Advertisement

Differential Proteome and Interactome Analysis Reveal the Basis of Pleiotropy Associated With the Histidine Methyltransferase Hpm1p

Open AccessPublished:May 21, 2022DOI:https://doi.org/10.1016/j.mcpro.2022.100249

      Highlights

      • Novel extraribosomal functional phenotypes found on loss of Hpm1p.
      • Large-scale in vivo quantitative and comparative crosslinking analysis performed.
      • Changes detected in structures and interactions, independent of protein abundance.
      • Phenotypes could be explained or predicted through crosslinking analysis.

      Abstract

      The methylation of histidine is a post-translational modification whose function is poorly understood. Methyltransferase histidine protein methyltransferase 1 (Hpm1p) monomethylates H243 in the ribosomal protein Rpl3p and represents the only known histidine methyltransferase in Saccharomyces cerevisiae. Interestingly, the hpm1 deletion strain is highly pleiotropic, with many extraribosomal phenotypes including improved growth rates in alternative carbon sources. Here, we investigate how the loss of histidine methyltransferase Hpm1p results in diverse phenotypes, through use of targeted mass spectrometry (MS), growth assays, quantitative proteomics, and differential crosslinking MS. We confirmed the localization and stoichiometry of the H243 methylation site, found unreported sensitivities of Δhpm1 yeast to nonribosomal stressors, and identified differentially abundant proteins upon hpm1 knockout with clear links to the coordination of sugar metabolism. We adapted the emerging technique of quantitative large-scale stable isotope labeling of amino acids in cell culture crosslinking MS for yeast, which resulted in the identification of 1267 unique in vivo lysine–lysine crosslinks. By reproducibly monitoring over 350 of these in WT and Δhpm1, we detected changes to protein structure or protein–protein interactions in the ribosome, membrane proteins, chromatin, and mitochondria. Importantly, these occurred independently of changes in protein abundance and could explain a number of phenotypes of Δhpm1, not addressed by expression analysis. Further to this, some phenotypes were predicted solely from changes in protein structure or interactions and could be validated by orthogonal techniques. Taken together, these studies reveal a broad role for Hpm1p in yeast and illustrate how crosslinking MS will be an essential tool for understanding complex phenotypes.

      Graphical Abstract

      Keywords

      Abbreviations:

      BDP–NHP (biotin–aspartate proline–N-hydroxyphthalimide), BL (Boca and Leek’s FDR regression method), CSM (cross-link spectral match), FDR (false discovery rate), GO (Gene Ontology), H243 (histidine 243), H243Me (methylation of H243), HCD (higher-energy collision dissociation), Hpm1p (histidine protein methyltransferase 1), IAA (iodoacetamide), MS (mass spectrometry), NCE (normalized collision energy), PDB (Protein Data Bank), PIR (protein interaction reporter), PPI (protein–protein interaction), PTM (post-translational modification), RT (room temperature), SC (synthetic complete), SCX (strong cation exchange), SILAC (stable isotope labeling of amino acids in cell culture), UNSW (University of New South Wales), URP (unique residue pair)
      Protein methylation is an important post-translational modification (PTM), which occurs predominantly on lysine and arginine residues in eukaryotic cells. Methylation of histidine residues (reviewed in Refs. (
      • Kwiatkowski S.
      • Drozak J.
      Protein histidine methylation.
      ,
      • Jakobsson M.E.
      Enzymology and significance of protein histidine methylation.
      )) can also occur and was initially described for only a small number of proteins, including actin, myosin, myosin kinase, S100A9, the large subunit ribosomal protein L3, and more recently, a family of zinc transporter proteins (
      • Lv M.
      • Cao D.
      • Zhang L.
      • Hu C.
      • Li S.
      • Zhang P.
      • Zhu L.
      • Yi X.
      • Li C.
      • Yang A.
      • Yang Z.
      • Zhu Y.
      • Zhang K.
      • Pan W.
      METTL9 mediated N1-histidine methylation of zinc transporters is required for tumor growth.
      ,
      • Davydova E.
      • Shimazu T.
      • Schuhmacher M.K.
      • Jakobsson M.E.
      • Willemen H.L.D.M.
      • Liu T.
      • Moen A.
      • Ho A.Y.Y.
      • Małecki J.
      • Schroer L.
      • Pinto R.
      • Suzuki T.
      • Grønsberg I.A.
      • Sohtome Y.
      • Akakabe M.
      • et al.
      The methyltransferase METTL9 mediates pervasive 1-methylhistidine modification in mammalian proteomes.
      ). The most well-studied example is the methylation of human β-actin H73, which has been shown to have important roles in regulating the stability of the actin cytoskeleton. The responsible methyltransferase SETD3 (
      • Kwiatkowski S.
      • Seliga A.K.
      • Vertommen D.
      • Terreri M.
      • Ishikawa T.
      • Grabowska I.
      • Tiebe M.
      • Teleman A.A.
      • Jagielski A.K.
      • Veiga-Da-Cunha M.
      • Drozak J.
      SETD3 protein is the actin-specific histidine N-methyltransferase.
      ,
      • Wilkinson A.W.
      • Diep J.
      • Dai S.
      • Liu S.
      • Ooi Y.S.
      • Song D.
      • Li T.M.
      • Horton J.R.
      • Zhang X.
      • Liu C.
      • Trivedi D.v.
      • Ruppel K.M.
      • Vilches-Moure J.G.
      • Casey K.M.
      • Mak J.
      • et al.
      SETD3 is an actin histidine methyltransferase that prevents primary dystocia.
      ) is implicated in a number of human diseases, including viral pathogenesis (
      • Diep J.
      • Ooi Y.S.
      • Wilkinson A.W.
      • Peters C.E.
      • Foy E.
      • Johnson J.R.
      • Zengel J.
      • Ding S.
      • Weng K.F.
      • Laufman O.
      • Jang G.
      • Xu J.
      • Young T.
      • Verschueren E.
      • Kobluk K.J.
      • et al.
      Enterovirus pathogenesis requires the host methyltransferase SETD3.
      ), liver tumorigenesis (
      • Cheng X.
      • Hao Y.
      • Shu W.
      • Zhao M.
      • Zhao C.
      • Wu Y.
      • Peng X.
      • Yao P.
      • Xiao D.
      • Qing G.
      • Pan Z.
      • Yin L.
      • Hu D.
      • Du H.N.
      Cell cycle-dependent degradation of the methyltransferase SETD3 attenuates cell proliferation and liver tumorigenesis.
      ), and lymphoma (
      • Chen Z.
      • Yan C.T.
      • Dou Y.
      • Viboolsittiseri S.S.
      • Wang J.H.
      The role of a newly identified SET domain-containing protein, SETD3, in oncogenesis.
      ), thus confirming the functional importance of histidine methylation. Because of the lack of available affinity reagents, our understanding of histidine methylation is less complete than that of lysine and arginine. However, a recent stringent reanalysis of ultra-deep human proteomic datasets has revealed that over 200 distinct proteins are histidine methylated (
      • Kapell S.
      • Jakobsson M.E.
      Large-scale identification of protein histidine methylation in human cells.
      ), highlighting the need to improve our understanding of this unique PTM.
      Histidine protein methyltransferase 1 (Hpm1p) was the first described histidine methyltransferase in any organism, and the only one known in the budding yeast Saccharomyces cerevisiae (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ). It monomethylates the Tau nitrogen of histidine 243 (H243) in the large ribosomal subunit protein Rpl3p (uL3; universal large subunit ribosomal protein 3) to stoichiometric levels. Recent work has confirmed functional conservation of Hpm1p, with the human ortholog METTL18 (human methyltransferase-like protein 18) shown to methylate histidine 245 of human Rpl3 (
      • Małecki J.M.
      • Odonohue M.F.
      • Kim Y.
      • Jakobsson M.E.
      • Gessa L.
      • Pinto R.
      • Wu J.
      • Davydova E.
      • Moen A.
      • Olsen J.v.
      • Thiede B.
      • Gleizes P.E.
      • Leidel S.A.
      • Falnes P.
      Human METTL18 is a histidine-specific methyltransferase that targets RPL3 and affects ribosome biogenesis and function.
      ). In addition, orthologs exist in other eukaryotes including Drosophila melanogaster, Caenorhabditis elegans, and Saccharomyces pombe (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ), suggesting that histidine methylation plays important roles across Eukarya. H243 is located in the buried tryptophan finger domain of yeast Rpl3p, which protrudes into the peptidyl transferase center of the assembled ribosome and makes extensive hydrogen-bonding contacts with the 25S rRNA. This finger is thought to function as a rocker, pushing aminoacyl-tRNAs between the A-site (aminoacyl site of protein synthesis in the translating ribosome) and the P-site (peptidyl site of protein synthesis in the translating ribosome) (
      • Meskauskas A.
      • Dinman J.D.
      Ribosomal protein L3 functions as a “rocker switch” to aid in coordinating of large subunit-associated functions in eukaryotes and Archaea.
      ,
      • Meskauskas A.
      • Dinman J.D.
      Ribosomal protein L3: Gatekeeper to the A site.
      ). Deletion of the hpm1 (YIL110W) gene is nonlethal, but targeted studies of this strain have identified effects on ribosome assembly, fidelity, and structure. This is manifest in decreased 60S subunit synthesis, increased stop-codon readthrough, programmed −1 ribosomal frameshifting, amino acid misincorporation, and increased resistance to ribosome-binding drugs, such as cycloheximide and anisomycin (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ).
      The hpm1 deletion strain is highly pleiotropic, with 48 distinct phenotypes known compared with a median of 14 for all other genes in yeast (
      • Cherry J.M.
      • Hong E.L.
      • Amundsen C.
      • Balakrishnan R.
      • Binkley G.
      • Chan E.T.
      • Christie K.R.
      • Costanzo M.C.
      • Dwight S.S.
      • Engel S.R.
      • Fisk D.G.
      • Hirschman J.E.
      • Hitz B.C.
      • Karra K.
      • Krieger C.J.
      • et al.
      Saccharomyces genome database: The genomics resource of budding yeast.
      ). Systematic screens of gene function have described phenotypes, which are related to ribosomal dysfunction, such as a decreased growth rate in yeast extract–peptone–dextrose or in synthetic media containing glucose (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ,
      • Steinmetz L.M.
      • Scharfe C.
      • Deutschbauer A.M.
      • Mokranjac D.
      • Herman Z.S.
      • Jones T.
      • Chu A.M.
      • Giaever G.
      • Prokisch H.
      • Oefner P.J.
      • Davis R.W.
      Systematic screen for human disease genes in yeast.
      ,
      • Breslow D.K.
      • Cameron D.M.
      • Collins S.R.
      • Schuldiner M.
      • Stewart-Ornstein J.
      • Newman H.W.
      • Braun S.
      • Madhani H.D.
      • Krogan N.J.
      • Weissman J.S.
      A comprehensive strategy enabling high-resolution functional analysis of the yeast genome.
      ,
      • Brown J.A.
      • Sherlock G.
      • Myers C.L.
      • Burrows N.M.
      • Deng C.
      • Wu H.I.
      • McCann K.E.
      • Troyanskaya O.G.
      • Brown J.M.
      Global analysis of gene function in yeast by quantitative phenotypic profiling.
      ,
      • Deutschbauer A.M.
      • Jaramillo D.F.
      • Proctor M.
      • Kumm J.
      • Hillenmeyer M.E.
      • Davis R.W.
      • Nislow C.
      • Giaever G.
      Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast.
      ) as well as a reduced cell volume (
      • Jorgensen P.
      • Nishikawa J.L.
      • Breitkreutz B.J.
      • Tyers M.
      Systematic identification of pathways that couple cell growth and division in yeast.
      ). However, there are many unrelated phenotypes reported for Δhpm1, for example, an improved growth rate in alternative carbon sources (ethanol (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ), xylose (
      • Bengtsson O.
      • Jeppsson M.
      • Sonderegger M.
      • Parachin N.S.
      • Sauer U.
      • Hahn-Hägerdal B.
      • Gorwa-Grauslund M.F.
      Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
      ), glycerol (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ), sucrose/fructose (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      )), decreased desiccation resistance (
      • Ratnakumar S.
      • Hesketh A.
      • Gkargkas K.
      • Wilson M.
      • Rash B.M.
      • Hayes A.
      • Tunnacliffe A.
      • Oliver S.G.
      Phenomic and transcriptomic analyses reveal that autophagy plays a major role in desiccation tolerance in Saccharomyces cerevisiae.
      ), altered vacuolar morphology (
      • Michaillat L.
      • Mayer A.
      Identification of genes affecting vacuole membrane fragmentation in Saccharomyces cerevisiae.
      ), and altered cell wall processes (increased caspofungin resistance (
      • Lesage G.
      • Sdicu A.M.
      • Ménard P.
      • Shapiro J.
      • Hussein S.
      • Bussey H.
      Analysis of β-1,3-glucan assembly in Saccharomyces cerevisiae using a synthetic interaction network and altered sensitivity to caspofungin.
      ), synthetic lethality with CHS1 (
      • Lesage G.
      • Shapiro J.
      • Specht C.A.
      • Sdicu A.M.
      • Ménard P.
      • Hussein S.
      • Tong A.H.Y.
      • Boone C.
      • Bussey H.
      An interactional network of genes involved in chitin synthesis in Saccharomyces cerevisiae.
      ,
      • Tong A.H.Y.
      • Lesage G.
      • Bader G.D.
      • Ding H.
      • Xu H.
      • Xin X.
      • Young J.
      • Berriz G.F.
      • Brost R.L.
      • Chang M.
      • Chen Y.Q.
      • Cheng X.
      • Chua G.
      • Friesen H.
      • Goldberg D.S.
      • et al.
      Global mapping of the yeast genetic interaction network.
      )). It is clear therefore that the function of Hpm1p, and the associated significance of Rpl3p H243 methylation, is not completely understood.
      In this study, we investigate the function/s of yeast Hpm1p with a focus on its extraribosomal phenome. First, we screened Δhpm1 cells for growth defects on an expanded set of stressors, which targeted pathways unrelated to the ribosome. This identified undescribed sensitivities to amphotericin-B, to cold, to salt, and oxidative stress. To investigate the basis of these phenotypes, we next compared the proteome of WT and Δhpm1 yeast through quantitative stable isotope labeling of amino acids in cell culture (SILAC)–based proteomics. This revealed a downregulation of trehalose biosynthetic enzymes and proteins involved in glucose metabolism. Finally, we adapted the emerging technique of quantitative large-scale crosslinking mass spectrometry (MS) (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ) to enable in vivo monitoring of the yeast interactome and structural proteome at low resolution. This revealed that hpm1 knockout, and hence loss of Rpl3 H243 methylation, locally and allosterically alters the structure of the ribosome. Deletion of hpm1 also affected the structure of some membrane proteins, proteins involved in chromatin compaction, and affected some protein–protein interactions (PPIs) involved in mitochondrial membrane remodeling. Importantly, these were independent of protein abundance. Orthogonal techniques confirmed alterations to processes seen in crosslinking analysis, including measurable differences to the extracellular acidification rate and mitochondrial activity. Together, these findings help explain the pleiotropy of Δhpm1, expand roles for Hpm1p in the regulation of carbon metabolic processes, and provide clues into the functional significance of histidine methylation in human and other eukaryotes.

