Methods to measure ubiquitin chain length and linkage

CHAPTER FIVE

Methods to measure ubiquitin chain length and linkage Fumiaki Ohtake†, Hikaru Tsuchiya†, Keiji Tanaka, Yasushi Saeki* Laboratory of Protein Metabolism, Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan *Corresponding author: e-mail address: [email protected]

Contents 1. Introduction 2. Ub-AQUA/PRM analysis to quantify ubiquitin linkages and K48/K63 branched chains 2.1 In-gel digestion and purification of ubiquitin peptides 2.2 In-gel digestion and purification of branched ubiquitin chain-derived peptides 2.3 Preparation and quality control of AQUA peptides 2.4 PRM for ubiquitin quantification 3. Ub-ProT analysis to measure ubiquitin chain length of in vitro and in vivo ubiquitin conjugates 3.1 Purification of TR-TUBE for Ub-ProT analysis 3.2 Ub-ProT for in vitro synthesized ubiquitin chains 3.3 Ub-ProT analysis of ubiquitin conjugates from Saccharomyces cerevisiae 4. Summary and conclusions References

106 109 109 114 116 118 123 123 126 129 130 131

Abstract To understand the biological roles of different ubiquitin chains, it is important to determine the types of ubiquitin linkages, the lengths of the polymers, and the combinations of ubiquitin chains attached to substrates. In this chapter, we describe a mass spectrometrybased quantification method of ubiquitin chains, named Ub-AQUA/PRM (ubiquitinabsolute quantification/parallel reaction monitoring), for direct and highly sensitive measurement of the stoichiometry of all eight ubiquitin-ubiquitin linkage types simultaneously. We also show a method to quantify the K48/K63 branched ubiquitin chain, a recently identified ubiquitin signal with a complex topology. In addition, we describe a method to measure ubiquitin chain length of ubiquitylated substrates using a chain protector and limited digestion of ubiquitin chains, named Ub-ProT (ubiquitin chain

†

Co-first authors.

Methods in Enzymology, Volume 618 ISSN 0076-6879 https://doi.org/10.1016/bs.mie.2018.12.019

#

2019 Elsevier Inc. All rights reserved.

105

106

Fumiaki Ohtake et al.

protection from trypsinization). These strategies will contribute to our knowledge of polymeric ubiquitin signals and permit investigation of new mechanisms concerning the ubiquitin code.

1. Introduction Posttranslational modification of cellular proteins with ubiquitin regulates diverse aspects of biological pathways. A major feature of ubiquitylation is that ubiquitin can be further modified with additional ubiquitin molecules forming amide bonds at any of its seven lysine residues or the first methionine, yielding eight types of homogeneous polyubiquitin chains (Komander & Rape, 2012). These ubiquitin chains direct substrate proteins to different pathways. Recent studies revealed that ubiquitin chains can be heterogeneous, comprising more than one linkage type. Such heterogeneous chains include branched chains, in which one ubiquitin moiety is modified with two different linkages, and mixed chains, in which ubiquitin is modified with different linkages in tandem (Swatek & Komander, 2016; Yau & Rape, 2016). Branched ubiquitin chains regulate various pathways depending on the combination of linkage types: K11/K48 branched chains regulate mitosis when generated by the E3 ubiquitin ligase APC/C, and regulate quality control of misfolded proteins (Meyer & Rape, 2014; Yau et al., 2017); K48/K63 branched chains enhance NF-κB signaling by stabilizing K63 linkages, and regulate proteasomal degradation of K63 linkagemodified substrates (Ohtake, Saeki, Ishido, Kanno, & Tanaka, 2016; Ohtake, Tsuchiya, Saeki, & Tanaka, 2018); K29/K48 branched ubiquitin chains facilitate proteasomal degradation of UFD (ubiquitin fusion degradation) substrates in yeast (Liu, Liu, Ye, & Li, 2017). The length of ubiquitin chains is also an important determinant of the “ubiquitin code,” and is reversibly regulated by ubiquitin elongating enzymes and deubiquitinases. Early studies indicated that at least four moieties of K48-linked ubiquitin chains are required for efficient targeting to the proteasome (Thrower, Hoffman, Rechsteiner, & Pickart, 2000). Moreover, ubiquitin chain length in cells is dynamically regulated by Cdc48/p97 in yeast (Tsuchiya et al., 2018). Taken together, the higher-order architecture of ubiquitin chains, characterized by linkages, branches, and length, is a critical component of the ubiquitin code. Mass spectrometry-based detection of ubiquitin chain linkages is an essential strategy for dissecting the function of the ubiquitin code. Trypsin digestion of ubiquitin chains generates signature peptides specific to

Methods to measure ubiquitin chain length and linkage

107

particular linkage types. While linkage-specific antibodies are commercially available for K11, K48, K63, and M1 linkages, mass spectrometric analyses enable simultaneous detection of all eight linkage types and direct comparison of the stoichiometry of each linkage across samples by means of absolute quantification (Ub-AQUA) (Fig. 1A). The Ub-AQUA strategy, in which isotopically labeled signature peptides (AQUA peptides) for the eight linkage types are added to samples as internal standards for absolute quantification, was established by the Gygi, Harper, and Kopito groups (Bennett et al., 2007; Kirkpatrick et al., 2006; reviewed in Ordureau, Munch, & Harper, 2015). The Ub-AQUA method allowed for detailed analyses of molecular mechanisms involving the ubiquitin code, including the ubiquitin chain linkage specificity of E3s, DUBs, decoders, and substrates. In this chapter, we describe methods to quantify ubiquitin chain linkages, branches, and length. For quantification of ubiquitin chain linkages, we follow methods developed by previous studies (Phu et al., 2011) with slight modifications. We use parallel reaction monitoring (PRM) instead of highresolution precursor mass spectrum (MS1) or multiple reaction monitoring for quantification (Tsuchiya et al., 2017; Tsuchiya, Tanaka, & Saeki, 2013). The advantage of PRM is its high sensitivity and accuracy, as described below, and the availability of Q Exactive instruments in many laboratories (Fig. 2). To address the complexity of ubiquitin chains, we recently

Fig. 1 Signature peptides for ubiquitin chain linkages. (A) Signature peptides from wildtype ubiquitin are shown. EST, TLS, and TITLE represent unmodified ubiquitin peptides comprising amino acids 64–72, amino acids 55–63, and amino acids 12–27, respectively. (B) Signature peptides from R54A ubiquitin are shown. The isotopically labeled amino acids in AQUA peptides are shown in italics.

108

Fumiaki Ohtake et al.

Fig. 2 Comparison of selected ion monitoring (SIM) and PRM. Ub-AQUA peptides (100, 250, or 500 amol) with yeast tryptic digests (500 ng) are separated using a 90-min gradient of LC (see Fig. 6A) and analyzed by MS1 quantification using SIM or MS2 quantification using PRM. Chromatograms of the K48 linkage signature peptide are shown. The y axis shows intensity of observed ions, and x axis shows retention times. In SIM, the parental precursor ions derived from K48 signature peptides (highlighted in orange columns) are buried among noise peaks. However, in PRM, fragment ions derived from K48 signature peptides (highlighted in orange column; five most intense fragment ions are shown) are robustly observed, which enables quantification of peak areas.

developed a method to quantify K48/K63 branched ubiquitin chains (Ohtake et al., 2016) and ubiquitin chain length (Tsuchiya et al., 2018). Here, we describe detailed methods for executing these strategies and considerations important for optimizing conditions.

