Endocytosis and intracellular localisation of type 1 ribosome-inactivating protein saporin-s6.

Saporin-S6 is a single-chain ribosome-inactivating protein (RIP) that has low toxicity in cells and animals. When the protein is bound to a carrier that facilitates cellular uptake, the protein becomes highly and selectively toxic to the cellular target of the carrier. Thus, saporin-S6 is one of the most widely used RIPs in the preparation of immunoconjugates for anti-cancer therapy. The endocytosis of saporin-S6 by the neoplastic HeLa cells and the subsequent intracellular trafficking were investigated by confocal microscopy that utilises indirect immunofluorescence analysis and transmission electron microscopy that utilises a direct assay with gold-conjugated saporin-S6 and an indirect immunoelectron microscopy assay. Our results indicate that saporin-S6 was taken up by cells mainly through receptor-independent endocytosis. Confocal microscopy analysis showed around 30% co-localisation of saporin-S6 with the endosomal compartment and less than 10% co-localisation with the Golgi apparatus. The pathway identified by the immunofluorescence assay and transmission electron microscopy displayed a progressive accumulation of saporin-S6 in perinuclear vesicular structures. The main findings of this work are the following: i) the nuclear localisation of saporin-S6 and ii) the presence of DNA gaps resulting from abasic sites in HeLa nuclei after intoxication with saporin-S6.

Increased serum level of xanthine oxidoreductase in liver transplanted patients.

Xanthine oxidoreductase (XOR) leakage into serum has been observed in various types of liver pathology as well as after liver transplantation (LT). We determined the amount of XOR associated with LT to investigate the changes in serum enzyme level during the LT procedure and the post-operative period. Additionally, we examined whether there was any correlation between XOR levels and the surgical technique. XOR levels were measured by a competitive ELISA. In a first group of patients, the portal vein was flushed before the liver and systemic reperfusions, which occurred simultaneously. In the second group, the graft was flushed with blood from the portal vein before the systemic reperfusion. XOR showed a marked elevation in the caval effluent collected during LT and was higher compared to control serum levels at all time points that were examined after LT. The XOR levels during LT were also higher than samples taken pre-LT or from the portal blood flush before reperfusion. The XOR level was higher in Group 2 than in Group 1. Enhancement of the XOR serum level during LT was not derived from enterocytes, and it should be attributed to enzyme leakage from graft liver cells. We report the elevation of serum XOR during the three weeks following LT for the first time, as well as the influence of the graft reperfusion technique on XOR serum levels.

Ribosome-inactivating proteins: progress and problems.

Ribosome-inactivating proteins (RIPs), mostly from plants, are enzymes which depurinate rRNA, thus inhibiting protein synthesis. They also depurinate other polynucleotide substrates. The biological activity of RIPs is not completely clarified, and sometimes independent of the inhibition of protein synthesis. There are differences in the cytotoxicity of RIPs and, consequently, in their toxicity to animals. Some RIPs are potent toxins, the best known being ricin, a potential biological weapon. New toxins have recently been identified. RIPs cause apoptotic and necrotic lesions, and induce production of cytokines causing inflammation. RIPs are potentially useful in agriculture and medicine because (i) they have antiviral activity and (ii) they are used for the preparation of conjugates with antibodies ('immunotoxins') or other carriers, rendering them specifically toxic to the cell target of the carrier, which may be helpful in therapy. The distribution, mechanism of action and role in nature of RIPs are not completely understood, and we can expect several future developments in their practical application.

Toxicity of xanthine oxidoreductase to malignant B lymphocytes.