      Experimental Procedures

      Yeast Strains

      WT S. cerevisiae were strain BY4742 (MATα [yeast maturation factor alpha] his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0). The BY4742 knockout strain of hpm1 (ordered locus name YIL110W) was derived from the EuroScarf Yeast Knockout Collection (
      • Winzeler E.A.
      • Shoemaker D.D.
      • Astromoff A.
      • Liang H.
      • Anderson K.
      • Andre B.
      • Bangham R.
      • Benito R.
      • Boeke J.D.
      • Bussey H.
      • Chu A.M.
      • Connelly C.
      • Davis K.
      • Dietrich F.
      • Dow S.W.
      • et al.
      Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis.
      ), with the following genotype: MATα; his3Δ1; leu2Δ0; lys2Δ0; ura3Δ0; YIL110WΔ::kanMX4. All gene and protein names throughout figures and text utilize the standard yeast protein nomenclature system, which refer to standard gene names for the encoded protein. For investigation of growth plate assay phenotype specificity, two strains were generated using homologous recombination–based genome editing in yeast (
      • Janke C.
      • Magiera M.M.
      • Rathfelder N.
      • Taxis C.
      • Reber S.
      • Maekawa H.
      • Moreno-Borchart A.
      • Doenges G.
      • Schwob E.
      • Schiebel E.
      • Knop M.
      A versatile toolbox for PCR-based tagging of yeast genes: New fluorescent proteins, more markers and promoter substitution cassettes.
      ): a pseudotype “WT∗” (BY4742 containing the URA3 gene integrated downstream of the Rpl3 locus) and a Rpl3 H243A mutant (also containing the URA3 gene insertion). A custom synthetic gBlock DNA fragment (Integrated DNA Technologies) containing (1) the 3′ coding region of Rpl3, (2) immediately adjacent 3′ intergenic region, and (3) the URA3 selection cassette, was amplified by PCR, using two primer pairs. These were the reverse primer 5′-TATTTAAATAAAATAGGAATAAGAATGCTCAATTAAAAAATGATTTATTTTCTAACAAAACTTCTTAGTTTTGCTGGCCGCA-3′ coupled with either 5′-GCTTCCGTCGTCAGAGTTTT-3′ for Rpl3-H243A (to amplify the entire gBlock, including H243A mutation) or 5′-AGGTTGCTTGTATTGGTGCT-3′ for WT∗ (to generate a smaller amplicon, which excluded the H243A mutation). The BY4742 WT yeast strain was then transformed with each amplicon, essentially as described (
      • Gietz R.D.
      • Schiestl R.H.
      High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method.
      ), and gene insertions and edits confirmed by PCR and Sanger sequencing using EuroScarf primers specific to the Rpl3 locus (
      • Winzeler E.A.
      • Shoemaker D.D.
      • Astromoff A.
      • Liang H.
      • Anderson K.
      • Andre B.
      • Bangham R.
      • Benito R.
      • Boeke J.D.
      • Bussey H.
      • Chu A.M.
      • Connelly C.
      • Davis K.
      • Dietrich F.
      • Dow S.W.
      • et al.
      Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis.
      ).

      Methyl-Peptide Analysis

      Overnight cultures of saturated stationary-phase (absorbance at 600 nm of ∼10) WT and Δhpm1 cells were pelleted, washed twice with MilliQ water, and then resuspended to 50 absorbance at 600 nm units/ml in Hepes lysis buffer (50 mm Hepes, 100 mm NaCl, 0.5% (v/v) Triton X-100, 2 mm DTT, 2 mm EDTA, pH 7.5) supplemented with Roche cOmplete Mini EDTA-free Protease Inhibitor Cocktail. Cells were lysed with three rounds of 30 s glass bead homogenization. The lysates were clarified by centrifugation at 20,000g for 30 min, and the supernatant resolved on a 4 to 12% Bis–Tris NuPAGE SDS-PAGE gel. In-gel digestion of a band containing the 44 kDa Rpl3p was performed with Promega Sequencing Grade Trypsin or Promega rLys-C, as described (
      • Hamey J.J.
      • Winter D.L.
      • Yagoub D.
      • Overall C.M.
      • Hart-Smith G.
      • Wilkins M.R.
      Novel N-terminal and lysine methyltransferases that target translation elongation factor 1A in yeast and human.
      ) but with the addition of reduction and alkylation steps (10 mM DTT at 37 °C for 1 h, followed by 55 mM iodoacetamide [IAA] at room temperature [RT] for 1 h). For LC–MS/MS analysis, online separation of peptides was performed using buffer A (0.1% (v/v) formic acid) and buffer B (0.1% (v/v) formic acid and 80% (v/v) acetonitrile), with a linear gradient of 2 to 45% buffer B over 30 min. For data-dependent analysis of the tryptic peptides, the Orbitrap Fusion Lumos Tribrid Mass Spectrometer (Thermo Fisher Scientific) was operated as described (
      • Separovich R.J.
      • Wong M.W.M.
      • Chapman T.R.
      • Slavich E.
      • Hamey J.J.
      • Wilkins M.R.
      Post-translational modification analysis of Saccharomyces cerevisiae histone methylation enzymes reveals phosphorylation sites of regulatory potential.
      ), with higher-energy collision dissociation;(HCD) fragmentation performed at normalized collision energy (NCE) of 30. Mascot was used to identify peptides, searching against the SwissProt reference yeast proteome (6049 sequences, accessed on December 3, 2020) for up to two missed cleavages (specificity to KR), carbamidomethylation of cysteine as a fixed modification, methylation of histidine and oxidation of methionine as variable modifications, error tolerance set to 20 ppm at the peptide level and 10 ppm at the fragment ion level, peptide charges of 2+/3+/4+ and ESI_HCD as the instrument. Only significant peptide identifications were considered (false discovery rate [FDR] <0.01), filtering by ion score (≥30) and expectation value (p < 0.05).

      Parallel Reaction Monitoring of Histidine-Methylated Peptides

      To assess the H243 modification status of Rpl3p in the WT and Δhpm1 strains, we performed a parallel reaction monitoring analysis of Lys-C peptides. The mass spectrometer was set to fragment charge states 1 to 4 of the unmodified (m/z = 867.5271, 434.2672, 289.8472, and 217.6372) and histidine-methylated (m/z = 881.5428, 441.2750, 294.5191, and 221.1412) peptide with the sequence THRGLRK. Skyline (
      • Pino L.K.
      • Searle B.C.
      • Bollinger J.G.
      • Nunn B.
      • MacLean B.
      • MacCoss M.J.
      The Skyline ecosystem: Informatics for quantitative mass spectrometry proteomics.
      ) (version 3.6.010493) was used to confirm the coelution of precursor and fragment ions. This analysis complies with tier 3 Molecular and Cellular Proteomics guidelines for targeted MS analysis.

      Reanalysis of Heavy Methyl SILAC Datasets for Methylated Histidine

      The entire heavy methyl SILAC labeling dataset (
      • Hart-Smith G.
      • Yagoub D.
      • Tay A.P.
      • Pickford R.
      • Wilkins M.R.
      Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
      ) was reanalyzed as described in the original study but extended to include histidine methylation as an additional variable modification. Because of the high FDR associated with methyl peptides, putative histidine-methylated peptides were examined for the presence and chromatographic coelution of light/heavy methyl pairs introduced by the 13CD3-methionine metabolic label using MethylQuant (
      • Tay A.P.
      • Geoghegan V.
      • Yagoub D.
      • Wilkins M.R.
      • Hart-Smith G.
      MethylQuant: a tool for sensitive validation of enzyme-mediated protein methylation sites from heavy-methyl SILAC data.
      ) and/or for the presence of the histidine methylation immonium ion (124.088 m/z) in the fragmentation spectra.

      Growth Plate Assays

      Synthetic complete (SC) agar medium [6.7 g/l yeast nitrogen base with ammonium sulphate, 1.92 g/l synthetic drop-out mix without uracil, 2% (w/v) glucose, 80 mg/l uracil, 3% (w/v) agar] was used for growth assays. Agar plates were prepared with or without the following chemical stressors: 2 mM H2O2, 0.1 μg/ml amphotericin B, 1 M NaCl, 1.5 M sorbitol, and 2 mM NaNO2. Low nitrogen SC plates were prepared with 1.7 g/l yeast nitrogen base without ammonium sulphate, 1.92 g/l synthetic drop-out mix without uracil, 50 μM ammonium sulphate, 2% (w/v) glucose, 80 mg/l uracil, and 3% (w/v) agar. WT and Δhpm1 BY4742 cells were cultured overnight to saturation in yeast extract–peptone–dextrose, washed twice with sterile water, and then adjusted to an absorbance of of 0.2 at 600 nm. Five 10-fold serial dilutions were performed for each strain, and 5 μl of each was spotted onto each plate. All plates were incubated at 30 °C, except for cold stressors (16 °C and 25 °C) and heat stressors (37 °C). Plates were imaged ∼72 h later, except for cold stress, salt, and H2O2, which required additional days of incubation to allow comparison of growth rates. Three biological replicates were performed. Growth plate assays containing all four strains (WT, Δhpm1, WT∗, and Rpl3-H243A) were also performed in three biological replicates as described previously, but only four 10-fold serial dilutions were prepared and spotted for each strain.

      2nSILAC Labeling

      The 2nSILAC (isotopic labeling of lysine and arginine in prototrophic yeast) method was used to metabolically label prototrophic BY4742 S. cerevisiae (
      • Dannenmaier S.
      • Stiller S.B.
      • Morgenstern M.
      • Lübbert P.
      • Oeljeklaus S.
      • Wiedemann N.
      • Warscheid B.
      Complete native stable isotope labeling by amino acids of Saccharomyces cerevisiae for global proteomic analysis.
      ). Briefly, cells were cultured for at least 12 doublings in SC labeling media containing 0.17% (w/v) yeast nitrogen base without amino acids and ammonium sulphate, 0.5% (w/v) ammonium sulphate, 20 mg/l each of histidine, leucine, methionine, tryptophan, adenine, and uracil, 200 mg/l proline, and 2% (w/v) glucose. For light-labeled cells, the media contained 287 μM l-arginine monohydrochloride and 342 μM l-lysine monohydrochloride. For heavy-labeled cells, the media contained equivalent molarity of 13C, 15N-labeled l-arginine monohydrochloride (+10 Da), and l-lysine monohydrochloride (+8 Da) (Silantes). To assess label incorporation efficiency, tryptic digests of heavy-labeled samples were subjected to tandem MS analysis as per the procedure described previously. Peptides were identified and quantified using MaxQuant (
      • Tyanova S.
      • Temu T.
      • Cox J.
      The MaxQuant computational platform for mass spectrometry-based shotgun proteomics.
      ) using search settings as described later, and the intensity of the heavy-labeled version of each peptide was compared with the equivalent-unlabeled peptide. Incorporation efficiency for each experiment was seen to be >99% in all 2nSILAC-labeled experiments.

      Differential Proteome Analysis

      To compare the proteome of WT and Δhpm1 BY4742 yeast, cells were cultured using 2nSILAC. Two biological replicates were generated in a label-swap design. Each strain was grown in 35 ml of heavy or light media and harvested at a midlog phase at an absorbance of ∼0.8 at 600 nm. To extract proteins, cells were resuspended at 50 absorbance at 600 nm units/ml of lysis buffer (8 M urea, 0.1 M ammonium bicarbonate, and pH 8), supplemented with Roche cOmplete Mini EDTA-free Protease Inhibitor Cocktail, and subjected to glass bead homogenization for three rounds of 30 s each. The lysates were clarified by centrifugation at 20,000g at 4 °C for 30 min. The Pierce Bicinchoninic Acid Protein Assay Kit was used to determine the protein content in each lysate. Lysates were mixed 1:1 in a label-swap design, to 1 mg total protein content (500 μg each). Reduction was performed using 5 mM DTT for 30 min at RT, followed by alkylation with 10 mM IAA for 30 min at RT in the dark. To facilitate digestion, each sample was diluted 10-fold with 50 mM ammonium bicarbonate and digested overnight at 37 °C with Promega Sequencing Grade Trypsin at a 1:100 protease:protein ratio. Peptides were desalted using C18 Sep-Pak columns, as described (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ), and then subjected to strong cation exchange (SCX) fractionation as described later in SCX to yield five fractions per combined light/heavy sample for analysis. Fractions resulting from SCX chromatography were desalted using Sep-Pak columns, again as described (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ). Desalted fractions were resuspended in 20 μl 0.1% (v/v) formic acid and injected in technical duplicate for LC–MS/MS analysis. Online fractionation was performed as described for methyl-peptide identification, albeit with a 120 min gradient. The Fusion Lumos mass spectrometer was operated with the following settings: MS1 spectra were acquired at 120,000 resolution, scan range 400 to 1500 m/z, RF lens 30%, and 50 ms maximum injection time. The most abundant ions were selected for fragmentation using monoisotopic precursor selection mode, 2.5E4 intensity threshold, z = 2 to 5, 20 s dynamic exclusion, across a 3 s duty cycle. Ions were fragmented using HCD fixed at 30% NCE with a 1.6 m/z isolation window, acquiring spectra with the ion trap mass analyzer in rapid mode.
      MaxQuant (
      • Tyanova S.
      • Temu T.
      • Cox J.
      The MaxQuant computational platform for mass spectrometry-based shotgun proteomics.
      ) (version 1.6.17.0) was used to both identify and quantify peptides, matching against the SwissProt reference yeast proteome (6049 sequences, downloaded December 3, 2020) supplemented with the default contaminants database. Relevant settings included use of the heavy label feature set to Arg10 and Lys8, variable modification of methionine oxidation and acetylation of the protein N terminus, fixed modification of cysteine carbamidomethylation, the use of trypsin/P specificity with a maximum of two missed cleavages, a minimum peptide length of 7, FDR limited to 1% at both the peptide and protein levels, and only unique + razor peptides (at least two) used for quantification. All other settings were default. An additional analysis for lysine acetylation was performed as previously mentioned one but with the addition of lysine acetylation as a variable modification. For differential abundance analysis, the resulting evidence.txt and proteingroups.txt files were analyzed using the aggregation, normalization, and differential expression analysis tools in ProteoSign, v2 (
      • Theodorakis E.
      • Antonakis A.N.
      • Baltsavia I.
      • Pavlopoulos G.A.
      • Samiotaki M.
      • Amoutzias G.D.
      • et al.
      ProteoSign v2: A faster and evolved user-friendly online tool for statistical analyses of differential proteomics.
      ). Default parameters were used, which only consider proteins quantified in both biological replicates and with at least two different peptides. Significance of expression ratios was assessed between biological replicates (at adjusted p ≤ 0.05), after determining ratios from technical replicates and between-fraction replicate identifications.

