Strategies in Relative and Absolute Quantitative Mass Spectrometry Based Proteomics

Home , Isotopic labeling, Phosphoproteomics, Quantitative proteomics, Tandem mass tag

Biol. Chem. 2017; 398(5-6): 687–699

Review Open Access

Claudia Lindemanna, Nikolas Thomaneka, Franziska Hundt, Thilo Lerari, Helmut E. Meyer, Dirk Wolters and Katrin Marcus* Strategies in relative and absolute quantitative mass spectrometry based proteomics

DOI 10.1515/hsz-2017-0104 Received January 18, 2017; accepted February 28, 2017; previously Introduction published online March 6, 2017 Liquid chromatography-mass spectrometry (LC-MS)- Abstract: Quantitative mass spectrometry approaches are based proteome quantification is an approach often used used for absolute and relative quantification in global pro- to study biological processes. Continuously, different teome studies. To date, relative and absolute quantifica- quantitative techniques coupled to LC-MS are developed tion techniques are available that differ in quantification that can address questions including protein-protein accuracy, proteome coverage, complexity and robustness. interactions (PPI) studies, protein abundances or post- This review focuses on most common relative or absolute translational modifications between two or more physi- quantification strategies exemplified by three experimen- ological states (e.g. treated/non-treated, healthy/disease). tal studies. A label-free relative quantification approach Common techniques can be generally summarized in was performed for the investigation of the membrane pro- relative quantification for comparing whole proteomes teome of sensory cilia to the depth of olfactory receptors or the relative amount of a high number of proteins or in in Mus musculus. A SILAC-based relative quantification absolute quantification to determine the absolute concen- approach was successfully applied for the identification tration of distinct proteins within a sample (see Figure 1) of core components and transient interactors of the per- (Bantscheff et al., 2012; Marcus, 2012). Both protein quan- oxisomal importomer in Saccharomyces cerevisiae. Fur- tification approaches uses either chemical or metabolic thermore, AQUA using stable isotopes was exemplified stable isotope labeling for protein quantification (label- to unraveling the prenylome influenced by novel pre- based) or quantification is performed without the intro- nyltransferase inhibitors. Characteristic enrichment and duction of stable isotopes (label-free). fragmentation strategies for a robust quantification of the Label-free quantification is a cost-efficient and easy prenylome are also summarized. handling relative quantification technique, where the number of measured spectra is determined (spectral count- Keywords: AQUA; label-free; LC-MS/MS; prenylation; pro- ing) or the average MS/MS intensities are analyzed (Liu tein-protein interactions; SILAC. et al., 2004; Asara et al., 2008). The biggest advantage of label-free quantification in comparison to other relative quantification approaches is the possibility to measure and compare an unlimited number of samples and the analysis of untreated proteins or peptides. However, in contrast to other technical labeling techniques, variations are higher, because samples are individually prepared and compari- aClaudia Lindemann and Nikolas Thomanek: These authors son occurs only during data analysis (Bantscheff et al., contributed equally to this article. *Corresponding author: Katrin Marcus, Ruhr-University Bochum, 2012). Label-based approaches granted more accurate Medizinisches Proteom-Center, Universitätsstraße 150, D-44801 relative protein quantification, including metabolic, chemi- Bochum, Germany, e-mail: [email protected] cal and enzymatic labeling. Stable isotope labeling with Claudia Lindemann, Nikolas Thomanek, Thilo Lerari and Helmut E. amino acids in cell culture (SILAC) and heavy 15N-labeling Meyer: Ruhr-University Bochum, Medizinisches Proteom-Center, are two of the best studied metabolic stable isotope-based Universitätsstraße 150, D-44801 Bochum, Germany Franziska Hundt and Dirk Wolters: Ruhr-University Bochum, relative quantification methods so far. Labeling of single Biomolecular Mass Spectrometry, Universitätsstraße 150, D-44801 amino acids like arginine and lysine or the replacement of Bochum, Germany all nitrogens for their heavy 15N variant during cell growth

©2017, Claudia Lindemann et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. 688 C. Lindemann et al.: Strategies in quantitative proteomics

concentration. Using this costly technique, the concentration of only a few peptides of interest can be determined and it is unsuited for large interactome or global proteome studies (Bantscheff et al., 2012). The review focus on most common relative or absolute quantification strategies exemplified by three different experimental examples: label-free relative quantification, investigation of the membrane proteome of sensory cilia to the depth of olfactory receptors in Mus musculus; SILAC-based relative quantification, identification of core components and transient interactors of the peroxisomal importomer in Saccharomyces cerevisiae (S. cerevisiae); and AQUA-based absolute quantification, analysis to unravel the prenylome influenced by e.g. novel prenyltransferase inhibitors. The subsequent parts will give background information about relative and absolute quantification techniques focused on label-free, SILAC-based and AQUA-based Figure 1: Common relative and absolute protein quantification quantification strategies. workflows. The earlier sample labeling and mixing is performed, the smaller is the technical variance. Time points of introducing the isotopic- labeled standard and/or mixing the sample is illustrated with a Label-free quantification gray circle. Common relative quantification labeling techniques are metabolic, chemical or label-free. When using metabolic labeling, Label-free relative quantification presents a high-through- cell cultures (white and gray) are differently isotopic-labeled during put method in the field of quantitative proteomics. Sample and mixed after cell cultivation. Chemical labeling and mixing of the preparation is managed without the application of labeling samples is performed on protein or peptide level after proteolytic reagents, which makes such approaches simple and cost effi- digestion. Samples are prepared and measured separately in a label-free approach and quantification occurs during data analysis. cient. Relative label-free methods can roughly be subdivided In an absolute label-based quantification approach, isotopic- into counting the number of peptide-to-spectrum matches labeled standard proteins or peptides with a known concentration (PSMs; spectral counting) obtained for each protein, as are added to the sample before LC-MS/MS and data analysis. No more abundant proteins are more likely to be observed in sample mixing takes place. LC-MS/MS, liquid chromatography peptide spectra (Washburn et al., 2001; Maerkens et al., tandem mass spectrometry. Revised from Bantscheff et al. (2012). 2013) or measurement of peptide signal intensities by using the extracted ion chromatogram (XIC) (Wiener et al., 2004; reduce variances and ensure a precise quantification accu- Megger et al., 2013a,b). Spectral counting is based on deter- racy as cells are labeled in a very early step in the work- mination and comparison of the number of tandem (MS/MS) flow. However, a limited number of different samples can fragment-ion spectra of peptides between different samples be compared and metabolic labeling can cause a growth without the consideration of physicochemical peptide prop- defect in some organisms (Crotty et al., 2011; Filiou et al., erties. This approach can easily be used for comparison of 2012). An alternative to this is chemical-based relative quan- a large number of datasets and therefore has gained popu- tification such as the isobaric tags for relative and absolute larity over the last years. Proteomic tools such as absolute quantitation (iTRAQ) or the tandem mass tags (TMT) that protein expression (APEX) and normalized spectral abun- are fused to proteins or peptides after proteolytic digestion. dance factor (NSAF) also incorporate the length of the corre- In comparison to metabolic labeling, chemical labeling is sponding protein for absolute quantification (Megger et al., more cost-efficient and suitable for all sorts of proteins or 2013a,b). The second label-free technique is based on the peptides, however, the variances in the samples are higher, measurement of MS-signal intensity of the area under the as sample mixing is performed at a later step in the sample chromatographic peak of the peptide precursor ion. In this preparation workflow (Thompson et al., 2003; Wiese et al., approach peptide quantification is typically accomplished 2007; Bantscheff et al., 2012). A powerful method to deter- by integration of ion intensities of any given peptide over mine the exact protein or peptide concentration is the its chromatographic elution profile. In differential studies usage of isotopic-labeled absolute quantification (AQUA) the integrated signal response of individual peptides is peptides, which are spiked into the sample in a known compared between LC-MS(/MS) runs of different samples. C. Lindemann et al.: Strategies in quantitative proteomics 689

