The Ribosomal Stalk

The ribosomal stalk represents a vital element within the ribosomal GTPase associated center. Stalk structure is composed of two distinct parts - the base of the stalk and its lateral elements [68]. The stalk base is constituted by conserved ribosomal proteins uL11 and uL10, which anchor the stalk to the rRNA [69,70]. The lateral part of the stalk has multimeric architecture and is built of multi-dimeric protein elements, which are unique for bacteria and eukaryotes. In bacteria, the bL12 proteins form a dimer, which is regarded as a basic structural element, and two, three, and even four dimers can form the lateral part anchored to the ribosome through uL10 [35,71]. In eukaryotes, the P1 and P2 proteins form a dimer, and two dimers are linked to the uL10 protein, forming pentameric architecture called the P-stalk, uL10-(P1-P2)2 [37,69,70,72–75]. It should be underlined that the stalk fulfils the same function on the ribosome, irrespectively of the life-domain origin, namely participation in stimulation of GTP hydrolysis [68]; however, the bL12 and P1/P2 proteins are not evolutionarily related and are regarded as analogous proteins [76,77]. The stalk is the only structure on the ribosome composed of multiple proteins. The eukaryotic stalk architecture has a complex nature. It is constituted by two P1/P2 protein dimers; each dimer is built of two domains: an N-terminal globular domain (NTD), responsible for dimerization and anchoring the dimer to uL10, and an unstructured C-terminal domain (CTD), regarded as a functional part interacting with trGTPases [70,73,78]. Both elements are connected through a highly flexible hinge region [73,79]. The most prominent feature of the eukaryotic P1/P2 stalk proteins is the highly conserved element present at the CTD, composed of a stretch of acidic and hydrophobic amino acids (EESEESDDDMGFGLFD) and regarded as the main functional element of the stalk. This element is involved in the interaction with trGTPases and toxins such as ribosome-inactivating proteins (RIPs) [22,80–84]. A unique feature of the eukaryotic stalk is multiplication of CTDs. The conserved CTD is also found on the uL10; therefore, five CTDs are present on the stalk: four coming from two P1/P2 dimers and one from uL10. The phenomenon of CTDs multiplication was functionally coupled with the qualitative aspect of ribosome action related to maintenance of translation accuracy [85]. It was proposed that the multiple CTDs might accelerate interaction with eEF1A, which is regarded as trGTPase with the highest GTP hydrolysis turnover. Interestingly, this feature has been hijacked by RIP toxins, and it has been shown that multiplication of P1/P2 proteins increase the interaction rate of the toxin [86].

#### *2.2. Mode of Ricin Interaction with Ribosome*

It has been established that ricin inhibits translation through its ability to remove/depurinate a specific adenine base of the universally conserved SRL [87], which is a crucial part of the GAC on the ribosome [27,38,88,89]. The SRL has been found as a primary target for ricin and other RIPs, and the specificity of the interaction with eukaryotic ribosomal proteins plays a critical role in ricin catalytic activity towards SRL. As shown over two decades ago, the e fficiency of rRNA depurination in the intact ribosome is much greater than the depurination of isolated 28S rRNA. The *k*cat of ricin against naked rRNA is more than 4 orders of magnitude lower than that of rRNA constituting a part of the ribosome [90–92]. Ricin depurinates the naked 23S rRNA from *E. coli* SRL, but not the intact ribosomes from *E. coli* [91], showing at the same time extraordinary specificity towards intact eukaryotic

