Two types of ribosome-inactivating proteins have been purified from plants: type I, which is made of a single polypeptide chain, and type II, which consists of two nonhomologous subunits. This chapter describes some of the properties of these proteins, named Luffa ribosome-inactivating proteins, and summarizes the cloning and expression of a related protein. A variety of single-chain ribosome-inactivating proteins have been isolated from plants and chemically linked to antibodies to form highly toxic, specific immunotoxins. Amino acid sequences were determined in a Beckman automated peptide sequencer. The fusion protein with ß-galactosidase did not show significant inhibition of protein synthesis in a cell-free system. Plaques showing a positive signal were then re-purified by plating at a lower density until all plaques in the dish were positive for the expression of fusion protein. In our cloning procedure, four rounds of plaque purification were necessary to reach that stage.
The proenzyme has very little adenosine diphosphate ribosyl transferase activity in vitro unless activated by simultaneous treatment with a denaturant, such as urea, and a reducing agent, such as dithiothreitol. Pseudomonas aeruginosa is a common, opportunistic, nosocomial pathogen infecting patients with weakened immunity, such as those with burns. Although a comprehensive functional map of the active site is still unavailable, several workers nevertheless have identified specific amino acids in domain III that are essential for the enzymic activity of the toxin. Since the discovery in 1966 that the lactose repressor protein is responsible for the selective induction of ß-galactosidase by E. coli in the presence of lactose, extensive studies have been made both on the lactose repressor, the bacteriophage lambda repressor, and other bacterial gene-regulating proteins. The transcriptional start site for the exotoxin A gene has been investigated by SI nuclease mapping.
Molecular cloning strategies depend largely upon what is already known about the protein or gene of interest, the size and abundance of the mRNA, and the availability of specific screening tools. Strategies beyond the scope of this brief overview include differential and subtractive cloning methods, functional and ligand screening assays, and polymerase chain reaction techniques for target genes where no sequence information is known. Straightforward polymerase chain reaction strategies are highlighted that can enhance progress in cooperative structural studies. Results of the SVS II study clearly illustrate the complementarity of protein chemistry and molecular biology. Direct analyses of this prototype plant toxin have revealed the mature protein structure and provided a sequence information critical for cloning, whereas DNA studies have demonstrated that the A and B chains are synthesized together as a single polypeptide containing both leader and linker sequences.
The DTB fragment is located at the carboxyl-terminus of the molecule, and contains a second disulfide bond between cysteine residues 461 and 471. The most commonly used method to introduce mutations at random into the DT gene is to expose the corynephage-carrying tox to N-methyl-N'-nitro-N-nitrosoguanidine (NG) sometimes after induction of lysogeny by exposure to ultraviolet (UV) light. CRMs with no enzymatic activity, or with reduced activity when compared on an equimolar basis with wild-type DT have been isolated; in general, the ability of a mutant to ADP-ribosylate EF-2 is related to its toxicity to sensitive cells. The position and number of arginine residues in a polypeptide has been shown to be crucial for its ability to interact in the correct manner with lipid membranes; it is possible that joint membrane insertion of the carboxyl-terminal end of DTA and the amino-terminal end of DTB initiates the translocation process, but only if the A fragment contains a single C-terminus arginine residue.
Pseudomonas exotoxin A (ETA) has interesting similarities to and differences with diphtheria toxin in its mechanism of killing cells. The N-terminal domain of ETA appears to include the cell surface-binding domain, although direct demonstration of this activity in a specific domain is lacking because of the difficulty in measuring ETA binding to cells. Domain II has been proposed to play a role in the entry of domain III into the cytosol based upon the loss of toxicity of domain II mutants without loss in cell surface-binding activity or loss of adenosine diphosphate (ADP)-ribosylation activity. Expression of C-terminal fragments of the toxin that correspond with domain III of the crystal structure demonstrates that the last 306 amino acids contain the ADP-ribosylation activity. The long history of research in DT mechanism relative to the recent burst of interest in ETA is reflected in the approaches taken to probe structure-function relationships.