Subsequently, estimation of differential protein abundance in vitro between the reagent and the peptides of interest is performed by an aggregation of differences measured for leads to a heavy-labeled peptide mixture. After LC-MS/ all (unique) peptides matching the respective protein. A MS a mass shift between two different isotopic-labeled step forward in the field of label-free proteomics presents samples is used for relative quantification. Chemical labe- MS/MS-based quantification methods like selected reaction ling is possible for a wide range of different sample types, monitoring (SRM). SRM includes targeted data dependent however, ideal labeling conditions are crucial and require acquisition (DDA) with a predefined parent mass list, which optimization. presents information about peptide fragmentation (transi- In contrast to chemical labeling, SILAC involve meta- tion signals) from compounds of interest. SRM is performed bolic labeling of proteins during their synthesis in vivo on triple quadrupole instruments, where the parent ions of (Ong et al., 2002). This technique minimizes variances interest are isolated in the first quadrupole, fragmented in between different labeled samples, since they are com- the second and the transition signals of interest are analyzed bined in an early step of the workflow after cell growth. in the third. The usage of modern high-resolution second This results in increased reproducibility and accuracy mass analyzers significantly increases the sensitivity of SRM compared to chemical or label-free methods (Ong et al., methods (Dillen et al., 2012). A further alternative label-free 2003a,b; Lau et al., 2014). The simple and robust SILAC method based on data independent acquisition (DIA) is strategy can be applied to various biological questions. sequential windowed acquisition of all theoretical fragment SILAC is mostly carried out by using arginine and lysine ion mass spectra (SWATH-MS) which is based on the analysis with stable isotopes of 2H, 13C and 15N. Additionally, amino of window of masses and stepping this window across the acids as methionine, histidine or leucine can be used for complete mass range of interest (Gillet et al., 2012). Label- SILAC as well (Beynon and Pratt, 2005). A typical SILAC free approaches do not support multiplexing of samples and experiment is performed with two biological populations every sample is handled and measured separately, which or cell cultures growing either in medium with unlabeled negatively affects the reproducibility of label-free approaches arginine/lysine (light) or in medium supplemented with and requires highly reproducible workflows and an appro- isotopic labeled arginine/lysine (heavy). The cultures are priate normalization. Moreover, variations of the retention mixed, proteins are extracted and proteolytic digested time or the m/z values have to be considered. Therefore, it to peptides prior to the measurement by LC-MS/MS. In is really important to reduce technical variances in all steps bottom-up proteomics, tryptic digestion results in pep- of the sample preparation and during LC-MS measurement. tides with C-terminal arginine and lysine. This ensures Furthermore, the impact of sample heterogeneity and con- a heavy-labeling of nearly all peptides. A heavy-labeled taminants in samples has to be considered (Braakman et al., peptide produces a specific mass shift compared to its 2012; Liu et al., 2012). For more details about pros and cons of light counterpart visible as SILAC peptide pairs in the mass label-free quantification, see Marcus (2012). Recently, label- spectrum. The calculation of the heavy to light-ratio per- free quantitative proteomics is increasingly used for global formed by software tools like MaxQuant, gives an evidence interaction studies. One example for protein-protein inter- in changes of protein abundance (Cox and Mann, 2008). action mapping is the label-free based quantitative analysis An accurate quantification requires a complete incorpora- of the membrane proteome of sensory cilia to the depth of tion of the heavy amino acids. Incomplete labeling can be olfactory receptors by (Kuhlmann et al., 2014). prevented by using organisms that are unable to synthe- size the employed heavy amino acids. Furthermore, the 13 13 conversion of C6-arginine to C5-proline has to be taken Label-based relative quantification into account and can be prevented by reducing the amount (focused on SILAC) of arginine and/or increasing the amount of proline (Ong et al., 2003a,b). In general, SILAC was successfully estab- Label-based approaches like chemical or metabolic labe- lished for mammalian cell culture (Ong et al., 2002), differ- ling allow a multiplexing of samples resulting in high ent model organisms such as mouse (Kruger et al., 2008), quantification accuracy and precision. Common chemi- fly (Sury et al., 2010) and worm (Fredens et al., 2011) as cal labeling regencies like the isotope-coded affinity tag well as cultured yeast (Oeljeklaus et al., 2014), bacteria (ICAT), isobaric tags for relative and absolute quantifica- Bacillus subtilis (Soufi et al., 2010) and plants (Gruhler tion (iTRAQ), tandem mass tag (TMT) or dimethyl labeling et al., 2005). The following SILAC-strategies have been present postharvest labeling methods (Gygi et al., 1999; described so far: dynamic SILAC (Amanchy et al., 2005), Hsu et al., 2003; Thompson et al., 2003; Wiese et al., 2007; heavy methyl SILAC (Ong et al., 2004), pulsed-SILAC Lindemann et al., 2013). Basically, a chemical reaction (Schwanhausser et al., 2009), super-SILAC (Geiger et al., 690 C. Lindemann et al.: Strategies in quantitative proteomics

