Advertisement

Single-domain antibodies neutralize ricin toxin intracellularly by blocking access to ribosomal P-stalk proteins

Open AccessPublished:February 16, 2022DOI:https://doi.org/10.1016/j.jbc.2022.101742
      During ricin intoxication in mammalian cells, ricin's enzymatic (RTA) and binding (RTB) subunits disassociate in the endoplasmic reticulum. RTA is then translocated into the cytoplasm where, by virtue of its ability to depurinate a conserved residue within the sarcin–ricin loop (SRL) of 28S rRNA, it functions as a ribosome-inactivating protein. It has been proposed that recruitment of RTA to the SRL is facilitated by ribosomal P-stalk proteins, whose C-terminal domains interact with a cavity on RTA normally masked by RTB; however, evidence that this interaction is critical for RTA activity within cells is lacking. Here, we characterized a collection of single-domain antibodies (VHHs) whose epitopes overlap with the P-stalk binding pocket on RTA. The crystal structures of three such VHHs (V9E1, V9F9, and V9B2) in complex with RTA revealed not only occlusion of the ribosomal P-stalk binding pocket but also structural mimicry of C-terminal domain peptides by complementarity-determining region 3. In vitro assays confirmed that these VHHs block RTA–P-stalk peptide interactions and protect ribosomes from depurination. Moreover, when expressed as “intrabodies,” these VHHs rendered cells resistant to ricin intoxication. One VHH (V9F6), whose epitope was structurally determined to be immediately adjacent to the P-stalk binding pocket, was unable to neutralize ricin within cells or protect ribosomes from RTA in vitro. These findings are consistent with the recruitment of RTA to the SRL by ribosomal P-stalk proteins as a requisite event in ricin-induced ribosome inactivation.

      Keywords

      Abbreviations:

      AUC (area under the curve), BSA–P2C11 (P2C11 peptide–bovine serum albumin BSA conjugate), CDR (complementarity-determining region), EF (elongation factor), ER (endoplasmic reticulum), HRP (horseradish peroxidase), IVT (in vitro translation), LNP (lipid nanoparticle), mAb (monoclonal antibody), PBST (PBS containing 0.1% Tween-20), PDB (Protein Data Bank), RIP (ribosome-inactivating protein), RTA (ricin toxin A subunit), RTB (ricin toxin B subunit), Sc (shape complementary), SPR (surface plasmon resonance), SRL (sarcin–ricin loop), TMB (3,3′,5,5′-tetramethylbenzidine), TNA (toxin-neutralizing activity)
      Ricin toxin is the archetype of the large and diverse family of medically important plant and bacterial ribosome-inactivating proteins (RIPs). A byproduct of castor beans (Ricinus communis), ricin toxin consists of two equal sized subunits, ricin toxin A subunit (RTA) and ricin toxin B subunit (RTB), that associate through electrostatic and covalent interactions (
      • Montfort W.
      • Villafranca J.E.
      • Monzingo A.F.
      • Ernst S.R.
      • Katzin B.
      • Rutenber E.
      • Xuong N.H.
      • Hamlin R.
      • Robertus J.D.
      The three-dimensional structure of ricin at 2.8 A.
      ,
      • Olsnes S.
      • Refsnes K.
      • Pihl A.
      Mechanism of action of the toxic lectins abrin and ricin.
      ). RTB is a bivalent Gal/GalNAc-specific lectin that traffics RTA from the plasma membrane to the endoplasmic reticulum (ER) (
      • Sowa-Rogozinska N.
      • Sominka H.
      • Nowakowska-Golacka J.
      • Sandvig K.
      • Slominska-Wojewodzka M.
      Intracellular transport and cytotoxicity of the protein toxin ricin.
      ). Within the ER, the interchain disulfide bond linking RTA and RTB is reduced by protein disulfide isomerase, and RTA is then retrotranslocated in an unfolded state across the ER membrane and into the cytoplasm (
      • Bellisola G.
      • Fracasso G.
      • Ippoliti R.
      • Menestrina G.
      • Rosen A.
      • Solda S.
      • Udali S.
      • Tomazzolli R.
      • Tridente G.
      • Colombatti M.
      Reductive activation of ricin and ricin A-chain immunotoxins by protein disulfide isomerase and thioredoxin reductase.
      ,
      • Nowakowska-Gołacka J.
      • Sominka H.
      • Sowa-Rogozińska N.
      • Słomińska-Wojewódzka M.
      Toxins utilize the endoplasmic reticulum-associated protein degradation pathway in their intoxication process.
      ,
      • Spooner R.A.
      • Watson P.D.
      • Marsden C.J.
      • Smith D.C.
      • Moore K.A.
      • Cook J.P.
      • Lord J.M.
      • Roberts L.M.
      Protein disulphide-isomerase reduces ricin to its A and B chains in the endoplasmic reticulum.
      ,
      • Wesche J.
      • Rapak A.
      • Olsnes S.
      Dependence of ricin toxicity on translocation of the toxin A-chain from the endoplasmic reticulum to the cytosol.
      ). From that point forward, RTA functions as a RIP, triggering the ribotoxic stress response and programmed cell death pathways (
      • Endo Y.
      • Mitsui K.
      • Motizuki M.
      • Tsurugi K.
      The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins.
      ,
      • Endo Y.
      • Tsurugi K.
      RNA N-glycosidase activity of ricin A-chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes.
      ,
      • Iordanov M.S.
      • Pribnow D.
      • Magun J.L.
      • Dinh T.H.
      • Pearson J.A.
      • Chen S.L.
      • Magun B.E.
      Ribotoxic stress response: Activation of the stress-activated protein kinase JNK1 by inhibitors of the peptidyl transferase reaction and by sequence-specific RNA damage to the alpha-sarcin/ricin loop in the 28S rRNA.
      ,
      • Vind A.C.
      • Genzor A.V.
      • Bekker-Jensen S.
      Ribosomal stress-surveillance: Three pathways is a magic number.
      ).
      The mode of action of RTA in inactivating mammalian ribosomes is well established. RTA catalyzes the hydrolysis of the N-glycosidic bond of a single adenine base within the sarcin–ricin loop (SRL) of the 28S rRNA, a conserved hairpin-like structure that interacts with eukaryotic elongation factors (EFs) 1 and 2, during ribosome translocation (
      • Montfort W.
      • Villafranca J.E.
      • Monzingo A.F.
      • Ernst S.R.
      • Katzin B.
      • Rutenber E.
      • Xuong N.H.
      • Hamlin R.
      • Robertus J.D.
      The three-dimensional structure of ricin at 2.8 A.
      ,
      • Olsnes S.
      • Refsnes K.
      • Pihl A.
      Mechanism of action of the toxic lectins abrin and ricin.
      ,
      • Endo Y.
      • Tsurugi K.
      RNA N-glycosidase activity of ricin A-chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes.
      ,
      • Carbone C.E.
      • Loveland A.B.
      • Gamper Jr., H.B.
      • Hou Y.M.
      • Demo G.
      • Korostelev A.A.
      Time-resolved cryo-EM visualizes ribosomal translocation with EF-G and GTP.
      ,
      • Endo Y.
      • Tsurugi K.
      The RNA N-glycosidase activity of ricin A-chain. The characteristics of the enzymatic activity of ricin A-chain with ribosomes and with rRNA.
      ,
      • Ready M.P.
      • Kim Y.
      • Robertus J.D.
      Site-directed mutagenesis of ricin A-chain and implications for the mechanism of action.
      ,
      • Moazed D.
      • Robertson J.M.
      • Noller H.F.
      Interaction of elongation factors EF-G and EF-Tu with a conserved loop in 23S RNA.
      ,
      • Monzingo A.F.
      • Robertus J.D.
      X-ray analysis of substrate analogs in the ricin A-chain active site.
      ). The active site of RTA consists of a large solvent-exposed cleft on one face of the molecule that accommodates adenine through a π-stacking network and hydrogen bonding (
      • Monzingo A.F.
      • Robertus J.D.
      X-ray analysis of substrate analogs in the ricin A-chain active site.
      ,
      • Mlsna D.
      • Monzingo A.F.
      • Katzin B.J.
      • Ernst S.
      • Robertus J.D.
      Structure of recombinant ricin A chain at 2.3 A.
      ). Site-directed mutagenesis has identified the key residues within the active site that are associated with RNA N-glycosidase activity (
      • Robertus J.D.
      • Yan X.
      • Ernst S.
      • Monzingo A.
      • Worley S.
      • Day P.
      • Hollis T.
      • Svinth M.
      Structural analysis of ricin and implications for inhibitor design.
      ). Namely, Tyr-80 and Tyr-123 stabilize the incoming adenine base via a π-stacking network, whereas Arg-180 is involved in protonation of the adenine-leaving group. Glu-177 participates in the ultimate hydrolysis of the N-glycosidic bond. There is neither evidence that residues outside the active site properly contribute to SRL depurination nor are there any reports that RTA enzyme kinetics are subject to allostery.
      That said, there is a growing body of evidence that SRL depurination is the second event in a two-step interaction between RTA and the ribosome (
      • Grela P.
      • Szajwaj M.
      • Horbowicz-Drozdzal P.
      • Tchorzewski M.
      How ricin damages the ribosome.
      ,
      • May K.L.
      • Yan Q.
      • Tumer N.E.
      Targeting ricin to the ribosome.
      ). According to the two-step model, RTA first associates with the C terminus of the acidic P-stalk proteins (RPLP1 and RPLP2; “P1” and “P2”) of the 60S ribosome before being guided to the SRL (
      • May K.L.
      • Yan Q.
      • Tumer N.E.
      Targeting ricin to the ribosome.
      ,
      • Li X.P.
      • Chiou J.C.
      • Remacha M.
      • Ballesta J.P.
      • Tumer N.E.
      A two-step binding model proposed for the electrostatic interactions of ricin a chain with ribosomes.
      ,
      • May K.L.
      • Li X.P.
      • Martinez-Azorin F.
      • Ballesta J.P.
      • Grela P.
      • Tchorzewski M.
      • Tumer N.E.
      The P1/P2 proteins of the human ribosomal stalk are required for ribosome binding and depurination by ricin in human cells.
      ). Targeted mutagenesis identified residues along the face of RTA normally occluded by RTB as being critical for ribosomal recruitment, with Arg235 being the most important (
      • Jetzt A.E.
      • Cheng J.S.
      • Li X.P.
      • Tumer N.E.
      • Cohick W.S.
      A relatively low level of ribosome depurination by mutant forms of ricin toxin A chain can trigger protein synthesis inhibition, cell signaling and apoptosis in mammalian cells.
      ,
      • Zhou Y.
      • Li X.P.
      • Chen B.Y.
      • Tumer N.E.
      Ricin uses arginine 235 as an anchor residue to bind to P-proteins of the ribosomal stalk.
      ). The structures of RTA in complex with C-terminal peptides of the P2 protein (“P2C11” and “P2C10”) were solved recently (
      • Fan X.
      • Zhu Y.
      • Wang C.
      • Niu L.
      • Teng M.
      • Li X.
      Structural insights into the interaction of the ribosomal P stalk protein P2 with a type II ribosome-inactivating protein ricin.
      ,
      • Shi W.W.
      • Mak A.N.
      • Wong K.B.
      • Shaw P.C.
      Structures and ribosomal interaction of ribosome-inactivating proteins.
      ). The structures revealed a mostly hydrophobic pocket on RTA formed by residues Tyr183, Arg235, Phe240, and Ile251 as the target of the P2 peptides. Short peptides corresponding to the P2 C terminus bind RTA with micromolar affinities (∼200–450 μM) and prevent RTA depurination of intact ribosomes, without interfering with the active site (
      • Li X.P.
      • Kahn J.N.
      • Tumer N.E.
      Peptide mimics of the ribosomal P stalk inhibit the activity of ricin A chain by preventing ribosome binding.
      ). Thus, the large hydrophobic face of RTA normally occluded by RTB is apparently “repurposed” in the cytoplasm to assist in recruitment to the ribosome.
      In this study, we report the identification and characterization of 11 alpaca-derived single-domain antibodies (VHHs) that recognize distinct epitopes on the hydrophobic surface of RTA normally occluded by RTB. We solved the cocrystal structures of four VHHs (V9E1, V9F9, V9B2, and V9F6) bound to RTA, revealing that three of the four antibodies not only physically occlude the ribosomal P-stalk binding pocket on RTA but also mimic key P-stalk peptide interactions. The fourth VHH, V9F6, recognizes an epitope adjacent to but not overlapping the ribosomal P-stalk binding pocket. Only the VHHs that directly obstruct the ribosomal P-stalk binding pocket neutralized RTA in cell-free translation assays and blocked RTA–P2 peptide interactions. Finally, we show that transient expression of V9E1 and other VHHs as intracellular antibodies (“intrabodies”) renders Vero cells resistant to ricin toxin challenge to levels equivalent to that previously achieved with VHHs targeting the active site of RTAs (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). Our findings support a model in which the interaction of RTA with ribosomal P-stalk proteins is a prerequisite to SRL depurination and raise the possibility of developing therapeutics that target two distinct sites on RTA.

