Engineered SUMO/protease system identifies Pdr6 as a bidirectional nuclear transport receptor

High-yield, plant-based production of an antimicrobial peptide with potent activity in a mouse model

Plants offer a promising chassis for the large-scale, cost-effective production of diverse therapeutics, including antimicrobial peptides (AMPs). However, key advances will reduce production costs, including simplifying the downstream processing and purification steps. Here, using Nicotiana benthamiana plants, we present an improved modular design that enables AMPs to be secreted via the endomembrane system and sequestered in an extracellular compartment, the apoplast. Additionally, we translationally fused an AMP to a mutated small ubiquitin-like modifier sequence, thereby enhancing peptide yield and solubilizing the peptide with minimal aggregation and reduced occurrence of necrotic lesions in the plant. This strategy resulted in substantial peptide accumulation, reaching around 2.9 mg AMP per 20 g fresh weight of leaf tissue. Furthermore, the purified AMP demonstrated low collateral toxicity in primary human skin cells, killed pathogenic bacteria by permeabilizing the membrane and exhibited anti-infective efficacy in a preclinical mouse (Mus musculus) model system, reducing bacterial loads by up to three orders of magnitude. A base-case techno-economic analysis demonstrated the economic advantages and scalability of our plant-based platform. We envision that our work can establish plants as efficient bioreactors for producing preclinical-grade AMPs at a commercial scale, with the potential for clinical applications.

The structural basis for RNA slicing by human Argonaute2

Argonaute (AGO) proteins associate with guide RNAs to form complexes that slice transcripts that pair to the guide. This slicing drives post-transcriptional gene-silencing pathways that are essential for many eukaryotes and the basis for new clinical therapies. Despite this importance, structural information on eukaryotic AGOs in a fully paired, slicing-competent conformation-hypothesized to be intrinsically unstable-has been lacking. Here we present the cryogenic-electron microscopy structure of a human AGO-guide complex bound to a fully paired target, revealing structural rearrangements that enable this conformation. Critically, the N domain of AGO rotates to allow the RNA full access to the central channel and forms contacts that license rapid slicing. Moreover, a conserved loop in the PIWI domain secures the RNA near the active site to enhance slicing rate and specificity. These results explain how AGO accommodates targets possessing the pairing specificity typically observed in biological and clinical slicing substrates.

The guide-RNA sequence dictates the slicing kinetics and conformational dynamics of the Argonaute silencing complex

The RNA-induced silencing complex (RISC), which powers RNA interference (RNAi), consists of a guide RNA and an Argonaute protein that slices target RNAs complementary to the guide. We find that, for different guide-RNA sequences, slicing rates of perfectly complementary bound targets can be surprisingly different (>250-fold range), and that faster slicing confers better knockdown in cells. Nucleotide sequence identities at guide-RNA positions 7, 10, and 17 underlie much of this variation in slicing rates. Analysis of one of these determinants implicates a structural distortion at guide nucleotides 6-7 in promoting slicing. Moreover, slicing directed by different guide sequences has an unanticipated, 600-fold range in 3'-mismatch tolerance, attributable to guides with weak (AU-rich) central pairing requiring extensive 3' complementarity (pairing beyond position 16) to more fully populate the slicing-competent conformation. Together, our analyses identify sequence determinants of RISC activity and provide biochemical and conformational rationale for their action.

The guide RNA sequence dictates the slicing kinetics and conformational dynamics of the Argonaute silencing complex

The RNA-induced silencing complex (RISC), which powers RNA interference (RNAi), consists of a guide RNA and an Argonaute protein that slices target RNAs complementary to the guide. We find that for different guide-RNA sequences, slicing rates of perfectly complementary, bound targets can be surprisingly different (>250-fold range), and that faster slicing confers better knockdown in cells. Nucleotide sequence identities at guide-RNA positions 7, 10, and 17 underlie much of this variation in slicing rates. Analysis of one of these determinants implicates a structural distortion at guide nucleotides 6-7 in promoting slicing. Moreover, slicing directed by different guide sequences has an unanticipated, 600-fold range in 3'-mismatch tolerance, attributable to guides with weak (AU-rich) central pairing requiring extensive 3' complementarity (pairing beyond position 16) to more fully populate the slicing-competent conformation. Together, our analyses identify sequence determinants of RISC activity and provide biochemical and conformational rationale for their action.