2010), spike-in SILAC (Geiger et al., 2011), triple-SILAC no standard peptides. Methods like emPAI is based on a (Bose et al., 2006) and affinity purification (AP)-MS SILAC so-called protein abundant index (PAI), calculating the (Blagoev et al., 2003; Oeljeklaus et al., 2012). SILAC allows observed peptides divided by the theoretically observed multiplexing from 2-plex up to 5-plex when using five dis- peptides (Ishihama et al., 2005). Another method is abso- tinct isotopic variants of arginine (Molina et al., 2009). A lute protein expression (APEX), which requires spectral step forward was the recently developed neutron encoding counting. Thereby, the number of observed divided by the (NeuCode) method, which combines SILAC and isobaric number of expected spectra is calculated including correc- tagging methods and was applied in a study of time- tion factors, which are based on prior MS-measurements resolved response of five signaling mutants in a single (Lu et al., 2007). Top3 is an additional method calculating 18-plex experiment (Merrill et al., 2014). With the devel- the relation of the three most abundant peptides between opment of super-SILAC (Geiger et al., 2010) and spike-in two or more samples. This is especially helpful when the SILAC (Geiger et al., 2011), SILAC can now be used for the protein of interest is digested incompletely, however, it quantification of samples from tissue or body fluids and requires a standard protein as reference (Silva et al., 2006). therewith has the potential to be applied for clinical appli- All these quantification techniques have a lower accuracy cations. Furthermore, false positive identification of regu- compared to quantification methods like AQUA, however, lated or interacting proteins in SILAC-based experiments they are cost and time efficient, as no heavy-labeled can also be avoided by the AP-MS SILAC technique, which standards are needed and comparison with an unlimited was successfully implemented by Oeljeklaus et al. (2012). number of different samples is ensured. AQUA was successfully used by Wotske et al. (2012).

Absolute quantification (focused on AQUA peptides) Relative and absolute quantification The most common label-based absolute quantification strategies method introduces heavy-labeled, so-called AQUA peptides, which serve as an internal standard on peptide level. Label-free relative quantification strategy for Here, synthetic peptides with a known concentration are added to the cell lysates during protein digestion. Similar the investigation of the membrane proteome to SILAC, stable isotope labeled amino acids containing 13C of sensory cilia to the depth of olfactory and 15N are used generating a predefined mass shift. The receptors in Mus musculus exact concentration of the peptide of interest is calculated using the intensity of the heavy-labeled AQUA standard The membrane of sensory cilia contains olfactory recep- peptide (Gerber et al., 2003). Although AQUA offers an tors (ORs) that extend from primary olfactory sensory accurate absolute quantification, some disadvantages neurons (OSNs), which covers the olfactory epithelium have to be considered using this technique. Synthesizing (OE) in the nasal cavity in vertebrates (McEwen et al., AQUA peptides is an expensive procedure only suitable for 2008). ORs belongs to G protein-coupled receptors binding the quantification of a few peptides of interest. Moreover, odorant and trigger a signal transduction cascade (Breer, prior information about the peptide, such as elution time, 2003). The signal transduction pathway is well estab- m/z ratio, charge state and the best fragmentation method, lished, but the molecular processes in OSN cilia are not is required to quantify peptides by AQUA. Peptides that fully understood. The study of Kuhlmann et al. focused on are not detectable by LC-MS/MS can never be quantified. the analysis of the mouse olfactory membrane ciliome to To avoid this problem absolute quantification techniques discover new resident proteins and low-abundant central such as full-length expressed stable isotope-labeled pro- components (mainly focused on ORs) in the OSNs (Kuhl- teins for absolute quantification (FLEXIQuant) (Brun mann et al., 2014). The used workflow includes a selective et al., 2007) or protein standard absolute quantification enrichment of olfactory ciliary membrane proteins, with (PSAQ) (Singh et al., 2009) were developed. Here, not only an established Ca2+/K+-shock method from transgenic peptides but entire heavy-labeled proteins with known mice expressing the gene coding for the green fluores- concentrations were added to the sample prior to tryptic cent protein under control of the olfactory marker protein digestion. Beside label-based absolute protein quantifi- (OMP)-promoter (Mayer et al., 2009). Next, the enriched cation, label-free absolute quantification is a consider- ciliary fractions were treated with Na2CO3 under alkaline able, software-based technique, which usually requires pH to enrich integral membrane proteins (Fujiki et al., C. Lindemann et al.: Strategies in quantitative proteomics 691

1982). After a successful verification of the enrichment of iBAQ values (estimated relative protein abundance level: olfactory cilia membranes by immunoblot assays, a gel- sum of all peptide intensities divided by the number of based and a gel-free (including strong cation exchange theoretically observable tryptic peptides of a protein) for (SCX)-chromatography) LC-MS/MS was performed. The the 62 ORs in the gel-free approach with an abundance usage of two different approaches (gel-free and gel-based) range more than two orders of magnitude were measured was carried out to ensure a comprehensive inventory of (Kuhlmann et al., 2014). All known constituents of the olfactory ciliary membranes to the depth of low abundant canonical pathway of olfactory signal transduction were ORs (Kuhlmann et al., 2014). In total, 4403 unique protein identified in the proteome study (ACIII, three CNG channel groups (false discovery rate <0.01) were identified, from subunits, heterotrimeric G protein subunits GNAL, GNB1 which 3750 protein groups (85.2%) were detected after and GNB13) with high iBAQ values, indicating a high the gel-based and 4257 protein groups (96.7%) after the expression level in olfactory cilia. Numerous proteins gel-free approach. Generally, the gel-free has more advan- of the Ca2+-signaling network, such as the Ca2+-binding tages compared to the gel-based analysis, because more proteins Annexin, were identified. Annexins ANAX1, ORs (60%, 37 of 62) were identified [including low-abun- ANAX2, ANAX5 and S100A5 were successfully verified by dant ORs; gel-based: only 6% (4 of 62)] and an overlay of immunoassays providing these proteins as potential new more than 80% of the protein IDs for soluble and integral functional targets in the investigation of olfactory signal membrane proteins within the biological replicates (n = 3) transduction. The identification of proteins involved in could be reached (see Figure 2). Furthermore, higher vesicle and intraflagellar transport (BBS7, BBS8: sorting or

