Physical and Functional Interaction between the Methyltransferase Bud23 and the Essential DEAH-Box RNA Helicase Ecm16

Abstract

The small ribosomal subunit assembles cotranscriptionally on the nascent primary transcript. Cleavage at site A2 liberates the pre-40S subunit. We previously identified Bud23 as a conserved eukaryotic methyltransferase that is required for efficient cleavage at A2. Here, we report that Bud23 physically and functionally interacts with the DEAH-box RNA helicase Ecm16 (also known as Dhr1). Ecm16 is also required for cleavage at A2. We identified mutations in ECM16 that suppressed the growth and A2 cleavage defects of a bud23Δ mutant. RNA helicases often require protein cofactors to provide substrate specificity. We used yeast (Saccharomyces cerevisiae) two-hybrid analysis to map the binding site of Bud23 on Ecm16. Despite the physical and functional interaction between these factors, mutations that disrupted the interaction, as assayed by two-hybrid analysis, did not display a growth defect. We previously identified mutations in UTP2 and UTP14 that suppressed bud23Δ. We suggest that a network of protein interactions may mask the loss of interaction that we have defined by two-hybrid analysis. A mutation in motif I of Ecm16 that is predicted to impair its ability to hydrolyze ATP led to accumulation of Bud23 in an ∼45S particle containing Ecm16. Thus, Bud23 enters the pre-40S pathway at the time of Ecm16 function.

INTRODUCTION

Ribosomes are intricately folded and highly complex ribonucleoprotein (RNP) particles. Although bacterial ribosomes can be assembled from their constituents in vitro and require only a small number of essential assembly factors in vivo (reviewed in reference 1), eukaryotic ribosomes require more than 200 trans-acting factors for their assembly, many of which are essential (2, 3). The assembly of ribosomes is a highly dynamic process (4) involving continuous remodeling of RNA-RNA, protein-protein, and protein-RNA interactions (5, 6). Structural and conformational changes mediated by these interactions in a timely manner are critical for ensuring the proper assembly of a functional ribosome (7–9).

The biogenesis of ribosomes in yeast (here Saccharomyces cerevisiae) begins with the transcription of 35S pre-rRNA in the nucleolus. The small subunit is assembled first, as 18S rRNA is encoded in the 5′ end of the polycistronic 35S precursor. Multiple trans-acting factors, including snoRNAs, assemble onto the growing transcript cotranscriptionally, leading to an early preribosomal precursor particle, the 90S precursor, which contains the small subunit (SSU) processome (10–12). The primary transcript undergoes extensive modification and processing (13). The major early cleavage events at sites A0 and A1 in the 5′-external transcribed spacer and at A2 in internal transcribed spacer 1 (ITS1) as well as RNA folding require U3 snoRNA (14–16). Although we know a considerable amount about the composition of the SSU processome, which is comprised of approximately 70 proteins (17), little is known about its disassembly.

Many remodeling events in ribosome biogenesis require energy-consuming enzymes, including DEAH/D-box RNA helicases, AAA-ATPases, and GTPases (7, 18–20). Indeed, there are 19 RNA helicases involved in ribosome biogenesis in yeast, 17 of which are essential (21–24). Several DEAH/D-box RNA helicases are required for the removal of snoRNAs that base pair with the pre-rRNAs (25–27) or remodel the RNPs by dissociating RNA-RNA or RNA-protein interactions (28). These structural rearrangements may be essentially irreversible or promote irreversible reactions, such as RNA cleavages (19, 29). Therefore, the regulation of the enzymatic activity of these helicases is important to ensure proper coordination of assembly events (7). However, most DEXD/H helicases do not have intrinsic substrate specificity (30). These enzymes contain, at a minimum, a core helicase module built of tandem RecA-like domains. Their substrate specificity typically comes from flanking N- and C-terminal domains that harbor binding sites for cofactors needed to recruit the helicase to substrates. Regulatory cofactors may also bind to these flanking sequences (31).

We previously identified Bud23 as the methyltransferase that modifies the guanosine base at position 1575 in 18S rRNA (32). Our initial characterization of Bud23 suggested that it acted late in the 40S biogenesis pathway, as it stably associated with pre-40S, after cleavage at A2 (32). More recently, we have shown that Bud23 functionally interacts with two SSU processome factors, Utp2 and Utp14, and that Bud23 is required for efficient cleavage at A2 (33), suggesting that it enters the 40S biogenesis pathway prior to A2 cleavage. Here we show that Bud23 functionally and physically interacts with the essential DEAH box RNA helicase Ecm16/Dhr1. We suggest that Bud23 is a component of a network of proteins that regulate Ecm16 activity.

MATERIALS AND METHODS

Strains, plasmids, and media.

All yeast strains used in this work are listed in Table 1. All strains were grown at 30°C in rich medium or synthetic dropout (SD) medium containing 2% glucose, unless otherwise indicated. The plasmids and oligonucleotides used in this work are listed in Tables 2 and 3, respectively.

TABLE 1.

S. cerevisiae strains used in this study

Strain	Genotype	Source or reference
AJY2161	MATa bud23Δ::KanMX6 his3Δ1 leu2Δ0 ura3Δ0 met15Δ0	32
AJY3321	MATa KanMX6-PGAL1-ECM16 bud23Δ::NatMX his3Δ1 leu2Δ0 met15Δ0 ura3Δ0	This study
AJY3325	MATα trm112Δ::KanMX6 trp1-901 leu2-3,112 ura3-52 his3Δ200 gal4Δ gal80Δ LYS2::GAL1-HIS3 GAL2-ADE met2::GAL7-LACZ	This study
AJY3447	MATa ECM16-GFP::HIS3MX6 leu2Δ0 ura3Δ0	48
AJY3517	MATα trm112Δ::G418r his3Δ1 leu2Δ0 ura3Δ0	39
AJY3711	MATa KanMX6-PGAL1-3×HA-ECM16 his3Δ1 leu2Δ0 met15Δ0 ura3Δ0	This study
BY4741	MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0	Open Biosystems
PJ69-4α	MATα trp1-901 leu2-3,112 ura3-52 his3Δ200 gal4Δ gal80Δ LYS2::GAL1-HIS3 GAL2-ADE met2::GAL7-LACZ	35

TABLE 2.

