1
Introduction and History

Fiorenzo Stirpe

Dipartimento di Medicina Specialistica, Università di Bologna, Italy

Introduction

The history of RIPs, especially the toxic ones, has been well reviewed recently.1 This present chapter will summarize the research steps that in the last 40 years have led to significant advancements in the knowledge of these proteins, of their mechanism of action, and of their possible practical applications in medicine and in agriculture.

Ribosome-inactivating proteins (RIPs), initially discovered in higher plants, have been the subject of numerous studies (reviews by Van Damme,2 Nielsen,3 Hartley,4 Girbés,5 Stirpe,6,7 Ng,8 and Puri9). More than 50 RIPs have been identified and purified, but it has become clear that they can, in some circumstances, be expressed in many plants and other organisms in which they have not been detected because of assay sensitivity or other reasons. Thus, they must have an important function to justify their persistence throughout the evolution of proteins, which are an expensive material to make. Furthermore, it is becoming more and more apparent that important uses of RIPs can be envisaged.

Identification and Distribution in Nature

The studies on the proteins that eventually were denominated ribosome-inactivating proteins (RIPs) began in Dorpat at the end of the nineteenth century when ricin, a potent toxin from the seeds of Ricinus communis (castor bean plant), was identified and isolated by Stillmark who described it in his thesis as a “ferment” (remarkably for the time!).10 Shortly afterwards, abrin, a toxin similar to ricin, was isolated from the seeds of Abrus precatorius.11

Research on these toxins was then rather scarce for a long period, until a revival of the studies on ricin and abrin was prompted by the report that these proteins were more toxic to malignant, than to normal, cells.12 The same authors found that the toxins inhibited protein synthesis by cells, a first step toward the discovery of their mechanism of action.13 Unfortunately, subsequent interest in the toxins, especially ricin, was stimulated also by the fear of a possible use for warfare or terrorist actions, as in the case of a Bulgarian journalist murdered with a micro-bullet probably containing ricin.14

Significant, pioneering progress in knowledge of the toxins was achieved by S. Olsnes, A. Pihl, and collaborators in the Institute of Biochemistry of the Norwegian Radium Hospital. They elucidated the structure of both ricin and abrin, establishing that they constituted two unequal polypeptide chains linked by a disulfide bond, an A chain which inhibited protein synthesis, and a B chain with the properties of a lectin specific for sugars with the structure of galactose. They found also that the latter chain binds to galactose residues on the surface of most cells, allowing the entry of the A chain, which exerts its toxic action.15 Furthermore, they confirmed that the toxins inhibit protein synthesis, not only in cells, but also in a cell-free system.16

The knowledge that ricin and abrin were produced by taxonomically unrelated plants, and still had a very similar structure and function, led our group to research whether other, similar, toxins were present in nature. In our laboratory this research was conducted by examining seed extracts from plants known to be toxic. On the basis of Olsnes’ observations, the extracts were tested for inhibition of protein synthesis in a cell-free system, a rabbit reticulocyte lysate, a much simpler and more rapid test to perform than to evaluate the effects on inhibition of protein synthesis by, or on viability of, cells. A number of seed extracts were screened in this way, starting from those from plants that in the old literature were reported as containing toxins similar to ricin, such as crotin and curcin.17 Much to our surprise (and for a while, disappointment!), we found that indeed in the seed extracts from many plants there were proteins that inhibited cell-free protein synthesis, but these were hardly toxic to cells.18 The meaning of these observations was elucidated when an antiviral protein (pokeweed antiviral protein, PAP) was isolated from the leaves of Phytolacca americana, that inhibited protein synthesis resembling the A chain of ricin.19, 20 Thus, it soon became clear that the proteins present in our and other plant extracts also had antiviral activity against both plant and animal viruses (review by Kaur et al.21) and were similar to the A chain of ricin. This led to classifying the ribosome-inactivating proteins into two types: type 2, consisting of an A and a B chain; and type 1, consisting of an A-like single polypeptide chain with enzymatic activity.22 Type 3 RIPs, consisting of an active moiety linked to a peptidic chain with unknown function, have also been described (reviewed in Peumans et al.23). A new classification of RIPs is proposed in Chapter 2. Another form of PAP was identified in pokeweed summer leaves,24 and another in pokeweed seeds.22 On the whole, these results indicated (i) that RIPs could be present in the same plant as isoforms, which subsequent research showed could be several, and (ii) that in some plants RIPs could be in different organs, and not only in seeds as ricin and abrin. It was found also that trichosanthin, a protein from Trichosanthes kirilowii and other plant proteins used in China to induce abortion, were RIPs25 (reviews by Ng et al.26 and Chapter 4).