A n = 3

Ciliary membrane- SCX LC-MS/MS and Olfactory epithelia Ciliary fraction enriched fraction chromatography data analysis

2+ + Ca /K -treatment Na2/Co3-treatment 25 fractions/replicate 75 samples in total

Σ 4257 Proteins Σ 213 Olfactory-specific proteins BC

Unknown Cell 763 (18%) membrane 758 (18%) Olfactory signaling and Non-olfactory regulation stimuli and cell 103 (48%) adhension Cytoplasm Organellar 35 (16%) 578 (14%) membrane 771 (18%) Membrane transport Organelle 29 (14%) 524 (12%) Unchar. membrane 444 (10%) Apoptosis and Membrane, vesicle cell survival and cytoskeleton Cytoskeleton Ciliary 9 (4%) organization 155 (4%) membrane Catalytical 10 (5%) Extracellular Cilia 76 (2%) processes Other or unknown Extracellular 116 (3%) intracellular 9 (4%) function milieu 72 (2%) 8 (4%) 10 (5%)

Figure 2: Experimental strategy and identification of novel olfactory-specific proteins. In panel (A) an experimental workflow to identify novel olfactory-specific proteins is shown. Celia were detached from mice olfactory 2+ + epithelia by using Ca /K -treatment and washed with Na2CO3. Next, ciliary membrane-enriched fractions were separated by SCX chromato graphy and 75 samples were measured by LC-MS/MS in total followed by data analysis. In (B) the localization of all 4257 identified proteins is illustrated. The absolute and relative amount of proteins is given in %. In (C) the function of all 213 olfactory cilium-specific membrane- associated and ciliary associated proteins are shown. The absolute and relative amount of proteins is given in %. All presented experiments (A, B and C) were performed with three biological replicates. SCX, strong cation exchange; LC-MS/MS, liquid chromatography-tandem mass spectrometry. Revised from Kuhlmann et al. (2014). 692 C. Lindemann et al.: Strategies in quantitative proteomics protein to the cilia), membrane transport processes (e.g. combination of both AP-experiments (AP-AM and AP-PM) K+- and Na+ -channels, Na+/K+-ATPase subunit) and cata- results in an improved discrimination of specific, non-spe- lytic processes (e.g. AK1 and ALDOA) indicate a communi- cific and transient interactions (Wang and Huang, 2008). cation between the extracellular milieu and the olfactory Data analysis of specific and non-specific interactors should cilia. In summary, Kuhlmann et al. presented the so far reveal the same light and heavy abundances in the AP-AM largest olfactory membrane ciliome study with detailed as well in the AP-PM strategy. In contrast, transient inter- results about the heterogeneity and relative abundance actors will show a higher abundance in their heavy form levels of native ORs in the olfactory cilia in Mus musculus in the AP-AM approach. Notably, the characterization of using a label-free relative quantification approach. The large membrane protein complexes by AP-MS presents a established workflow provides a huge potential for further challenging task. Due to their hydrophobicity, purifica- analysis of stimulus-induced changes in the olfactory tion of such complexes in their native form requires novel cilia membrane proteome to investigate the adaptation of strategies (Oeljeklaus et al., 2009). In a dual-track SILAC peripheral olfactory system to external stimuli over time. approach, which combines the AP-AM and AP-PM strategies, Oeljeklaus and colleagues successfully investigated the peroxisomal importomer in S. cerevisiae (Oeljeklaus et al., SILAC-based relative quantification strategy 2012). During the biogenesis of peroxisomes, proteins of the for the identification of core components importomer mediate the import of matrix proteins across and transient interactors of the peroxisomal the organelle membrane. Pex14p represents the central importomer in S. cerevisiae component of the multiprotein importomer complex. Using the dual-track SILAC approach, Pex14p tagged with TEVcs- The large-scale analysis of PPI networks can be realized with Protein A (PA) was overproduced and isolated from yeast techniques such as the yeast two-hybrid system or AP-MS cells grown in light SILAC medium (‘tagged’-culture). Yeast (Fields and Sternglanz, 1994; Rigaut et al., 1999). In contrast cells synthesizing endogenous Pex14p served as control and to two-hybrid systems, where interactions of only two pro- were metabolic labeled with the corresponding heavy SILAC teins are tested at a single time point, AP-MS can be used for form (‘control’ culture). Lysates of differentially treated the investigation of multiprotein complexes. The isolation cells were subjected to AP-AM or AP-PM experiments using of complexes is carried out under native conditions using a tagged Pex14p as bait protein in equal concentrations (see tagged bait protein subjected to one or more affinity purifi- Figure 3). The dual-track AP-MS SILAC strategy resulted in cation steps followed by LC-MS/MS (von Mering et al., 2002). the identification of known components of the importomer Additionally, the combination of AP-MS with SILAC allows (Pex13p, Pex14p, Pex17p), proteins closely associated with the quantification of interaction partners inside protein the importomer (Dyn2) or involved in peroxisome prolifera- complexes and can be used to probe dynamic changes tion (Pex11p). Additionally, a group of transient interaction (Brand et al., 2004). In an AP-MS SILAC experiment the partners including peroxins, peroxisomal membrane, and ‘tagged’ bait protein is overproduced in light SILAC medium. matrix proteins were identified. The results allow a clear In a second culture the ‘control’ protein is overproduced in discrimination between stable and transient interaction heavy SILAC medium. The two differently grown cultures are partners of Pex14p. Thereby, Oeljeklaus et al. identified the mixed before AP, termed as affinity purification after mixing most detailed Pex14p interactome with nine core and twelve (AP-AM) (Wang and Huang, 2008). During mixing, an on/ transient components. Furthermore, the successful applica- off exchange of light- and heavy-labeled transient bait inter- tion on large membrane protein complexes was carried out acting proteins can occur. In contrast, specific light-labeled without organelle isolation and additional purification or interactors are constantly bound to the bait protein. Fur- washing steps. Thus, the dual-track SILAC approach allows thermore, non-specific binding of light- and heavy-labeled further investigations of membrane-embedded macromo- proteins to the affinity matrices can occur. After purification lecular complexes. and LC-MS/MS, specific interactors are identified mainly in the light-labeled state, whereas non-specific and transient interactors will appear in the light- and heavy-labeled form. AQUA-like absolute quantification Therefore, transient interacting proteins can be falsely cat- of prenylated Rab proteins – a unique egorized as co-purified contaminants after AP. This results posttranslational modification in false positive or false negative identification of interactors (Gingras et al., 2007). Affinity purification prior to mixing In comparison to other Ras superfamily members, Rab (AP-PM) prevents an exchange of transient interactors. A GTPases are also posttranslational modified with prenyl C. Lindemann et al.: Strategies in quantitative proteomics 693