      SCX of Peptides and Cross-Linked Peptides

      Desalted peptides were fractionated using the Agilent 1260 Infinity II HPLC, coupled to a Phenomenex Luna 5um SCX 100Å 250 × 10 mm LC column, and using buffer A containing 7 mM KH2PO4, pH 2.6, and 30% (v/v) acetonitrile, and buffer B containing 7 mM KH2PO4, pH 2.6, 30% (v/v) acetonitrile, and 350 mM KCl. The gradient was as operated at 1.5 ml/min with the following program: 0% B at 0 min, 5% B at 7.5 min, 60% B at 47.5 min, 100% B at 67.5 min, 100% B at 77.5 min, 0% B at 77.51 min, and 0% B at 97.5 min. For the proteomics experiments, individual 1-min fractions were combined into five final pooled fractions from the following elution times: F1 = 10 to 39 min, F2 = 39 to 48 min, F3 = 48 to 58 min, F4 = 58 to 72 min, and F5 = 72 to 89 min. For the crosslinking experiments that are described in later sections, later eluting peptide fractions enriched for cross-linked peptides were combined into five final fractions as follows: F6–F7 = 48 to 53 min, F8–F9 = 58 to 63 min, F10 = 68 to 73 min, F11 = 73 to 78 min, F12 = 78 to 83 min, and F13 to F15 = 83 to 93 min.

      Functional Enrichment Analysis

      Panther Gene Ontology (GO) Term Over Representation analysis (
      • Mi H.
      • Muruganujan A.
      • Thomas P.D.
      PANTHER in 2013: Modeling the evolution of gene function, and other gene attributes, in the context of phylogenetic trees.
      ) was performed on the results of the differential proteome analysis, using an expanded list of proteins (adjusted p ≤ 0.1) The background was set to all quantified proteins. Biological Process Complete GO terms were analyzed, with the Fisher’s exact test and FDR calculated and controlled to 5%. Only the significant child terms were reported to avoid redundancy.

      Biotin–Aspartate Proline–N-Hydroxyphthalimide Cross-Linker Synthesis

      The peptide-based cross-linker biotin–aspartate proline–N-hydroxyphthalimide (BDP–NHP) was synthesized using Fmoc chemistry as described (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ) but with the use of Rink amide ProTide resin (CEM) instead of Fmoc-Gly-Wang resin. Amino acids were coupled to the resin in the following order: Lys(biotin)-Lys-Pro2-Asp2-succinate2-N-hydroxyphthalimide (NHP)2, using a Liberty Lite peptide synthesizer (CEM). Activated NHP esters were incorporated by incubating the resin containing the BDP molecule with a 12-fold molar excess of N-(trifluoroacetoxy)-phthalimide–NHP in pyridine. The reaction was carried out for 20 min at RT. The resin was then transferred to a PolyPrep column (Bio-Rad) and coupled to a vacuum flask. The resin was washed extensively with dimethylformamide followed by extensive washes with dichloromethane. The crosslinker was cleaved from the resin by incubation with 95% N-(trifluoroacetoxy), 2.5% dichloromethane, and 2.5% H2O for 3 h at RT. The crosslinker was then precipitated in cold diethyl ether and washed extensively with fresh cold diethyl ether. The resulting pellet was dried by vacuum centrifugation. The resulting product was dissolved in dimethyl sulfoxide to a concentration of 200 mM, determined by UV–visible absorbance.

      In Vivo CrossLinking MS

      WT and Δhpm1 cells were cultured using the 2nSILAC approach, with each strain grown in 200 ml of light- or heavy-labeled media. Cells were harvested at a midlog absorbance of ∼0.4 at 600 nm and washed with water. An aliquot of each was taken to confirm label incorporation efficiency, which was found to be >99%. Cells were then mixed 1:1, based on an absorbance at 600 nm, in a label-swap design, and then followed by an additional wash with water. To enzymatically digest the cell wall, cells were first pretreated with 10 mM DTT and 100 mM Tris for 10 min at 30 °C with shaking. After washing with water, cells were resuspended in 2 ml 1.2 M sorbitol supplemented with 300 U/ml zymolyase and incubated for 30 min at 30 °C with light shaking. The progression of cell wall digestion (spheroplasting) was monitored via microscopy and through monitoring of lysis with a 1:20 dilution in water. Spheroplasts were pelleted at 750g, washed with 1.2 M sorbitol twice, and then resuspended in crosslinking buffer (1.2 M sorbitol, 170 mM Na2HPO4, pH 8). Protein interaction reporter (PIR) (amide-BDP–NHP) crosslinker was added to 10 mM, and the reaction allowed to proceed for 30 min at RT with shaking. Cross-linked spheroplasts were washed five times with 1.2 M sorbitol until the supernatant was no longer yellow and then resuspended in lysis buffer (8 M urea, 0.1 M ammonium bicarbonate, and pH 8.0). This was followed by lysis via glass bead homogenization for three rounds of 30 s each. A bicinchoninic acid assay was used to quantify protein content, and then 3 mg of each lysate was reduced with 5 mM DTT for 30 min at RT, followed by alkylation with 10 mM IAA for 30 min at RT in the dark. The sample was diluted 1:10 with 0.1 M ammonium bicarbonate and digested with Promega Sequencing Grade Trypsin at a protease:protein ratio of 1:50, overnight at 37 °C with shaking 650 rpm. Peptides were desalted using a Sep-Pak column, and a 100 μg aliquot taken for protein abundance level quantification, using the same MS scheme described earlier for the proteomics analysis. The remaining peptides were fractionated by SCX as described previously, generating five final fractions per light/heavy sample. Cross-linked peptides from each fraction were enriched with Pierce Monomeric Avidin UltraLink beads, as described (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ), via the biotin group in the BDP–NHP crosslinker.
      Each of the five avidin-enriched fractions was injected twice and analyzed on the 120 min liquid chromatography gradient described previously. The Fusion Lumos mass spectrometer was operated with the following settings: MS1 spectra were acquired in positive ion mode, 60,000 resolution with quadrupole isolation, scan range 400 to 2000 m/z, RF lens 30%, 100 ms maximum injection time, automatic gain control target set to “Standard.” Ions were selected for fragmentation using monoisotopic precursor selection mode, 2.5E4 intensity threshold, z = 4 to 8, 20 s dynamic exclusion, across a 2 s duty cycle. Ions were fragmented using stepped HCD (27/30/33 NCE) with a 1.6 m/z isolation window, acquiring spectra with the Orbitrap analyzer at 60,000 resolution.
      The Mango/Comet/XLinkProphet pipeline was used to identify cross-linked peptides. RAW files were first converted to mzXML (with peak picking at both MS levels) using MSConvert (ProteoWizard, version 3.0.1908) (
      • Kessner D.
      • Chambers M.
      • Burke R.
      • Agus D.
      • Mallick P.
      ProteoWizard: Open source software for rapid proteomics tools development.
      ). Spectra that contained putatively cross-linked peptides were extracted using the Mango tool (version 2017.01 rev 0 beta 3) (
      • Mohr J.P.
      • Perumalla P.
      • Chavez J.D.
      • Eng J.K.
      • Bruce J.E.
      Mango: A general tool for collision induced dissociation-cleavable cross-linked peptide identification.
      ), with the reporter neutral mass set to 693.399601 Da and the mass tolerance at the relationship level set to 10 ppm and peptide level set to 20 ppm. Fragment ions in these spectra were matched against the yeast reference proteome (6049 sequences, downloaded December 3, 2020), alongside a reverse decoy database, using the search engine Comet (version 2018.01 rev 2) (
      • Eng J.K.
      • Jahan T.A.
      • Hoopmann M.R.
      Comet: An open-source MS/MS sequence database search tool.
      ), performing two individual searches for each file to identify completely light or heavy isotopically labeled peptides. Key Comet parameters were set as follows: 20 ppm parent and fragment ion tolerance, cysteine carbamidomethylation as a fixed, methionine oxidation as variable, and the cross-linker stump mass of 197.032422 Da on lysines as variable, up to two missed tryptic (KR, negated by P) cleavages allowed. XLinkProphet (
      • Keller A.
      • Chavez J.D.
      • Bruce J.E.
      Increased sensitivity with automated validation of XL-MS cleavable peptide cross-links.
      ) was used with default recommended parameters to recombine spectral matches and perform FDR control of identifications to a 1% error rate at the nonredundant cross-linked peptide level. These identifications were uploaded to the XLinkDB (
      • Schweppe D.K.
      • Zheng C.
      • Chavez J.D.
      • Navare A.T.
      • Wu X.
      • Eng J.K.
      • Bruce J.E.
      XLinkDB 2.0: Integrated, large-scale structural analysis of protein cross-linking data.
      ) under the name “YeastHpm1_Wilkins.” For mapping of crosslinks onto structure, all possible structures involving the cross-linked residues were downloaded from the Protein Data Bank (PDB) (rscb.org (
      • Berman H.M.
      • Westbrook J.
      • Feng Z.
      • Gilliland G.
      • Bhat T.N.
      • Weissig H.
      • Shindyalov I.N.
      • Bourne P.E.
      The protein data bank.
      )) and crosslinks mapped in every chain pair configuration using Xwalk (
      • Kahraman A.
      • Malmström L.
      • Aebersold R.
      Xwalk: Computing and visualizing distances in cross-linking experiments.
      ), to calculate the minimum Cɑ–Cɑ Euclidean distances imposed by the given crosslinks. The SIFTS resource was used to determine PDB resolved residues (
      • Dana J.M.
      • Gutmanas A.
      • Tyagi N.
      • Qi G.
      • O’Donovan C.
      • Martin M.
      • Velankar S.
      SIFTS: Updated Structure integration with function, taxonomy and sequences resource allows 40-fold increase in coverage of structure-based annotations for proteins.
      ). A random lysine–lysine pair list was generated by randomly selecting any lysine residue in each protein member of an experimentally identified crosslink that mapped to that structure. The Euclidean distances of these links were also evaluated.
      Cross-link spectral matches (CSMs) from these identified cross-linked peptides were subjected to analysis with MethylQuant (
      • Tay A.P.
      • Geoghegan V.
      • Yagoub D.
      • Wilkins M.R.
      • Hart-Smith G.
      MethylQuant: a tool for sensitive validation of enzyme-mediated protein methylation sites from heavy-methyl SILAC data.
      ) tool (version 1.0) to automatically extract ion chromatograms. The derived light and heavy m/z values were used as input, using default program parameters, but without the use of the in-built mass shift function. The MethylQuant output was filtered for only crosslinks with successful “H/L Ratio #2” ratios (which are MethylQuant ratios that passed an internal quality control based on the presence of highly correlated isotope coelutions) and a MethylQuant score greater than 10. Crosslinks involving methionine were not analyzed further. The cross-link ratios within each biological replicate were normalized (to correct for unequal mixing in sample preparation) by dividing each ratio, before log transformation, by a protein mixing factor. The mixing ratio factor was determined from a matched proteomic analysis of the samples before fractionation, with each sample acquired in the same manner as the deep proteome MS/MS described previously except in one shot. Note that this proteomic dataset was also used to determine changes in protein abundances, for comparing crosslink changes with changes in protein abundance.
      To summarize changes in crosslinks to the level of unique residue pair (URP), redundant normalized ratios from cross-linked peptides were averaged to form a single ratio per biological replicate (where redundant ratios could represent multiple peptide pairs representing the same URP, quantification in different fractions, or technical replicates). Only URPs with ratios successfully calculated from both biological replicates were considered further, whereby the average of biological replicates was used for statistical testing. To determine statistical significance, two-tailed unpaired t tests were performed by comparing ratios to 0. Multiple testing correction was performed using Boca and Leek’s FDR regression method (BL) as implemented in the swfdr R package (
      • Boca S.M.
      • Leek J.T.
      A direct approach to estimating false discovery rates conditional on covariates.
      ). Briefly, a linear regression with spline smoothing (two degrees of freedom) was modeled on t test p values and three covariates to generate q values. The three covariates were correlated to the quality of the quantifications themselves: (1) the standard deviation of all technical replicate redundant ratios for the URP, (2) the total number of redundant ratios calculated for the URP, and (3) the maximum number of well-correlated isotopic distribution heavy/light pair coelutions (as determined by MethylQuant) for any quantification representing the URP. To compare URP ratios with protein-level change, the URP ratios for significantly changing URPs (q ≤ 0.05, log2 ≥ 0.3 and ≤−0.3) were compared with the mean normalized protein ratio for the given protein/s, determining statistical significance using the two-tailed unpaired t test and then BL multiple testing correction as described previously. Cytoscape (
      • Shannon P.
      • Markiel A.
      • Ozier O.
      • Baliga N.S.
      • Wang J.T.
      • Ramage D.
      • Amin N.
      • Schwikowski B.
      • Ideker T.
      Cytoscape: A software environment for integrated models of biomolecular interaction networks.
      ) was used to visualize crosslinks as either a protein interaction network or a residue-pair network. PyXlinkViewer (
      • Schiffrin B.
      • Radford S.E.
      • Brockwell D.J.
      • Calabrese A.N.
      PyXlinkViewer: A flexible tool for visualization of protein chemical cross-linking data within the PyMOL molecular graphics system.
      ) was used to map crosslinks onto the ribosome structure (PDB: 6GQV) (
      • Pellegrino S.
      • Demeshkina N.
      • Mancera-Martinez E.
      • Melnikov S.
      • Simonetti A.
      • Myasnikov A.
      • Yusupov M.
      • Yusupova G.
      • Hashem Y.
      Structural insights into the role of diphthamide on elongation factor 2 in mRNA reading-frame maintenance.
      ).

      Monitoring of Media Acidification Rate

      To monitor changes to media pH over time, three independent colonies each of WT and Δhpm1 were separately cultured overnight to saturation in SC medium (6.7 g/l yeast nitrogen base without amino acids, 1.92 g/l synthetic drop-out mix without uracil, 2% [w/v] glucose, and 80 mg/l uracil). For each sample, 50 ml cultures of SC media were inoculated at an initial absorbance at 600 nm of ∼0.1 and cultured at 30 °C with 250 rpm shaking for 11 h. An aliquot was taken every hour to determine an absorbance at 600 nm and, following pelleting of cells, the pH of media was measured by a tabletop pH probe. For each strain, a one-phase nonlinear regression curve was fitted to data from the three biological replicates using GraphPad Prism 9.3.1 (GraphPad Software, Inc). To determine whether the two models differed (whether one model could adequately explain all data), the extra-sum-of-squares F test was used, comparing all equation parameters.

      Flow Cytometric Analysis of Mitochondrial Activity

      To determine the degree of mitochondrial superoxide production in the WT and hpm1 knockout strains, yeast cells were stained with MitoSOX Red Mitochondrial Superoxide Indicator dye (Thermo Fisher Scientific) and assessed by flow cytometry. Briefly, three independent colonies of each strain were grown separately to saturation overnight in SC medium, cultured throughout the day to produce a growth curve, and inoculated into 25 ml of fresh SC for overnight culture. Cells were harvested the following morning at an absorbance of ∼0.7 at 600 nm, washed twice with PBS, and then resuspended to a density of 10 absorbance at 600 nm units/ml in PBS. Unstained and stained (5 μM MitoSOX 10 min at RT in the dark) cells were prepared for each strain, with two additional PBS washes performed before analysis on the BD LSRFortessa SORP Cell Analyzer (BD Biosciences). Following gating to select for single cells, the mean intensity of the staining (10,000 cells per sample) was measured using the 561 nm Yellow/Green laser with 610/20 band pass filter. The mean population intensity of unstained samples was subtracted from the mean population intensity of the matched labeled samples, and a two-tailed unpaired t test was used to assess statistical significance.