      Results

      Identification of a collection of VHHs that target the hydrophobic face of RTA normally occluded by RTB

      As part of a VHH phage-display screening strategy targeting diverse epitopes on ricin toxin's A and B subunits (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ,
      • Poon A.Y.
      • Vance D.J.
      • Rong Y.
      • Ehrbar D.
      • Mantis N.J.
      A supercluster of neutralizing epitopes at the interface of ricin's enzymatic (RTA) and binding (RTB) subunits.
      ,
      • Rudolph M.J.
      • Poon A.Y.
      • Kavaliauskiene S.
      • Myrann A.G.
      • Reynolds-Peterson C.
      • Davis S.A.
      • Sandvig K.
      • Vance D.J.
      • Mantis N.J.
      Structural analysis of toxin-neutralizing, single-domain antibodies that bridge ricin's A-B subunit interface.
      ,
      • Vance D.J.
      • Tremblay J.M.
      • Rong Y.
      • Angalakurthi S.K.
      • Volkin D.B.
      • Middaugh C.R.
      • Weis D.D.
      • Shoemaker C.B.
      • Mantis N.J.
      High-resolution epitope positioning of a large collection of neutralizing and nonneutralizing single-domain antibodies on the enzymatic and binding subunits of ricin toxin.
      ), we identified 11 VHHs that recognized RTA but not ricin holotoxin (Table 1). The VHHs grouped into five families based on complementarity-determining region 3 (CDR3) amino acid sequences (Table 1). The binding affinities of the 11 VHHs for RTA ranged from potent to ultrapotent, as determined by surface plasmon resonance (SPR) (Table 1 and Fig. S1). VHHs in families 1, 2, and 5 (n = 4 total) had equilibrium dissociation constants (KD) of ∼0.1 nM (0.073–0.22 nM; Table 1) with on (ka) and off (kd) rates ranging from 3.5 × 106 to 6.5 × 106 M−1 s1 and 2.2 × 10−4 to 1.3 × 10−3 s−1, respectively. Six of the seven VHHs in families 3 and 4 had ultrahigh affinities for RTA with estimated KDs of <0.03 nM, with on rates (ka) ranging from 3.2 × 106 to 6.1 × 108 M−1 s−1 and off rates (kd) too slow to be accurately measured by SPR with no dissociation observed up to 10 min after VHH–RTA complex formation. The seventh VHH (V9D12) had a KD of ∼0.07 nM.
      Table 1Summary of V9 VHH families identified in this study
      FamilyVHHCompetitionKD (nM)CDR3GenBank
      SyH7JD4
      1V9E10.11AADRDRLPSAITYEYNYMW389192
      2V9E5+0.075AGDRDTTAQAMGLFGARGDYMW389193
      3V9B2+<0.03ATEEVCTLGIFGHGPDDYMW389188
      4V9B8+<0.03AAADPLPLICTEADEYNYMW389189
      4V9D12+0.070AAADPLPLICTEPDEYTYMW389191
      4V9E9+<0.03AAADPLPLICTEADEYNYMW389194
      4V9F9+<0.03AAADPLPLVCTWGDEYDYMW389196
      4V9G11+<0.03AAADPLPLLCTEADEYDYMW389198
      4V9H10+<0.03AAADPLPLICTEADEYNYMW389199
      5V9F6+0.22AAGSYAAILYAPSYMW389195
      5V9D5+0.073AAGSYAAILYAPSYMW389190
      Underline indicates VHHs whose structures were solved in complex with RTA in this study.
      The reactivity of the VHHs with RTA but not ricin holotoxin suggested that the V9 series of VHHs might recognize the face of RTA that is normally occluded by RTB. In support of this notion, the VHHs failed to recognize a derivative of RTA (RVEc) that lacks the folding domain (residues 200–267) that normally associates with RTB (Fig. S2). Moreover, we performed competition ELISAs with two monoclonal antibodies (mAbs) (SyH7 and JD4) that recognize epitopes on opposite sides of the RTA–RTB interface (Table 1 and Fig. S3) (
      • Toth R.T.I.
      • Angalakurthi S.K.
      • Van Slyke G.
      • Vance D.J.
      • Hickey J.M.
      • Joshi S.B.
      • Middaugh C.R.
      • Volkin D.B.
      • Weis D.D.
      • Mantis N.J.
      High-definition mapping of four spatially distinct neutralizing epitope clusters on RiVax, a candidate ricin toxin subunit vaccine.
      ). V9E1, the sole VHH in family 1, was not competed by either SyH7 or JD4, whereas VHHs in families 2 (n = 1), 3 (n = 1), and 4 (n = 6) were competed by SyH7 but not JD4. The two VHHs in family 5 were competed by JD4 but not SyH7. Thus, the five VHH families fall within three distinct competition groups that likely represent spatially distinct epitopes on the “underside” of RTA that is exposed only when RTA is liberated from RTB and translocated, across the ER membrane into the cell cytoplasm.
      We next evaluated the 11 VHHs for toxin-neutralizing activity (TNA) in a cytotoxicity assay in which ricin was mixed with each VHH across a range of antibody concentrations (0.01–1000 nM) and applied to Vero cells for 2 h. The cells were then washed to remove unbound antibody, and cell viability was measured 48 h later. As anticipated, none of the V9 VHHs had any measurable TNA in this assay, even at >1000-fold molar excess over ricin. In contrast, V5E1, a previously described VHH with potent TNA, neutralized ricin with an IC50 of ∼5 nM (Fig. S2) (
      • Rudolph M.J.
      • Vance D.J.
      • Cassidy M.S.
      • Rong Y.
      • Mantis N.J.
      Structural analysis of single domain antibodies bound to a second neutralizing hot spot on ricin toxin's enzymatic subunit.
      ). We conclude that the V9 VHHs lack TNA in a conventional Vero cell cytotoxicity assay.

      Structural analysis of VHH–RTA complexes

      We employed X-ray crystallography to identify the epitopes on RTA recognized by representative VHH family members. Toward this end, the structures of four VHHs (V9E1, V9B2, V9F9, and V9F6) complexed with RTA were solved at resolutions ranging from 1.3 to 2.3 Å (Table 2 and S1 and S2). Each VHH assumed a typical immunoglobulin fold that consisted of nine β-strands arranged in two β-sheets comprised of five and four β-strands, with the exception of V9E1 as the first β-strand in V9E1's four β-strand sheet formed a loop instead of a β-strand. CDRs 1, 2, and 3 were located on one face of each VHH molecule, and each VHH antibody contained two, three, or four 310 helices (Fig. 1). In all four VHH–RTA complexes, CDR elements 1 to 3 made contact with RTA, although the interaction was invariably dominated by CDR3. VHHs V9B2, V9F6, and V9F9 exhibited the canonical disulfide bond between Cys-22 and Cys-96 (Cys-22 and Cys-95 in V9F9) linking FR1 and FR3 (
      • Muyldermans S.
      Applications of nanobodies.
      ). The analogous disulfide bond was absent in V9E1, despite the presence of the requisite cysteine residues. V9B2 and V9F9 each had a second less common disulfide bridge between CDR2 with CDR3: residues Cys-50 and Cys-102 in V9B2 and Cys-49 and Cys-105 in V9F9. The structure of RTA in each complex consisted of seven α-helices (A–G), up to three 310 helices (310a–310c), and 10 β-strands (a–j). Specific RTA secondary structure annotations are shown in Fig. S4 for reference.
      Table 2VHH–RTA binding data and interface information
      VHHKd
      Nanometer, determined in in vitro translation assay.
      EC50
      Nanometer, determined in in vitro translation assay.
      H-bonds
      Hydrogen bonds.
      SB
      Salt bridges.
      SC
      Shape complementarity.
      Total BSA
      Buried surface area.
      TotalCDR1/2/3
      V9B2<0.0343.021/0/150.721897
      V9F9<0.0319.151/3/170.70 (0.73
      Second VHH–RTA complex in crystallographic asymmetric unit.
      )
      2028 (1994
      Second VHH–RTA complex in crystallographic asymmetric unit.
      )
      V9E10.114.090/1/560.611864
      V9F60.221.9115/2/400.731130
      a Nanometer, determined in in vitro translation assay.
      b Hydrogen bonds.
      c Salt bridges.
      d Shape complementarity.
      e Buried surface area.
      f Second VHH–RTA complex in crystallographic asymmetric unit.
      Figure thumbnail gr1
      Figure 1Structures of VHH–RTA complexes. Structures of RTA (green) in complex with (A) V9B2, (B) V9E1, (C) V9F9, and (D) V9F6 depicted as ribbon diagrams. Each VHH is colored cyan. RTA, ricin toxin subunit A.
      Among the four VHH–RTA structures solved, V9B2–RTA and V9E1–RTA were most similar to each other, as evidenced by an RMSD of 1.8 Å over 352 Cα atoms. By comparison, V9B2–RTA had an RMSD of 2.0 Å over 347 Cα atoms with V9F9–RTA, whereas V9E1–RTA and V9F9–RTA had an RMSD of 2.0 Å over 305 Cα atoms (Fig. S5). The most dissimilar structure was V9F6–RTA as V9F6 principally interacted with a structurally distinct RTA epitope. No conformational changes were evident in RTA when bound by antibody, as RTA in all four VHH–RTA complexes had RMSD values of 0.4 to 0.6 Å when all RTA Cα atoms were superimposed on RTA (Protein Data Bank [PDB] ID: 1RTC).

      Occlusion of ribosomal P-stalk binding site of RTA

      Three of the four VHHs (V9E1, V9B2, and V9F9), representing families 1, 3, and 4, recognized epitopes that overlap with the ribosomal P-stalk binding pocket of RTA (Figs. 2 and S5, A and B). Specifically, V9B2, V9E1, and V9F9 engaged RTA α-helix F (residues 183–194), β-strand i and j, and loop i–j (residues 232–236 and residues 240, 242, and 244) as well as helix 310c (residues 246–249). V9E1 in addition contacts loop G–i on RTA (residue 223). The three VHHs make slightly different contacts outside the P-stalk binding pocket, including with RTA α-helix A (residues 18, 19, 22–23, and 26), loop b–c (residue 41), loop E–F (residue 182), loop F–g (residue 196), α-helix G (residues 203 and 207), and several residues in helix 310c (residues 250–251), and immediately C-terminal to helix 310c (residue 253).
      Figure thumbnail gr2
      Figure 2Relative localization of V9F6, V9B2, V9E1, and V9F9 epitopes on RTA. Molecular surface of RTA (gray) with the red surface highlighting the V9F6 epitope and the green surface depicting the combined epitopes of V9B2, V9E1, and V9F9. The blue surface depicts the overlapping epitopes of all four VHHs. The P-stalk peptide highlighting the P-stalk binding site is drawn as gray sticks with carbon atoms gray, nitrogen atoms blue, and oxygen atoms red. RTA, ricin toxin subunit A.
      V9F6, in contrast, does not contact the P-stalk protein binding site (Figs. 2 and S5C). The epitope of V9F6 principally involves the C-terminal portion of α-helix A (residues 26, 29–30), loop–b (residues 34–37), β-strand b (residues 38–39), loop b–c (residues 40–41), β-strand c (residue 43), loop E–F (residue 182), and two residues in helix 310c (residues 249–250) as well as residues immediately proximal to helix 310c (residues 253, 261–262). That said, V9F6 does share a few contact points with the epitopes of V9B2, V9E1, and V9F9, including RTA α-helix A (residue 26), loop b–c (residue 41), loop E–F (residue 182), and residues 246, 249 to 250 within helix 310c, along with residue 253 located immediately C-terminal to helix 310c (Fig. 2).

