Coordinated ribosomal L4 protein assembly into the pre-ribosome is regulated by its eukaryote-specific extension

SUMMARY

Eukaryotic ribosome biogenesis requires nuclear import and hierarchical incorporation of ~80 ribosomal proteins (RPs) into the ribosomal RNA core. In contrast to prokaryotes, many eukaryotic RPs possess long extensions that interdigitate in the mature ribosome. RpL4 is a prime example with a ~80 residue long surface extension of unknown function. Here, we identify assembly chaperone Acl4 that initially binds the universally conserved internal loop of newly synthesized RpL4 via its superhelical TPR domain, thereby restricting RpL4 loop insertion at its cognate nascent rRNA site. RpL4 release from Acl4 is orchestrated with pre-ribosome assembly, during which the eukaryote-specific RpL4 extension makes several distinct interactions with the 60S surface including a co-evolved site on neighboring RpL18. Consequently, mutational inactivation of this contact site, on either RpL4 or RpL18, impairs RpL4-Acl4 disassembly and RpL4 pre-ribosome incorporation. We propose that hierarchical ribosome assembly can be achieved by eukaryotic RP extensions and dedicated assembly chaperones.

INTRODUCTION

Ribosomes are huge (~4 MDa) ribonucleoprotein complexes, which translate the genetic code into polypeptide chains enabling the cell to manipulate their cellular environment. ~80 different ribosomal proteins (RP) are assembled coincident with transcription of pre-ribosomal RNA (rRNA) molecules, which are subsequently processed to mature rRNA molecules (25S, 18S, 5.8S and 5S) mediated by a complex interplay of more than 200 trans-acting assembly factors (Ben-Shem et al., 2011; Lafontaine and Tollervey, 2001; Woolford and Baserga, 2013). This highly energy consuming assembly process is tightly regulated within the cell to guarantee the rapid production of more than one million ribosomes per cell generation in mammalian cells (Grummt, 1999). In yeast, ~14 million RPs are generated per generation that translocate through the nuclear pore complex to meet the maturing ribosome in the nucleus (Schutz et al., 2014). Recent work demonstrates that RPs assemble in a hierarchical and cooperative manner onto the pre-rRNAs, which allow a rough classification of RPs into primary, secondary and tertiary binders (Gamalinda et al., 2014; Shajani et al., 2011). Ill-timed binding of RPs to the pre-ribosome or altered stoichiometry of RPs can disturb ribosome biogenesis resulting in ribosomal stress. This stress signal can be transduced to other cellular pathways, potentially resulting in disease development including cancer (James et al., 2014; McCann and Baserga, 2013; Teng et al., 2013). Thus, synthesis of RPs must be tightly coordinated with their timely assembly into the pre-ribosome to avoid inadvertent defects in ribosomal biogenesis (Wang et al., 2014). Eukaryotic RPs often have composite protein structures located on the mature ribosome with folded and unfolded domains and/or long extensions. During evolution eukaryotic RPs acquired extensions on the ribosomal surface, which contact large stretches of the rRNA and other RPs (Ben-Shem et al., 2010; Melnikov et al., 2012). While the biological reason for the evolution of these RP extensions remains to be established, their involvement in translation, ribosome assembly, and masking the eukaryote-specific rRNA expansion segments (ESs) was suggested (Ben-Shem et al., 2011; Melnikov et al., 2012). The molecular mechanisms by which eukaryotic cells stabilize delicate largely unstructured RPs and coordinate their incorporation into the pre-ribosome are largely unknown. Dedicated chaperones were identified that bind to a number of RPs, disassemble karyopherin•RP import complexes, or synchronize nuclear RP import (Eisinger et al., 1997; Holzer et al., 2013; Iouk et al., 2001; Jäkel et al., 2002; Koch et al., 2012; Kressler et al., 2012; Schaper et al., 2001; Schutz et al., 2014).