Plasmids used in this study

Plasmid	Description	Source or reference
pAJ2037	ecm16-M744T URA3 CEN ARS	This study
pAJ2039	ecm16-E831K URA3 CEN ARS	This study
pAJ2291	Gal4BD–c-myc–bud23-G152E TRP1 2μ	This study
pAJ2292	Gal4BD–c-myc–bud23-D99G TRP1 2μ	This study
pAJ2293	Gal4BD–c-myc–bud23-L42P TRP1 2μ	This study
pAJ2294	Gal4BD–c-myc–bud23-L37P TRP1 2μ	This study
pAJ2295	Gal4BD–c-myc–bud23-L170P TRP1 2μ	This study
pAJ2296	Gal4BD–c-myc–bud23-I51S TRP1 2μ	This study
pAJ2297	Gal4BD–c-myc–bud23-S110P TRP1 2μ	This study
pAJ2299	Gal4BD–c-myc–bud23-D94G TRP1 2μ	This study
pAJ2300	Gal4BD–c-myc–bud23-R107L TRP1 2μ	This study
pAJ2301	Gal4BD–c-myc–bud23-D112G TRP1 2μ	This study
pAJ2302	Gal4BD–c-myc–bud23-F178S TRP1 2μ	This study
pAJ2303	Gal4BD–c-myc–bud23-W73R TRP1 2μ	This study
pAJ2311	ECM16-13myc LEU2 CEN ARS	This study
pAJ2312	ECM16-C–HIS6 in pET21a	This study
pAJ2558	BUD23-TAP LEU2 CEN ARS	This study
pAJ2593	WT ECM16 URA3 CEN ARS	This study
pAJ2594	ecm16-E430K URA3 CEN ARS	This study
pAJ2762	Gal4AD-HA-ecm16 LEU2 2μ	This study
pAJ2768	Gal4AD-HA-WT BUD23 LEU2 2μ	This study
pAJ2773	Gal4BD–c-myc–bud23 TRP1 2μ	This study
pAJ2774	Gal4BD–c-myc–myc–bud23ΔN TRP1 2μ	This study
pAJ2775	Gal4BD–c-myc–bud23ΔC TRP1 2μ	This study
pAJ2787	6×His-Bud23 Trm112 in pRSFDuet-1	This study
pAJ2795	Gal4AD–HA–Ecm16 aa467–1267 gene LEU2 2μ	This study
pAJ2796	Gal4AD–HA–Ecm16 aa1–142 gene LEU2 2μ	This study
pAJ2797	Gal4AD–HA–Ecm16 aa1–467 gene LEU2 2μ	This study
pAJ2895	TRM112–Gal4BD–c-myc TRP1 2μ	This study
pAJ2922	Gal4BD–c-myc–WT ECM16 TRP1 2μ	This study
pAJ2930	Gal4BD–c-myc–ecm16-N231D TRP1 2μ	This study
pAJ2932	Gal4BD–c-myc–ecm16-V252A TRP1 2μ	This study
pAJ2933	Gal4BD–c-myc–ecm16-K200 STOP TRP1 2μ	This study
pAJ2934	Gal4BD–c-myc–ecm16-F195-FSa TRP1 2μ	This study
pAJ2935	Gal4BD–c-myc–ecm16-K171-FS TRP1 2μ	This study
pAJ2937	Gal4BD–c-myc–ecm16-K237E TRP1 2μ	This study
pAJ2938	Gal4BD–c-myc–ecm16-K245-FS TRP1 2μ	This study
pAJ2945	Gal4BD–c-myc–ecm16-K247D TRP1 2μ	This study
pAJ3053	ecm16-N231D URA3 CEN ARS	This study
pAJ3054	ecm16-V252A URA3 CEN ARS	This study
pAJ3055	ecm16-K237E URA3 CEN ARS	This study
pAJ3056	ecm16-F229S E296G Q334R F374L URA3 CEN ARS	This study
pAJ3057	ecm16-N247D URA3 CEN ARS	This study
pAJ3063	ecm16-N231D K237E N247D URA3 CEN ARS	This study
pAJ3064	ecm16-N231D K237E V252A URA3 CEN ARS	This study
pAJ3065	Gal4AD–HA–bud23-G152E LEU2 2μ	This study
pAJ3066	Gal4AD–HA–bud23-D99G LEU2 2μ	This study
pAJ3067	Gal4AD–HA–bud23-L42P LEU2 2μ	This study
pAJ3068	Gal4AD–HA–bud23-L37P LEU2 2μ	This study
pAJ3069	Gal4AD–HA–bud23-L170P LEU2 2μ	This study
pAJ3071	Gal4AD–HA–bud23-S110P LEU2 2μ	This study
pAJ3072	Gal4AD–HA–bud23-D94G LEU2 2μ	This study
pAJ3073	Gal4AD–HA–bud23-R107L LEU2 2μ	This study
pAJ3074	Gal4AD–HA–bud23-D112G LEU2 2μ	This study
pAJ3075	Gal4AD–HA–bud23-F178S LEU2 2μ	This study
pAJ3076	Gal4AD–HA–bud23-W73R LEU2 2μ	This study
pAJ3081	ecm16-K420A–13myc LEU2 CEN ARS	This study
pAJ3082	WT ECM16 LEU2 CEN ARS	This study
pACT2	LEU2 2μ	Clontech
pGBKT7	TRP1 2μ	34
pRS416	URA3 CEN ARS

^aFS, frameshift.

TABLE 3.

Oligonucleotides used in this study

Oligonucleotide	Target(s)	Sequence
AJO603	35S, 27SA2, 23S, 21S	TGTTACCTCTGGGCCCCGATTG
AJO130	20S	TCTTGCCCAGTAAAAGCTCTCATGC
AJO190	18S	GTCTGGACCTGGTGAGTTTCCC
AJO962	U2	GCGACCAAAGTAAAAGTCAAGAACGACTCCACAAGTGCGAGGGTCGCGAC
AJO1566	ECM16	GATCCTCGAGGCTGAGGGCTTATTCTTGCC
AJO1567	ECM16	GGTCACTAGTGGCTCATCGTCACTATATTGG
AJO1575	TRM112	GCGCCATGGAGTTCTTAACCACCAACTTCC
AJO1576	TRM112	CGCGGATCCTTATACCAGGTGTGGAGGTAAC
AJO1633	Gal4BD	GATAATGTGAATAAAGATGCCG
AJO1636	ECM16	CGGCAGATCTCGAGTTTTTTTTCTTTCTCTTCACCTGTG
AJO1640	ECM16	CTATGATAATGGAGGAATATTTAG
AJO1646	BUD23	GATCATATGGGCAGCAGCCACCACCATCACCATCACGGCAGCATGTCACGTCCTGAGGAG
AJO1647	BUD23	GCTTTAATTAACTAGAACCTGTGTCTTCTTTTC
AJO1686	U3	TAGATTCAATTTCGGTTTCTC
AJO1730	ECM16	CCTCATATGGGTACTTACAGAAAAAGGTTTAATGAAAAAGCCAGATCCGGCCACATGGCC
AJO1850	22S	CCCACCTATTCCCTCTTGC

Suppressing mutations in ECM16.

