Routing of internalized ricin and ricin conjugates to the Golgi complex

The history of ricin, abrin and related toxins

Ricin, abrin and related plant toxins have played interesting and important roles in the history of clinical medicine and biomedical research. The use of these proteins in medical treatment since ancient times is reviewed. Later the proteins played important roles in the early days of immunological research and some of the fundamental principles of immunology were discovered with toxic proteins of this group. During the last three decades the mechanism of action of the toxins was elucidated. This led to a major effort to target the toxins to malignant cells. Ricin has been used in bioterrorism. Recently, the toxins have played important roles as experimental models to elucidate the intracellular trafficking of endocytosed proteins.

Cytotoxic ribosome-inactivating lectins from plants

A class of heterodimeric plant proteins consisting of a carbohydrate-binding B-chain and an enzymatic A-chain which act on ribosomes to inhibit protein synthesis are amongst the most toxic substances known. The best known example of such a toxic lectin is ricin, produced by the seeds of the castor oil plant, Ricinnus communis. For ricin to reach its substrate in the cytosol, it must be endocytosed, transported through the endomembrane system to reach the compartment from which it is translocated into the cytosol, and there avoid degradation making it possible for a few molecules to inactivate a large proportion of the ribosomes and hence kill the cell. Cell entry by ricin involves the following steps: (i) binding to cell-surface glycolipids and glycoproteins bearing beta-1,4-linked galactose residues through the lectin activity of the B-chain (RTB); (ii) uptake by endocytosis and entry into early endosomes; (iii) transfer by vesicular transport to the trans-Golgi network; (iv) retrograde vesicular transport through the Golgi complex and into the endoplasmic reticulum (ER); (v) reduction of the disulfide bond connecting the A- and B-chains; (vi) a partial unfolding of the A-chain (RTA) to enable it to translocate across the ER membrane via the Sec61p translocon using the pathway normally followed by misfolded ER proteins for targeting to the ER-associated degradation (ERAD) machinery; (vi) refolding in the cytosol into a protease-resistant, enzymatically active structure; (vii) interaction with the sarcin-ricin domain (SRD) of the large ribosome subunit RNA followed by cleavage of a single N-glycosidic bond in the RNA to generate a depurinated, inactive ribosome. In addition to the highly specific action on ribosomes, ricin and related ribosome-inactivating proteins (RIPs) have a less specific action in vitro on DNA and RNA substrates releasing multiple adenine, and in some instances, guanine residues. This polynucleotide:adenosine glycosidase activity has been implicated in the general antiviral, and specifically, the anti HIV-1 activity of several single-chain RIPs which are homologous to the A-chains of the heterodimeric lectins. However, in the absence of clear cause and effect evidence in vivo, such claims should be regarded with caution.

Antigen transport and cytoskeletal characteristics of a distinct enterocyte population in inflammatory bowel diseases

Intestinal antigen uptake is enhanced in inflammatory bowel disease. We analyzed transcellular transport routes of antigens in different compartments of normal enterocytes and atypical intestinal epithelial cells called "rapid antigen uptake into the cytosol enterocytes" (RACE cells). These cells constitute a recently described population of enterocyte-derived cells, which are increased in inflammatory bowel disease. Mucosa of freshly resected specimens were incubated with the antigens ovalbumin or horseradish peroxidase. Ultrastructural labeling patterns of differentiation-dependent proteins, the brush-border enzyme sucrase-isomaltase and the cytoskeleton proteins villin and actin, were determined in enterocytes. Apoptosis was investigated biochemically and ultrastructurally by cleavage of caspase-3. Both antigens were transported to late endosomes and to trans-Golgi vesicles of enterocytes in inflammatory bowel disease and control specimens. Quantitative evaluation revealed a significantly increased transepithelial antigen transport in both compartments of RACE relative to normal enterocytes. Labeling densities for sucrase-isomaltase, villin, and actin were decreased in RACE relative to normal enterocytes. Caspase-3 was not increased in RACE cells relative to controls. RACE cells are characterized by increased antigen transport to late endosomes and the trans-Golgi network, a disassembled cytoskeleton and lower concentrations of proteins that are markers of cell differentiation.

Endosomal ricin transport: involvement of Rab4- and Rab5-positive compartments

Transport of the ribosome-inactivating protein ricin through endosomes was studied in A431 cells expressing Rab5-, Rab4-, and Rab11-GFP. It was shown that Rab5- and Rab4-positive functional domains of early endosomes are involved in ricin transport. Ricin enters cells by both clathrin-dependent and clathrin-independent mechanisms. The main pool of internalized toxin accumulates in early endosomes and remains associated with them for a long time. In contrast to earlier observations, current observations indicate that the majority of ricin avoids transport to lysosomes. The low level of ricin association with Rab11 as well as with transferrin accumulated in the pericentriolar recycling compartment shows that the compartment is not responsible for keeping ricin away from degradation in lysosomes. Escape from degradation in lysosomes is assumed to result from the potentiality of ricin to form assemblies within compartments.

Membrane destabilization by ricin

Ricin is a promising candidate for the treatment of cancer because it can be selectively targeted to tumor cells via linkage to monoclonal antibodies. Biochemical evidence suggests that escape of ricin or its ribosome-inactivating subunit from an intracellular compartment is mediated by retrograde transport to the endoplasmic reticulum and subsequent direction into the ER-associated degradation pathway. Alternatively, lipase activity of ricin may facilitate leakage from endocytic vesicles. We have observed ricin-mediated release of macromolecular dyes from lipid vesicles that mimic the composition of endosomal membranes. Release of small molecules occurs to the same extent, suggesting an all-or-none mechanism due to bilayer destabilization. The level of accompanying membrane fusion depends on vesicle composition. Since it takes 24 h of incubation before the first traces of lysolipids are detectable by matrix-assisted laser desorption/ionization mass spectrometry, membrane destabilization is not due to the lipase activity of ricin.

Identification of the ricin lipase site and implication in cytotoxicity

Ricin is a heterodimeric plant toxin and the prototype of type II ribosome-inactivating proteins. Its B-chain is a lectin that enables cell binding. After endocytosis, the A-chain translocates through the membrane of intracellular compartments to reach the cytosol where its N-glycosidase activity inactivates ribosomes, thereby arresting protein synthesis. We here show that ricin possesses a functional lipase active site at the interface between the two subunits. It involves residues from both chains. Mutation to alanine of catalytic serine 221 on the A-chain abolished ricin lipase activity. Moreover, this mutation slowed down the A-chain translocation rate and inhibited toxicity by 35%. Lipase activity is therefore required for efficient ricin A-chain translocation and cytotoxicity. This conclusion was further supported by structural examination of type II ribosome-inactivating proteins that showed that this lipase site is present in toxic (ricin and abrin) but is altered in nontoxic (ebulin 1 and mistletoe lectin I) members of this family.

