The box C/D snoRNP assembly factor Bcd1 interacts with the histone chaperone Rtt106 and controls its transcription dependent activity

Abstract

Biogenesis of eukaryotic box C/D small nucleolar ribonucleoproteins initiates co-transcriptionally and requires the action of the assembly machinery including the Hsp90/R2TP complex, the Rsa1p:Hit1p heterodimer and the Bcd1 protein. We present genetic interactions between the Rsa1p-encoding gene and genes involved in chromatin organization including RTT106 that codes for the H3-H4 histone chaperone Rtt106p controlling H3K56ac deposition. We show that Bcd1p binds Rtt106p and controls its transcription-dependent recruitment by reducing its association with RNA polymerase II, modulating H3K56ac levels at gene body. We reveal the 3D structures of the free and Rtt106p-bound forms of Bcd1p using nuclear magnetic resonance and X-ray crystallography. The interaction is also studied by a combination of biophysical and proteomic techniques. Bcd1p interacts with a region that is distinct from the interaction interface between the histone chaperone and histone H3. Our results are evidence for a protein interaction interface for Rtt106p that controls its transcription-associated activity.

Fig. 1. Data from the GIM screen.

Two independent screens were performed, one with a mutant strain in which the RSA1 gene was deleted (rsa1Δ::Nat^R) and the other with a mutant used as reference strain. Two independent experiments were performed for each of these screens. The query strain and the reference strain were mated with a pool of strains containing all the viable strains from the haploid gene deletion collection. After selection of heterozygous diploids, sporulation and selection of the haploid, double mutants were grown for ~18 generations in rich liquid medium (YPD). Microarrays were used to measure the relative abundance of double mutants with query versus the reference populations. Normalized results are expressed as log2(Q/R) (see Methods section). Negative values indicate a synthetic growth defect. Positive values reveal either epistatic (buffering) or suppressive (alleviating) interactions between RSA1 and the selected genes. The genetic interactions with RSA1 are indicated by green arrows for the positive log2(Q/R) values and by red arrows for the negative values. The exact values are given in parentheses. HC heterochromatin.

Fig. 2. Bcd1p and histone chaperone Rtt106p form a stable heterodimer.

a Schematic representation of the Rtt106p and Bcd1p domain organization. DD dimeric domain, PH Plekstrin homology, PH1 + PH2 = MD middle domain, AR acidic region, ZHD zinc finger HIT domain, RBD Rtt106 binding domain, DR disorder region. Y2H assay: full-length Bcd1p (Bcd1p_FL) fused to the Gal4 binding domain (BD) interacts with the full-length Rtt106p (Rtt106p_FL) or fragment spanning amino acids 65–320 (Rtt106p-M) fused to the Gal4 activation domain (AD) as evidenced by growth on a His deprived medium supplemented with increased concentrations of 3-amino-1, 2, 4-triazol (3-AT). No interaction was observed with homologous fragments of FACT chaperone components Spt16p and Pob3p (Spt16p-M and Pob3p-M). b Interaction of Bcd1p_FL with Rtt106p_FL in yeast. Co-immunoprecipitation (co-IP) was performed on GAL1::3HA-BCD1 × RTT106-TAP cells expressing 3xHA-tagged Bcd1p_FL and TAP-tagged Rtt106p_FL. Cells expressing the nontagged Bcd1p were used as negative control. Extracts were incubated with anti-HA beads. The co-immunoselected proteins were analyzed by SDS-PAGE and western blotting. 10% of total proteins used per assay were loaded in the input lane. Tagged proteins were detected with PAP antibodies for Rtt106p and anti-HA antibodies for Bcd1p. The Dps1 protein used to control protein loading was detected using specific anti-Dps1p antibodies. The two panels correspond to a cropping of two sections of the same membrane. The full-length membrane is presented in the Source data file. The experiment was independently repeated three times with similar results. c Interaction of recombinant Bcd1p and Rtt106p in E. coli. His₆-tagged full-length or M domain of Rtt106p were co-expressed with Bcd1p_FL. His₆-tagged Bcd1p_FL was co-expressed with Rtt106p_FL or Rtt106p-M. Complexes were selected from crude extract by immobilized metal ion affinity chromatography (IMAC), fractionated by SDS-PAGE and revealed by Coomassie blue staining. The results correspond to co-expression with high salt (400 mM) buffer. Molecular weight markers (MW) in kilo Dalton (kDa) were loaded on the left. The experiment was repeated twice with similar results. The identity of the proteins in bands 1 A, 1B, 1 C, 1D, 1E, 1 F, 1 G, and 1H was confirmed by in-gel digestion of gel slices and mass spectrometry (MS) analysis of the peptide extract (Supplementary Table 1). d Bcd1p and Rtt106p interacting domains. ITC data for the interaction of Bcd1p_FL with Rtt106p-M (on the left) and Bcd1p_120-303 with Rtt106p_65-301 (on the right) recorded at 293 K in buffer containing 10 mM NaPi at pH 7.5, 150 mM NaCl and 0.5 mM TCEP. The calculated affinities Kd, and thermodynamic parameters as variations in enthalpy (ΔH) and entropy (ΔS) are indicated. e Nondenaturing MS characterization of the complex formed upon incubation of recombinant Bcd1p_FL with fragment Rtt106p-M. NanoESI mass spectra performed under nondenaturing conditions confirmed the presence of a 1:1 binding stoichiometry of Bcd1p_FL:Rtt106p-M complex (Da = Dalton). Source data for panel b are provided as a Source Data file.