Here, we report that newly synthesized RpL4 is chaperoned by an uncharacterized protein, Ydr161w, which we name Assembly Chaperone of RpL4 (Acl4). We show that Acl4 binds to a protruding evolutionarily conserved loop of RpL4, which in the mature ribosome is buried deep in the rRNA core structure. Moreover, we demonstrate that the C-terminal RpL4 extension not only is involved in nuclear import of the Acl4•RpL4 complex, but also facilitates ribosome incorporation of RpL4 and associated disassembly of the Acl4•RpL4 complex. Thus, the eukaryotic cell acquired the ability to regulate ribosome maturation through C-terminal RP extensions and dedicated ribosomal assembly chaperones.

RESULTS AND DISCUSSION

Nascent RpL4 Association with Acl4 Ensures Efficient 60S Ribosome Biogenesis

RpL4 is a typical RP with an ~80 residue C-terminal eukaryote-specific extension and an ~60 residue exposed evolutionarily conserved loop (Ben-Shem et al., 2011; Gamalinda and Woolford, 2014; Ramakrishnan and White, 1998). The loop protrudes from the globular folded core and deeply projects into the 25S rRNA core, lining the peptide exit tunnel of the mature ribosome (Figure 1A) (Zhang et al., 2013). The RpL4 extension meanders ~140 Å across the 60S surface with contacts to rRNA and RPs (Figure 1A).

Figure 1. Acl4 is a Chaperone of Nascent RpL4.

(A) Structure of RpL4 as observed in the S. cerevisiae 80S ribosome (PDB 3U5E, 3U5D) (Ben-Shem et al., 2011). Surface of 25S and 5S rRNA are shown in grey. RpL4, RpL18, RpL7 and RpL20 are shown in red, blue, green and yellow, respectively. Core, loop and C-terminal extension of RpL4 are indicated.

(B) Epitope pulse-chase analysis of RpL4 in yeast cells. HA-RpL4-Flag-ProtA was pulsed for 0 (lane 1) or 5 minutes (lane 2), and chased for 5 (lane 3) and 19 minutes (lane 4). Newly synthesized RpL4 was tandem affinity-purified and analyzed by SDS-PAGE and Coomassie staining. Indicated proteins were identified by mass-spectrometry. Asterisk indicates keratin contaminant.

(C) Tandem affinity-purification (Flag-ProtA) of chromosomal tagged Acl4 from yeast cells. Eluates were analyzed by SDS-PAGE, Coomassie staining and mass-spectrometry. Asterisk indicates RpL4 breakdown.

(D) Epitope pulse-chase analysis of RpL4ΔExt. HA-RpL4ΔExt-Flag-ProtA was pulsed for 0 (lane 1) or 5 minutes (lane 2), and chased for 5 (lane 3) and 19 minutes (lane 4). Asterisks indicate keratin and IgG contaminants.

(E) Acl4 localization in yeast cells. Acl4 was chromosomal tagged with GFP and the localization was assessed by fluorescence microscopy. Scale bar is 5 μm.

(F) Growth analysis of Δacl4 strain. The impaired growth of Δacl4 was partially rescued by expression of an additional copy of RPL4 from a centromeric plasmid.

(G) Polysome profiles of ACL4 and Δacl4 cells. Deletion of ACL4 impairs the synthesis of the 60S ribosomal subunit. Free 80S, 60S, 40S and polysomes are indicated. Arrows point to the observed halfmer polysomes. See also Figure S1.