Ricin

Ricin is a heterodimeric protein produced in the seeds of the castor oil plant (Ricinus communis). It is exquisitely potent to mammalian cells, being able to fatally disrupt protein synthesis by attacking the Achilles heel of the ribosome. For this enzyme to reach its substrate, it must not only negotiate the endomembrane system but it must also cross an internal membrane and avoid complete degradation without compromising its activity in any way. Cell entry by ricin involves a series of steps: (i) binding, via the ricin B chain (RTB), to a range of cell surface glycolipids or glycoproteins having beta-1,4-linked galactose residues; (ii) uptake into the cell by endocytosis; (iii) entry of the toxin into early endosomes; (iv) transfer, by vesicular transport, of ricin from early endosomes to the trans-Golgi network; (v) retrograde vesicular transport through the Golgi complex to reach the endoplasmic reticulum; (vi) reduction of the disulphide bond connecting the ricin A chain (RTA) and the RTB; (vii) partial unfolding of the RTA to render it translocationally-competent to cross the endoplasmic reticulum (ER) membrane via the Sec61p translocon in a manner similar to that followed by misfolded ER proteins that, once recognised, are targeted to the ER-associated protein degradation (ERAD) machinery; (viii) avoiding, at least in part, ubiquitination that would lead to rapid degradation by cytosolic proteasomes immediately after membrane translocation when it is still partially unfolded; (ix) refolding into its protease-resistant, biologically active conformation; and (x) interaction with the ribosome to catalyse the depurination reaction. It is clear that ricin can take advantage of many target cell molecules, pathways and processes. It has been reported that a single molecule of ricin reaching the cytosol can kill that cell as a consequence of protein synthesis inhibition. The ready availability of ricin, coupled to its extreme potency when administered intravenously or if inhaled, has identified this protein toxin as a potential biological warfare agent. Therapeutically, its cytotoxicity has encouraged the use of ricin in 'magic bullets' to specifically target and destroy cancer cells, and the unusual intracellular trafficking properties of ricin potentially permit its development as a vaccine vector. Combining our understanding of the ricin structure with ways to cripple its unwanted properties (its enzymatic activity and promotion of vascular leak whilst retaining protein stability and important immunodominant epitopes), will also be crucial in the development of a long awaited protective vaccine against this toxin.

Transport of exogenous growth factors and cytokines to the cytosol and to the nucleus

In recent years a number of growth factors, cytokines, protein hormones, and other proteins have been found in the nucleus after having been added externally to cells. This review evaluates the evidence that translocation takes place and discusses possible mechanisms. As a demonstration of the principle that extracellular proteins can penetrate cellular membranes and reach the cytosol, a brief overview of the penetration mechanism of protein toxins with intracellular sites of action is given. Then problems and pitfalls in attempts to demonstrate the presence of proteins in the cytosol and in the nucleus as opposed to intracellular vesicular compartments are discussed, and some new approaches to study this are described. A detailed overview of the evidence for translocation of fibroblast growth factor, HIV-Tat, interferon-gamma, and other proteins where there is evidence for intracellular action is given, and translocation mechanisms are discussed. It is concluded that although there are many pitfalls, the bulk of the experiments indicate that certain proteins are indeed able to enter the cytosol and nucleus. Possible roles of the internalized proteins are discussed.

Differences in endocytosis and intracellular sorting of ricin and viscumin in 3T3 cells

Ricin and viscumin are heterodimeric protein toxins. Their A-chain is enzymatically active and removes an adenine residue from the 28S rRNA, the B-chain has lectin activity and binds to terminal galactose residues of cell surface receptors. The toxins reveal a high degree of identity in their amino acid sequences. Nevertheless, uptake into 3T3 cells occurs via different receptors and endocytotic pathways. This has been revealed by enzyme linked based analysis of ricin competition with viscumin, and by fluorochrome-labeled toxins (viscumin-FITC, ricin-Alexa 568), which were added simultaneously or separately to living cells. Then the uptake was followed by confocal laser scanning microscopy. Ricin immediately is delivered to the tubular and vesicular structures of endosomes in the perinuclear area while viscumin becomes endocytosed into small vesicles preferentially in the cell periphery. After about 60 min both these toxins may be found in tubo-vesicular structures of endosomes where the sorting process can directly be observed. The fact that this sorting takes place is a strong argument for the assumption that the toxins are bound to membrane proteins, either to their original receptors or to other proteins inside the endosomal compartment exhibiting terminal galactose residues. The toxins are biologically fully active as has been proven by binding and by toxicity experiments, thus the differences in targeting do not arise from labeling.

Selective regulation of the Rab9-independent transport of ricin to the Golgi apparatus by calcium

Transport of ricin from endosomes to the Golgi apparatus occurs, in contrast to the transport of the mannose 6-phosphate receptor, by a Rab9-independent process. To characterize the pathway of ricin transport to the Golgi apparatus, we investigated whether it was regulated by calcium. As shown here, our data indicate that calcium is selectively involved in the regulation of ricin transport to the Golgi apparatus. Thapsigargin, which inhibits calcium transport into the ER, and the calcium ionophore A23187 both increased the transport of ricin to the Golgi apparatus by a factor of 20. By contrast, transport of the mannose 6-phosphate receptor to the Golgi apparatus was unaffected. Ricin and mannose 6-phosphate receptor transport were measured by quantifying the sulfation of modified forms of ricin and the mannose 6-phosphate receptor. The increased transport of ricin was reduced by wortmannin and LY294002, suggesting that phosphoinositide 3-kinase might be involved in transport of ricin to the Golgi apparatus. Together, these findings indicate that the different pathways to the Golgi apparatus utilized by ricin and the mannose 6-phosphate receptor are regulated by different mechanisms.

Ricin

The plant toxin ricin consists of two disulfide-linked polypeptides with different functions. The A-chain enters the cytosol and inactivates the ribosomes enzymatically, whereas the B-chain has lectin properties and binds to carbohydrates at the cell surface. This binding is a requirement for translocation of the A-chain to the cytosol. The bound toxin is endocytosed and transported retrograde through the Golgi apparatus to the endoplasmic reticulum where it appears to be translocated to the cytosol by the sec61p complex.