      Experimental Design and Statistical Rationale

      Four independent cultures (two of each of the control WT and Δhpm1 strains in heavy or light media) were used in SILAC expression or crosslinking experiments, producing two biological replicates in a label-swap design (forward: light = WT, heavy = Δhpm1; reverse: light = Δhpm1, heavy = WT). Five SCX fractions were generated per light/heavy combined biological replicate, and each SCX fraction was analyzed in technical duplicate, being injected twice for LC–MS/MS analysis. Data processing and statistical analyses for each investigation type are described in detail previously.

      Results

      Confirmation of the Rpl3p H243 Methyl Site and Determination of Stoichiometry

      Amino acid analysis and mutational analysis had previously shown that methyltransferase Hpm1p catalyzes the methylation of H243 (H243Me) on the large ribosomal subunit protein L3 (Rpl3p) (Fig. 1A). However, efforts to detect this modification by MS had only localized it to an 18-residue window of G234 to C251 (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ). To confirm Hpm1p-mediated H243Me, in-gel tryptic digestion of Rpl3p from whole-cell lysates was performed, and the resultant peptides were analyzed by LC–MS/MS. This identified the methylated peptide KTHMeRGLR and precisely localized the methylation to residue 243 (supplemental Fig. S1). Notably, the identified spectrum contained a 124.088 m/z immonium ion (supplemental Fig. S1, inset), which has recently been described as diagnostic for histidine methylation (
      • Kapell S.
      • Jakobsson M.E.
      Large-scale identification of protein histidine methylation in human cells.
      ). To generate a peptide with no missed cleavages and thus more suitable for quantification, we digested proteins with Lys-C. Targeted MS/MS analysis of the peptide THMeRGLRK showed chromatographic coelution of the parent and fragment ions and further confirmed that the methylation site is localized to H243 (Fig. 1B). Analysis of the unmethylated and methylated forms of the LysC-generated peptide from WT cells confirmed that this modification reaches stoichiometric (>99.9%) levels on Rpl3p in vivo (Fig. 1C). By contrast, methylation on this peptide in hpm1 knockout cells is completely absent, proving that Hpm1p is the sole methyltransferase responsible for this modification.
      Figure thumbnail gr1
      Fig. 1Identification of Rpl3p H243Me by bottom–up mass spectrometry. A, the histidine methyltransferase Hpm1p monomethylates histidine 243 of the large ribosomal subunit protein Rpl3p. The methylation site of Rpl3p directly contacts the 25S rRNA and is in proximity to the ribosomal peptidyl transferase center. B, PRM analysis of the LysC generated, Rpl3p methylated peptide THMeRGLRK (2+, 443.2672 m/z), showing codetection of the fragments at the same retention time. C, extracted ion chromatograms of the precursors for the THRGLRK (2+, 441.2750 m/z) and THMeRGLRK (2+, 443.2672 m/z) peptides, generated from LysC digest of the Rpl3p band in WT or Δhpm1 lysate. H243Me, methylation of histidine 243; Hpm1p, histidine protein methyltransferase 1; PRM, parallel reaction monitoring.
      There have been conflicting reports (
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ) on whether Rpl3p is the only substrate of Hpm1p. We therefore investigated whether additional histidine methylation sites could be detected, by reanalyzing a published heavy-methyl-SILAC, yeast deep proteome dataset (
      • Hart-Smith G.
      • Yagoub D.
      • Tay A.P.
      • Pickford R.
      • Wilkins M.R.
      Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
      ). Heavy-methyl-SILAC enables high confidence discovery of methylated residues by analyzing a mixture of unlabeled and 13CD3-methionine-labeled peptides (
      • Ong S.E.
      • Mittler G.
      • Mann M.
      Identifying and quantifying in vivo methylation sites by heavy methyl SILAC.
      ). FDRs, a known issue in the discovery of methylation sites (
      • Hart-Smith G.
      • Yagoub D.
      • Tay A.P.
      • Pickford R.
      • Wilkins M.R.
      Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
      ), can be reduced through correlative analysis of isotopic peaks in SILAC pairs and their chromatographic coelution profiles (
      • Tay A.P.
      • Geoghegan V.
      • Yagoub D.
      • Wilkins M.R.
      • Hart-Smith G.
      MethylQuant: a tool for sensitive validation of enzyme-mediated protein methylation sites from heavy-methyl SILAC data.
      ). Our analysis did not detect any histidine-methylated peptides (data not shown). This included the THMeRGLRK Rpl3p peptide; however, this is produced from two missed tryptic cleavages and therefore difficult to detect within complex lysates. Further analysis of all data for the 124.088 m/z diagnostic immonium ion also did not detect any histidine-methylated peptides. Therefore, at least in a comprehensive tryptic peptide set, we were not able to find any evidence that Hpm1p catalyzes methylation of histidine residues anywhere in the yeast proteome, except on Rpl3p as described above and in Ref. (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ).

      Deletion of the hpm1 Gene Results in Sensitivity to Environmental Stressors

      The deletion of hpm1, and hence loss of Rpl3p H243 methylation, is known to result in ribosome-related defects (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ). To further characterize the phenotype of this knockout, we subjected WT and Δhpm1 yeast to a range of environmental stressors (Fig. 2). This revealed growth defects of Δhpm1 under oxidative stress (2 mM H2O2), membrane stress (0.1 μg/ml amphotericin B), salt stress (1 M NaCl), and cold stress (16 °C, RT). All other stressors tested, including heat stress (37 °C), low nitrogen media, osmotic stress (1.5 M sorbitol), and nitrosative stress (2 mM NaNO2), did not increase or decrease the growth rate of the hpm1 deletion mutant. To further investigate whether H243 methylation on Rpl3p specifically mediated these responses, we constructed the Rpl3p-H243A point mutant and compared its growth with a matched WT control that also contained the URA3 selection cassette used in our genome editing (supplemental Fig. S2). Interestingly, this revealed that the H243A mutant phenocopied the hpm1 knockout sensitivity for some stressors (cold and salt stress) but that only the hpm1 knockout was sensitive to other stressors (H2O2 and amphotericin). This observation has parallels with those of the study by Al-Hadid et al. (
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ), who saw perturbed levels of ribosomal subunits in hpm1 knockout cells but not in Rpl3p-H243A. Our results show that loss of Hpm1p is associated with sensitivity to a number of environmental stressors and clear extraribosomal phenotypes.
      Figure thumbnail gr2
      Fig. 2hpm1 knockout sensitizes yeast to cold, to oxidative, and cell membrane stress. WT and hpm1 knockout BY4742 Saccharomyces cerevisiae were cultivated in liquid cultures overnight to saturation, adjusted to 0.2 absorbance at 600 nm, and serially diluted 1:10 five times. Five microliters of each dilution was spotted onto synthetic complete agar plates with cold, heat, oxidative, salt, osmotic, or cell membrane stressors. Hpm1p, histidine protein methyltransferase 1.

      Protein Abundance Profiling Identifies a Role for Hpm1p in Carbon Metabolism

      To investigate the reasons for the aforementioned Δhpm1 growth phenotypes, and others previously reported, we compared the proteome of WT and Δhpm1 yeast (Fig. 3A). To do so, we cultured each strain in media containing either unlabeled lysine and arginine (light) or heavy-isotope labeled lysine and arginine (heavy, +8 and +10 Da, respectively). We confirmed efficient (>99%) incorporation of heavy lysine and heavy arginine in isotopically labeled cultures (supplemental Fig. S3) and controlled for mixing by normalizing ratios. In a double label-swap SILAC design, MaxQuant (
      • Tyanova S.
      • Temu T.
      • Cox J.
      The MaxQuant computational platform for mass spectrometry-based shotgun proteomics.
      ) identified 55,113 unique peptides (length ≥7 amino acids) at 1% FDR from five SCX fractions per biological replicate, analyzed in technical duplicate. This corresponds to 3801 unique protein groups with at least two unique peptides (supplemental Table S1). The knockout status of the Δhpm1 BY4742 was confirmed, with ion intensity from peptides in Hpm1p (LVTIPNLVITWAK, LTFEEFQSPIYR, EQWYALQKDECEDIPINNEELLLTSK) only identifiable and quantifiable in the WT strain (data not shown). Proteins with at least two successfully quantified peptides were subjected to statistical analysis, which revealed 11 proteins changing with significance (adjusted p value ≤0.05), all of which had log2 fold change >±1 (supplemental Table S2). One protein was upregulated, Tdp1p. The downregulated proteins included glycolysis enzymes Emi2p, Hxk1p, and Glk10p, trehalose synthesizing enzymes Tps2p and Tsl1p, glucose-regulated protein Ygp1p, and proteins in other systems, such as Rep2p, Pho84p, Sse2p, and Rtc3p (Fig. 3B). Functional enrichment analysis on a more extensive list of 24 proteins with adjusted p values ≤0.1 identified several processes broadly related to carbon metabolism with statistical significance (Fig. 3C). These were cellular response to desiccation (GO: 0071465), trehalose biosynthetic process (GO: 0005992), cellular glucose homeostasis (GO: 0001678), carbohydrate phosphorylation (GO: 0046835), glucose 6-phosphate metabolic process (GO: 0051156), and carbohydrate catabolic process (GO: 0016052). Furthermore, mapping of proteome changes to all members of the trehalose synthetic pathway (Tsl1p, Tps1p, Tps2p, and Tps3p) revealed general downregulation of this metabolic pathway in Δhpm1 (Fig. 3D). Our results relate to previous observations from large-scale screens that Δhpm1 yeast grow faster than WT cells on media containing nonfermentable carbon sources, such as xylose (
      • Bengtsson O.
      • Jeppsson M.
      • Sonderegger M.
      • Parachin N.S.
      • Sauer U.
      • Hahn-Hägerdal B.
      • Gorwa-Grauslund M.F.
      Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
      ), glycerol, and ethanol (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ). These carbon sources allow cells to either entirely avoid glycolysis or be more reliant on metabolic pathways such as oxidative phosphorylation. All four known members of the trehalose synthetic pathway (Tsl1p, Tps1p, Tps2p, and Tps3p) were downregulated in Δhpm1 (Fig. 3D).
      Figure thumbnail gr3
      Fig. 3Proteomic analysis identifies changes to sugar metabolism upon hpm1 knockout. A, label-swap replicates prepared by 2nSILAC (arginine +10, lysine +8) were used to compare the steady-state proteome of log-phase WT and Δhpm1. The tryptic digests were fractionated by SCX and then subjected to high-resolution HCD MS/MS analysis on a Fusion Lumos. B, results of the differential proteome abundance analysis. Labeled proteins have an adjusted p value ≤0.05 and log2 fold change ≥1 or ≤−1. C, statistically significant child terms resulting from a Panther Biological Process GO term overrepresentation analysis using 24 proteins as input (adjusted p≤ 0.1), with all quantified proteins used as background. Changing GO terms are annotated, and the asterisk ∗ represents enrichment fold changes >100. D, proteins involved in the trehalose biosynthetic pathway, with protein log2 changes overlaid by color. 2nSILAC, isotopic labeling of lysine and arginine in prototrophic yeast; GO, Gene Ontology; HCD, higher-energy collision dissociation; SCX, strong cation exchange.