To reveal the molecular details of how primarily unstructured RpL4 is assembled into the pre-ribosome and establish whether this process is assisted by additional biogenesis factors, we performed pulse-chase analyses combined with affinity-purification to study the fate of RpL4 during its early life (Stelter et al., 2012). After a 5 minute pulse newly synthesized RpL4 (for complementation, see Figure S1A; for yeast strains and constructs, see Table S1 and S2) co-precipitated an uncharacterized ~45 kDa protein (Ydr161w) (Figure 1B). Later, during the chase, RpL4 associated with other ribosomal proteins and assembly factors with concomitant dissociation of Ydr161w. At the end of the chase, pulse-labeled RpL4 was part of mature ribosomes, as indicated by co-precipitation of both 60S and 40S RPs (Figures 1B, 3B). Consistent with a transient interaction, affinity-purification of Ydr161w from yeast cells strongly co-enriched RpL4 (Figure 1C). We tested whether the eukaryote-specific RpL4 extension is linked to Ydr161w, for which a prokaryotic homolog was not found. However, Ydr161w rather being impaired in the interaction did not dissociate from newly synthesized RpL4 lacking its C-terminal extension (RpL4ΔExt) during the pulse-chase and was not assembled into the 60S subunit (Figure 1D). In contrast to a recent report (Gamalinda and Woolford, 2014), we observe that the RpL4 C-terminal extension is essential for cell growth and its absence from RpL4 induces a dominant growth defect, particular at higher temperatures (Figure S1B). Altogether, we conclude that Ydr161w, which we named Assembly Chaperone of L4 (Acl4), binds newly synthesized RpL4, whose release from Acl4 and subsequent ribosome incorporation depends on its eukaryote-specific extension domain (Figures 1B, 1D). Consistent with this interpretation, Acl4 directly binds RpL4, forming a stable complex that was reconstituted in vitro using either Chaetomium thermophilum or yeast orthologs (Figure 2A). In vivo, Acl4 is distributed throughout the cell but is enriched in the nucleus (Figure 1E). Acl4 is not essential but acl4Δ cells display a slow growth phenotype (Figure 1F). The growth defect of acl4Δ cells can be partly rescued by expressing an extra RPL4 copy (Figure 1F) and appears to be the consequence of a defective ribosome synthesis, indicated by reduction of free 60S relative to 40S subunits and the appearance of halfmer polysomes in ribosome profiles (Figure 1G).

Figure 3. The Conserved Loop of Nascent RpL4 is Protected by Acl4 until Ribosome Insertion.

(A) rpl4AΔ/rpl4BΔ knockout strain was transformed with an empty vector or vectors containing RPL4A and RPL4A R95E/R98E. The RPL4A wild-type copy on the URA-plasmid was shuffled out on FOA. The respective clones were spotted on YPD plates and incubated at 23 °C, 30 °C and 37 °C for 2 days.

(B) Epitope pulse-chase analysis of RpL4, RpL4 R95E/R98E and RpL4 Δ63–87. Wild-type (lanes 1–3) or mutants (lanes 4–9) of RpL4 were pulsed for 0 minutes (lanes 1, 4, 7) or 5 minutes (lanes 2, 5, 8) as a GAL::tcapt-HA-RpL4-Flag-ProtA version and subsequently chased for 19 minutes (lanes 3, 6, 9). HA-RpL4-Flag-ProtA was affinity-purified and resolved on a SDS-PAGE gel and visualized by silver staining or western blot (lower panel). Indicated proteins were identified by mass-spectrometry. Acl4 was not found in the purification of RpL4 R95E/R98E and RpL4Δ63–87 (lanes 5, 8). The original blot and silver SDS-PAGE gel were sliced and put together from one gel/blot image (original blot and SDS-PAGE gel, see Figure S5). Accordingly, wild-type RpL4 (lanes 1–3) in Figures 3B and 4B are derived from the same gel.

(C) Part of the 60S subunit with RpL4, RpL18 and the extended sequence of 25S RNA (ES7) are depicted. The RpL4 residues I289, I290 and I295 (violet) contacting RpL18 residues L32 and V129 (cyan), and RpL4 residues K332 and F334 (violet) contacting ES7 RNA (goldenrod) are shown.

(D) The C-terminal extension (residues 277–362) of RpL4 targets the attached 3xyEGFP reporter to the nucleus. A 44-residue region of the C-terminal RpL4 extension (residues 301–345), containing two potential PY-NLS sequences (consensus is indicated above the amino acid sequence), is sufficient for efficient nuclear import. Mutation of three lysine residues (K314A, K315A, K319A) in the extended NLS (residues 277–362) or shorter constructs, ranging from residues 303–320 or 311–333, were also tested for NLS activity by monitoring nuclear accumulation of GFP. Scale bar is 5 μm.