RIBOSOME-INACTIVATING PROTEINS: A Plant Perspective

Ribosome-inactivating proteins (RIPs) are toxic N-glycosidases that depurinate the universally conserved alpha-sarcin loop of large rRNAs. This depurination inactivates the ribosome, thereby blocking its further participation in protein synthesis. RIPs are widely distributed among different plant genera and within a variety of different tissues. Recent work has shown that enzymatic activity of at least some RIPs is not limited to site-specific action on the large rRNAs of ribosomes but extends to depurination and even nucleic acid scission of other targets. Characterization of the physiological effects of RIPs on mammalian cells has implicated apoptotic pathways. For plants, RIPs have been linked to defense by antiviral, antifungal, and insecticidal properties demonstrated in vitro and in transgenic plants. How these effects are brought about, however, remains unresolved. At the least, these results, together with others summarized here, point to a complex biological role. With genetic, genomic, molecular, and structural tools now available for integrating different experimental approaches, we should further our understanding of these multifunctional proteins and their physiological functions in plants.

An interaction between ricin and calreticulin that may have implications for toxin trafficking

Here we demonstrate that ricin is able to interact with the molecular chaperone calreticulin both in vitro and in vivo. The interaction occurred with ricin holotoxin, but not with free ricin A chain; and it was prevented in the presence of lactose, suggesting that it was mediated by the lectin activity of the ricin B chain. This lectin is galactose-specific, and metabolic labeling with [(3)H]galactose or treating galactose oxidase-modified calreticulin with sodium [(3)H]borohydride indicated that Vero cell calreticulin possesses a terminally galactosylated oligosaccharide. Brefeldin A treatment indicated that the intracellular interaction occurred initially in a post-Golgi stack compartment, possibly the trans-Golgi network, whereas the reductive separation of ricin subunits occurred in an earlier part of the secretory pathway, most probably the endoplasmic reticulum (ER). Intoxicating Vero cells with ricin whose A chain had been modified to include either a tyrosine sulfation site or the sulfation site plus available N-glycosylation sites, in the presence of Na(2)35SO(4), confirmed that calreticulin interacted with endocytosed ricin that had already undergone retrograde transport to both the Golgi and the ER. Although we cannot exclude the possibility that the interaction between ricin and calreticulin is an indirect one, the data presented are consistent with the idea that calreticulin may function as a recycling carrier for retrograde transport of ricin from the Golgi to the ER.

Reconstitution of clathrin-independent endocytosis at the apical domain of permeabilized MDCK II cells: requirement for a Rho-family GTPase

This paper studies the endocytosis of ricin at the apical pole of polarized MDCK II cells after permeabilization of the cells basolaterally with streptolysin O. Ricin endocytosis after the addition of cytosol with an ATP-regenerating system was 2-3-fold higher than after the addition of a transport medium. A similar increase in ricin endocytosis was obtained by reconstitution of dialyzed cytosol with the nonhydrolyzable GTP analog, GTP gamma S, in the presence of an ATP-regenerating system. The nonhydrolyzable GDP analog, GDP beta S, did not increase ricin uptake. In contrast to the data obtained with ricin, GTP gamma S was found to inhibit apical transferrin uptake in MDCK II cells transfected with the human transferrin receptor, and the data thus imply that GTP gamma S supports clathrin-independent endocytosis. Electron microscopy (EM) demonstrated that free endocytic vesicles were formed from the apical pole of permeabilized MDCK II cells in the presence of GTP gamma S and that both a ricin-HRP conjugate, HRP, and cationized gold were endocytosed. Ricin endocytosis in the presence of intact cytosol, as well as GTP gamma S-stimulated ricin uptake, was inhibited by Clostridium botulinum C3 transferase, an enzyme found to inactivate Rho proteins. The data demonstrate that apical clathrin-independent endocytosis functions in the presence of GTP gamma S, and suggest that one or more of the small GTP binding proteins of the Rho family is involved in regulation of the apical clathrin-independent endocytosis in MDCK II cells.

Endosome to Golgi transport of ricin is regulated by cholesterol

We have here studied the role of cholesterol in transport of ricin from endosomes to the Golgi apparatus. Ricin is endocytosed even when cells are depleted for cholesterol by using methyl-beta-cyclodextrin (m beta CD). However, as here shown, the intracellular transport of ricin from endosomes to the Golgi apparatus, measured by quantifying sulfation of a modified ricin molecule, is strongly inhibited when the cholesterol content of the cell is reduced. On the other hand, increasing the level of cholesterol by treating cells with mbetaCD saturated with cholesterol (m beta CD/chol) reduced the intracellular transport of ricin to the Golgi apparatus even more strongly. The intracellular transport routes affected include both Rab9-independent and Rab9-dependent pathways to the Golgi apparatus, since both sulfation of ricin after induced expression of mutant Rab9 (mRab9) to inhibit late endosome to Golgi transport and sulfation of a modified mannose 6-phosphate receptor (M6PR) were inhibited after removal or addition of cholesterol. Furthermore, the structure of the Golgi apparatus was affected by increased levels of cholesterol, as visualized by pronounced vesiculation and formation of smaller stacks. Thus, our results indicate that transport of ricin from endosomes to the Golgi apparatus is influenced by the cholesterol content of the cell.

Modulation of antigen trafficking to MHC class II-positive late endosomes of enterocytes

BACKGROUND & AIMS

Oral tolerance is recognized as a central immunoregulatory phenomenon. The mechanisms of its induction remain unclear, and the role of the intestinal epithelial cells that are able to present antigens to T lymphocytes is poorly understood. In this report, we analyze under in vivo conditions the intracellular targeting of mucosally administered ovalbumin (OVA) to major histocompatibility complex (MHC) class II antigen containing compartments of enterocytes and compare these pathways between BALB/c and SCID mice, the latter being unable to generate a transferable tolerogenic moiety after a feed of OVA.

METHODS

OVA, lysosome-associated membrane proteins (LAMP-1), and MHC class II antigens were localized in jejunal biopsy specimens of BALB/c and SCID mice at 0, 5, 10, 20, 40, 60, and 120 minutes after a single feed with OVA by fluorescence and electron microscopy.

