eL15 and eL22 proteins and their functional environments during ribosome assembly in Saccharomyces cerevisiae

eL15 and eL22 proteins and their functional environments during ribosome assembly in Saccharomyces cerevisiae D O C T O R A L T H E S I S 2 0 2 3 José Fernández Fernández Departamento de Genética Universidad de Sevilla eL15 and eL22 proteins and their functional environments during ribosome assembly in Saccharomyces cerevisiae Memoria presentada para aspirar al grado de Doctor en Biología Molecular, Biomedicina e Investigación Clínica V.° B.°: El director de la Tesis El doctorando Jesús de la Cruz Díaz José Fernández Fernández This work was funded by an FPI fellowship (BES-2017-080876), project grants from the Spanish Government (BFU2016-75352-P and PID2019-103850GB-I00) and the Andalusian Regional Government (P20_00581), and by an EMBO shortterm fellowship (8468). A Vane y Nico, que me robaron el corazón INDEX RESUMEN …………………………………..……………………………………………………………………………… 1 SUMMARY ……………………………………………………..………………………………………………………….. 5 INTRODUCTION ……………………………………………………………………………………………………….… 9 • Structure, function and evolution of the ribosome ………………………………………… 11 • Ribosomal proteins …………………………………………..………………………………..………… 14 • Ribosome biogenesis …………………………………………………………………………..……….. 18 • Processing of precursor rRNAs …………………………………………………..……….…………. 20 • Formation of the SSU processome or 90S pre-ribosomal particle ………………..… 23 • Disassembly of the SSU processome …………………………………………..…………………. 26 • Maturation of pre-40S ribosomal particles ……………………….…………………………… 27 • Assembly of pre-60S ribosomal particles ……………………………………………………….. 29 o Very early nucleolar assembly of pre-60S r-particles …………………………….…. 30 o Early nucleolar assembly of pre-60S ribosomal particles ……………….………… 31 o Late nucleolar/early nucleoplasm transition of pre-60S ribosomal particles ……………………………………………………………………………………………...… 34 o Late nucleoplasmic formation of pre-60S r-particles ……………………….……… 37 o Export of the pre-60S ribosomal particle to the cytoplasm ……………………... 40 o Cytoplasmic assembly of pre-60S ribosomal particles ……………………………… 41 OBJECTIVES ……………………………………………………………………………………………….…………….. 45 CHAPTER 1 – Role of the yeast ribosomal protein eL22 in ribosome biogenesis ……... 49 • INTRODUCTION ……………………………………………………………………………………….…… 51 • MATERIALS AND METHODS ……………………………………………………………….…………. 55 • RESULTS …………………………………………………………………………………………………..…… 65 o Loss of eL22 results in a mild growth defect that is exacerbated at low temperatures …………………………………………………………….………….. 65 4 indican que la presencia de eL22 es necesaria para el correcto procesamiento del prerRNA 27S. Además, hemos observado que eL22 juega un papel importante en el citoplasma, donde su presencia es probablemente necesaria para el reciclaje eficaz de los factores de ensamblaje Arx1 y Alb1. Con respecto a la proteína ribosómica eL15, en esta tesis se demuestra que esta proteína es esencial para la producción de ribosomas maduros y, por tanto, para la supervivencia y el crecimiento celular. En concreto, hemos demostrado que eL15 es necesaria para el correcto procesamiento de los pre-rRNAs 27SA3 y 27SB en el nucleolo. Esto a su vez se debe al papel de eL15 en la estabilización de partículas ribosómicas tempranas que permitan la unión estable de diversas proteínas ribosómicas de su propio grupo funcional (eL8, eL36 y eL13) y numerosos factores de ensamblaje, denominados factores A3 y B que son necesarios para el procesamiento de los pre-rRNAs 27SA3 y 27SB, respectivamente. 5 SUMMARY 6 7 Ribosomes are ribonucleoprotein complexes in charge of protein translation. Translation is an essential process in all organisms that involves decoding the information contained in mRNAs (messenger RNAs) to generate proteins. Similarly, the process of ribosome biogenesis represents a critical pathway for all cells. Ribosomes are comprised of two different subunits, one large and one small, also named 60S and 40S, respectively, in eukaryotes, among them Saccharomyces cerevisiae, which is the model organism employed in this thesis. As ribosome biogenesis is well conserved among different eukaryotes, S. cerevisiae is a perfectly appropriate organism for this study. Ribosomes of S. cerevisiae are composed of 79 ribosomal proteins and 4 ribosomal RNAs (rRNAs), three of them in the large (25S, 5.8S y 5S) and one in the small ribosomal subunit (18S). The biogenesis of ribosomes is an extraordinarily complex process, in which in addition of the four rRNAs and the 79-80 ribosomal proteins, take part about 100 small nucleolar RNAs (snoRNAs) and more than 300 assembly protein factors, which are not present in the mature subunits and therefore are also known as trans-acting factors. The synthesis of ribosomes is a sequential process highly regulated and organized that starts in the nucleolus and orderly progresses through the nucleoplasm; almost mature ribosomal particles are exported to the cytoplasm, where the last steps of maturation take place, previously to their involvement in the translation process. The role of the different components participating in the biogenesis of the ribosomal subunits has been broadly studied through last decades. However, there are numerous assembly factors and some ribosomal proteins which still await characterization. This was the case for the ribosomal proteins eL22 and eL15 during the assembly of the 60S ribosomal subunit before the study performed in this doctoral thesis. With respect to the eL22 ribosomal protein, in this thesis, we have observed that, although, in laboratory conditions, this protein is not essential for survival, cell growth and ribosome production are mildly affected in strains lacking eL22. In this context, we have proved that the synthesis of ribosomes lacking of eL22 is partially blocked in the nucleus. Thus, our results indicate that the presence of eL22 is necessary for the correct processing of the 27S pre-rRNAs. Moreover, we have observed that eL22 performs an 8 important role in the cytoplasm, where its presence is likely needed for the efficient recycling of assembly factors Arx1 and Alb1. With respect to the eL15 ribosomal protein, in this thesis, we have shown that this protein is essential for the production of mature ribosomes and, thus, for the cell survival and cell growth. Specifically, we have demonstrated that eL15 is necessary for the precise processing of the 27SA3 and 27SB pre-rRNAs in the nucleolus. This is due to the role of eL15 in the stabilization of early ribosomal particles which permits the stable assembly of proteins of its own functional cluster (eL8, eL36 y eL13) and association of numerous assembly proteins, known as A3- and B-factors, which are necessary for the processing of the 27SA3 and 27SB pre-rRNAs, respectively. 9 INTRODUCTION 10 11 Structure, function and evolution of the ribosome The ribosome is the macromolecular ribonucleoprotein complex responsible for catalysis of the translation process. During translation, ribosomes decode the information contained in messenger RNAs (mRNAs) in order to synthesize the corresponding proteins. Therefore, ribosomes are essential players in the process of gene expression [1]. Ribosomes are constituted of 54-80 ribosomal proteins (r-proteins) and 3-4 specific RNAs known as ribosomal RNAs (rRNA) distributed in two ribosomal subunits (rsubunits), a large subunit (LSU) and a small one (SSU). These subunits are named according to their sedimentation coefficient; the LSU is termed 50S in prokaryotes and 60S in eukaryotes. In turn, the SSU is named 30S in prokaryotes and 40S in eukaryotes. Both subunits together form the 70S ribosome in prokaryotes and the 80S in eukaryotes [2,3]. There are also specific ribosomes in mitochondria and chloroplasts [4,5]. Both subunits exert different functions; the LSU catalyses the formation of peptide bonds in the peptidyl-transferase centre (PTC), while the SSU is responsible for decoding mRNAs by base pairing of their codons with the transfer RNAs (tRNAs) in the decoding centre [6]. Translation initiates when an SSU, an LSU and an mRNA come together building an active ribosome. In the subunit interface, the A (acceptor), P (peptidyl) and E (exit) sites are formed. The A-site is the first binding site for tRNAs in the ribosome while the P-site is the second and the E-site the last one. The A-site is the place in the ribosome where the charged aminoacyl-tRNAs enters. The P-site holds a tRNA with the growing polypeptide chain for peptide bond formation. The E-site is the site at which the deacylated tRNAs bind before they dissociate from the ribosome. The mRNA moves along the groove of the SSU as it is being decode while the nascent protein in formation exits through the polypeptide exit tunnel (PET) that is located in the LSU [6]. In this work, we have used the budding yeast Saccharomyces cerevisiae as a model organism. This is an eukaryotic organism with the advantages of low complexity, fast growth and available powerful biochemical, genetic and molecular tools. The process of ribosome biogenesis has been studied in several other eukaryotes, including humans, but most of our current knowledge in this process comes from yeast. In S. 12 cerevisiae, the LSU is composed of 46 r-proteins and 3 rRNAs named 25S, 5.8S and 5S. The SSU is composed of 33 r-proteins and one rRNA named 18S [3]. Regarding the structure of ribosomes, a repertoire of techniques such as cryogenic electron microscopy (cryo-EM) and X-ray crystallography has provided the field with detailed models at atomic or subatomic resolution (Figure 1). Although rRNAs and r-proteins are distributed all intertwined along the r-subunits, proteins are more commonly situated at the solvent exposed surface while rRNAs are mainly in the subunit interface, where play crucial role in the functional centres. Several structural domains can be highlighted in the r-subunits (Figure 1). In the SSU, it can be distinguished the domains named beak, head, platform and body. The groove of mRNA is formed between the head and the body and platform, where the mRNA moves while is being decoded. In the LSU, the main domains are the central protuberance (CP), where is situated the PTC, the L1-stalk and the P-stalk [7-9]. Figure 1. Architecture of the yeast r-subunits. The figure shows the small r-subunit or SSU (left) and the large r-subunit or LSU (right). The model shows the location of the different rproteins. Main structural features are highlighted: head, beak, platform and body in the small r-subunit; central protuberance (CP), L1-stalk and P-stalk in the large r-subunit. (A) Interface view. (B) Solvent exposed view. Adapted from [10]. 13 Ribosomes are well conserved in evolution. However, it can be appreciated an increase in complexity and size from prokaryotes to eukaryotes (Figure 2). There is a conserved common core, which involves the functional centres and that corresponds mainly to the prokaryote ribosome. Over this core, evolution has added a "shell", layer that is formed by some expansion segment (ES) of rRNA, r-proteins, and eukaryotespecific r-protein extensions. This layer is distributed towards the periphery of the ribosome, mainly in the solvent exposed surface. Although the biological function of rRNA ESs is not fully understood, their accessibility on the surface of the LSU raises the possibility that they could play a role in the recruitment of ribosome assembly factors and r-proteins during assembly. Most of the r-protein extensions interact with other eukaryote-specific r-proteins and rRNA ESs, suggesting that extensions of the conserved proteins may facilitate the assembly of eukaryote-specific ribosome components. These specific elements also provide new interactions with elements involved in translation; indeed, translation initiation presents several differences between kingdoms and therefore, many eukaryotic specific components have been found to participate in this process [7,11]. Figure 2. Comparative composition of bacterial and eukaryotic ribosomes. The common core between bacteria and eukaryotes is shown in light red (proteins) and light blue (rRNA). Both, bacteria and eukaryotes contain specific r-proteins and/or extensions and insertions in those r-proteins that are conserved (both in red). The expansion segments of rRNA are shown in blue. Dashed lines around the common core represent the position of the stalks that are used to be flexible in X-ray structures. Adapted from [11]. 20 results in a group of inherited pathologies known as ribosomopathies, which are linked to mutations affecting r-proteins and rRNA processing and assembly factor genes, or regulatory elements. In addition, mutation and misregulation of r-proteins genes are common features in several types of cancer [42,43]. Figure 5. Regulation of ribosome biogenesis in S. cerevisiae. The TORC1 and PKA pathways regulate the synthesis of ribosomes. These signalling pathways control the transcription levels of r-proteins, protein assembly factors and rDNA. Post-transcriptionally, they also regulate pre-rRNA processing, ribosome assembly, translation and degradation at different levels. Besides, the process itself provides feedback regulation at different steps during the expression of pre-rRNAs, r-proteins and assembly factors. Adapted from [41]. Processing of precursor rRNAs The yeast 35S pre-rRNA is the first detectable precursor transcribed by RNAPI. This harbours internal transcribed spacers (ITS1 and ITS2) and external transcribed spacers (5' ETS and 3' ETS). The 5S rRNA is transcribed by RNAPIII as a short precursor extended few nucleotides in its 3' end. These precursors (pre-rRNAs) undergo a complex stepwise process of modification and processing to produce the mature 18S, 5.8S, 25S and 5S rRNAs. The modification and processing reactions occur in the context of preribosomal particles on the road of maturation [44]. Figure 6 summarises the primary structure of the 35S and pre-5S pre-rRNAs and the different pre-rRNA processing steps in yeast. 21 Figure 6. rRNAs maturation pathway. (A) Representation of a repeat unit of rDNA containing the two independently transcribed genes, the 35S rDNA is transcribed by RNA polymerase I into a polycistronic pre-rRNA, which encodes the 18S, 5.8S and 25S rRNAs whereas the 5S rDNA is transcribed by RNA polymerase III. Internal and external spacers (ETSs and ITSs) are indicated. Mature rRNAs are represented with bars and spacers with lines. The start of transcription is indicated by red arrows and the location of different processing sites are shown. (B) Pre-rRNA processing pathway. The 35S transcript is subjected to co- or post-transcriptional processing generating the 20S and 27SA2 pre-rRNAs, which are part of pre-40S or pre-60S r-particles, respectively. These pre-rRNAs are processed into the mature 18S, 5.8S and 25S rRNAs. The 5S pre-rRNA is processed into the mature 5S rRNA, which is a component of the LSU. Different steps, pre-rRNAs intermediaries and nucleases involved are indicated. Note that some cleavages still remain unassigned to specific nucleases. Adapted from [44]. 