1993; Farnsworth et al., 1994; Stein et al., 2003; Rak et al., 2004). After geranylgeranylation, the Rab GGTase leaves the REP complex. Finally, the REP escorts the Rab protein to the membrane of its final compartment (Stein et al., 2003). During vesicular transport and protein trafficking, Rab GTPases are bound via the geranylgeranyl moieties to a lipid membrane of a vesicle, an organelle, or the plasma membrane (Stein et al., 2003). After completing vesicular transport, the Rab protein is recognized by Rab GDP dissociation inhibitor (Rab GDI) for return transport of the Rab protein to its starting membrane (Stein et al., 2003; Goody et al., 2005; Hutagalung and Novick, 2011). Therefore, Rab GDI internalizes the geranylgeranyl moieties and extracts the protein from the lipid membrane (Goody et al., 2005). The Rab protein is returned to its starting membrane for a new initialization of a vesicular transport. Dysfunction of Rab GTPases also plays crucial roles in serious human diseases (Goitre et al., 2014). Dif- ferent to other members of the Ras superfamily, patho- Figure 3: Quantitative dual-track AP-MS SILAC approach for the genesis is not developed by influences of mutations in investigation of the interactome of Pex14p, the central component of the Rab protein itself. Furthermore, interacting proteins the peroxisomal importomer. like REP, Rab GDI or other effectors cause diseases such Yeast cultures producing either PA-TEV-tagged Pex14p (Tag-Pex14p) as choroideremia (Andres et al., 1993), X-linked mental or non-tagged Pex14p were cultured in light SILAC medium (‘tagged’) retardation (D’Adamo et al., 1998) or Griscelli syndrome or heavy SILAC medium (‘control’). Oleic acid was supplemented, inducing peroxisome proliferating. Light- and heavy-labeled cells (Menasche et al., 2003a,b). Rab dysregulation or aberrant were mixed either after (AP-AM) or prior (AP-PM) affinity purifica- expression was observed in cancer, due to the support of tion. Purified samples using both strategies were analyzed by tumor progression (Recchi and Seabra, 2012; Zhen and SDS-PAGE followed by tryptic digestion and LC-MS/MS analysis. Stenmark, 2015). Therefore, Rab proteins are interesting Statistical data analysis allowed discrimination between specific clinical targets. Since no Rab-specific drugs were devel- (diamond), non-specific (triangle) and transient (square) interacting oped so far, Rab GGTase inhibitors were used to prevent proteins of Pex14p. TEV, Tobacco etch virus; SILAC, stable isotope labeling with amino acids in cell culture; AP-AM, affinity purification geranylgeranylation (Recchi and Seabra, 2012; Triola after mixing; AP-PM, affinity purification prior mixing; LC-MS/MS, et al., 2012). This assures non-function of the aberrant liquid chromatography-tandem mass spectrometry. Revised from expressed Rab proteins due to non-membranous localiza- Oeljeklaus et al. (2012). tion. As prenylated Rabs are of low abundance, strategies were developed to target this class of proteins. A strategy prior to mass spectrometric analysis was the online chro- moieties attached to the C-terminus of the protein. Dif- matographic separation technique multidimensional ferently to other Ras GTPases, which bear a CAAX-box protein identification technology (MudPIT) (Washburn (with C = cysteine, A = aliphatic amino acid, X = any et al., 2001; Franzel and Wolters, 2011). In MudPIT, pep- amino acid; defining the type of attached prenyl group, tides prepared after a tryptic digestion of complex protein either farnesyl or geranylgeranyl) as modification signal, samples are two-dimensionally separated. Eluting pep- Rab GTPases possess a CXC- or CC-motif. Both cysteines tides from the SCX phase reach an analytical reverse- of the CXC- or CC-motif bear a geranylgeranyl moiety; phase for hydrophobic separation. Using this technique, therefore Rab GTPases are often doubly geranylgera- we were able to separate farnesylated peptides in complex nylated (Farnsworth et al., 1994; Wennerberg et al., 2005; cell samples. We demonstrated that prenylated Rab pep- Triola et al., 2012). Before geranylgeranylation, each tides elute much later than any other non-prenylated newly synthesized Rab GTPase needs to be recognized peptide on the reverse phase of our MudPIT column, by a Rab escort protein (REP). This protein presents the which paves the way for clean MS/MS analysis without bound Rab GTPase to the Rab geranylgeranyltransferase interfering peptides (see Figure 4) (Wotske et al., 2012). (Rab GGTase), which attaches the geranylgeranyl moie- Moreover, prenylated peptides elute exclusively with ties to the cysteines of the CXC- or CC-motif (Andres et al., high salt step conditions (NH4Ac), which adds another 694 C. Lindemann et al.: Strategies in quantitative proteomics dimension for robust identification and thus quantifica- see Figure 5 in green); and RabGTPases that control tion even without or with crude enrichment. the traffic in and to the Golgi apparatus (Rab 1A,B, and Interestingly, prenylated peptides show a loss of the Rab2A,B, see Figure 5 in blue). Eventually, profiling the prenyl group during ESI and MS/MS fragmentation. This entire prenylome to follow the mechanism and potency characteristic event can be used for specific quantifica- of different prenyltransferase inhibitors, for example, tion approaches. For the farnesylated peptides, the proto- becomes feasible. nated farnesyl group (205 Da) serves as a marker ion and a neutral loss of the corresponding ions in MS/MS spectra can be observed (Wotske et al., 2012). Conclusions Alexandrov and coworkers introduced a strategy that tags Rab proteins with a biotin-geranyl moiety and a sub- Recently, different quantitative proteomics techniques sequently isolation via streptavidin (Nguyen et al., 2009). were successfully established for several biological For this purpose they prevented geranylgeranylation of systems and scientific questions. First, a label-free MS- Rab proteins in COS-7 cells by using different inhibitors intensity-based quantification method was used for the (compactin, BMS, etc.) and finally attached the biotin- investigation of the membrane proteome of sensory cilia geranyl moieties via recombinant REP-1 and Rab GGTase, to the depth of olfactory receptors in Mus musculus. This which recognize the artificial substrate biotinylated work presents the largest olfactory membrane ciliome geranyl pyrophosphate (BGPP). Using MudPIT, Wotske study to date that includes results about heterogeneity et al. identified and quantified 42 different Rab proteins and relative abundance levels of native ORs in the olfac- in COS-7 cells. By using this strategy, farnesylated and tory cilia in Mus musculus. The established workflow singly geranylgeranylated proteins can also be iden- allows further analysis of stimulus-induced changes and tified when adding recombinant farnesyltransferase long term adaptation of the peripheral olfactory system, or geranylgeranyltransferase for studying the whole including low-abundant signaling components like ORs. prenylome. For quantification of the prenylated Rab Second, the dual-track SILAC approach allowed the iden- GTPases they used an AQUA-like approach. In contrast tification of highly transient interactors resulting in the to synthetic stable isotope peptides in AQUA, they used most detailed interactome study of Pex14p to date. Addi- 15N-labeled overproduced Rab22A, which was spiked into tionally, the strategy enabled the analysis of the mem- the compactin-treated and BMS-treated COS-7 cell lysate brane-localized importomer without additional AP steps as an internal standard for absolute quantification. Total or organelle isolation. This opens up future in-depth amount of Rabs in the BMS-treated sample was normal- investigation of membrane protein complexes. Addition- ized against the total amount of Rabs in the compactin- ally, we investigated the quantification of less prominent treated sample. By this approach they elucidated that the posttranslational modification, namely protein prenyla- abundant RabGTPases form three clusters: GTPases that tion and exemplified an AQUA-like strategy. For this control the early steps of endocytosis (Rab5A,B,C, Rab14, quantification approach we developed suitable enrich- Rab21, and Rab35, see Figure 5 in red); GTPases that ment, separation and MS strategies. The demonstrated control secretion (Rab3D, Rab10, Rab11A,B, and Rab18, high potential of the established approaches advanced