(E) The scKap104 interaction with scRpL4 is dependent on its C-terminal extension. GST-scKap104, His₆-scAcl4•scRpL4, His₆-scAcl4•scRpL4ΔExt, His₆-scAcl4 and His₆-scSyo1 (positive control for scKap104 binding) (Kressler et al., 2012) were expressed in E. coli and affinity-purified (lanes 1–5, input). GST-scKap104 (lanes 5–9) and GST (lanes 10–14) were immobilized on GSH-beads and incubated with excess of purified His₆-scAcl4•scRpL4, His₆-scAcl4•scRpL4ΔExt, His₆-scAcl4 and His₆-scSyo1. Samples were analyzed by SDS-PAGE and Coomassie staining. See also Figure S4 and S5.

Figure 2. Biochemical and Structural Analysis of the Acl4-RpL4 Interaction.

(A) Size-exclusion chromatography coupled to multiangle light scattering (SEC-MALS) analysis of the ctAcl4•ctRpL4•ctKap104 trimeric nuclear import complex. The absorbance at 280 nm is plotted against the elution volume of a Superdex 200 10/300 GL size exclusion column and overlaid with the molecular mass of the different proteins. Fractions that were resolved on a SDS-PAGE gel and visualized by Coomassie staining are indicated by a gray bar. Degradation products are labeled with asterisks.

(B) Domain representation and crystal structure of ctAcl4. Blue, basic N-terminal region; grey, α-helical region; red, acidic C-terminal region. The crystallized fragment is indicated with a black bar. The ctAcl4 crystal structure is shown in cartoon representation in two different orientations.

(C) Surface representation of ctAcl4 in four different orientations colored according to a multi-species sequence alignment (Figure S3A). Sequence conservation is shaded from white (< 40 % similarity) to yellow (40 % similarity) to red (100 % identity).

(D) Surface representation colored according to electrostatic potential from −10 k_BT/e (red) to +10 k_BT/e (blue).

(E) GST pull-down of ctAcl4 variants. Red, GST-ctAcl4 variants (bait); black, ctRpL4 variants. Loaded (top, soluble lysate fraction) and pulled-down (bottom) fractions were analyzed by SDS-PAGE and Coomassie staining. ctAcl4 fragment boundaries are shown above each lane and are depicted in a cartoon.

(F) GST pull-down of ctRpL4 variants. Red, GST-ctAcl4 (bait); black, ctRpL4 variants. ctRpL4 fragment boundaries are shown above each lane and indicated in the scRpL4 structure extracted from the S. cerevisiae ribosome (PDB 1VXY) (Svidritskiy et al., 2014).

(G) His-scAcl4, GST-scRpL4 and GST-scRpL4 R95E/R98E were purified from E. coli. GST, GST-scRpL4 and GST-scRpL4 R95E/R98E beads were incubated with excess of imidazole eluted His-scAcl4. Beads were boiled and eluates were resolved on a SDS-PAGE gel and visualized by Coomassie Brilliant Blue staining.

(H) Negative stain electron microscopic analysis of recombinant purified ctAcl4•ctRpL4 complex. Two-dimensional class average of ctAcl4•ctRpL4 complex (top row) determined by multivariate statistical analysis matching with the projections of the final 3D model (middle row) and surface representations of equivalent orientations (bottom row).

(I) Model of the Acl4•RpL4 complex. See also Figure S2 and S3.

RpL4 Interacts with Crescent-Shaped Acl4 via an Exposed Loop

To gain structural insight into the Acl4-RpL4 interaction, we crystallized a Chaetomium thermophilum Acl4 fragment (for E. coli expression constructs, see Table S3), lacking the unstructured acidic C-terminal extension (residues 1–338), and solved its structure at a 2.9-Å resolution (Figure 2B, Table 1). ctAcl4 exhibits a superhelical tetratricopeptide repeat (TPR) fold in its central α-helical region, which adopts an overall crescent-shaped structure (Figure 2B). ctAcl4 is composed of 13 α-helices (αA-Αm), which are arranged in a zig-zag fashion with a right-handed superhelical twist, forming 6.5 TPRs (Figure 2B). Whereas TPRs 1, 2, 5, and 6 adopt canonical TPRs, repeats 3 and 4 possess atypical extended helices (αF, αG) that form a characteristic tower in the middle of the domain, separating the protein into two halves (Figure S2C). No electron density was observed for the N-terminal basic region (residues 1–28) and thus this region is presumed to be disordered.