      Detailed analysis of VHH interactions with the ribosomal P-stalk binding pocket of RTAs

      The interface between V9B2 and RTA buries 1897 Å2 and consists of two hydrogen bonds and five salt bridges (Table 2). The two hydrogen bonds form between the carbonyl oxygen atoms of Asp-31 in CDR1 and Arg-193 in RTA, along with the carbonyl oxygen atom from Phe-107 in CDR3 and Arg-235 of RTA. The five salt bridges occur between CDR1 Asp-31 and RTA Arg-193, CDR2 Arg-57 with RTA Glu-41, CDR3 Asp-100 with RTA Arg-189, and Arg-193 along with CDR3 Asp-112 with RTA-Arg196 (Fig. 3A). V9B2s CDR3 residue Phe-107 also forms a key π-stacking interaction between RTA residues Tyr-183 and Phe-240. This interaction mimics the π-stacking interaction that occurs between RTA residues Tyr-183 and Phe-240 with Phe-10 in the ribosomal P-stalk peptide (sequence SDDDMGFGLFD), as revealed in the RTA–P-stalk peptide crystal structure (Figs. 4A and S6B) (
      • Shi W.W.
      • Tang Y.S.
      • Sze S.Y.
      • Zhu Z.N.
      • Wong K.B.
      • Shaw P.C.
      Crystal structure of ribosome-inactivating protein ricin A chain in complex with the C-terminal peptide of the ribosomal stalk protein P2.
      ). Further similarities include V9B2 CDR3 residue Ile-106 that burrows into a hydrophobic patch on RTA lined by Val-242, Ile-247, and Leu-248, in a similar fashion to Leu-9 within the P-stalk peptide in complex with RTA (Figs. 4A and S6A). V9B2 has a shape complementary (Sc) score of 0.72 with RTA.
      Figure thumbnail gr3
      Figure 3Close-up of the VHH interactions with RTA. Zoom in of the interface between the ribbon diagrams of RTA (green) in complex with (A) V9B2, (B) V9E1, (C) V9F9, and (D) V9F6 depicting key VHH contacts with RTA. Interacting VHH and RTA residues are drawn as sticks and color coordinated to their respective main-chain color with nitrogen atoms blue and oxygen atoms red. Hydrogen bonds and salt bridges are represented as red dashes. RTA, ricin toxin subunit A.
      Figure thumbnail gr4
      Figure 4VHH interactions with the P-stalk binding site. Close-up view of the ribbon diagrams of RTA (green) in complex with (A) V9B2, (B) V9E1, (C) V9F9 depicting key VHH contacts with the P-stalk peptide, and (D) V9F6 highlighting the distant position of V9F6 relative to P-stalk binding site on RTA. Interacting VHH and RTA residues are drawn as sticks and color coordinated to their respective main-chain color with nitrogen atoms blue and oxygen atoms red. The P-stalk peptide is drawn as gray sticks with carbon atoms gray, nitrogen atoms blue, and oxygen atoms red. Hydrogen bonds and salt bridges are represented as red dashes. RTA, ricin toxin subunit A.
      The interface between V9E1 and RTA buries 1864 Å2 and involves nine hydrogen bonds and six salt bridges (Table 2). Two key hydrogen bonds are present between CDR3 residues Ile-106 and Tyr-108 with RTA residues Arg-234 and Arg-235, respectively. FR residue Glu-43 and CDR3 residue Glu-109 established two salt bridges with RTA residues Arg-234 and Arg-235, respectively (Fig. 3B). Arg-234 and Arg-235 are significant in that they form salt bridges with the C-terminal Asp residue in the P-stalk protein when bound to RTA (PDB ID: 5GU4) (Fig. S6C). V9E1 also inserts CDR3 residue Ile-106 between RTA residues Tyr-183 and Phe-240 similarly to P-stalk peptide residue Phe-10 (Fig. 4B). V9E1 has an Sc score of 0.62 with RTA.
      The interface between V9F9 and RTA buries 2028 Å2 (1994 Å2 for the second V9F9–RTA complex within the crystallographic asymmetric unit) and consists of five hydrogen bonds and seven salt bridges, with contributions from each CDR. V9F9 CDR1 residue Asp-29 forms a salt bridge with RTA residue Arg-193, CDR2 residue Ser-53 H-bonded with RTA residue Glu-185, and CDR3 residue Asp-99 H-bonded with RTA residue Thr-190. Two salt bridges were also evident between V9F9 CDR3 residue Glu-110 and RTA residues Arg-234 and Arg-235 (Fig. 3C). As noted previously, Arg-234 and Arg-235 form key contacts with the P-stalk peptide residue Asp-11 (Fig. S6C). Finally, V9F9 CDR3 residue Leu-103 also mimics P-stalk peptide residue Phe-10 inserting between RTA residues Tyr-183 and Phe-240 (Fig. 4C). V9F9 has an Sc score of 0.70 with RTA (0.73 for the second V9F9–RTA complex within the crystallographic asymmetric unit).
      V9F6 established a total of 11 hydrogen bonds between CDR1, CDR2, and CDR3 residues in V9F6 with RTA. Some of the key interactions included hydrogen bonding between the amide proton in CDR1 Arg-27 with the main-chain carbonyl oxygen of Ala-36 in RTA, the amide proton in CDR2 Leu-53 with Glu-41 in RTA, the amide proton in CDR3 Tyr-101 with the main-chain carbonyl oxygen of Glu-41 in RTA, and Tyr-110 with the main-chain carbonyl oxygen of Ala-36 in RTA (Fig. 3D). V9F6 does not contact the P-stalk protein binding site on RTA (Figs. 2 and S5C). In fact, the closest V9F6 residue Tyr-101 is 6.0 Å away from the N-terminal amine nitrogen atom of the superpositioned P-stalk peptide bound to RTA (Fig. 4D). Overall, the focal point of V9F6’s interaction with RTA centers on β-strands b and c, loop b and c, and α-helix A with the rest of the antibody angling away from V9B2, V9E1, and V9F9 when bound to RTA (Fig. S5). V9F6 has an Sc score of 0.73.

      Inhibition of RTA's RIP activity in vitro

      The overlap between V9B2, V9F9, and V9E1’s epitopes with the ribosomal P-stalk binding site on RTA prompted us to examine whether the VHHs interfere with the RIP activity of RTA. We were particularly interested in the comparison between V9E1 and V9F6 because the two VHHs have similar binding affinities for RTA, but V9E1 blocks the P-stalk binding site on RTA, whereas V9F6 does not.
      To address this experimentally, the two VHHs were assessed for their ability to neutralize RTA in an established in vitro translation (IVT) assay (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). Specifically, V9E1 and V9F6 were incubated with increasing amounts of RTA (0.02–13.6 nM) and then added to a master mixture containing mRNA template encoding firefly luciferase (Fig. 5A). In the absence of RTA, the reaction yielded >106 relative light units, whereas the addition of RTA resulted in a dose-dependent reduction in translation, as evidenced by >1000-fold reduction in luciferase activity (Fig. S7). The addition V9F6 to the IVT assay had little to no effect on the RIP activity of RTA. In contrast, the addition of V9E1 (13.6 nM) resulted in a significant restoration of relative light unit, as compared with RTA treatment alone (Fig. 5B). In fact, V9E1’s activity was similar to that of V2A11, a high-affinity VHH (KD = 0.31 nM) that occupies the active site of RTA (data not shown) (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). We conclude that V9E1 can indeed neutralize the RIP activity in vitro of RTA.
      Figure thumbnail gr5
      Figure 5Inhibition of RIP activity by VHHs targeting the ribosomal P-stalk binding pocket of RTA. A, in vitro translation assays were performed by mixing 13.6 nM VHH with serial dilutions of RTA and then adding to a mixture containing rabbit reticulocyte lysate and luciferase mRNA. B, a standard curve (open circles) was established absent VHH addition. Shown are the results and nonlinear regression analyses of each VHH (filled circles or squares) normalized to positive controls without RTA added (100%). Data represent the mean ± SD of three (V9F6) or nine (V9E1) biological replicates. Significance was determined at each RTA concentration by two-way ANOVA. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. RIP, ribosome-inactivating protein; RTA, ricin toxin subunit A.

      Inhibition of RTA–P2 stalk recognition in vitro

      To test whether V9E1 interferes with the ability of RTA to associate with ribosome P-stalk proteins, we developed a solid phase binding assay in which microtiter plates were coated with a P2C11 peptide–bovine serum albumin BSA conjugate (BSA–P2C11) and then probed with RTA in the absence or the presence of antibody (Fig. 6). The addition of V9E1 resulted in a dose-dependent reduction of RTA attachment to BSA–P2C11 (Fig. 6). In fact, V9E1’s EC50 corresponded to a 2:1 RTA:VHH stoichiometry, whereas RTA binding to BSA–P2C11 was fully abolished at a 1:1 RTA:VHH stoichiometry. In contrast, V9F6 had no effect on the ability of RTA to interact with BSA–P2C11, even when the antibody was in molar excess. These results are consistent with predictions from the crystal structures in that V9E1 blocks access of RTA to the ribosomal P-stalk proteins, whereas V9F6, whose epitope is slightly offset vis a vis the ribosomal P-stalk binding site on RTA, has no effect.
      Figure thumbnail gr6
      Figure 6V9E1 but not V9F6 prevents RTA-stalk binding. A, schematic of the RTA-stalk competition ELISA. B, RTA (303 nM) was incubated with serial dilutions of either V9E1 or V9F6 and added to BSA–P2C11-coated ELISA plates. Platebound RTA was detected with a noncompeting monoclonal antibody and normalized to control wells without VHH added (100%). Vertical dotted line indicates the 1:1 M ratio of RTA:VHH added to the plate. Each point represents the mean ± SD of three (V9F6) or four (V9E1) biological replicates. BSA–P2C11, P2C11 peptide–bovine serum albumin BSA conjugate; RTA, ricin toxin subunit A.