Reactive oxygen species (ROS) generated by xanthine oxidoreductase (XOR) were toxic to B lymphoma-derived Raji cells (positive for 8A monoclonal antibody, mAb). The sensitivity of these malignant cells to the hypoxanthine/XOR system was higher than that observed in peripheral human lymphocytes. The understanding of the mechanisms of cytotoxicity induced by XOR-produced ROS is essential in view of a possible clinical application. Cell death mostly had the feature of apoptosis and post-apoptotic necrosis and depended on the activity of XOR. Catalase, but not superoxide dismutase, protected cells from the toxicity of XOR, thus indicating that cell damage depended on the production of hydrogen peroxide. The toxicity of ROS was selectively targeted to malignant Raji cells by antibody-XOR conjugation, either directly, with an 8A-XOR conjugate, or indirectly, with an 8A mAb plus an anti-mouse IgG-XOR. Both direct and indirect immunotoxins induced apoptotic death to target cells in a dose-dependent manner. These conjugates showed no aspecific cytotoxicity in conditions very similar to the ex vivo treatment of cell suspension for bone marrow transplantation. Moreover, the prevalence of apoptotic death over necrosis may reduce the in vivo inflammatory response and its local and systemic consequences, thus becoming relevant in the construction of immunotoxins with therapeutic potential.

CD38 as a target of IB4 mAb carrying saporin-S6: design of an immunotoxin for ex vivo depletion of hematological CD38+ neoplasia.

An anti-CD38 mAb (IB4) coupled to saporin-S6, a type 1 ribosome-inactivating protein (RIP), was designed for ex vivo or loco-regional therapeutical applications in myeloma and lymphoma. The ability of this immunotoxin to eliminate CD38+ cells was studied in vitro on selected CD38+ human cell lines (Raji, HBL6, L540 and CEM) and on CD38+ neoplastic cells from a Non Hodgkin Lymphoma (NHL) patient. HBL6, Raji and L540 cells resulted very sensitive to the IB4/saporin-S6 conjugate, concentrations as low as 100 pM of the immunotoxin completely inhibited protein synthesis. CD38+ neoplastic cells from the NHL patient were completely eliminated after treatment with immunotoxin at 10 nM concentration. CFU-c rescue by bone marrow precursors was maintained after exposure to the immunotoxin. These results indicate that IB4/saporin-S6 is endowed with strong and specific cytotoxic effects on selected CD38+ tumor cells lineages. Consequently, it is reasonable to propose a clinical use of the IB4/saporin-S6 for ex vivo purging of unwanted cells (e.g. depletion of contaminating neoplastic cells in aphereses obtained from G-CSF-treated patients) or for loco-regional therapies of CD38+ tumors.

Interaction of volkensin with HeLa cells: binding, uptake, intracellular localization, degradation and exocytosis.

Among two-chain ribosome-inactivating proteins (RIPs), volkensin is the most toxic to cells and animals, and is retrogradely axonally transported in the rat central nervous system, being an effective suicide transport agent. Here we studied the binding, endocytosis, intracellular routeing, degradation and exocytosis of this RIP. The interaction of volkensin with HeLa cells was compared to that of nigrin b, as an example of a type 2 RIP with low toxicity, and of ricin, as a reference toxin. Nigrin b and volkensin bound to cells with comparable affinity (approx. 10(-10) M) and had a similar number of binding sites (2 x 10(5)/cell), two-log lower than that reported for ricin. The cellular uptake of volkensin was lower than that reported for nigrin b and ricin. Confocal microscopy showed the rapid localization of volkensin in the Golgi stacks with a perinuclear localization similar to that of ricin, while nigrin b was distributed between cytoplasmic dots and the Golgi compartment. Consistently, brefeldin A, which disrupts the Golgi apparatus, protected cells from the inhibition of protein synthesis by volkensin or ricin, whereas it was ineffective in the case of nigrin b. Of the cell-released RIPs, 57% of volkensin and only 5% of ricin were active, whilst exocytosed nigrin b was totally inactive. Despite the low binding to, and uptake by, cells, the high cytotoxicity of volkensin may depend on (i) routeing to the Golgi apparatus, (ii) the low level of degradation, (iii) rapid recycling and (iv) the high percentage of active toxin remaining after exocytosis.

Ricin toxicity to microglial and monocytic cells.