Figure 4: Base peak chromatograms from the analyses of 5 pmol Rab1B-CVIM in 5 pmol digest of HeLa cell proteins. The 125 mm ammonium acetate salt step is displayed. # and * indicate the retention times of the farnesylated C-terminal peptide IDSTPVKPAGGGC(farnesylated)VIM (*) and the farnesylated C-terminal peptide with an oxidized methionine (#). RT, retention time; MS, mass spectrometry. Revised from Wotske et al. (2012). C. Lindemann et al.: Strategies in quantitative proteomics 695

Figure 5: Identification and quantification of Rab GTPases from compactin-treated (full bars) and BMS-treated (striped bars) COS-7 cells. The protein abundance is given in pmol and individual Rab proteins are colored according to their function: Rabs controlling the biosyn- thetic pathways are blue, Rabs controlling secretion are green, and Rabs controlling endocytic and recycling pathways are red. Revised from Nguyen et al. (2009).

the knowledge in the individual projects. In general, rela- MS (Geiger et al., 2010, 2011; Shenoy and Geiger, 2015). tive and absolute quantitative techniques and applica- During data analysis, samples are quantified relatively tion continuously develop. SILAC quantification evolved against the internal standard and fold changes are cal- by the construction of so-called spike-in or super-SILAC culated between an isotopic-labeled peptide of the standards. Both techniques are based on internal, iso- standard and the non-labeled peptide of interest. The topic-labeled protein standards of one cell line (spike-in techniques were applied to cell culture as well as tissues SILAC) or two or more different cell lines (super-SILAC), or body fluids, which indicates the high potential of dif- which are added to the samples of interest before LC-MS/ ferent SILAC approaches in the fields of clinical research 696 C. Lindemann et al.: Strategies in quantitative proteomics in the future (Chen et al., 2015). Another alternative to Beynon, R.J. and Pratt, J.M. (2005). Metabolic labeling of proteins for common metabolic or chemical isotopic labeling strate- proteomics. Mol. Cell. Proteomics 4, 857–872. Blagoev, B., Kratchmarova, I., Ong, S.E., Nielsen, M., Foster, L.J., gies was introduced by the usage of so-called N-succin- and Mann, M. (2003). A proteomics strategy to elucidate func- imidyloxycarbonylmethyl tris(2,4,6-trimethoxyphenyl) tional protein-protein interactions applied to EGF signaling. phosphonium bromide (TMPP)-labeled peptides, which Nat. Biotechnol. 21, 315–318. reacts with the N-terminal of proteins or peptides. TMPP- Bose, R., Molina, H., Patterson, A.S., Bitok, J.K., Periaswamy, B., labeling reaches similar quantification accuracy com- Bader, J.S., Pandey, A., and Cole, P.A. (2006). Phosphoprot- pared to SILAC, however the increasing hydrophobicity eomic analysis of Her2/neu signaling and inhibition. Proc. Natl. Acad. Sci. USA 103, 9773–9778. of peptides lead to enhanced ionization efficiency (Shen Braakman, R.B., Tilanus-Linthorst, M.M., Liu, N.Q., Stingl, C., Dek- et al., 2015). Recently, a step forward in label-free quan- ker, L.J., Luider, T.M., Martens, J.W., Foekens, J.A., and Umar, A. tification was the development of DIA-based SWATH-MS (2012). Optimized nLC-MS workflow for laser capture microdis- introducing the concept of combining a protein library sected breast cancer tissue. J. Proteomics 75, 2844–2854. and individual digital maps. Analog to other targeted MS- Brand, M., Ranish, J.A., Kummer, N.T., Hamilton, J., Igarashi, K., Francastel, C., Chi, T.H., Crabtree, G.R., Aebersold, R., and methods, the quantification occurs on MS/MS level with Groudine, M. (2004). Dynamic changes in transcription factor similar accuracy and consistency. Furthermore, SWATH- complexes during erythroid differentiation revealed by quanti- MS improves the identification of small and low abun- tative proteomics. Nat. Struct. Mol. Biol. 11, 73–80. dant proteins showing a high potential in the discovery of Breer, H. (2003). Sense of smell: recognition and transduction of biomarker and clinical diagnosis of human tissue, where olfactory signals. Biochem. Soc. Trans. 