      V9E1 intrabodies render Vero cells resistant to ricin intoxication

      While V9E1 had no detectable ricin toxin neutralizing activity when applied to Vero cells extracellularly, we postulated that it might protect cells if expressed within the cell cytoplasm where it can encounter free RTA prior to engagement with the ribosome. In a recent study, we developed a protocol to successfully express RTA-specific VHHs as intracellular antibodies or “intrabodies” in Vero cells (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ,
      • Czajka T.F.
      • Mantis N.J.
      Single-domain antibodies for intracellular toxin neutralization.
      ). Using that approach, the V9E1 and V9F6 coding sequences were cloned into the mammalian expression vector pcDNA3 and delivered into Vero cells by lipid nanoparticle (LNP) transfection. Both V9E1 and V9F6 were detected in cell lysates at comparable levels, as determined using an RTA-capture ELISA (Fig. S8).
      To assess the impact of the V9E1 and V9F6 intrabodies on the sensitivity of Vero cells to ricin intoxication, transiently transfected cells were treated with escalating amounts of ricin toxin for 2 h, washed, and then assessed for viability 48 h later (Fig. 7). A botulinum neurotoxin–specific VHH, ciA-H7, was used as a control for these studies (
      • Tremblay J.M.
      • Kuo C.L.
      • Abeijon C.
      • Sepulveda J.
      • Oyler G.
      • Hu X.
      • Jin M.M.
      • Shoemaker C.B.
      Camelid single domain antibodies (VHHs) as neuronal cell intrabody binding agents and inhibitors of Clostridium botulinum neurotoxin (BoNT) proteases.
      ). Vero cells transfected with V9E1 were ∼100-fold more resistant to ricin, as compared with cells transfected with LNP. In contrast, Vero cells transfected with V9F6 remained as sensitive to ricin as control cells (Fig. 7). This result demonstrates that targeting the ribosomal P-stalk binding site of RTA is sufficient to protect the ribosomes from ricin’s RIP activity.
      Figure thumbnail gr7
      Figure 7Protection of Vero cells from ricin by intracellular V9E1. Vero cells were seeded in 96-well plates for 1 day and then transiently transfected with VHH intrabodies. Cells were treated with dilutions of ricin 1 day after transfection, and viability was assessed 2 days post-treatment. Left, viabilities of cells transfected with vehicle control (LNP; open circles) or intrabodies (filled circles or squares) were measured as a percentage of “live” control cells not treated with ricin. Shown are the mean ± SD of at least three biological replicates representing three technical replicates each. In the absence of a BoNT-specific ELISA, expression of ciA-H7 intrabody was confirmed by Western blot (data not shown). Significance was determined at each ricin concentration by two-way ANOVA. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. Right, epitopes on RTA highlighted for V9E1 (green) and V9F6 (pink) with the stalk peptide shown in magenta. BoNT, botulinum neurotoxin; LNP, lipid nanoparticle; RTA, ricin toxin subunit A.

      Inactivation of RTA by the remaining V9 families of VHHs

      We next turned our attention to the other V9 VHH family members (Table 1). V9B2 and V9F9 (families 3 and 4) were of particular interest because they occlude the ribosome P-stalk binding site of RTA like V9E1 but bind RTA with significantly higher affinity (Table 1 and Fig. 4). As predicted, V9B2 and V9F9 inhibited RTA from binding to P2C11 in a dose-dependent manner (Fig. S9). Three other VHHs from family 4 (V9B8/V9E9/V9G11) also prevented RTA–P2C11 interactions, whereas V9D5, a member of family 5, did not (Fig. S9). Similarly, in the IVT assay, there was a clear hierarchy of RTA neutralization, with V9B2 and V9F9 and other VHHs in family 4 (V9B8/V9E9/V9G11) being the most potent inhibitors, followed by V9E1 and V9E5 (families 1 and 2) (Fig. 8). In contrast, V9D5 (family 5) was ineffective at neutralizing RTA in the IVT assay. These results further demonstrate that direct obstruction of, and not just proximity to, the ribosomal recruitment site of RTA is necessary to block P2 stalk binding and to neutralize RTA RIP activity.
      Figure thumbnail gr8
      Figure 8Inhibition of the activity of RTA by VHH families 2, 3, and 4. A, in vitro translation assays were performed with VHHs from families 2, 3, 4, and the remaining family 5 VHH (V9D5). Shown are mean ± SD of three replicate experiments. Significance at each RTA concentration was determined by two-way ANOVA. B, dose–response curves for each VHH–RTA combination were transformed, and area under the curve (AUC) analysis was performed. The dashed line represents the standard curve AUC (x = 118). The high-affinity active site VHH, V2A11, is also shown (gray). Significance was calculated using a one-way ANOVA with comparisons to the standard curve AUC shown. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. RTA, ricin toxin subunit A.
      From these experiments, we expected that, when expressed as intrabodies, V9B2 and V9F9 and three other VHHs in family 4 would be more effective than V9E1 at protecting Vero cells from ricin intoxication. To test this hypothesis, Vero cells were transiently transfected with pcDNA3 expression vectors encoding families 3 and 4 VHHs (V9B2, V9B8, V9E9, V9F9, and V9G11) and then challenged with ricin toxin 1 day later. While each of the VHHs rendered Vero cells more resistant to ricin than the mock-transfected cells, their activities were lower than anticipated considering their ultra–high-binding affinities (Fig. 9). In fact, area under the curve (AUC) analysis indicated that V9E1 (family 1) was superior to VHHs in families 3 and 4 (data not shown). VHHs in families 3 and 4 differ from V9E1 (family 1) in that they have (or are suspected to have) a second disulfide bond that tethers CDR2 and CDR3, which could reduce protein expression (Fig. S10) (
      • Muyldermans S.
      Applications of nanobodies.
      ).
      Figure thumbnail gr9
      Figure 9Protection of Vero cells from ricin by family 2, 3, and 4 intrabodies. Cytotoxicity assays were performed with intrabodies from families 2, 3, 4, and the remaining family 5 intrabody (V9D5). Viabilities of cells transfected with vehicle control (LNP; open circles) or intrabodies (filled circles or squares) were measured as a percentage of “live” control cells not treated with ricin. Shown are the mean ± SD of at least three biological replicates representing three technical replicates each. Significance was determined at each ricin concentration by two-way ANOVA. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. LNP, lipid nanoparticle.
      To examine this experimentally, Vero cells were transfected as noted previously, and then cell lysates were subjected to RTA ELISA. As predicted, V9B2 was expressed at low levels, as compared with VHHs in families 1, 2, and 5 (Fig. S8). Indeed, all five VHHs with a predicted second disulfide bond were expressed at low levels (Fig. S8). Nonetheless, V9B2-transfected cells were significantly more resistant to ricin, as compared with control cells, revealing that even small amount of antibody has the capacity to neutralize ricin (Fig. 9).
      Finally, we evaluated V9E5 (family 2) and V9D5 (family 5) as intrabodies. Neither V9E5 nor V9D5 contain a second disulfide bond, and, correspondingly, both expressed well in transfected cells (Fig. S8). V9E5-transfected cells were significantly resistant to ricin compared with LNP cells (>22-fold increase in EC50), whereas V9D5 transfection conferred no protection from the toxin (Fig. 9).

      Correlation between epitope specificity and binding affinity in RTA inhibition

      With the recent discovery of the P-stalk binding site and demonstration that preventing P-stalk binding reduces the enzymatic activity of RTA, there is interest in developing small molecules against RTA that target P-stalk binding site (
      • Li X.P.
      • Harijan R.K.
      • Cao B.
      • Kahn J.N.
      • Pierce M.
      • Tsymbal A.M.
      • Roberge J.Y.
      • Augeri D.
      • Tumer N.E.
      Synthesis and structural characterization of ricin inhibitors targeting ribosome binding using fragment-based methods and structure-based design.
      ). Those approaches have yielded compounds with high specificity that mimic interactions of Phe-10 within the P2C11 peptide and RTA as revealed by structural analysis, similar to the mimicry we observed for V9B2, V9E1, and V9F9. These molecules are also capable of inhibiting the depurination activity of RTA and can reduce ricin cytotoxicity in a Vero cell model, albeit at high concentrations because of their low affinities for RTA. From our structural and functional analyses, we reasoned that the primary drivers for recruitment site–targeted neutralization are binding strength, proximity to the P-stalk binding site, and the ability to express intracellularly. With this in mind, we reasoned that a competition assay with V9B2, an ultra–high-affinity RTA binder that overlaps P2C6 binding on RTA, would enable accurate prediction of the neutralization strength of inhibitors targeting the P-stalk binding site.
      In this assay, V9B2 was coated on the ELISA plate, and each potential competitor VHH was serially diluted and mixed with a fixed concentration of RTA (0.46 nM). There was a clear separation of VHH competition profiles into three clusters with unique binding site and strength differences, as predicted (Fig. S11). Ultra–high-affinity VHHs in families 3 and 4 (V9B2, V9B8, V9E9, V9F9, and V9G11) competed strongly with V9B2 as evidenced by 50% binding inhibition at 1:1 VHH:RTA ratios. Next, V9E1 and V9E5, high-affinity VHHs in families 1 and 2, competed more weakly with V9B2 for RTA with IC50 values at >25:1 VHH:RTA ratios. Conversely, neither of the high-affinity VHHs in family 5 (V9F6 and V9D5) demonstrated notable competition, likely because of the fact that their epitopes are offset relative to the P-stalk binding pocket. AUC analyses of these data correlate strongly with cell-free RTA neutralization results using a simple linear regression model, showing an apparent delineation of these three clusters by both V9B2 competition and IVT results (Fig. S11). Multiple linear regression analysis accurately predicted intrabody-based ricin resistance from each of the nine V9 VHHs examined. To correct for expression level differences, we incorporated V9B2 competition data as well as intrabody expression data into this analysis. We observed a strong correlation (r2 = 0.84) between both ELISAs and cytotoxicity results (Fig. S11). These results demonstrate that intracellular neutralization of the toxin by targeting the ribosomal recruitment event of RTA can be strongly predicted based on binding site specificity and affinity.