22 Regarding pre-rRNA processing, the 35S pre-rRNA could be processed co- or post-transcriptionally. In rapidly growing cells, about 70% of the 35S pre-rRNA processing occurs in a co-transcriptional manner [45]. The 35S pre-rRNA processing initiates in the context of the particle called SSU processome. The first reactions remove the 5' ETS segment by cleavage at site A0 and immediately after, at site A1. It is not known the nuclease responsible for the cleavage at site A0 but Utp24 has been suggested to be the endonuclease that cleaves at sites A1 and A2, although some authors had previously given this role to the Rcl1 endonuclease [46-48]. At its 3' end, processing of 35S prerRNA is poorly understood but seems to occur concomitantly to the processing reactions at sites A0 and A1 when processing is occurring post-transcriptionally. It seems that the endonuclease Rnt1 cleaves at the site B0 and then, the 3'-5' exonuclease Rex1 trims from this site to the site B2, which is the 3' end of mature 25S rRNA [49]. The cleavage at site A1 is coupled and triggers that at site A2 to generate the 20S and the 27SA2 pre-rRNAs [47]. This cleavage permits the separation of the SSU processome into the pre-40S and pre-60S r-particles that will undergo apparently independent maturation. Cleavage at site A2 could occur co- or post-transcriptionally, so in the case of the co-transcriptional process, the early pre-60S r-particles is formed on nascent 27S pre-rRNA. In the cytoplasm, the 20S pre-rRNA is cleaved at site D to produce the mature 18S rRNA by the endonuclease Nob1 [50,51]. This is one of the last steps of maturation of pre-40S particles and occurs in the context of proofreading 80S-like particles, which are non-translating ribosomal particles formed by the association between a pre-40S r-particle and a mature LSU [52-54]. In slowly growing cells, cleavage at site A3 is favoured versus cleavage at site A2. This step is carried out by the MRP endonuclease and results in the formation of a 23S pre-rRNA. Whether or not processing of 23S pre-rRNA is productive (leading to a functional 18S rRNA) is still controversial [44,55]. The processing of the 27A2 pre-rRNA can be achieved in two ways. In about 85% of cases, this precursor is cleaved by MRP in the A3 site, producing 27SA3 pre-rRNA that is then trimmed until the B1s site to generate the 27SBs pre-rRNA by 5'-3' exonucleases Rrp17, Rat1 and Xrn1 [56-58]. In the remaining 15% of the cases, 27SA2 pre-rRNA is cleaved by a still unknown endonuclease at the B1L site producing the 27SBL pre-rRNA. 23 The two isoforms of 27SB pre-rRNAs are processed in an identical way, being cleaved at the C2 site in the middle of ITS2 by the endonuclease Las1. This reaction separates the 27SB pre-rRNAs in 25.5S and 7SS or 7SL pre-rRNAs [59,60]. In the nucleus, the 25.5S pre-rRNA is 5'-3' trimmed to mature 25S rRNA by the exonucleases Rrp17, Rat1 and Xrn1. In contrast, 7SL/S pre-rRNAs are processed in a sequential and multistep process. First, the 3' end of 7SL/S pre-rRNAs is trimmed in the nucleus by an exonucleolytic multi-protein complex named exosome to generate first the 5.8SL/S+30 pre-rRNAs and then the 5.8SL/S+8 pre-rRNAs, which are better known as 6SL/S pre-rRNAs. In the cytoplasm, the Rex1-3 exonucleases trim the 6SL/S pre-rRNAs generating the 5.8S L/S+5 pre-rRNAs, which are finally processed by the nuclease Ngl2 into mature 5.8SS or 5.8SL rRNAs [60-64]. On the other hand, the pre-5S rRNA is only subjected to minor processing. It is trimmed at the 3' end by 3'-5' exonucleases Rex1-3 to produce the mature 5S rRNA [65]. Concomitantly to processing, specially at very early steps, the pre-rRNAs are subjected to an extensive process of modification. Most of these modifications are conversions of uridines into pseudouridines and 2'-O-ribose methylations [66,67], which are executed by specific enzymes (Cbf5 and Nop1, respectively) that are specifically guided to their targets by a subset of about 80 snoRNAs. These snoRNAs are in complex with proteins forming small nucleolar ribonucleoproteins (snoRNPs). However, some other modifications are independent of snoRNAs. Interestingly, modifications are clustered in strategical regions of the rRNAs comprising the functional centres in the subunit interface; this fact clearly suggests that rRNA modifications play important role during translation [68,69]. Formation of the SSU processome or 90S pre-ribosomal particle Co-transcriptional assembly of 35S pre-rRNA forms the primordial SSU processome, also known as 90S pre-ribosomal particle. Advances in cryo-EM have identified different r-particles that are typically referred as 90S, pre-40S or pre-60S rparticles, leading to the wrong idea of static snapshots and well-defined complexes. However, each of them encompasses a wide variety of maturation states. These states 24 present differences in the presence/absence of r-proteins and, which is more informative, protein assembly factors. These factors have been broadly identified in systematic compositional studies by co-immunoprecipitation with different r-proteins and selected assembly factors. In turn, the information obtained in these compositional studies has been the base to perform further cryo-EM studies of selected complexes, whose comparison has permitted to deeply know the structure and sequence of the maturation events. However, it has to still be considered that the study of these static particles has allowed to deduce certain dynamics of ribosome maturation. The SSU processome is a cryo-EM particle that includes 5' ETS and a pre-rRNA precursor fragment containing 18S rRNA. The rest of the regions of the 35S pre-rRNA were not resolved in the structure as likely they are still flexible and unstructured at this point. In this particle, the sequence comprising 18S rRNA is organised in four topological domains: 5', central, 3' major and 3' minor domains, which are formed in a hierarchical 5' to 3' manner (Figure 7). Prior folding, 35S pre-rRNA must be extensively modified by pseudouridylation and 2'-O-ribose methylation reactions, which are carried out by selected snoRNPs. These complexes must be released for further folding of the pre-rRNA and optimal r-protein assembly. Release of snoRNPs involves several enzymes including RNA helicases [40,70]. Figure 7. Arrangement of the 18S rRNA domains in the structured pre-rRNA in the SSU processome. Relative orientation of the different 18S rRNA domains in the yeast SSU processome and in the cytoplasmic pre-40S particle. The 5', central, 3' major and 3' minor domains and the U3 snoRNA are indicated. Adapted from [38]. 25 The analysis of the SSU processome structure indicates the formation of different sub-complexes. The transcription of the 5' ETS permits the recruitment of two large modules: UTP-A (Utp4, Utp8, Utp9, Utp10, Utp15, Pol5 and Nam1) and UTP-B (Utp6, Utp13, Utp18, Utp21, Dip2 and Pwp2) sub-complexes. These modules form an scaffold where the nascent particle can continue formation by the recruitment of the U3 snoRNP, the Mpp10 complex (Mpp10-Imp3-Imp4) and other factors such as Utp3 [71-74]. After this, each domain of the 18S rRNA is independently associated with a subset of rproteins and assembly factors (Figure 8). At this moment, the U3 snoRNP plays a crucial role as central structural organiser. U3 snoRNP seems not to display any modification function, although the methyltransferase Nop1 is part of this complex. Instead, U3 snoRNP makes interactions with the UTP-A and UTP-B modules and other factors and contacts with the 5' ETS and regions of 18S rRNA within 35S pre-rRNA to bring them together. Besides, U3 snoRNP forms an heteroduplex with 18S rRNA, protecting the site A1 from being prematurely cleaved. By this way, U3 snoRNP locks the 90S pre-ribosomal particle in an intermediary state, permitting ulterior events to take place [48,75,76]. Figure 8. Structure of the SSU processome. (A) Two views of the cryo-EM model of the SSU processome from Chaetomium thermophilum with most identified factors, r-proteins and pre-rRNA in their assigned location. (B) Simplified representation of the SSU processome highlighting the main modules of pre-rRNA differently coloured, as well as the position of the UTP-A and UTP-B complexes, the U3 snoRNP and the 5' ETS region of 35S pre-rRNA. Adapted from [75]. 26 Disassembly of the SSU processome The SSU processome is not just a static particle. Instead, the earliest particle must go through a complex assortment of maturation states to allow cleavage at the A1 site. These steps require folding and high compaction of the pre-rRNA (Figure 7) [77]. Among the different structural rearrangements compacting the pre-rRNA is included the formation of the so-called H1 stem loop in the 5' domain of the nascent 18S rRNA. The formation of this loop allows the A1 site to get close to the endonuclease Utp24, thus favouring its cleavage. After this reaction, the SSU processome is dismantled, however, this does not occur as a block event but in a stepwise manner (Figure 9). The accepted model suggests that first the Sfa1 module and Utp6 are removed, later UTP-A, U3 snoRNP and part of UTP-C, and finally the UTP-B and the remaining UTP-C factors [48]. The 5' ETS region of 35S pre-rRNA is further processed by the exosome, whose binding to the particle provides the remodelling force necessary for the release of this 5' ETS region and most assembly factors that will be recycled to form a new SSU processome [78]. Once Utp24 has cleaved at the site A1, the partial disassembly of the particle permits this nuclease to reach the A2 site in the flexible ITS1 region. Cleavage at site A2 leads to the disassembly of the rest of the SSU processome leading to the generation of the initial pre-40S particles [48]. As the SSU processome does not contain components of the LSUs, it is assumed this structure correspond to ribosomes being matured through the co-transcriptional processing of pre-rRNA, thus, the corresponding initial pre-60S particle will be formed on the nascent 27S pre-rRNA. Recently, the isolation of a primordial pre-60S particle, composed by a specific set of assembly factors and snoRNAs has been reported. This particle, which still lacks of the enough atomic resolution, seems to reflect a 90S particle with a bipartite pre-40S-pre-60S structure, thus, being related to post-transcriptional processing of pre-rRNA [79]. 27 Figure 9. Dismantling of the yeast SSU processome. The figure shows the current model for the conversion of the SSU processome in an early pre-40S intermediate based on the structure of different identified cryo-EM particles. Top panel: Cryo-EM maps of distinct states of SSU processome to pre-40S particle transition. The main assembly factors and modules are indicated. Changes in the different proposed steps are depicted. Middle panel: Different view of the top panel, showing the molecular models and the resolution obtained for the different particles. Bottom panel: Representation of the pre-18S rRNA density within the models. The different domains of the 18S rRNA, represented as its secondary structure, are shown in the boxes. Adapted from [48]. Maturation of pre-40S ribosomal particles Once the SSU processome has been disassembled, the primordial pre-40S rparticle emerges. This earliest pre-40S r-particle contains a defined composition, incorporating some additional assembly factors such as Tsr1, Rio2, Ltv1 or Dim1 and keeping other factors such as Pno1, Enp1 or Rrp12 that were already present in the 90S r-particle. Differently to the complex pre-60S r-particle maturation, the pre-40S rparticle shows a conformation almost close to the mature state that does not need extensive rearrangements. Soon after formation, nucleoplasmic pre-40S r-particles are rapidly exported to the cytoplasm where they acquire translation competence [80]. In early nucleoplasmic pre-40S r-particles, the 5' and central domains of the prerRNA are in a conformation almost mature. However, the 3' domains still need of certain remodelling, mainly related to the head formation and the beak completion. Once the 28 head is formed, it presents a highly rotated state when compared to mature SSU and only acquires its mature conformation in final steps in the cytoplasm. Functional sites as the mRNA binding groove and the decoding centre are not constructed yet. Instead, some factors as Tsr1, Enp1, Rio2, and Pno1 control their maturation, preventing premature formation until particles reach the cytoplasm (Figure 10) [81-83]. Figure 10. Structure of the cytoplasmic pre-40S particle. Representation of the structure of the pre-40S particle from the intersubunit interface (left) and solvent exposed interface (right). The 20S pre-rRNA is shown in grey and the r-proteins in light blue. Tsr1, Rio2, Pno1, Enp1 and Ltv1 are highlighted in different colours. The different structural domains, typical of SSUs, are also indicated. Adapted from [83]. Transport of the pre-40S r-particles from the nucleus to the cytoplasm through the nuclear pore complex (NPC) is facilitated by several and overlapping mechanisms. This redundancy provides efficiency for a rapid export. Different assembly factors such as Rio2, Pno1 and Ltv1 harbour nuclear export signals (NESs) that are recognized by selected exportins. Ltv1 and Rio2 are associated to nuclear pre-40S r-particles while Pno1 is present in the 90S particle but in a conformation that makes its NES still not accessible, thus, preventing premature export [80]. The most relevant exportin is Crm1 that binds to NES containing proteins, Ran-GTP and nucleoporins [80,84]. Besides, the heterodimer formed by Mex67 and Mtr2 and the Rrp12 factor, which are also important during export of pre-60S r-particles, also play a role in the export of pre-40S r-particles [85,86]. 29 Once pre-40S r-particles reach the cytoplasm, the 3' major domain still requires maturation, in part due to the presence of Enp1 and Ltv1. These factors occupy the binding site of the eS10 r-protein in the still immature beak and stabilize the pre-rRNA in a conformation that impedes the stably association of the uS3 r-protein. Therefore, the correct association and completion of the beak depends on the release of Enp1 and Ltv1 in the cytoplasm [82,83]. Furthermore, the dissociation of these factors is thought to trigger several additional conformational changes in the 3' major rRNA, such as the head rotation. Besides, the presence of ribosomal assembly factors in cytoplasmic pre- 40S particles has been postulated to be a quality control mechanism that impedes immature particles to engage prematurely in translation [82,83,87]. The head rotation induces the release of Rio2 and Tsr1 from the functional sites [83]. Then, Rio1 binds to the particle in a similar way than Rio2 and mediates the formation of the so-called 80S-like particle, which involves a transient union with a mature LSU. The formation of the 80S-like particle has been interpreted as part of a proofreading mechanism to ensure that only properly matured SSUs gain translation competence [52,88,89]. Furthermore, the release of Rio2 and Tsr1 results in some conformational changes that ultimately generates the displacement of Pno1 that releases the site D in 20S pre-rRNA to allow the endonucleolytic reaction by Nob1 to form the mature 18S rRNA [82,83]. Assembly of pre-60S ribosomal particles The maturation of LSUs is much more elaborated than that of SSUs as the subunits themselves are also more complex. Within LSUs, the 25S rRNA forms six RNA domains (I-VI) of secondary structure that are more intertwined that those present in the 18S rRNA [3,90]. In general, domains I, II and the 5.8S rRNA form the solvent exposed face of the LSU while domains IV and V form the subunit interface. Domains III and VI mainly act as bridges between the former ones. The 5S rRNA docks to domains II and V. The assembly of different LSU domains does not follow a clear 5' to 3' pattern as occurring for those of SSU, although certain 5' to 3' order can be envisaged (Figure 11) [38]. As for SSUs, the incorporation of the different r-proteins to pre-60S r-particles 36 At this maturation state, the pre-rRNA region that will form the future L1-stalk starts its definition by interactions with a set of factors included Ebp2, Nip7, Noc3, Nsa2 and Nop2. Furthermore, this pre-rRNA region of the future L1-stalk is displaced far from its final position and located at the top of the particle occupying the anchor site of the 5S-RNP and avoiding the formation of the CP. Although the 5S-RNP had already been attached to domain V at earlier stages, it still remains flexible (Figure 11 and 14) [92,121,123]. Figure 14. Rea1-dependent nuclear maturation events of pre-60S r-particles. The ATPase mediates two different reactions. First, Rea1 allows the release of Erb1-Ytm1 to permit the binding of Nop53, which eventually leads to the foot and ITS2 removal. Release of Erb1 induces the release of factors, step that is related to the stabilisation and conformation of the L1-stalk. The 5S RNP can stably bind to the core particle as part of these reactions. Second, Rea1 induces the rotation of 5S RNP within the central protuberance, its own release and that of Rsa4. Adapted from [123]. Following maturation, the large AAA ATPase Rea1 binds to Ytm1 and its ATPase activity induces the release of this factor together with Erb1, which is its interaction partner. The release of Erb1, which has allowed the earlier establishment of multitude of contacts within the pre-60S r-particles, seems to destabilise several local regions of the particles triggering novel remodelling events; these lead to the release of Has1, 37 Nop16, Nip7, Nop2, Noc3, Brx1 and Ebp2 and the relevant re-accommodation of the N- terminal extension of Rlp7 in an alternative position [92,93,111,112,124]. This allows the association of the Nop53 factor to ITS2 next to Nop7 at the foot. The association of Nop53 is a key event for later nucleoplasmic removal of ITS2 [92,115,125]. At this point, the release of Spb1 permits the L1-stalk to acquire a mature like conformation and to attach the particle. Concomitantly the r-proteins uL2 and eL43 associate and fix the L1- stalk (Figure 14) [123]. Late nucleoplasmic formation of pre-60S r-particles As above mentioned, cryo-EM analysis has deduced that the release of Erb1 triggers an important sequence of remodelling events to the nucleoplasmic pre-60S particles. At