ribosomes [20,21,93], what underlines the role of ribosomal proteins in the process. The same applies to other related RIPs, such as Shiga toxin 1 (Stx1) [93,94], Shiga toxin 2 (Stx2) [95,96], trichosanthin (TCS) [97–99], and maize RIP [100], which specifically depurinate the SRL on the eukaryotic ribosome. In the case of ricin, the mechanistic model of molecular recognition of the ribosome assumes a double-step mechanism, involving first slow and nonspecific electrostatic-based interactions with the ribosome and then fast specific interactions based on the ribosomal stalk interplay, leading to its attack on the SRL rRNA [90]. Although the SRL is highly conserved among ribosomes in all species, the P-proteins determine the specificity of ricin and other RIPs toward eukaryotic ribosomes [12,22]. The deletion of stalk P-proteins from ribosomes greatly reduces the depurination activity and cellular sensitivity to ricin, indicating that binding to the P-stalk is a critical step in depurination of the SRL and in the toxicity of ricin [101–103]. The structural investigations provide significant insights into the mode of interaction between P-proteins and the ricin or trichosanthin (TCS), which hijacks the translational factor recruitment function of the ribosomal P-stalk to reach its target site on the ribosome [22,84]. Especially, the interaction site of P-proteins with RIPs was mapped to a short conserved 11-mer peptide, SDDDMGFGLFD, present at the CTDs of all P-proteins [98]. This interaction is required for the full activity of ricin and other RIPs, and biochemical analyses confirmed that positively charged residues, especially the cluster of arginines, play a key role [101]. As was shown in a TCS study, the interaction of TCS is primarily mediated by the electrostatic interactions of K173, R174, and K177 in the C-terminal domain of TCS with the conserved DDD residues in the CTDs of P-proteins. However, hydrophobic interactions also play a vital role in stabilization of the bilateral interplay between TCS and the conserved C-terminal peptide of P-proteins [93,99]. The tertiary structure of the catalytic subunit of ricin (RTA) with a short peptide corresponding to the last six conserved residues of the stalk proteins (GFGLFD) showed that the peptide docks into a hydrophobic pocket at the C-terminus of RTA [84,104]. The structural superposition of TCS-P-protein and RTA-P-protein complexes demonstrated that the short C-terminal peptide of P-proteins adopts distinct orientations and slightly di fferent interaction modes with the two di fferent RIPs, suggesting that the flexibility of the CTD facilitates accommodation of di fferent class of RIPs to the ribosome [84,104]. The kinetic studies showed that the P1-P2 heterodimeric conformation of P-proteins in the stalk pentamer represents an optimal binding site for RTA, where individual P-protein CTDs play non-equivalent roles with a pivotal role of P1 CTD [65]. Additionally, previous results obtained using yeas<sup>t</sup> as an experimental system showed that the two dimers, P1A-P2B and P1B-P2A, do not interact equally with RTA [102], suggesting that these dimers may have a di fferent architecture and their CTDs may not be equally accessible to external factors, such as RTA or other RIPs. The high specificity of ricin interaction with the P-stalk is also reflected by the measured dissociation constant, which is in a nanomolar range [65,80,86]. Thus, the kinetic model of RTA interaction with the ribosome and SRL depurination assumes that the toxin initially interacts with the P-protein stalk, and it allows orienting the active site of the toxin toward the SRL, which in turn places it in correct orientation for binding to the target adenine. It is also proposed that the P-stalk binding event allosterically stimulates the catalysis of ribosome depurination by RTA, explaining the extraordinary specificity of the toxin toward eukaryotic ribosomes [101].

#### **3. Toxic Action of Ricin on The Translational Process**

Ricin is composed of two subunits, RTA and RTB, covalently linked through a disulfide bond. In the form of holotoxin, it does not exhibit catalytic activity toward ribosome [105]. When RTA is separated from RTB, the cluster of arginine residues located at the interface domain between RTA and RTB is exposed to the solvent and serves as an interaction platform for P-stalk proteins. The RTA-stalk interaction stimulates the toxin to trigger its enzymatic activity by orienting the active site of RTA (opposite to the arginine interface) toward the SRL [101]. RTA is an RNA N-glycosidase (EC 3.2.2.22) that hydrolyzes the N-glycosidic bond between a specific adenine on the SRL and the sugar backbone [91]. The specificity of rRNA depurination by RTA is determined by the conformation of the topical part of the SRL loop structure, and the GAGA sequence with the prominent key adenine

base is recognized as the major element [106,107]. During the catalysis of the SRL depurination process by ricin, the conserved adenine on the tip of the sarcin–ricin loop is inserted between two Tyr residues (in the RTA catalytic center - Tyr80 and Tyr123) to form π-stacking interactions [108,109]. Additionally, the adenine position is stabilized by hydrogen binding with RTA Gly121, Val81, Glu177, and Arg180 residues [110]. It has been shown that two RTA residues, i.e. Glu177 and Arg180, play a crucial role in the hydrolysis of N-glycosidic bonds by stabilizing the transition state during catalysis of the depurination reaction (Figure 4) [111,112].