      Discussion

      The efficiency by which RTA inactivates mammalian ribosomes is remarkable. The underlying chemistry involves hydrolysis of a single N-glycosidic bond of a conserved adenosine residue situated in the SRL of 28S rRNA (
      • Endo Y.
      • Tsurugi K.
      RNA N-glycosidase activity of ricin A-chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes.
      ). The catalytic center of RTA constitutes a solvent-exposed cleft on one side of the molecule that accommodates a protruding adenine base. While the molecular basis of the depurination reaction has been recognized for decades, only recently have we gained insight into how RTA is recruited to the ribosome in the first place (
      • Grela P.
      • Szajwaj M.
      • Horbowicz-Drozdzal P.
      • Tchorzewski M.
      How ricin damages the ribosome.
      ,
      • Zhou Y.
      • Li X.P.
      • Chen B.Y.
      • Tumer N.E.
      Ricin uses arginine 235 as an anchor residue to bind to P-proteins of the ribosomal stalk.
      ,
      • Fan X.
      • Zhu Y.
      • Wang C.
      • Niu L.
      • Teng M.
      • Li X.
      Structural insights into the interaction of the ribosomal P stalk protein P2 with a type II ribosome-inactivating protein ricin.
      ,
      • Grela P.
      • Li X.P.
      • Horbowicz P.
      • Dźwierzyńska M.
      • Tchórzewski M.
      • Tumer N.E.
      Human ribosomal P1-P2 heterodimer represents an optimal docking site for ricin A chain with a prominent role for P1 C-terminus.
      ,
      • Horbowicz-Drożdżal P.
      • Kamel K.
      • Kmiecik S.
      • Borkiewicz L.
      • Tumer N.E.
      • Shaw P.C.
      • Tchórzewski M.
      • Grela P.
      Phosphorylation of the conserved C-terminal domain of ribosomal P-proteins impairs the mode of interaction with plant toxins.
      ,
      • Zhou Y.
      • Li X.P.
      • Kahn J.N.
      • McLaughlin J.E.
      • Tumer N.E.
      Leucine 232 and hydrophobic residues at the ribosomal P stalk binding site are critical for biological activity of ricin.
      ). Indeed, there is now compelling evidence that the RIP activity of RTA is actually a two-step event, wherein step 1 involves interactions with ribosome by P-stalk proteins before the second engagement with the SRL. The X-ray crystal structures of RTA in a complex with peptides derived from human P2 stalk protein revealed a putative ribosomal P-stalk binding pocket located on the opposite side of the molecule relative to the active site (
      • Fan X.
      • Zhu Y.
      • Wang C.
      • Niu L.
      • Teng M.
      • Li X.
      Structural insights into the interaction of the ribosomal P stalk protein P2 with a type II ribosome-inactivating protein ricin.
      ,
      • Shi W.W.
      • Mak A.N.
      • Wong K.B.
      • Shaw P.C.
      Structures and ribosomal interaction of ribosome-inactivating proteins.
      ,
      • Shi W.W.
      • Tang Y.S.
      • Sze S.Y.
      • Zhu Z.N.
      • Wong K.B.
      • Shaw P.C.
      Crystal structure of ribosome-inactivating protein ricin A chain in complex with the C-terminal peptide of the ribosomal stalk protein P2.
      ). Genetic and biochemical studies had already implicated this exact region of RTA as being important in associating with purified ribosomes (
      • Li X.P.
      • Chiou J.C.
      • Remacha M.
      • Ballesta J.P.
      • Tumer N.E.
      A two-step binding model proposed for the electrostatic interactions of ricin a chain with ribosomes.
      ,
      • Zhou Y.
      • Li X.P.
      • Kahn J.N.
      • McLaughlin J.E.
      • Tumer N.E.
      Leucine 232 and hydrophobic residues at the ribosomal P stalk binding site are critical for biological activity of ricin.
      ,
      • Li X.P.
      • Kahn P.C.
      • Kahn J.N.
      • Grela P.
      • Tumer N.E.
      Arginine residues on the opposite side of the active site stimulate the catalysis of ribosome depurination by ricin A chain by interacting with the P-protein stalk.
      ). However, the importance of the P-stalk binding event within the context of intoxicated mammalian cells has remained unclear.
      The results of our current study reveal that ribosome P-stalk binding is an essential step in the RIP activity of RTA. This conclusion is based on the demonstration that surgically targeting the ribosomal P-stalk binding pocket on RTA with high-affinity single-domain antibodies was sufficient to knock out the RIP activity of RTA in cell-free assays, as well as within the context of intoxicated cells. The comparison between two VHHs, V9E1 and V9F6, proved especially informative. V9E1 and V9F6 have comparable binding affinities for RTA (0.11 and 0.22 nM, respectively) but slightly different epitopes relative to the P-stalk binding site. V9E1 occludes the P-stalk binding pocket and forms hydrogen bonds and π-stacking interactions with residues (e.g., Arg234, Arg235, Tyr183, and Phe240) involved in P-stalk peptide recognition (
      • Fan X.
      • Zhu Y.
      • Wang C.
      • Niu L.
      • Teng M.
      • Li X.
      Structural insights into the interaction of the ribosomal P stalk protein P2 with a type II ribosome-inactivating protein ricin.
      ,
      • Shi W.W.
      • Tang Y.S.
      • Sze S.Y.
      • Zhu Z.N.
      • Wong K.B.
      • Shaw P.C.
      Crystal structure of ribosome-inactivating protein ricin A chain in complex with the C-terminal peptide of the ribosomal stalk protein P2.
      ). V9F6, by comparison, recognizes an epitope immediately adjacent to but not obstructing the P-stalk binding pocket. The relatively small difference in epitope location had significant consequences in terms of interfering with RTA activity: V9E1 inhibited RTA–P2C11 interactions, whereas V9F6 did not. V9E1 protected ribosomes from RTA in cell-free translation assays, whereas V9F6 did not. Finally, V9E1 conferred resistance to ricin toxin when expressed as an intrabody in Vero cells, whereas V9F6 did not. These results demonstrate that direct occlusion of the ribosomal P-stalk peptide binding site is sufficient to neutralize the RIP activity of RTA within the cell cytoplasm. Moreover, they suggest that ribosomal P-stalk protein interactions are likely confined to a relatively small surface area on RTA (i.e., essentially outlined by the footprint of V9E1).
      Moreover, blocking access to the P-stalk binding pocket of RTA attenuates the RIP activity of RTA to a degree equivalent to that achieved by interfering with the active site of RTA. This conclusion is based on the comparison of V9E1 with V2A11. V2A11 is one of seven VHHs identified that target the active site pocket of RTA (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). The binding affinity of V2A11 for RTA is similar to V9E1 (0.11 nM versus 0.31 nM). The two antibodies have identical Sc scores (Sc = 0.61) and similar buried surface areas (1953 versus 1864 Å2). In the cell-free translation assays, V2A11 and V9E1 had identical RTA inhibition values, and, when comparing across studies of RTA, similar profiles in terms of conferring resistance to ricin toxin when expressed as intrabodies in Vero cells. In essence, docking with the ribosomal P-stalk proteins is a requisite step in the RIP activity, and as others have proposed, as important as the actual depurination event itself (
      • Li X.P.
      • Chiou J.C.
      • Remacha M.
      • Ballesta J.P.
      • Tumer N.E.
      A two-step binding model proposed for the electrostatic interactions of ricin a chain with ribosomes.
      ,
      • May K.L.
      • Li X.P.
      • Martinez-Azorin F.
      • Ballesta J.P.
      • Grela P.
      • Tchorzewski M.
      • Tumer N.E.
      The P1/P2 proteins of the human ribosomal stalk are required for ribosome binding and depurination by ricin in human cells.
      ,
      • Li X.P.
      • Kahn P.C.
      • Kahn J.N.
      • Grela P.
      • Tumer N.E.
      Arginine residues on the opposite side of the active site stimulate the catalysis of ribosome depurination by ricin A chain by interacting with the P-protein stalk.
      ,
      • Chiou J.C.
      • Li X.P.
      • Remacha M.
      • Ballesta J.P.
      • Tumer N.E.
      The ribosomal stalk is required for ribosome binding, depurination of the rRNA and cytotoxicity of ricin A chain in Saccharomyces cerevisiae.
      ).
      In many ways, the two-step model, as put forth by Tumer et al. (
      • Li X.P.
      • Chiou J.C.
      • Remacha M.
      • Ballesta J.P.
      • Tumer N.E.
      A two-step binding model proposed for the electrostatic interactions of ricin a chain with ribosomes.
      ), explains the efficiency of RTA as a RIP. A longstanding question in the field is how RTA, following unfolding and translocation across the ER membrane, finds its substrate within the densely packed environment of the cell cytoplasm. Molecular chaperones like Hsc70 have been proposed to engage RTA following ER dislocation, whereas others have suggested that refolding may be facilitated by ribosomes themselves (
      • Argent R.H.
      • Parrott A.M.
      • Day P.J.
      • Roberts L.M.
      • Stockley P.G.
      • Lord J.M.
      • Radford S.E.
      Ribosome-mediated folding of partially unfolded ricin A-chain.
      ,
      • Spooner R.A.
      • Hart P.J.
      • Cook J.P.
      • Pietroni P.
      • Rogon C.
      • Hohfeld J.
      • Roberts L.M.
      • Lord J.M.
      Cytosolic chaperones influence the fate of a toxin dislocated from the endoplasmic reticulum.
      ). Irrespective of the how and where the folding event occurs, the sheer efficiency of RTA as a ribotoxin indicates that it wastes no time in accessing its substrate. Imai et al. (
      • Imai H.
      • Uchiumi T.
      • Kodera N.
      Direct visualization of translational GTPase factor pool formed around the archaeal ribosomal P-stalk by high-speed AFM.
      ) recently provided evidence that P-stalk proteins function as “factor-pooling platforms” that increase the local concentration of translational GTPases (“EFs”) near the ribosome surface to ensure optimal translation efficiency. Seemingly, RTA exploits (and competes for) the pathways used by EFs to engage with the GTPase-associated center and increase likelihood of encountering the SRL. This strategy may be conserved considering that other RIPs, such as Shiga toxin 2, have also been shown to bind to the C-terminal domain of ribosome P-stalk proteins (
      • Rudolph M.J.
      • Davis S.A.
      • Tumer N.E.
      • Li X.P.
      Structural basis for the interaction of Shiga toxin 2a with a C-terminal peptide of ribosomal P stalk proteins.
      ).
      While our focus up to this point has been on V9E1, it is just one of the nine VHHs identified in this study capable of neutralization of RTA in the IVT assay and interfering with RTA–P2C11 association. Structural analysis of two of those additional VHHs, namely V9B2 and V9F9, revealed occupancy of ribosome P-stalk binding pocket of RTA in a manner similar to V9E1. Six of the nine VHHs (including V9B2 and V9F9) are remarkable in that their binding affinities for RTA were estimated to be in the low picomolar range (<0.03 nM). With extremely slow off-rates, V9B2 and V9F9 (and the other VHHs) were expected to be exceptionally potent inhibitors of ricin toxin when expressed as intrabodies. While they did render Vero cells more resistant to ricin than controls, the levels were not commensurate with binding affinities. As it turns out, V9B2 and V9F9 (as well as others) expressed poorly as intrabodies, most likely because of the fact that they carry a second noncanonical disulfide bond between CDR2 and CDR3. Indeed, it is generally recognized that VHHs and single-chain variable fragments carrying a second disulfide bond are poorly expressed because of the reducing environment of the cytoplasm (
      • Seo M.J.
      • Jeong K.J.
      • Leysath C.E.
      • Ellington A.D.
      • Iverson B.L.
      • Georgiou G.
      Engineering antibody fragments to fold in the absence of disulfide bonds.
      ,
      • Worn A.
      • Pluckthun A.
      An intrinsically stable antibody scFv fragment can tolerate the loss of both disulfide bonds and fold correctly.
      ). Our preliminary efforts to engineer out the pertinent cysteine residues in V9B2 and V9F9 using methodologies described by others have proven unsuccessful to date (T. Czajka, unpublished results) (
      • Worn A.
      • Pluckthun A.
      An intrinsically stable antibody scFv fragment can tolerate the loss of both disulfide bonds and fold correctly.
      ,
      • Proba K.
      • Worn A.
      • Honegger A.
      • Pluckthun A.
      Antibody scFv fragments without disulfide bonds made by molecular evolution.
      ).
      The remarkably high-binding affinities reported here for six of the V9 VHHs are certainly unusual but not unprecedented in the case of ricin toxin or other agents (
      • Rudolph M.J.
      • Vance D.J.
      • Cassidy M.S.
      • Rong Y.
      • Mantis N.J.
      Structural analysis of single domain antibodies bound to a second neutralizing hot spot on ricin toxin's enzymatic subunit.
      ). For example, several VHHs with picomolar affinities for the spike protein of severe acute respiratory syndrome coronavirus 2 were isolated from immune camelid libraries (
      • Guttler T.
      • Aksu M.
      • Dickmanns A.
      • Stegmann K.M.
      • Gregor K.
      • Rees R.
      • Taxer W.
      • Rymarenko O.
      • Schunemann J.
      • Dienemann C.
      • Gunkel P.
      • Mussil B.
      • Krull J.
      • Teichmann U.
      • Gross U.
      • et al.
      Neutralization of SARS-CoV-2 by highly potent, hyperthermostable, and mutation-tolerant nanobodies.
      • Xiang Y.
      • Nambulli S.
      • Xiao Z.
      • Liu H.
      • Sang Z.
      • Duprex W.P.
      • Schneidman-Duhovny D.
      • Zhang C.
      • Shi Y.
      Versatile and multivalent nanobodies efficiently neutralize SARS-CoV-2.
      ). There are likely multiple factors that give rise of such tight binders, including a hydrophobic target interface and the addition of noncanonical disulfide bonds within CDR elements that minimize the entropic cost of antigen engagement (
      • Govaert J.
      • Pellis M.
      • Deschacht N.
      • Vincke C.
      • Conrath K.
      • Muyldermans S.
      • Saerens D.
      Dual beneficial effect of interloop disulfide bond for single domain antibody fragments.
      ,
      • Kunz P.
      • Zinner K.
      • Mucke N.
      • Bartoschik T.
      • Muyldermans S.
      • Hoheisel J.D.
      The structural basis of nanobody unfolding reversibility and thermoresistance.
      ). The relative binding affinities of VHHs can be further enhanced through oligomerization as dimers, trimers, or even higher order oligomers (
      • Czajka T.F.
      • Vance D.J.
      • Mantis N.J.
      Slaying SARS-CoV-2 one (single-domain) antibody at a time.
      ,
      • Laursen N.S.
      • Friesen R.H.E.
      • Zhu X.
      • Jongeneelen M.
      • Blokland S.
      • Vermond J.
      • van Eijgen A.
      • Tang C.
      • van Diepen H.
      • Obmolova G.
      • van der Neut Kolfschoten M.
      • Zuijdgeest D.
      • Straetemans R.
      • Hoffman R.M.B.
      • Nieusma T.
      • et al.
      Universal protection against influenza infection by a multidomain antibody to influenza hemagglutinin.
      ,
      • Muyldermans S.
      A guide to: Generation and design of nanobodies.
      ). Toward this end, we have found that the potency of V9E1–V2A11 heterodimers greatly exceeds that of any single VHH monomer tested to date (T. Czajka and N. Mantis, manuscript in preparation). Such constructs if delivered by mRNA technologies may offer an entirely new strategy for ricin toxin postexposure therapeutics (
      • Thran M.
      • Mukherjee J.
      • Ponisch M.
      • Fiedler K.
      • Thess A.
      • Mui B.L.
      • Hope M.J.
      • Tam Y.K.
      • Horscroft N.
      • Heidenreich R.
      • Fotin-Mleczek M.
      • Shoemaker C.B.
      • Schlake T.
      mRNA mediates passive vaccination against infectious agents, toxins, and tumors.
      ).