the late nucleoplasmic states, the subunit interface progresses with the compaction of domain IV and the association of factors as Cgr1, Nog2 and Rsa4. The PET completes its maturation with the assembly of the eL39 r-protein. Interestingly, at this stage, but not before, the C-terminal extension of Nog1 totally occupy the pipe of PET and the heterodimer formed by Arx1 and its interaction partner Alb1 binds obstructing the exit of PET [92,115,123,126]. The dynamics of association of the C-terminal extension of Nog1 with the pre-60S r-particles suggests multiple functions for Nog1. First, Nog1 can act as a scaffold to form the tunnel. Second, Nog1 can probe the proper formation of the PET. Third, Nog1 can occupy the tunnel to avoid the entry of other proteins and prevents premature translation. Furthermore, as the presence of Nog1 is necessary for the cleavage at site C2, it is probably that the appropriate formation of the PET would be a pre-requisite for ITS2 removal [127,128]. As discussed later, the pipe of PET will be occupied by other factors in cytoplasmic pre-60S r-particles (Figure 17) [126,129,130]. At this stage, the stably assembly of the 5S rRNP complex occurs. In yeast, 5S rRNA first assembles as a complex with the r-proteins uL5 and uL18. This 5S RNP is bound in block to the still flexible domain V at an early stage in the nucleolus by the anchor function of Rpf2 and Rrs1 assembly factors but remains undocked to the core pre-60S rparticle until later maturation stages [131]. Syo1 is the dedicated chaperone for uL18 38 that escorts the 5S rRNP to the nucleus and helps in its initial assembly too [25,132,133]. Once in the nucleoplasm and after stabilisation of the L1-stalk, the 5S RNP can stably bind to the pre-60S r-particle in the base of this stalk to constitute the main component of the CP. However, in this initial conformation, the 5S RNP presents a 180° rotated conformation in comparison to the position it adopts in the mature LSU, thus, needing rotation prior to pre-60S r-particles export to the cytoplasm (Figure 14). This step is coupled to the removal of the Rix1 complex and the Rsa4 factor by the ATPase Rea1 [92,121,123]. One of the last steps in the nucleoplasmic maturation of pre-60S r-particles involves the ATPase Rea1 and the Rix1 complex, whose association to the pre-60S rparticles triggers a sequence of events that ultimately induces ITS2 removal. Rea1, in a new round of action, binds now to its second target Rsa4. Moreover, the Rea1- interacting Rix1 complex also associates at this late nucleoplasmic state to the pre-60S r-particle. The Rix1 complex situates on top of L1-stalk and the Sda1 assembly factor, next to the CP (Figure 14). This positioning of Rea1-Rix1 also triggers the release of Rpf2 and Rrs1 and the concomitant rotation of the CP. After that, the ATPase activity of Rea1 induces the release of Rsa4 and the own Rea1-Rix1 machinery. Then, the assembly of uL2 and eL43 r-proteins promotes the completion and stabilization of the CP. As the CP rotation precedes ITS2 processing and the foot structure removal, it is possible that this rotation somehow induces the subsequent cleavage at C2 site [121,123,134]. This prerRNA processing step is carried out by the Las1 complex, which is formed by the endonuclease Las1, the kinase Grc3, the exonuclease Rat1 and its cofactor Rai1. After CP rotation, Grc3 drives Las1 to site C2, then Las1 cleaves at this site generating the 7S and 25.5S pre-rRNAs. After that, Grc3 phosphorylates the 5' OH of the 25.5S pre-rRNA which enables the trimming of this pre-rRNA by Rat1 (Figure 15) [59,135-138]. Two other 5' to 3' exonucleases, Rrp17 and Xrn1, are also capable to trim the 25.5S pre-rRNA to mature 25S rRNA [56,57]. On the other hand, once cleavage at site C2 has occurred, Nop53 could recruit the exosome through its Mtr4 cofactor to the particle [125,139-141]. Mtr4 is an RNA helicase which forms part of the Trf4/5-Air1/2-Mtr4 polyadenylation complex (TRAMP). The polyadenylation is thought to provide a grip for Mtr4 to unwind ITS2 rRNA and 39 maybe for restructuration of some proteins to make the ITS2 accessible for the exosome [142-145]. The exosome acts in two steps. First, it removes all but the last 30 nucleotides of ITS2 from the 3' end of the 7S pre-rRNAs. In a second step, it degrades all but the last 8 nucleotides producing the 6S pre-rRNAs. The 6S pre-rRNAs will be finally processed in the cytoplasm to mature 5.8S rRNA. Just before or during the exosome binding, the foot is completely remodelled and only Nop7 remains stably associated (Figure 15) [61,146]. Figure 15. Removal of ITS2 from the pre-60S r-particles. The endonuclease Las1 cleaves the ITS2 at site C2, which generates 7S and 25.5S (also 26S) pre-rRNAs from 27SB pre-rRNA. The 25.5S pre-rRNA is 5' to 3' trimmed by the exonuclease Rat1. Nop53 recruits the exosome by mean of the Mtr4 cofactor to allow 3' to 5' processing of 7S pre-rRNAs. Concomitantly to these steps, the assembly factors of the foot are released from particles. Adapted from [77]. 40 Export of the pre-60S ribosomal particle to the cytoplasm The nucleoplasmic pre-60S r-particle needs to get remodelled before gaining export competence. Thus, one of these steps is the release of Nog2. This GTPase act as a hub protein situated at the centre of the immature PTC interacting with several factors as Nog1, Nsa2, Rpf2, Rsa4, Bud20 or Dbp10 and with pre-rRNA from domains II and IV. The association of Nog2 with the PTC suggests a role as a controller of its maturation [126,147]. The release of Nog2 is dependent of two activities, the ATPase activity of Rea1 and the GTPase activity of the own Nog2 [148]. Nog2 deeply interacts with the Ipi1 component of the Rix1 complex and with the Nsa2 factor (Figure 14). As the release of Rsa4 and the Rea1-Rix1 machinery occurs before of that of Nog2, it is likely that this step, in a direct or indirect manner, triggers the activation of the GTPase activity of Nog2 and its consequent release [123]. Curiously, the binding site of Nog2 in pre-60S rparticles overlaps with that of the export factor Nmd3, so the latter can only bind after the release of the former. By this way, Nog2 release functions as a checkpoint for acquiring export competence [130,148,149]. There are several other factors involved in the transport of the pre-60S particle to the cytoplasm, among them the Mex67-Mtr2 complex, which has some overlapping with Mrt4, the placeholder of uL10 r-protein [150], the dimer Arx1-Alb1, Rrp12, Bud20, Npl3, Ecm1 and Gle2 (Figure 16) [151-154]. However, Nmd3 is the main responsible for the export, whereas the rest of transport factors increases the efficiency of this process. Nmd3 is the only NES containing adaptor; the NES of Nmd3 is essential for nuclear export, which is dependent of the exportin Crm1 [151]. Arx1-Alb1, Bud20 and Ecm1 are thought to directly interact with nucleoporins to facilitate the translocation to the cytoplasm [155-158]. Npl3 interacts with nucleoporins and it is also thought to shield the particle from the hydrophobic environment by interacting with uL23 r-protein and pre-rRNA [159]. Gle2 mediates the export by interacting specifically with the nucleoporin Nup116 [160]. Mex67-Mtr2 form a dimer that directly interacts with nucleoporins and together with Rrp12 are the only adapter factors shared with pre-40S r-particles [85,161-163]. All these factors are located at different positions in the pre- 60S r-particle. This abundance of export factors reflects the need of a rapid export of 41 pre-ribosomal particles for the high cellular demand of ribosomes [35]. In addition, it seems that most of these factors also neutralise the partial negative charge of the subunit interface of particles, as some important r-proteins in this area did not still assemble [122]. Before getting export competence, several assembly factors are released and the particle gets compacted to fit the diameter of the NPC [122,164]. Figure 16. Representation of the export of pre-40S and pre-60S r-particles through the NPC. The cartoons show all the export factors (yellow) that have been identified as ensuring the rapid export of pre-ribosomal particles to the cytoplasm, and how these factors interact with hydrophobic repeats of some special nucleoporins from the nucleopore channel. Adapted from [151]. Cytoplasmic assembly of pre-60S ribosomal particles In the cytoplasm, the last steps of the maturation take place. There, the last rRNA processing events occur, the construction of the functional centres is completed and proofread to ensure the functionality of the mature LSU. Nine r-proteins are incorporated in the cytoplasm, and all the remaining assembly factors are released via the activity of selected ATPases and GTPases, in a sequential manner (Figure 17) [87,165]. The cytoplasm maturation of pre-60S r-particles probably initiates with the recruitment of the ATPase Drg1, which depends on the factor Rlp24. The ATPase activity of Drg1 induces its own release and that of Rlp24. Rlp24 is a paralog of the eL24 r-protein and act as its placeholder, thus, once Rlp24 is released, the assembly of eL24 could take place [165-167]. The ATPase activity of Drg1 is also necessary for the release of the nearby Nog1, Nsa2 and Nug1 factors. As Rlp24 interacts with Nog1, it is likely that the 42 release of Rlp24 pulls Nog1 out of the pipe of PET [167,168]. Once eL24 is assembled, pre-60S r-particles can interact and probably recruit Rei1 [169]. Interestingly, the C- terminal extension of Rei1 penetrates deep into the PET similarly as does the C-terminal extension of Nog1. As Nog1, Rei1 probes the exit of the PET impeding its premature occupation [129]. However, Rei1 presents an additional role in the release of the dimer Arx1-Alb1. Thus, Rei1 together with Jjj1 and the ATPase Ssa1 are necessary for the release of Arx1-Alb1, which is accompanied by the own release of all these factors [169- 173]. Once Rei1 leaves the cytoplasmic pre-60S r-particle, its homolog Reh1 occupies the PET. However, the exact role of this replacement is unclear [129,130,174]. Figure 17. Final cytoplasmic steps of pre-60S r-particle maturation. (A) The cartoons show two parallel pathways for the maturation of the PTC-CP and the PET functional centres of pre-60S r-particles. The Drg1 activity facilitates the replacement of Nog1 by Rei1. Rei1 together with Jjj1 and Ssa proteins enables the release of the export receptor Arx1, located at the exit of PET. Lsg1 GTPase activity is responsible for the Nmd3 release. Finally, the activity of the GTPase Efl1 together with its cofactor Sdo1 induces the release of Tif6 from the subunit joining face of the particle. It is not known how Reh1 exits the particle. (B) Cartoon showing how the PET is proofread by the sequential association of Nog1, Rei1 and Reh1 interacting with Arx1 at the exit of the PET. Adapted from [126,130]. 43 One of the more interesting cytoplasmic steps is the formation of the P-stalk. Apparently in the cytoplasm, the factor Yvh1 allows the dissociation of Mrt4, which is the placeholder of uL10. Although some authors have discussed that this step can occur in the nucleoplasm, it seems this exchange is a predominantly cytoplasmic event [150,175-177]. Upon the assembly of uL10 at the base of P-stalk, the two heterodimers of P1-P2 r-proteins can join the almost mature LSU and complete the formation of the P-stalk. Furthermore, the formation of the P-stalk is apparently a pre-requisite for the downstream release of Tif6 and Nmd3 [165,177]. Nmd3 and the assembly of the eL40 r-protein assist the assembly of the uL16 rprotein [178]. An internal loop of uL16 that embraces the P site of the ribosome is a central controller of the transitions between rotated and non-rotated states that occurs in the mature and elongating ribosome [179]. At this stage, the P site is occupied by Nmd3 but the abovementioned loop can now compete and contribute to displace Nmd3. This, together with the activity of the GTPase Lsg1, allows the release of Nmd3 from particles [178,180-182]. The incorporation of uL16 and the release of Nmd3 are essential for the completion of the PTC region of LSUs [178]. Before gaining final translation competence, almost mature LSUs still need to dissociate the Reh1 and Tif6 factors. In the case of Reh1, it is not known the timing and the mechanism of its release. Tif6 is an anti-association factor that prevents the joining of the pre-60S r-particles with SSUs. In this case, the internal loop of uL16 that assists the release of Nmd3 is also responsible for the recruitment of the Sdo1 factor to the P site of LSUs [178,179]. Sdo1 is the cofactor of the GTPase Efl1, which activity is responsible for the release of Tif6 [183-187]. Efl1 binds the GTPase activating centre (GAC). After Efl1 binding, Sdo1 is repositioned and stimulates the GTPase activity of Efl1, which induces a conformational change allowing the release of Tif6 by competing for its binding site in pre-60S r-particles [188]. Once Tif6 has abandoned the pre-60S ribosomal particle, it immediately recycles back to the nucleolus and the mature LSU acquires the translation competence. 44 45 OBJECTIVES 52 Figure 1. Position of eL22 and its neighbourhood in the three-dimensional structure of the cryo-EM reconstruction of Rei1 and Arx1-Alb1 bound to yeast 60S r-subunits. The rproteins eL22 (blue), eL31 (marine blue), and eL38 (marine blue) and the assembly factors Rei1 (yellow), Arx1 (red) and Alb1 (purple) are highlighted. The rest of r-proteins are coloured in light blue, the 5.8S rRNA in coloured in orange, the 5S rRNA in pale orange and the 25S rRNA in pale grey. The representation was generated with the UCSF Chimera program, using the atomic model for the structure of the 60S-Arx1-Alb1-Rei1 complex (PDB ID: 5APO [129]). Upper particles: interface view of the 60S r-subunit (left) that is rotated around the Y-axis 90° twice to obtain the solvent-side view (right). At the bottom, result of rotated around the X-axis 270° the structure in the interface view. Due to the key role of ribosomes in cellular physiology, deletions of most rprotein genes are detrimental to cells, causing non-viability or severe growth defects [30]. However, previous studies have shown that yeast cells lacking eL22 results in just a moderate growth defect mainly caused by the absence of the RPL22A gene, which represents the major contribution to the mRNA and protein levels for eL22 [30,190-194]. The differences on levels of eL22A and eL22B r-protein variants are due to an expression regulatory mechanism based on the intronic regions of their mRNAs. In this sense, the splicing of the RPL22B mRNA is specifically inhibited by the own eL22 protein. The unspliced mRNA is then efficiently eliminated by the Nonsense Mediated Decay (NMD) and Xrn1 dependent 5’-3’ cytoplasmic degradation pathway [193,195,196]. Moreover, 53 the splicing efficiency of the RPL22B mRNA varies upon different stress conditions [195,197]. Besides, RPL22A mRNA is also downregulated by eL22 but to a lesser extent, allowing efficient synthesis of the eL22A r-protein [193,196]. Other studies have shown that eL22 has additional roles in general splicing and therefore in the expression of genes involved in many different cellular processes, among them ribosome biogenesis. In this sense, high levels of eL22 result in a negative feedback that downregulates ribosome biogenesis [191-194,197]. In yeast, the absence of eL22 or a decrease on its levels have been also related to an increased replicative lifespan, increased resistance to endoplasmic reticulum stress and capability of entering in meiosis [30,192,194,198]. Furthermore, some studies have also found that the absence or decreased levels of eL22 lead to aberrant polysome profiles, pointing out a possible role during ribosome maturation and/or translation [30,192]. In humans, there is a family of unrelated developmental rare diseases known as ribosomopathies that are linked to mutations in a wide range of r-proteins and assembly factors [43]. These diseases display a variety of different phenotypes with significant different clinical presentations and potential for treatment, but most cause defects during haematopoiesis and an increased susceptibility to cancer [42]. In fact, as shown in Figure 2, human eL22 is the second r-protein more misregulated during cancer just after uL18 [42]. The role as tumour suppressor of the uL18 r-protein together with that of its partner in the 5S RNP complex, the uL5 r-protein, has been well characterised. It has been reported that these two r-proteins bind to and inactivate the E3 ubiquitin ligase Mdm2, therefore avoiding p53 degradation. Although eL22 has also been described to participate in this inhibitory circuit by binding to both uL18 and Mdm2, its role during tumorigenesis remains unclear. Apparently, the positive selection of RPL22 mutations provides a significant advantage to cells containing wild-type p53, as these mutations could impair the activation of p53 and therefore its signalling, thus enhancing the susceptibility to cancer [199]. In addition, eL22 have been described to participate in other cellular processes in eukaryotes, such as the invasion of the Epstein-Barr virus in human cells and the integration of certain transposon elements in Drosophila [200,201]. 