Microglial cells, like macrophages, are very sensitive to ricin, a galactose-specific toxic lectin belonging to the family of ribosome-inactivating proteins. This toxin can be taken up by most cells through the binding of its B chain to galactose-containing molecules on the cell membrane. In macrophagic cell types it can be internalised also by mannose receptors which are present on the surface of these cells. Endocytosis of the toxin by either pathway was evaluated by ricin toxicity to primary cultures of rat microglial cells and to a microglial N11 cell line in the presence or absence of lactose and mannan, which compete for the endocytosis via the ricin lectin chain or cellular mannose receptors, respectively. Results were compared with those obtained in cultures of mouse macrophages, human monocytes, and a monocytic JM cell line. All cultures were protected from ricin toxicity more by lactose than by mannan, indicating that ricin endocytosis via its lectin B chain is prevalent over that mediated by cellular mannose receptors. However, a partial protection by mannan was observed in all cases but not-stimulated N11 cells, either in the form of direct protection or of significant additional protection over that afforded by lactose. Mannose receptor expression by N11 cells was negative before, and positive after, treatment with endotoxin, as assessed by the specific binding of 125I-mannose-bovine serum albumin. Moreover, a partial protection from ricin toxicity by mannan was induced in the N11 microglial line after stimulation, consistently with an inducible expression of the mannose receptor by activated cells switched towards a microglial phenotype.

Serum xanthine oxidase in human liver disease.

OBJECTIVES: High concentrations of serum xanthine oxidase (XO) have been reported during human liver disease and hepatocyte injury in experimental settings. However, it is unclear whether this elevation reflects hepatocyte necrosis or has a different meaning. METHODS: The serum level of XO in 64 patients with chronic liver disease (17 patients with cirrhosis, 30 with chronic hepatitis, and 17 with cholestatic disorders) and in 12 control subjects was determined by a competitive ELISA. Conventional serum markers of liver damage were assessed in all patients, and grading and staging were scored in the chronic hepatitis group according to Knodell. RESULTS: The XO serum levels were significantly higher in the patients than in the controls. The differences were also significant when controls were compared to patients with chronic hepatitis and cholestatic disorders separately, but not when compared to the cirrhosis group. Patients with cholestatic disorders had XO values higher than those of patients with cirrhosis or chronic hepatitis. XO levels did not correlate with stage and grade in chronic hepatitis group. We found a weak but significant positive correlation in patients between XO serum level and gamma-glutamyl transpeptidase (r = 0.37). This correlation was stronger when chronic hepatitis (r = 0.42) and, especially cholestatic disorders (r = 0.71), were separately tested, but was absent in the cirrhosis group. The XO values positively correlated with alkaline phosphatase in patients with cholestatic disorders. A level of serum XO >32 microg/ml specifically identified cholestatic disorders in our study population. CONCLUSIONS: A marked elevation of serum XO in patients with chronic liver disease seems to reflect the presence of cholestasis. No correlation between XO levels and histological or serum evidence of hepatocyte necrosis was found in these patients.

Oxidative stress to human lymphocytes by xanthine oxidoreductase activity.

The in vitro toxicity of the reactive oxygen species generating enzyme xanthine oxidoreductase (XOR) to human peripheral blood lymphocytes was studied after stimulation with phytohaemoagglutinin or anti-CD3/CD28 antibodies. Apoptosis and necrosis were induced by the XOR/hypoxanthine system in a time- and concentration-dependent manner. CD8+ lymphocytes showed a higher sensitivity than CD4+ cells to the XOR/hypoxanthine system. The occurrence of apoptosis was demonstrated by annexin-V binding to injured cell membrane, which was the most precocious alteration observed, followed by the increment of transglutaminase activity, which was significant at the lowest XOR concentration used. Nuclear damage was assessed by the increased hypodiploid nuclei and by DNA migration on gel electrophoresis, which turned to an apoptotic pattern before the occurrence of cell membrane necrotic lesions. Apoptosis was induced by XOR activity proportionally to substrate concentration and was prevented by the competitive enzyme inhibitor, allopurinol. The hydrogen peroxide scavenging enzyme, catalase, gave a higher protection than superoxide dismutase from the toxicity caused by the XOR/hypoxanthine system. Necrosis occurs in a variable percentage indicating that reactive oxygen species may trigger both apoptosis and necrosis in proliferating human lymphocytes, mostly depending on XOR concentration.