Methods to measure ubiquitin chain length and linkage

109

2. Ub-AQUA/PRM analysis to quantify ubiquitin linkages and K48/K63 branched chains PRM is a targeted proteomics method using a quadrupole-equipped Orbitrap instrument such as the Q Exactive (Ordureau et al., 2015; Peterson, Russell, Bailey, Westphall, & Coon, 2012). PRM yields quantitative data over a wide dynamic range from complex biological samples; it measures fragment ions (MS2) by a high-resolution Orbitrap analyzer enabling high sensitivity and accuracy (Fig. 2). We follow previously developed Ub-AQUA methods (Phu et al., 2011), but with some modifications, such as use of PRM quantification (Section 2.4), a strategy to avoid nonspecific adsorption of hydrophobic peptides (Sections 2.1 and 2.2), and a strategy to measure very hydrophilic K29 linkage peptides (Section 2.4). A limitation of conventional Ub-AQUA is its inability to measure branched linkages, with the exception of branches at neighboring lysines such as K6/K11, K11/K27, and K27/K29. We developed a strategy to measure K48/K63 branched linkages, composed of K48 and K63 linkages, which are the two most abundant linkage types in mammalian cells as well as in yeast. Our method allows for simultaneous quantification of K48/K63 branched linkages, K48 linkages not branched at K63 (unbranched K48 linkages), and K63 linkages not branched at K48 (unbranched K63 linkages) (Section 2.2) (Fig. 1B).

2.1 In-gel digestion and purification of ubiquitin peptides An essential prerequisite for accurate quantification of ubiquitin chain linkages is complete digestion of ubiquitin chains by trypsin without miscleavage. We confirmed the trypsin digestion efficiency for bulk ubiquitylated proteins in two ways. First, we performed a shotgun MS analysis of the tryptic digests of bulk ubiquitylated proteins, and found that uncleaved ubiquitin peptides were only rarely detected. We next analyzed putative uncleaved peptides at the K63 locus, including “TLSDYNIQKESTLHLVLR,” by an MS1-based quantification, and confirmed that the cleavage efficiency at the K63 locus was >99% (Tsuchiya et al., 2017). Proper purification of ubiquitin chains from cell lysates is also very important. The optimal method will differ depending on the purpose and experimental design. For example, purification of the protein of interest (POI) under denaturing conditions is necessary

110

Fumiaki Ohtake et al.

to avoid contamination by ubiquitin chains attached to proteins that associate noncovalently with the POI. These considerations are described elsewhere (Emmerich & Cohen, 2015).

2.1.1 Equipment • SDS-PAGE apparatus • Vacuum centrifuge • Shaker

2.1.2 Buffers and reagents • LC–MS grade water • LC–MS grade acetonitrile (ACN) • 50 mM ammonium bicarbonate (AMBC)/5% ACN • 50 mM AMBC/30% ACN • 50 mM AMBC/50% ACN • Sequencing grade trypsin (Promega) • 0.1% trifluoroacetic acid (TFA) • 0.1% TFA/70% ACN • 1% H2O2 • Bio-Safe Coomassie stain (Bio-Rad) • ProteoSave tubes (AMR or Sumitomo Bakelite Co., Ltd.) • ProteoSave vials (AMR)

2.1.3 Procedure 1. Separate proteins by SDS-PAGE, and stain bands with Bio-Safe Coomassie. We typically load 10 μg of protein from whole cell lysates. For immunoprecipitates and in vitro reactants, we check the abundance of ubiquitin chains by Western blotting with anti-ubiquitin antibody (P4D1 or Dako). 2. Wash the gels thoroughly in Milli-Q water, and incubate overnight at 4°C. 3. Excise gel regions corresponding to molecular weights of interest. Dice 7the gel fragments into 1-mm3 pieces, and add to 1.5 mL ProteoSave tubes. 4. Wash the gel pieces in 1 mL of 50 mM AMBC/30% ACN for >2 h with agitation, and wash again in 1 mL of 50 mM AMBC/50% ACN

Methods to measure ubiquitin chain length and linkage

5.

6.

7.

8. 9. 10.

111

for >1 h. Dehydrate the gel pieces in 200 μL of 100% ACN for 15 min, with gentle tapping every 5 min. Prepare trypsinization solution by diluting 20 ng/μL modified sequence-grade trypsin (Promega) in 50 mM AMBC/5% ACN, pH 8.0. Add the trypsinization solution to gel pieces (20 ng/μL, 15 μL), and incubate on ice for 30 min. Add another small volume of trypsinization solution to gel pieces (approximately 5 μL), and incubate at 37°C for 12–15 h. Extract the digested peptides. Add 40 μL of 0.1% TFA/70% ACN to samples, and vortex for 20 min. Transfer the extracted peptide solution to a new ProteoSave tube. Repeat the extraction three more times by adding 40 μL of 0.1% TFA/70% ACN to the gels, and combine the extracted peptide solution. Although it is optional, sonicating the sample tubes in a bath sonicator for 5 s after the third and fourth extractions might improve recovery of hydrophobic peptides. Add AQUA peptides. Thaw the AQUA stock solution, prepared as described in Section 2.3, on ice, and vortex. Do not leave the stock solution for a long time because this may cause adsorption of the peptides on tubes. Because PRM analysis is typically performed using 10–50 fmol of AQUA peptides in a 5 μL injection, add 40–200 fmol of AQUA peptides to the extracted peptide solution. Concentrate the extracted peptides to a volume <10 μL by vacuum centrifugation. Be careful not to dry the peptides completely because this may cause incomplete resuspension of hydrophobic peptides. Prepare the peptides in 20 μL of 0.1% TFA containing 0.05% H2O2, and incubate for 12 h at 4°C to oxidize methionine residues. Centrifuge at 20,000  g for 20 min to precipitate contaminants. Transfer 7 μL of supernatant to ProteoSave vial (AMR) for LC–MS analysis.

2.1.4 Notes 1. For analysis of all the ubiquitin species from whole cell lysates, we typically run the SDS-PAGE for 1 cm using 4%–20% gradient gels. Because ubiquitin species generally range from 8.7 to >250 kDa, we excise the whole gel regions stained by Coomassie. For analysis of ubiquitin chains attached to the POI, we excise high molecular weight regions, e.g., corresponding to proteins larger than 50 kDa. 2. The peptides are treated with 0.05% H2O2 for 12 h to oxidize the methionine-containing peptides. The only methionine in ubiquitin itself is M1; therefore, the only ubiquitin peptides containing methionine after

112

Fumiaki Ohtake et al.

complete trypsinization are from chains that had M1 linkages or K6 linkages or from unmodified 6-residue N-terminal peptides (cleaved after K6), as described in a previous study (Phu et al., 2011). We confirmed that >95% of methionine-containing peptides were converted to their oxidized form by MS1-based selected ion monitoring (SIM) analysis (Fig. 3). 3. Nonspecific adsorption of hydrophobic peptides, especially the K63 linkage signature peptide, to tubes and vials occurs during sample preparation (Fig. 4A). To avoid this, we compared several tubes and found that using hydrophilic tubes and vials, such as ProteoSave MS tubes and vials (AMR) or ProteoSave tubes (Sumitomo Bakelite Co., Ltd.), minimizes nonspecific adsorption.

