galactose binding sites of ricin by its linkage to antibody. Specific Cytotoxicity effects of the conju in vivo than free ricin.4 Also, we have previously shown in a gates. Eur. J. Biochem., 140: 63-71, 1984. similar system that a chemotherapeutic drug (cisplatin) linked 17. Pietersz, G. A., Kanellos, J., and McKezie, F. C. Novel synthesis and in vitro characterization of disulfide-linked ricin-monoclonal antibody conjugates de to avidin via a dextran bridge could be immunotargeted to KB void of galactose binding activity. Cancer Res., 48: 4469-4476, 1988. tumor xenografts in nude mice using b-mAb 108 (22). Tumor 18. Bursa, P., Fietribiasi. F., Bussolati, G., Dosio, F., Arione, R., Comoglio, P. development was specifically inhibited when the drug-dextran- M., Prat, M., and Cattel, L. Blocked and not blocked whole-ricin-antibody immunotoxin: intraperitoneal therapy of human tumor xenografted in nude avidin was introduced in vivo 24 h following the administration mice. Cancer Immunol. Immunother., 29: 185-192, 1989. of b-mAb 108. Such a procedure may offer several advantages 19. Lambert, J. M., Mclntyre, G., Gauthier, M. N., Zullo, D., Rao, V., Sleeves, R. M., Goldmacher, V. S., and Blattler, W. A. The galactose-binding sites of when used in vivo: (a) biotinylation of ricin or antibody is a the cytotoxic lectin ricin can be chemically blocked in high yield with reactive simple procedure resulting in high yields and retention of the ligands prepared by chemical modification of glycopeptides containing tri biological activity of either conjugate; (/>)the immunotargeting antennary AMinked oligo-saccharides. , 30: 3234-3247, 1991. 20. Greenwood, F. C., Hunter, W. M., and Glover, J. S. The preparation of system might be simplified since one pool of biotinyl-ricin may 125I-labeled human growth hormón of high specific radioactivity. Biochem. be used in combination with different biotinyl-antibody prepa J., «9:114-123, 1963. 21. Schlessinger, J. Allosteric regulation of the epidermal growth factor receptor rations; (c) the system can be manipulated by increasing anti kinase. J. Cell Biol., 103: 2067-2072, 1986. body but not necessarily ricin doses, i.e., leaving the streptavi 22. Schechter, B., Arnon, R., Wilchek, M., Schlessinger, J., Hurwitz, E., Aboud- din-b-ricin complex at subtoxic levels. Application of this Pirak, E., and Sela, M. Indirect immunotargeting of cis-Pl to human epider moid carcinoma KB using the avidin-biotin system. Int. J. Cancer, 48: 167- approach to tumor systems where the antibodies are not readily 172, 1991. internalized may create a biotin-rich environment at the tumor 23. Green, N. M. A spectrophotometric assay for avidin and biotin based on site to which safe doses of streptavidin-b-ricin may then be binding of dyes by avidin. Biochem. J., 94: 23c-24c, 1965. 24. Wilchek, M., and Bayer, E. A. The avidin-biotin complex in bioanalytical targeted. applications (review). Anal. Biochem.. 171: 1-32, 1988. 25. Wilchek, M., and Bayer E. A. Avidin-biotin technology ten years on: has it lived up to its expectations? Trends Biol. Sci., 14: 408-412, 1989. REFERENCES 26. 26. Schechter, B., Silberman, R., Arnon, R.. and Wilchek, M. Tissue distri 1. Olsnes, S., Refens, K., and Pihl, A. Mechanism of action of the toxic lectins bution of avidin and streptavidin injected to mice. Eur. J. Biochem., 189: abrin and ricin. Nature (Lond.), 249: 627-631, 1974. 327-331, 1990. 27. Bayer, A. B., Grootjans, J. J., Alón,R., and Wilchek. M. Affinity cleavage and targeted catalysis of proteins using the avidin-biotin system. Biochemis 4 Unpublished data. try, 29: 11274-11279, 1990.

4452

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research. Cytotoxicity of Streptavidin-blocked Biotinyl-Ricin Is Retrieved by in Vitro Immunotargeting via Biotinyl Monoclonal Antibody

Bilha Schechter, and Meir Wilchek

Cancer Res 1992;52:4448-4452.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/52/16/4448

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/52/16/4448. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1992 American Association for Cancer Research.