54 Our group is interested in understanding the contribution of specific 60S rproteins to ribosome biogenesis. There are dispersed reports suggesting a role of eL22 during the biogenesis of LSUs. In this thesis work, we have undertaken a systematic and specific functional analysis of yeast eL22 in ribosome synthesis to demonstrate its relevance during this process. Figure 2. Main r-protein genes with recurrent somatic mutations in human cancer. The figure shows that eL22 is the second r-protein with mutations found in different cancer types. The incidence percentages of the r-protein defects in specific cancers are shown. Typography font size is proportional to the incidence of the r-protein mutations in the cancer types where they have been described. CLL stands for chronic lymphocytic leukemia, and T-AA for T-cell acute lymphoblastic leukemia. Taken from [42]. 55 MATERIALS AND METHODS Strains and microbiological methods All the strains of Saccharomyces cerevisiae used in this thesis work are listed in Table 1. The strains JFY010 (rpl22A::kanMX4), JFY011 (rpl22B::kanMX4), JFY084 (rpl38::kanMX4), JFY449 (rpl31A::kanMX4), JFY535 (alb1::KANMX4), JFY540 (arx1::kanMX4) and JFY551 (rps9B::kanMX4) are meiotic segregants from their corresponding heterozygous diploid strains obtained from the Euroscarf collection. In order to generate double and triple deletion mutants, the corresponding strains were crossed, sporulated and tetrads were dissected. Phenotypes involving alb1∆ and arx1∆ deletions were re-checked using strains harbouring these deletions from the Fromont- Racine group. Strain JFY012 (rpl22A::kanMX4 rpl22B::kanMX4) is a haploid segregant derived from crossing JFY010 and JFY011. Strains JFY114 (rpl22A::kanMX4 rpl22B::kanMX4 rpl38::kanMX4), JFY497 (rpl22A::kanMX4 rpl22B::kanMX4 rei1::natMX4), JFY501 (rpl22A::kanMX4 rpl22B::kanMX4 alb1::natMX4), JFY524 (rpl22A::kanMX4 rpl22B::kanMX4 arx1::HPH) and JFY531 (rpl22A::kanMX4 rpl22B::kanMX4 rpl31A::kanMX4) are haploid segregants derived from crossing JFY012 with JFF084, LMA523, LMA525, LMA539 and JFY449, respectively. Strains JFY563 (rpl22A::kanMX4 rpl22B::kanMX4 arx1::kanMX4) and JFY564 (rpl22A::kanMX4 rpl22B::kanMX4 alb1::kanMX4) are haploid segregants derived from crossing JFY012 with JFY540 and JFY535, respectively. The correctness of all deletions was verified by PCR. Strains were grown at the permissive temperature of 30 °C by default. The low temperature of 22 °C and the high temperature of 37 °C were used in the indicated cases. Cells were grown in rich YP(A)D medium (typically 1% yeast extract, 2% glucose, 2% peptone with a supplementation of 0.2% adenine) or in synthetic minimal SD medium (0.5% ammonium sulphate, 0.15% yeast nitrogen base, 2% glucose) supplemented with the appropriate amino acids and bases. When required, 2% glucose was substituted by 2% galactose as carbon source (SGal media). For the preparation of solid medium, 2% agar was added. When needed, cells were transformed with 56 appropriate vectors using the lithium acetate method [202]. Tetrads were dissected using a Singer MSM200 micromanipulator. DNA manipulation techniques were done according to established protocols [203]. Growth and handling of yeast and preparation of yeast media were done according to established procedures [204]. Plasmids and oligonucleotides The DH5α strain of Escherichia coli was used for cloning and propagation of all plasmid vectors of this work. These are listed in Table 2. The construction of the plasmids YCplac111-RPL22A and YCplac111-RPL22A-GFP was carried out by amplification of the appropriate DNA fragments by PCR using distinct oligonucleotides. The PCR products were subsequently cloned in the appropriate vectors and verified by DNA sequencing. The oligonucleotides used in this study are listed in Table 3. Other plasmids have been previously reported in the studies that are appropriately indicated. Translation accuracy assays To quantify the efficiency of amino acid misincorporation and readthrough, the JFY012 strain (rpl22A::kanMX4 rpl22B::kanMX4) and the positive control strains YDB339 (upf1::hisG [PSI+]) and JFY551 (rps9B::kanMX4 [pRS315-rps9B-D94N]) were transformed with different dual luciferase reporter plasmids [205,206], which are also listed in the Table 2 These plasmids allow to monitor translation fidelity in vivo by measuring luminescence in a luciferase assay. Cells were grown until mid-log phase in SD-Ura medium at 30 °C and luciferase activity was measured using the Dual Glo Luciferase Assay System (Promega) in a luminometer CLARIOstar 1.2 microplate reader (BMG Labtech). Misreading or readthrough rates were calculated as the percentage of firefly/Renilla activity of the reporter plasmid divided by the firefly/Renilla activity of the reference plasmid. For each strain, the experiment was done in quadruplicate replica and the results show as the mean ± the standard deviation. 57 Polysome profile analysis Cell extracts for polysome analysis were obtained accordingly to previously described procedures [207]. Yeast cultures of 200 ml were grown to mid-log phase, then translation was blocked by adding 1 ml of cycloheximide (20 mg/ml) and finally cells were collected by centrifugation. Then, cells were resuspended in 500 μl of lysis buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 30 mM MgCl2, 100 μg/ml cycloheximide, 200 μg/ml heparin, 0.2 μl/ml DEPC) and 500 μl of glass beads (0.45 μm diameter) were added. Cells were lysed by vortexing 8 min at maximum speed. Cells extracts were clarified by centrifugation at 13,000 rpm for 10 min. Glycerol to a 10% of final concentration was added and samples were quantified and stored at -80 °C. Ten A260 units of cell extract were separated by centrifugation in a Beckman Coulter rotor SW41Ti at 39,000 rpm in a 7-50% sucrose gradient for 2 h and 45 min. Then, the A254 of the gradients was continuously recorded by an ISCO UA-6 system in order to obtain the polysome profiles. All the protocol, except the acquisition of the profiles, was conducted at 4 °C. Visualization of proteins tagged with a fluorescent reporter by microscopy In order to check the localisation of eL22 assembly, the JFY012 strain (rpl22A::kanMX4 rpl22B::kanMX4) was transformed with two plasmids: one expressing a yEGFP-tagged eL22A protein and another expressing the Nmd3-∆100 protein, which dominantly inhibits the nuclear transport of LSUs [153]. Cells were grown in liquid SD- Ura-Leu medium and shifted for 24 h to SGal-Ura-Leu medium. Then, cells were stained with Hoechst and inspected by fluorescence microscopy. In order to study the subcellular location of pre-ribosomal particles upon absence of eL22, the strains BY7471 (wild type), JFY010 (rpl22A::kanMX4), JFY011 (rpl22B::kanMX4) and JFY012 (rpl22A::kanMX4 rpl22B::kanMX4) were transformed with plasmids expressing the tagged proteins Nop1-mRFP and uL23-yEGFP or uS3-yEGFP. In this case, the different strains were grown in SD-Ura medium until mid-log phase and then washed with PBS buffer (140 mM NaCl, 8 mM Na2HPO4, 1.5 mM KH2PO4, 2.75 mM 58 KCl, pH 7.3); cells were visualized using an Olympus BX61 fluorescence microscopy and the acquired images were analysed with the Olympus cellSens software. RNA extractions and steady-state RNA levels analysis Total RNA was extracted from samples corresponding to 10 OD600 units of midlog phase cultures by the acid-phenol method [208]. Cells were first collected and resuspended in 400 μl of TES buffer (10 mM Tris-HCl, pH 7.5, 10 mM EDTA, pH 8.0, 0.5% SDS). Then, 400 μl of acid phenol was added, vigorously mixed, and samples were incubated for 60 min at 65 °C with agitation. Samples were centrifuged and washed once with acid phenol, another time with chloroform and then precipitated by overnight incubation with 1 ml ethanol 96% and 40 μl 3 M sodium acetate at -20 °C, followed by 15 min of centrifugation at maximum speed in a microfuge. Finally, samples were resuspended in 100 μl of distilled water and RNA quantified by using a NanoDrop spectrophotometer (Thermo Fisher Scientific). Equal amounts of total RNA (normally, 5 μg) were loaded and resolved by electrophoresis in 7% polyacrylamide-8 M gel (for low molecular-weight RNAs) and 1.2% agarose-6% formaldehyde gel (for high molecular-weight RNAs). After migration, RNAs were transferred to a nylon membrane (HybondN, GE Healthcare) and crosslinked as previously described [209]. For northern hybridization, specific antisense oligonucleotides (Table 3) were labelled with [-32P] ATP (6000 Ci/mmol; Perkin Elmer) and used as probes to detect distinct pre-rRNAs and mature rRNAs. For primer extension analysis, equal amounts of total RNA were subjected to reverse transcription using the specific labelled oligonucleotide probe g ([-32P] ATP) as primer. After primer extension, reverse-transcribed DNA was separated by electrophoresis in a 7% polyacrylamide-8 M gel, and the gel placed on a sheet of Whatman filter paper to dry down under vacuum. Nylon membranes or Whatman filter papers were exposed to Phosphorimager screens (Fuji, BAS-IP MS; GE Healthcare) for different times. Visualization was done using a Typhoon™ FLA-9000 imaging system (GE Healthcare). When required, signal intensities were quantified using the GelQuant.NET software. 59 Pulse-chase labelling of pre-rRNAs Pulse-chase labelling experiments were done as previously described [210]. Strains BY4741 (wild type) and JFY012 (rpl22A::kanMX4 rpl22B::kanMX4) were first transformed with an empty YCplac33 plasmid in order to acquire prototrophy for uracil. Cells were grown in liquid SD-Ura medium at 30 °C until mid-log phase and then 40 0D600 units were pulse-labelled for 2 min with 100 Ci of [5,6-3H] uracil (45-50 Ci/mmol, PerkinElmer). The pulse was finished by adding a large excess of unlabelled uracil (1 mg/ml). Then, cells were chased for 0, 5, 15, 30, and 60 min. Cells were washed with cold water and frozen until RNA extraction. Total RNA was extracted by the hot acidic phenol-chloroform procedure described above. Incorporation of labelled uracil was measured by a scintillation counter and about 3000 cpm per RNA sample was loaded and resolved by electrophoresis in a 7% polyacrylamide-8 M urea gel or a 1.2% agarose-6% formaldehyde gel. RNA was transferred to a nylon membrane (HybondN, GE Healthcare), then crosslinked to the membrane and exposed to a tritium screen (BAS-IP TR2040E; GE Healthcare) for different times. Finally, the screen was revealed using a Typhoon™ FLA-9000 imaging system (GE Healthcare). Affinity purification of complexes containing eL22A-GFP The strain JFY012 (rpl22A::kanMX4 rpl22B::kanMX4) was transformed with either YCplac111-L22AGFP or YCplac111-L22A. Then, yEGFP-tagged eL22 was precipitated following the established one-step GFP-Trap_A procedure (Chromotek) [211]. About 40 OD600 units of cells growing at mid-log phase in SD-Leu medium were collected and washed twice with cold water. Cells were resuspended in 500 μl of lysis buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 1.5 mM MgCl2, 1 mM DTT, 0.5% NP-40, complete Protease Inhibitor (Roche)) and 300 μg of glass beads (0.45 μm diameter) were added. Cells were lysed using a FastPrep-24 homogenizer (MP Biomedicals) and then clarified by 15 min of centrifugation at the maximum speed. About 50 μl of GFP-Trap_A beads were added to each cell extracts and incubated for 2 h at 4 °C with end-over-end tube rotation. Then, the beads were washed three times 60 with 1 ml of lysis buffer and collected. Finally, RNA was purified by the phenolchloroform procedure and analysed by northern hybridization as described above. RNA was also extracted from total cell extracts and the samples included in the northern blot analysis. 61 Table 1. Yeast strains used in this study Strain Relevant genotype Source BY7471 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 Euroscarf Y22672 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 RPL22A/rpl22A::kanMX4 Euroscarf Y25844 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 RPL22B/rpl22B::kanMX4 Euroscarf Y23329 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 RPS9B/rps9B::kanMX4 Euroscarf JFY010 MATα his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 This work JFY011 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22B::kanMX4 This work JFY012 MATα his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 This work YDB339 MATa ura2–52 his3Δ200 trp1-Δ901 ade2-101 leu2-3,112 pep4::LEU2 ste6::HIS3 upf1::hisG [PSI+] [206] JFY551 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rps9B::kanMX4 This work LMA523 MATa his3∆1 leu2∆0 ura3∆0 rei1::natMX4 [169] LMA525 MATa his3∆1 leu2∆0 ura3∆0 alb1::natMX4 [169] LMA539 MATa his3∆1 leu2∆0 ura3∆0 arx1::HPH [169] Y25234 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 RPL38/rpl38::kanMX3 Euroscarf Y23772 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 RPL31A/rpl31A::kanMX4 Euroscarf Y21303 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 ALB1/alb1::kanMX4 Euroscarf Y24036 MATa/MATα his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0 ARX1/arx1::kanMX4 Euroscarf JFF084 MATa his3∆1 leu2∆0 ura3∆0 rpl38::kanMX4 This work JFY449 MATa his3∆1 leu2∆0 ura3∆0 rpl31A::kanMX4 This work JFY535 MATa his3∆1 leu2∆0 ura3∆0 alb1::kanMX4 This work JFY540 MATa his3∆1 leu2∆0 ura3∆0 arx1::kanMX4 This work JFY497 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 rei1::natMX4 This work JFY501 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 alb1::natMX4 This work JFY524 MATα his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 arx1::HPH This work JFY531 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 rpl31A::kanMX4 This work JFY564 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 alb1::kanMX4 This work JFY563 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 arx1::kanMX4 This work JFY114 MATa his3∆1 leu2∆0 met15∆0 ura3∆0 rpl22A::kanMX4 rpl22B::kanMX4 rpl38::kanMX4 This work 68 eL22 associates with pre-60S r-particles during middle-nucleolar stages of assembly Assembly of most r-proteins occurs in the nucle(ol)us, although a few LSU rproteins appear to stably load only with cytoplasmic pre-60S r-particles. In the nucle(ol)us, r-proteins can associate to pre-ribosomal intermediates at different stages, from early to late ones [91,218]. The exact timing of eL22 assembly was initially unknown. eL22 has been found among the r-proteins present in nucleoplasmic pre-60S r-particles analysed by cryo-EM methods [134]. However, the comparative compositional analysis of various affinity purified particles spanning early to late nuclear pre-60S r-particles did result inconclusive in the case of eL22 [91]. To further study the timing of assembly of eL22, we expressed an eL22A protein tagged at its C-terminal end with a yeast enhanced GFP (yEGFP) molecule from its own promoter on a CEN plasmid. This eL22-yEGFP protein was fully functional as its expression restored the growth defects depicted by the rpl22 null strain at 30 °C. Moreover, the eL22-yEGFP protein also complemented the LSU formation defect of the rpl22 null strain (see next section and Figure 5). We monitored the localisation of the eL22-yEGFP construct upon expression of the Nmd3-Δ100 dominant negative variant protein under the control of an inducible GAL promoter. Nmd3 is an essential factor required for the nucleo-cytoplasmic export of pre- 60S r-particles [153,219]; the dominant-negative NMD3-Δ100 allele produces a truncated protein that lacks the nuclear export signal (NES) of the factor, resulting in the trap of pre-60S r-particles in the nucle(ol)us [153]. As a result, we observed that eL22- yEGFP accumulated in the nucleus under inducible conditions of the NMD3-Δ100 expression but, as expected for a r-protein, it was found in the cytoplasm under noninducible conditions of the NMD3-Δ100 allele (Figure 6). These results confirmed that eL22 is assembled in the nucleus, as strongly suggested by the previous structural analysis studies [92,134]. 