      Experimental procedures

      Chemicals and biological reagents

      Ricin toxin (R. communis agglutinin II; RCA60) and RTA were purchased from Vector Laboratories and dialyzed against PBS to remove residual sodium azide prior to use in cytotoxicity assays. Unless noted otherwise, all reagents were purchased from Sigma–Aldrich.

      VHH ELISAs

      Anti-RTA mAbs were coated in 96-well plates at a concentration of 1 μg/ml in PBS overnight at 4 °C. Plates were then blocked with 2% goat serum in PBS containing 0.1% Tween-20 (PBST) for 2 h at room temperature. Ricin, RTA, or RVEc was then applied to the plates at 1 μg/ml to be captured by the anti-RTA antibodies for 1 h, and unbound protein was washed away with PBST. VHHs were then applied at a concentration of 330 nM in PBS for 1 h. Unbound VHH was washed away with PBST, and bound VHH was detected for 1 h with an anti-E-horseradish peroxidase (HRP) secondary antibody (Bethyl Labs). 3,3′,5,5′-Tetramethylbenzidine (TMB; SureBlue; SeraCare) was used to develop the plates and was then quenched with 1 M phosphoric acid. Absorbance at 450 nm was read on a VersaMax microplate reader (Molecular Devices).

      Cloning, expression, and protein purification for structural analysis

      The PCR amplicons for the four VHHs were subcloned into the pSUMO expression vector encoding an N-terminal decahistidine and SUMO tag. RTA (residues 1–268) was subcloned into the N-terminally decahistidine-tagged MCSG7 expression vector. All clonings were performed using a standard ligase-independent cloning protocol. All VHHs and RTA were expressed in Escherichia coli strain BL21(DE3)-pRARE. The transformed bacteria were grown at 37 °C in terrific broth medium and induced at 20 °C with 0.1 mM (IPTG) at an absorbance at 600 nm of 0.6 for ∼16 h at 20 °C. After induction, cells were harvested and resuspended in 20 mM Hepes (pH 7.5) and 150 mM NaCl. The cell suspension was sonicated and centrifuged at 30,000g for 30 min. After centrifugation, the protein-containing supernatant was purified by nickel-affinity and size-exclusion chromatography on an AKTAxpress System (GE Healthcare), which consisted of a 1 ml nickel affinity column followed by a Superdex 200 16/60 gel filtration column. The elution buffer consisted of 0.5 M imidazole in binding buffer, and the gel filtration buffer consisted of 20 mM Hepes (pH 7.5), 150 mM NaCl, and 20 mM imidazole. Fractions containing each VHH and RTA were pooled and subject to tobacco etch virus protease cleavage (1:10 weight ratio) for 3 h at room temperature in order to remove their respective fusion tags. The cleaved proteins were passed over a 1 ml Ni–NTA agarose (Qiagen) gravity column to remove tobacco etch virus protease, cleaved residues, and uncleaved fusion protein. To generate each VHH–RTA protein complex, RTA was mixed with each VHH in a 1:1 stoichiometry then concentrated to 10 mg/ml.

      Crystallization and data collection

      VHH–RTA crystals were grown by sitting drop vapor diffusion at 20 °C using a protein to reservoir volume ratio of 1:1 with total drop volume of 0.2 μl. Crystallization solutions are shown in Table S1. Crystals were flash frozen in liquid nitrogen after a short soak in the respective crystallization buffers supplemented with 25% ethylene glycol. Data were collected at the 24-ID-C and 24-ID-E beamlines at the Advanced Photon Source, Argonne National Labs. All data were indexed, merged, and scaled using HKL2000 (
      • Otwinowski Z.
      • Minor W.
      Processing of X-ray diffraction data collected in oscillation mode.
      ) and then converted to structure factors using CCP4 (
      • Winn M.D.
      • Ballard C.C.
      • Cowtan K.D.
      • Dodson E.J.
      • Emsley P.
      • Evans P.R.
      • Keegan R.M.
      • Krissinel E.B.
      • Leslie A.G.
      • McCoy A.
      • McNicholas S.J.
      • Murshudov G.N.
      • Pannu N.S.
      • Potterton E.A.
      • Powell H.R.
      • et al.
      Overview of the CCP4 suite and current developments.
      ).

      Structure determination and refinement

      The VHH–RTA complex structures were solved by molecular replacement using the program Phaser (
      • McCoy A.J.
      • Grosse-Kunstleve R.W.
      • Adams P.D.
      • Winn M.D.
      • Storoni L.C.
      • Read R.J.
      Phaser crystallographic software.
      ). Molecular replacement calculations were performed using the coordinates of the ricin A chain as a search model for RTA (PDB ID: 1RTC) in all four VHH–RTA complexes. The VHH coordinates used as a search model for all four VHH–RTA complexes were D10 (PDB ID: 4LGR) or A9 (PDB ID: 6CWG) with all three of the CDRs removed from the search model. The resulting phase information from molecular replacement was used to autobuild the polypeptide chain of the VHH within each complex using automated refinement procedure (
      • Morris R.J.
      • Perrakis A.
      • Lamzin V.S.
      ARP/wARP and automatic interpretation of protein electron density maps.
      ). Further manual model building was performed with the open source software, Coot (
      • Emsley P.
      • Lohkamp B.
      • Scott W.G.
      • Cowtan K.
      Features and development of Coot.
      ), combined with structural refinement employing the PHENIX package (
      • Adams P.D.
      • Afonine P.V.
      • Bunkoczi G.
      • Chen V.B.
      • Davis I.W.
      • Echols N.
      • Headd J.J.
      • Hung L.W.
      • Kapral G.J.
      • Grosse-Kunstleve R.W.
      • McCoy A.J.
      • Moriarty N.W.
      • Oeffner R.
      • Read R.J.
      • Richardson D.C.
      • et al.
      PHENIX: A comprehensive Python-based system for macromolecular structure solution.
      ). Data collection and refinement statistics are listed in Table S2. Molecular graphics were prepared using PyMOL (Schrodinger, DeLano Scientific LLC).

      P2C11 peptide competition

      To evaluate VHH/ribosomal stalk competition for RTA binding, 96-well ELISA plates (Immulon 4HBX) were coated overnight at 4 °C with 10 μg/ml of BSA–P2C11 (Genemed Synthesis, Inc) in PBS. Plates were blocked with PBS containing goat serum (2% v/v) and Tween-20 (0.1% v/v) for 2 h at room temperature. During block, RTA (303 nM) was incubated with twofold serial dilutions of VHH in 96-well dilution plates for 30 to 60 min with 12 wells per plate receiving only RTA and block buffer to serve as 100% binding controls. Following block, VHH–RTA samples were transferred to the ELISA plates for 1 h. 1 μg/ml PB10 (anti-RTA mAb) (reference) was added for 1 h followed by HRP-conjugated goat antimouse immunoglobulin G secondary antibody (1:2000 dilution; SouthernBiotech) for 30 min. TMB (SureBlue; Kirkegaard & Perry Labs) was added for 6 to 10 min followed by stop solution (1 M phosphoric acid). ELISA plates were analyzed using a SpectraMax iD3 spectrophotometer equipped with Softmax Pro 7 software (Molecular Devices) at an absorbance at 450 nm. Unbound RTA, PH12, and 2° antibody were removed after each incubation period by washing plates with PBS-Tween (0.1%). RTA, VHHs, and antibodies were diluted in block buffer.

      V9E1 and V9B2 competition assays

      96-Well ELISA plates were coated overnight at 4 °C with 1 μg/ml of V9E1 or V9B2 in PBS. Competition ELISAs were performed as described for the stalk competition, with the exception of the RTA and VHH concentrations. During block, RTA (4.6 nM for V9E1 competition or 0.46 nM for V9B2 competition) was incubated with threefold serial dilutions of VHH in 6-well dilution plates for 30 to 60 min with 12 wells per plate receiving only RTA and block buffer to serve as 100% binding controls.

      Intrabody detection

      Intrabody detection ELISAs were performed as described previously with minor alterations (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). 96-Well ELISA plates were coated overnight at 4 °C with the mAb PH12 (1 μg/ml in PBS) (
      • Van Slyke G.
      • Angalakurthi S.K.
      • Toth R.T.T.
      • Vance D.J.
      • Rong Y.
      • Ehrbar D.
      • Shi Y.
      • Middaugh C.R.
      • Volkin D.B.
      • Weis D.D.
      • Mantis N.J.
      Fine-specificity epitope analysis identifies contact points on ricin toxin recognized by protective monoclonal antibodies.
      ). Plates were washed and blocked for 2 h at room temperature. Following block, RTA (1 μg/ml in PBS) was applied to ELISA plates for 1 h. Transfected cell lysate was serially diluted in duplicate and added to plates for 1 h. Plates were washed and HRP-conjugated anti-E-tag antibody (1:10,000 dilution; Bethyl Laboratories; catalog no.; A190-132P) was applied for 1 h. Plates were washed, and 100 μl TMB was added for 5 to 10 min followed by stop solution. ELISA plates were analyzed using a SpectraMax iD3 spectrophotometer equipped with Softmax Pro 7 software at an absorbance at 450 nm. Purified VHH protein with a C-terminal E-tag was used as a positive binding control for each transfection. Plates were washed following each step in PBS-Tween (0.1%). Cell lysates and secondary antibody were diluted in block buffer.

      In vitro RTA inhibition

      VHHs at a stock concentration of 218.2 nM in dimethyl sulfoxide were diluted in 1% BSA in PBS (w/v) and mixed with threefold serial dilutions of RTA in PBS from 54.5 to 0.07 nM, to a final concentration of 13.6 nM VHH and 13.6 to 0.019 nM RTA. This was added to an ice-cold mixture containing luciferase mRNA (3.7 μg/ml; Ambion, Inc) and Retic Lysate IVT Kit (Thermo Fisher Scientific). The cocktail was incubated for 90 min at 30 °C and then chilled on ice for 5 min before being transferred to wells of an opaque 96-well microtiter plate with an equal volume (20 μl) of Bright-Glo luciferase substrate (Promega) at room temperature. Luminescence was detected immediately using a SpectraMax iD3. Luciferase translation was determined as a percentage of positive control reactions without RTA added. Standard curves were generated for RTA dilutions with dimethyl sulfoxide, and no VHH was added.