The Ecm16-E430K extragenic suppressor was identified by high-throughput SOLiD sequencing of genomic DNA from a spontaneous mutant that arose in AJY2161. Sequencing was carried out by the Genome Sequencing and Analysis Facility at the University of Texas at Austin.

Screen to identify additional mutants of ECM16 as suppressors of the bud23Δ mutant.

ECM16 was amplified by PCR using oligonucleotides AJO1566 and AJO1567 with Taq DNA polymerase. The PCR product was cotransformed with MscI-digested pAJ2593 into the bud23Δ mutant strain (AJY2161). The transformants were selected on synthetic dropout medium lacking uracil (SD ura−) at 30°C. Plasmids were isolated from fast-growing transformants and retransformed into AJY2161 to confirm the plasmid dependence of the suppression phenotype. The suppressing clones were sequenced, and in cases where multiple mutations were identified, the individual mutations responsible for suppression were identified.

Yeast two-hybrid interaction assay.

For two-hybrid analysis, the indicated Gal4 activation domain (Gal4AD) and binding domain (Gal4BD) plasmids (34) were transformed into PJ69-4α (35) and selected on synthetic dropout medium lacking leucine and tryptophan (SD Leu− Trp− medium). The transformants were then patched on SD Leu− Trp− His−to test for two-hybrid interaction.

(i) Reverse two-hybrid screen to identify Bud23 mutants defective for interaction with Ecm16.

BUD23 was amplified by PCR with Taq DNA polymerase. The PCR product was cotransformed with NdeI- and NcoI-digested pAJ2773 into PJ69-4α containing pAJ2762 (AD-Ecm16). Transformants were selected on SD Leu− Trp− medium at 30°C. Ten thousand yeast colonies were screened by plating on SD Leu− Trp− medium and replica plating onto SD Leu− Trp− His− medium at 30°C. Colonies that did not grow on SD Leu− Trp− His− medium were considered to be candidates for interaction-defective Bud23. Plasmids were recovered and transformed back into PJ69-4α containing pAJ2762 (AD-Ecm16) to confirm loss of the two-hybrid interaction between Bud23 and Ecm16. The mutations in BUD23 were identified by sequencing.

(ii) Reverse two-hybrid screen to identify Ecm16 mutants defective for interaction with Bud23.

ECM16 was randomly mutagenized with Taq DNA polymerase using oligonucleotides AJO1633 and AJO1640. The PCR product was cotransformed with SmaI- and PshAI-digested pAJ2922 into strain PJ69-4α containing pAJ2768 (AD-Bud23) and selected on SD Leu− Trp− medium. Transformants were screened for loss of growth on SD Leu− Trp− His− medium by replica plating as described above. Putative candidates were isolated, retested for loss of two-hybrid interaction with Bud23, and tested for stability and expression by Western blotting. Mutations in ECM16 were identified by sequencing.

(iii) Two-hybrid interaction assay in trm112Δ background.

The trm112Δ::KanMX cassette was amplified from AJY3517 and transformed into strain PJ69-4α. The resulting strain, AJY3325 was confirmed for deletion of TRM112 by PCR. pAJ2773 and pAJ2762 were transformed into AJY3325 together or with the corresponding control vectors and tested for growth on SD Leu− Trp− His− medium containing 3-amino-1,2,4-triazole (3-AT).

Sucrose density gradient sedimentation.

Cells were grown at 30°C to an optical density at 600 nm (OD₆₀₀) of 0.3. Cycloheximide (100 μg/ml) was added to the cultures, followed by incubation in the 30°C shaker for 10 min. The cells were then poured onto ice and collected by centrifugation. All steps were carried out at 0 to 4°C. Cells were washed with lysis buffer (100 mM KCl, 50 mM Tris-HCl [pH 7.5], 5 mM MgCl₂, 150 μg/ml cycloheximide, 7 mM β-mercaptoethanol (βME), 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 μg/ml leupeptin, 1 μg/ml pepstatin A), and lysed by vortexing in the presence of glass beads. Extracts were centrifuged for 10 min at 15,000 × g, and 9 OD₂₆₀ units was loaded onto 7% to 47% sucrose gradients prepared in lysis buffer and centrifuged for 2.5 h at 40,000 rpm in a Beckman SW40 rotor. Gradients were fractionated using an Isco model 640 fractionator with continuous monitoring at 254 nm. Fractions were precipitated with 10% trichloroacetic acid (TCA) overnight at −20°C. The pellets were resuspended in Laemmli buffer and heated at 99°C for 5 min prior to separation on 8% SDS-PAGE gels, transferred to a nitrocellulose membrane, and subjected to Western blot analysis.

IP and Western blotting.

For immunoprecipitations (IPs), 200- to 250-ml cultures were grown in Leu− galactose to an OD₆₀₀ of 0.08 at 30°C, followed by the addition of 2% glucose for 6 h. Cells were resuspended in 500 μl of IP buffer (100 mM NaCl, 50 mM Tris-HCl [pH 7.5], 1.5 mM MgCl₂, 0.15% NP-40, 1 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml pepstatin A), lysed by vortexing with glass beads, and clarified by centrifugation at 15,000 × g at 4°C. Immunoprecipitation with the tandem affinity purification (TAP) tag was performed by incubating extracts with IgG-Sepharose beads (Amersham IgG-Sepharose 6 Fast Flow) for 2 h at 4°C, followed by tobacco etch virus (TEV) enzyme cleavage at 16°C for 2 h. The eluted proteins were precipitated by adding 10% TCA and incubated overnight at −20°C. The precipitated proteins were resuspended in Laemmli buffer, heated at 99°C for 5 min, and separated on an 8% SDS-PAGE gel. Immunoprecipitation for the 13×myc (13myc) tag was performed by incubating extracts with anti-myc monoclonal antibody (9e10) for 2 h at 4°C, followed by addition of protein A-conjugated beads and an additional incubation for 1 h. The beads were washed three times with buffer, and the immunoprecipitated proteins were eluted in Laemmli buffer by being heated at 99°C for 5 min and separated on an 8% SDS-PAGE gel. For Northern blotting from immunoprecipitated samples, TEV eluates were subjected to acid phenol-chloroform extraction. The RNA in the aqueous phase was precipitated with 2.5 volumes of ethanol and 1/10 volume sodium acetate (pH 5) at −20°C for 24 h. Proteins in the organic phase were precipitated with 2 volumes of acetone at −20°C for 24 h.

Northern blotting was performed as described previously (36). The hybridization signals were detected by phosphorimaging and quantified using Quantity One (Bio-Rad). Oligonucleotide probes are listed in Table 3.