31, 113–116. chemical or metabolic labeling is often not applicable Brun, V., Dupuis, A., Adrait, A., Marcellin, M., Thomas, D., Court, M., Vandenesch, F., and Garin, J. (2007). Isotope-labeled protein (Anjo et al., 2017). standards: toward absolute quantitative proteomics. Mol. Cell. Proteomics 6, 2139–2149. Acknowledgments: All work was supported by the Chen, X., Wei, S., Ji, Y., Guo, X., and Yang, F. (2015). Quantitative German Research Foundation within the research pro- proteomics using SILAC: principles, applications, and develop- ject SFB642: GTP- and ATP-dependent membrane pro- ments. Proteomics 15, 3175–3192. Cox, J. and Mann, M. (2008). MaxQuant enables high peptide iden- cesses. We thank Prof. Dr. Bettina Warscheid, Dr. Silke tification rates, individualized p.p.b.-range mass accuracies Oeljeklaus and Dr. Katja Kuhlmann for their contribution and proteome-wide protein quantification. Nat. Biotechnol. 26, to the successful work within SFB642. The prenylation 1367–1372. work was carried out in close collaboration with Prof. Crotty, F.V., Blackshaw, R.P., and Murray, P.J. (2011). Differential Roger Goody and other members of the Max Planck Insti- growth of the fungus Absidia cylindrospora on 13C/15N-labelled tute of Molecular Physiology in Dortmund, Germany. media. Rapid Commun. Mass Spectrom. 25, 1479–1484. D’Adamo, P., Menegon, A., Lo Nigro, C., Grasso, M., Gulisano, M., Tamanini, F., Bienvenu, T., Gedeon, A.K., Oostra, B., Wu, S.K., et al. (1998). Mutations in GDI1 are responsible for X-linked References non-specific mental retardation. Nat. Genet. 19, 134–139. Dillen, L., Cools, W., Vereyken, L., Lorreyne, W., Huybrechts, T., de Amanchy, R., Kalume, D.E., and Pandey, A. (2005). Stable isotope Vries, R., Ghobarah, H., and Cuyckens, F. (2012). Comparison labeling with amino acids in cell culture (SILAC) for studying of triple quadrupole and high-resolution TOF-MS for quantifica- dynamics of protein abundance and posttranslational modification of peptides. Bioanalysis 4, 565–579. tions. Sci. STKE 2005, pl2. Farnsworth, C.C., Seabra, M.C., Ericsson, L.H., Gelb, M.H., and Andres, D.A., Milatovich, A., Ozcelik, T., Wenzlau, J.M., Brown, M.S., Glomset, J.A. (1994). Rab geranylgeranyl transferase catalyzes Goldstein, J.L., and Francke, U. (1993). cDNA cloning of the two the geranylgeranylation of adjacent cysteines in the small subunits of human CAAX farnesyltransferase and chromosomal GTPases Rab1A, Rab3A, and Rab5A. Proc. Natl. Acad. Sci. USA mapping of FNTA and FNTB loci and related sequences. Genom- 91, 11963–11967. ics 18, 105–112. Fields, S. and Sternglanz, R. (1994). The two-hybrid system: an assay Anjo, S.I., Santa, C., and Manadas, B. (2017). SWATH-MS as a tool for protein-protein interactions. Trends Genet. 10, 286–292. for biomarker discovery: From basic research to clinical appli- Filiou, M.D., Varadarajulu, J., Teplytska, L., Reckow, S., Maccarrone, cations. Proteomics 17, 3–4. G., and Turck, C.W. (2012). The 15N isotope effect in Escheri- Asara, J.M., Christofk, H.R., Freimark, L.M., and Cantley, L.C. (2008). chia coli: a neutron can make the difference. Proteomics 12, A label-free quantification method by MS/MS TIC compared to 3121–3128. SILAC and spectral counting in a proteomics screen. Proteom- Franzel, B. and Wolters, D.A. (2011). Advanced MudPIT as a next step ics 8, 994–999. toward high proteome coverage. Proteomics 11, 3651–3656. Bantscheff, M., Lemeer, S., Savitski, M.M., and Kuster, B. (2012). Fredens, J., Engholm-Keller, K., Giessing, A., Pultz, D., Larsen, M.R., Quantitative mass spectrometry in proteomics: critical review Hojrup, P., Moller-Jensen, J., and Faergeman, N.J. (2011). Quan- update from 2007 to the present. Anal. Bioanal. Chem. 404, titative proteomics by amino acid labeling in C. elegans. Nat. 939–965. Methods 8, 845–847. C. Lindemann et al.: Strategies in quantitative proteomics 697