Fig. 3 Oxidation of Met1-containing peptides. M1 or K6 linkage signature peptides were incubated with H2O2 (0.01% or 0.05%) for 12 h at 4°C (A) or with 0.05% H2O2 for 6, 12, or 18 h at 4°C (B), and analyzed by SIM. Data show intensities of precursor ions derived from M1 or K6 signature peptides, each of which is extracted from chromatograms. This shows conversion of M1 and K6 signature peptides to methionine-oxidized forms with increased H2O2 concentrations or incubation times.

Fig. 4 Nonspecific adsorption of hydrophobic peptides during sample preparation. (A) Ub-AQUA peptides (10 fmol at 2 fmol/μL) prepared in 0.1% TFA were incubated in uncoated tubes for 6 or 12 h at 4°C, and analyzed by PRM. Whereas the EST peptide is retained, the K63 linkage signature peptide is adsorbed during incubation. (B) Ub-AQUA peptides (25 fmol at 5 fmol/μL) prepared in 0.1% TFA were incubated in uncoated tubes, ProteoSave MS tubes (AMR), or ProteoSave tubes (Sumitomo Bakelite Co., Ltd.) for 24 h at 4°C and analyzed by PRM. The data show intensities of fragmental ions derived from the indicated signature peptides, each of which is extracted from chromatograms as in Fig. 2.

114

Fumiaki Ohtake et al.

2.2 In-gel digestion and purification of branched ubiquitin chain-derived peptides Ubiquitin peptides bearing K48/K63 branched linkages, detected as remnant GG peptides on both K48 and K63 in a single peptide, are generally not detected using a bottom-up proteomics strategy such as Ub-AQUA because of the trypsin-sensitive R54 residue in between these two lysines. We therefore developed a method to measure them. In our strategy, the arginine-54 residue (R54) is mutated to alanine, such that trypsin digestion of a K48/K63 branched ubiquitin linkage would generate a single signature peptide modified with double glycine (GG) at both K48 and K63 residues (Fig. 1B). Using this strategy, we successfully quantified K48/K63 branched linkages from whole cell lysates as well as immunoprecipitated substrate proteins (Ohtake et al., 2016, 2018). R54A ubiquitin (UbR54A) can be transiently or stably expressed in normal mammalian cell lines, or can replace endogenous (wild-type) ubiquitin by using shRNAs targeting endogenous ubiquitin genes and an shRNAresistant UbR54A expression vector such as the Tet-inducible rUb in the U2OS cell line (Xu, Skaug, Zeng, & Chen, 2009). For quantification of K48/K63 branched linkages from in vitro reactions, recombinant UbR54A should be used instead of wild-type ubiquitin. Our strategy allows for simultaneous detection of K48/K63 branched linkages, unbranched K48 linkages, and unbranched K63 linkages (Fig. 1B). However, the signature peptide for unbranched K63 linkages possesses an N-terminal glutamine residue, a portion of which is spontaneously converted to pyroglutamate during analysis, hampering accurate quantification. Therefore, we use glutaminyl-peptide cyclotransferase (QPCT) to enzymatically convert the N-terminal glutamine residue into pyroglutamate. We confirmed that a majority of the unbranched K63 linkage-derived signature peptides were converted (Fig. 5). 2.2.1 Buffers and reagents • Buffers and reagents listed in Section 2.1.2 • Recombinant glutaminyl-peptide cyclotransferase (QPCT) (R&D Systems) • Detergent removal column (Thermo Scientific) • Desalting column (Thermo Scientific) • C18 tips (Thermo Scientific) • GL-GC tips (GL Science) • Escherichia coli matrix (MassPREP E. coli Digest Standard, 186003196, Waters)

Methods to measure ubiquitin chain length and linkage

115

Fig. 5 Strategy to quantify unbranched K63 linkages from R54A ubiquitin chains. (A) Schematic illustration of the QPCT reaction. (B) Without QPCT reaction, approximately two-thirds of unbranched K63 linkage signature peptides are in the N-terminal glutamine form. The QPCT reaction converts a majority of the peptides into pyroglutamate form.

2.2.2 Procedure 1. Follow steps 1–8 of the procedure in Section 2.1.3. 2. Purify 1 μg of recombinant glutaminyl-peptide cyclotransferase (QPCT) from the 0.1 μg/μL frozen stock using detergent removal and desalting columns to remove Tris, NaCl, and glycerol. Elute proteins with 50 μL of 50 mM AMBC. 3. Dilute the peptides obtained from step 1 with 50 mM AMBC for neutralization to a volume of 17 μL, and confirm that the pH is approximately 7.0–8.0 using pH paper. 4. Incubate peptides with 3 ng/μL QPCT (3 μL of 20 ng/μL QPCT) in a 20 μL reaction for at least 5 h at 37°C. 5. Add 4 μL 1% TFA to the reaction to adjust the pH to <4. 6. Purify peptides using C18 tips and elute the peptides with 20 μL of 0.1% TFA/80% ACN twice. Save the flow-through for step 7. 7. Recover hydrophilic peptides in the flow-through from step 6 using GL-GC tips and elute with 40 μL of 0.1% TFA/80% ACN.

116

Fumiaki Ohtake et al.

8. Combine the eluates from steps 6 and 7, and concentrate to a volume <10 μL by vacuum centrifugation. 9. Follow steps 9–10 of the procedure in Section 2.1.3. 2.2.3 Notes 1. Previous yeast genetic studies showed that the UbR54A mutation does not affect cell growth, heat and cold sensitivity, or endocytosis (Roscoe, Thayer, Zeldovich, Fushman, & Bolon, 2013; Sloper-Mould, Jemc, Pickart, & Hicke, 2001). We confirmed that in vitro ubiquitin chain synthesis by several E2 enzymes as well as assembly of K48 and K63 chains in human 293F cells was not affected by UbR54A mutation. Nonetheless, it should be empirically tested whether the function of your POI is retained in the context of UbR54A. For example, we tested the interaction of TAB2 (NZF domain) and HUWE1 (UBA–UIM domain), along with RAP80/UIMC1 (tUIM domain) as representative ubiquitin-binding domains (UBDs), with K63-linked chains derived from either wild-type ubiquitin or UbR54A in pull-down assays (Ohtake et al., 2016). The data showed that UbR54A behaves basically like wild-type ubiquitin in these interactions. 2. Because the signature peptides for K48/63 branched linkages and unbranched K63 linkages are highly hydrophobic, sample loss due to adsorption is a major cause of low signal intensity in PRM. We recommend using ProteoSave tubes throughout the procedure (Fig. 4B). For samples with low protein quantity, adding 100–500 ng of E. coli matrix (Waters) after trypsinization may prevent adsorption of ubiquitin peptides. In addition, avoid drying peptides completely during vacuum concentration.