Table 1.

Crystallographic analysis

Data collection
Protein	ctAcl41−338	ctAcl428−338	ctAcl428−338	ctAcl428−338
PDB code	4YNW	4YNV
Synchrotron	APSa	APS	APS	APS
Beamline	GM/CA-CAT	GM/CA-CAT	GM/CA-CAT	GM/CA-CAT
Space group	P1	P1	P1	P1
Cell parameters
a, b, c (Å)	49.1, 49.6, 80.3	49.5, 49.1, 80.1	49.5, 49.1, 80.0	49.5, 49.1, 80.0
α, β, γ (°)	98.6, 100.1, 98.5	98.7, 97.9, 100.0	98.7, 97.8, 100.0	98.7, 97.8, 100.0
	Native	Se peak	Se inflection	Se remote
Wavelength (Å)	0.97939	0.97936	0.97961	0.94937
Resolution (Å)	50.0 – 2.9	20.0 – 2.95	20.0 – 3.0	20.0 – 3.0
Rsym (%)b	4.6 (77.5)	5.9 (10.9)	6.1 (78.5)	6.5 (80.9)
<I> / <σ/I>b	40.3 (2.0)	22.4 (1.9)	21.2 (1.8)	20.5 (1.7)
Completeness (%)b	97.2 (85.8)	99.0 (98.3)	99.0 (98.3)	99.1 (98.6)
No. observations	125,935	119,531	115,067	113,041
No. unique reflectionsb	16,367 (1,424)	30,229 (3,037)	29,064 (2,926)	28,582 (2,895)
Redundancyb	7.7 (6.6)	4.0 (3.9)	4.0 (3.9)	4.0 (3.9)
Refinement
Resolution (Å)	20.0 – 2.9	20.0 – 2.95
No. reflections total	15,661	30,019
No. reflections test set	1,577	1,497
Rwork / Rfree (%)	24.0 / 26.1	21.9 / 25.6
No. atoms	4,432	4,432
Protein	4,432	4,432
B-factors
Protein	128.6	102.6
R.m.s. deviations
Bond lengths (Å)	0.002	0.003
Bond angles (°)	0.5	0.6
Ramachandran plotc
Favored (%)	94.7	94.5
Additionally allowed (%)	5.3	5.5
Outliers (%)	0.0	0.0
Rotamer outliers (%)c	0.0	0.0
Cβ outliers (%)c	0.0	0.0
Clash scorec	2.2	1.72
MolProbity scorec	1.36	1.30

^aAPS, Advanced Photon Source

^bHighest resolution shell is shown in parentheses

^cAs determined by MolProbity (Davis et al., 2007)

A multi-species sequence alignment shows that Acl4 is evolutionarily conserved with orthologs in fungi, insects, mollusks, worms, fish and plants (Figure S3A). As previously observed for other ribosome assembly chaperones (Holzer et al., 2013), no mammalian Acl4 orthologs could be identified in database searches, suggesting that mammalian Acl4 protein sequences are evolutionarily more distant, or acquired an alternative shielding mechanism for nascent RpL4. Analysis of the conservation and electrostatic potential of the ctAcl4 surface reveals that the concave and bottom surface are evolutionarily conserved and display a strong negative surface potential (Figures 2C, D).

These findings are in line with structure-based truncations of ctAcl4 and ctRpL4, which revealed that the C-terminal ctAcl4 half (residues 156–398), comprising the central long helix αG, and the protruding ctRpL4 loop, which is an evolutionarily conserved feature in all prokaryotic and eukaryotic RpL4s, are necessary and sufficient structural elements for the ctAcl4-ctRpL4 interaction (Figures 2E, F). To identify mutants that impair scAcl4 binding, we designed site-specific charge-swap mutations of invariant positively charged residues in the stem (R95E, R98E) and deleted the tip (Δ63–87) of the scRpL4 loop, and found that they indeed disrupted the interaction (Figures 2G, S3B).