Determination of xanthine oxidase in human serum by a competitive enzyme-linked immunosorbent assay (ELISA).

Xanthine oxidase was purified from human milk and used to immunise rabbits. A competitive immunoenzymatic assay with purified enzyme and rabbit antiserum was optimised to measure xanthine oxidase in human serum, the lowest detectable amount being 0.03 pmol of enzymatic protein. Thus, the test (i) is sensitive enough to determine xanthine oxidase in human serum, being more sensitive than the spectrophotometric method, (ii) it is more convenient for clinical laboratories than other sensitive tests and (iii) it has the advantage over the enzyme activity-based assays of also detecting inactive enzyme molecules. A competitive enzyme-linked immunosorbent assay (ELISA) was used to measure the serum xanthine oxidase level in healthy donors and in patients with liver diseases, and it was found that any concentration below 1 mg/L is in the normal range.

Different sensitivity of CD30+ cell lines to Ber-H2/saporin-S6 immunotoxin.

The in vitro sensitivity of cells to a Ber-H2(anti-CD30)/saporin-S6 immunotoxin has been investigated. The CD30+ cell lines, K562, L428 and L540, were used to study cell binding, uptake and degradation of the immunotoxin. K562 cells were less sensitive than L428 and L540 cells to the immunotoxin by approximately one order of magnitude. The difference in cytotoxicity correlated with the intracellular accumulation and with the ratio of degraded over total internalized Ber-H2/saporin-S6, regardless of the immunotoxin binding to the cells. After 6 h incubation, the less sensitive K562 cells (i) accumulated only one third and one tenth of the immunotoxin accumulated by the more sensitive L428 and L540 cells, respectively, and (ii) degraded two thirds of the internalized protein versus one third degraded by either L428 or L540 cells. Ammonium chloride and chloroquine reduced the cytotoxicity of the immunotoxin towards K562 but not to L540 cells. This effect correlated with the increment of immunotoxin catabolism by K562 cells in the presence of chloroquine. In conclusion, uptake alone of an immunotoxin by target cells is not sufficient to assure its efficacy which might also depend on intracellular routing. Only a cytotoxicity test may be really predictive.

Simulated ischaemia-reperfusion conditions increase xanthine dehydrogenase and oxidase activities in rat brain slices.

Xanthine dehydrogenase and oxidase activities increased by 87% in rat brain slices after 30 min in vitro ischaemia. A further 41% increase was induced by 30 min simulated reperfusion of ischaemic slices. No conversion from the dehydrogenase to the oxidase activity was observed. The increment of enzyme activity was not due to neosynthesis of the enzyme, since it was not affected by the addition of cycloheximide during the ischaemic incubation. The increased oxygen-dependent form of the enzyme could aggravate the ischaemic brain injury by free radicals production, in particular after reperfusion.

Toxicity of ricin and volkensin, two ribosome-inactivating proteins, to microglia, astrocyte, and neuron cultures.

Ricin and volkensin, two potent toxins belonging to the family of ribosome-inactivating proteins (RIPs), have been largely exploited in recent years in in vivo experiments of neuronal degeneration consequent to suicide transport or immunolesioning. We have determined both the toxicity of, and the inhibition of, protein synthesis by ricin and volkensin in in vitro cultures enriched in microglial cells, astrocytes, or neurons. In microglial cultures, 50% of toxicity (estimated by LDH released from dead cells) after 24 h exposure to RIPs was obtained with volkensin at 2.2x10(-12) M concentration and 50% of protein synthesis inhibition at 2x10(-14) M concentration. Both values were higher by about one order of magnitude in astrocyte-enriched cultures. Toxicity of, and inhibition of, protein synthesis by, ricin were lower for both cell types by about 1 order of magnitude as compared to volkensin. Cerebellar granule neurons in culture survived remarkably well to 24 h exposure to ricin or volkensin, although their protein synthesis was effectively inhibited by the two toxins with a potency similar to that found for astrocytes. These results demonstrate that glial cells, in particular microglia, are very sensitive to RIPs toxicity and should, therefore, be a primary target of these toxins when injected in vivo. Thus, the damage observed after in vivo experiments could be partly related to diffusion of toxic substances from early-affected glial cells.