2.3 Preparation and quality control of AQUA peptides The properties of signature peptides, such as solubility and stability during preparation, and sensitivity in PRM measurement, are highly variable. Moreover, the lower limit of quantification may differ depending on the MS instrument. Therefore, for accurate quantification it is essential to analyze AQUA peptides and establish a standard curve before measurement of your sample. 2.3.1 Equipment • Bath sonicator

Methods to measure ubiquitin chain length and linkage

117

2.3.2 Buffers and reagents • LC–MS grade water • LC–MS grade ACN • LC–MS grade formic acid • AQUA peptides, quantification grade, 1 nmol (Sigma-Aldrich) 2.3.3 Procedure 1. For wild-type ubiquitin-derived peptides, add 200 μL of 0.1% formic acid/10% ACN to the vial containing individual AQUA peptides to yield a 5 pmol/μL stock solution. For hydrophobic peptides, such as signature peptides for K48/K63 branched linkages and for unbranched K63 linkages, add 0.1% TFA/50% ACN. 2. Vortex for 20 min, then use a bath sonicator for 5 s for complete solubilization. 3. Mix all the AQUA peptides and dilute with 0.1% formic acid/10% ACN to yield 200 fmol/μL AQUA peptide mixture. For R54A-derived signature peptides, we prepare a 1 pmol/μL mixture of AQUA peptides. 4. Prepare 10 μL aliquots of the AQUA peptide mixture, flash-freeze in liquid nitrogen, and store at 80°C. Avoid repeated freeze–thaw cycles. 5. Analyze the AQUA peptide mixture using PRM (see Section 2.4). Establish a standard curve by performing PRM analysis on 5 μL samples containing 0.05, 0.1, 1, 10, and 100 fmol of each Ub-AQUA peptide and 100 ng of E. coli matrix. 2.3.4 Notes 1. Hydrophobic peptides tend to adsorb onto the tube during handling. To test which tube better prevents adsorption of peptides, we incubated AQUA peptides (5 fmol/μL) in each tube for 24 h at 4°C before PRM analysis (Fig. 4B). We found that adsorption of peptides strongly depends on the hydrophobicity of peptides; the measured intensity of K48 linkages was not significantly different between uncoated tubes and the ProteoSave tube. However, peptides for K48/K63 branched linkages and unbranched K63 linkages are mostly lost in uncoated tubes but retained in ProteoSave tubes. 2. The Ub-AQUA/PRM method is also applicable to measure posttranslational modifications of ubiquitin. For example, we designed AQUA peptides acetylated at K6 or K48 to quantify the abundance of acetylated

118

Fumiaki Ohtake et al.

ubiquitin in mammalian cells (Ohtake et al., 2015). Similarly, AQUA peptides phosphorylated at S65 were used to quantify phosphorylated ubiquitin (Koyano et al., 2014; Ordureau et al., 2014).

2.4 PRM for ubiquitin quantification 2.4.1 Equipment • Nanoflow UHPLC such as the EASY-nLC 1000 (Thermo Fisher Scientific) • Quadrupole-equipped Orbitrap MS instrument such as the Q Exactive (Thermo Fisher Scientific) • Xcalibur software (Thermo Fisher Scientific) • Pinpoint software 1.3 (Thermo Fisher Scientific) 2.4.2 Buffers and reagents • LC–MS grade 0.1% formic acid/water (solvent A) • LC–MS grade 0.1% formic acid/ACN (solvent B) • C18 analytical columns (ReproSil-Pur 3 μm, 75 μm id  12 cm packed tip column, Nikkyo Technos Co., Ltd.) 2.4.3 Procedure 1. Set the EASY-nLC system with an 80-min three-step gradient (0%–10% solvent B in 5 min, 10%–25% in 70 min, and 25%–80% in 5 min) with a flow rate of 300 nL/min. Typical settings for quantification of normal ubiquitin peptides and branched ubiquitin chains are shown in Fig. 6A and B. 2. Set the Q Exactive instrument in targeted MS2 mode using Xcalibur software. Typical parameters are shown in Fig. 6C. 3. Prepare transition lists for time-scheduled acquisition of pairs of isotopically labeled AQUA peptides and endogenous peptides in a 10 min retention time window. Lists for ubiquitin peptides and their typical retention times are shown in Table 1. 4. Follow steps 1–10 of the procedure in Section 2.1.3, then transfer 7 μL of the supernatant to a ProteoSave vial. Run the sequence. 5. Process the data using Pinpoint software. The area under the curve (AUC) of the selected fragment ion is calculated for each point based on the coelution profiles of light- or heavy-labeled peptides. The AUCs of each individual PRM transition are then summed to obtain AUCs at the peptide level. Product ions used for peptide quantification are shown in Table 1.

Methods to measure ubiquitin chain length and linkage

119

Fig. 6 Usage settings for the EASY-nLC and Q Exactive instruments. LC settings for quantification of Ub(WT)-derived linkages (A) and Ub(R54A)-derived linkages (B) are shown. LC is run with a 90 min gradient of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). Settings for targeted MS2 of Ub(R54A) linkages are shown in (C).

2.4.4 Notes 1. Scheduled acquisition of heavy and light peptide pairs is typically performed with a retention time window of 10 min. 2. The signature peptide for K29 linkage is very hydrophilic and is therefore not efficiently trapped to a precolumn and is eluted as a broad peak in the nanoflow UHPLC system. To quantify K29 linkage, we use a one-column setting in which peptide samples are directly loaded onto the C18 analytical column. We also use 0.1% TFA instead of formic acid for sample preparation, which improves the peak shape of the K29 linkage signature peptide (Fig. 7). 3. We used the three most reliable peptides at the K63 locus (TLSDYNIQK, ESTLHLVLR, and K63 linkage) for calculation of total ubiquitin. 4. We typically performed Ub-AQUA quantification with three to five biological replicates.

Table 1 List of ubiquitin peptides used in PRM analysis Abbreviation

Peptide sequence

Precursor m/z

Charge state

Product ions used for PRM

RT (min) for Ub(WT)a

RT (min) for Ub(R54A)a

K6 linkage

M[Oxid]QIFVK[di-GlyGly]TLTGK

465.927

3

y4+, y5 +, y6+, y7+, y8 +

37.77

29.61

M[Oxid]QIFVK[GG]TL[HeavyL]TGK

468.266

TLTGK[di-GlyGly]TITLEVEPSDTIENVK

801.427

3

y8+, y9 +, y10+, y11 +, y12 +, 57.01 y13 +

45.18

TLTGK[GG]TITLEVEPSDTIENV[HeavyV]K

803.431

TITLEVEPSDTIENVK[di-GlyGly]AK

701.039

3

y6+, y7 +, y8+, y9+, y10 +, y11 +, y12 +

50.87

40.82

TITLEVEPSDTIENV[HeavyV]K[GG]AK

703.044

AK[di-GlyGly]IQDK

408.732

2

y3+, y4 +, y5+

17.08

14.26

AK[GG]I[HeavyI]QDK

412.241

IQDK[di-GlyGly]EGIPPDQQR

546.613

2

y6+, y8 +, y9+

21.76

19.03

IQDK[GG]EGIP[HeavyP]PDQQR

548.618

LIFAGK[di-GlyGly]QLEDGR

487.6

3

y4+, y5 +, y6+, y7+, y8+, y9 +

39.92

32.56

LIFAGK[GG]QL[HeavyL]EDGR

489.939

TLSDYNIQK[di-GlyGly]ESTLHLVLR

748.738

3

y5+, y6 +, y7+, y8+, y9 +, y10 +

59.84

46.31

TLSDYNIQK[GG]ESTLHLVL[HeavyL]R

751.077

GGM[Oxid]QIFVK

448.239

2

y3+, y4 +, y5+, y6+

25.69

22.81

GGM[Oxid]QIFV[HeavyV]K

451.246

TITLEVEPSDTIENVK

894.4672 2

y6+, y9 +, y10+, y11 +, y12 +, 56.01 y13 +

45.41

TITLEVEPSDTIENV[HeavyV]K

897.4741

K11 linkage

K27 linkage

K29 linkage (human)

K33 linkage

K48 linkage

K63 linkage

M1 linkage

TITLE (human)

EST

TLS

K29 linkage (yeast)