      Vero cell culture, transfection, and lysis

      Vero cells (American Type Culture Collection) were cultured in Dulbecco’s minimal essential medium with fetal bovine serum (10% v/v) and penicillin/streptomycin at 37 °C (5% CO2). For ELISAs and Western blotting, cells were transfected in 6-well plates and lysed 48 h after transfection, as described previously (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). For cytotoxicity assays, cells were transfected in 96-well plates and treated with ricin 24 h after transfection, as described (
      • Rudolph M.J.
      • Czajka T.F.
      • Davis S.A.
      • Thi Nguyen C.M.
      • Li X.P.
      • Tumer N.E.
      • Vance D.J.
      • Mantis N.J.
      Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
      ). Viability was determined using Cell Titer-Glo (Promega) and a SpectraMax iD3 for luminescence detection. Viability was measured as a percentage of live control cells (transfected but not treated with ricin).

      Statistical analysis

      Statistical analyses were performed using GraphPad Prism 9.1 software (GraphPad Software, Inc) for Windows. For IVT assays, a one-way ANOVA was with the Tukey's post hoc test to compare areas under the curve for each condition, and a two-way ANOVA was used with the Sidak post hoc test to compare the percent translation between VHH and the RTA standard curves at each RTA concentration. For intrabody ELISA expression analysis, a one-way ANOVA with the Tukey's post hoc test was used to compare each transfection condition. For intrabody cytotoxicity results, a two-way ANOVA was used with the Sidak post hoc test to compare transfected and LNP-vehicle control cell viabilities at each ricin concentration. A simple linear regression and Pearson correlation analysis were performed between the AUC data for competition ELISAs and IVT dose–response curves. To predict intrabody-based Vero cell protection, a multiple linear regression least-squares analysis was performed with cytotoxicity AUC as the dependent outcome and intrabody ELISA and V9B2 competition areas under the curve as independent variables.

      PDB accession numbers

      The structures generated in this study were deposited in the PDB (http://www.rcsb.org/pdb/) under accession numbers 7TGF for V9B2–RTA, 7TGI for V9E1–RTA, 7TH3 for V9F6–RTA, and 7TH2 for V9F9–RTA as described in Table S1.

      Data availability

      All data associated with the results presented in this article are included herein or within the supporting information.

      Supporting information

      This article contains supporting information.

      Conflict of interest

      The authors declare that they have no conflicts of interest with the contents of this article.

      Acknowledgments

      We are grateful to Leslie Eisele and Renjie Song of the Wadsworth Center’s Biochemistry and Immunology Core facility for assistance with SPR. We thank the staff in the Wadsworth Center’s Media and Tissue Culture core for providing reagents, the Applied Genomics Technologies core for DNA sequencing. We thank members of the Mantis laboratory for technical assistance and Beth Cavosie for administrative support. Finally, we gratefully acknowledge the 24-ID-C and 24-ID-E beamline staff at the Advanced Photon Source for their assistance in data collection.

      Author contributions

      T. F. C., D. J. V., S. D., M. J. R., and N. J. M. conceptualization; T. F. C., D. J. V., M. J. R., and N. J. M. methodology; T. F. C., D. J. V., M. J. R., and N. J. M. software; T. F. C., D. J. V., and M. J. R. validation; T. F. C., D. J. V., M. J. R., and N. J. M. formal analysis; T. F. C., D. J. V., S. D., M. J. R., and N. J. M. investigation; T. F. C., D. J. V., and M. J. R. resources; M. J. R. data curation; T. F. C., D. J. V., M. J. R., and N. J. M. writing–original draft; T. F. C., D. J. V., M. J. R., and N. J. M. writing–review & editing; T. F. C., D. J. V., and M. J. R. visualization; M. J. R. and N. J. M. supervision; N. J. M. project administration; N. J. M. funding acquisition.

      Funding and additional information

      This work was supported by contract no. HHSN272201400021C and grant R01 AI125190 from the National Institute of Allergy and Infectious Diseases ( National Institutes of Health ). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the article.