Fig. 3. Interaction of Rtt106p with selected transcripts and loci.

a RNA immunoprecipitation (RIP) assays were performed on extracts prepared from RTT106-TAP; GAL1:3HA-BCD1 cells disrupted (rsa1Δ) or not of the RSA1 ORF. The cells were cultivated in the presence of galactose (YPG) or shifted to a glucose-containing medium (YPD) for 6 h before preparation of the extract. IP was performed using IgG-sepharose beads. RNAs retained on beads were quantified by RT-qPCR as described in the Methods section. The snoRNAs and pre-snoRNAs analyzed are indicated. b Chromatin immunoprecipitation (ChIP) assays were performed on yeast GAL1:3HA-BCD1 cells transformed with recombinant plasmid FLAG-RTT106; with (rsa1Δ) or without the disruption of the RSA1 ORF. Cells were cultivated in YPG or shifted to YPD medium before ChIP assays. Cells expressing nontagged Rtt106p (GAL1:3HA-BCD1 and GAL1:3HA-BCD1; rsa1Δ) were used as controls and IP was performed using anti-FLAG antibody. The loci analyzed by qPCR analysis are indicated. c Chip assays performed as in panel b on GAL1:3HA-BCD1 cells transformed with recombinant plasmid FLAG-RTT106. Primers were used for qPCR analysis on both side and across the U3 encoding gene (SNR17A). Data reported in the panels a and c are mean values plus standard error of the mean of three biological replicates (n = 3). Data reported in panel b are mean values plus standard error of the mean of three (for GAL1::3HA-BCD1, rsa1Δ + p413 (−) and GAL1::3HA-BCD1, rsa1Δ + p413::FLAG-RTT106; n = 3) or four (for GAL1::3HA-BCD1 + p413 (−) and GAL1::3HA-BCD1 + p413::FLAG-RTT106; n = 4) biological replicates. Two-tailed t-tests: *P < 0.05, **P < 0.01, ***P < 0.001. In panel b, for FLAG-RTT106 YPG versus YPD: * = 0.012 for HTA/B1, *** = 0.0006 for SNR128 (U14), * = 0.045 for SNR52, ** = 0.002 for SNR32, * = 0.032 for HMR a1. Note that SNR17A (U3) is close to significance: P = 0,056. For FLAG-RTT106 YPG versus rsa1Δ; FLAG-RTT106 YPG: * = 0.02 for SNR128 (U14) and * = 0.017 for SNR32. In panel c, for FLAG-RTT106 YPG versus YPD: ** = 0.006, * = 0.031, and ** = 0.001 from the left to the right. Source data for all these panels are provided as Source Data files.