**Figure 4.** Model of sarcin–ricin loop (SRL) depurination by RTA. Removal of key A2660 residue (*E. coli* numbering) impairs the intricate interaction network responsible for stabilization of the active state of trGTPases, resulting in the inhibition of translation process. The base of A2660 is bound to the catalytic center of RTA with π-stacking interactions and its position is stabilized by hydrogen bonds with Gly121, Val81 and Arg180 (green). Glu177 and Arg180 (green) play crucial role in catalysis by transition state stabilization. Arg180 residue protonates base of A2660 causing delocalization of ring electrons. Glu177 polarizes water molecule (blue) and resultant hydroxide ion attacks positive center on ribose which leads to hydrolysis of N-glycosidic bond; prepared based on [110].

As already discussed, the structural stability of the SRL is provided mostly by the π-stacking interaction network [52,113,114]. Removal of the key adenine at the tip of the SRL may destabilize this type of interactions, thereby affecting the SRL structure stability and abolishing an extended interaction network responsible for stabilization of the active state of trGTPases (Figure 4).

It was already observed in the 1970s that ricin inhibits translation in mammalian cells [115], as confirmed in an in vitro experimental system [53,116–119]. Early analyses in in vitro protein synthesis systems have shown that the presence of ricin blocks the synthesis of polypeptides, and this was mainly associated with the elongation step of the translational cycle [116,117,120–122]. It has also been reported that ricin does not affect the synthesis of peptide bonds [53,116,123,124], but a significant inhibition of GTP hydrolysis was associated with ricin action [53,54,116,124–130]. Numerous analyses have shown that eEF2, i.e. a factor involved in the translocation event during the elongation cycle, binds less efficiently to ribosomes modified by ricin [54,125,126,128,131]. Additionally, it has also been shown that treatment of ribosomes with ricin decreased the level of GTP hydrolysis by the eEF2 factor [53,116]. Further analyses demonstrated that ricin inhibited translocation, and the effect was dependent on the eEF2 concentration used [123]. Additional evidence for the inhibitory effects of ricin on translocation was provided by applying test with the use of diphtheria toxin, showing that the ribosomes treated with both toxins mainly paused at the beginning of the mRNA [119]. It was also shown that the rate of formation of pre-initiation complexes, i.e. the attachment of the 60S subunit to the pre-initiation complex 40S, was decreased under the influence of ricin [119]. Thus, these analyses have provided evidence that the modification of the 60S subunit by ricin resulted not only in inhibition of elongation at the translocation step, but also in reduction of the initiation rate [119]. In vitro analyses on the bacterial model have confirmed the experiments performed on the eukaryotic system, showing that ribosomes lacking adenine in the topical part of SRL are unable to stimulate GTP hydrolysis by EF-G, which is a bacterial trGTPase homologous to eukaryotic eEF2 [66]. All these experiments laid

the foundation for a general notion that depurination of the SRL brings deleterious e ffects for the translational machinery, linking the toxic e ffect of ricin with blockage of protein synthesis in the cell. However, in vivo analyses have shown that there is no clear cross-correlation between the ribosome depurination, translation inhibition, and cell death [102,103], leaving the issue of ricin toxicity at the molecular level as an open question. All currently available information concerning ricin toxicity suggests that the toxic e ffect on cell metabolism has a multifactorial nature, involving induction of numerous pathways leading to cell death [25], but the molecular trigger is still obscure.