A checkpoint function for Nup98 in nuclear pore formation suggested by novel inhibitory nanobodies

Nuclear pore complex (NPC) biogenesis is a still enigmatic example of protein self-assembly. We now introduce several cross-reacting anti-Nup nanobodies for imaging intact nuclear pore complexes from frog to human. We also report a simplified assay that directly tracks postmitotic NPC assembly with added fluorophore-labeled anti-Nup nanobodies. During interphase, NPCs are inserted into a pre-existing nuclear envelope. Monitoring this process is challenging because newly assembled NPCs are indistinguishable from pre-existing ones. We overcame this problem by inserting Xenopus-derived NPCs into human nuclear envelopes and using frog-specific anti-Nup nanobodies for detection. We further asked whether anti-Nup nanobodies could serve as NPC assembly inhibitors. Using a selection strategy against conserved epitopes, we obtained anti-Nup93, Nup98, and Nup155 nanobodies that block Nup-Nup interfaces and arrest NPC assembly. We solved structures of nanobody-target complexes and identified roles for the Nup93 α-solenoid domain in recruiting Nup358 and the Nup214·88·62 complex, as well as for Nup155 and the Nup98 autoproteolytic domain in NPC scaffold assembly. The latter suggests a checkpoint linking pore formation to the assembly of the Nup98-dominated permeability barrier.

A nanobody-based strategy for rapid and scalable purification of human protein complexes

The isolation of proteins in high yield and purity is a major bottleneck for the analysis of their three-dimensional structure, function and interactome. Here, we present a streamlined workflow for the rapid production of proteins or protein complexes using lentiviral transduction of human suspension cells, combined with highly specific nanobody-mediated purification and proteolytic elution. Application of the method requires prior generation of a plasmid coding for a protein of interest (POI) fused to an N- or C-terminal GFP or ALFA peptide tag using a lentiviral plasmid toolkit we have designed. The plasmid is then used to generate human suspension cell lines stably expressing the tagged fusion protein by lentiviral transduction. By leveraging the picomolar affinity of the GFP and ALFA nanobodies for their respective tags, the POI can be specifically captured from the resulting cell lysate even when expressed at low levels and under a variety of conditions, including detergents and mild denaturants. Finally, rapid and specific elution of the POI (in its tagged or untagged form) under native conditions is achieved within minutes at 4 °C, using the engineered SUMO protease SENPEuB. We demonstrate the wide applicability of the method by purifying multiple challenging soluble and membrane protein complexes to high purity from human cells. Our strategy is also directly compatible with many widely used GFP-expression plasmids, cell lines and transgenic model organisms. Finally, our method is faster than alternative approaches, requiring only 8 d from plasmid to purified protein, and results in substantially improved yields and purity.

Synapsin condensation controls synaptic vesicle sequestering and dynamics

Neuronal transmission relies on the regulated secretion of neurotransmitters, which are packed in synaptic vesicles (SVs). Hundreds of SVs accumulate at synaptic boutons. Despite being held together, SVs are highly mobile, so that they can be recruited to the plasma membrane for their rapid release during neuronal activity. However, how such confinement of SVs corroborates with their motility remains unclear. To bridge this gap, we employ ultrafast single-molecule tracking (SMT) in the reconstituted system of native SVs and in living neurons. SVs and synapsin 1, the most highly abundant synaptic protein, form condensates with liquid-like properties. In these condensates, synapsin 1 movement is slowed in both at short (i.e., 60-nm) and long (i.e., several hundred-nm) ranges, suggesting that the SV-synapsin 1 interaction raises the overall packing of the condensate. Furthermore, two-color SMT and super-resolution imaging in living axons demonstrate that synapsin 1 drives the accumulation of SVs in boutons. Even the short intrinsically-disordered fragment of synapsin 1 was sufficient to restore the native SV motility pattern in synapsin triple knock-out animals. Thus, synapsin 1 condensation is sufficient to guarantee reliable confinement and motility of SVs, allowing for the formation of mesoscale domains of SVs at synapses in vivo.

