Cancer Immunol Immunother (1989) 29: 185-192 ancer mmunolggy mmunotherapy © Springer-Verlag 1989 Blocked and not blocked whole-ricin-antibody immunotoxins: Intraperitoneal therapy of human tumour xenografted in nude mice* P. Brusa 1, F. Pietribiasi 2, G. Bussolati 2, F. Dosio 1, R. Arione 3, P. M. Comoglio 2, M. Prat 2, and L. Cattel 1 Applied Medicinal Chemistry Institute, University of Turin, Italy 2 Department of Biomedical Sciences and Oncology, University of Turin, Italy • 3 Microbiology Institue, University of Turin, Italy Summary. A blocked immunotoxin, consisting of ricin and AR-3 monoclonal antibody joined by a short thioether bond, was previously synthesized. This conjugate had lost the ability to bind the galactosidic residues of Sepharose 6B, probably because of the steric restraint of the antibody molecule on the ricin B chain. In in vitro assays immuno- toxin was active only on cells expressing the correspond- ing AR-3 epitope. The in vivo activity of our blocked im- munotoxin was assessed by injecting it directly into the peritoneal cavity of tumour-bearing nude mice. The ani- mals were i.p. grafted with the HT-29 cell line, which was derived from a human colorectal adenocarcinoma express- ing the antigen CAR-3, against which the AR-3 monoclo- nal antibody is directed. The best protocol tested, to arrive at the optimal regimen for the i.p. blocked immunotoxin therapy, required the administration of the immunotoxin (2 lxg) on days 4 and 6 after the graft. The mice were killed on different subsequent days to determine the therapeutic effects. Histological sections of the different organs were prepared and stained with haematoxylin/eosin and were also examined by an immunocytochemical method with AR-3 monoclonal antibody to confirm the presence of the relating antigen on the tumour cell surface. The blocked immunotoxin substantially suppressed tumour growth of the grafted HT-29 cells, without showing any undesirable ricin toxicity. Most importantly, established transplanted HT-29 tumour cells treated with blocked immunotoxin al- most completely regressed, while under the same condi- tions the not blocked immunotoxin, an irrelevant immu- notoxin, ricin, and the AR-3 alone failed to inhibit tumour growth. Introduction Immunotoxins are conjugates of highly active protein tox- ins and monoclonal antibodies [2, 12, 36, 37]. The aim of using immunotoxins in cancer therapy is to concentrate the toxin within tumour tissue causing the specific killing of the neoplastic cells [14]. * This work was supported by grants from the Ministry of Public Education, the Regione Piemonte and the Associazione Italiana per la Ricerca sul Cancro Offprint requests to: L. Cattel, Istituto di Chimica Farmaceutica Applicata, corso Raffaello 31, I-10125, Torino, Italy Ricin is a toxic plant glycoprotein composed of two subunits (A and B) linked by a disulphide bond [26]. The A chain catalytically inhibits protein synthesis by inactivat- ing 60S ribosomal subunits, so that very few A chain mole- cules are sufficient to kill a cell [7, 8, 25]. The B chain binds the toxin to a cell-surface galactosidic receptor and facilitates A chain translocation [34, 42, 43]. Linkage of a tumour-cell-specific monoclonal antibody to the A chain alone or to the whole ricin molecule gener- ates a hybrid molecule which, in theory, combines the cancer specificity of the antibody with the extreme cyto- toxicity of the toxin. However, A-chain conjugates are not as toxic as whole-ricin conjugates and may even be inac- tive [24, 34]. Nevertheless intact ricin immunotoxins have non-specific effects, since the B-chain galactose-binding site is active [2]. This non-specific binding can be limited in vitro with high concentrations of lactose or galactose [26, 41]; such intact ricin-antibody conjugates have been successfully used in vitro to treat bone marrow cells, prior to transplantation, for the removal of T cells [9] or leukemic cells [22]. Recently a disulphide-linked whole-ricin-antibody conjugate was prepared, which failed to bind to lactose- Sepharose, implying the blocking of the galactose-binding site on the B chain [27]. This conjugate showed cytotoxicity in vitro equal to that of native ricin, while in vivo it was only moderately effective against tumours (growing s.c. in mice), since it caused some liver and kidney damage. However, injection of the immunotoxins directly into the tumours resulted in regression of the tumours [28]. Another approach, to reduce the non-specific binding, was to pre- pare intact ricin immunotoxins, thioether-linked; the B-chain galactose-binding site appeared to be sterically hindered by the antibody [33]. This immunotoxin, al- though highly specific to target cells in vitro [2, 5], has not so far been tested in vivo. Following the same strategy we prepared [4] a new thioether-linked ricin-antibody immunotoxin in which the galactose-binding site of the ricin was sterically hindered. The immunotoxin was synthesized by linking the AR-3 monoclonal antibody, previously derivatized with N-hy- droxysuccinimidyl-S-acetyl thioacetate (SATA), to ricin iodoacetylated with the N-hydroxysuccinimidylester of iodoacetic acid (SIA). After the last step of purification on Sepharose 6B, we obtained two populations of immuno- toxins; one, referred to as blocked, which had lost the abil- ity to bind the galactosidic residues of the Sepharose 6B,