Protein expression and purification.

The synonymous nucleotide mutation T48C was introduced to disrupt the NdeI site within ECM16. The ECM16 open reading frame was then amplified by PCR using oligonucleotides AJO1730 and AJO1636. The PCR product was cloned into the NdeI and XhoI sites of pET21a (Novagen) to make pAJ2312, which expresses Ecm16 with a C-terminal 6×His tag. The protein was expressed in BL21-CodonPlus(DE3)-RIL cells (Stratagene) by induction with isopropyl-β-d-thiogalactopyranoside (IPTG) overnight at 15°C. For protein purification, all steps were carried out at 0 to 4°C. Cell pellets were washed once in and resuspended in extraction buffer (20 mM Tris-HCl [pH 8.0], 500 mM NaCl, 10% (vol/vol) glycerol, 2 mM βME, 1 mM PMSF, and 1 μM (each) leupeptin and pepstatin). Lyzosyme was added to 0.05 mg/ml, and cells were incubated on ice for 30 min. Cells were disrupted by sonication, and extracts were clarified by centrifugation at 50,000 × g for 20 min. Imidazole was added to 10 mM, and the extract was loaded onto Ni-nitrilotriacetic acid (NTA) resin. The column was washed extensively with extraction buffer containing 20 mM imidazole followed by preelution buffer (20 mM Tris-HCl [pH 8.0], 150 mM NaCl, 10% [vol/vol] glycerol, 2 mM βME, 1 mM PMSF and 1 μM [each] leupeptin and pepstatin), and Ecm16 was eluted in preelution buffer containing 250 mM imidazole.

Bud23 and Trm112 were coexpressed from pAJ2787, derived from pRSFDuet-1. TRM112 was amplified from genomic DNA by PCR using the primers AJO1575 and AJO1576 and cloned into the NcoI and BamHI sites of pRSFDuet-1. BUD23 was amplified from genomic DNA using primers AJO1646 and AJO1647 and cloned into the NdeI and PacI sites of pRSFDuet-1 to give pAJ2727. Bud23 and Trm112 were coexpressed in BL21-CodonPlus(DE3)-RIL cells (Stratagene). Cells were grown and proteins purified as described previously (14).

Native agarose gel electrophoresis was performed by using 0.8% agarose gel in native gel buffer A (25 mM Tris-HCl, 19.2 mM glycine [pH 9] 10 μM ZnCl₂). Proteins were incubated separately or together at room temperature for 10 min, before being mixed with an equal volume of 2× sample loading buffer (20% glycerol, 0.2% bromophenol blue, 0.12 M Tris-HCl [pH 9]). Proteins were electrophoresed at 5 V/cm for 1 h at room temperature. Gels were then stained with Coomassie blue or transferred to nitrocellulose for Western blotting.

RESULTS

Loss of BUD23 is suppressed by dominant mutations in ECM16.

We previously reported the identification of spontaneous extragenic suppressors of bud23Δ in the SSU processome proteins Utp2 and Utp14 (33). We identified an additional suppressing mutation that did not map to either of these genes. Genomic sequencing identified a mutation in ECM16 (DHR1) encoding a DEAH-box RNA helicase that is required for the release of U3 snoRNA from pre-18S (R. Sardana, X. Liu, S. Granneman, J. Zhu, M. Gill, D. Tollervey, C. C. Correll, and A. W. Johnson, unpublished data). The suppressing mutation, ecm16-E430K, is close to the Walker A motif in the catalytic core of the protein. The ecm16-E430K mutation was dominant (data not shown), suggesting a gain of function. We screened for additional dominant suppressing mutations with the expectation that their location in Ecm16 would shed light on how these mutations affect Ecm16 activity.

The coding region of ECM16 was mutagenized using Taq DNA polymerase and cotransformed with a gapped vector to allow homologous recombination in a bud23Δ strain. Because ecm16-E430K was dominant, we were able to screen in a wild-type (WT) ECM16 background. From this screen and subsequent subcloning to separate mutations, we identified nine additional single point mutations (E360K, E402G, F567L, H593Y, R596C, M744T, E831G, E831K, and F837L) in ECM16 that were dominant suppressors of the growth defect of a bud23Δ mutant. The growth suppression of a subset of these mutants is shown in Fig. 1A.

FIG 1.

Mutations in ECM16 suppress the growth defect of a bud23Δ mutant. (A) Tenfold serial dilutions of AJY3321 (P_GAL-ECM16 bud23Δ::KanMX) containing empty vector (pAJ100) (row 1), AJY3711 (P_GAL-ECM16 BUD23 WT) containing wild-type ECM16 (pAJ2593) (row 2), and AJY3321 containing wild-type ECM16 (pAJ2593) (row 3), ecm16-E430K (pAJ2594) (row 4), ecm16-M744T (pAJ2037) (row 5), or ecm16-E831K (pAJ2039) (row 6) were spotted on SD ura− medium containing glucose and incubated at 30°C for 2 days. (B) Primary structure of Ecm16 showing the region modeled on Prp43 in color, conserved RecA-like domains, the winged helix, ratchet domain, and the OB folds as indicated. (C) Ecm16 (aa 319 to 1192) threaded on the Prp43 structure (PDB no. 3KX2) using the Phyre2 server (37). ADP as red sticks and residues that were mutated in bud23Δ suppressing mutants are indicated as magenta spheres.

That dominant ECM16 mutants suppress a bud23Δ mutant suggests that they bypass the function of Bud23 and implies that bud23Δ mutants are defective for Ecm16 activity. However, ECM16 is essential, whereas BUD23 is not. Thus, in the absence of Bud23, Ecm16 protein must be partially functional. One model is that Bud23 activates Ecm16 function and mutations in Ecm16 can bypass the need for Bud23 activation. DEAH-helicases have a conserved organization of the domains comprising the helicase activity flanked by N- and C-terminal extensions that are thought to provide interaction surfaces for accessory factors to provide substrate specificity (31). The helicase core contains tandem RecA domains, with the P-loop in RecA1, a ratchet domain, a degenerate winged-helix domain, and an oligonucleotide binding (OB) fold. RNA binds in the cleft between the RecA domains and the ratchet domain (37). We threaded the sequence of Ecm16 (amino acids [aa] 319 to 1192) on the crystal structure of Prp43, a related DEAH-box RNA helicase (37), and mapped the suppressing mutations on the structure (Fig. 1B and C). The mutations clustered in the conserved RecA1 and RecA2 domains of the helicase core. The affected residues were predominantly in solvent-exposed positions rather than in the catalytic center or the RNA binding cleft. Mutations in the RecA1 and RecA2 domains could alter the intrinsic activity of Ecm16 or could alter its interaction with binding partners that regulate its activity. In addition, several mutations (M744T, F567L, and F837L) are at the solvent-exposed edges of the interface between the RecA domains, where they could affect the dynamics of conformational changes in the protein. Like other DEAH-box RNA helicases, Ecm16 displays poly(A)-dependent ATPase activity (Sardana et al., unpublished). To ask if these mutations affect the intrinsic activity of Ecm16, we expressed and purified Ecm16-E831K and Ecm16-M744T and compared their RNA-dependent ATPase activity to that of wild-type (WT) Ecm16. The ATPase activity of the mutant proteins was not significantly different from that of the WT, suggesting that the suppressing mutations do not directly affect the catalytic function of Ecm16 (data not shown).