RESULTS

Ten minutes after oral administration, OVA was transported to the proximity of MHC class II antigens within LAMP-1-positive vacuoles and to the basolateral membrane of enterocytes from BALB/c strain mice. However, in SCID mice, OVA reached the paracellular spaces during the same time period through LAMP-1-negative vacuoles of enterocytes, which lacked MHC class II antigens.

CONCLUSIONS

Orally administered OVA is rapidly targeted to late endosomes containing LAMP-1 and MHC class II antigens in enterocytes of BALB/c mice but not in SCID mice bred on a BALB/c background. We suggest that this targeting process within the enterocytes is one of the requirements for the induction of oral tolerance.

Dependence of ricin toxicity on translocation of the toxin A-chain from the endoplasmic reticulum to the cytosol

Ricin acts by translocating to the cytosol the enzymatically active toxin A-chain, which inactivates ribosomes. Retrograde intracellular transport and translocation of ricin was studied under conditions that alter the sensitivity of cells to the toxin. For this purpose tyrosine sulfation of mutant A-chain in the Golgi apparatus, glycosylation in the endoplasmic reticulum (ER) and appearance of A-chain in the cytosolic fraction was monitored. Introduction of an ER retrieval signal, a C-terminal KDEL sequence, into the A-chain increased the toxicity and resulted in more efficient glycosylation, indicating enhanced transport from Golgi to ER. Calcium depletion inhibited neither sulfation nor glycosylation but inhibited translocation and toxicity, suggesting that the toxin is translocated to the cytosol by the pathway used by misfolded proteins that are targeted to the proteasomes for degradation. Slightly acidified medium had a similar effect. The proteasome inhibitor, lactacystin, sensitized cells to ricin and increased the amount of ricin A-chain in the cytosol. Anti-Sec61alpha precipitated sulfated and glycosylated ricin A-chain, suggesting that retrograde toxin translocation involves Sec61p. The data indicate that retrograde translocation across the ER membrane is required for intoxication.

Ricin RCA60: evidence of its phospholipase activity

The present work documents, on a qualitative and quantitative basis, the lipolytic activity of ricin protein RCA60 on glycerophospholipids. RCA60 demonstrates a low level of hydrolysis towards radioactive dipalmitoyl-glycerophosphatidylcholine. This observation was confirmed on a better substrate, palmitoyl-oleoyl-glycerophosphatidylcholine, after analysis of the reaction products by thin-layer and gas chromatography. A comparable hydrolytic activity was observed when palmitoyl-oleoyl-glycerophosphatidylethanolamine was used as substrate. The nature of the hydrolysis products supports the conclusion that RCA60 demonstrates phospholipase A1 and A2 activities as well as a lysophospholipase activity of A1 and A2 type. The insensitivity of this lipolytic activity towards calcium ions and the presence of the already described consensus sequence of lipases, Gly-Xaa-Ser-Xaa-Gly, in the primary sequence of the B-chain of RCA60 support the idea that the lipolytic activity of RCA60 is more related to the lipase family than to the phospholipases A. We hypothesize that such activity contributes to the mechanism which underlies the expression of the cytotoxicity of RCA60.

Differences in cytotoxicity of native and engineered RIPs can be used to assess their ability to reach the cytoplasm

Ricin is a heterodimeric cytotoxin composed of RTB, a galactose binding lectin, and RTA, an enzymatic N-glycosidase. The toxin is endocytosed, and after intracellular routing, RTA is translocated to the cytoplasm where it inactivates ribosomes resulting in a loss of host cell protein synthesis and cell death. We show for the first time that the cytotoxicity against cultured T cells by several RTA mutants is directly proportional to the enzyme activity of RTA, suggesting this is a reliable system to measure translocation effects. Large discrepancies between cytotoxicity and enzyme action for a given pair of toxins are therefore attributable to differences in cell binding, uptake, or membrane translocation. Fluid phase uptake and cytotoxicity of isolated RTA are essentially identical to that of the single chain toxin PAP. This important finding suggests that RTA, and the A chain of class 2 RIPs in general, has not evolved special translocation signals to complement the increased target cell binding facilitated by RTB. Experiments with the lectin RCA and with ebulin suggest those toxins have diminished cytotoxicity probably mediated by comparative deficiencies in B chain binding. Addition of a KDEL sequence to RTA increases fluid phase uptake, consistent with the notion that transport to the ER is important for cytotoxicity. Fusion of MBP or GST to the amino terminus of RTA has little effect on enzyme action or cytotoxicity. This result is not altered by protease inhibitors, suggesting the fusion proteins are probably not cleaved prior to translocation of the toxic A chain and implying that the toxins can carry large passenger proteins into the cytoplasm, an observation with interesting potential for analytical and therapeutic chemistry.

Expression of mutant dynamin inhibits toxicity and transport of endocytosed ricin to the Golgi apparatus

Endocytosis and intracellular transport of ricin were studied in stable transfected HeLa cells where overexpression of wild-type (WT) or mutant dynamin is regulated by tetracycline. Overexpression of the temperature-sensitive mutant dynG273D at the nonpermissive temperature or the dynK44A mutant inhibits clathrin-dependent endocytosis (Damke, H., T. Baba, A.M. van der Blieck, and S.L. Schmid. 1995. J. Cell Biol. 131: 69-80; Damke, H., T. Baba, D.E. Warnock, and S.L. Schmid. 1994. J. Cell Biol. 127:915-934). Under these conditions, ricin was endocytosed at a normal level. Surprisingly, overexpression of both mutants made the cells less sensitive to ricin. Butyric acid and trichostatin A treatment enhanced dynamin overexpression and increased the difference in toxin sensitivity between cells with normal and mutant dynamin. Intoxication with ricin seems to require toxin transport to the Golgi apparatus (Sandirg, K., and B. van Deurs. 1996. Physiol. Rev. 76:949-966), and this process was monitored by measuring the incorporation of radioactive sulfate into a modified ricin molecule containing a tyrosine sulfation site. The sulfation of ricin was much greater in cells expressing dynWT than in cells expressing dynK44A. Ultrastructural analysis using a ricin-HRP conjugate confirmed that transport to the Golgi apparatus was severely inhibited in cells expressing dynK44A. In contrast, ricin transport to lysosomes as measured by degradation of 125I-ricin was essentially unchanged in cells expressing dynK44A. These data demonstrate that although ricin is internalized by clathrin-independent endocytosis in cells expressing mutant dynamin, there is a strong and apparently selective inhibition of ricin transport to the Golgi apparatus. Also, in cells with mutant dynamin, there is a redistribution of the mannose-6-phosphate receptor.