A selectivity filter in the ER membrane protein complex limits protein misinsertion at the ER

Tail-anchored (TA) proteins play essential roles in mammalian cells, and their accurate localization is critical for proteostasis. Biophysical similarities lead to mistargeting of mitochondrial TA proteins to the ER, where they are delivered to the insertase, the ER membrane protein complex (EMC). Leveraging an improved structural model of the human EMC, we used mutagenesis and site-specific crosslinking to map the path of a TA protein from its cytosolic capture by methionine-rich loops to its membrane insertion through a hydrophilic vestibule. Positively charged residues at the entrance to the vestibule function as a selectivity filter that uses charge-repulsion to reject mitochondrial TA proteins. Similarly, this selectivity filter retains the positively charged soluble domains of multipass substrates in the cytosol, thereby ensuring they adopt the correct topology and enforcing the "positive-inside" rule. Substrate discrimination by the EMC provides a biochemical explanation for one role of charge in TA protein sorting and protects compartment integrity by limiting protein misinsertion.

Molecular basis of RanGTP-activated release of Histones H2A-H2B from Importin-9

Imp9 is the primary importin for shuttling H2A-H2B from the cytoplasm to the nucleus. It employs an unusual mechanism where the binding of RanGTP is insufficient to release H2A-H2B. The resulting stable RanGTP·Imp9·H2A-H2B complex gains nucleosome assembly activity with H2A-H2B able to be deposited into an assembling nucleosome in vitro. Using hydrogen-deuterium exchange coupled with mass spectrometry (HDX), we show that Imp9 stabilizes H2A-H2B beyond the direct-binding site, like other histone chaperones. HDX also shows that binding of RanGTP releases H2A-H2B contacts at Imp9 HEAT repeats 4-5, but not 18-19. DNA- and histone-binding surfaces of H2A-H2B are exposed in the ternary complex, facilitating nucleosome assembly. We also reveal that RanGTP has a weaker affinity for Imp9 when H2A-H2B is bound. Imp9 thus provides a connection between the nuclear import of H2A-H2B and its deposition into chromatin.

Efficient in planta production of amidated antimicrobial peptides that are active against drug-resistant ESKAPE pathogens

Antimicrobial peptides (AMPs) are promising next-generation antibiotics that can be used to combat drug-resistant pathogens. However, the high cost involved in AMP synthesis and their short plasma half-life render their clinical translation a challenge. To address these shortcomings, we report efficient production of bioactive amidated AMPs by transient expression of glycine-extended AMPs in Nicotiana benthamiana line expressing the mammalian enzyme peptidylglycine α-amidating mono-oxygenase (PAM). Cationic AMPs accumulate to substantial levels in PAM transgenic plants compare to nontransgenic N. benthamiana. Moreover, AMPs purified from plants exhibit robust killing activity against six highly virulent and antibiotic resistant ESKAPE pathogens, prevent their biofilm formation, analogous to their synthetic counterparts and synergize with antibiotics. We also perform a base case techno-economic analysis of our platform, demonstrating the potential economic advantages and scalability for industrial use. Taken together, our experimental data and techno-economic analysis demonstrate the potential use of plant chassis for large-scale production of clinical-grade AMPs.

Genetic encoding of 3-nitro-tyrosine reveals the impacts of 14-3-3 nitration on client binding and dephosphorylation

14-3-3 proteins are central hub regulators of hundreds of phosphorylated "client" proteins. They are subject to over 60 post-translational modifications (PTMs), yet little is known how these PTMs alter 14-3-3 function and its ability to regulate downstream signaling pathways. An often neglected, but well-documented 14-3-3 PTM found under physiological and immune-stimulatory conditions is the conversion of tyrosine to 3-nitro-tyrosine at several Tyr sites, two of which are located at sites considered important for 14-3-3 function: Y130 (β-isoform numbering) is located in the primary phospho-client peptide-binding groove, while Y213 is found on a secondary binding site that engages with clients for full 14-3-3/client complex formation and client regulation. By genetically encoding 3-nitro-tyrosine, we sought to understand if nitration at Y130 and Y213 effectively modulated 14-3-3 structure, function, and client complexation. The 1.5 Å resolution crystal structure of 14-3-3 nitrated at Y130 showed the nitro group altered the conformation of key residues in the primary binding site, while functional studies confirmed client proteins failed to bind this variant of 14-3-3. But, in contrast to other client-binding deficient variants, it did not localize to the nucleus. The 1.9 Å resolution structure of 14-3-3 nitrated at Y213 revealed unusual flexibility of its C-terminal α-helix resulting in domain swapping, suggesting additional structural plasticity though its relevance is not clear as this nitrated form retained its ability to bind clients. Collectively, our data suggest that nitration of 14-3-3 will alter downstream signaling systems, and if uncontrolled could result in global dysregulation of the 14-3-3 interactome.