Negative-stain electron microscopy of the reconstituted ctAcl4•ctRpL4 complex revealed a horseshoe-shaped structure that is large enough to accommodate the crescent shaped ctAcl4, binding to ctRpL4 with its C-terminal half (Figures 2H, S2D). In this arrangement, the concave surface of ctAcl4 could bind the protruding ctRpL4 loop and shield it (Figure 2I). Consistent with this notion, the ctRpL4 loop is protected against trypsin digestion only when bound to ctAcl4 (Figure S4SA). In contrast, the ctRpL4 C-terminal extension, which is dispensable for the interaction with ctAcl4 (Figure 2F), is sensitive to proteolysis in the ctAcl4•ctRpL4 complex (Figure S4A). Thus, the C-terminal extension is accessible for other interactions in the cell (see below).

RpL4 Loop Mutants Deficient in Acl4 Binding Enter Ribosome Biogenesis but Display a Growth Defect

We investigated how RPL4 loop mutants impaired in Acl4 binding affect yeast growth and cellular pathways. RPL4 R95E R98E was able to complement the non-viable rpl4Δ null strain, but with a slower growth at higher temperatures (Figure 3A). As shown in the pulse-chase assay RpL4 R95E R98E was assembled into mature 60S subunits even though Acl4 was not found on nascent RpL4 in the 5 min pulse (Figure 3B). In contrast, the tip deletion variant rpl4Δ63–87 induced non-viability (Figure S1B). However, RpL4Δ63–87 was not present in mature but rather in pre-60S particles that arrested at a late stage, as suggested by the co-enrichment of nuclear export factor Nmd3 and cytoplasmic assembly factors Lsg1 and Yvh1 (Figure 3B). Apparently, 60S subunits carrying RpL4Δ63–87 cannot finally mature, indicating that correct insertion of the RpL4 loop into the lining of the peptide exit tunnel is linked to a unknown checkpoint control implemented in a late 60S maturation step. Altogether, the data indicate that Acl4 has an assisting role in RpL4 ribosome assembly.

The RpL4 C-Terminal Extension Harbors a Nuclear Import Signal and Plays a Pivotal Role in the Disassembly of the Acl4•RpL4 Complex at the Pre-Ribosome

Owing to the finding that the RpL4 C-terminal extension is not required for Acl4 binding but is essential for cell growth (see Figures 3B, S1B), we looked for other eukaryote-specific functions. The RpL4 extension contains two putative PY-NLSs in series (residues 301–345), which induce a strong nuclear accumulation when fused to GFP (Figure 3D). Mutation of three lysine residues (K314A, K315A, K319A) in this extended NLS (residues 277–362) significantly diminished the NLS activity. Moreover, shorter constructs of this sequence (ranging from residues 303–320 or 311–333), which still carry these crucial lysine residues, have lost nuclear targeting activity. Altogether this data suggests that RpL4 contains an extended and complex NLS in its C-terminal extension.

Prompted by these findings, we tested interaction of the Acl4•RpL4 complex with nuclear import receptors. It was possible to bind in vitro the ctAcl4•ctRpL4 heterodimer to the PY-NLS receptor ctKap104 (Kressler et al., 2012; Suel and Chook, 2009), forming a stoichiometric ctAcl4•ctRpL4•ctKap104 complex as determined by size-exclusion chromatography coupled to multiangle light scattering (Figure 2A). However, scKap104 was not bound to the scAcl4•scRpL4ΔExt dimer, lacking the PY-NLS region (residues 277–362) (Figure 3E). These data suggest that Kap104 binds to the C-terminal extension of RpL4 to mediate nuclear import of the Acl4•RpL4 complex.