69 Figure 5. Functional analysis of the eL22A-yEGFP construct. (A) Growth comparison of the isogenic wild type and rpl22Δ strains transformed with an empty YCplac111 centromeric plasmid and the rpl22Δ strain transformed with a YCplac111 plasmid harbouring a RPL22A- yEGFP allele that expressed GFP-tagged eL22A from its own promoter. Cells were grown in SD-Leu at 30 °C until mid-log phase, diluted to an OD600 of 0.05 and then 10-fold series of dilutions were spotted on SD-Leu plates. Plates were incubated for two days at 30 °C. (B) The above strains were grown in SD-Leu media until mid-log phase at 30 °C. Then, whole cell extracts were prepared, and 10 A260 units of each extract were resolved in 7-50% sucrose gradients. The A254 was continuously measured. Peaks corresponding to free SSUs (40S) and LSUs (60S), vacant ribosomes/monosomes (80S) and polysomes are indicated. Half-mer polysomes are labelled with arrows. 70 Figure 6. Nucle(ol)ar assembly of eL22. Localisation of GFP-tagged eL22 upon induction of the dominant-negative NMD3-Δ100 allele, which expresses a truncated Nmd3-∆100 protein lacking the nuclear export signal (NES) from a GAL promoter. Cells of the rpl22 null strain expressing eL22-yEGFP from a plasmid were transformed with the pRS316-GAL-NMD3∆100 plasmid. Transformants were grown in liquid SD-Ura-Leu medium (Glc) at 30 °C until midlog phase and shifted for 24 h to SGal-Ura-Leu (Gal) to induce the expression of the Nmd3- ∆100 protein. Then, cells were stained with Hoechst and inspected by fluorescence microscopy. Arrows point to nuclei of representative cells. Approximately 200 cells were examined per experiment; the percentage of cells showing nucle(ol)ar accumulation of eL22-yEGFP is depicted. Scale bar, 10 μm. Next, we affinity purified particles containing eL22-yEGFP using GFP-Trap beads and determined which pre-rRNAs co-purified with them by northern blot hybridization. As shown in Figure 7, eL22-yEGFP was able to co-purify with 27SB pre-rRNAs and subsequent pre-rRNA intermediates as 7S pre-rRNAs; as expected for a r-protein, there was also substantial co-purification of the mature rRNAs. However, we could not detect the 27SA2 pre-rRNA, which is the main intermediate of earliest pre-60S r-particles. Thus, this result strongly suggests that eL22 stably associates with early to middle nucleolar pre-60S r-particles but not at very earlier steps, as again suggested by the previous structural analysis studies [92]. 71 Figure 7. eL22-yEGFP stably assembles with pre-60S particles containing 27SB pre-rRNAs. Cells from the rpl22 null strain expressing an untagged eL22 or a yEGFP-tagged eL22 protein from a plasmid were grown in SD-Leu media at 30 °C until mid-log phase. Whole cell extracts were performed and ribosomal particles affinity purified using the GFP-Trap immunoprecipitation method. Total RNA was extracted from whole cell extracts (T) and immunoprecipitates (IP) and analysed by northern blotting. Probes used to detect the different pre- and mature rRNAs are indicated between parentheses. Loss of eL22 results in a shortage of 60S r-subunits To investigate whether or not the absence of eL22 affects the production of mature LSUs, we analysed the polysome profiles from the rpl22AΔ and rpl22BΔ single mutants and the rpl22 null strain at 30 °C. As shown in Figure 8, both rpl22AΔ and rpl22Δ mutants displayed aberrant polysome profiles compared to those of an isogenic wildtype strain, evidenced by the clear decreased levels of free LSUs versus free SSUs and the presence of half-mer polysomes. These profiles are a common feature of mutants specifically impaired in the biogenesis of LSUs. However, as expected, rpl22BΔ cells presented a normal profile similar to that obtained for the isogenic wild-type strain. 72 Figure 8. Absence of eL22 results in a deficit in 60S r-subunits. Polysome profiles analysis of the wild-type and rpl22 null strains. Cells were grown in liquid YP(A)D media at 30 °C until mid-log phase, whole cell extracts were obtained and 10 A260 units of each extract were resolved in 7-50% sucrose gradients. The A254 was continuously measured. Peaks corresponding to free SSUs (40S) and LSUs (60S), vacant ribosomes/monosomes (80S) and polysomes are indicated. Half-mer polysomes are labelled with arrows. We also analysed how the absence of eL22 affected the production of mature LSUs at 22 and 37 °C. As shown in Figure 9, the extent of the LSU defect did not aggravate when rpl22 null cells were grown at 22 °C (Figure 9A) but was almost completely eliminated when cells lacking eL22 were grown at 37 °C (Figure 9B). Thus, we conclude that eL22 plays an important role for the production of LSUs; we hypothesize that eL22 must exert its function through the proper folding of pre-rRNA. At low temperatures, pre-rRNA may be trapped in an unproductive conformation in the absence of eL22, circumstance that might be bypassed at high temperatures, thus allowing more intermediate pre-60S r-particles to mature and impacting growth to a lesser extent. 73 Figure 9. The 60S r-subunit shortage led by absence of eL22 is less severe at high temperatures. Polysome profiles analysis of the wild-type and rpl22 null strains. Cells were grown in liquid YP(A)D media until mid-log phase either at 22 °C (A) or 37 °C (B). Whole cell extracts were obtained and 10 A260 units of each extract were resolved in 7-50% sucrose gradients. The A254 was continuously measured. Peaks corresponding to free SSUs (40S) and LSUs (60S), vacant ribosomes/monosomes (80S) and polysomes are indicated. Half-mer polysomes are labelled with arrows. The absence of eL22 leads to a mild defect on the nuclear export of pre-60S r-particles To further evaluate the consequences of the absence of eL22 on LSU maturation, we determined whether the rpl22Δ mutant strain was impaired in nuclear export of pre- 60S r-particles. To do so, we analysed the localization of uL23-yEGFP, which was used as an LSU reporter in the rpl22 null and the isogenic wild-type strains grown at 30 °C and 22 °C. As shown in Figure 10, when cells were grown at 22 °C the steady-state levels of uL23-yEGFP were cytoplasmic in the wild-type strain, as expected for a r-protein. In addition, an apparent similar localization was found in the rpl22BΔ strain (data not shown). However, we observed a mild nuclear retention of the fluorescent signal of the uL23-yEGFP reporter on some cells (ca. 30%) of the rpl22AΔ (data not shown) and rpl22 null strain (Figure 10). The comparison with the localization of the mRFP-Nop1 nucleolar reporter led us to conclude that this retention was mostly nucleolar. This retention phenotype was specific for the LSU reporter since when we analysed the localisation of 74 uS3-yEGFP, which was used as a SSU reporter, it localised in the cytoplasm in both wildtype and rpl22 null cells. Very faint differences in the localisation of either uL23-yEGFP was observed at 30 °C when wild-type and rpl22 null cells were compared (data not shown). Figure 10. eL22 is required for an efficient nuclear-cytoplasmic export of pre-60S rparticles. Isogenic wild-type and rpl22 null cells were transformed with plasmids that coexpressed mRFP-Nop1 and either uL23-yEGFP or uS3-yEGFP from their cognate promoters. uL23 r-protein was used as a reporter for pre- and mature 60S r-particles while uS3 r-protein was used as a report for pre- and mature 40S r-particles. Selected transformants were grown in liquid SD-Leu medium at 30 °C until mid-log phase, and then shifted to 22 °C for 3 h. The subcellular localisation of the GFP-tagged r-proteins and the mRFP-Nop1 nucleolar marker was analysed by fluorescence microscopy. Selected nucleoli are pointed by arrows. Approximately 200 cells were examined per experiment; the percentage of cells showing nucle(ol)ar accumulation of either uL23-yEGFP or uS3-yEGFP is depicted. Scale bar, 10 μm. It is strongly unlikely to expect that eL22 would have a direct role in the nucleocytoplasmic export of pre-60S r-particles, therefore, we assume that this defect must arise from the activation of a sort of nuclear quality control mechanism that prevents exit of abnormal pre-60S r-particles from the nucle(ol)us to the cytoplasm due to the absence of the correct assembly of eL22. A similar conclusion has previously been reported upon the deletion, depletion or mutation in other r-protein genes (see Discussion). 75 eL22 is required for normal 27SB pre-rRNA processing To determine whether the LSU maturation defects observed in the absence of eL22 could be attributed to an impairment of a specific step of the pre-rRNA processing pathway, we assessed the steady-state levels of pre- and mature rRNAs from the single and double rpl22 mutants by northern blot hybridization and primer extension analyses. To do this, we first extracted and analysed RNA from wild-type cells, rpl22AΔ and rpl22BΔ single mutants and the rpl22 null strain grown in YP(A)D at 30 °C. Consistent with the polysome profile data, no apparent defects were observed in the rpl22BΔ single mutant compared to the isogenic wild-type strain (Figure 11). In contrast, the rpl22AΔ single mutant and the rpl22 null strain showed mild defects in pre-rRNA processing, which were similar in both cases. In these mutants, a significant accumulation of 27SB pre-rRNA was observed and the amount of mature 25S rRNA mildly decreased (Figure 11A). However, the accumulation of 27SB pre-rRNAs was not accompanied by apparent alterations in the steady-state levels of 7S pre-rRNAs and of mature 5.8S and 5S rRNAs (Figure 11B). Primer extension confirmed that the levels of both 27SBL and 27SBS prerRNAs increased, while levels of 27SA2 and 27SA3 pre-rRNAs remained practically unaltered (Figure 11C). In the rpl22AΔ and rpl22 null strains, we also observed a slight accumulation of both 35S and 23S pre-rRNAs (Figure 11A) that, as discussed later, could be the reflect of an increment of postin detriment of co-transcriptional pre-rRNA processing and the consequent increase in premature cleavage at site A3 [55]. Alternatively, as mentioned in other studies (see Discussion), these accumulations could be the indirect effect of limiting relevant assembly factors required for early processing of 35S pre-rRNA that cannot properly recycle from abnormal pre-60S r-particles as the consequence of the lack of assembly of the eL22 r-protein. 76 Figure 11. Absence of eL22 affects 27SB pre-rRNA processing. The indicated isogenic strains were grown in liquid YP(A)D media at 30 °C until mid-log phase. Total RNA was extracted and equal amounts (5 μg) were subjected to northern blotting or primer extension analysis. (A) Northern blot analysis of high-molecular-mass pre- and mature rRNAs. (B) Northern blot analysis of low-molecular-mass pre- and mature rRNAs. (C) Primer extension analysis using the probe g as primer. Signal intensities of the different RNAs were measured by phosphorimager scanning (indicated below each panel) and normalised to the signals obtained for the wild-type samples, arbitrarily set to 1.0. The probes used for hybridizations and that for primer extension analysis are indicated between parentheses and described in the Materials and Methods section. As mature rRNAs are rather stable, to ascertain the contribution of eL22 to the dynamic conversion of pre-rRNAs to mature rRNAs, we also studied the kinetics of rRNA formation by [5,6-3H] uracil pulse-chase analysis. This experiment was done with the rpl22 null strain and its isogenic wild-type counterpart grown at 30 °C. As a result, both strains displayed very similar kinetics of rRNA synthesis, but rpl22 null cells presented a very mild delay in the conversion of 27SB pre-rRNAs to mature 25S rRNA and a slight delay in that of 7S pre-rRNAs to mature 5.8S rRNAs (Figure 12). However, these delays have only slight consequences in the efficiency of production of the corresponding mature rRNAs. No apparent impairment in either the processing of 20S pre-rRNA to 77 mature 18S rRNA or the synthesis of 5S rRNA was detected. Altogether, these results indicate that eL22 is not required for the maturation of rRNA. Instead, the absence of eL22 induces only a very minor defect on the specific branch of the pre-rRNA processing pathway that reduce the efficiency of the conversion of 27SB pre-rRNAs to mature 25S and 5.8S rRNAs. Figure 12. Absence of eL22 leads to delayed processing of 27S pre-rRNAs. Wild-type and rpl22Δ cells were transformed with an empty YCplac33 (CEN URA3) plasmid and then grown in liquid SD-Ura at 30 °C until mid-log phase. Cells were pulse-labelled for 2 min with [5,6- 3H] uracil, washed and chased for 5 to 60 minutes with a large excess of unlabelled uracil. Total RNA was extracted and about 3000 cpm per sample were loaded and separated on (A) a 1.2% agarose-6% formaldehyde gel or (B) a 7% polyacrylamide-8 M urea gel. RNA was transferred to nylon membranes, exposed to a 3H-sensitive phosphor screen and analysed by phosphorimager scanning. The position of the different pre- and mature rRNAs is indicated. 84 deletion of RPL22B, levels of the RPL22A mRNA are few affected by the feedback regulation, explaining the fact that the phenotypes present in the rpl22B∆ strain are practically identical to those of its wild-type counterpart [193]. Interestingly, this mechanism of autoregulation seems to be conserved in higher eukaryotes, where RPL22 inhibits the expression of its paralog RPL22L1 by binding to its mRNA [222]. Further studies have been done on the autoregulation mechanism of the yeast eL22 expression. Indeed, as eL22A and eL22B are proteins with almost 90% of identity, it has been shown that the key differentiator element in the negative feedback of eL22 expression is present in the respective intronic regions of the RPL22 mRNAs [193,196]. It is also worthy to mention that other studies have found additional extra-ribosomal roles for eL22 in the regulation of a high variety of mRNAs from factors involved in different processes as meiotic induction, one-carbon metabolic pathways or stress response; mechanistically, this regulation is again differentially dependent of the different paralogs of eL22, both in yeast and higher eukaryotes [192-194]. In our work, we have shown that the absence of eL22 led to a mild slow growth that is, however, particularly severe at low temperatures. This result is consistent with previous studies that showed cell cycle arrest at 19 °C and cell death at 4 °C linked to the absence of eL22 [192]. We assume that these growth defects are related to the role of eL22 as a component part of the ribosome; however, as eL22 has been shown to be an important regulator for global gene expression, we cannot initially exclude the possibility that these defects are due to the impact in the extra-ribosomal regulatory function of eL22 rather than in the role that it could have in ribosome biogenesis and function. Indeed, our results show that eL22-lacking ribosomes do not significantly affect translation accuracy; moreover, as we explain below, analysis of ribosome biogenesis by polysome profile, pulse-chase, northern blot and primer extension, and fluorescence microscopy indicates that only a modest deficit of LSU is induced in the absence of eL22. Clearly, only further genetic experiments could provide a definitive answer to this issue by the generation of specific dissociative mutants that could affect one function (e.g. ribosome biogenesis) without affecting the other (i.e. regulation of gene expression) or vice versa. 