      Supporting information

      References

        • Montfort W.
        • Villafranca J.E.
        • Monzingo A.F.
        • Ernst S.R.
        • Katzin B.
        • Rutenber E.
        • Xuong N.H.
        • Hamlin R.
        • Robertus J.D.
        The three-dimensional structure of ricin at 2.8 A.
        J. Biol. Chem. 1987; 262: 5398-5403
        • Olsnes S.
        • Refsnes K.
        • Pihl A.
        Mechanism of action of the toxic lectins abrin and ricin.
        Nature. 1974; 249: 627-631
        • Sowa-Rogozinska N.
        • Sominka H.
        • Nowakowska-Golacka J.
        • Sandvig K.
        • Slominska-Wojewodzka M.
        Intracellular transport and cytotoxicity of the protein toxin ricin.
        Toxins (Basel). 2019; 11: 350
        • Bellisola G.
        • Fracasso G.
        • Ippoliti R.
        • Menestrina G.
        • Rosen A.
        • Solda S.
        • Udali S.
        • Tomazzolli R.
        • Tridente G.
        • Colombatti M.
        Reductive activation of ricin and ricin A-chain immunotoxins by protein disulfide isomerase and thioredoxin reductase.
        Biochem. Pharmacol. 2004; 67: 1721-1731
        • Nowakowska-Gołacka J.
        • Sominka H.
        • Sowa-Rogozińska N.
        • Słomińska-Wojewódzka M.
        Toxins utilize the endoplasmic reticulum-associated protein degradation pathway in their intoxication process.
        Int. J. Mol. Sci. 2019; 20: 1307
        • Spooner R.A.
        • Watson P.D.
        • Marsden C.J.
        • Smith D.C.
        • Moore K.A.
        • Cook J.P.
        • Lord J.M.
        • Roberts L.M.
        Protein disulphide-isomerase reduces ricin to its A and B chains in the endoplasmic reticulum.
        Biochem. J. 2004; 383: 285-293
        • Wesche J.
        • Rapak A.
        • Olsnes S.
        Dependence of ricin toxicity on translocation of the toxin A-chain from the endoplasmic reticulum to the cytosol.
        J. Biol. Chem. 1999; 274: 34443-34449
        • Endo Y.
        • Mitsui K.
        • Motizuki M.
        • Tsurugi K.
        The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins.
        J. Biol. Chem. 1987; 262: 5908-5912
        • Endo Y.
        • Tsurugi K.
        RNA N-glycosidase activity of ricin A-chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes.
        J. Biol. Chem. 1987; 262: 8128-8130
        • Iordanov M.S.
        • Pribnow D.
        • Magun J.L.
        • Dinh T.H.
        • Pearson J.A.
        • Chen S.L.
        • Magun B.E.
        Ribotoxic stress response: Activation of the stress-activated protein kinase JNK1 by inhibitors of the peptidyl transferase reaction and by sequence-specific RNA damage to the alpha-sarcin/ricin loop in the 28S rRNA.
        Mol. Cell. Biol. 1997; 17: 3373-3381
        • Vind A.C.
        • Genzor A.V.
        • Bekker-Jensen S.
        Ribosomal stress-surveillance: Three pathways is a magic number.
        Nucleic Acids Res. 2020; 48: 10648-10661
        • Carbone C.E.
        • Loveland A.B.
        • Gamper Jr., H.B.
        • Hou Y.M.
        • Demo G.
        • Korostelev A.A.
        Time-resolved cryo-EM visualizes ribosomal translocation with EF-G and GTP.
        Nat. Commun. 2021; 12: 7236
        • Endo Y.
        • Tsurugi K.
        The RNA N-glycosidase activity of ricin A-chain. The characteristics of the enzymatic activity of ricin A-chain with ribosomes and with rRNA.
        J. Biol. Chem. 1988; 263: 8735-8739
        • Ready M.P.
        • Kim Y.
        • Robertus J.D.
        Site-directed mutagenesis of ricin A-chain and implications for the mechanism of action.
        Proteins. 1991; 10: 270-278
        • Moazed D.
        • Robertson J.M.
        • Noller H.F.
        Interaction of elongation factors EF-G and EF-Tu with a conserved loop in 23S RNA.
        Nature. 1988; 334: 362-364
        • Monzingo A.F.
        • Robertus J.D.
        X-ray analysis of substrate analogs in the ricin A-chain active site.
        J. Mol. Biol. 1992; 227: 1136-1145
        • Mlsna D.
        • Monzingo A.F.
        • Katzin B.J.
        • Ernst S.
        • Robertus J.D.
        Structure of recombinant ricin A chain at 2.3 A.
        Protein Sci. 1993; 2: 429-435
        • Robertus J.D.
        • Yan X.
        • Ernst S.
        • Monzingo A.
        • Worley S.
        • Day P.
        • Hollis T.
        • Svinth M.
        Structural analysis of ricin and implications for inhibitor design.
        Toxicon. 1996; 34: 1325-1334
        • Grela P.
        • Szajwaj M.
        • Horbowicz-Drozdzal P.
        • Tchorzewski M.
        How ricin damages the ribosome.
        Toxins (Basel). 2019; 11: 241
        • May K.L.
        • Yan Q.
        • Tumer N.E.
        Targeting ricin to the ribosome.
        Toxicon. 2013; 69: 143-151
        • Li X.P.
        • Chiou J.C.
        • Remacha M.
        • Ballesta J.P.
        • Tumer N.E.
        A two-step binding model proposed for the electrostatic interactions of ricin a chain with ribosomes.
        Biochemistry. 2009; 48: 3853-3863
        • May K.L.
        • Li X.P.
        • Martinez-Azorin F.
        • Ballesta J.P.
        • Grela P.
        • Tchorzewski M.
        • Tumer N.E.
        The P1/P2 proteins of the human ribosomal stalk are required for ribosome binding and depurination by ricin in human cells.
        FEBS J. 2012; 279: 3925-3936
        • Jetzt A.E.
        • Cheng J.S.
        • Li X.P.
        • Tumer N.E.
        • Cohick W.S.
        A relatively low level of ribosome depurination by mutant forms of ricin toxin A chain can trigger protein synthesis inhibition, cell signaling and apoptosis in mammalian cells.
        Int. J. Biochem. Cell Biol. 2012; 44: 2204-2211
        • Zhou Y.
        • Li X.P.
        • Chen B.Y.
        • Tumer N.E.
        Ricin uses arginine 235 as an anchor residue to bind to P-proteins of the ribosomal stalk.
        Sci. Rep. 2017; 7: 42912
        • Fan X.
        • Zhu Y.
        • Wang C.
        • Niu L.
        • Teng M.
        • Li X.
        Structural insights into the interaction of the ribosomal P stalk protein P2 with a type II ribosome-inactivating protein ricin.
        Sci. Rep. 2016; 6: 37803
        • Shi W.W.
        • Mak A.N.
        • Wong K.B.
        • Shaw P.C.
        Structures and ribosomal interaction of ribosome-inactivating proteins.
        Molecules. 2016; 21: 1588
        • Li X.P.
        • Kahn J.N.
        • Tumer N.E.
        Peptide mimics of the ribosomal P stalk inhibit the activity of ricin A chain by preventing ribosome binding.
        Toxins (Basel). 2018; 10: 371
        • Rudolph M.J.
        • Czajka T.F.
        • Davis S.A.
        • Thi Nguyen C.M.
        • Li X.P.
        • Tumer N.E.
        • Vance D.J.
        • Mantis N.J.
        Intracellular neutralization of ricin toxin by single-domain antibodies targeting the active site.
        J. Mol. Biol. 2020; 432: 1109-1125
        • Poon A.Y.
        • Vance D.J.
        • Rong Y.
        • Ehrbar D.
        • Mantis N.J.
        A supercluster of neutralizing epitopes at the interface of ricin's enzymatic (RTA) and binding (RTB) subunits.
        Toxins (Basel). 2017; 9: 378
        • Rudolph M.J.
        • Poon A.Y.
        • Kavaliauskiene S.
        • Myrann A.G.
        • Reynolds-Peterson C.
        • Davis S.A.
        • Sandvig K.
        • Vance D.J.
        • Mantis N.J.
        Structural analysis of toxin-neutralizing, single-domain antibodies that bridge ricin's A-B subunit interface.
        J. Mol. Biol. 2021; 433: 167086
        • Vance D.J.
        • Tremblay J.M.
        • Rong Y.
        • Angalakurthi S.K.
        • Volkin D.B.
        • Middaugh C.R.
        • Weis D.D.
        • Shoemaker C.B.
        • Mantis N.J.
        High-resolution epitope positioning of a large collection of neutralizing and nonneutralizing single-domain antibodies on the enzymatic and binding subunits of ricin toxin.
        Clin. Vaccine Immunol. 2017; 24e00236-17
        • Toth R.T.I.
        • Angalakurthi S.K.
        • Van Slyke G.
        • Vance D.J.
        • Hickey J.M.
        • Joshi S.B.
        • Middaugh C.R.
        • Volkin D.B.
        • Weis D.D.
        • Mantis N.J.
        High-definition mapping of four spatially distinct neutralizing epitope clusters on RiVax, a candidate ricin toxin subunit vaccine.
        Clin. Vaccine Immunol. 2017; 24e00237-17
        • Rudolph M.J.
        • Vance D.J.
        • Cassidy M.S.
        • Rong Y.
        • Mantis N.J.
        Structural analysis of single domain antibodies bound to a second neutralizing hot spot on ricin toxin's enzymatic subunit.
        J. Biol. Chem. 2017; 292: 872-883
        • Muyldermans S.
        Applications of nanobodies.
        Annu. Rev. Anim. Biosci. 2021; 9: 401-421
        • Shi W.W.
        • Tang Y.S.
        • Sze S.Y.
        • Zhu Z.N.
        • Wong K.B.
        • Shaw P.C.
        Crystal structure of ribosome-inactivating protein ricin A chain in complex with the C-terminal peptide of the ribosomal stalk protein P2.
        Toxins (Basel). 2016; 8: 296
        • Czajka T.F.
        • Mantis N.J.
        Single-domain antibodies for intracellular toxin neutralization.
        Methods Mol. Biol. 2022; 2446: 469-487
        • Tremblay J.M.
        • Kuo C.L.
        • Abeijon C.
        • Sepulveda J.
        • Oyler G.
        • Hu X.
        • Jin M.M.
        • Shoemaker C.B.
        Camelid single domain antibodies (VHHs) as neuronal cell intrabody binding agents and inhibitors of Clostridium botulinum neurotoxin (BoNT) proteases.
        Toxicon. 2010; 56: 990-998
        • Li X.P.
        • Harijan R.K.
        • Cao B.
        • Kahn J.N.
        • Pierce M.
        • Tsymbal A.M.
        • Roberge J.Y.
        • Augeri D.
        • Tumer N.E.
        Synthesis and structural characterization of ricin inhibitors targeting ribosome binding using fragment-based methods and structure-based design.
        J. Med. Chem. 2021; 64: 15334-15348
        • Grela P.
        • Li X.P.
        • Horbowicz P.
        • Dźwierzyńska M.
        • Tchórzewski M.
        • Tumer N.E.
        Human ribosomal P1-P2 heterodimer represents an optimal docking site for ricin A chain with a prominent role for P1 C-terminus.
        Sci. Rep. 2017; 7: 5608
        • Horbowicz-Drożdżal P.
        • Kamel K.
        • Kmiecik S.
        • Borkiewicz L.
        • Tumer N.E.
        • Shaw P.C.
        • Tchórzewski M.
        • Grela P.
        Phosphorylation of the conserved C-terminal domain of ribosomal P-proteins impairs the mode of interaction with plant toxins.
        FEBS Lett. 2021; 595: 2221-2236
        • Zhou Y.
        • Li X.P.
        • Kahn J.N.
        • McLaughlin J.E.
        • Tumer N.E.
        Leucine 232 and hydrophobic residues at the ribosomal P stalk binding site are critical for biological activity of ricin.
        Biosci. Rep. 2019; 39BSR20192022
        • Li X.P.
        • Kahn P.C.
        • Kahn J.N.
        • Grela P.
        • Tumer N.E.
        Arginine residues on the opposite side of the active site stimulate the catalysis of ribosome depurination by ricin A chain by interacting with the P-protein stalk.
        J. Biol. Chem. 2013; 288: 30270-30284
        • Chiou J.C.
        • Li X.P.
        • Remacha M.
        • Ballesta J.P.
        • Tumer N.E.
        The ribosomal stalk is required for ribosome binding, depurination of the rRNA and cytotoxicity of ricin A chain in Saccharomyces cerevisiae.
        Mol. Microbiol. 2008; 70: 1441-1452
        • Argent R.H.
        • Parrott A.M.
        • Day P.J.
        • Roberts L.M.
        • Stockley P.G.
        • Lord J.M.
        • Radford S.E.
        Ribosome-mediated folding of partially unfolded ricin A-chain.
        J. Biol. Chem. 2000; 275: 9263-9269
        • Spooner R.A.
        • Hart P.J.
        • Cook J.P.
        • Pietroni P.
        • Rogon C.
        • Hohfeld J.
        • Roberts L.M.
        • Lord J.M.
        Cytosolic chaperones influence the fate of a toxin dislocated from the endoplasmic reticulum.
        Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 17408-17413
        • Imai H.
        • Uchiumi T.
        • Kodera N.
        Direct visualization of translational GTPase factor pool formed around the archaeal ribosomal P-stalk by high-speed AFM.
        Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 32386-32394
        • Rudolph M.J.
        • Davis S.A.
        • Tumer N.E.
        • Li X.P.
        Structural basis for the interaction of Shiga toxin 2a with a C-terminal peptide of ribosomal P stalk proteins.
        J. Biol. Chem. 2020; 295: 15588-15596
        • Seo M.J.
        • Jeong K.J.
        • Leysath C.E.
        • Ellington A.D.
        • Iverson B.L.
        • Georgiou G.
        Engineering antibody fragments to fold in the absence of disulfide bonds.
        Protein Sci. 2009; 18: 259-267
        • Worn A.
        • Pluckthun A.
        An intrinsically stable antibody scFv fragment can tolerate the loss of both disulfide bonds and fold correctly.
        FEBS Lett. 1998; 427: 357-361
        • Proba K.
        • Worn A.
        • Honegger A.
        • Pluckthun A.
        Antibody scFv fragments without disulfide bonds made by molecular evolution.
        J. Mol. Biol. 1998; 275: 245-253
        • Guttler T.
        • Aksu M.
        • Dickmanns A.
        • Stegmann K.M.
        • Gregor K.
        • Rees R.
        • Taxer W.
        • Rymarenko O.
        • Schunemann J.
        • Dienemann C.
        • Gunkel P.
        • Mussil B.
        • Krull J.
        • Teichmann U.
        • Gross U.
        • et al.
        Neutralization of SARS-CoV-2 by highly potent, hyperthermostable, and mutation-tolerant nanobodies.
        EMBO J. 2021; 40e107985
        • Xiang Y.
        • Nambulli S.
        • Xiao Z.
        • Liu H.
        • Sang Z.
        • Duprex W.P.
        • Schneidman-Duhovny D.
        • Zhang C.
        • Shi Y.
        Versatile and multivalent nanobodies efficiently neutralize SARS-CoV-2.
        Science. 2020; 370: 1479-1484
        • Govaert J.
        • Pellis M.
        • Deschacht N.
        • Vincke C.
        • Conrath K.
        • Muyldermans S.
        • Saerens D.
        Dual beneficial effect of interloop disulfide bond for single domain antibody fragments.
        J. Biol. Chem. 2012; 287: 1970-1979
        • Kunz P.
        • Zinner K.
        • Mucke N.
        • Bartoschik T.
        • Muyldermans S.
        • Hoheisel J.D.
        The structural basis of nanobody unfolding reversibility and thermoresistance.
        Sci. Rep. 2018; 8: 7934
        • Czajka T.F.
        • Vance D.J.
        • Mantis N.J.
        Slaying SARS-CoV-2 one (single-domain) antibody at a time.
        Trends Microbiol. 2020; 29: 195-203
        • Laursen N.S.
        • Friesen R.H.E.
        • Zhu X.
        • Jongeneelen M.
        • Blokland S.
        • Vermond J.
        • van Eijgen A.
        • Tang C.
        • van Diepen H.
        • Obmolova G.
        • van der Neut Kolfschoten M.
        • Zuijdgeest D.
        • Straetemans R.
        • Hoffman R.M.B.
        • Nieusma T.
        • et al.
        Universal protection against influenza infection by a multidomain antibody to influenza hemagglutinin.
        Science. 2018; 362: 598-602
        • Muyldermans S.
        A guide to: Generation and design of nanobodies.
        FEBS J. 2021; 288: 2084-2102
        • Thran M.
        • Mukherjee J.
        • Ponisch M.
        • Fiedler K.
        • Thess A.
        • Mui B.L.
        • Hope M.J.
        • Tam Y.K.
        • Horscroft N.
        • Heidenreich R.
        • Fotin-Mleczek M.
        • Shoemaker C.B.
        • Schlake T.
        mRNA mediates passive vaccination against infectious agents, toxins, and tumors.
        EMBO Mol. Med. 2017; 9: 1434-1447
        • Otwinowski Z.
        • Minor W.
        Processing of X-ray diffraction data collected in oscillation mode.
        Methods Enzymol. 1997; 276: 307-326
        • Winn M.D.
        • Ballard C.C.
        • Cowtan K.D.
        • Dodson E.J.
        • Emsley P.
        • Evans P.R.
        • Keegan R.M.
        • Krissinel E.B.
        • Leslie A.G.
        • McCoy A.
        • McNicholas S.J.
        • Murshudov G.N.
        • Pannu N.S.
        • Potterton E.A.
        • Powell H.R.
        • et al.
        Overview of the CCP4 suite and current developments.
        Acta Crystallogr. D Biol. Crystallogr. 2011; 67: 235-242
        • McCoy A.J.
        • Grosse-Kunstleve R.W.
        • Adams P.D.
        • Winn M.D.
        • Storoni L.C.
        • Read R.J.
        Phaser crystallographic software.
        J. Appl. Crystallogr. 2007; 40: 658-674
        • Morris R.J.
        • Perrakis A.
        • Lamzin V.S.
        ARP/wARP and automatic interpretation of protein electron density maps.
        Methods Enzymol. 2003; 374: 229-244
        • Emsley P.
        • Lohkamp B.
        • Scott W.G.
        • Cowtan K.
        Features and development of Coot.
        Acta Crystallogr. D Biol. Crystallogr. 2010; 66: 486-501
        • Adams P.D.
        • Afonine P.V.
        • Bunkoczi G.
        • Chen V.B.
        • Davis I.W.
        • Echols N.
        • Headd J.J.
        • Hung L.W.
        • Kapral G.J.
        • Grosse-Kunstleve R.W.
        • McCoy A.J.
        • Moriarty N.W.
        • Oeffner R.
        • Read R.J.
        • Richardson D.C.
        • et al.
        PHENIX: A comprehensive Python-based system for macromolecular structure solution.
        Acta Crystallogr. D Biol. Crystallogr. 2010; 66: 213-221
        • Van Slyke G.
        • Angalakurthi S.K.
        • Toth R.T.T.
        • Vance D.J.
        • Rong Y.
        • Ehrbar D.
        • Shi Y.
        • Middaugh C.R.
        • Volkin D.B.
        • Weis D.D.
        • Mantis N.J.
        Fine-specificity epitope analysis identifies contact points on ricin toxin recognized by protective monoclonal antibodies.
        Immunohorizons. 2018; 2: 262-273