      Differential In Vivo Crosslinking of S. cerevisiae Proteins

      hpm1 deletion cells are hypersensitive to membrane, oxidative, and salt stress (Fig. 2). These phenotypes were not readily explained by changes in protein abundance but may arise because of changes in PPIs or gross changes in protein structures. To investigate this, we adapted the PIR in vivo crosslinking approach (
      • Chavez J.D.
      • Mohr J.P.
      • Mathay M.
      • Zhong X.
      • Keller A.
      • Bruce J.E.
      Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
      ) for use in S. cerevisiae, which employs the lysine-reactive and MS-cleavable crosslinker BDP–NHP (Fig. 4A). This was achieved through (1) use of 2n native SILAC metabolic labeling and (2) addition of an enzymatic digestion step of whole yeast with zymolyase to permeabilize the cell wall to the PIR crosslinker. Our experiment represents the first in vivo large-scale crosslinking MS analysis of budding yeast, and one which uses an enrichable and MS-cleavable crosslinker.
      Figure thumbnail gr4
      Fig. 4Large-scale crosslinking mass spectrometry (MS) for the monitoring of in vivo protein structure and interactions in Saccharomyces cerevisiae. A, experimental overview of quantitative crosslinking MS to compare WT and Δhpm1 cells, adapted for use with S. cerevisiae by the enzymatic digestion of the yeast cell wall to increase permeability to the MS-cleavable crosslinker BDP–NHP (PIR). A sample was taken before SCX fractionation, for single-shot proteomic analysis and to normalize crosslinking data for changes in protein abundance. B, overlap of unique residue pair (URP) identifications between biological replicates. C, density histogram of minimum distances for 543 interprotein and intraprotein URPs, and random lysine-lysine pairs, when mapped to 794 possible cognate PDB structures. D, protein interaction network (including unambiguous homodimers) represented by the identified URPs. Nodes are unique proteins, and edges are URPs. BDP–NHP, biotin–aspartate proline–N-hydroxyphthalimide; PDB, Protein Data Bank; PIR, protein interaction reporter; SCX, strong cation exchange.
      To compare WT and Δhpm1 cells using this approach, cells were metabolically labeled and then mixed in a label-swap design before downstream spheroplasting, crosslinking, and protein extraction. Tryptic digests of lysates were fractionated using SCX and crosslinker-reacted peptides then enriched with monomeric avidin beads, generating five fractions per biological replicate. These were analyzed with LC–MS/MS by stepped HCD in technical duplicate. The Mango/Comet/XLinkProphet pipeline confidently identified 1889 nonredundant cross-linked peptide pairs at a 1% FDR (controlled at the nonredundant cross-linked peptide level) (supplemental Table S3). This represented 1267 URPs, of which 261 (21%) were interprotein links and 1006 (79%) were intraprotein links. Redundancy observed in raw CSMs was due to SILAC labeling (doubled by the presence of heavy-labeled CSMs) and a high degree of URP overlap (46%) between biological replicates (Fig. 4B).
      We first explored the identified URPs in terms of structural information and PPIs. The intralink URPs represented 1006 in vivo structural constraints for 251 unique proteins. Mapping of all (interprotein and intraprotein) URPs onto an ensemble of PDB structures showed that 93% of the 543 mappable URPs fell within the 42 Å theoretical distance constraint of the PIR crosslinker (Fig. 4C). Notably, the distribution of Euclidean distances imposed by random lysine–lysine links in these PDB structures was much broader, with only 49% falling within the PIR distance cutoff. Unmappable intraprotein URPs represent valuable information for proteins, which have no experimental structure or where part of the available structure is unresolved. An additional 261 URPs represented 146 unique in vivo PPIs, including homodimers (Fig. 4D). Twenty-six unique PPIs were of very high confidence (with more than one CSM) and also novel (without curation in the APID protein interaction database (
      • Alonso-López Di.
      • Campos-Laborie F.J.
      • Gutiérrez M.A.
      • Lambourne L.
      • Calderwood M.A.
      • Vidal M.
      • de Las Rivas J.
      APID Database: Redefining Protein-Protein Interaction Experimental Evidences and Binary Interactomes.
      )). An example is the interaction between the uncharacterized protein YKR015Cp and histones H3 and H2B.
      To investigate the effect of hpm1 knockout on the structural proteome and interactome, as measured by crosslinking MS, light- and heavy-labeled cross-linked peptides were quantified by generating extracted ion chromatograms in MethylQuant (
      • Tay A.P.
      • Geoghegan V.
      • Yagoub D.
      • Wilkins M.R.
      • Hart-Smith G.
      MethylQuant: a tool for sensitive validation of enzyme-mediated protein methylation sites from heavy-methyl SILAC data.
      ). This was possible for 877 URPs, with 390 URPs quantified in both replicates (Figs. 5A, S4 and supplemental Table S4). Of these 390, 60 URP ratios changed significantly (q value ≤0.05, log2 ratios ≥0.3 and ≤−0.3), indicating the modulation of a protein’s structural conformation or PPI. We investigated whether there were URPs that were unique in the WT or hpm1 knockout strains. To do this, we asked if there were exclusively quantifiable URPs in one strain, and furthermore, if they were reproducibly unique in label-swap replicates. We were not able to detect any such crosslinks in our analysis. We also asked whether there was evidence of crosslinking between heavy- and light-labeled peptides, which may arise from the mixing of proteins from cells before crosslinking occurred. There were only 51 of 21,978 (0.23%) spectra with PIR ions where the individual peptides were of different label (at a PSM FDR of 1%).
      Figure thumbnail gr5
      Fig. 5Quantitative large-scale crosslinking reveals modulation of the structural proteome and interactome in hpm1 knockout yeast. A, volcano plot showing Δhpm1/WT ratios for 390 URPs, successfully quantified in both biological replicates. Crosslinks were averaged across biological replicates to form the URP log2 ratio, after normalization for mixing as determined by the matched proteomics experiment. Statistical significance was determined using a two-tailed unpaired t test, with BL multiple testing correction. Colored URPs have q values ≤0.05 and log2 ratios ≥0.3 or ≤−0.3. About 37 URPs with extreme q values been excluded from this plot to aid visualization of most URPs. B, lysine–lysine network view of URPs with statistically significant change, after adjustment for changes in cognate protein(s). Nodes contain lysine residue numbers, with the associated protein name annotated in proximity. Ratios of change are overlaid in color for both protein levels (on nodes) and URPs (on edges). BL, Boca and Leek’s FDR regression method; URP, unique residue pair.
      To confirm that changes in cross-link abundance were not simply a result of differential protein abundance, we also analyzed the proteome of the cross-linked spheroplasts (Fig. 4A and supplemental Table S5). Note that this is a separate analysis to that which was presented previously (Fig. 3), which was performed on the non–cross-linked and nonspheroplasted proteome. This revealed that most proteins were not changing in abundance between WT and Δhpm1 yeast (supplemental Fig. S5). A direct comparison of the 60 significant URP ratios to the mean protein level ratio of the cognate proteins, obtained from the parallel proteomic analysis, revealed that 44 URPs changed (q value ≤0.05) beyond the level of protein abundance changes (Fig. 5B and supplemental Tables S6 and S7).

      Quantitative Changes in Crosslinks are Associated With Phenotypes of Δhpm1

      We examined the cases where URPs changed in abundance beyond detected changes in abundance of cognate proteins. The ratios of changed interprotein and intraprotein URPs and of their proteins are given in supplemental Tables S6 and S7, respectively. Interestingly, some of the observed changes related to key phenotypes of the Δhpm1 knockout. These included a phenotype of decreased expression of many proteins in the proteome (Fig. 3B), and changes to proteins in membranes, the mitochondria, and some metabolic processes. For example, our crosslinking indicated that Δhpm1 cells may modulate some transcriptional processes (and hence the proteome) through changes to chromatin. URP interlinks involving the histone subunits increased in abundance in the knockout strain (Fig. 5B), suggesting chromatin compaction. Consistent with these observations, reinvestigation of our proteomic data revealed that acetylation of histone Hta1/2p lysine 8 (H2AK7ac, from the tryptic peptide SGGKGGKAcAGSAAK), a transcriptionally activating epigenetic modification (
      • Dai H.
      • Wang Z.
      Histone modification patterns and their responses to environment.
      ,
      • Rando O.J.
      • Winston F.
      Chromatin and transcription in yeast.
      ), was downregulated in the knockout strain, with a Δhpm1/WT log2 ratio of −1.03 (data not shown). As noted, this reflects what we observed in our non–cross-linked proteome analyses where most differentially abundant proteins, and indeed the majority of other proteins we quantified, were downregulated on hpm1 knockout (Fig. 3B). By contrast, no clear changes were detected in the abundance of other acetylated histone peptides including in Htb1p-K12 (H2BK11ac, from KPASKAcAPAEK) and Htb2p-K7 and Htb2p-K8 (H2BK6ac from SSAAEKAcKPASK and H2BK7ac from SSAAEKKAcPASK, respectively).
      In a different phenotype, wherein Δhpm1 yeasts were sensitive to the membrane stressor amphotericin B (Fig. 2), we noted a relationship with changed URPs. URP intralinks identified within the plasma membrane–embedded ATPase Pma1p, and in the endoplasmic reticulum chaperone Kar2p, were greatly increased in the Δhpm1 strain. In addition, there were changes to intralink URPs within the Yet3p and Sey1p proteins, which also reside in the endoplasmic reticulum. In a further example, there was an intralink URP change in the plasma membrane sodium transporter Ena2p, which may reflect the increased sensitivity of this strain to salt stress (Fig. 2). Interestingly, we were able to detect structural changes in enzymes involved in glucose metabolism, which may reflect the changes to metabolic processes detected in Figure 3. This included the cytochrome b reductase Cbr1p, the pyruvate kinase Cdc19p, and phosphopyruvate hydratase Eno2p, which are involved in gluconeogenesis and glycolysis.
      Mitochondrial dysfunction is known to increase the sensitivity of yeast to reactive oxygen species, such as that in the 2 mM H2O2 treatment, which reduced Δhpm1 growth in our plate assays (Fig. 2). The mitochondrion is also essential for respiration, necessary for growth on alternative carbon sources. Interestingly, some URPs involved in mitochondrial membrane protein sorting (Phb1p and Phb2p) showed significant change (Fig. 5B). For the proteins prohibitin 1 and 2, the cross-linked peptide FLVEKAEQER-AAFVVDKAR represented in the Phb1p-203:Phb2p-232 URP, increased in Δhpm1 yeast (Δhpm1/WT log2 ratio of 1.1). However, the Phb1p and Phb2p proteins themselves did not change in abundance, with Δhpm1/WT log2 ratios of −0.08 and −0.01, respectively. Hence, the PPI between prohibitin 1 and 2, known to be involved in many mitochondrial membrane functions including protein chaperoning (
      • Signorile A.
      • Sgaramella G.
      • Bellomo F.
      • de Rasmo D.
      Prohibitins: A critical role in mitochondrial functions and implication in diseases.
      ), increased in Δhpm1 yeast. Other mitochondrial protein URPs were also increased, including in the 2-oxodicarboxylate transporter Odc2p, protein import chaperone Ssc1p, and mitochondrial resident protein Cbr1p (Fig. 5B).

      Functional Validation of Processes in hpm1 Knockout That Showed Altered Intraprotein or Interprotein Crosslinks

      Having found associations, aforementioned, of changed protein and cross-link abundance with some hpm1 knockout phenotypes, we asked whether we could predict and then functionally validate further hpm1 knockout phenotypes solely from changes in intraprotein or interprotein crosslinks. Here, we provide three examples, first through considering crosslinks for a single protein, then by considering crosslinks in a large protein complex, and then through considering changes in a eukaryotic organelle.
      Pma1p is a plasma membrane proton pump whose role is to directly regulate and acidify the extracellular pH, often in response to environmental and nutrient cues. This is done as a function of its structure (
      • Portillo F.
      Regulation of plasma membrane H+-ATPase in fungi and plants.
      ). Interestingly, many crosslinks within Pma1p were found to be significantly increased in the hpm1 deletion strain (Fig. 5, A and B and supplemental Table S6), raising the prospect that it has an altered structure and function. To test whether the rate of media acidification differed between WT and Δhpm1 strains, we measured media pH as a function of absorbance during exponential growth (Fig. 6A). Δhpm1 cells exhibited a significantly slower acidification rate compared with the WT strain during its growth (p < 0.0001, comparing one-phase decay nonlinear regressions using an extra-sum-of-squares F test). Our crosslinking experiments were performed in cells harvested at midlog growth, a time where changes in media pH of strains were becoming evident. The changes in Pma1p cross-link abundances on hpm1 deletion (Fig. 5, A and B and supplemental Table S6), likely associated with a change in protein structure, could here predict a phenotype—being a difference in the rate of media acidification.
      Figure thumbnail gr6
      Fig. 6Changes in intraprotein and interprotein crosslinks can explain or predict cellular phenotypes. A, the media pH and absorbance at 600 nm of WT and Δhpm1 strains grown in SC media were monitored over time to assess the rate of media acidification. This phenotype is affected by Pma1p. Statistical significance reflects the results of comparing nonlinear regression models using an extra-sum-of-squares F test, where ∗∗∗∗ = p < 0.0001, and shading indicates the 95% confidence interval of each model. B, mapping of URPs onto a cryo-EM structure of the yeast ribosome in complex with mRNA, tRNA, and eukaryotic elongation factor 2 (eEF2) (PDB: 6GQV). Only crosslinks that were quantified in both replicates are shown, with blue crosslinks determined to be significantly downregulated upon hpm1 knockout and independent of changes to the underlying protein levels. C, to assay for changes to mitochondrial activity and dysfunction, the mean intensity of MitoSOX staining per cell was determined by flow cytometry analysis of 10,000 events in WT and Δhpm1 cells. The mean population intensity for each replicate was normalized against the matched unlabeled control sample. Statistical analysis was performed using an unpaired two-tailed t test where ∗∗ = p < 0.01. PDB, Protein Data Bank; SC, synthetic complete; URP, unique residue pair.
      Hpm1p is a methyltransferase that methylates H243 of ribosomal protein Rpl3p. Many of the significantly changing URPs did involve ribosomal proteins, with the knockout of hpm1 resulting in a downregulation of these crosslinks (Fig. 5, A and B and supplemental Table S6). When ribosomal URPs successfully quantified in both replicates were mapped to the cryo-EM structure of the yeast ribosome complexed with mRNA, tRNA, and eukaryotic elongation factor 2 (PDB: 6GQV) (
      • Pellegrino S.
      • Demeshkina N.
      • Mancera-Martinez E.
      • Melnikov S.
      • Simonetti A.
      • Myasnikov A.
      • Yusupov M.
      • Yusupova G.
      • Hashem Y.
      Structural insights into the role of diphthamide on elongation factor 2 in mRNA reading-frame maintenance.
      ) (Fig. 6B), we found some URPs proximal to Rpl3p with significant change. There was also evidence of some allosteric changes further from Rpl3p, in the 40S subunit. The evaluation of these findings, alone, would suggest changes to ribosome function. Perhaps unsurprisingly, and as noted earlier in this article, the literature shows that hpm1 deletion leads to increased stop-codon readthrough, programmed −1 ribosomal frameshifting, amino acid misincorporation, and increased resistance to ribosome binding drugs such as cycloheximide and anisomycin (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ).
      Finally, we asked whether crosslinks that change in an organelle's proteins could provide clues to altered function of that organelle. Crosslinks within proteins involved in mitochondrial processes were significantly altered on hpm1 knockout (discussed in detail in the aforementioned section) (Fig. 5, A and B). An increased burden of mitochondrial activity/dysfunction results in an increased production of superoxide by the organelle, for example, in respiratory growth when cells require larger and remodeled mitochondria and the ability to perform oxidative phosphorylation (
      • di Bartolomeo F.
      • Malina C.
      • Campbell K.
      • Mormino M.
      • Fuchs J.
      • Vorontsov E.
      • Gustafsson C.M.
      • Nielsen J.
      Absolute yeast mitochondrial proteome quantification reveals trade-off between biosynthesis and energy generation during diauxic shift.
      ). To determine whether altered mitochondrial PPIs and structures reflected a change in mitochondrial function, we determined whether the degree of superoxide production differed between WT and hpm1 deletion strains (Fig. 6C). Flow cytometry analysis of cells stained with the mitochondria-specific superoxide sensor MitoSOX revealed a significantly higher degree of staining, and hence higher burden of mitochondrial stress, in hpm1 deletion cells (p < 0.01 resulting from an unpaired two-tailed t test). This finding agrees with our crosslinking results of altered mitochondrial processes and in addition may relate to an increased propensity for respiratory growth as suggested by our proteomics data (Fig. 3C). More generally, this example shows how changes to multiple protein structures and interactions within an organelle may provide evidence for perturbations to change of function.