85 In this work, we have addressed the timing of eL22 assembly in pre-60S rparticles by using different complementary approaches, since previous studies on this regard were not completely conclusive. We first demonstrated using a GFP-tagged version of eL22 that eL22 assembles in the nucle(ol)us, as initially expected [91]. Second, we demonstrated using the GFP-trap procedure that eL22 stably assembles within nucle(ol)ar pre-60S r-particles that contain 27SB pre-rRNAs and subsequent pre-rRNAs but not earlier 27SA2 pre-rRNA. In agreement with these data, eL22 has been found in early pre-60S r-particles, such as those obtained upon split-tag affinity purification of Nsa1-TAP and Flag-Ytm1 and classified as state D particles by cryo-EM analysis [92]. With respect to the role of eL22 in ribosome biogenesis, we have observed a clear and specific reduction in the amount of mature LSUs by analysing polysome profiles, similarly as previously shown [192]. Therefore, eL22 is necessary for the optimal production of the LSUs. However, since eL22 is not essential under laboratory conditions, this need is not strict, specially at high temperatures where polysome profiles returned to an almost wild-type situation. At low temperatures, however, the defects might be stronger, circumstance that has been detected for several other rprotein gene mutants (i.e. [223,224]). Thus, we have observed that cells lacking eL22 display a suboptimum export of LSUs from the nucleus, specially at low temperatures, in consonance with a stronger LSU assembly failure at 22 °C than at 30 °C. By northern blot analysis, our results showed a mild delay in the processing of 27SB and 7S pre-rRNAs in cells lacking eL22. These processing defects denote that both the Las1 and the exosome complexes are unable of efficiently remove the ITS2 from their pre-rRNA substrates. As eL22 is located far away from the foot structure where the ITS2 is situated, we speculate that this phenotype should be due to an indirect reason such as the inefficient interaction and/or function of a single or a set of ribosome biogenesis factors required for these steps; for instance, Nog1 is a perfect candidate for this scenario as it has been shown to interact with eL22 in middle pre-60S r-particles [115] and it has been reported that it is required for 27SB pre-rRNA processing [225]. In the final part of this chapter, we have performed some genetic interaction analysis. Thus, we first studied the relationship between eL22 and the r-proteins eL38 and eL31, as eL22 is sandwiched between these two proteins on mature LSUs. Our 86 results show a very weak enhancement of the slow-growth phenotype of the rpl22 null strain by deleting the RPL31A gene whereas, in contrast, a mild enhancement by deleting the RPL38 one, specially at low temperatures and after exposition to sublethal concentrations of selected antibiotics that impair translation. The significance of these interactions requires further experiments, but it seems reasonable to speculate that if eL22 is important for proper folding of its surrounded rRNA area, pre-ribosomal particles lacking simultaneously eL22 and other r-proteins as eL38 situated in the same region may lead to a more severe conformation defect than that occurring when eL22 is missing alone. Second, we have analysed the relationship between eL22 and a set of ribosome assembly factors that contact with or map close to it in cytoplasmic pre-60S r-particles, such as Rei1 and Arx1, respectively [173]. Recently, a complex of the yeast LSU with Rei1 and the Arx1-Alb1 heterodimer has been reconstituted and its structure studied at nearatomic resolution by cryo-EM analysis [129]. This structure revealed that Rei1 binds near the exit of the PET contacting Arx1 and eL22. During cytoplasmic maturation of LSUs, association of Rei1 occurs concomitantly to the assembly of eL24 r-protein after the release of the assembly factor Rlp24 by the Drg1 ATPase [167] and that of the assembly factor Nog1 [165]. As Nog1, Rei1 inserts its C-terminal tail into the PET [126,129]. Interestingly, loss-of-function of Rei1 impairs release and nuclear recycling of Arx1 and blocks subsequent maturation of cytoplasmic pre-60S r-particles [129,169-172]. Our results show that eL22 and Rei1 functionally interacts as demonstrated by the fact that the introduction of the REI1 deletion into the rpl22 null strain leads to an enhancement of its slow-growth phenotype, specially at low temperatures. This result suggests that eL22 modulates the association and/or function of Rei1 during cytoplasmic maturation of pre-60S r-particles, although clearly further experiments are necessary to validate this possibility. Supporting this model, we have also found that, as previously shown for the link between Rei1 or Jjj1 and Arx1-Alb1, where the absence of either Arx1 or Alb1 is able to suppress the growth defect of the rei1∆ or jjj1∆ strains [129,169-172], the deletion of ARX1 or ALB1 genes in the rpl22 null strain, especially that of ALB1, significantly suppresses the growth defect of the latter strain at either 30 or 22 °C. These results are at least curious as Arx1 and Alb1 are thought to work together and redundantly as an 87 heterodimer, being Alb1 the modulator of Arx1, and Arx1 an export adaptor factor for LSUs that also play a role during the functional check of the PET [155,157,165]. From the functional link between Rei1 or Jjj1 and Arx1 it has been concluded that the trap of Arx1 in cytoplasmic pre-60S r-particles is the main reason of the growth defects observed in the rei1∆ or jjj1∆ cells. This trap prevents Arx1-Alb1 recycling to the nucleus and avoids progression of cytoplasmic pre-60S r-particle maturation, which includes the release of Tif6. This suggest that the recycling of these proteins is functionally connected [129,165,169-172]. Giving the location of eL22 in LSUs and its physical interaction with Rei1, we speculate that the above functional interactions between eL22, Rei1, Arx1 and Alb1 are also linked to all these cytoplasmic releasing events. Further experiments are clearly required to establish whether this hypothesis is correct. 88 89 CHAPTER 2 EARLY ROLE OF RIBOSOMAL PROTEIN eL15 DURING THE ASSEMBLY OF 60S RIBOSOMAL SUBUNITS IN Saccharomyces cerevisiae 90 91 INTRODUCTION As previously mentioned in the main introduction section, r-proteins are assembled in an ordered and hierarchical manner. Eukaryotic ribosomes have several specific r-proteins, among them, eL15. In this thesis work, we have further investigated the functional relevance of the poorly characterised yeast eL15 r-protein in the biogenesis of LSUs. eL15 is one of the r-proteins which are specific from eukaryotes. In S. cerevisiae, it is encoded by two paralogous genes, RPL15A (YLR029C) and RPL15B (YMR121C), which, in contrast to most yeast r-protein genes, lack of introns [189]. Moreover, it has previously been reported that the RPL15A gene is essential whereas the RPL15B is dispensable. Both genes encode functional eL15 proteins but RPL15B is not expressed, at least at normal growth laboratory conditions [226]. The coding regions of these two genes are nearly identical, differing in few nucleotides which results in two amino acid changes between eL15A (glutamine in position 11 and aspartic acid in position 153) and eL15B (glutamic acid in position 11 and asparagine in position 153). The compositional analysis of different pre-ribosomal particles and the solved structures of different pre-60S intermediates by cryo-EM (i.e. [91,92]) have suggested that eL15 is assembled during early nucleolar steps of biogenesis in the surrounding area of the foot structure, interacting with the r-proteins eL8 and eL36, which are also eukaryote-specific proteins (Figure 1). In humans, the misregulation of the genes encoding eL15 has been associated with several kind of cancers [227-233]. For example, overexpression of human eL15 has been shown to promote cell proliferation in gastric and colon cancers and it has been suggested to be a potential target for anticancer therapy [227,231]. In contrast, the levels of eL15 are dramatically decreased in some sorts of pancreatic adenocarcinomas associated with the invasiveness of tumoral cells [233]. Furthermore, mutations in the human RPL15 gene have been described as the molecular cause for some cases of the ribosomopathy named Diamond-Blackfan anemia. These mutations resulted in an 92 impairment on the biogenesis of LSUs and delayed ITS1 processing, which are suggestive of a very early assembly of eL15 also in human ribosomes [234]. Our group is interested in understanding the contribution of specific r-proteins from either the LSU or SSU to ribosome biogenesis. Specifically, at the beginning of this thesis work, the role of eL15 r-protein in pre-rRNA processing and ribosome biogenesis remained undefined, despite its essentiality for yeast cell growth. Herewith, we have undertaken a systematic and specific functional analysis of yeast eL15 in ribosome synthesis to reveal its important role during this process as well as its in vivo assembly. Among other features, we have investigated the interdependence of the incorporation of eL15 and its neighbour eL36 and eL8 r-proteins into pre-60S r-particles. 93 MATERIALS AND METHODS Strains and microbiological methods All Saccharomyces cerevisiae strains used in this work are listed in Table 1. Most of them are derived from BY4743 (MATa/α his3∆1/his3∆1 leu2∆0/leu2∆0 met15∆0/MET15 LYS2/lys2∆0 ura3∆0/ura3∆0) [235]. The JFY144 (rpl15A::kanMX4) and JFY149 (rpl15B::kanMX4) strains are meiotic segregants of the heterozygous diploid strains Y21584 and Y26562, respectively (Euroscarf). The JFY144 and JFY149 strains were crossed and the obtained diploid sporulated; then, tetrads were dissected and tetratypes were selected. Due to the essential nature of eL15, the diploid strains were transformed with the YCplac33-RPL15A plasmid (CEN URA3 RPL15A) prior to meiosis induction. The JFY168 strain (rpl15A::kanMX4 rpl15B::kanMX4 [YCplac33-RPL15A]) is a spore clone used in further experiments. This double mutant was transformed with the YCplac111-GAL-RPL15A plasmid (CEN LEU2 GAL::RPL15A) to generate the JFY355 strain (rpl15A::kanMX4 rpl15B::kanMX4 [YCplac111-GAL-RPL15A]) by counter-selection of the YCplac33-RPL15A plasmid on 5-FOA-containing SGal plates. The presence of the different deletion cassettes in the strains of this study was verified by PCR of the corresponding loci. A similar strategy was performed to obtain the JFY519 strain (rpl36A::kanMX4 rpl36B::kanMX4 [pAS25-RPL36B]). The strains JFY156 (rpl36A::kanMX4) and JFY178 (rpl36B::kanMX4) are meiotic segregants of the heterozygous diploids Y20779 and Y27724, respectively (Euroscarf). Then, they were crossed, sporulated and spore clones with the appropriate markers selected. In this case, prior to meiosis, the diploids were transformed with the pAS25-RPL36B (CEN URA3 GAL::RPL36B), which expresses eL36B from a GAL promoter. Spore clones were selected on YP(A)Gal plates. The JFY440 strain (rpl15A∆::kanMX4 rpl15B∆::kanMX4 noc2::NOC2- TAP-URA3 [YCplac111-GAL-RPL15A]) and the JFY559 strain (rpl36A::kanMX4 rpl36B::kanMX4 noc2::NOC2-TAP-URA3 [pAS25-RPL36B]) are meiotic segregants derived from the diploid resulting of crossing JFY355 and JFY519 with the TY1879 strain (noc2::NOC2-TAP-URA3). 100 Table 2. Plasmid vectors used in this work Name Relevant information Source YCplac33 CEN, URA3 [212] YCplac111 CEN, LEU2 [212] YCplac33-RPL15A RPL15A, CEN, URA3 This study YCplac111-GAL-RPL15A GAL-RPL15A, CEN, LEU2 This study pRS316-RPL25-eGFP NOP1-mRFP RPL25-yEGFP, mRFP-NOP1, CEN URA3 [213] pRS316-RPS3-eGFP NOP1-mRFP RPS3-yEGFP, mRFP-NOP1, CEN, URA3 [213] pAS25-RPL36B GAL-HA-RPL36B, CEN, URA3 This study YCplac111-HA-RPL36B HA-RPL36B, CEN, LEU2 This study 101 Table 3. Oligonucleotides used in this work Name 5'-3' sequence Use L15A UP GCATGCTCACTCTCTTGATTAACCCA Cloning and PCR verification L15A DOWN GAATTCCCTTGGCTAAGTCTAATAAA Cloning and PCR verification L15A-pAS24 UP GCATGCATGGGTGCCTACAAATATTT Cloning and sequencing verification L15A-pAS24 DOWN AAGCTTGATACGAACATTGTTCTCTA Cloning L15A SEC2 AGGTCTCTTTCTGTTACCAC Sequencing verification L15A SEC3 TGAACTCCTACTGGGTTAAC Sequencing verification L15B UP GCATGCAAGCGGAAGCGGAAGT PCR verification L15B DOWN AAGCTTGCCTTTAATGCCAATCTGT PCR verification L36B-pAS25 UP GTCGACATGGCTGTCAAGACTGGTAT Cloning L36B-pAS25 DOWN GCATGCGGGCTGGTTATGAAACAACG Cloning L36A UP AATATAGGTCAGTAGGGAAG PCR verification L36A DOWN CCACCGTATAGTAAACGATT PCR verification L36B UP GAATTCAAAACCACCAACTA PCR verification L36B DOWN TCTAGATTCCACAGAGGACATTGTAA PCR verification NOC2 UP CGTGAAGTTAAGGAAGAAAAGGCC PCR verification NOC2 DOWN GCTTGGAGTGCTCACATGTTAGC PCR verification Probe b (18S) CATGGCTTAATCTTTGAGAC 18S rRNA hybridization Probe c (3-D/A2) GACTCTCCATCTCTTGTCTTCTTG Pre-rRNA hybridization Probe d (A2/A3) TGTTACCTCTGGGCCC Pre-rRNA hybridization Probe e (5.8S) TTTCGCTGCGTTCTTCATC 5.8S rRNA hybridization Probe f (E/C2) GGCCAGCAATTTCAAGTTA Pre-rRNA hybridization Probe g (C1/C2) GAACATTGTTCGCCTAGA Pre-rRNA hybridization and primer extension Probe h (25S) CTCCGCTTATTGATATGC 25S rRNA hybridization Probe 5S GGTCACCCACTACACTACTCGG 5S rRNA hybridization 102 103 RESULTS Yeast eL15 is required for LSU production Yeast eL15 is an essential r-protein ([226] and Figure 2) of 204 amino acids with a predicted molecular mass of 24.5 kDa. As most yeast r-proteins, eL15 is encoded by two genes, RPL15A (YLR029C) and RPL15B (YMR121C). The coding regions of these two genes are nearly identical, differing in few nucleotides that result in two practically identical eL15A and eL15B r-proteins, differing in two amino acid, which curiously involve amino acids of different properties (polar to negatively charged or vice versa) in the same position. Thus, eL15A contains a glutamine in position 11 and an aspartic acid in position 153 while eL15B contains a glutamic acid in position 11 and an asparagine in position 153. Figure 2. Growth analysis of rpl15 deleted strains. (A) Growth test of the rpl15A and rpl15B null mutants compared to their wild-type control strain. Isogenic strains BY4741 (Wild type), JFY144 (rpl15A∆), JFY149 (rpl15B∆) and JFY168 (rpl15A∆ rpl15B∆), harbouring an YCPlac33- RPL15A plasmid were grown in liquid SD-Ura medium to mid-log phase at 30 °C, diluted to an OD600 of 0.05 and spotted in 10-fold serial dilutions onto SD-Ura and SD containing 5-FOA plates for counterselection of the plasmid. Plates were incubated 3 days at 30 °C. (B) Depletion of eL15. Growth comparison of the isogenic strains BY4741 (Wild type), JFY168 harbouring the YCplac33-RPL15A plasmid (RPL15) and JFY355 harbouring the YCplac111- GAL-RPL15A plasmid (GAL::RPL15). Cells were grown in YP(A)Gal to mid-log phase at 30 °C, diluted to an OD600 of 0.05 and spotted in 10-fold serial dilutions onto YP(A)Gal and YPD plates, which were incubated 3 days at 30 °C. 104 Interestingly, as shown in Figure 2A, RPL15A is an essential gene while RPL15B is apparently fully dispensable under laboratory conditions, in agreement with previous reports [30,226]. Moreover, it has been demonstrated that both RPL15 genes are unequally expressed: while RPL15A is expressed under normal growth conditions, RPL15B is not expressed likely because its promoter lacks essential elements for binding of transcriptional factors [226]. Therefore, to properly study the effects of the loss-of- function of eL15 in ribosome biogenesis, avoiding an undesirable suppression event due to the presence of RPL15B, in this work we constructed a strain with both RPL15 genes deleted and expressing as the sole source of eL15, a plasmid-encoded variant of eL15A under the control of a GAL promoter (hereafter, GAL::RPL15 strain). The presence of this plasmid was able to sustain wild-type growth on YP(A)Gal plates, but was unable to do it on YP(A)D plates. As a control, we also cloned the RPL15A gene expressed from its own promoter on a centromeric plasmid, which allowed apparent wild-type growth in both YP(A)Gal and YP(A)D medium (Figure 2B). After shifting a culture of the GAL::RPL15 strain in mid-log phase from liquid YP(A)Gal to YP(A)D medium, the growth rate significantly slowed down in few hours and cultures stopped growing in less than 6 h in YP(A)D medium (Figure 3A). However, western blot revealed practically no reduction on the protein levels of eL15 when the GAL::RPL15 strain was subjected to a shift from galactose- to glucose-containing rich medium for even 12 h (Figure 3B), likely reflecting the stability of pre-existing ribosomes (see also Figure 6A). To start examining the general effects of eL15 depletion in ribosome biogenesis, we analysed polysome profiles from cells of the GAL::RPL15 strain grown in YP(A)Gal or shifted for 3 h to YP(A)D. When grown in YP(A)Gal, the strain showed a normal polysome profile, however, when shifted to YP(A)D for 3 h, it showed an aberrant profile, consisting of a clear decrease in the levels of free LSUs versus SSUs, a decrease in the 80S peak and the appearance of half-mer polysomes (Figure 4). Half-mer polysomes correspond to mRNAs associated with integral numbers of ribosomes plus a stalled 48S pre-initiation complex at the initiation codon [240,241]. They have been linked to translation initiation and/or LSU accumulation defects (e.g. [241-244]). As eL15 is a rprotein, we assumed that half-mer polysomes arose from a decrease in LSU accumulation, thus, calculating the overall LSU/SSU ratio of cells could help confirming 105 this assumption. To do so, we estimated the relative amount of each r-subunit by analysing cell extracts under ribosomal dissociation conditions using low Mg2+ run-off sucrose gradients. These experiments revealed a 25% reduction in the amount of LSUs versus that of SSUs upon 3 h depletion of eL15. Taken together, these results indicate that eL15 is an essential protein required for normal accumulation of LSUs. Figure 3. In vivo depletion of eL15 inhibits cell growth. (A) The strains BY4741 (Wild type; filled black circles) and JFY355 harbouring the YCplac111-GAL-RPL15A plasmid (GAL::RPL15, filled blue squares) were grown in liquid YP(A)Gal medium to mid-log phase at 30 °C, diluted to an OD600 of 0.4 and shifted to YP(A)D medium. The OD600 was measured at the indicated times. (B) Detection of eL15 by western blotting. The above strains were grown in liquid YP(A)Gal and shifted to YP(A)D medium for up to 12 h. Cell extracts were prepared from samples harvested at the indicated times and assayed by western blotting. Equal amounts of total protein (ca. 70 µg) were loaded in each lane. Polyclonal rabbit anti-eL15 and antieS26 were used to detected eL15 and eS26, respectively. Pgk1, detected with anti-Pgk1 monoclonal antibodies, were used as a loading control. Depletion of eL15 impairs export of pre-60S r-particles from the nucleus to the cytoplasm As eL15 is assembled in the nucle(ol)us ([92] and J.F.