TITLE (yeast)

K48-K63 branched linkage (R54A)

unbranched K48 linkage (R54A)

unbranched K63 (R54A)

ESTLHLVLR

534.314

2

y3+, y4 +, y5+, y6+, y7 +

38.51

ESTLHLVL[HeavyL]R

537.8226

TLSDYNIQK

541.2798 2

y3+, y4 +, y5+, y6+, y7 +

26.26

TL[HeavyL]SDYNIQK

544.7884

SK[di-GlyGly]IQDK

416.73

2

y2+, y3 +, y4+, y5+

21.44

SK[GG]I[HeavyI]QDK

420.238

TITLEVESSDTIDNVK

882.449

2

y3+, y5 +, y6+, y8+, y10 +, y11 +, y12+, y14+

53.76

TITLEVEDSDTIDNV[HeavyV]K

885.456

LIFAGK[di-GlyGly]QLEDGATLSDYNIQK [di-GlyGly]ESTLHLVLR

900.9809 4

LIFAGK[di-GlyGly]QLEDGATLSDYNIQK [di-GlyGly]ESTLHLVL[HeavyL]R

902.7352

LIFAGK[di-GlyGly]QLEDGATLSDYNIQK

813.4235 3

LIFAGK[di-GlyGly]QLEDGATL[HeavyL] SDYNIQK

815.7626

Q[Pyroglutamate]LEDGATLSDYNIQK[di-GlyGly] 947.4856 3 ESTLHLVLR

31.43

y5+, y8 +, y9+, y10 +

59.93

y5+, y7 +, y8+, y9+, y11 +, y12 +

50.39

y8+, y9 +, y10+, y11 +

60.56

Q[Pyroglutamate]LEDGATLSDYNIQK[di-GlyGly] 949.8247 ESTLHLVL[HeavyL]R a Retention time for Ub(WT) and Ub(R54A) corresponds to LC gradient (A) and (B) of Fig. 6, respectively. Sequence, precursor (parental) m/z, charge state, product (fragmental) ions used for PRM, and typical retention time (RT) for light and heavy ion pairs of signature peptides are shown.

Fig. 7 Quantification of hydrophilic K29 linkages. Using 0.1% TFA instead of formic acid for sample preparation improves the peak shape of the K29 linkage signature peptide. The data show chromatograms of precursor ion from K29 signature peptide. The y axis shows intensity of ions and x axis shows retention times.

Methods to measure ubiquitin chain length and linkage

123

3. Ub-ProT analysis to measure ubiquitin chain length of in vitro and in vivo ubiquitin conjugates Despite the fundamental importance of ubiquitin chain length in ubiquitin signaling, there had been no technique to determine chain lengths in biological samples. In many cases, endogenous substrates have multiple ubiquitylation sites (Kim et al., 2011), and attached chains may have heterogeneous lengths. Therefore, one cannot determine ubiquitin chain length simply by analyzing gel mobility of specific conjugated proteins. To overcome this problem, we designed an approach using trypsin and a ubiquitin chain-binding protector named Ub-ProT (ubiquitin chain protection from trypsinization) (Fig. 8A). When polyubiquitylated proteins are subjected to trypsinization under native conditions, the substrate proteins are almost completely digested, but the polyubiquitin chains are partially digested or only cleaved at Arg74 of ubiquitin. However, in the presence of a ubiquitin chain protector such as TR-TUBE, substrate-attached chains are protected from trypsinization, and we could estimate the ubiquitin chain length. Although our strategy has some limitations, it is the first method to measure the length of ubiquitin chains attached to a substrate in vivo.

3.1 Purification of TR-TUBE for Ub-ProT analysis Tandem ubiquitin-binding entity (TUBE), which consists of four tandem repeats of the UBA domain of UBQLN1, is a valuable tool for purifying and protecting ubiquitylated proteins from cell extracts (Hjerpe et al., 2009). We use a modified TUBE to protect ubiquitin chains from trypsinization. To capture ubiquitylated proteins more efficiently and to prevent trypsinization of the TUBE itself, we mutated all arginine residues in the UBA domain to alanine and created a gene encoding four, six, or eight tandem repeats of this modified UBA domain (Tsuchiya et al., 2018; Yoshida et al., 2015). The gene encoding trypsin-resistant (TR)-TUBE was inserted into a modified pRSET-A vector, which also codes for a hexahistidine tag for purification and a cysteine residue for biotinylation at the N-terminus of TR-TUBE. 3.1.1 Equipment • MicroSpin Empty columns (GE Healthcare) • ImageQuant LAS4000 (GE Healthcare) • ImageQuant TL software (version 8.1; GE Healthcare)

124

Fumiaki Ohtake et al.

Fig. 8 Concept of Ub-ProT assay and preparation of TR-TUBE. (A) Schematic illustration of our method for determining ubiquitin chain length. When ubiquitylated proteins are trypsinized under nondenaturing conditions, the substrate proteins are almost completely digested, but ubiquitin chains are only partially digested; they are cleaved into ubiquitin monomers (upper panel). In the presence of a ubiquitin chain protector, ubiquitin chains are protected from trypsinization (lower panel). (B) Purified ubiquitin chain protector for Ub-ProT assay. TR-TUBE was purified from E. coli cells transformed with pRSET-(Cys)-TR-TUBE. A Coomassie-stained 4%–12% NuPAGE gel with approximately 2 μg protein per lane is shown.

• • •

€ Superdex 75 10/100 GL and AKTA protein purification system (GE Healthcare) French press (Ohtake Inc.) Standard LDS-NuPAGE equipment, 4%–12% NuPAGE gels

Methods to measure ubiquitin chain length and linkage

125

3.1.2 Buffers and reagents • pRSET-A (+Cys)-TR-TUBE (4, 6 , and 8  UBAs) (Addgene) • E. coli BL21 (DE3) competent cells (Agilent) • TALON metal affinity resin (Clontech) • EZ-Link maleimide-PEG2-biotin (Thermo Fisher Scientific) • LB medium (supplemented with 50 μg/mL ampicillin) • 1 M isopropyl β-D-1-thiogalactopyranoside (IPTG) • Lysis buffer: 50 mM sodium phosphate pH 7.0, 300 mM NaCl, 10% glycerol, 1 mM Tris (2-carboxyethyl)phosphine hydrochloride (TCEP) • Elution buffer: 50 mM HEPES pH 7.1, 100 mM NaCl, and 0.2 M imidazole • Biotin reagent: 2 mg EZ-Link maleimide-PEG2-biotin in 200 μL of 50 mM HEPES pH 7.1 • Bio-Safe Coomassie G-250 Stain (Bio-Rad) 3.1.3 Procedure 1. Grow a 1 L culture of E. coli BL21 (DE3) transformed with pRSETTUBE in LB medium to an OD600 of 0.5 using standard methods. 2. Cool cells to 22°C and induce with 0.2 mM IPTG for 15 h at 22°C. 3. Harvest cells by centrifugation (5000 rpm for 10 min) and store at 80°C until needed. 4. Suspend cells in 45 mL lysis buffer and lyse in a French pressure cell with a mechanically operated pump. 5. Add Triton X-100 to a concentration of 0.1% and incubate on ice for 10 min. 6. Clear the lysate by centrifuging at 29,300  g for 30 min and recover the supernatant. 7. Add 500 μL TALON resin that has been preequilibrated with lysis buffer, and rotate for 1 h at 4°C. 8. Collect TALON beads by centrifuging at 3000 rpm for 5 min and wash with 10 mL lysis buffer supplemented with 0.1% Triton X-100. 9. Transfer the beads to a MicroSpin Empty column and wash five times with 500 μL of lysis buffer supplemented with 0.1% Triton X-100. 10. Incubate with 250 μL elution buffer for 15 min at 4°C twice. 11. Analyze 2 μL of the purified protein, using bovine serum albumin as a standard, by SDS-PAGE. Stain the gel with Coomassie Brilliant Blue and scan the stained gel using an ImageQuant LAS4000 (GE Healthcare). Quantify the protein bands using ImageQuant TL software.