Involvement of ATP-dependent Pseudomonas exotoxin translocation from a late recycling compartment in lymphocyte intoxication procedure

Pseudomonas exotoxin (PE) is a cytotoxin which, after endocytosis, is delivered to the cytosol where it inactivates protein synthesis. Using diaminobenzidine cytochemistry, we found over 94% of internalized PE in transferrin (Tf) -positive endosomes of lymphocytes. When PE translocation was examined in a cell-free assay using purified endocytic vesicles, more than 40% of endosomal 125I-labeled PE was transported after 2 h at 37 degrees C, whereas a toxin inactivated by point mutation in its translocation domain was not translocated. Sorting of endosomes did not allow cell-free PE translocation, whereas active PE transmembrane transport was observed after > 10 min of endocytosis when PE and fluorescent-Tf were localized by confocal immunofluorescence microscopy within a rab5-positive and rab4- and rab7-negative recycling compartment in the pericentriolar region of the cell. Accordingly, when PE delivery to this structure was inhibited using a 20 degrees C endocytosis temperature, subsequent translocation from purified endosomes was impaired. Translocation was also inhibited when endosomes were obtained from cells labeled with PE in the presence of brefeldin A, which caused fusion of translocation-competent recycling endosomes with translocation-incompetent sorting elements. No PE processing was observed in lymphocyte endosomes, the full-sized toxin was translocated and recovered in an enzymatically active form. ATP hydrolysis was found to directly provide the energy required for PE translocation. Inhibitors of endosome acidification (weak bases, protonophores, or bafilomycin A1) when added to the assay did not significantly affect 125I-labeled PE translocation, demonstrating that this transport is independent of the endosome-cytosol pH gradient. Nevertheless, when 125I-labeled PE endocytosis was performed in the presence of one of these molecules, translocation from endosomes was strongly inhibited, indicating that exposure to acidic pH is a prerequisite for PE membrane traversal. When applied during endocytosis, treatments that protect cells against PE intoxication (low temperatures, inhibitors of endosome acidification, and brefeldin A) impaired 125I-labeled PE translocation from purified endosomes. We conclude that PE translocation from a late receptor recycling compartment is implicated in the lymphocyte intoxication procedure.

Retrograde transport: going against the flow

Certain protein toxins act by catalytically modifying substrates in the cytosol of mammalian cells. To reach this compartment, these proteins undergo retrograde transport from the cell surface, via the Golgi complex, to the endoplasmic reticulum.

Differential toxicity of ricin and diphtheria toxin for bloodstream forms of Trypanosoma brucei

The cytotoxicity of ricin and diphtheria toxin was studied in culture-adapted bloodstream forms of Trypanosoma brucei. Although ricin is endocytosed at a rate comparable to that of other internalized macromolecules, it is nontoxic to bloodstream-form trypanosomes. The resistance lies partly in low susceptibility of the targeted ribosomes: T. brucei cell-free protein biosynthesis is only partially inhibited by ricin A chain. In addition, ricin is degraded before it reaches the ribosomes, as the toxin is delivered to lysosomes. In contrast, diphtheria toxin shows similar cytotoxicities for bloodstream-form trypanosomes and mouse myeloma cells. Both trypanosome and myeloma cells are more than 1000-fold less sensitive to the action of the toxin than most other mammalian cell lines, although nicked reduced diphtheria toxin inhibits cell-free protein synthesis of T. brucei and myeloma cells to the same extent as that of a rabbit reticulocyte lysate. The effect of diphtheria toxin on T. brucei in vitro translation is NAD+ dependent, suggesting that ADP-ribosylation of elongation factor 2 could be the cause of the inhibition as it is in mammalian cells. Thus, the toxic moiety of diphtheria toxin is suitable for preparation of cell-type-specific cytotoxic reagents directed towards trypanosomes.

Synthesis of green fluorescent protein-ricin and monitoring of its intracellular trafficking

We performed genetic engineering to fuse enhanced green fluorescent protein (EGFP) to the N terminus of RTA, expressed the fusion protein in Escherichia coli, purified and reassociated EGFP-RTA with plant RTB, and purified EGFP-ricin by size exclusion HPLC. The fusion heterodimer was able to bind galactosides, intoxicate cells, and show strong fluorescence. Mammalian cells incubated with EGFP-ricin showed strong cell surface fluorescence at 4 degrees C and, on incubation at 37 degrees C, distributed initially to endosomes and then to Golgi vesicles. Variable sensitivity of mammalian cells to ricin and ricin fusion proteins may be due in part to different patterns of intracellular routing. Cells were incubated with ricin or EGFP-ricin, and inhibition of protein synthesis was measured. Human hepatocellular carcinoma Hep3B cells were 10-fold more sensitive to ricin and 85-fold more sensitive to EGFP-ricin than human epidermoid carcinoma KB cells. Epifluorescence microscopy of cells incubated with EGFP-ricin showed greater localization of the fluorescence signal in the Golgi compartments in Hep3B cells than in KB cells. These data support a model requiring a Golgi-dependent step in cell intoxication by ricin. The work further identifies the usefulness of green fluorescent protein fusions in the study of retrograde transport of internalized peptides.

Lectin-deficient ricin toxin intoxicates cells bearing the D-mannose receptor

Ricin toxin with genetic or chemical modification of lectin sites has been previously reported to show markedly reduced cytotoxicity to cells following uptake by several receptors including the mannose receptor. Investigators have hypothesized that an intracellular galactoside-binding function was required for optimal intracellular targeting of ricin for these receptors. We have prepared insect-derived mutant ricin toxin B chain (RTB) with modifications of three lectin side domains (1 alpha, 1 beta, and 2 gamma) yielding a 1000-fold reduced galactoside avidity. After reassociation with plant RTA, the recombinant heterodimer and plant ricin were tested for cytotoxicity on mammalian cells expressing (mouse peritoneal macrophages, J774E cells, and MMR61 cells) or not expressing (KB cells) the D-mannose receptor. Receptor expression was confirmed by immunofluorescence microscopy. Lactose was included in the media to block cell-surface galactoside binding, and mannan was added as a control in each experiment to confirm mannose receptor-specific targeting. Plant ricin A chain (RTA) and E. coli-derived RTA were also tested for cytotoxicity on J774E and KB cells. Both wild-type and lectin-deficient ricin displayed mannose-receptor mediated cell cytotoxicity. This is the first report of a genetically modified ricin showing that RTB intracellular galactose binding activity is not required for ricin cytotoxicity. Sensitivity of mannose-receptor bearing cells, but not control cells, to mannosylated RTA, but not unglycosylated RTA, confirmed these observations. These results imply fusion toxins employing ricin can be prepared with maximal reductions in normal tissue binding.