Molecular basis of RanGTP-activated nucleosome assembly with Histones H2A-H2B bound to Importin-9

Padavannil et al. 2019 show that Importin-9 (Imp9) transports Histones H2A-H2B from the cytoplasm to the nucleus using a non-canonical mechanism whereby binding of a GTP-bound Ran GTPase (RanGTP) fails to evict the H2A-H2B cargo. Instead, a stable complex forms, comprised of equimolar RanGTP, Imp9, and H2A-H2B. Unlike the binary Imp9•H2A-H2B complex, this RanGTP•Imp9•H2A-H2B ternary complex can release H2A-H2B to an assembling nucleosome. Here, we define the molecular basis for this RanGTP-activated nucleosome assembly by Imp9. We use hydrogen-deuterium exchange coupled with mass spectrometry and compare the dynamics and interfaces of the RanGTP•Imp9•H2A-H2B ternary complex to those in the Imp9•H2A-H2B or Imp9•RanGTP binary complexes. Our data are consistent with the Imp9•H2A-H2B structure by Padavannil et al. 2019 showing that Imp9 HEAT repeats 4-5 and 18-19 contact H2A-H2B, as well as many homologous importin•RanGTP structures showing that importin HEAT repeats 1 and 3, and the h8 loop, contact RanGTP. We show that Imp9 stabilizes H2A-H2B beyond the direct binding site, similar to other histone chaperones. Importantly, we reveal that binding of RanGTP releases H2A-H2B interaction at Imp9 HEAT repeats 4-5, but not 18-19. This exposes DNA- and histone-binding surfaces of H2A-H2B, thereby facilitating nucleosome assembly. We also reveal that RanGTP has a weaker affinity for Imp9 when H2A-H2B is bound. This may ensure that H2A-H2B is only released in high RanGTP concentrations near chromatin. We delineate the molecular link between the nuclear import of H2A-H2B and its deposition into chromatin by Imp9.

Significance

Imp9 is the primary importin for shuttling H2A-H2B from the cytoplasm to the nucleus. It employs an unusual mechanism where the binding of RanGTP alone is insufficient to release H2A-H2B. The resulting stable RanGTP•Imp9•H2A-H2B complex gains nucleosome assembly activity as H2A-H2B can be deposited onto an assembling nucleosome. We show that H2A-H2B is allosterically stabilized via interactions with both N- and C-terminal portions of Imp9, reinforcing its chaperone-like behavior. RanGTP binding causes H2A-H2B release from the N-terminal portion of Imp9 only. The newly-exposed H2A-H2B surfaces can interact with DNA or H3-H4 in nucleosome assembly. Imp9 thus plays a multi-faceted role in histone import, storage, and deposition regulated by RanGTP, controlling histone supply in the nucleus and to chromatin.

Tagging Recombinant Proteins to Enhance Solubility and Aid Purification

Protein fusion technology has had a major impact on the efficient production and purification of individual recombinant proteins. The use of genetically engineered affinity and solubility-enhancing polypeptide "tags" has a long history, and there is a considerable repertoire of these that can be used to address issues related to the expression, stability, solubility, folding, and purification of their fusion partner. In the case of large-scale proteomic studies, the development of purification procedures tailored to individual proteins is not practicable, and affinity tags have become indispensable tools for structural and functional proteomic initiatives that involve the expression of many proteins in parallel. In this chapter, the rationale and applications of a range of established and more recently developed solubility-enhancing and affinity tags is described.

MTCH2 is a mitochondrial outer membrane protein insertase

In the mitochondrial outer membrane, α-helical transmembrane proteins play critical roles in cytoplasmic-mitochondrial communication. Using genome-wide CRISPR screens, we identified MTCH2, and its paralog MTCH1, and showed that it is required for insertion of biophysically diverse tail-anchored (TA), signal-anchored, and multipass proteins, but not outer membrane β-barrel proteins. Purified MTCH2 was sufficient to mediate insertion into reconstituted proteoliposomes. Functional and mutational studies suggested that MTCH2 has evolved from a solute carrier transporter. MTCH2 uses membrane-embedded hydrophilic residues to function as a gatekeeper for the outer membrane, controlling mislocalization of TAs into the endoplasmic reticulum and modulating the sensitivity of leukemia cells to apoptosis. Our identification of MTCH2 as an insertase provided a mechanistic explanation for the diverse phenotypes and disease states associated with MTCH2 dysfunction.