Growth factor-toxin fusion proteins may also be used to regulate immune responses by eliminating subsets of lymphocytes or mononuclear cells that express specific cytokine receptors. A variety of genetic constructions employing several types of microbial-expression plasmids have been used successfully to produce growth factor-toxin fusion proteins. In particular, the optimal transcriptional promoters, ribosome-binding sites, and distances between the promoter and translational start site for each genetic construct are often different for different growth factor-toxin fusion proteins. The major obstacle to bringing several growth factor-toxin fusion proteins to clinical trials is the production and purification of sufficient quantities of biologically active protein. One of the most critical requirements for successful large-scale isolation of a recombinant fusion protein from bacteria is proper formation of disulfide bonds. An appropriate refolding step is generally essential for recovery of biologically active foreign proteins from bacteria.
Bacterial toxins such as diphtheria toxin and Pseudomonas exotoxin A and plant toxins such as ricin and abrin are able to kill their target cells by delivering a catalytically active polypeptide into the cell cytosol. In the bacterial toxins, the ADP-ribosylating peptide and the cell-binding peptide are initially synthesized as distinct domains within a single chain protein molecule. The chimeric proteins are endocytosed by their target cells and are converted through the action of intracellular proteases into their cytotoxic heterodimeric forms. The approach of making functionally cytotoxic, single-chain chimeric proteins by the generation and expression of novel DNA fusions cannot be directly extended to the plant toxins. The alternative approach we have successfully used is to employ site-directed mutagenesis to alter residues within the ricin linker sequence to generate a specific protease recognition site. One advantage of recombinant chimeras containing the diphtheria toxin loop is that target cell protease(s) can process the molecule intracellularly to a cytotoxic heterodimeric form.
The application of current molecular biological techniques has permitted dissection and manipulation of the gene, thus permitting a detailed analysis of the protein and a way to wield the extreme potency of the toxin for targeting to discrete cell populations. In 1888, the bacterium Corynebacterium diphtheriae was shown to secrete a protein toxin which contributed to the pathogenicity of diphtheria. It took over 65 years to demonstrate an association between bacteria producing diphtheria toxin and lysogenization by bacteriophage. The lack of spontaneous conversion of nontoxinogenic strains to toxinogenicity in the absence of infection by bacteriophage, and the rapid appearance of toxinogenic cells following infection, argued against the selection of a "mutant" population in favor of an induced change associated with the phage. In order to clone the diphtheria toxin gene, phage genome restriction fragments bearing tox were identified so that the smallest fragment containing the entire diphtheria toxin gene could be cloned with a minimum amount of extraneous phage DNA.
Activation may be an issue when testing cytokine fusion toxins in which the cytokine still may have the capability of triggering activation prior to delivery of a toxic signal. Several factors might contribute to conjugate toxicity in vivo. In the majority of clinical trials, the vascular leak syndrome has occurred. In the case of cytokine fusion toxins, it could be feasible for the cytokine to trigger its receptor without enough conjugate being present to trigger toxicity. Historically, animal models have played an important role in the clinical development of immunoconjugates. Activation of cells triggered by the binding of mAb or cytokine, especially to cells of the immune system, can have toxic consequences. Fusion toxins are constructed by fusing cytokine genes or antibody genes to toxin genes. Molecular biology has given new life to the immunoconjugate field. On the one hand, toxin conjugates have major limitations in achieving doses necessary for saturation of clinical target sites, curing a given disease.
The production of ribotoxins as potent as a-Sarcin and restrictocin by aspergini raises questions about the ability of the microbes to protect themselves during production. However, in spite of some promising indications, the antitumor activity of the fungal protein toxins was considered to be too limited and the toxicity too great to merit continued investigation. The ribotoxins a-Sarcin, restrictocin, and mitogillin form a specific class of ribosome-inactivating protein (RIP), they differ in structure and mode of action from the plant RIPs. The plant RIPs exist as single chain or double chain; the latter are substantially more potent in vivo since one of the chains is a cell surface binding protein which targets the toxin. However, the fact that ribonucleolytic activity and sequence homology with "typical" ribonucleases can be found associated with proteins that have diverse biological roles related to this activity suggests that such proteins may play specific roles in cell function.
Trichosanthin (TCS) and a-momorcharin are plant proteins isolated from Trichosanthes kirilowii and Momordica charantia, respectively. In China, TCS has been administered for abortion and the treatment of ectopic pregnancy, hydatidiform mole, invasive mole, and choriocarcinoma. In order to elucidate the structure-function relationship of these two proteins and to improve their pharmacological usage, we cloned and expressed the cDNA of these proteins. To improve expression, we then removed the nonessential 0.4 kb from the encoded propeptide of TCS up to the transcriptional terminal signal of the vector to generate the expression vector pTRC58210. The N-terminal 23-amino acid is a signal peptide as it is hydrophobic and does not exist in the mature a-MMC protein.