      Discussion

      Hpm1p is the sole histidine methyltransferase in yeast. Previous studies have established the importance of this enzyme for H243Me in Rpl3p and its role in regulating ribosomal assembly and translational fidelity (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ). Here, we have confirmed the position and stoichiometry of the Rpl3p-catalyzed H243 methyl site. We have also expanded the biology of Hpm1p and histidine methylation by describing new phenotypes for hpm1 knockout cells. Our observations are consistent with previous systematic screens, which reported ribosomal and nonribosomal phenotypes for such cells (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ,
      • Steinmetz L.M.
      • Scharfe C.
      • Deutschbauer A.M.
      • Mokranjac D.
      • Herman Z.S.
      • Jones T.
      • Chu A.M.
      • Giaever G.
      • Prokisch H.
      • Oefner P.J.
      • Davis R.W.
      Systematic screen for human disease genes in yeast.
      ,
      • Breslow D.K.
      • Cameron D.M.
      • Collins S.R.
      • Schuldiner M.
      • Stewart-Ornstein J.
      • Newman H.W.
      • Braun S.
      • Madhani H.D.
      • Krogan N.J.
      • Weissman J.S.
      A comprehensive strategy enabling high-resolution functional analysis of the yeast genome.
      ,
      • Brown J.A.
      • Sherlock G.
      • Myers C.L.
      • Burrows N.M.
      • Deng C.
      • Wu H.I.
      • McCann K.E.
      • Troyanskaya O.G.
      • Brown J.M.
      Global analysis of gene function in yeast by quantitative phenotypic profiling.
      ,
      • Deutschbauer A.M.
      • Jaramillo D.F.
      • Proctor M.
      • Kumm J.
      • Hillenmeyer M.E.
      • Davis R.W.
      • Nislow C.
      • Giaever G.
      Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast.
      ,
      • Jorgensen P.
      • Nishikawa J.L.
      • Breitkreutz B.J.
      • Tyers M.
      Systematic identification of pathways that couple cell growth and division in yeast.
      ,
      • Bengtsson O.
      • Jeppsson M.
      • Sonderegger M.
      • Parachin N.S.
      • Sauer U.
      • Hahn-Hägerdal B.
      • Gorwa-Grauslund M.F.
      Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
      ,
      • Ratnakumar S.
      • Hesketh A.
      • Gkargkas K.
      • Wilson M.
      • Rash B.M.
      • Hayes A.
      • Tunnacliffe A.
      • Oliver S.G.
      Phenomic and transcriptomic analyses reveal that autophagy plays a major role in desiccation tolerance in Saccharomyces cerevisiae.
      ,
      • Michaillat L.
      • Mayer A.
      Identification of genes affecting vacuole membrane fragmentation in Saccharomyces cerevisiae.
      ,
      • Lesage G.
      • Sdicu A.M.
      • Ménard P.
      • Shapiro J.
      • Hussein S.
      • Bussey H.
      Analysis of β-1,3-glucan assembly in Saccharomyces cerevisiae using a synthetic interaction network and altered sensitivity to caspofungin.
      ,
      • Lesage G.
      • Shapiro J.
      • Specht C.A.
      • Sdicu A.M.
      • Ménard P.
      • Hussein S.
      • Tong A.H.Y.
      • Boone C.
      • Bussey H.
      An interactional network of genes involved in chitin synthesis in Saccharomyces cerevisiae.
      ,
      • Tong A.H.Y.
      • Lesage G.
      • Bader G.D.
      • Ding H.
      • Xu H.
      • Xin X.
      • Young J.
      • Berriz G.F.
      • Brost R.L.
      • Chang M.
      • Chen Y.Q.
      • Cheng X.
      • Chua G.
      • Friesen H.
      • Goldberg D.S.
      • et al.
      Global mapping of the yeast genetic interaction network.
      ). To investigate the molecular underpinnings of these phenotypes, we used the complementary techniques of proteomics and crosslinking MS. Strikingly, we observed distinct yet functionally convergent changes to many aspects of the proteome including protein abundance, PTMs, structural conformations, and PPIs. From this, we identified a role for Hpm1p in carbon metabolism and membrane biology. Importantly, we found that changes to the interactions or structure of proteins, as detected by crosslinking MS, could explain or predict aspects of Hpm1p biology not seen with expression-based proteomics.

      Monitoring Protein Structure and Interactions Generates Novel Insights Into Mechanisms and Phenotypes

      A focus of this study was the use of large-scale quantitative crosslinking MS to investigate protein function. Our results highlight several benefits of such an approach, including its orthogonality to protein expression analysis, ability to molecularly explore phenotype in an unbiased manner, and ability to generate mechanistic insights.
      An interesting case study is the structure–function changes of Pma1p. Glucose levels can directly regulate its phosphorylation status, its structural conformation and interactions, resulting in modulation of its activity (summarized in Refs. (
      • Morsomme P.
      • Slayman C.W.
      • Goffeau A.
      Mutagenic study of the structure, function and biogenesis of the yeast plasma membrane H+-ATPase.
      ,
      • Lecchi S.
      • Nelson C.J.
      • Allen K.E.
      • Swaney D.L.
      • Thompson K.L.
      • Coon J.J.
      • Sussman M.R.
      • Slayman C.W.
      Tandem phosphorylation of Ser-911 and Thr-912 at the C terminus of yeast plasma membrane H+-ATPase leads to glucose-dependent activation.
      )). This plasma membrane ATPase did not change in abundance upon hpm1 knockout in our deep proteome analysis (Fig. 3). However, changes in Pma1p crosslinks (Fig. 5B) on hpm1 knockout were observed; all were intraprotein crosslinks, capturing changes to the structural conformation of the protein. The discovery of these changes prompted a functional investigation into extracellular pH, which is affected by the proton pumping of Pma1p. Experiments revealed a measurable difference in phenotype, with a lower extracellular acidification rate in Δhpm1 (Fig. 6A). Given our observation that hpm1 knockout cells are sensitive to the membrane stressor amphotericin B (Fig. 2), and that changes to pH are known to modulate membrane fidelity (
      • Brandão R.L.
      • Rosa J.C.C.
      • Nicoli J.R.
      • Almeida M.V.S.
      • do Carmo A.P.
      • Queiros H.T.
      • Castro I.M.
      Investigating acid stress response in different Saccharomyces strains.
      ), altered Pma1p structure may be a mechanism contributing to this phenotype.
      The discovery, by crosslinking (Fig. 5, A and B), of changes in a PPI between two mitochondrial proteins led to a hypothesis of altered mitochondrial activity. Phb1p and Phb2p interact to form the PHB complex; an essential protein scaffold hub that coordinates proteostasis, membrane organization, chaperoning, and the assembly and stabilization of oxidative phosphorylation complexes (reviewed in Ref. (
      • Artal-Sanz M.
      • Tavernarakis N.
      Prohibitin and mitochondrial biology.
      )). To determine whether this changed interaction affects mitochondrial activity, we used flow cytometry and functional dyes. This detected an increase in mitochondrial superoxide (Fig. 6C) in the hpm1 knockout, which is a byproduct of a high mitochondrial activity and load (reviewed in Ref. (
      • Kitagaki H.
      • Takagi H.
      Mitochondrial metabolism and stress response of yeast: Applications in fermentation technologies.
      )). Interestingly, this may explain the high sensitivity of Δhpm1 to oxidative stress (Fig. 2). The changes seen to mitochondrial PPIs, and structures of Phb1p, Phb2p, Odc2p, and Ssc1p, may reflect a reorganization of mitochondrial membranes, which is known to occur upon a shift to respiratory growth (
      • di Bartolomeo F.
      • Malina C.
      • Campbell K.
      • Mormino M.
      • Fuchs J.
      • Vorontsov E.
      • Gustafsson C.M.
      • Nielsen J.
      Absolute yeast mitochondrial proteome quantification reveals trade-off between biosynthesis and energy generation during diauxic shift.
      ). To summarize this example, we were able to detect and validate a functional effect, indicated by a change in PPI, and which was detected by crosslinking MS.

      Alterations to Carbon Metabolism and Membranes in Δhpm1 Cells

      In the hpm1 knockout, there was evidence for the modulation of sugar synthesis and metabolism. Our proteomic expression analysis revealed that some enzymes involved in glucose metabolism and fermentative growth were specifically downregulated in Δhpm1 compared with WT (Fig. 3). These changes may reflect a metabolic priming toward respiratory growth, also evidenced by the fact that Δhpm1 cells have a decreased growth rate on glucose but an increased growth rate on alternative carbon sources, such as ethanol (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ), xylose (
      • Bengtsson O.
      • Jeppsson M.
      • Sonderegger M.
      • Parachin N.S.
      • Sauer U.
      • Hahn-Hägerdal B.
      • Gorwa-Grauslund M.F.
      Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
      ), glycerol (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ), and sucrose or fructose (
      • Qian W.
      • Ma D.
      • Xiao C.
      • Wang Z.
      • Zhang J.
      The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
      ). Notably, the protein Ygp1p, which was significantly downregulated in Δhpm1, is involved in physical and behavioral changes associated with nutrient sensing. Specifically, Ygp1p is implicated in the regulation of biofilm formation and is itself regulated by extracellular glucose levels (
      • Vandenbosch D.
      • de Canck E.
      • Dhondt I.
      • Rigole P.
      • Nelis H.J.
      • Coenye T.
      Genomewide screening for genes involved in biofilm formation and miconazole susceptibility in Saccharomyces cerevisiae.
      ,
      • Destruelle M.
      • Holzer H.
      • Klionsky D.J.
      Identification and characterization of a novel yeast gene: The YGP1 gene product is a highly glycosylated secreted protein that is synthesized in response to nutrient limitation.
      ). We did not detect a concomitant increase in respiratory proteins. However, in our study, cells were cultured in media containing glucose as the sole carbon source. They were harvested at midlog phase, before high levels of ethanol could accumulate from fermentation (the diauxic shift) and before substantial metabolic remodeling occurs. We did however directly observe alterations to the structure of two enzymes involved in gluconeogenesis and glycolysis, Cdc19p and Eno2p, in our crosslinking data (Fig. 5B).
      We found that the trehalose biosynthetic pathway (Tsl1p, Tps1p, Tps2p, and Tps3p) was downregulated upon hpm1 knockout (Fig. 3, C and D). The disaccharide trehalose is up to 15% of the dry cell weight of yeast (
      • Lillie S.H.
      • Pringle J.R.
      Reserve carbohydrate metabolism in Saccharomyces cerevisiae: Responses to nutrient limitation.
      ) and acts not only as a carbohydrate reservoir but also as a structural protectant against physical stressors such as freezing and drying (reviewed in Refs. (
      • Lillie S.H.
      • Pringle J.R.
      Reserve carbohydrate metabolism in Saccharomyces cerevisiae: Responses to nutrient limitation.
      ,
      • Eleutherio E.
      • Panek A.
      • de Mesquita J.F.
      • Trevisol E.
      • Magalhães R.
      Revisiting yeast trehalose metabolism.
      )). The downregulation of the trehalose pathway is consistent with our observations that Δhpm1 yeasts are sensitive to cold stress (Fig. 2) and the observations by others of a decreased tolerance to desiccation (
      • Ratnakumar S.
      • Hesketh A.
      • Gkargkas K.
      • Wilson M.
      • Rash B.M.
      • Hayes A.
      • Tunnacliffe A.
      • Oliver S.G.
      Phenomic and transcriptomic analyses reveal that autophagy plays a major role in desiccation tolerance in Saccharomyces cerevisiae.
      ). Interestingly, the trehalose precursor trehalose-6-phosphate, which is catalyzed by Tps1p from glucose-6-phosphate and UDP-glucose, has also been shown to directly regulate the ability of yeast to undergo glycolysis (
      • Vicente R.L.
      • Spina L.
      • Gómez J.P.L.
      • Dejean S.
      • Parrou J.L.
      • François J.M.
      Trehalose-6-phosphate promotes fermentation and glucose repression in Saccharomyces cerevisiae.
      ,
      • Thevelein J.M.
      • Hohmann S.
      Trehalose synthase: Guard to the gate of glycolysis in yeast?.
      ), and Δtps1 cells entirely lose the ability to grow on glucose (
      • Thevelein J.M.
      • Hohmann S.
      Trehalose synthase: Guard to the gate of glycolysis in yeast?.
      ).
      We found that the deletion of hpm1 resulted in alterations to membrane integrity and function. Our plate assays identified severe growth defects in the membrane stressor 0.1 μg/ml amphotericin-B upon hpm1 knockout (Fig. 2), and our crosslinking and follow-up validation studies revealed a change to Pma1p structure and function (Figs. 5B and 6A). There are several other molecular links to membrane dysfunction in our study. Amphotericin-B sequesters ergosterol (
      • Gray K.C.
      • Palacios D.S.
      • Dailey I.
      • Endo M.M.
      • Uno B.E.
      • Wilcock B.C.
      • Burke M.D.
      Amphotericin primarily kills yeast by simply binding ergosterol.
      ), which is important for the structure and function of membranes, and to membrane-bound organelles including vacuoles. Indeed, hpm1 knockouts are reported to have alterations to both vacuolar structure and function (
      • Michaillat L.
      • Mayer A.
      Identification of genes affecting vacuole membrane fragmentation in Saccharomyces cerevisiae.
      ,
      • Bonangelino C.J.
      • Chavez E.M.
      • Bonifacino J.S.
      Genomic screen for vacuolar protein sorting genes in Saccharomyces cerevisiae.
      ), which may be associated with our observed change in the abundance of the vacuolar protein Pho84p, which was highly downregulated in the knockout strain. Dysfunctions in sterol processes are known to result in a decreased tolerance to salt stress (
      • Kodedová M.
      • Sychrová H.
      Changes in the sterol composition of the plasma membrane affect membrane potential, salt tolerance and the activity of multidrug resistance pumps in Saccharomyces cerevisiae.
      ), and we found that Δhpm1 yeast also exhibited significant growth defects under 1 M NaCl. Alternatively, the salt stress phenotype could be related to the changes in the intralink URP we observed in the Ena2p sodium transporter protein. We were also able to identify likely changes in structure within other membrane proteins (which were not changing in abundance themselves) associated with Δhpm1 phenotypes. These proteins included the plasma membrane ATPase Pma1p (as noted previously), the endoplasmic reticulum protein importer Kar2p, and mitochondrial membrane protein sorters Phb1p and Phb2p.