Retrograde transport of mutant ricin to the endoplasmic reticulum with subsequent translocation to cytosol

Translocation of ricin A chain to the cytosol has been proposed to take place from the endoplasmic reticulum (ER), but attempts to visualize ricin in this organelle have failed. Here we modified ricin A chain to contain a tyrosine sulfation site alone or in combination with N-glycosylation sites. When reconstituted with ricin B chain and incubated with cells in the presence of Na(2)(35)SO(4), the modified A chains were labeled. The labeling was prevented by brefeldin A and ilimaquinone, and it appears to take place in the Golgi apparatus. This method allows selective labeling of ricin molecules that have already been transported retrograde to this organelle. A chain containing C-terminal N-glycosylation sites became core glycosylated, indicating retrograde transport to the ER. In part of the toxin molecules, the A chain was released from the B chain and translocated to the cytosol. The finding that glycosylated A chain was present in the cytosol indicates that translocation takes place after transport of the toxin to the ER.

Ricin toxin contains three lectin sites which contribute to its in vivo toxicity

Ricin intoxication of mammalian cells is initiated by B chain (RTB) binding to cell surface galactosides. Recombinant insect-derived RTB mutants with modifications in lectin-site subdomains 2 gamma, 1 alpha/2 gamma, and 1 alpha/1 beta/2 gamma were reassociated with plant RTA and tested for lethality in C57B1/6 6-8 weeks old mice. The LD50 of intraperitoneally injected castor bean ricin was 75 ng per 18 g mouse. The LD50 of single-site 2 gamma mutant heterodimer was 100 ng: the LD50 of the double-site 1 alpha/2 gamma mutant heterodimer was 500 ng, and the LD50 of the triple-site 1 alpha/1 beta,2 gamma mutant heterodimer was > 10 micrograms. Plant RTA alone had an LD50 of 300 micrograms. Animals died between 1 and 10 days post-injection. Histopathological examination of morbid animals receiving an LD50 dose of each toxin revealed only apoptosis in the thymus and spleen. The present data provide clear evidence for participation of three lectin sites in ricin in vivo toxicity. These results suggest an origin for some of the normal tissue toxicities observed with clinical trials of doubly blocked ricin conjugates and suggest modification of at least three RTB subdomains will be necessary in genetically engineered ricin fusion proteins.

Thapsigargin-induced transport of cholera toxin to the endoplasmic reticulum

Cholera toxin is normally observed only in the Golgi apparatus and not in the endoplasmic reticulum (ER) although the enzymatically active A subunit of cholera toxin has a KDEL sequence. Here we demonstrate transport of horseradish peroxidase-labeled cholera toxin to the ER by electron microscopy in thapsigargin-treated A431 cells. Thapsigargin treatment strongly increased cholera toxin-induced cAMP production, and the formation of the catalytically active A1 fragment was somewhat increased. Binding of cholera toxin to the cell surface and transport of toxin to the Golgi apparatus were not changed in thapsigargin-treated cells, suggesting increased retrograde transport of cholera toxin from the Golgi apparatus to the ER. The data demonstrate that retrograde transport of cholera toxin can take place and that the transport is under regulation. The results are consistent with the idea that retrograde transport can be important for the action of cholera toxin.

Importance of glycolipid synthesis for butyric acid-induced sensitization to shiga toxin and intracellular sorting of toxin in A431 cells

The human epidermoid carcinoma cell line A431 becomes highly sensitive to Shiga toxin upon treatment with butyric acid. This strong sensitization (> 1000-fold) is accompanied by an increase in the fraction of cell-associated toxin transported to the Golgi apparatus and to the endoplasmic reticulum (ER). Furthermore, our previous work showed that the length of the fatty acyl chain of Gb3, the Shiga toxin receptor, also was changed (longer fatty acids). We have not investigated the importance of this change by testing whether glycolipid synthesis is required for the changed intracellular sorting and the toxin sensitivity. We demonstrate here that inhibition of glycosphingolipid synthesis by inhibition of N-acyltransferase with fumonisin B1, by inhibition of glucosylceramide synthetase by PDMP or PPMP, or by inhibition of serine palmitoyl transferase by beta-fluoroalanine, inhibited the butyric acid-induced change in sensitivity and the increase in the fraction of cell-associated Shiga toxin transported to the Golgi apparatus and the ER. The block in butyric acid-induced sensitization caused by beta-fluoroalanine could be abolished by simultaneous addition of sphinganine or sphingosine. Thus, the data suggest that the fatty acyl chain length of glycosphingolipids is important for intracellular sorting and translocation of Shiga toxin to the cytosol.

A saporin-insulin conjugate: synthesis and biochemical characterization

Saporin, a single-chain, non-cytotoxic, ribosome-inactivating protein from Saponaria officinalis, was chemically linked to the hormone insulin in a 1:1 complex. To follow by dynamic video microscopy the endocytosis and intracellular transport in vivo, a second covalent conjugate with a saporin derivative labelled with fluorescein isothiocyanate was also prepared. Both conjugates were characterized with reference to homogeneity, stoichiometry, optical spectroscopy and toxicity. Both were found to exhibit scarce toxicity toward both CHO and HEP G2 cells; optical video microscopy on living cells indicates that reduced toxicity may be (partly) due to a very limited binding of the saporin-insulin conjugate to membrane receptors. These results suggest a strategy for new possible covalent conjugates of saporin with alternative and specific macromolecular carriers.

Mannose receptor dependent uptake of ricin A1 and A2 chains by macrophages

The ricin A chain, the toxic subunit of ricin, consists of two forms which differ in sugar content. The major component A1 contains one high mannose chain while the minor component A2 contains an additional high mannose chain. Endocytosis of this toxin occurs in macrophages via the mannose receptor. To study the role of the sugar residues in ricin A chain cytotoxicity, we have purified the two forms by ion-exchange chromatography. The uptake of A1 and A2 by a macrophage cell line was concentration and time dependent. The total amount of A2 internalized was approximately twice the amount of A1, indicating a higher affinity of A2 for the mannose receptor. Ricin A2 was four times more toxic to macrophages than A1, in agreement with the higher affinity of A2 compared to the A1. These experiments suggest that the high mannose chains on the A chain promote mannose-receptor-mediated endocytosis by providing the initial binding to the cell surface. Once the toxin is accumulated inside the cell however, the carbohydrates do not seem to influence intracellular transport and/or translocation of the ricin A chain into the cytoplasm.