We showed previously that bud23Δ mutants are defective for A2 cleavage in internal transcribed spacer 1 (ITS1), which generates 27SA2 and results in reduced production of 18S rRNA (33) (Fig. 2, compare lanes 3 and 2). This cleavage event separates the pre-40S particle from the pre-60S particle during ribosome biogenesis. We asked if suppression of mutations in ECM16 restored A2 cleavage in a bud23Δ mutant. Indeed, the ecm16-suppressing mutations E430K, M744T, and E831K partially restored the levels of 27SA2 RNA, indicating restored processing at the A2 site (Fig. 2, compare lanes 4 to 6 to lane 3).

FIG 2.

Pre-rRNA processing in the bud23 mutant suppressed by mutations in ECM16. AJY3321 (P_GAL-ECM16 bud23Δ::KanMX) containing empty vector (pAJ100) (lane 1), AJY3711 (P_GAL-ECM16) containing wild-type ECM16 (pAJ2593) (lane 2), and AJY3321 containing wild-type ECM16 (pAJ2593) (lane 3), ecm16-E430K (pAJ2594) (lane 4), ecm16-M744T (pAJ2037) (lane 5), and ecm16-E831K (pAJ2039) (lane 6) mutants were grown in SD ura− galactose medium and then shifted to SD ura− glucose medium for 6 h at 30°C. All cultures were harvested at an OD₆₀₀ of ∼0.3. Total RNA was extracted using phenol-chloroform, separated on a 1% agarose–formaldehyde denaturing gel, transferred to a membrane, and probed with the indicated oligonucleotide probes to identify rRNA processing intermediates. U2 RNA was used as a loading control. The 27SA2 signal relative to U2 was quantified using ImageJ (NIH) and normalized to the wild type (27SA2/WT).

Physical interaction between Bud23 and Ecm16.

Because of the functional interaction between BUD23 and ECM16, we used two-hybrid analysis to ask if they also showed a physical interaction. Bud23 was expressed as an N-terminal fusion of the Gal4 DNA-binding domain (Gal4BD) in pGBKT7 and Ecm16 with an N-terminal fusion of the Gal4 activation domain (Gal4AD) in pACT2. A robust two-hybrid interaction between BUD23 and ECM16 was seen by growth on medium lacking histidine (Fig. 3A).

FIG 3.

Bud23 and Ecm16 interact. (A) Cells of yeast strain PJ69-4α containing Gal4BD-Bud23 (pAJ2773) and Gal4AD-Ecm16 (pAJ2762) were patched on SD Leu− Trp− and SD Leu− Trp− His− plates. Corresponding controls containing only Gal4BD-Bud23 and empty vector (pACT2) or Gal4AD-Ecm16 and empty vector (pGBKT7) were also patched on selective media to rule out the possibility of self-activation of the HIS3 reporter. Plates were incubated at 30°C for 2 days. (B) Extracts from cells expressing Ecm16-GFP (AJY 3447) and Bud23-TAP (pAJ2558) were immunoprecipitated with IgG-Sepharose. Extracts prepared from cells expressing Ecm16-GFP and untagged Bud23 were used as the negative control. Immunoprecipitated proteins were resolved on an 8% SDS-PAGE gel and Western blotted using anti-GFP–horseradish peroxidase (HRP) and anti-Rps8 antibodies to detect Ecm16-GFP and Rps8, respectively. (C) Extracts from cells expressing Ecm16-13myc (pAJ2311) or untagged Ecm16 (pAJ3082) were incubated with anti-myc antibody, followed by addition of protein A-agarose beads to immunoprecipitate Ecm16-associated protein complexes. Immunoprecipitated proteins were resolved on a 8% SDS-PAGE gel and Western blotted using anti-myc and anti-Bud23 antibodies to detect Ecm16-13myc and Bud23, respectively.

We carried out reciprocal coimmunoprecipitation experiments (Fig. 3B and C) to test if Bud23 and Ecm16 were in the same complex in vivo. A plasmid expressing C-terminal TAP-tagged Bud23 was transformed into a strain containing genomic C-terminally green fluorescent protein (GFP)-tagged Ecm16. Whole-cell extracts were immunoprecipitated against the TAP tag using IgG-Sepharose and analyzed for Ecm16-GFP by Western blotting using anti-GFP antibody. Ecm16-GFP was specifically immunoprecipitated with TAP-tagged Bud23 (Fig. 3B). Using a reciprocal approach, cells expressing C-terminally 13myc-tagged Ecm16 were immunoprecipitated with anti-myc antibody and analyzed for the presence of Bud23 by Western blotting using anti-Bud23 antibody. Indeed, Bud23 also specifically immunoprecipitated with Ecm16-13myc (Fig. 3C).

To support a direct interaction between Bud23 and Ecm16, we tested binding in vitro using purified proteins. Ecm16 was expressed with a C-terminal 6×His tag and purified from bacterial cells (Fig. 4A). C-terminally tagged Ecm16 is functional in yeast and the purified protein displayed poly(A)-dependent ATPase activity (data not shown). Because Bud23 is insoluble in the absence of Trm112 (38), we coexpressed N-terminal 6×His-tagged Bud23 and native Trm112 as described previously (38) (Fig. 4A). Under the conditions used for electrophoresis of native proteins (pH 9), free Ecm16 migrated as a diffuse band and more rapidly than the Bud23/Trm112 complex (Fig. 4B, left panel), consistent with their predicted pIs of 6.5 and 7.5, respectively. In the presence of Bud23/Trm112, Ecm16 appeared as a more discrete band, and its mobility was retarded (Fig. 4B, left panel). In the presence of Ecm16, no free Bud23/Trm112 complex was observed, and Western blotting revealed that Bud23 was present in the Ecm16-containing complex (Fig. 4B, right panel). These results suggest that Ecm16 and Bud23 interact physically as well as functionally.

FIG 4.