Ribosome-inactivating lectins with polynucleotide:adenosine glycosidase activity.

Lectins from Aegopodium podagraria (APA), Bryonia dioica (BDA), Galanthus nivalis (GNA), Iris hybrid (IRA) and Sambucus nigra (SNAI), and a new lectin-related protein from Sambucus nigra (SNLRP) were studied to ascertain whether they had the properties of ribosome-inactivating proteins (RIP). IRA and SNLRP inhibited protein synthesis by a cell-free system and, at much higher concentrations, by cells and had polynucleotide:adenosine glycosidase activity, thus behaving like non-toxic type 2 (two chain) RIP. APA and SNAI had much less activity, and BDA and GNA did not inhibit protein synthesis.

Toxicity and cytotoxicity of nigrin b, a two-chain ribosome-inactivating protein from Sambucus nigra: comparison with ricin.

Nigrin b, a lectin isolated from the bark of elderberry (Sambucus nigra L.), has structure and enzymatic activity similar to that of ricin and other type 2 ribosome inactivating proteins (RIPs), and yet is much less toxic to cells and animals. In an attempt to explain this difference, we studied (1) the cytotoxicity of both lectins at 18 and 37 degrees C, and in the presence of substances interfering with intracellular routing, and (2) the binding of nigrin b to, and its uptake and degradation by HeLa cells, in parallel with ricin. As compared with the latter, (1) less nigrin b was bound and more was degraded by cells, with a resulting lower concentration remaining inside the cells, and (2) there is evidence for a different intracellular routing followed by the two lectins. These results may explain at least partly the different cytotoxicity and consequently the lower toxicity to mice of nigrin b compared with ricin.

New ribosome-inactivating proteins with polynucleotide:adenosine glycosidase and antiviral activities from Basella rubra L. and bougainvillea spectabilis Willd.

New single-chain (type 1) ribosome-inactivating proteins (RIPs) were isolated from the seeds of Basella rubra L. (two proteins) and from the leaves of Bougainvillea spectabilis Willd. (one protein). These RIPs inhibit protein synthesis both in a cell-free system, with an IC50 (concentration causing 50% inhibition) in the 10(-10) M range, and by various cell lines, with IC50S in the 10(-8)-10(-6) M range. All three RIPs released adenine not only from rat liver ribosomes but also from Escherichia coli rRNA, polyadenylic acid, herring sperm DNA, and artichoke mottled crinkle virus (AMCV) genomic RNA, thus being polynucleotide:adenosine glycosidases. The proteins from Basella rubra had toxicity to mice similar to that of most type 1 RIPs (Barbieri et al., 1993, Biochim Biophys Acta 1154: 237-282) with an LD50 (concentration that is 50% lethal) < or = 8 mg.kg-1 body weight, whilst the RIP from Bougainvillea spectabilis had an LD50 > 32 mg.kg-1. The N-terminal sequence of the two RIPs from Basella rubra had 80-93% identity, whereas it differed from the sequence of the RIP from Bougainvillea spectabilis. When tested with antibodies against various RIPs, the RIPs from Basella gave some cross-reactivity with sera against dianthin 32, and weak cross-reactivity with momordin I and momorcochin-S, whilst the RIP from Bougainvillea did not cross-react with any antiserum tested. An RIP from Basella rubra and one from Bougainvillea spectabilis were tested for antiviral activity, and both inhibited infection of Nicotiana benthamiana by AMCV.

Hepatoxicity of ricin, saporin or a saporin immunotoxin: xanthine oxidase activity in rat liver and blood serum.