-F., unpublished results), we were interested to test whether the depletion of eL15 impaired LSU export from the nucleus to the cytoplasm. Nuclear export retention of pre-60S r-particles is common to most loss-of-function LSU assembly factors and/or LSU r-proteins (e.g. [118,245-247]) and it has been interpreted as the consequence of a quality control mechanism that retains in the nucle(ol)us defective pre-60S r-particles unable to acquire nuclear export competence; this control mechanism impedes this way the involvement of malfunctioning LSUs in cytoplasmic translation [77,247,248]. To test whether eL15 106 depletion also impaired nuclear export of pre-60S r-particles, we assessed the subcellular localization of the LSU reporter uL23-yEGFP in the GAL::RPL15 strain before or after a shift to glucose-containing medium for 3 h. As expected for the steady-state levels of a r-protein, uL23-yEGFP was predominantly localised in the cytoplasm of GAL::RPL15 cells grown in selective galactose-containing medium (Figure 5), indicating that nascent pre-60S r-particles were efficiently exported to the cytoplasm. In contrast, when GAL::RPL15 cells were shifted to selective glucose-containing medium for 3 h, uL23-yEGFP was found to accumulate in the nucleus, mostly restricted to the nucleolus, which was detected with the nucleolar marker mRFP-Nop1 (Figure 5). We did not observe nucleolar accumulation of the uL23-yEGFP reporter in the isogenic wild-type control grown in either galactose- or glucose-containing medium (data not shown). Moreover, neither nuclear accumulation was observed when we studied the localisation of the SSU reporter uS2-yEGFP following eL15 depletion (data not shown). Assuming that assembly of uL23 r-protein is not strictly dependent on previous assembly of eL15 (see later Figure 9), these results strongly suggest that both intra-nuclear and nucleocytoplasmic transport of pre-60S r-particles is impaired early after depletion of eL15; this block is apparently specific as the equivalent transport of pre-40S r-particles is unaffected following depletion of eL15. Figure 4. The depletion of eL15 results in a shortage of 60S ribosomal subunits. The strain JFY355 (rpl15A∆ rpl15B∆ [YCplac111-GAL-RPL15A]) was grown at 30 °C in YP(A)Gal medium until mid-log phase and shifted to YP(A)D medium for 3 h to deplete eL15 r-protein. Cell extracts were prepared and 10 A260 units were subjected to polysome profile analysis in 7- 50% sucrose gradients. The peaks of free 40S and 60S r-subunits, 80S free couples/monosomes and polysomes are indicated. Half-mer polysomes are labelled by arrows. 107 Figure 5. Depletion of eL15 leads to nuclear retention of pre-60S ribosomal particles upon eL15 depletion. JFY355 (rpl15A∆ rpl15B∆ [YCplac111-GAL-RPL15A]) cells co-expressing the 60S r-subunit reporter uL23-yEGFP and the nucleolar marker mRFP-Nop1 from a plasmid were grown at 30 °C in liquid SGal-Leu-Ura medium until mid-log phase and shifted to SD- Leu-Ura for 3 h. The subcellular localisations of the GFP-tagged r-protein and the mRFP- Nop1 marker were analysed by fluorescence microscopy. Arrows point to nucleolar fluorescence. Approximately 200 cells were examined and practically all cells gave the results shown in the pictures. Depletion of eL15 leads to an impairment in 27SA3 pre-rRNA processing As the transport of pre-60S r-particles lacking eL15 r-protein was blocked into the nucleus, we wondered whether or not eL15 was also required for pre-rRNA processing. Thus, we first analysed changes in steady-state levels of pre- and mature rRNAs in isogenic wild-type and GAL::RPL15 strains grown in YP(A)Gal medium or at various time points after a shift to YP(A)D medium by northern blotting and primer extension. As shown in Figure 6A, ongoing depletion of eL15 resulted in a strong accumulation of 35S pre-rRNA and a concomitant increase of 23S pre-rRNA; however, levels of 20S pre-rRNA remained apparently unaffected. Accordingly, levels of 27SA2 prerRNA slightly diminished, while interestingly, those of 27SB pre-rRNAs clearly decreased. In agreement with the high stability of mature ribosomes, levels of 25S and 18S rRNAs remained practically unaffected during the 6 h shift of duration of the experiment, suggesting that likely these levels corresponded mainly to pre-existing rather than nascent mature LSUs. Analysis of low-molecular mass pre- and mature rRNAs revealed 108 a significant decrease in the levels of 7S pre-rRNAs, without alteration in either the levels of 5.8S and 5S rRNAs or the ratio of 5.8SL versus 5.8SS rRNAs (Figure 6B). To unambiguously examine the steady-state levels of each 27S pre-rRNA intermediate, we also performed primer extension experiments. As shown in Figure 6C, our results showed that upon depletion of eL15, levels of 27SA2 pre-rRNA did not practically change. However, a clear increase in 27SA3 pre-rRNA was observed. This increase was accompanied by a slight accumulation of 27SBL pre-rRNA but a significant reduction in 27SBS and 25.5S pre-rRNAs. Figure 6. Depletion of eL15 negatively affects the steady-state levels of pre- and mature rRNA species. The strains BY4741 (Wild type) and JFY355 (GAL::RPL15) were grown in YP(A)Gal medium at 30 °C and shifted to YP(A)D medium. Total RNA was extracted from the cultures at the indicated times after the shift and ca. 5 μg were used for northern blotting or primer extension analysis. (A) Northern blot analysis of high-molecular-mass or (B) lowmolecular-mass pre- and mature rRNAs. Probes used for northern blotting, between parentheses, are indicated and described in Table 3. (C) Primer extension analysis using the probe g, which is complementary to sequences in ITS2. This probe allows detection of 27SA2, 27SA3, both 27SB pre-rRNAs and 25.5S pre-rRNA. We also assayed the kinetics of synthesis and turnover of pre-rRNAs by 5,6-3H- uracil pulse-chase labelling experiments. For this purpose, the GAL::RPL15 strain and an isogenic wild-type control were first transformed with the YCplac33 (CEN URA3) plasmid; then, transformants were grown in SGal-Ura medium and shifted to SD-Ura medium for 3 h. Cells were pulsed with tritiated uracil for 2 min and chased with an 109 excess of unlabelled uracil for 60 min. Total RNA was extracted from cells after the pulse and at different times (5, 15, 30 and 60 min) after the initiation of the chase, separated by gel electrophoresis, transferred to a nylon membrane and visualised using imaging screens. As shown in Figure 7A, in wild-type cells, the 35S pre-rRNA was rapidly converted to 20S and 27SA2 pre-rRNAs, which were efficiently converted to mature 18S and 25S rRNAs, respectively. In contrast, in GAL::RPL15 cells, processing of 35S pre-rRNA was delayed although substantial amounts of 20S pre-rRNA and mature 18S rRNA were produced. However, only traces of 27S pre-rRNAs were visualised and no mature 25S rRNA could be detected (Figure 7A). Analysis of low-molecular mass RNAs revealed no 5.8S rRNA production but no alteration on the synthesis of mature 5S rRNA and tRNAs in the GAL::RPL15 cells compared to the isogenic wild-type control (Figure 7B). Figure 7. Depletion of eL15 induces turnover of pre-rRNAs. Wild type BY4741 (Wild type) and JFY355 (rpl15A∆ rpl15B∆ [YCplac111-GAL-RPL15A]) (GAL::RPL15) strains were transformed with an empty YCplac33 (CEN URA3) plasmid, grown at 30 °C in SGal-Ura to mid-log phase and shifted to SD-Ura for 3 h. Cells were pulse-labelled with [5,6-3H] uracil for 2 min followed by a chase with a large excess of SD-Ura for the indicated times. Total RNA was extracted from each sample and 3,000 cpm was loaded and separated on (A) a 1.2% agarose-6% formaldehyde gel or (B) a 7% polyacrylamide-8M urea gel, blotted to nylon membranes and exposed to a screen. The positions of the different pre- and mature rRNAs are indicated. Together, these results indicate that eL15 is required for conversion of 27S prerRNAs into mature 25S and 5.8S rRNAs, more specifically for processing of 27SA3 prerRNA into 27SBS pre-rRNA. Although 27SBL pre-rRNA continues to be produced, it fails to be converted to 7SL pre-rRNA. Furthermore, upon depletion of eL15, there might be 116 has been shown to cross-linked within ITS2 and 5.8S rRNA, close to the eL15 r-protein binding site [92,261]. Noc3 forms a dimer with Noc2, which during maturation replaces the Noc1-Noc2 complex in early pre-60S r-particles [249]; Noc3 binds close to the eL15 r-protein binding site at early pre-60S r-particles [92]. Strikingly, as shown in Figures 12 and 13, most of the assembly factors identified in our iTRAQ analysis have been found to be component of stable early pre-60S ribosomal intermediates (state A to state D intermediates) resolved at high resolution by cryo-EM analysis [92]. Figure 12. Trans-acting factors affected upon depletion of eL15 on early pre-60S ribosomal particles (state A). The cartoons and legend of colour is exactly the same used in Figure 10. Particles are shown from the intersubunit interface (A) or solvent exposed interface (B). The A3-factors are highlighted in different tones of purple/pink. Other factors are coloured in khaki. Pre-rRNA is shown in pale grey and 5.8S rRNA included a region of ITS2 in orange. Only the surfaces of eL8 (green), eL15 (blue) and eL36 (red) are shown. Co-depleted rproteins other than eL15 and eL36 upon depletion of eL15 or eL36 are shown in yellow. The rest of r-proteins are coloured in pale blue. 117 Taken together, we conclude that depletion of eL15 aborts formation of early stable pre-60S r-particles as those identified by the snapshots states A and D resolved by cryo-EM analysis. These particles cannot properly progress in their maturation through removal of ITS1 (cleavage at site A3) and ITS2 (processing at B1 sites), which explain the pre-rRNA processing and nucleo-cytoplasmic export defects observed upon depletion of eL15. Figure 13. Trans-acting factors affected upon depletion of eL15 on early pre-60S ribosomal particles (state E). The cartoons and legend of colour is exactly the same used in Figure 11. Particles are shown from the intersubunit interface (A) or solvent exposed interface (B). The A3-factors are highlighted in different tones of purple/pink and the B-factor in different tones of red. The rest of factors are coloured in khaki. Pre-rRNA is shown in pale grey, 5.8S rRNA in orange and a region of ITS2 in black. Only the surfaces of eL8 (green), eL15 (blue) and eL36 (red) are shown. Co-depleted r-proteins other than eL15 and eL36 upon depletion of eL15 or eL36 are shown in yellow. The rest of r-proteins are coloured in pale blue. 118 Hierarchical and interdependent assembly of the eL15 neighbourhood environment We compared our iTRAQ results with those of the literature and found remarkable similarities with those previously shown upon depletion of eL8. These studies reported that depletion of eL8 leads to practically identical 27S pre-rRNA processing than depletion of eL15. More interestingly, depletion of eL8 also negatively impacts the assembly of both eL15 and eL36 r-proteins, as well as prevents the stable association of A3- and B-factors with pre-60S r-particles [239,252,265]. However, although another study showed that eL36 is also required for processing of 27SA3 prerRNA [251], no data was available for the dynamics of association of r-proteins and protein assembly factors with pre-ribosomal intermediates upon depletion of eL36. Thus, to further study the relationship between this triad of interconnected r-proteins, we also compared the changes in pre-rRNA and protein composition of Noc2-TAP containing particles before or after depletion of eL36. To do so, we first constructed a GAL::RPL36 strain (see Materials and Methods). This strain showed similar growth to that of an isogenic wild-type strain on YP(A)Gal plates but was unable to grow on YP(A)D plates (Figure 14A). Reproducibly, after shifting a culture of the GAL::RPL36 strain at mid-log phase from liquid YP(A)Gal medium to YP(A)D medium, the growth progressively decreased stopping 6 to 9 h after the shift (Figure 14B). Figure 14. In vivo depletion of eL36 inhibits cell growth. (A) Growth comparison of the isogenic strains BY4741 (Wild type) and JFY519 harbouring the pAS25-RPL36B plasmid (GAL::RPL36) or YCplac111-HA-RPL36B (RPL36). Cells were grown in YP(A)Gal to mid-log phase at 30 °C, diluted to an OD600 of 0.05 and spotted in 10-fold serial dilutions onto YP(A)Gal and YPD plates, which were incubated 3 days at 30 °C. (B) Growth curves of the above strains (Wild type; filled black circles. GAL::RPL36, filled red squares) in YP(A)D medium at 30 °C. Cells were first grown in liquid YP(A)Gal medium to mid-log phase at 30 °C, diluted to an OD600 of 0.4 and shifted to YP(A)D. The OD600 was measured at the indicated times. Note that the data corresponding to the wild-type strain are the previously shown in Figure 3. 119 We first analysed the pre- and mature rRNAs that associate with Noc2-TAP by northern blotting. As shown in Figure 15, Noc2-TAP was able to efficiently co-purified 27SB pre-rRNAs both before and after in vivo depletion of eL36, indicative for its association with early to intermediate pre-60S r-particles. Figure 15. Pre-rRNA composition of Noc2-TAP affinity purified particles before or after in vivo depletion of eL15. Strains TY1879 (NOC2-TAP), JFY519 (GAL::RPL36), and JFY559 (NOC2-TAP GAL::RPL36) were grown at 30 °C in YP(A)Gal to mid-log phase (Gal) and shifted to YP(A)D for 3 h (Glc) to shut down the expression of eL36. Total cell extracts were prepared for each strain and each condition and the complexes associated with Noc2-TAP affinity purified. Total RNA was extracted from the affinity-purified samples (lanes IP) and from an amount of total extracts corresponding to 1/50 of that used for purification (lanes Input), separated on a denaturing agarose gel and subjected to northern blotting. Pre- and mature rRNAs were detected using the oligonucleotide probes indicated between parentheses. When r-proteins were assessed by iTRAQ mass spectrometry, as shown in Figure 16A, and as expected, eL36 itself was the r-protein whose levels were most severely diminished from Noc2-TAP-containing particles upon depletion of eL36. The levels of other r-proteins were also significantly reduced, specially uL1 uL22, eL14, eL19, uL24, eL13 and uL23. The uL1 r-protein forms the uL1-stalk together with the tip of helix H76 of 25S rRNA, whose base interacts with eL36 [266]. As above described, eL13 maps close 120 to eL15, and thus to eL36. uL24, but also uL22, uL23 and eL19 are located in the nearby environment of eL36 in both mature LSUs and early pre-60S r-particles ([255]; see also Figures 10 and 11). As for eL15, the relationship of eL14 with the depletion of eL36 is also unclear. Curiously, neither eL15 nor eL8 were among the top five most affected rproteins, in contrast to the results obtained upon depletion of eL15 (see above) or of eL8 [239,252,265]. This fact suggests a hierarchical manner for the assembly of these triad of r-proteins (see Discussion). Few r-proteins were enriched upon depletion of eL36 (Figure 16A); as observed upon depletion of eL15, these included P2A and uL10, most likely due to some contamination of mature LSUs in particles purified from depleted cells. We also studied how the depletion of eL36 affected the association of protein ribosome assembly factors with Noc2-TAP purified particles. Similarly, as occurred upon depletion of eL15 or eL8, most A3-factors (10 out of 12 factors: Brx1, Drs1, Ebp2, Erb1, Has1, Nop7, Nop15, Nsa3, Rlp7, and Ytm1) and B-factors (9 out of 12 factors: Dbp10, Nip7, Nog1, Nop2, Rlp24, Rpf2, Rsr1, Sbp4, and Tif6) were greatly diminished in our iTRAQ analysis upon depletion of eL36 (Figure 16B). Other factors such as Spb1 or Nug1 were also affected upon depletion of eL8, eL15 and eL36 (this work; [239]). Interestingly, the most negatively affected factor after depletion of eL36 was Nop16 (Figure 16B), which also showed reduced levels upon depletion of eL8 [239]. This behaviour might be linked to the interaction of Nop16 with eL8 and eL13 and its role interconnecting RNA elements within 5.8S/25S rRNA domain I (see Figure 13). Few assembly factors were overrepresented in Noc2-TAP particles following depletion of eL36, among them Pwp1, Noc3, Nop12 and Noc4; interestingly, most of these factors also increased upon depletion of eL15 (see Figure 9) or eL8 [239]. As suggested above, these factors must associate with pre-60S r-particles before the stable assembly of any of this triad of r-proteins and leave them only once the assembly of the r-proteins occurs (i.e. [79]). Altogether, these results suggest that depletion of eL36, as well as that of eL15 (this study) and eL8 [239] prevent the stable formation of early pre-60S r-particles. In the absence of any of these r-proteins, the association of most assembly factors that 121 bind domain I such as Nop16 and A3-factors, the subsequent association of particular B- factors and the assembly of a set of r-proteins that mainly lie close to them fail. Moreover, the assembly of eL8, eL15 and eL36 seems to occur in an interdependent manner but apparently following a particular hierarchy (see Discussion). Figure 16. Changes in composition of pre-ribosomal particles upon depletion of eL36. Strain JFY559 (NOC2-TAP GAL::RPL36) was grown at 30 °C in YP(A)Gal to mid-log phase and shifted to YP(A)D for 3 h (Glc) to shut down the expression of eL36. Whole cell extracts were prepared, Noc2-TAP-associated ribosomal particles were affinity-purified and protein abundance was analysed by iTRAQ semiquantitative mass spectrometry. The ratios of the proteins purified after versus before depletion of eL36 are shown. Data are the average of three independent experiments with their standard deviation. Only proteins identified by more than one peptide are indicated. Ratios were normalized to that of Noc2, the bait protein, whose ratio was set arbitrarily to one. (A) Relative enrichment/deprivation of rproteins purified upon depletion of eL36. (B) Relative enrichment/deprivation of protein trans-acting factors purified upon depletion of eL36. 