High-resolution NMR study of a GdAGA tetranucleotide loop that is an improved substrate for ricin, a cytotoxic plant protein

Ricin is a cytotoxic plant protein that inactivates ribosomes by hydrolyzing the N-glycosidic bond at position A4324 in eukaryotic 28S rRNA. Recent studies showed that a four-nucleotide loop, GAGA, can function as a minimum substrate for ricin (the first adenosine corresponds to the site of depurination). We previously clarified the solution structure of this loop by NMR spectroscopy [Orita et al. (1993) Nucleic Acids Res. 21, 5670-5678]. To elucidate further details of the structural basis for recognition of its substrate by ricin, we studied the properties of a synthetic dodecanucleotide, r1C2U3C4A5G6dA7G8A9U10G11A12G (6dA12mer), which forms an RNA hairpin structure with a GdAGA loop and in which the site of depurination is changed from adenosine to 2'-deoxyadenosine. The N-glycosidase activity against the GdAGA loop of the A-chain of ricin was 26 times higher than that against the GAGA loop. NMR studies indicated that the overall structure of the GdAGA loop was similar to that of the GAGA loop with the exception of the sugar puckers of 6dA and 7G. Therefore, it appears that the 2'-hydroxyl group of adenosine at the depurination site (6A) does not participate in the recognition by ricin of the substrate. Since the 2'-hydroxyl group can potentially destabilize the developing positive charge of the putative transition state intermediate, an oxycarbonium ion, the electronic effect may explain, at least in part, the faster rate of depurination of the GdAGA loop compared to that of GAGA loop. We also show that the amino group of 7G is essential for substrate recognition the ricin A-chain.

Characterization of single site ricin toxin B chain mutants

DNA encoding ricin B chain was modified by site-directed mutagenesis, and eight separate mutant RTB cDNAs including four novel mutants were ligated into the baculovirus transfer vector, pAcGP67A. Cotransfection of S. frugiperda Sf9 cells with BaculoGold DNA was followed by limiting dilution isolation of recombinant baculoviruses. Infection of Sf9 cells at a multiplicity of infection of 5 in the presence of 25 mM lactose produced 0.05-1 mg/L of soluble, glycosylated 34 kDa proteins immunoreactive with monoclonal and polyclonal antibodies to ricin B chain. Mutant ricin B chains were partially purified by monoclonal antibody immunoaffinity chromatography to 10-50% purity in near milligram quantities. The mutant ricin B chains had decreased lectin binding relative to plant ricin B chain as measured by binding to immobilized lactose and asialofetuin and cell binding immunofluorescence. The mutant ricin B chains reassociated with plant RTA similarly to plant RTB, and the recombinant heterodimers had slightly reduced cell cytotoxicity relative to ricin.

Point mutations in the hydrophobic C-terminal region of ricin A chain indicate that Pro250 plays a key role in membrane translocation

A series of mutations have been made in the carboxyl terminus of ricin A chain, centred on the hydrophobic region between amino acid residues Val245 and Val256. The mutant ricin A chains were expressed to a high level in an Escherichia coli system and the proteins purified to homogeneity. The enzymic activity of each of these A chain molecules was tested on rabbit reticulocyte ribosomes; in all cases, the activities were found to be comparable to wild-type recombinant ricin A chain. Following reassociation of these A chains to ricin B chain, Vero cells were challenged with these holotoxins and the cytotoxicities determined. Mutant ricin A chain with Ile247-->Ala was unable to reassociate and form holotoxin, indicating the importance of this residue in the interaction with ricin B chain. Mutant ricin A chain with Pro250-->Ala readily reassociated with ricin B chain, forming holotoxin with a 170-fold reduction in cytotoxicity to Vero cells. Other mutations in this region also produced A chain proteins which gave marked reductions in holotoxin cytotoxicity. We propose therefore that the C-terminal hydrophobic region of ricin A chain may be involved in membrane interactions prior to the translocation of this subunit into the cytosol, and that Pro250 plays a key role in one or both of these steps.

Furin-induced cleavage and activation of Shiga toxin

Shiga toxin has a single A subunit non-covalently associated with a pentamer of B subunits. The toxin has a trypsin-sensitive region near the COOH-terminal end of the A-chain, and upon cleavage, two disulfide bonded fragments, A1 and A2, are generated. These fragments are also formed upon incubation with cells. The disulfide loop contains the sequence (Arg-X-X-Arg), which is a consensus motif for cleavage by the membrane-anchored protease furin. We found that a soluble form of furin cleaves intact A-chain producing A1 and A2 fragments, and furin also seems to be responsible for rapid cellular cleavage of Shiga toxin. LoVo cells, which normally do not produce functional furin, cleave intact A-chain very efficiently when transfected with furin (LoVo/fur), whereas a control cell (LoVo/neo) cleaves the toxin very slowly. To investigate the role of this cleavage for intoxication of cells, we studied the ability of unnicked and furin-nicked toxin to inhibit protein synthesis in LoVo/fur and LoVo/neo cells. LoVo/fur cells were intoxicated equally well with unnicked and nicked toxin, whereas in LoVo/neo cells nicked toxin was about 20 times more active than unnicked toxin. The results suggest that cleavage of Shiga toxin is important for intoxication of cells, and they indicate that furin can cleave and thereby activate Shiga toxin in cells.

Different cytotoxic activity and intracellular fate of an anti-CD5-momordin immunotoxin in normal compared to tumour cells

We investigated the different sensitivity of peripheral blood mononuclear cells (PBMC) and human T cell leukaemias (Jurkat and CEM) to an anti-CD5-momordin immunotoxin. In a short-term assay, the immunotoxin displayed different cytotoxic activity on normal and tumour cells: for leukaemic cell lines an incubation time of 72 h was necessary for the immunotoxin to reach the IC50 of 41-53 pM, compared to the 1 h sufficient for 6 pM immunotoxin to inhibit 50% of PBMC protein synthesis. In a long-term clonogenic assay (15 days), the immunotoxin demonstrated a comparable efficacy of clonogenic cell killing for both cell types. We investigated the immunotoxin internalization pathway by a flow-cytometric method and our data seem to indicate that the molecules meet a different intracellular fate in the two cell populations. It may be assumed that the low cytotoxic activity of immunotoxins on tumour cells, detected in the short-term assay, is due to inefficient delivery to their cytoplasmatic target, while a longer exposure of the cells to the immunotoxin promotes adequate intracellular distribution.