126

Fumiaki Ohtake et al.

12. Add TCEP to a final concentration of 1 mM and biotin reagent equivalent to a 20-fold excess and rotate overnight at 4°C. 13. Perform gel filtration on a Superdex 75 10/100 GL column preequilibrated with 50 mM HEPES pH 7.5, 100 mM NaCl, and 10% glycerol to separate purified protein from free biotin. 14. Analyze 5 μL of the purified protein, using bovine serum albumin as a standard, by SDS-PAGE. Store small aliquots of purified protein at 80°C. 3.1.4 Notes • We typically obtained approximately 1–2 mg of 4  and 6  TR-TUBE using this procedure (Fig. 8B). 8  TR-TUBE is expressed at lower levels, and the final protein amount obtained was approximately 0.5–1 mg. • We constructed TR-TUBEs with 4 , 6 , and 8  ubiquitin-binding domains. All three constructs seem to have equal binding ability, at least to unanchored M1-linked ubiquitin chains (Tsuchiya et al., 2018). We mainly used biotinylated 6  TR-TUBE, but 4  and 8  TR-TUBEs may be more useful for specific ubiquitin chain topologies.

3.2 Ub-ProT for in vitro synthesized ubiquitin chains To analyze chain length from substrates ubiquitylated in vitro, we tried two different procedures: a simple one-step Ub-ProT assay in solution and a pull-down of ubiquitylated substrate by TR-TUBE followed by trypsinization on beads. 3.2.1 Equipment • Standard LDS-NuPAGE equipment, 4%–12% NuPAGE gels 3.2.2 Buffers and reagents • K48-linked polyubiquitin chains (polyubiquitin/Ub1-Ub7 WT chains, K48-linked, UC-240, Boston Biochem) • K63-linked polyubiquitin chains (polyubiquitin/Ub1-Ub7 WT chains, K63-linked, UC-340, Boston Biochem) • M1-linked polyubiquitin chains, prepared using petite-LUBAC as described previously (Sakamoto et al., 2015) • Biotinylated 6  TR-TUBE • Phosphate-buffered saline (PBS) with 0.05% Tween 20 (PBS-T) • Dynabeads MyOne Streptavidin C1 (Life Technologies)

Methods to measure ubiquitin chain length and linkage

• • • • •

127

Ammonium bicarbonate (AMBC) A MS compatible detergent: RapiGest SF (Waters) Trypsin Gold (Promega) Trypsin solution: 50 mM AMBC, 0.01% RapiGest SF, 2.5 ng/μL Trypsin Gold 1  NuPAGE LDS sample buffer (Thermo Fisher Scientific), with 5% 2-mercaptoethanol

3.2.3 Procedure Ub-ProT assay in solution 1. Mix 100 ng of polyubiquitin chains (M1-, K48-, or K63-linked Ub chains) and 5 μg of TR-TUBE in 50 mM AMBC. 2. Rotate for 1 h at 4°C. 3. Add RapiGest SF to a concentration of 0.01%. 4. Add 100 ng Trypsin Gold (final sample volume: 20 μL). 5. Rotate overnight at 37°C. 6. Quench the reaction with 1 LDS sample buffer. Ub-ProT assay on beads 1. Mix 100 ng of polyubiquitin chains and 5 μg of TR-TUBE in buffer A containing 0.1% Triton X-100. 2. Rotate for 30 min at 4°C. 3. Add 1 mg of Dynabeads MyOne Streptavidin C1. 4. Rotate for 45 min at 4°C. 5. Wash with PBS-T three times. 6. Add 20 μL of trypsin solution. 7. Rotate overnight at 37°C. 8. Wash the beads with PBS-T twice. 9. Elute the TR-TUBE-protected polyubiquitin chains by incubating with 1  NuPAGE LDS sample buffer for 30 min at 37°C. 10. Subject the eluted product and free polyubiquitin chains to electrophoresis on a NuPAGE 12% gel in MES buffer and immunoblot using an anti-ubiquitin antibody. 3.2.4 Notes 1. Both procedures can protect polyubiquitin chains from trypsinization (Fig. 9A). The Ub-ProT assay on beads yields stable measured products and has good reproducibility. 2. In order for the TR-TUBE to effectively protect polyubiquitin chains from trypsinization, it is necessary to minimize the amount of trypsin

128

Fumiaki Ohtake et al.

Fig. 9 Ub-ProT assay of free ubiquitin chains and yeast lysate. (A) Ub-ProT assay for unanchored polyubiquitin chains. 100 ng of polyubiquitin chains were subjected to the Ub-ProT assay (procedures described in Section 3.2). Ubiquitin was detected by immunoblotting with a monoclonal ubiquitin antibody (P4D1). The number of ubiquitin molecules in the chains is labeled at the right of each panel. (B) Ub-ProT assay of yeast lysate. S. cerevisiae cells were harvested and soluble fractions were prepared. Ubiquitylated proteins were captured by TR-TUBE and trypsinized. Samples were separated by SDS-PAGE on 12% NuPAGE gels and immunoblotted with anti-ubiquitin antibody. Free ubiquitin chains were used as markers.

used. Also, the trypsin sensitivity of proteins varies with their structural properties. Therefore, the amount of trypsin should be titrated in each experimental setup. We found that 50 ng of trypsin was required for complete cleavage of 500 ng polyubiquitin chains (Tsuchiya et al., 2018). 3. Elute samples without boiling in order to avoid aggregation of polyubiquitin chains. 4. Because monomeric ubiquitin is more resistant to trypsin than chains, ubiquitin monomers were occasionally observed after trypsinization in some controls, but this did not affect our analysis of chain length or composition. Thus, incomplete digestion is unlikely to be a source of error in our experiments. 5. Ub-ProT may slightly underestimate the occurrence of certain linkage types; TR-TUBE can bind and protect all ubiquitin linkage types, but the degree of protection varies up to twofold. In a Ub-ProT assay using di-ubiquitin, we confirmed that K11-, K48-, K63-, and M1-linked chains were almost completely protected from trypsinization (approximately 90%), but K6-, K27-, K29-, and K33-linked ubiquitins were only 40%–60% protected (Tsuchiya et al., 2018). We also found that heterogeneous K48/K63 linkages can be protected from trypsinization by Ub-ProT, although there is a decrease in protection efficiency.