In vitro interaction between Bud23 and Ecm16. (A) Purified His-tagged Ecm16 and Bud23/Trm112 complex were separated by 8 to 15% SDS-PAGE and stained with Coomassie blue. The positions of molecular mass markers are indicated. Proteolytic fragments of Ecm16 are indicated by asterisks. (B) Binding reaction mixtures contained 1.2 μg of Ecm16 and/or 0.7 μg of Bud23/Trm112 complex as indicated. Samples were then electrophoresed under native conditions, and proteins were visualized by staining with Coomassie blue (left panel) or Western blotting for Bud23 (right panel).

Mapping the Bud23 and Ecm16 interaction domains: Bud23 interacts with the N terminus of Ecm16.

To define the region of Ecm16 responsible for interaction with Bud23, we performed truncation analysis of Ecm16. The conserved helicase core of Ecm16 extends from amino acids 319 to 1192. This is flanked by N-terminal (aa 1 to 319) and C-terminal (aa 1192 to 1267) regions that are not conserved among RNA helicases but conserved within the Ecm16 family. Taking advantage of existing restriction sites in the ECM16 sequence, we made three truncation constructs: one containing the first 142 aa of the N terminus, Ecm16 aa1–142, and one containing the entire N terminus plus part of the helicase core region, Ecm16 aa1–467, or containing the rest of the helicase core and the nonconserved C terminus, Ecm16 aa467–1267. Only the truncation mutant expressing aa 1 to 467, containing the entire N terminus of Ecm16, displayed two-hybrid interaction with Bud23 (Fig. 5A).

FIG 5.

Mapping the Bud23 binding site on Ecm16. (A) Strain PJ69-4α was transformed with Gal4BD-Bud23 (pAJ2773) and another plasmid encoding Gal4AD fusions of either full-length Ecm16 (pAJ2762), Ecm16 aa1–142 (pAJ2796), Ecm16 aa1–467 (pAJ2797), or Ecm16 aa467–1267 (pAJ2795). The transformants were patched on SD Leu− Trp− and SD Leu− Trp− His− media and incubated for 2 days at 30°C. (B) Relative positions of truncations (indicated by truncated lines) or missense mutations (solid dots). FS, frameshift. Relative growth on SD Trp− Leu− His− plates is indicated by + or −.

To identify point mutations within this region of Ecm16 that disrupt its interaction with Bud23, we mutagenized the Ecm16 aa1–467 construct and screened for mutations that disrupted two-hybrid interaction. We used random PCR mutagenesis of the N terminus of Ecm16 and cotransformed the PCR product with a gapped vector to allow in vivo recombination of the mutations into the two-hybrid vector. The transformants were screened for loss of growth on media lacking histidine. Candidates were scored for protein expression and stability by Western blotting and sequenced. From this screen, we identified multiple point mutations that disrupted the two-hybrid interaction with Bud23. The results are summarized in Fig. 5B. We also sequenced truncation mutants expressing the largest noninteracting fragments as a second means to identify a boundary for an interaction domain. The largest noninteracting truncation mutant contained a frameshift at position 195. A stop codon at position 200 gave weak growth, suggesting a boundary between amino acids 195 and 200. Point mutations that disrupted the interaction between Ecm16 and Bud23 were identified from aa 229 to 252. We conclude that amino acids 195 to 252 in the N-terminal extension of Ecm16 comprise an interaction surface with Bud23. Because this region of Ecm16 is not conserved between Ecm16 and any other DEAH box helicase for which there is a crystal structure, it is not currently possible to map these mutations on the structure of Ecm16.

Screen for Bud23 mutants that are defective for interaction with Ecm16.

We next sought mutations in Bud23 that disrupted interaction with Ecm16. The N-terminal region of Bud23 (aa 1 to 219) contains the methyltransferase domain, while the positively charged C terminus is required for nucleolar localization of Bud23. We cloned these two domains, the C terminus (bud23ΔN) and the N terminus (bud23ΔC), separately as Gal4BD fusions using pGBKT7. Only the N terminus of Bud23 supported two-hybrid interaction with Ecm16 (Fig. 6A). We then mutagenized the methyltransferase domain of Bud23 by random PCR mutagenesis and cotransformed it with gapped Gal4BD vector to allow in vivo recombination in our two-hybrid reporter strain carrying Ecm16 in a Gal4AD vector. Transformants were obtained and replica plated to screen for histidine auxotrophy to identify Bud23 mutants that were defective for interaction with Ecm16. The mutants were tested for expression by Western blotting to eliminate nonsense mutations and only candidates expressing full-length Bud23 were sequenced. The identified point mutations are listed in Fig. 6B.

FIG 6.

Mapping the Ecm16 binding surface on Bud23. (A) Strain PJ69-4α was transformed with Gal4AD-Ecm16 (pAJ2762) and a second plasmid encoding Gal4BD fusions of full-length Bud23 (pAJ2773), ΔN Bud23 (pAJ2774), or ΔC Bud23 (pAJ2775). The transformants were patched on SD Leu− Trp− and SD Leu− Trp− His− media and incubated for 2 days at 30°C. (B) Table of bud23 point mutants that showed loss of two-hybrid interaction with ECM16. Ntd, nucleotide.

We and others have recently shown that Bud23 requires the small methyltransferase partner protein Trm112 for stability (38, 39). This raised the possibility that the mutations we mapped on Bud23 affect Trm112 interaction, reducing Bud23 stability and only indirectly affecting Ecm16 interaction. In fact, some of the identified mutations were the same as those we previously identified as deficient in the Bud23-Trm112 interaction (39). Indeed, Bud23-Ecm16 two-hybrid interaction could not be detected in a trm112Δ mutant (data not shown), probably because Bud23 levels are significantly reduced in the absence of Trm112 (39). To tease apart the Bud23-Ecm16 and Bud23-Trm112 interacting residues, we asked if any of the bud23 mutants that showed a loss of interaction with Ecm16 retained interaction with Trm112. To do this, we tested two-hybrid interaction between Trm112 and the various Bud23 mutants on increasing concentrations of 3-AT. Bud23 mutants that maintained interaction with Trm112 in the presence of 3-AT are likely to be defective specifically for Ecm16 interaction and not Trm112 interaction. Four mutants showed 3-AT-resistant Trm112 interaction: those with the L37P, W73R, D94G, and D99G mutations (Fig. 7). We mapped these mutations to a modeled Bud23/Trm112 structure and compared their positions to those of mutations that disrupted interaction with Trm112 (Fig. 7B). The Ecm16 interaction mutants (Fig. 7B, green spheres) appear to define a surface distinct from that identified by two-hybrid assay as important for Trm112 interaction (Fig. 7B, yellow spheres). Surprisingly, D94 and D99 are squarely in the predicted Bud23-Trm112 interface but have a greater impact on Ecm16 interaction than on Trm112 interaction. In addition, the Trm112 interaction mutations extend along the bottom of Bud23, as viewed in Fig. 7B, away from the predicted Trm112 interface, suggesting an interaction that is not evident in the modeling. We note that residues 33 to 44 of Trm112 could not be modeled and could extend along this surface. That D94 and D99 appear to be important for Ecm16 interaction and also comprise part of the Trm112 interface suggests that Trm112 binding to this surface of Bud23 is mutually exclusive with Ecm16 binding.