We showed that MTCH2 was both necessary and sufficient for insertion of diverse α-helical proteins into the mitochondrial outer membrane, and was the defining member of a family of insertases that have co-opted the SLC25 transporter fold.

PEX5 translocation into and out of peroxisomes drives matrix protein import

Peroxisomes are ubiquitous organelles whose dysfunction causes fatal human diseases. Most peroxisomal enzymes are imported from the cytosol by the receptor PEX5, which interacts with a docking complex in the peroxisomal membrane and then returns to the cytosol after monoubiquitination by a membrane-embedded ubiquitin ligase. The mechanism by which PEX5 shuttles between cytosol and peroxisomes and releases cargo inside the lumen is unclear. Here, we use Xenopus egg extract to demonstrate that PEX5 accompanies cargo completely into the lumen, utilizing WxxxF/Y motifs near its N terminus that bind a lumenal domain of the docking complex. PEX5 recycling is initiated by an amphipathic helix that binds to the lumenal side of the ubiquitin ligase. The N terminus then emerges in the cytosol for monoubiquitination. Finally, PEX5 is extracted from the lumen, resulting in the unfolding of the receptor and cargo release. Our results reveal the unique mechanism by which PEX5 ferries proteins into peroxisomes.

Mechanical control of nuclear import by Importin-7 is regulated by its dominant cargo YAP

Mechanical forces regulate multiple essential pathways in the cell. The nuclear translocation of mechanoresponsive transcriptional regulators is an essential step for mechanotransduction. However, how mechanical forces regulate the nuclear import process is not understood. Here, we identify a highly mechanoresponsive nuclear transport receptor (NTR), Importin-7 (Imp7), that drives the nuclear import of YAP, a key regulator of mechanotransduction pathways. Unexpectedly, YAP governs the mechanoresponse of Imp7 by forming a YAP/Imp7 complex that responds to mechanical cues through the Hippo kinases MST1/2. Furthermore, YAP behaves as a dominant cargo of Imp7, restricting the Imp7 binding and the nuclear translocation of other Imp7 cargoes such as Smad3 and Erk2. Thus, the nuclear import process is an additional regulatory layer indirectly regulated by mechanical cues, which activate a preferential Imp7 cargo, YAP, which competes out other cargoes, resulting in signaling crosstalk.

The translation of mechanical cues into gene expression changes is dependent on the nuclear import of mechanoresponsive transcriptional regulators. Here the authors identify that Importin-7 drives the nuclear import of one such regulator YAP while YAP then controls Importin-7 response to mechanical cues and restricts Importin-7 binding to other cargoes.

Karyopherin-mediated nucleocytoplasmic transport

Efficient and regulated nucleocytoplasmic trafficking of macromolecules to the correct subcellular compartment is critical for proper functions of the eukaryotic cell. The majority of the macromolecular traffic across the nuclear pores is mediated by the Karyopherin-β (or Kap) family of nuclear transport receptors. Work over more than two decades has shed considerable light on how the different Kap family members bring their respective cargoes into the nucleus or the cytoplasm in efficient and highly regulated manners. In this Review, we overview the main features and established functions of Kap family members, describe how Kaps recognize their cargoes and discuss the different ways in which these Kap-cargo interactions can be regulated, highlighting new findings and open questions. We also describe current knowledge of the import and export of the components of three large gene expression machines - the core replisome, RNA polymerase II and the ribosome - pointing out the questions that persist about how such large macromolecular complexes are trafficked to serve their function in a designated subcellular location.