Correlation of the structure of ricin A chain with its catalytic activity will aid in the understanding of the mechanism by which it inactivates protein synthesis. The use of ricin A chain as a component of immunotoxins in the treatment of a number of human diseases may present new problems, including immuogenicity, toxicity, and localization to the desired target. Results from trinitrophenylation of whole ricin D suggests that free amino groups exposed on the surface of the whole toxin may be involved in the enzymatic activity of the A chain. Modification of arginine residues in isolated ricin A chain by either phenylglyoxal or 1,2-cyclohexanedione inactivates its inhibitor activity in a cell-free protein synthesis system. When considering the effect of amino acid substitutions or mutations on the catalytic activity of ricin A chain, the background within which the mutations are made must be considered.
This chapter discusses the structure of the heterodimeric toxin ricin, aimed at serving as a guide to the rational engineering of the protein. Biochemists and medical researchers have made extensive use of both forms of protein, which we shall refer to interchangeably as ribosome-inactivating proteins or plant toxins, in the search for therapeutic agents. The structure is also vital to guide the rational design of specific genetic mutations to be created by molecular biology techniques. The linear sequence alignments were used to guide amino acid substitutions, insertions, and deletions made to the three-dimensional RTA model using an interactive graphics system. The galactosyl moiety of lactose is oriented and secured by hydrogen bonds to the side chains of several polar residues which lie at the back of the pocket, whereas the glucosyl moiety extends freely into solvent and makes no specific interaction with the protein.
Diphtheria toxin is produced and secreted by strains of Corynebacterium diphtheriae that are lysogenic for one of a number of toxigenic coryne-bacteriophages. The N-terminal 21,167 fragment of toxin, fragment A, is the catalytically active toxophore responsible for the adenosine diphosphorylribosylation of elongation factor 2 within the cytosol of intoxicated cells. As long as the targeted receptors were internalized by receptor-mediated en-docytosis, the diphtheria toxin-based conjugate proteins containing the fragment B hydrophobic membrane-associating domains should be biologically active. The recombinant toxin should then be internalized and intoxicate only target cells. In addition to the minimal structural features necessary for assembly of these new receptor directed toxins, it was also clear that these recombinant fusion proteins had to be expressed in Escherichia coli in a protease-resistant form.
The widespread occurrence of single-chain ribosome-inactivating proteins (RIP) throughout the plant kingdom is an established fact. Several type I RIPs were isolated from Saponaria officinalis. However, whether these differences are due to selective action on ribosomes and not to some form of differential susceptibility of the various RIPs to the experimental conditions employed by the authors remains to be ascertained. Most of the toxin cloning studies have revealed multiple related toxin genes in plants. To date, very few studies have been reported on the genomic organization and sequence of type I RIPs and their transcripts. Two clones were selected for further analysis and were characterized by subcloning into EcoRl- digested M13 mp8 in both directions of insertion. The biosynthesis of ricin has been investigated in considerable detail, showing a close relationship between its biosynthetic pathway and that of several storage proteins and lectins accumulating in the seeds of Ricinus communis and of other plants.
Chimeric toxins are generated by removing or crippling the receptor-binding domains of bacterial or plant toxins and substituting in their place proteins or peptides that bind mammalian cells. The cell-binding protein will then dictate which cell type the chimeric toxin will bind and kill. Gene fusions are then introduced into an appropriate expression system and recombinant chimeric toxins produced. Recombinant chimeric toxins have been made with diphtheria toxin and Pseudomonas exotoxin in combination with a variety of hormones, antibodies, and other cell-binding proteins. To be cytotoxic, chimeric toxins must have binding, translocating, and enzymatic activities. Of the chimeric toxins examined, most bind 3to 10-fold less well than the corresponding native ligands (34,38,45). The chimeric toxin IL-2-PE40 is toxic for cells and cell lines that display IL-2 receptors on their surface. IX A major clinical focus for the use of chimeric toxins is the treatment of cancer and immunological disorders.