      The Substrates and Function of Hpm1p

      In addition to revealing new biology for Hpm1p, our study has also generated insights into the direct function of Hpm1p. These include the possibility of undescribed substrates for the enzyme. The question of substrate proteins for Hpm1p, that exist in addition to Rpl3p, remains unresolved. In one study (
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ), radiolabeled methylhistidines were detected by amino acid analysis and largely attributed to ribosomal fractions. A small amount of radiolabel was however detected in membrane-bound organelles. This was attributed to the process of ribosome biogenesis within the nucleus, as SDS-PAGE revealed only one band, corresponding to Rpl3p in size, that was absent in Δhpm1 fractions (
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ). A later study by the same group (
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ) constructed an Rpl3p H243A mutant to detect radiolabeled methylation in potential non-Rpl3p substrates. This revealed two potential substrates in the postmitochondrial cellular fraction, although their identities were not determined (
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ). While we did not find any evidence for alternative Hpm1p substrates in our reanalysis of published methyl-SILAC–LC–MS/MS data (
      • Hart-Smith G.
      • Yagoub D.
      • Tay A.P.
      • Pickford R.
      • Wilkins M.R.
      Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
      ), novel methyl-histidine sites could be either entirely absent or present at low stoichiometry in the growth conditions sampled. Equally, they could only be amenable to discovery with different endoproteinases to those used (
      • Hart-Smith G.
      • Yagoub D.
      • Tay A.P.
      • Pickford R.
      • Wilkins M.R.
      Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
      ). Our investigation of the H243A mutant (supplemental Fig. S2) revealed some evidence for additional Hmp1p substrates, in that the mutant does not completely phenocopy the hmp1 knockout. This phenomenon was also seen previously (
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ). But the possibility also remains that Hpm1p protein itself has additional molecular functions to histidine methyltransferase activity. Interestingly, the stressors that just affected Δhpm1 but not Rpl3p H243A (oxidative stress, amphotericin B) related to the mitochondrial and membrane processes that we could functionally validate.
      Another means by which Hpm1p may directly mediate phenotypes, that we and others have observed, is through the modulation of ribosome heterogeneity. Alterations to the ribosome’s protein/RNA composition, PTM, and subcellular localization have been demonstrated to modulate translational specificity and fidelity (reviewed in Ref. (
      • Emmott E.
      • Jovanovic M.
      • Slavov N.
      Ribosome stoichiometry: From form to function.
      )). Our study, and those of others (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ) have only studied Rpl3p H243 methyl site stoichiometry in cells grown in media containing glucose as the sole carbon source. In these conditions, the site appears to be constitutively present. However, in another link to metabolism, it is known that the levels of hpm1 transcripts are regulated by metabolic conditions; for example, they are specifically downregulated in response to xylose (
      • Bengtsson O.
      • Jeppsson M.
      • Sonderegger M.
      • Parachin N.S.
      • Sauer U.
      • Hahn-Hägerdal B.
      • Gorwa-Grauslund M.F.
      Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
      ). As the H243 site lies within the decoding center of the ribosome, and makes direct contact with rRNAs, modulation of Rpl3p histidine methylation has potential to allosterically alter the structure of this complex, and thus alter translational specificity. In fact, modulation of translation specificity through Rpl3p histidine methylation has been observed in the knockout of hpm1’s human ortholog METTL18, which displays altered ribosome occupancies for specific groups of genes, including those involving vacuolar and metabolic processes (
      • Małecki J.M.
      • Odonohue M.F.
      • Kim Y.
      • Jakobsson M.E.
      • Gessa L.
      • Pinto R.
      • Wu J.
      • Davydova E.
      • Moen A.
      • Olsen J.v.
      • Thiede B.
      • Gleizes P.E.
      • Leidel S.A.
      • Falnes P.
      Human METTL18 is a histidine-specific methyltransferase that targets RPL3 and affects ribosome biogenesis and function.
      ). It has also been shown to affect ribosomal translocation at tyrosine residues (
      • Matsuura-Suzuki E.
      • Shimazu T.
      • Takahashi M.
      • Kotoshiba K.
      • Suzuki T.
      • Kashiwagi K.
      • Sohtome Y.
      • Akakabe M.
      • Sodeoka M.
      • Dohmae N.
      • Ito T.
      • Shinkai Y.
      • Iwasaki S.
      METTL18-mediated histidine methylation on RPL3 modulates translation elongation for proteostasis maintenance.
      ). There is some functional evidence of change to the structure of the ribosome in hpm1 knockout cells, which display increased resistance to translational inhibitors such as cycloheximide, verrucarin, and anisomycin (
      • Webb K.J.
      • Zurita-Lopez C.I.
      • Al-Hadid Q.
      • Laganowsky A.
      • Young B.D.
      • Lipson R.S.
      • Souda P.
      • Faull K.F.
      • Whitelegge J.P.
      • Clarke S.G.
      A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Munroe W.
      • Dzialo M.C.
      • Chanfreau G.F.
      • Clarke S.G.
      Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
      ,
      • Al-Hadid Q.
      • Roy K.
      • Chanfreau G.
      • Clarke S.G.
      Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
      ,
      • Al-Hadid Q.
      • White J.
      • Clarke S.
      Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
      ). In our study, we were able to directly observe specific changes to the structure of some ribosomal proteins, both locally surrounding Rpl3p and distally (Fig. 6B). These allosteric changes were expected as the H243 residue makes direct contact with the rRNA core of the assembled ribosome.

      Conclusions

      Our study has shown how untargeted screens of PPIs and low-resolution measures of protein structure, combined with functional readouts and traditional quantitative proteomics, can shed new light on complicated phenotypes. It has substantially expanded the biological role of histidine methylation in S. cerevisiae and raised the possibility of alternative Hmp1p substrates. This study has also raised interesting questions about the role of histidine methylation as a possible modulator of ribosomal heterogeneity and how it may alter translational specificity. Technical advances, such as the development of histidine methylation affinity reagents and the use of orthogonal proteomics methods, will ultimately decipher the enigmatic functions of Hpm1p.

      Data Availability

      All MS data and spectral identifications have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository (
      • Perez-Riverol Y.
      • Bai J.
      • Bandla C.
      • García-Seisdedos D.
      • Hewapathirana S.
      • Kamatchinathan S.
      • Kundu D.J.
      • Prakash A.
      • Frericks-Zipper A.
      • Eisenacher M.
      • Walzer M.
      • Wang S.
      • Brazma A.
      • Vizcaíno J.A.
      The PRIDE database resources in 2022: a hub for mass spectrometry-based proteomics evidences.
      ) with the dataset identifier PXD027953. Annotated spectra showing the best CSM for each cross-linked peptide pair has been provided as supplemental data (supplemental Fig. S6). Targeted MS analyses have been uploaded to Panorama, available at https://panoramaweb.org/0M3Z0c.url. HTML RMarkdown reports and R scripts are available for the crosslinking analyses, accessible at https://github.com/TaraBartolec/Hpm1.

      Supplemental data

      This article contains supplemental data.

      Conflict of interest

      The authors declare no competing interests.

      Acknowledgments

      We acknowledge the staff at the Bioanalytical Mass Spectrometry Facility at the University of New South Wales (UNSW), especially Dr Ling Zhong and Dr Mark Raftery, for maintenance of mass spectrometers, Dr Helene Lebhar from the Recombinant Products Facility at the UNSW for assistance with HPLC method development, Dr Maitrea Dunham and Cindy Yeh at the University of Washington for the supply of strains from the Yeast Knockout Collection, Ryan Separovich from the UNSW for assistance with editing for clarity, and Jared Mohr from the University of Washington for sharing automated spectra annotation scripts. We also acknowledge the assistance of Florence Bartlett-Tomasetig at the Katharina Gaus Light Microscopy Facility for supply of fluorophores, Dr Emma Johansson Beves for assistance in flow cytometry experimental design and analysis, and Dr Xabier Vázquez-Campos for assistance with statistics.

      Funding and additional information

      T. K. B. acknowledges the support of an Australian Government Research Training Program scholarship and the UNSW Adrian Lee Travel scholarship. M. R. W. acknowledges support from the Australian Research Council, Australia .

      Author contributions

      T .K. B., J. J. H., and M. R. W. conceptualization; T. K. B., A. K., and J. D. C. methodology; T. K. B. and A. K. software; T. K. B. validation; T. K. B. formal analysis; T. K. B. investigation; A. K. and J. E. B. resources; T. K. B. and A. K. data curation; T. K. B. writing–original draft; J. J. H., A. K., J. D. C., and M. R. W. writing–review & editing; T. K. B. visualization; J. J. H., J. D. C., J. E. B., and M. R. W. supervision; J. E. B. and M .R. W. project administration; M. R. W. funding acquisition.