In the nucleus, RpL4 has to dissociate from Acl4 to be incorporated into the nascent ribosome. Previous studies established that RpL4 incorporation occurs early in ribosome formation together with several other RPs, including RpL18, RpL7 and RpL20 (Ben-Shem et al., 2011; Gamalinda and Woolford, 2014), which all reside in the vicinity of the C-terminal RpL4 extension in the fully assembled 60S ribosomal subunit (Figures 1A, 3C). Hence, we hypothesized that the interaction of the RpL4 extension with the pre-ribosomal surface could trigger the disassembly of the Acl4•RpL4 complex and would allow the insertion of the RpL4 molecule into the nascent pre-60S ribosome. To test this hypothesis, we mutated residues in two regions of the RpL4 extension that directly contact either RpL18 (corresponding mutation in RpL4: I289A, I290A, I295A) or the eukaryote-specific expansion segment 7 (ES7) of the 25S rRNA (corresponding mutation in RpL4: K332E and F334A) (Figure 3C). In vivo, both sets of RpL4 extension mutants were efficiently imported into the nucleus, but exhibited a slow growth phenotype at elevated temperatures (Figures 4A, S4B). Pulse-chase assays showed that the RpL4 mutant deficient in RpL18 binding was inefficiently released from Acl4 with a delayed assembly into the 60S subunit (Figure 4B). The RpL4 mutant deficient in ES7 binding showed a similar albeit weaker defect, underlining that the RpL4 C-terminal extension uses multiple contact sites for recruitment to the nascent 60S subunit.

Figure 4. The C-Terminal Extension of RpL4 Coordinates the Incorporation of RpL4 into the Pre-Ribosome.

(A) Growth analysis of RPL4 (wt), RPL4-RPL18 contact and RPL4-ES7 contact mutants.

(B) Epitope pulse-chase analysis of RpL4 (lanes 1–3) and RpL4-RpL18 contact mutant (I289A, I290A, and I295A) (lane 4–6) and RpL4-ES7 contact mutant (K332E and F334A) (lanes 7–9). RpL4 and RpL4 mutants were pulsed for 0 minutes (lanes 1, 4, 7) or 5 minutes (lanes 2, 5, 8) as a GAL::tcapt-HA-RpL4-Flag-ProtA version and subsequently chased for 19 minutes (lanes 3, 6, 9), tandem affinity-purified (TAP) and separated by SDS-PAGE and analyzed by silver staining and western blot. Blot and silver gel were sliced and put together from one gel/blot image (original blot and gel, see Figure S5). Accordingly, wild-type RpL4 (lanes 1–3) in Figures 3B and 4B are derived from the same gel.

(C) Close up view of the RpL4-RpL18 interaction, as observed in the S. cerevisiae 80S ribosome (PDB 3U5E) (Ben-Shem et al., 2011). Surface depiction of RpL18 (grey) and ribbon illustration of the RpL18 contact region of the C-terminal RpL4 extension (residues 277–301, red). Mutated RpL18 residues are indicated in blue (PDB 3U5E, 3U5D) (Ben-Shem et al., 2011).

(D) Ribosome profiles of wild-type and rpl4 I289A I290A I295A or rpl18 L32E V129D (RpL4-RpL18 contact mutant) strains analyzed by sucrose gradient centrifugation of cell homogenate.

(E) Tandem affinity-purification (TAP) of chromosomal tagged RpL4 in RPL18 (lanes 1 and 2) and rpl18 L32E V129D (lanes 3 and 4) strains. Purified samples were separated by SDS-PAGE and analyzed by Coomassie staining and western blot (lower panel). Cells (lanes 2 and 4) were shifted to 37 °C for 2.5 hours. The band labeled as Acl4 was identified by mass spectrometry.

(F) Model of coordinated RpL4 assembly into the maturing ribosome. Details as discussed in the text. See also Figure S4 and S5.

To verify that RpL18 is a critical factor assisting in the release of RpL4 from Acl4, we mutated hydrophobic RpL18 residues (L32E, V129D) that are in contact with hydrophobic RpL4 residues on the mature ribosome (Figures 3C, 4C). In vivo, the RpL18 mutant deficient in RpL4 binding showed similar defects as the RpL4 mutant deficient in RpL18 binding, with impaired cell growth at 37 °C and a defective 60S subunit synthesis (Figures 4A, 4D). Finally, the purification of chromosomally TAP-tagged RpL4 from strains harboring either wild-type RpL18 or the RpL4-binding deficient mutant revealed that the latter indeed showed an enrichment of Acl4 (Figure 4E). This data suggests that the C-terminal extension of RpL4 delivers the Acl4•RpL4 complex to the pre-ribosome, triggering RpL4 release from Acl4 and incorporation into the 60S ribosomal subunit.