Fig. 4. Bcd1p controls transcription-dependent activity of Rtt106p.

a Interaction of Bcd1p with the RNA polymerase II large subunit Rpb1p in yeast. Co-immunoprecipitation (co-IP) was performed on GAL1::3HA-BCD1 × RPB1-TAP cells transformed with p416GDP::FLAG-Rtt106p expressing 3xHA-tagged Bcd1p, TAP-tagged Rpb1p and FLAG-Rtt106p. Cells were transformed with empty vector p416GDP as a control. Cells were grown in YPG (+Bcd1p) or YPD (−Bcd1p). Procedure was as described in Fig. 2b but with the IP performed with anti-TAP antibodies. The image corresponds to a cropping of different sections of the same membrane. The full-length membrane incubated with different antibodies is presented in the Source data file. Histogram on the right presents quantification of Western blots obtained from six independent co-IP experiments. Quantification was performed using Fusion Solo (Vilber Lourmat) and Fusion-Capt Advance Solo 4 software. The FLAG signal (Rtt106p) was normalized to the TAP signal (Rpb1p). The data represent mean values plus standard error to the mean of four biological replicates. b Depletion of Bcd1p affects H3K56ac levels at several RNA polymerase II-dependent genes. ChIP H3K56ac enrichment (IP/INPUT) for parental BY4741 strain and the GAL1::3HA-BCD1 strain grown in YPG (+Bcd1p) or YPD (−Bcd1p) is presented. Signal specificity was controlled with IgG antibodies. The histogram represents mean values plus standard error to the mean of three biological replicates. Two-tailed t-tests: *P < 0.05, **P < 0.01, ***P < 0.001. In panel a, *** = 0.0001; in panel b, * = 0.011 for SNR17A (U3), ** = 0.002 for HTA/B1, * = 0.039 for ALG9. Source data for these panels are provided as Source Data files.

Fig. 5. Solution 3D structure of Bcd1p_120-303.

a ¹H-¹⁵N-HSQC spectra of Bcd1p_120-303. The assigned peaks are labeled (ω = frequency). The box shows a zoom of the center of the spectra. b Ribbon representation of the 20 best NMR solutions for the 3D structure of Bcd1p_120-303. Flexible loops are circled in gray and labelled. Secondary structure elements are α-helices (in dark green) and β-strands (in light green). N and C are the N-terminal and C-terminal extremities of the protein, respectively. c Two opposite views 180° apart in a cartoon representation of Bcd1p_120-303 with secondary structures labeled and numbered. The color code is the same as in panel b. d NMR heteronuclear nOe. Residue numbers are indicated at the bottom. Secondary structure elements are represented at the top in the same colors as in panel b. The flexible internal regions are highlighted in gray, and reported in panel b.

Fig. 6. Bcd1p_120-303 interacts with the PH1 domain of Rtt106p_65-301.

a Ribbon representation of the crystal structure of the complex between Rtt106p_65-301 and Bcd1p_120-303. Rtt106p is in orange and Bcd1p in green. b Superimposition of the PH1 domain of Rtt106p bound to Bcd1p (in orange) and in a free state (in blue, entry PDB code 3TW1 [https://www.rcsb.org/structure/3TW1]). H = Helix, S = Strand, N = N-terminal extremity, C = C-terminal extremity. c Residues located at the molecular surface of Rtt106p that are buried upon Bcd1p binding are in magenta. d Residues located at the molecular surface of Bcd1p that are buried in the Rtt106p:Bcd1p interface are in magenta. e–g Hydrophobic contacts, salt-bridges, and hydrogen bonds at the Rtt106p:Bcd1p interface: e Hydrophobic cluster (magenta) at the interface of the heterodimer. f Charged residues located at the interface form ionic interactions between Rtt106p and Bcd1p. g Network of hydrogen bonds between Rtt106p and Bcd1p. h, i Comparison of the 3D structures of Pop3p (in blue, entry PDB code 4PQ0 [https://www.rcsb.org/structure/4PQ0]) and Spt16p (in yellow, entry PDB code 4IOY [https://www.rcsb.org/structure/4IOY]). The sequence spanning amino acids 162–182 in Rtt106p, which differs strongly from Pob3p and Spt16p, is in red.

Fig. 7. Summary of the XL-MS and HDX-MS experiments.

Relative fractional uptake (RFU) differences export on the crystal structure of the complex between Rtt106p-M and Bcd1p_FL determined by hydrogen deuterium exchange mass spectrometry (HDX-MS)_. Export is realized for 2 min deuteration experiments. Differences are color scaled on Bcd1p_FL from blue (deprotection) to red (protection) upon Rtt106p-M binding (−10% to 10% RFU difference range). Differences are color scaled on Rtt106p-M from green (deprotection) to magenta (protection) upon Bcd1p_FL binding (−10% to 10% RFU difference range). The Cα-Cα distances of intercross-linked peptides are represented with black dotted lines. Orange residues and secondary structures represent the most affected regions on Bcd1p_FL while the purple ones represent the most affected regions on Rtt106p-M.