      References

        • Kwiatkowski S.
        • Drozak J.
        Protein histidine methylation.
        Curr. Protein Pept. Sci. 2020; 21: 675-689
        • Jakobsson M.E.
        Enzymology and significance of protein histidine methylation.
        J. Biol. Chem. 2021; 297: 101130
        • Lv M.
        • Cao D.
        • Zhang L.
        • Hu C.
        • Li S.
        • Zhang P.
        • Zhu L.
        • Yi X.
        • Li C.
        • Yang A.
        • Yang Z.
        • Zhu Y.
        • Zhang K.
        • Pan W.
        METTL9 mediated N1-histidine methylation of zinc transporters is required for tumor growth.
        Protein Cell. 2021; 12: 965-970
        • Davydova E.
        • Shimazu T.
        • Schuhmacher M.K.
        • Jakobsson M.E.
        • Willemen H.L.D.M.
        • Liu T.
        • Moen A.
        • Ho A.Y.Y.
        • Małecki J.
        • Schroer L.
        • Pinto R.
        • Suzuki T.
        • Grønsberg I.A.
        • Sohtome Y.
        • Akakabe M.
        • et al.
        The methyltransferase METTL9 mediates pervasive 1-methylhistidine modification in mammalian proteomes.
        Nat. Commun. 2021; 12: 891
        • Kwiatkowski S.
        • Seliga A.K.
        • Vertommen D.
        • Terreri M.
        • Ishikawa T.
        • Grabowska I.
        • Tiebe M.
        • Teleman A.A.
        • Jagielski A.K.
        • Veiga-Da-Cunha M.
        • Drozak J.
        SETD3 protein is the actin-specific histidine N-methyltransferase.
        eLife. 2018; 7e37921
        • Wilkinson A.W.
        • Diep J.
        • Dai S.
        • Liu S.
        • Ooi Y.S.
        • Song D.
        • Li T.M.
        • Horton J.R.
        • Zhang X.
        • Liu C.
        • Trivedi D.v.
        • Ruppel K.M.
        • Vilches-Moure J.G.
        • Casey K.M.
        • Mak J.
        • et al.
        SETD3 is an actin histidine methyltransferase that prevents primary dystocia.
        Nature. 2019; 565: 372-376
        • Diep J.
        • Ooi Y.S.
        • Wilkinson A.W.
        • Peters C.E.
        • Foy E.
        • Johnson J.R.
        • Zengel J.
        • Ding S.
        • Weng K.F.
        • Laufman O.
        • Jang G.
        • Xu J.
        • Young T.
        • Verschueren E.
        • Kobluk K.J.
        • et al.
        Enterovirus pathogenesis requires the host methyltransferase SETD3.
        Nat. Microbiol. 2019; 4: 2523-2537
        • Cheng X.
        • Hao Y.
        • Shu W.
        • Zhao M.
        • Zhao C.
        • Wu Y.
        • Peng X.
        • Yao P.
        • Xiao D.
        • Qing G.
        • Pan Z.
        • Yin L.
        • Hu D.
        • Du H.N.
        Cell cycle-dependent degradation of the methyltransferase SETD3 attenuates cell proliferation and liver tumorigenesis.
        J. Biol. Chem. 2017; 292: 9022-9033
        • Chen Z.
        • Yan C.T.
        • Dou Y.
        • Viboolsittiseri S.S.
        • Wang J.H.
        The role of a newly identified SET domain-containing protein, SETD3, in oncogenesis.
        Haematologica. 2013; 98: 739-743
        • Kapell S.
        • Jakobsson M.E.
        Large-scale identification of protein histidine methylation in human cells.
        NAR Genom. Bioinform. 2021; 3https://doi.org/10.1093/nargab/lqab045
        • Webb K.J.
        • Zurita-Lopez C.I.
        • Al-Hadid Q.
        • Laganowsky A.
        • Young B.D.
        • Lipson R.S.
        • Souda P.
        • Faull K.F.
        • Whitelegge J.P.
        • Clarke S.G.
        A novel 3-methylhistidine modification of yeast ribosomal protein Rpl3 is dependent upon the YIL110W methyltransferase.
        J. Biol. Chem. 2010; 285: 37598-375606
        • Małecki J.M.
        • Odonohue M.F.
        • Kim Y.
        • Jakobsson M.E.
        • Gessa L.
        • Pinto R.
        • Wu J.
        • Davydova E.
        • Moen A.
        • Olsen J.v.
        • Thiede B.
        • Gleizes P.E.
        • Leidel S.A.
        • Falnes P.
        Human METTL18 is a histidine-specific methyltransferase that targets RPL3 and affects ribosome biogenesis and function.
        Nucl. Acids Res. 2021; 49: 3185-3203
        • Meskauskas A.
        • Dinman J.D.
        Ribosomal protein L3 functions as a “rocker switch” to aid in coordinating of large subunit-associated functions in eukaryotes and Archaea.
        Nucl. Acids Res. 2008; 36: 6175-6186
        • Meskauskas A.
        • Dinman J.D.
        Ribosomal protein L3: Gatekeeper to the A site.
        Mol. Cell. 2007; 25: 878-888
        • Al-Hadid Q.
        • Roy K.
        • Munroe W.
        • Dzialo M.C.
        • Chanfreau G.F.
        • Clarke S.G.
        Histidine methylation of yeast ribosomal protein Rpl3p is required for proper 60S subunit assembly.
        Mol. Cel. Biol. 2014; 34: 2903-2916
        • Al-Hadid Q.
        • Roy K.
        • Chanfreau G.
        • Clarke S.G.
        Methylation of yeast ribosomal protein Rpl3 promotes translational elongation fidelity.
        RNA. 2016; 22: 489-498
        • Al-Hadid Q.
        • White J.
        • Clarke S.
        Ribosomal protein methyltransferases in the yeast Saccharomyces cerevisiae: Roles in ribosome biogenesis and translation.
        Biochem. Biophys. Res. Commun. 2016; 470: 552-557
        • Cherry J.M.
        • Hong E.L.
        • Amundsen C.
        • Balakrishnan R.
        • Binkley G.
        • Chan E.T.
        • Christie K.R.
        • Costanzo M.C.
        • Dwight S.S.
        • Engel S.R.
        • Fisk D.G.
        • Hirschman J.E.
        • Hitz B.C.
        • Karra K.
        • Krieger C.J.
        • et al.
        Saccharomyces genome database: The genomics resource of budding yeast.
        Nucl. Acids Res. 2012; 40: D700-D705
        • Qian W.
        • Ma D.
        • Xiao C.
        • Wang Z.
        • Zhang J.
        The genomic landscape and evolutionary resolution of antagonistic pleiotropy in yeast.
        Cell Rep. 2012; 2: 1399-1410
        • Steinmetz L.M.
        • Scharfe C.
        • Deutschbauer A.M.
        • Mokranjac D.
        • Herman Z.S.
        • Jones T.
        • Chu A.M.
        • Giaever G.
        • Prokisch H.
        • Oefner P.J.
        • Davis R.W.
        Systematic screen for human disease genes in yeast.
        Nat. Genet. 2002; 31: 400-404
        • Breslow D.K.
        • Cameron D.M.
        • Collins S.R.
        • Schuldiner M.
        • Stewart-Ornstein J.
        • Newman H.W.
        • Braun S.
        • Madhani H.D.
        • Krogan N.J.
        • Weissman J.S.
        A comprehensive strategy enabling high-resolution functional analysis of the yeast genome.
        Nat. Met. 2008; 5: 711-718
        • Brown J.A.
        • Sherlock G.
        • Myers C.L.
        • Burrows N.M.
        • Deng C.
        • Wu H.I.
        • McCann K.E.
        • Troyanskaya O.G.
        • Brown J.M.
        Global analysis of gene function in yeast by quantitative phenotypic profiling.
        Mol. Syst. Biol. 2006; 2https://doi.org/10.1038/msb4100043
        • Deutschbauer A.M.
        • Jaramillo D.F.
        • Proctor M.
        • Kumm J.
        • Hillenmeyer M.E.
        • Davis R.W.
        • Nislow C.
        • Giaever G.
        Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast.
        Genetics. 2005; 169: 1915-1925
        • Jorgensen P.
        • Nishikawa J.L.
        • Breitkreutz B.J.
        • Tyers M.
        Systematic identification of pathways that couple cell growth and division in yeast.
        Science. 2002; 297: 395-400
        • Bengtsson O.
        • Jeppsson M.
        • Sonderegger M.
        • Parachin N.S.
        • Sauer U.
        • Hahn-Hägerdal B.
        • Gorwa-Grauslund M.F.
        Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering.
        Yeast. 2008; 25: 835-847
        • Ratnakumar S.
        • Hesketh A.
        • Gkargkas K.
        • Wilson M.
        • Rash B.M.
        • Hayes A.
        • Tunnacliffe A.
        • Oliver S.G.
        Phenomic and transcriptomic analyses reveal that autophagy plays a major role in desiccation tolerance in Saccharomyces cerevisiae.
        Mol. BioSyst. 2011; 7: 139-149
        • Michaillat L.
        • Mayer A.
        Identification of genes affecting vacuole membrane fragmentation in Saccharomyces cerevisiae.
        PLoS One. 2013; 8e54160
        • Lesage G.
        • Sdicu A.M.
        • Ménard P.
        • Shapiro J.
        • Hussein S.
        • Bussey H.
        Analysis of β-1,3-glucan assembly in Saccharomyces cerevisiae using a synthetic interaction network and altered sensitivity to caspofungin.
        Genetics. 2004; 167: 35-49
        • Lesage G.
        • Shapiro J.
        • Specht C.A.
        • Sdicu A.M.
        • Ménard P.
        • Hussein S.
        • Tong A.H.Y.
        • Boone C.
        • Bussey H.
        An interactional network of genes involved in chitin synthesis in Saccharomyces cerevisiae.
        BMC Genet. 2005; 6: 8
        • Tong A.H.Y.
        • Lesage G.
        • Bader G.D.
        • Ding H.
        • Xu H.
        • Xin X.
        • Young J.
        • Berriz G.F.
        • Brost R.L.
        • Chang M.
        • Chen Y.Q.
        • Cheng X.
        • Chua G.
        • Friesen H.
        • Goldberg D.S.
        • et al.
        Global mapping of the yeast genetic interaction network.
        Science. 2004; 303: 808-813
        • Chavez J.D.
        • Mohr J.P.
        • Mathay M.
        • Zhong X.
        • Keller A.
        • Bruce J.E.
        Systems structural biology measurements by in vivo cross-linking with mass spectrometry.
        Nat. Protoc. 2019; 14: 2318-2343
        • Winzeler E.A.
        • Shoemaker D.D.
        • Astromoff A.
        • Liang H.
        • Anderson K.
        • Andre B.
        • Bangham R.
        • Benito R.
        • Boeke J.D.
        • Bussey H.
        • Chu A.M.
        • Connelly C.
        • Davis K.
        • Dietrich F.
        • Dow S.W.
        • et al.
        Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis.
        Science. 1999; 285: 901-906
        • Janke C.
        • Magiera M.M.
        • Rathfelder N.
        • Taxis C.
        • Reber S.
        • Maekawa H.
        • Moreno-Borchart A.
        • Doenges G.
        • Schwob E.
        • Schiebel E.
        • Knop M.
        A versatile toolbox for PCR-based tagging of yeast genes: New fluorescent proteins, more markers and promoter substitution cassettes.
        Yeast. 2004; 21: 947-962
        • Gietz R.D.
        • Schiestl R.H.
        High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method.
        Nat. Protoc. 2007; 2: 31-34
        • Hamey J.J.
        • Winter D.L.
        • Yagoub D.
        • Overall C.M.
        • Hart-Smith G.
        • Wilkins M.R.
        Novel N-terminal and lysine methyltransferases that target translation elongation factor 1A in yeast and human.
        Mol. Cell Proteomics. 2016; 15: 164-176
        • Separovich R.J.
        • Wong M.W.M.
        • Chapman T.R.
        • Slavich E.
        • Hamey J.J.
        • Wilkins M.R.
        Post-translational modification analysis of Saccharomyces cerevisiae histone methylation enzymes reveals phosphorylation sites of regulatory potential.
        J. Biol. Chem. 2021; 296: 100192
        • Pino L.K.
        • Searle B.C.
        • Bollinger J.G.
        • Nunn B.
        • MacLean B.
        • MacCoss M.J.
        The Skyline ecosystem: Informatics for quantitative mass spectrometry proteomics.
        Mass Spectrom. Rev. 2020; 39: 229-244
        • Hart-Smith G.
        • Yagoub D.
        • Tay A.P.
        • Pickford R.
        • Wilkins M.R.
        Large scale mass spectrometry-based identifications of enzyme-mediated protein methylation are subject to high false discovery rates.
        Mol. Cell Proteomics. 2016; 15: 989-1006
        • Tay A.P.
        • Geoghegan V.
        • Yagoub D.
        • Wilkins M.R.
        • Hart-Smith G.
        MethylQuant: a tool for sensitive validation of enzyme-mediated protein methylation sites from heavy-methyl SILAC data.
        J. Proteome Res. 2018; 17: 359-373
        • Dannenmaier S.
        • Stiller S.B.
        • Morgenstern M.
        • Lübbert P.
        • Oeljeklaus S.
        • Wiedemann N.
        • Warscheid B.
        Complete native stable isotope labeling by amino acids of Saccharomyces cerevisiae for global proteomic analysis.
        Anal. Chem. 2018; 90: 10501-10509
        • Tyanova S.
        • Temu T.
        • Cox J.
        The MaxQuant computational platform for mass spectrometry-based shotgun proteomics.
        Nat. Protoc. 2016; 11: 2301-2319
        • Theodorakis E.
        • Antonakis A.N.
        • Baltsavia I.
        • Pavlopoulos G.A.
        • Samiotaki M.
        • Amoutzias G.D.
        • et al.
        ProteoSign v2: A faster and evolved user-friendly online tool for statistical analyses of differential proteomics.
        Nucl. Acids Res. 2021; 49: W573-W577
        • Mi H.
        • Muruganujan A.
        • Thomas P.D.
        PANTHER in 2013: Modeling the evolution of gene function, and other gene attributes, in the context of phylogenetic trees.
        Nucl. Acids Res. 2013; 41: D377-D386
        • Kessner D.
        • Chambers M.
        • Burke R.
        • Agus D.
        • Mallick P.
        ProteoWizard: Open source software for rapid proteomics tools development.
        Bioinformatics. 2008; 24: 2534-2536
        • Mohr J.P.
        • Perumalla P.
        • Chavez J.D.
        • Eng J.K.
        • Bruce J.E.
        Mango: A general tool for collision induced dissociation-cleavable cross-linked peptide identification.
        Anal. Chem. 2018; 90: 6028-6034
        • Eng J.K.
        • Jahan T.A.
        • Hoopmann M.R.
        Comet: An open-source MS/MS sequence database search tool.
        Proteomics. 2013; 13: 22-24
        • Keller A.
        • Chavez J.D.
        • Bruce J.E.
        Increased sensitivity with automated validation of XL-MS cleavable peptide cross-links.
        Bioinformatics. 2019; 35: 895-897
        • Schweppe D.K.
        • Zheng C.
        • Chavez J.D.
        • Navare A.T.
        • Wu X.
        • Eng J.K.
        • Bruce J.E.
        XLinkDB 2.0: Integrated, large-scale structural analysis of protein cross-linking data.
        Bioinformatics. 2016; 32: 2716-2718
        • Berman H.M.
        • Westbrook J.
        • Feng Z.
        • Gilliland G.
        • Bhat T.N.
        • Weissig H.
        • Shindyalov I.N.
        • Bourne P.E.
        The protein data bank.
        Nucl. Acids Res. 2000; 28: 235-242
        • Kahraman A.
        • Malmström L.
        • Aebersold R.
        Xwalk: Computing and visualizing distances in cross-linking experiments.
        Bioinformatics. 2011; 27: 2163-2164
        • Dana J.M.
        • Gutmanas A.
        • Tyagi N.
        • Qi G.
        • O’Donovan C.
        • Martin M.
        • Velankar S.
        SIFTS: Updated Structure integration with function, taxonomy and sequences resource allows 40-fold increase in coverage of structure-based annotations for proteins.
        Nucl. Acids Res. 2019; 47: D482-D489
        • Boca S.M.
        • Leek J.T.
        A direct approach to estimating false discovery rates conditional on covariates.
        PeerJ. 2018; 6e6035
        • Shannon P.
        • Markiel A.
        • Ozier O.
        • Baliga N.S.
        • Wang J.T.
        • Ramage D.
        • Amin N.
        • Schwikowski B.
        • Ideker T.
        Cytoscape: A software environment for integrated models of biomolecular interaction networks.
        Genome Res. 2003; 13: 2498-2504
        • Schiffrin B.
        • Radford S.E.
        • Brockwell D.J.
        • Calabrese A.N.
        PyXlinkViewer: A flexible tool for visualization of protein chemical cross-linking data within the PyMOL molecular graphics system.
        Protein Sci. 2020; 29: 1851-1857
        • Pellegrino S.
        • Demeshkina N.
        • Mancera-Martinez E.
        • Melnikov S.
        • Simonetti A.
        • Myasnikov A.
        • Yusupov M.
        • Yusupova G.
        • Hashem Y.
        Structural insights into the role of diphthamide on elongation factor 2 in mRNA reading-frame maintenance.
        J. Mol. Biol. 2018; 430: 2677-2687
        • Ong S.E.
        • Mittler G.
        • Mann M.
        Identifying and quantifying in vivo methylation sites by heavy methyl SILAC.
        Nat. Met. 2004; 1: 119-126
        • Alonso-López Di.
        • Campos-Laborie F.J.
        • Gutiérrez M.A.
        • Lambourne L.
        • Calderwood M.A.
        • Vidal M.
        • de Las Rivas J.
        APID Database: Redefining Protein-Protein Interaction Experimental Evidences and Binary Interactomes.
        Database, Oxford2019 (2019)
        • Dai H.
        • Wang Z.
        Histone modification patterns and their responses to environment.
        Curr. Environ. Health Rep. 2014; 1: 11-21
        • Rando O.J.
        • Winston F.
        Chromatin and transcription in yeast.
        Genetics. 2012; 190: 351-387
        • Signorile A.
        • Sgaramella G.
        • Bellomo F.
        • de Rasmo D.
        Prohibitins: A critical role in mitochondrial functions and implication in diseases.
        Cells. 2019; 8: 71
        • Portillo F.
        Regulation of plasma membrane H+-ATPase in fungi and plants.
        Biochim. Biophys. Acta - Rev. Biomem. 2000; 1469: 31-42
        • di Bartolomeo F.
        • Malina C.
        • Campbell K.
        • Mormino M.
        • Fuchs J.
        • Vorontsov E.
        • Gustafsson C.M.
        • Nielsen J.
        Absolute yeast mitochondrial proteome quantification reveals trade-off between biosynthesis and energy generation during diauxic shift.
        Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 7524-7535
        • Morsomme P.
        • Slayman C.W.
        • Goffeau A.
        Mutagenic study of the structure, function and biogenesis of the yeast plasma membrane H+-ATPase.
        Biochim. Biophys. Acta - Rev. Biomem. 2000; 1469: 133-157
        • Lecchi S.
        • Nelson C.J.
        • Allen K.E.
        • Swaney D.L.
        • Thompson K.L.
        • Coon J.J.
        • Sussman M.R.
        • Slayman C.W.
        Tandem phosphorylation of Ser-911 and Thr-912 at the C terminus of yeast plasma membrane H+-ATPase leads to glucose-dependent activation.
        J. Biol. Chem. 2007; 282: 35471-35481
        • Brandão R.L.
        • Rosa J.C.C.
        • Nicoli J.R.
        • Almeida M.V.S.
        • do Carmo A.P.
        • Queiros H.T.
        • Castro I.M.
        Investigating acid stress response in different Saccharomyces strains.
        J. Mycol. 2014; 2014: 1-9
        • Artal-Sanz M.
        • Tavernarakis N.
        Prohibitin and mitochondrial biology.
        Trends Endocrinol. Metab. 2009; 20: 394-401
        • Kitagaki H.
        • Takagi H.
        Mitochondrial metabolism and stress response of yeast: Applications in fermentation technologies.
        J. Biosci. Bioeng. 2014; 117: 383-393
        • Vandenbosch D.
        • de Canck E.
        • Dhondt I.
        • Rigole P.
        • Nelis H.J.
        • Coenye T.
        Genomewide screening for genes involved in biofilm formation and miconazole susceptibility in Saccharomyces cerevisiae.
        FEMS Yeast Res. 2013; 13: 720-730
        • Destruelle M.
        • Holzer H.
        • Klionsky D.J.
        Identification and characterization of a novel yeast gene: The YGP1 gene product is a highly glycosylated secreted protein that is synthesized in response to nutrient limitation.
        Mol. Cel. Biol. 1994; 14: 2740-2754
        • Lillie S.H.
        • Pringle J.R.
        Reserve carbohydrate metabolism in Saccharomyces cerevisiae: Responses to nutrient limitation.
        J. Bacteriol. 1980; 143: 1384-1394
        • Eleutherio E.
        • Panek A.
        • de Mesquita J.F.
        • Trevisol E.
        • Magalhães R.
        Revisiting yeast trehalose metabolism.
        Curr. Genet. 2015; 61: 263-274
        • Vicente R.L.
        • Spina L.
        • Gómez J.P.L.
        • Dejean S.
        • Parrou J.L.
        • François J.M.
        Trehalose-6-phosphate promotes fermentation and glucose repression in Saccharomyces cerevisiae.
        Microb. Cell. 2018; 5: 444-459
        • Thevelein J.M.
        • Hohmann S.
        Trehalose synthase: Guard to the gate of glycolysis in yeast?.
        Trends Biochem. Sci. 1995; 20: 3-10
        • Gray K.C.
        • Palacios D.S.
        • Dailey I.
        • Endo M.M.
        • Uno B.E.
        • Wilcock B.C.
        • Burke M.D.
        Amphotericin primarily kills yeast by simply binding ergosterol.
        Proc. Natl. Acad. Sci. U. S. A. 2012; 109: 2234-2239
        • Bonangelino C.J.
        • Chavez E.M.
        • Bonifacino J.S.
        Genomic screen for vacuolar protein sorting genes in Saccharomyces cerevisiae.
        Mol. Biol. Cell. 2002; 13: 2486-2501
        • Kodedová M.
        • Sychrová H.
        Changes in the sterol composition of the plasma membrane affect membrane potential, salt tolerance and the activity of multidrug resistance pumps in Saccharomyces cerevisiae.
        PLoS One. 2015; 10e0139306
        • Emmott E.
        • Jovanovic M.
        • Slavov N.
        Ribosome stoichiometry: From form to function.
        Trends Biochem. Sci. 2019; 44: 95-109
        • Matsuura-Suzuki E.
        • Shimazu T.
        • Takahashi M.
        • Kotoshiba K.
        • Suzuki T.
        • Kashiwagi K.
        • Sohtome Y.
        • Akakabe M.
        • Sodeoka M.
        • Dohmae N.
        • Ito T.
        • Shinkai Y.
        • Iwasaki S.
        METTL18-mediated histidine methylation on RPL3 modulates translation elongation for proteostasis maintenance.
        Elife. 2022; 11: e72780
        • Perez-Riverol Y.
        • Bai J.
        • Bandla C.
        • García-Seisdedos D.
        • Hewapathirana S.
        • Kamatchinathan S.
        • Kundu D.J.
        • Prakash A.
        • Frericks-Zipper A.
        • Eisenacher M.
        • Walzer M.
        • Wang S.
        • Brazma A.
        • Vizcaíno J.A.
        The PRIDE database resources in 2022: a hub for mass spectrometry-based proteomics evidences.
        Nucleic Acids Res. 2022; 50: D543-D552