FIG 7.

Separating Trm112 and Ecm16 interaction mutants. (A) Strain PJ69-4α containing Trm112-Gal4BD (pAJ2895) and empty vector, Gal4AD-WT Bud23 (pAJ2768), or the indicated Gal4AD-bud23 mutants on plasmids was patched on SD Leu− Trp− or SD Leu− Trp− His− plates containing 0, 2, 5, 10, or 20 mM 3-AT. Plates were incubated at 30°C for 2 days. Mutations that were resistant to 3-AT and maintained strong Trm112 interaction are indicated by green circles. These mutations preferentially disrupt interaction with Ecm16. Mutations that were sensitive to 3-AT, indicated by yellow circles, likely affect Bud23-Trm112 interaction. (B) Mutations in panel A were mapped to a model of Bud23/Trm112. A model of Bud23 (aa 1 to 219) (32) was manually docked into the structure of Encephalitozoon cuniculi Trm112/Mtq2 complex, replacing Mtq2 (PDB no. 3Q87). Residues identified as specific for Ecm16 interaction are indicated as green spheres, and residues directly affecting the Bud23-Trm112 interaction are shown as yellow spheres.

ecm16 mutants defective for Bud23 interaction can complement function in vivo.

To determine whether the ecm16 mutants defective for Bud23 interaction confer a biological defect in vivo, we subcloned five of the mutants into a low-copy-number vector under the control of their native promoter and asked if they could complement the growth defect from repressing P_GAL-ECM16. The single point mutants identified in the two-hybrid screen (Fig. 5B) were able to support growth on SD ura− medium containing glucose at 20, 30, and 37°C (Fig. 8A) (data not shown). We also combined mutations (N231D K237E N247D) and (N231D K237E V252A) to make two different triple mutants. The triple mutants grew indistinguishably from the WT (Fig. 8B). The lack of phenotype of ecm16 mutants that are weakened for interaction with Bud23, assayed by two-hybrid analysis, may suggest that loss of interaction in the context of the preribosome is buffered by the presence of a network of multiple protein interactions.

FIG 8.

ecm16 mutants defective for interaction with Bud23 complement function. (A) The yeast strain AJY3711 (P_GAL-ECM16) was transformed with plasmids harboring the indicated ecm16 mutant alleles and selected on SD ura− medium containing galactose. ecm16 FEQF was included despite containing multiple point mutations because it contained the F229S mutation present in the region like the other single point mutants. Tenfold serial dilutions of the mutants along with the corresponding controls were spotted on SD ura− medium containing galactose or glucose as the carbon source. Plates were incubated at 30°C for 2 days. (B) Tenfold serial dilutions of AJY3711 transformed with empty vector or vectors containing wild-type ECM16 or two triple mutant ecm16 alleles (ecm16-N231D K237E N247D or ecm16-N231D K237E V252A) were spotted on SD ura− medium containing galactose or glucose. Plates were incubated at 30°C for 2 days.

Bud23 enters the preribosomal particle before Ecm16 is released.

We initially characterized Bud23 as a protein that entered late in the pre-40S pathway (32), after release of the SSU processome and after A2 cleavage. However, genetic interaction with UTP2 and UTP14 and its role in A2 cleavage indicated that Bud23 is likely to enter the particle before A2 cleavage (33). Consequently, we considered the possibility that Ecm16 recruits Bud23 to the pre-40S particle or vice versa. The potential cosedimentation of Bud23 with pre-40S particles in WT or Ecm16-depleted cells was assayed by sedimentation through sucrose density gradients. Extracts were prepared from cells expressing WT Ecm16 or P_GAL-3×HA-Ecm16 depleted of Ecm16 by growth in glucose. Ecm16 tagged N terminally with 3 copies of the hemagglutinin epitope (3×HA) was reduced to levels below detection after 4.5 h in glucose (data not shown). Bud23 was detected in preribosomes in both strains (Fig. 9A). Thus, Bud23 does not depend on Ecm16 for recruitment to the preribosome. However, in the absence of Ecm16, the sedimentation of Bud23 was shifted to slightly lighter fractions, from approximately 80S in wild-type cells to ∼45S. Similarly, we asked if Bud23 was needed for recruitment of Ecm16 to the pre-40S particle. Because WT Ecm16 sediments primarily as free protein, and not with preribosomes (Fig. 9C), we needed a means to trap Ecm16 in the preribosome once it was recruited. For this, we employed an ATPase-defective mutant of Ecm16 (Ecm16-K420A) that stably accumulates in ∼45S particles containing U3 and core SSU processome components (Sardana et al., unpublished). This mutant was expressed in WT and bud23Δ mutant cells, and sedimentation of mutant Ecm16 was monitored by Western blotting. We saw no difference in sedimentation of Ecm16-K420A in the WT or bud23Δ mutant; in both cases, Ecm16-K420A sedimented at ∼45S (Fig. 9B). We asked if Bud23 also accumulated in this particle. Indeed, a fraction of Bud23 showed altered sedimentation (Fig. 9C), shifting to the position of Ecm16-K420A at ∼45S; this position is similar to that observed in the absence of Ecm16 (Fig. 9A). This result suggests that Bud23 enters the pre-40S particle before Ecm16 is released.

FIG 9.

Sedimentation analysis of Bud23 and Ecm16. (A) Whole-cell extracts from cycloheximide-treated AJY3711 (P_GAL-ECM16) grown in YP-galactose or shifted to YPD (glucose) medium for 6 h to deplete ECM16, were subjected to sucrose density gradient ultracentrifugation. Proteins were precipitated from fractions and subjected to SDS-PAGE. Bud23 and Rps8 were detected using anti-Bud23 and anti-Rps8 antibody, respectively. (B) pAJ3081 (ecm16-K420A–13myc) was transformed in wild-type and bud23Δ cells, and whole-cell lysates were subjected to sucrose gradient sedimentation analysis as described in panel A. Ecm16-K420A–13myc and Rps8 were detected using anti-myc and anti-Rps8 antibody, respectively. (C) Sedimentation analysis of cells expressing WT Ecm16-13myc (pAJ2311) or Ecm16-K420A (pAJ3081) from plasmids in AJY3711 shifted to glucose medium for 6 h to allow for the expression of mutant ecm16. Ecm16-13myc, Bud23, and Rps8 were detected using anti-myc, anti-Bud23, and anti-Rps8 antibody, respectively. The positions of 40S, 60S, and 80S are indicated.