Retrograde transport from the Golgi complex to the ER of both Shiga toxin and the nontoxic Shiga B-fragment is regulated by butyric acid and cAMP

Endocytosed Shiga toxin is transported from the Golgi complex to the endoplasmic reticulum in butyric acid-treated A431 cells. We here examine the extent of this retrograde transport and its regulation. The short B fragment of Shiga toxin is sufficient for transport to the ER. The B fragment of cholera toxin, which also binds to glycolipids, is transported to all the Golgi cisterns, but cannot be localized in the ER even after butyric acid treatment. Under all conditions the toxic protein ricin was found predominantly in the trans-Golgi network. There is no transport of endocytosed fluid to the Golgi apparatus or to the ER even after butyric acid treatment and in the presence of Shiga toxin, indicating that transport to the ER, through the trans-Golgi network and the cisterns of the Golgi apparatus, involves several sorting stations. Since Shiga toxin receptors (Gb3) in butyric acid-treated A431 cells seem to have a ceramide moiety with longer fatty acids than in untreated cells, the possibility exists that fatty acid composition of the receptor is important for sorting to the ER. Both retrograde transport and intoxication with Shiga toxin can also be induced by cAMP, supporting the idea that retrograde transport from the Golgi to the ER is required for intoxication. The data suggest that transport to the ER in cells in situ may depend on fatty acid composition and is regulated by physiological signals.

Ultrastructural alterations induced in quail ciliary neurons by postganglionic nerve crush and by Ricinus toxin administration, separately and in combination

The response to postganglionic nerve crush and Ricinus toxin administration by the ciliary neurons of the quail ciliary ganglion was investigated at the ultrastructural level. The toxin was either applied at the crush site on the postganglionic nerves or injected into the anterior eye chamber without any other operative intervention. Crush of postganglionic nerves without toxin administration and saline injection into the anterior eye chamber served as controls for the two toxin administration procedures. Postganglionic nerve crush caused a distinct chromatolytic reaction, accompanied by massive detachment of the preganglionic axon terminals from the ciliary neurons and loss of most of the synapses, both chemical and electrical. This process does not induce cell death and is reversible. Saline injection in the anterior eye chamber caused a moderate retrograde reaction in some of the ciliary neurons, presumably as a consequence of paracentesis. The changes consisted mainly of an increase of perikaryal neurofilaments with, at most, a minor detachment of the preganglionic boutons from a small portion of the cell body at the nuclear pole. Ricinus toxin administration induced neuronal degeneration following a pattern common to both delivery modes. The degenerative process consisted of disruption and detachment of polyribosomes from the rough endoplasmic reticulum, an increase of smooth cisterns and tubules, a dramatic increase of neurofilament bundles, compartmentalization of the cytoplasmic organelles and, finally, karyorrhexis and cell lysis. The final stages of Ricinus toxin degeneration involve a progressive accumulation of extracellular flocculo-filamentous material and cell lysis. After administration of Ricinus toxin to the crush site, ricin-affected neurons showed withdrawal of the preganglionic boutons from a portion of the ciliary neuron, especially at the nuclear pole. After Ricinus toxin injection into the anterior eye chamber, however, the bouton shell surrounding the affected ciliary neurons remained intact in the early stages of degeneration. Detachment of the preganglionic terminals and disruption of the cell junctions, therefore, is the consequence of nerve crush and not of the toxin itself. This study demonstrated that quail ciliary neurons are a suitable model for experimental neuropathology and neurotoxicology.

Intracellular dynamics of ricin followed by fluorescence microscopy on living cells reveals a rapid accumulation of the dimeric toxin in the Golgi apparatus

The intracellular dynamics of fluorescent conjugates of the toxic lectin ricin was followed by video fluorescence microscopy on living CHO cells, demonstrating that the ricin heterodimer and its isolated B chain, after binding to the plasma membrane receptors, migrate to and accumulate in the Golgi apparatus following internalization. A ricin derivative labelled with fluorescein on the A chain and rhodamine on the B chain did not display significant splitting of the A-B heterodimer during translocation of the toxin to the Golgi; this novel finding provides support for the hypothesis that further processing of ricin takes place in this cellular compartment.

Entamoeba histolytica: electron-dense granule secretion, collagenase activity and virulence are altered in the cytoskeleton mutant BG-3

HM-1:IMSS, a pathogenic strain of Entamoeba histolytica, and its mutant BG-3, identified by resistance to cytochalasin D, were tested for their capacity to: (i) secrete electron-dense granules; (ii) adhere and digest native type I collagen gels; and (iii) produce liver abscesses in new-born hamsters. The results demonstrate that the mutant has low adherence to collagen, low electron-dense granule secretion and collagenolytic activity, and low capacity to produce liver lesions in vivo, compared with the parental strain HM1:IMSS.

High-resolution NMR study of a synthetic oligoribonucleotide with a tetranucleotide GAGA loop that is a substrate for the cytotoxic protein, ricin

Ricin is a cytotoxic protein that inactivates ribosomes by hydrolyzing the N-glycosidic bond at position A4324 in eukaryotic 28S rRNA. Its substrate domain forms a double helical stem and a 17-base loop that includes the sequence GAGA, the second adenosine of which corresponds to A4324. Recently, studies of mutant RNAs have shown that the four-nucleotide loop, GAGA, can function as a substrate for ricin. To investigate the structure that is recognized by ricin, we studied the properties of a short synthetic substrate, the dodecaribonucleotide r-CUCAGAGAUGAG, which forms a RNA hairpin structure with a GABA loop and a stem of four base pairs. The results of NMR spectroscopy allowed us to construct the solution structure of this oligonucleotide by restrained molecular-dynamic calculations. We found that the stem region exists as an A-form duplex. 5G and 8A in the loop region form an unusual G:A base pair, and the phosphodiester backbone has a turn between 5G and 6A. This turn seems to help ricin to gain access to 6A which is the only site of depurination in the entire structure. The overall structure of the GAGA loop is similar to those of the GAAA and GCAA loops that have been described but that are not recognized by ricin. Therefore, in addition to the adenosine at the depurination site, the neighboring guanosine on the 3' side (7G) may also play a role in the recognition mechanism together with 5G and 8A.