The structure of ricin, and subsequently of other RIPs of both types, was elucidated by crystallographic studies (review by Robertus and Monzingo27 and Chapter 8), which led to the identification of active and sugar-binding sites.

The search for other toxins similar to ricin continued, and modeccin, a highly toxic protein described as a “toxalbumin”28 and already at that time suspected to be similar to ricin, was purified and characterized almost at the same time in Oslo29 and in Bologna.30 A galactose-specific haemagglutinating lectin purified from mistletoe31 turned out to be a type 2 RIP.32 Other type 2 RIPs were found in several other plants,5 including some belonging to the genus Adenia.33 From Adenia stenodactyla stenodactylin was isolated, probably the most potent toxin of plant origin, with an LD50 <1 µg/kg of body weight.34 Lists of RIPs are given in several reviews.2, 5, 6, 9, 35

From the number of RIPs isolated, it seemed that type 1 RIPs were very common in plants, whilst type 2 RIPs were rather scarce. This belief was somewhat changed: two barely toxic tetrameric agglutinins consisting of two A and two B chains are present in the seeds of Ricinus communis and Abrus precatiorius;36 from Sambucus nigra a lectin, nigrin b, was purified and had structure and biochemical properties similar to those of type 2 RIPs, and still a very low toxicity37 (Chapter 5), and other non-toxic type 2 RIPs have been isolated (see lists in2, 5, 6), so it is possible that other, possibly many, similar proteins are present in nature.

An evolutionary relationship between single-chain and two-chain RIPs was elaborated by Ready et al.38 and the precursors in the synthesis of ricin were identified.39

Ribosome-inactivating proteins seemed particularly frequent in plants belonging to some families, namely Cucurbitaceae, Euphorbiaceae, and Caryophyllaceae. However, most results came from a search aimed at finding proteins present at a level sufficient to allow easy purification of significant quantities, and consequently the search was concentrated on plants taxonomically close to those already known to contain RIPs. Furthermore, a number of plants were not considered to contain a RIP when their extracts had protein synthesis inhibitory activity below a pre-set arbitrary level: thus it is possible that in many other plants RIPs are present, although at a low level or in an inactive form. This notion was supported by the study of plant genomes (Chapter 9) and by the observation that the expression of RIPs is stimulated or even appears under several conditions of stress.40 However, the hypothesis that RIPs could be ubiquitous in plants was dismissed after it was reported that in the complete genome of Arabidopsis taliana there are no sequences encoding for RIPs.2

There are reports, some of them controversial, that RIPs or RIP-like proteins are produced by organisms other than higher plants, namely algae, fungi, and microorganisms.5 Among bacteria, Shighella shigae and strains of Escherichia coli produce, respectively, Shiga toxin and Shiga-like toxins which have RIP activity (reviews by Reyes et al.41 and Chapter 7), and the Streptomyces coelicolor genome appears to encode a type 1 ribosome-inactivating protein41 (see Chapter 9).

Mechanism of Action

Our laboratory also became involved in the study of the mechanism of action of ricin and related toxins. The fact that RIPs inhibited cell-free protein synthesis led to an investigation on their effect in a system of purified ribosomes with necessary enzymes and cofactors. The experiments by L. Montanaro and S. Sperti established that ricin damaged ribosomes,42 more precisely their larger subunit,43 rendering ribosomes incapable of binding Elongation Factor 2.44 This effect was irreversible and occurred at a less-than-equimolar toxin : ribosome ratio, indicating that the toxins acted catalytically, that is enzymatically. The effect of the enzymatic activity was discovered by Endo and...