In conclusion, we propose a model how a nascent ribosomal protein (RpL4) can be incorporated into the pre-ribosome in a hierarchical fashion (Figure 4F). The key to this coordinated process is an assembly chaperone, Acl4, which shields RpL4 until timely release and insertion into the pre-ribosome is possible. RpL4 shielding could encompass the prevention of non-productive interactions or cellular degradation. RpL4 dissociation from Acl4 is triggered by the eukaryote-specific extension of RpL4, which contacts co-evolved sites on the pre-60S surface. The ~100 fold higher abundance of RpL4 suggests that Acl4 enters a new assembly cycle after RpL4 delivery. Such a mechanism could explain how eukaryotic cells achieve coordinated assembly of interdependent RPs into the maturing ribosome and shines light on the evolution of eukaryotic RP extensions regarding a role in ribosome assembly. The finding that no obvious mammalian Acl4 ortholog could be identified despite its sequence conservation in many eukaryotic species requires further investigation to allow for the generalization of this mechanism, but at the same time opens the door for the development of novel anti-fungal agents.

EXPERIMENTAL PROCEDURES

Yeast Strains and Genetic Methods

The S. cerevisiae strains used in this study are listed in Table S1. Gene disruption and C-terminal tagging were performed as recently described (Janke et al., 2004; Longtine et al., 1998).

Plasmid Constructs

Recombinant DNA techniques were performed according to standard procedures. Plasmids used in this study are listed in Table S2 and S3.

Epitope Pulse-Chase Assay of RpL4

The epitope pulse-chase experiments were performed according to the recently described protocols (Stelter and Hurt, 2014; Stelter et al., 2012). Yeast strains were pre-cultured in SDC-LEU/-TRP and diluted into 2 L YP-raffinose medium with a starting OD₆₀₀ of 0.4. At OD₆₀₀ of 1.5 cells were induced with galactose for 18 minutes and pulsed for 5 minutes with Ome-tyrosine and chased for 19 minutes with tetracycline-glucose. HA-RpL4-FPA was purified using lysis-buffer (20 mM TRIS-HCl (pH 7.5), 150 mM NaCl, 50 mM potassium acetate, 2 mM magnesium acetate, 0.01 % (w/v) NP-40) according to the standard purification protocol and eluates were resolved on a SDS-PAGE gel and visualized by Coomassie Brilliant Blue or silver staining. Associated proteins were analyzed by mass-spectrometry or western blot using the indicated antibodies.

Protein Purification, Crystallization, and Structure Determination

Proteins were expressed and purified from E. coli using standard methods. Crystals of ctAcl4¹⁻³³⁸ and ctAcl4²⁸⁻³³⁸ were obtained in hanging drops at 15 mg/ml in 0.2 M ammonium citrate tribasic (pH 7.0) and 20 % (w/v) PEG 3350 and 0.2 M potassium formate and 20 % (w/v) PEG 3350, respectively. X-ray diffraction data were collected at 100 K at beamline GM/CA-CAT 23ID-D at the Advanced Photon Source (APS). The ctAcl4 structure was solved by experimental phasing. For details of the data collection and refinement statistics see Table 1.

Miscellaneous

Further methods used in this study and previously described were sucrose gradient profiles to obtain ribosomal and polysome profiles, ribosomal export assays using the 60S reporter (RpL3-GFP) and import assay using 3xyEGFP fused to RpL4 domains were monitored by fluorescent microscopy, tryptic digest of recombinant ctAcl4•ctRpL4 complex (Kressler et al., 2012; Kressler et al., 2008). Live yeast cells were imaged by fluorescence microscopy using either a Zeiss Imager Z1 or an Olympus BX54 microscope.