Targeted truncated TGF-β receptor type II delivery to fibrotic liver by PDGFβ receptor-binding peptide modification for improving the anti-fibrotic activity against hepatic fibrosis in vitro and in vivo

Truncated transforming growth factor-β receptor type II (tTβRII) is a promising anti-fibrotic candidate because it attenuates excessive transforming growth factor-β1 (TGF-β1) and then blocks TGF-β1 activity in hepatic fibrosis. However, its use has been greatly limited due to the fact that it is expensive to chemically synthesize and it does not specifically target to the lesion site. In this study, we describe that platelet- derived growth factor β receptor (PDGFβR)-binding peptide BiPPB modified tTβRII (BiPPB-tTβRII) was prepared from the cleavage of SUMO-BiPPB-tTβRII by digestion with SUMO-specific protease. Moreover, compared to the unmodified tTβRII, the target protein BiPPB-tTβRII not only highly specific targeted activated hepatic stellate cells (HSCs) and fibrotic liver tissue, but also significantly inhibited the protein levels of fibrosis-related genes in TGF-β1-induced HSC-T6 cells and CCl₄-induced liver fibrosis in mice. Furthermore, BiPPB-tTβRII markedly ameliorated liver morphology, fibrotic responses and the damage of liver function in fibrosis animal. More importantly, BiPPB-tTβRII showed a much lesser extent in binding to quiescent HSCs and non-fibrotic liver tissue. Taken together, our results suggested that the target protein BiPPB-tTβRII, with its high specific fibrotic liver-targeting potential and its improved anti-fibrotic activity in liver fibrosis, may be a potential therapeutic agent for liver fibrosis.

WNK1 is an assembly factor for the human ER membrane protein complex

The assembly of nascent proteins into multi-subunit complexes is a tightly regulated process that must occur at high fidelity to maintain cellular homeostasis. The ER membrane protein complex (EMC) is an essential insertase that requires seven membrane-spanning and two soluble cytosolic subunits to function. Here, we show that the kinase with no lysine 1 (WNK1), known for its role in hypertension and neuropathy, functions as an assembly factor for the human EMC. WNK1 uses a conserved amphipathic helix to stabilize the soluble subunit, EMC2, by binding to the EMC2-8 interface. Shielding this hydrophobic surface prevents promiscuous interactions of unassembled EMC2 and directly competes for binding of E3 ubiquitin ligases, permitting assembly. Depletion of WNK1 thus destabilizes both the EMC and its membrane protein clients. This work describes an unexpected role for WNK1 in protein biogenesis and defines the general requirements of an assembly factor that will apply across the proteome.

Structural basis for membrane insertion by the human ER membrane protein complex

A defining step in the biogenesis of a membrane protein is the insertion of its hydrophobic transmembrane helices into the lipid bilayer. The nine-subunit endoplasmic reticulum (ER) membrane protein complex (EMC) is a conserved co- and posttranslational insertase at the ER. We determined the structure of the human EMC in a lipid nanodisc to an overall resolution of 3.4 angstroms by cryo-electron microscopy, permitting building of a nearly complete atomic model. We used structure-guided mutagenesis to demonstrate that substrate insertion requires a methionine-rich cytosolic loop and occurs via an enclosed hydrophilic vestibule within the membrane formed by the subunits EMC3 and EMC6. We propose that the EMC uses local membrane thinning and a positively charged patch to decrease the energetic barrier for insertion into the bilayer.

Structural basis for the nuclear import and export functions of the biportin Pdr6/Kap122

Importins ferry proteins into nuclei while exportins carry cargoes to the cytoplasm. In the accompanying paper in this issue (Vera Rodriguez et al. 2019. J. Cell Biol. https://doi.org/10.1083/jcb.201812091), we discovered that Pdr6 is a biportin that imports, e.g., the SUMO E2 ligase Ubc9 while depleting the translation factor eIF5A from the nuclear compartment. In this paper, we report the structures of key transport intermediates, namely, of the Ubc9•Pdr6 import complex, of the RanGTP•Pdr6 heterodimer, and of the trimeric RanGTP•Pdr6•eIF5A export complex. These revealed nonlinear transport signals, chaperone-like interactions, and how the RanGTPase system drives Pdr6 to transport Ubc9 and eIF5A in opposite directions. The structures also provide unexpected insights into the evolution of transport selectivity. Specifically, they show that recognition of Ubc9 by Pdr6 differs fundamentally from that of the human Ubc9-importer Importin 13. Likewise, Pdr6 recognizes eIF5A in a nonhomologous manner compared with the mammalian eIF5A-exporter Exportin 4. This suggests that the import of Ubc9 and active nuclear exclusion of eIF5A evolved in different eukaryotic lineages more than once and independently from each other.