DISCUSSION

Functional and physical interaction between Bud23 and Ecm16.

We identified dominant mutations in Ecm16 that suppressed the growth defect of the bud23Δ mutant. Moreover, they suppressed the rRNA processing defect of the bud23Δ mutant, a failure to cleave at site A2, an event that is also dependent on SSU processome components, including U3 snoRNA and Ecm16. Previously, we reported that mutations in two other SSU components, Utp2 and Utp14, also suppressed bud23Δ (33). Together, these results further support the idea that Bud23 is connected to SSU processome function. Dominant gain-of-function mutations could arise from activation of the native function of a protein or from novel, off target effects. Because we have identified suppressing mutations in multiple SSU processome components, we believe it is likely that these mutations activate or bypass the native function of a component of the SSU processome. In cytoplasmic maturation of 60S subunits, there are multiple examples of mutations in trans-acting factors that weaken their affinity for the preribosome, thereby bypassing the need for a releasing factor (40–42). Adopting this logic, we suggest that Bud23 is required for the efficient disassembly of components of the SSU processome and that mutations in SSU processome components bypass the in vivo function of Bud23.

The mutations that we identified in Ecm16 all map to the RecA1 and RecA2 domains but not to the nucleotide or RNA binding sites. Consequently, they may not directly impair or stimulate the ATPase or helicase activities of Ecm16. Consistent with this notion, these mutations did not alter the intrinsic RNA-dependent ATPase activity of the protein. The suppressing mutations do map to solvent-exposed surfaces and cluster in two regions of the protein, where it is easy to imagine they affect protein-protein interactions. However, which proteins these mutations affect remain to be identified.

Ecm16 interacts with Bud23, but disruption of this interaction does not have an obvious effect on cell growth.

In addition to the functional interaction between Ecm16 and Bud23, we also identified a physical interaction through two-hybrid and in vitro protein binding assays. We mapped the interaction to the N-terminal extension of Ecm16. The N- and C-terminal domains that flank the core helicase domains of RNA helicases typically confer specificity to the helicase through interaction with accessory proteins. A cofactor could recruit a helicase to its target complex or physically modulate the activity of the enzyme (31). Examples of both modes of action have been demonstrated. Binding of the G-patch protein Ntr1 to Prp43 helicase recruits it to function in spliceosome disassembly, whereas PfaI activates Prp43 ATPase activity (43, 44). Binding of Esf2 to Dbp8 helicase, involved in 40S biogenesis, stimulates the ATPase activity of Dbp8 (45). Although we propose that Bud23 is such an accessory factor for Ecm16, mutations in Ecm16 that abolished its interaction with Bud23, as measured by two-hybrid analysis, did not noticeably alter growth of cells. We suggest that either the physical interaction between Ecm16 and Bud23 is not required for their functional interaction or, in the context of the SSU processome, additional contacts buffer the effect of these mutations. We have reported previously that mutations in other SSU processome components (Utp2 and Utp14) can suppress the growth defect of a bud23Δ mutant (33). These factors may establish a network of interactions within the processome, and disruption of any one interaction may not be sufficient to result in a growth phenotype in vivo. Alternatively, the binding of Bud23 may induce a conformational change in the rRNA or in the protein-RNA environment of Ecm16 that activates its helicase activity. We have shown previously that Bud23 protein, but not its enzymatic function, is important for 40S biogenesis (32).

While mapping the Bud23 binding surface on Ecm16 was straightforward, mapping the Ecm16 interface on Bud23 was complicated by the fact that Bud23 also interacts with Trm112 (38, 39). Loss of Trm112 interaction results in loss of Bud23 and resulted in false positives for mutations affecting Ecm16 interaction. We used resistance of two-hybrid interactions to 3-AT to distinguish between Bud23 mutations that disrupted Trm112 interaction and those that disrupted Ecm16 interaction. Mutations that retained Trm112 interaction at high 3-AT concentrations but lost Ecm16 interaction were regarded to specifically disrupt interaction with Ecm16. Most of these mutations mapped to a surface distinct from that defined by mutations that disrupted Trm112 interaction (Fig. 7). Although a crystal structure for the Bud23/Trm112 complex is not yet available, the complex can be modeled on the structure of Trm112 in complex with a related methyltransferase, Mtq2 (46). Two of the residues that we identified as specific for Ecm16 interaction (D94 and D99) map to the predicted Trm112 binding surface. In addition, multiple mutations in Bud23 that are not predicted to be at the Bud23-Trm112 interface disrupted interaction with Trm112. These results suggest that the interaction with Trm112 may be more extensive than that predicted by our modeling. Alternatively, it is also possible that Bud23, Trm112, and Ecm16 form a heterotrimeric complex or that Trm112 bridges some of the interactions between Bud23 and Ecm16. Indeed, physical interaction identified by high-throughput affinity capture has been reported between Trm112 and Ecm16 (47), although we were unable to detect a two-hybrid interaction between Trm112 and Ecm16 (data not shown). However, it is also possible that the Bud23-Trm112 interaction is dynamic. We have previously concluded that Bud23 is not continuously associated with Trm112: possibly only free Bud23, not interacting with its substrate, requires Trm112 for stability (39). Thus, it is possible that Trm112 is released from Bud23 when it binds to its substrate, allowing Ecm16 to interact with a surface that also interacts with Trm112. Further experiments using in vitro reconstitution of this complex should settle this issue.

Does Bud23 monitor the status of assembly of the subunit?

Bud23 methylates G1575 in 18S rRNA. This residue is in the P-site of the small subunit in an RNA element that stacks coaxially on the central pseudoknot. In separate work, we have provided evidence that Ecm16 is responsible for the release of U3 snoRNA from 18S rRNA (Sardana et al., unpublished). U3 snoRNA base pairs with the 5′ end of 18S rRNA, and the U3-18S duplex precludes formation of the central pseudoknot. Thus, release of U3 is required for completion of the folding of the central pseudoknot. Bud23 is not a canonical component of the SSU processome. However, it does enter the small subunit processing pathway, while Ecm16 is associated with the preribosome, and the absence of Bud23 results in failure of cleavage at A2, a phenotype shared with other SSU processome components. We propose that Bud23 enters the SSU assembly pathway to trigger dissociation of the SSU processome, only after completion of transcription and folding of the majority of the small subunit. Our data support a model that Bud23 recruitment promotes SSU processome disassembly, possibly by stimulating the RNA helicase Ecm16.