122 123 DISCUSSION We are interested in understanding the role of yeast r-proteins in the assembly and function of ribosomes. Although the contribution of practically all r-proteins from both r-subunits in pre-rRNA processing and nucleo-cytoplasmic transport have been successfully studied (e.g. [223,267], for a review, see [218]), few r-proteins still await characterization or have been poorly studied. In this work, we have undertaken the functional analysis in ribosome assembly of one of them, the yeast eL15 r-protein. eL15 is an essential protein conserved in both archaea and eukaryotes, although the eukaryotic eL15 r-protein contains a specific C-terminal extension, which is absent in the archaeal orthologs [7,189]. In yeast, eL15 is encoded by duplicated genes, RPL15A and RPL15B, but while RPL15A is an essential gene, RPL15B is apparently fully dispensable for growth under laboratory conditions. Indeed, it has been reported that RPL15B is transcriptionally inactive [226]. Yeast eL15A and eL15B proteins have practically identical sequences, differing only in two amino acids, Q11 and D153 in eL15A versus N11 and E153 in eL15B. Whether or not these particular residues are involved in distinct functional-relevant interactions/repulsions with nearby residues of other r-proteins, ribosomal assembly factors or rRNA remains unknown. Interestingly, both eL15 variant proteins are functional, as nicely demonstrated by the expression of either an plasmidencoded eL15B r-protein from the promoter region of the RPL15A gene or by the expression of an eL15A[Q11N D153E] point mutant as the sole sources of eL15 [226]. Regarding the eukaryote-specific C-terminal extension of eL15, it has also been shown that while the last four amino acids of eL15A can be deleted without no effect on function, deletion of last 12 amino acids is fully detrimental for growth, in part due to an interference with the nuclear import of the truncated protein since a predicted bipartite NLS with significant high score is found within this region [226,268]. In this work, we have addressed the role of the eL15 r-protein using a GAL::RPL15 strain that conditionally expresses eL15A. As expected for an essential 60S r-protein, depletion of eL15 led to a strong deficit of 60S r-subunits (see Figure 4). Study of prerRNA processing upon eL15 depletion clearly indicates that this deficit was due to a main impairment of the 27S pre-rRNAs conversion into mature 25S and 5.8S rRNAs. Thus, 124 northern and primer extension analyses showed that while the steady-state levels of 27SA2 pre-rRNA remained practically invariable or slightly decreased upon depletion of eL15, those of 27SA3 pre-rRNA accumulated. Moreover, levels of the subsequent 27SBS pre-rRNA rapidly and considerably decreased but those from 27SBL pre-rRNA, which derived directly from 27SA2 pre-rRNA, did not do it (see Figure 6). Nevertheless, the 27SBL pre-rRNA species seemed not to be productively processed, as neither 7S prerRNA form were produced. Consistently, no change in the ratio 5.8SL:5.8SS could be observed following eL15 depletion, contrary to what occurs with loss-of-function mutations of components of RNase MRP, which is the A3-site endonuclease (e.g. [269,270]). We conclude, therefore, that both processing of 27SA3 to 27SBS pre-rRNA at ITS1 and maturation of both species of 27SB pre-rRNA at ITS2 are impaired upon depletion of eL15. Northern analysis also revealed that depletion of eL15 mildly compromised early processing of 35S pre-rRNA since both, this intermediate and an aberrant 23S pre-rRNA, accumulated with ongoing eL15 depletion. Moreover, as deduced by pulse-chase experiments, although some 35S pre-rRNA was detected and normal amounts of 20S pre-rRNA and mature 18S rRNA were produced, all intermediates destined for mature 25S and 5.8S rRNAs and these mature rRNAs themselves underwent significant turnover following eL15 depletion (see Figure 7). In addition to pre-rRNA processing defects, very rapidly upon eL15 depletion (3 h), there was an accumulation of pre-60S r-particle intermediates in the nucle(ol)us as deduced by the nucle(ol)ar accumulation of the uL23-yEGFP reporter in these conditions (see Figure 5). This result suggests that abortive pre-60S r-particles containing uL23-yEGFP but lacking eL15 were being retained in the nucle(ol)us due to impaired intra-nuclear and nucleo-cytoplasmic export, most likely due to the fact that they could not gain export competence. Similar pre-rRNA processing defects, namely accumulation of 27SA3 and 27SBL pre-rRNAs, rapid turnover of 27S pre-rRNA intermediates to mature 25S and 5.8S rRNAs, delayed pre-rRNA processing at the early A0-A2 sites and nucle(ol)ar accumulation of pre-60S r-particles, have previously been observed upon depletion of other 60S rproteins and protein assembly factors involved in 60S r-subunit biogenesis. Thus, practically identical defects to those herein described upon eL15 depletion have been 125 reported upon depletion of eL8 [239,252] and eL36 [251] or upon depletion of the subset of 60S r-subunit assembly factors known as A3-factors ([105,107- 109,113,251,260,261,264,271-280], for a review see [36]). Alike pre-rRNA processing defects have been reported upon mutation in or depletion of other r-proteins, such as uL4 [254,281], eL6 [254], eL7 [239], eL14 [246], uL13 [91,252,282], eL18 [252,265], eL20 [252,265] and eL33 [252,283]; depletion of eL13 leads to accumulation of 27SA3 prerRNA without a parallel decrease in the levels of 27SBS pre-rRNA [252,254]. In order to rationale why these ribosome biogenesis phenotypes emerge upon depletion of eL15, which is an early assembling r-protein (e.g. [91,92]), and distinguish which ones are general and which ones specific of the loss of this particular r-protein, we studied the pre-rRNA and protein composition of early pre-ribosomal particles purified with the trans-acting factor Noc2 fused to a C-terminal TAP tag before and after depletion of eL15. This approach has been successfully carried out in previous analyses after depletion of or mutation in many ribosome assembly factors and other different rproteins, included eL8 (e.g. [91,107,223,237-239,246,254,265,274,284-286]). For comparative studies, we performed similar experiments upon depletion of eL36, which is together with eL8, the closest neighbouring r-protein to eL15 in mature ribosomes [7]. We used the TAP-tagged Noc2 factor for performing the affinity-purification analyses of pre-ribosomal particles containing or lacking each r-protein. Noc2 is present in 90S and early pre-60S r-particles as member of the consecutive Noc2-Noc1-Rrp5 and Noc2-Noc3 subcomplexes [96,249], thus, being suitable to reveal those compositional changes related to the pre-rRNA processing defects observed upon the depletion of eL15 or eL36. Importantly, a Noc2-TAP construct was also used as the purification bait for akin analyses performed following depletion of eL8 [239,265], which allowed us to directly compare our results with those obtained for eL8. As proven in our co-IP analyses (see Figures 8 and 15), Noc2-TAP was still able to associate with pre-ribosomal intermediates upon eL15 or eL36 depletion; most likely, these particles correspond to those that were retained in the nucle(ol)us upon depletion of eL15 (see Figure 5). Indeed, as our analysis confirmed, uL23 seems to still be stably associated with Noc2-TAP-containing particles after the depletion of eL15 (Figure 9). Moreover, both northern hybridisation and iTRAQ mass spectrometry experiments suggest that, upon depletion of eL15 or eL36, there was 132 133 CONCLUSIONS 134 135 1. The non-essential ribosomal protein eL22 is required for efficient production of 60S ribosomal subunits. Specifically, eL22 is important for the correct processing of 27SB pre-rRNA in the nucleus. 2. eL22 has a cytoplasmic role, being important for the correct recycling of the Arx1 and Alb1 ribosomal assembly factors from the cytoplasm to the nucleus. 3. Despite its function on ribosome biogenesis, ribosomes lacking eL22 do not apparently display translation accuracy errors. 4. The ribosomal protein eL15 is essential for the production of pre-60S ribosomal particles at nucleolar stages of ribosome assembly. 5. eL15 is necessary for the processing of 27SA3 and 27SB pre-rRNAs. Upon depletion of eL15, pre-60S r-particles are subjected to rapid turnover. 6. eL15 is essential for the formation of domain I of 25S/5.8S rRNA. Its assembly displays a hierarchical interdependence with that of eL8 and eL36. In addition, assembly of eL15 is important for the stable association of A3- and B-assembly factors with early pre-60S ribosomal particles. 136 137 BIBLIOGRAPHY 138 139 1. Green, R. & Noller, H.F. Ribosomes and translation. Annu. Rev. Biochem. 66, 679- 716 (1997). 2. Shajani, Z., Sykes, M.T. & Williamson, J.R. Assembly of bacterial ribosomes. Annu. Rev. Biochem. 80, 501-526 (2011). 3. Klinge, S., Voigts-Hoffmann, F., Leibundgut, M. & Ban, N. Atomic structures of the eukaryotic ribosome. Trends Biochem. Sci. 37, 189-198 (2012). 4. Greber, B.J. & Ban, N. Structure and function of the mitochondrial ribosome. Annu. Rev. Biochem. 85, 103-132 (2016). 5. Zoschke, R. & Bock, R. Chloroplast translation: structural and functional organization, operational control, and regulation. Plant Cell 30, 745-770 (2018). 6. Schmeing, T.M. & Ramakrishnan, V. What recent ribosome structures have revealed about the mechanism of translation. Nature 461, 1234-1242 (2009). 7. Ben-Shem, A., Garreau de Loubresse, N., Melnikov, S., Jenner, L., Yusupova, G. & Yusupov, M. The structure of the eukaryotic ribosome at 3.0 Å resolution. Science 334, 1524-1529 (2011). 8. Klinge, S., Voigts-Hoffmann, F., Leibundgut, M., Arpagaus, S. & Ban, N. Crystal structure of the eukaryotic 60S ribosomal subunit in complex with initiation factor 6. Science 334, 941-948 (2011). 9. Rabl, J., Leibundgut, M., Ataide, S.F., Haag, A. & Ban, N. Crystal structure of the eukaryotic 40S ribosomal subunit in complex with initiation factor 1. Science 331, 730-736 (2011). 10. Yusupova, G. & Yusupov, M. Crystal structure of eukaryotic ribosome and its complexes with inhibitors. Philos. Trans. R. Soc. Lond. B Biol. Sci. 372(2017). 11. Melnikov, S., Ben-Shem, A., Garreau de Loubresse, N., Jenner, L., Yusupova, G. & Yusupov, M. One core, two shells: bacterial and eukaryotic ribosomes. Nat. Struct. Mol. Biol. 19, 560-567 (2012). 12. Reuveni, S., Ehrenberg, M. & Paulsson, J. Ribosomes are optimized for autocatalytic production. Nature 547, 293-297 (2017). 140 13. Tye, B.W., Commins, N., Ryazanova, L.V., Wuhr, M., Springer, M., Pincus, D. & Churchman, L.S. Proteotoxicity from aberrant ribosome biogenesis compromises cell fitness. eLife 8, e43002 (2019). 14. Pillet, B., Mitterer, V., Kressler, D. & Pertschy, B. Hold on to your friends: Dedicated chaperones of ribosomal proteins: dedicated chaperones mediate the safe transfer of ribosomal proteins to their site of pre-ribosome incorporation. Bioessays 39, 1-12 (2017). 15. Sung, M.K., Porras-Yakushi, T.R., Reitsma, J.M., Huber, F.M., Sweredoski, M.J., Hoelz, A., Hess, S. & Deshaies, R.J. A conserved quality-control pathway that mediates degradation of unassembled ribosomal proteins. eLife 5, e19105 (2016). 16. Sung, M.K., Reitsma, J.M., Sweredoski, M.J., Hess, S. & Deshaies, R.J. Ribosomal proteins produced in excess are degraded by the ubiquitin-proteasome system. Mol. Biol. Cell 27, 2642-2652 (2016). 17. Lafontaine, D.L. A 'garbage can' for ribosomes: how eukaryotes degrade their ribosomes. Trends Biochem. Sci. 35, 267-277 (2010). 18. Cole, S.E., LaRiviere, F.J., Merrikh, C.N. & Moore, M.J. A convergence of rRNA and mRNA quality control pathways revealed by mechanistic analysis of nonfunctional rRNA decay. Mol. Cell 34, 440-450 (2009). 19. Grousl, T., Ivanov, P., Frydlova, I., Vasicova, P., Janda, F., Vojtova, J., Malinska, K., Malcova, I., Novakova, L., Janoskova, D., Valasek, L. & Hasek, J. Robust heat shock induces eIF2alpha-phosphorylation-independent assembly of stress granules containing eIF3 and 40S ribosomal subunits in budding yeast, Saccharomyces cerevisiae. J. Cell. Sci. 122, 2078-2088 (2009). 20. Fujii, K., Kitabatake, M., Sakata, T., Miyata, A. & Ohno, M. A role for ubiquitin in the clearance of nonfunctional rRNAs. Genes Dev. 23, 963-974 (2009). 21. Fujii, K., Kitabatake, M., Sakata, T. & Ohno, M. 40S subunit dissociation and proteasome-dependent RNA degradation in nonfunctional 25S rRNA decay. EMBO J. 31, 2579-2589 (2012). 22. Kraft, C., Deplazes, A., Sohrmann, M. & Peter, M. Mature ribosomes are selectively degraded upon starvation by an autophagy pathway requiring the Ubp3p/Bre5p ubiquitin protease. Nat. Cell Biol. 10, 602-610 (2008). 141 23. Heeseon An, H. & Harper, J.W. Ribosome abundance control via the ubiquitinproteasome system and autophagy. J. Mol. Biol. 432, 170-184 (2020). 24. Melnikov, S., Manakongtreecheep, K. & Soll, D. Revising the structural diversity of ribosomal proteins across the three domains of life. Mol. Biol. Evol. 35, 1588- 1598 (2018). 25. Pausch, P., Singh, U., Ahmed, Y.L., Pillet, B., Murat, G., Altegoer, F., Stier, G., Thoms, M., Hurt, E., Sinning, I., Bange, G. & Kressler, D. Co-translational capturing of nascent ribosomal proteins by their dedicated chaperones. Nat. Commun. 6, 7494 (2015). 26. Fernández-Pevida, A., Martín-Villanueva, S., Murat, G., Lacombe, T., Kressler, D. & de la Cruz, J. The eukaryote-specific N-terminal extension of ribosomal protein S31 contributes to the assembly and function of 40S ribosomal subunits. Nucleic Acids Res. 44, 7777-7791 (2016). 27. Mazumder, B., Sampath, P., Seshadri, V., Maitra, R.K., DiCorleto, P.E. & Fox, P.L. Regulated release of L13a from the 60S ribosomal subunit as a mechanism of transcript-specific translational control. Cell 115, 187-198 (2003). 28. Deisenroth, C. & Zhang, Y. Ribosome biogenesis surveillance: probing the ribosomal protein-Mdm2-p53 pathway. Oncogene 29, 4253-4260 (2010). 29. Kellis, M., Birren, B.W. & Lander, E.S. Proof and evolutionary analysis of ancient genome duplication in the yeast Saccharomyces cerevisiae. Nature 428, 617-624 (2004). 30. Steffen, K.K., McCormick, M.A., Pham, K.M., Mackay, V.L., Delaney, J.R., Murakami, C.J., Kaeberlein, M. & Kennedy, B.K. Ribosome deficiency protects against ER stress in Saccharomyces cerevisiae. Genetics 191, 107-118 (2012). 31. Ghulam, M.M., Catala, M. & Abou Elela, S. Differential expression of duplicated ribosomal protein genes modifies ribosome composition in response to stress. Nucleic Acids Res. 48, 1954-1968 (2020). 32. Lu, H., Zhu, Y.F., Xiong, J., Wang, R. & Jia, Z. Potential extra-ribosomal functions of ribosomal proteins in Saccharomyces cerevisiae. Microbiol Res 177, 28-33 (2015).