Methods to measure ubiquitin chain length and linkage

129

3.3 Ub-ProT analysis of ubiquitin conjugates from Saccharomyces cerevisiae Earlier in vitro studies showed that tetra-ubiquitin is the minimal recognition signal for proteasomal degradation of folded proteins (Thrower et al., 2000). However, a recent study suggested that monoubiquitylation, multiple short chains with different linkages, and branched chains can also induce proteasomal degradation (Lu, Lee, King, Finley, & Kirschner, 2015; Martinez-Fonts & Matouschek, 2016). Thus, the length of ubiquitin chains is a key parameter of ubiquitination, but it has not been carefully examined in vivo. Here, we describe the Ub-ProT method using soluble fractions obtained from S. cerevisiae. 3.3.1 Equipment • Screw cap micro tubes, 2 mL (Sarstedt) • Dynabeads MyOne Streptavidin C1 (Life Technologies) • Glass beads (0.5 mm diameter) • Multi-Beads Shocker (Yasui Kikai Co.) • Standard LDS-NuPAGE equipment, 12% NuPAGE gels 3.3.2 Buffers and reagents • S. cerevisiae strain W303 • YPD medium: 1% yeast extract, 2% peptone, and 2% glucose • Buffer A: 50 mM Tris–HCl pH 7.5, 100 mM NaCl, 10% glycerol • Lysis buffer: buffer A supplemented with 10 μM MG132, 10 mM iodoacetamide, and 1 complete protease inhibitor cocktail (EDTA free) (Roche) • Pierce BCA Protein Assay kit (Thermo Fisher Scientific) • PBS-T: PBS with 0.05% Tween 20 • Dynabeads MyOne Streptavidin C1 (Life Technologies) • Trypsin solution: 50 mM AMBC, 0.01% RapiGest SF (Waters), and 15 ng/μL Trypsin Gold (Promega) • 1  NuPAGE LDS sample buffer: NuPAGE LDS sample buffer (Thermo Fisher Scientific) with 5% 2-mercaptoethanol • Ubiquitin antibody (P4D1), HRP-conjugated (Santa Cruz Biotechnology) 3.3.3 Procedure 1. Grow a 50 mL culture of S. cerevisiae in YPD to log-phase ( 0.8–1.0 OD600). 2. Harvest 30 OD600 units by centrifuging at 3000 rpm for 10 min in 50 mL tubes.

130

Fumiaki Ohtake et al.

3. Suspend the cells with milli-Q water and transfer to 2 mL screw cap tubes. 4. Suspend cells in 300 μL of lysis buffer, and add glass beads. 5. Lyse cells by subjecting them to six cycles of vortexing for 30 s at 2600 rpm and chilling for 30 s at 4°C in a Multi-Beads Shocker. 6. Remove glass beads by piercing the bottom of the tube with an 18G needle, and collect the lysate in a new tube by centrifuging at 860  g for 3 min. 7. Clear the lysate by centrifuging at 20,000  g for 10 min and transferring the supernatant to a new 1.5 mL tube (repeat three times). 8. Determine the protein concentration using a BCA assay kit. 9. Incubate 100–500 μg of protein with 10 μg of biotinylated TR-TUBE for 1 h at 4°C. 10. Add 1 mg of Dynabeads and incubate for 45 min at 4°C. 11. Wash the beads with PBS-T three times. 12. Wash the beads with 50 mM AMBC twice and add 100 μL of trypsin solution, and incubate overnight at 37°C with rotator. 13. Wash the beads with PBS-T twice. 14. Elute the TR-TUBE-protected polyubiquitin chains by incubating with 1  NuPAGE LDS sample buffer for 30 min at 37°C. 15. Subject the eluted product and free polyubiquitin chains (described in Section 3.2) to electrophoresis on NuPAGE 12% gels in MES buffer and immunoblot with anti-ubiquitin antibody (Fig. 9B). 3.3.4 Notes 1. We found that streptavidin was not digested by trypsin under the described conditions, so the polyubiquitin chains were retained on the beads via the TR-TUBE/streptavidin complex after trypsinization. 2. Because ubiquitin chains with different linkage types exhibit different gel mobilities, free ubiquitin chain ladders can be used to estimate the lengths of substrate-cleaved ubiquitin chains. 3. By combining Ub-ProT analysis with Ub-AQUA/PRM, we can determine ubiquitin chain lengths and composition.

4. Summary and conclusions Ub-AQUA is a powerful strategy that enables quantification of ubiquitin linkages, and has been utilized by many researchers to dissect ubiquitin signaling. Complementary methods such as immunoblotting

Methods to measure ubiquitin chain length and linkage

131

and immunocytochemistry provide useful information such as mobility shifts of substrates and cellular localization, respectively, and are enabled by the use of linkage-specific antibodies available for K11, K48, K63, and M1 chains (Newton et al., 2008). The methods described here also allow for measurement of ubiquitin chain length and branching at K48/K63, providing insight into the complexities of ubiquitin chains. One caveat of the Ub-AQUA technique is that it focuses on the composition of linkages and is not suitable for investigating the higher-order architecture of ubiquitin chains. Middle-down mass spectrometric analysis of ubiquitin chains is being developed to measure branching of ubiquitin chains (Valkevich, Sanchez, Ge, & Strieter, 2014). Further technical advances, such as a top-down mass spectrometry-based strategies, are needed for elucidating ubiquitin chain architecture. In addition, ubiquitin chain cleavage using linkage-specific deubiquitinases (UbiCRest) (Mevissen et al., 2013) and heterotypic linkage-specific antibodies (Yau et al., 2017) can be used for analysis of heterotypic ubiquitin chains. Combined use of these complementary methods will be important for increasing our understanding of the ubiquitin code.

References Bennett, E. J., Shaler, T. A., Woodman, B., Ryu, K. Y., Zaitseva, T. S., Becker, C. H., et al. (2007). Global changes to the ubiquitin system in Huntington’s disease. Nature, 448(7154), 704–708. https://doi.org/10.1038/nature06022. Emmerich, C. H., & Cohen, P. (2015). Optimising methods for the preservation, capture and identification of ubiquitin chains and ubiquitylated proteins by immunoblotting. Biochemical and Biophysical Research Communications, 466(1), 1–14. https://doi.org/ 10.1016/j.bbrc.2015.08.109. Hjerpe, R., Aillet, F., Lopitz-Otsoa, F., Lang, V., England, P., & Rodriguez, M. S. (2009). Efficient protection and isolation of ubiquitylated proteins using tandem ubiquitinbinding entities. EMBO Reports, 10(11), 1250–1258. https://doi.org/10.1038/embor. 2009.192. Kim, W., Bennett, E. J., Huttlin, E. L., Guo, A., Li, J., Possemato, A., et al. (2011). Systematic and quantitative assessment of the ubiquitin-modified proteome. Molecular Cell, 44(2), 325–340. https://doi.org/10.1016/j.molcel.2011.08.025. Kirkpatrick, D. S., Hathaway, N. A., Hanna, J., Elsasser, S., Rush, J., Finley, D., et al. (2006). Quantitative analysis of in vitro ubiquitinated cyclin B1 reveals complex chain topology. Nature Cell Biology, 8(7), 700–710. https://doi.org/10.1038/ncb1436. Komander, D., & Rape, M. (2012). The ubiquitin code. Annual Review of Biochemistry, 81, 203–229. https://doi.org/10.1146/annurev-biochem-060310-170328. Koyano, F., Okatsu, K., Kosako, H., Tamura, Y., Go, E., Kimura, M., et al. (2014). Ubiquitin is phosphorylated by PINK1 to activate parkin. Nature, 510, 162–166. https://doi.org/10.1038/nature13392. Liu, C., Liu, W., Ye, Y., & Li, W. (2017). Ufd2p synthesizes branched ubiquitin chains to promote the degradation of substrates modified with atypical chains. Nature Communications, 8, 14274. https://doi.org/10.1038/ncomms14274.

132

Fumiaki Ohtake et al.