Male Wistar rats each received an i.p injection of the ribosome-inactivating proteins ricin or saporin, or a Ber-H2 (anti-CD30)-saporin immunotoxin at a dose corresponding to three times the LD50 calculated for mice. Animals were killed 24, 48 or 72 h after treatment. Histological examination showed hepatic necrosis in all treated animals, although the sinusoidal lining was affected only in ricin-poisoned rats. The activities of xanthine dehydrogenase (D-form) and oxidase (O-form) were determined spectrophotometrically in liver and serum samples. In ricin-treated animals the liver enzyme was progressively converted from the D- to the O-form, which accounted for more than 60% of total activity after 48 h of poisoning, whilst no change in the xanthine oxidase activity was found in the serum. In the liver of rats treated with free or Ber-H2-conjugated saporin, the D-form was more than 75%, as in normal animals. In the same animals the serum xanthine oxidase activity was up to three-fold control values. The determination of serum xanthine oxidase may prove helpful in the evaluation of liver damage in patients treated with immunotoxins. It may become a diagnostic tool for the differential diagnosis of liver diseases.

Toxicity of ribosome-inactivating proteins-containing immunotoxins to a human bladder carcinoma cell line.

Immunotoxins were prepared by linking the type 1 ribosome-inactivating proteins (RIP) momordin I, pokeweed antiviral protein from seeds (PAP-S) and saporin-S6 to the 48-127 monoclonal antibody (MAb) recognising a glycoprotein (gp54) expressed on all human bladder tumours tested and on human bladder carcinoma cell lines, in particular on the T24 cell line. T24 cells required a 2 hr contact with immunotoxins to ensure binding and endocytosis. A time course of exposure, followed by further incubation without the immunotoxins, showed that maximum inhibition of protein synthesis by T24 cells was reached after 2 hr of contact followed by 3 days without the immunotoxins. Under optimal conditions, 48-127/RIP immunotoxins at nanomolar concentrations inhibited by 50% protein synthesis of target T24 cells. No toxicity was observed if (i) target cells were treated with non-conjugated RIP, (ii) target cells were treated with momordin I- or PAP-S-containing immunotoxins made with an irrelevant antibody and (iii) a non-target cell line was treated with the same 2 RIP conjugated to 48-127 antibody. The in vitro selective toxicity of these immunotoxins encourages further studies in view of a possible use in clinical trials for the local therapy of human bladder carcinomas.

Excitotoxic increase of xanthine dehydrogenase and xanthine oxidase in the rat olfactory cortex.

Excitotoxic lesions induced by systemic injection of kainic acid, resulted in 2-3-fold increase of xanthine dehydrogenase and xanthine oxidase activities in the rat olfactory cortex 48-72 h after drug administration. A significant increase of the xanthine oxidase/dehydrogenase ratio was also observed at 4 and 48 h post-injection. No similar changes were noticed in the hippocampus. The enhancement of enzyme activity seems to be primarily a consequence of the altered cell composition in damaged area. Free radicals produced by the increased oxygen-dependent form of the enzyme could in turn aggravate the excitotoxic brain injury.

Ribosome-inactivating proteins (RNA N-glycosidases) from the seeds of Saponaria ocymoides and Vaccaria pyramidata.

From the seeds of the Caryophyllaceae Saponaria ocymoides and Vaccaria pyramidata two proteins were purified which have the properties of the type-1 (single-chain) ribosome-inactivating proteins [reviewed by Barbieri, L., Battelli, M. G. & Stirpe, F. (1993) Ribosome-inactivating proteins from plants, Biochim. Biophys. Acta 1154, 237-282]. The proteins have molecular masses of 30.2 kDa (S. ocymoides) and 28.0 kDa (V. pyramidata) and pI greater than 9.5, their N-terminal amino acid sequences are similar to those of saporin-S6 and dianthin 30, ribosome-inactivating proteins from other Caryophyllaceae, and they partially cross-react with sera against these proteins. Both proteins inhibit protein synthesis by a rabbit-reticulocyte lysate with IC50 (concentrations giving 50% inhibition) below 10(-10) M, have a smaller effect on poly(U)-directed phenylalanine polymerisation by rat liver ribosomes (nanomolar IC50, approximately) and on protein synthesis by various cell lines (IC50 ranging from 4 nM to > 3000 nM) and possess rRNA N-glycosidase activity, releasing 1 mol adenine/ribosome.