Lu, Y., Lee, B. H., King, R. W., Finley, D., & Kirschner, M. W. (2015). Substrate degradation by the proteasome: A single-molecule kinetic analysis. Science, 348(6231), 1250834. https://doi.org/10.1126/science.1250834. Martinez-Fonts, K., & Matouschek, A. (2016). A rapid and versatile method for generating proteins with defined ubiquitin chains. Biochemistry, 55(12), 1898–1908. https://doi.org/ 10.1021/acs.biochem.5b01310. Mevissen, T. E., Hospenthal, M. K., Geurink, P. P., Elliott, P. R., Akutsu, M., Arnaudo, N., et al. (2013). OTU deubiquitinases reveal mechanisms of linkage specificity and enable ubiquitin chain restriction analysis. Cell, 154(1), 169–184. https://doi.org/10.1016/j. cell.2013.05.046. Meyer, H. J., & Rape, M. (2014). Enhanced protein degradation by branched ubiquitin chains. Cell, 157(4), 910–921. https://doi.org/10.1016/j.cell.2014.03.037. Newton, K., Matsumoto, M. L., Wertz, I. E., Kirkpatrick, D. S., Lill, J. R., Tan, J., et al. (2008). Ubiquitin chain editing revealed by polyubiquitin linkage-specific antibodies. Cell, 134(4), 668–678. https://doi.org/10.1016/j.cell.2008.07.039. Ohtake, F., Saeki, Y., Ishido, S., Kanno, J., & Tanaka, K. (2016). The K48-K63 branched ubiquitin chain regulates NF-kB signaling. Molecular Cell, 64(2), 251–266. Ohtake, F., Saeki, Y., Sakamoto, K., Ohtake, K., Nishikawa, H., Tsuchiya, H., et al. (2015). Ubiquitin acetylation inhibits polyubiquitin chain elongation. EMBO Reports, 16(2), 192–201. https://doi.org/10.15252/embr.201439152. Ohtake, F., Tsuchiya, H., Saeki, Y., & Tanaka, K. (2018). K63 ubiquitylation triggers proteasomal degradation by seeding branched ubiquitin chains. Proceedings of the National Academy of Sciences of the United States of America, 115(7), E1401–E1408. https://doi.org/ 10.1073/pnas.1716673115. Ordureau, A., Munch, C., & Harper, J. W. (2015). Quantifying ubiquitin signaling. Molecular Cell, 58(4), 660–676. https://doi.org/10.1016/j.molcel.2015.02.020. Ordureau, A., Sarraf, S. A., Duda, D. M., Heo, J. M., Jedrychowski, M. P., Sviderskiy, V. O., et al. (2014). Quantitative proteomics reveal a feedforward mechanism for mitochondrial PARKIN translocation and ubiquitin chain synthesis. Molecular Cell, 56(3), 360–375. https://doi.org/10.1016/j.molcel.2014.09.007. Peterson, A. C., Russell, J. D., Bailey, D. J., Westphall, M. S., & Coon, J. J. (2012). Parallel reaction monitoring for high resolution and high mass accuracy quantitative, targeted proteomics. Molecular & Cellular Proteomics, 11(11), 1475–1488. https://doi.org/ 10.1074/mcp.O112.020131. Phu, L., Izrael-Tomasevic, A., Matsumoto, M. L., Bustos, D., Dynek, J. N., Fedorova, A. V., et al. (2011). Improved quantitative mass spectrometry methods for characterizing complex ubiquitin signals. Molecular & Cellular Proteomics, 10(5). https://doi.org/10.1074/ mcp.M110.003756M110.003756. Roscoe, B. P., Thayer, K. M., Zeldovich, K. B., Fushman, D., & Bolon, D. N. (2013). Analyses of the effects of all ubiquitin point mutants on yeast growth rate. Journal of Molecular Biology, 425(8), 1363–1377. https://doi.org/10.1016/j.jmb.2013.01.032. Sakamoto, H., Egashira, S., Saito, N., Kirisako, T., Miller, S., Sasaki, Y., et al. (2015). Gliotoxin suppresses NF-kappaB activation by selectively inhibiting linear ubiquitin chain assembly complex (LUBAC). ACS Chemical Biology, 10(3), 675–681. https:// doi.org/10.1021/cb500653y. Sloper-Mould, K. E., Jemc, J. C., Pickart, C. M., & Hicke, L. (2001). Distinct functional surface regions on ubiquitin. The Journal of Biological Chemistry, 276(32), 30483–30489. https://doi.org/10.1074/jbc.M103248200. Swatek, K. N., & Komander, D. (2016). Ubiquitin modifications. Cell Research, 26(4), 399–422. https://doi.org/10.1038/cr.2016.39. Thrower, J. S., Hoffman, L., Rechsteiner, M., & Pickart, C. M. (2000). Recognition of the polyubiquitin proteolytic signal. The EMBO Journal, 19(1), 94–102. https://doi.org/ 10.1093/emboj/19.1.94.

Methods to measure ubiquitin chain length and linkage

133

Tsuchiya, H., Burana, D., Ohtake, F., Arai, N., Kaiho, A., Komada, M., et al. (2018). Ub-ProT reveals global length and composition of protein ubiquitylation in cells. Nature Communications, 9(1), 524. https://doi.org/10.1038/s41467-018-02869-x. Tsuchiya, H., Ohtake, F., Arai, N., Kaiho, A., Yasuda, S., Tanaka, K., et al. (2017). In vivo ubiquitin linkage-type analysis reveals that the Cdc48-Rad23/Dsk2 axis contributes to K48-linked chain specificity of the proteasome. Molecular Cell, 66(4). https://doi.org/ 10.1016/j.molcel.2017.04.024488–502.e7. Tsuchiya, H., Tanaka, K., & Saeki, Y. (2013). The parallel reaction monitoring method contributes to a highly sensitive polyubiquitin chain quantification. Biochemical and Biophysical Research Communications, 436(2), 223–229. https://doi.org/10.1016/j.bbrc.2013. 05.080. Valkevich, E. M., Sanchez, N. A., Ge, Y., & Strieter, E. R. (2014). Middle-down mass spectrometry enables characterization of branched ubiquitin chains. Biochemistry, 53(30), 4979–4989. https://doi.org/10.1021/bi5006305. Xu, M., Skaug, B., Zeng, W., & Chen, Z. J. (2009). A ubiquitin replacement strategy in human cells reveals distinct mechanisms of IKK activation by TNFalpha and IL-1beta. Molecular Cell, 36(2), 302–314. https://doi.org/10.1016/j.molcel.2009. 10.002. Yau, R. G., Doerner, K., Castellanos, E. R., Haakonsen, D. L., Werner, A., Wang, N., et al. (2017). Assembly and function of heterotypic ubiquitin chains in cell-cycle and protein quality control. Cell, 171(4), 918–933.e20. https://doi.org/10.1016/j.cell.2017.09.040. Yau, R., & Rape, M. (2016). The increasing complexity of the ubiquitin code. Nature Cell Biology, 18(6), 579–586. https://doi.org/10.1038/ncb3358. Yoshida, Y., Saeki, Y., Murakami, A., Kawawaki, J., Tsuchiya, H., Yoshihara, H., et al. (2015). A comprehensive method for detecting ubiquitinated substrates using TR-TUBE. Proceedings of the National Academy of Sciences of the United States of America, 112(15), 4630–4635. https://doi.org/10.1073/pnas.1422313112.