Fujiki, Y., Hubbard, A.L., Fowler, S., and Lazarow, P.B. (1982). Isola- Lindemann, C., Lupilova, N., Muller, A., Warscheid, B., Meyer, H.E., tion of intracellular membranes by means of sodium carbonate Kuhlmann, K., Eisenacher, M., and Leichert, L.I. (2013). Redox treatment: application to endoplasmic reticulum. J. Cell Biol. proteomics uncovers peroxynitrite-sensitive proteins that help 93, 97–102. Escherichia coli to overcome nitrosative stress. J. Biol. Chem. Geiger, T., Cox, J., Ostasiewicz, P., Wisniewski, J.R., and Mann, M. 288, 19698–19714. (2010). Super-SILAC mix for quantitative proteomics of human Liu, H., Sadygov, R.G., and Yates, J.R., 3rd. (2004). A model for ran- tumor tissue. Nat. Methods 7, 383–385. dom sampling and estimation of relative protein abundance in Geiger, T., Wisniewski, J.R., Cox, J., Zanivan, S., Kruger, M., Ishi- shotgun proteomics. Anal. Chem. 76, 4193–4201. hama, Y., and Mann, M. (2011). Use of stable isotope labeling Liu, N.Q., Braakman, R.B., Stingl, C., Luider, T.M., Martens, J.W., by amino acids in cell culture as a spike-in standard in quanti- Foekens, J.A., and Umar, A. (2012). Proteomics pipeline for bio- tative proteomics. Nat. Protoc. 6, 147–157. marker discovery of laser capture microdissected breast cancer Gerber, S.A., Rush, J., Stemman, O., Kirschner, M.W., and Gygi, S.P. tissue. J. Mammary Gland Biol. Neoplasia 17, 155–164. (2003). Absolute quantification of proteins and phosphopro- Lu, P., Vogel, C., Wang, R., Yao, X., and Marcotte, E.M. (2007). teins from cell lysates by tandem MS. Proc. Natl. Acad. Sci. USA Absolute protein expression profiling estimates the relative 100, 6940–6945. contributions of transcriptional and translational regulation. Gillet, L.C., Navarro, P., Tate, S., Rost, H., Selevsek, N., Reiter, L., Nat. Biotechnol. 25, 117–124. Bonner, R., and Aebersold, R. (2012). Targeted data extraction Maerkens, A., Kley, R.A., Olive, M., Theis, V., van der Ven, P.F., of the MS/MS spectra generated by data-independent acquisi- Reimann, J., Milting, H., Schreiner, A., Uszkoreit, J., Eisenacher, tion: a new concept for consistent and accurate proteome M., et al. (2013). Differential proteomic analysis of abnormal analysis. Mol. Cell. Proteomics 11, O111 016717. intramyoplasmic aggregates in desminopathy. J. Proteomics Gingras, A.C., Gstaiger, M., Raught, B., and Aebersold, R. (2007). 90, 14–27. Analysis of protein complexes using mass spectrometry. Nat. Marcus, K. (2012). Quantitative Methods in Proteomics. Methods Rev. Mol. Cell Biol. 8, 645–654. Mol. Biol. 893, 85–341. Goitre, L., Trapani, E., Trabalzini, L., and Retta, S.F. (2014). The Ras Mayer, U., Kuller, A., Daiber, P.C., Neudorf, I., Warnken, U., superfamily of small GTPases: the unlocked secrets. Methods Schnolzer, M., Frings, S., and Mohrlen, F. (2009). The proteome Mol. Biol. 1120, 1–18. of rat olfactory sensory cilia. Proteomics 9, 322–334. Goody, R.S., Durek, T., Waldmann, H., Brunsveld, L., and Alexan- McEwen, D.P., Jenkins, P.M., and Martens, J.R. (2008). Olfactory drov, K. (2005). Application of protein semisynthesis for the cilia: our direct neuronal connection to the external world. Curr. construction of functionalized posttranslationally modified rab Top. Dev. Biol. 85, 333–370. GTPases. Methods Enzymol. 403, 29–42. Megger, D.A., Bracht, T., Kohl, M., Ahrens, M., Naboulsi, W., Weber, Gruhler, A., Schulze, W.X., Matthiesen, R., Mann, M., and Jensen, F., Hoffmann, A.C., Stephan, C., Kuhlmann, K., Eisenacher, M., O.N. (2005). Stable isotope labeling of Arabidopsis thaliana et al. (2013a). Proteomic differences between hepatocellular cells and quantitative proteomics by mass spectrometry. Mol. carcinoma and nontumorous liver tissue investigated by a com- Cell. Proteomics 4, 1697–1709. bined gel-based and label-free quantitative proteomics study. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H., and Mol. Cell. Proteomics 12, 2006–2020. Aebersold, R. (1999). Quantitative analysis of complex protein Megger, D.A., Bracht, T., Meyer, H.E., and Sitek, B. (2013b). Label- mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17, free quantification in clinical proteomics. Biochim. Biophys. 994–999. Acta 1834, 1581–1590. Hsu, J.L., Huang, S.Y., Chow, N.H., and Chen, S.H. (2003). Stable- Menasche, G., Feldmann, J., Houdusse, A., Desaymard, C., Fischer, isotope dimethyl labeling for quantitative proteomics. Anal. A., Goud, B., and de Saint Basile, G. (2003a). Biochemical and Chem. 75, 6843–6852. functional characterization of Rab27a mutations occurring in Hutagalung, A.H. and Novick, P.J. (2011). Role of Rab GTPases in Griscelli syndrome patients. Blood 101, 2736–2742. membrane traffic and cell physiology. Physiol. Rev. 91, 119–149. Menasche, G., Ho, C.H., Sanal, O., Feldmann, J., Tezcan, I., Ersoy, Ishihama, Y., Oda, Y., Tabata, T., Sato, T., Nagasu, T., Rappsilber, J., F., Houdusse, A., Fischer, A., and de Saint Basile G. (2003b). and Mann, M. (2005). Exponentially modified protein abun- Griscelli syndrome restricted to hypopigmentation results from dance index (emPAI) for estimation of absolute protein amount a melanophilin defect (GS3) or a MYO5A F-exon deletion (GS1). in proteomics by the number of sequenced peptides per pro- J. Clin. Invest. 112, 450–456. tein. Mol. Cell. Proteomics 4, 1265–1272. Merrill, A.E., Hebert, A.S., MacGilvray, M.E., Rose, C.M., Bailey, D.J., Kruger, M., Moser, M., Ussar, S., Thievessen, I., Luber, C.A., Forner, Bradley, J.C., Wood, W.W., El Masri, M., Westphall, M.S., Gasch, F., Schmidt, S., Zanivan, S., Fässler, R., and Mann, M. (2008). A.P., et al. (2014). NeuCode labels for relative protein quantifi- SILAC mouse for quantitative proteomics uncovers kindlin-3 cation. Mol. Cell. Proteomics 13, 2503–2512. as an essential factor for red blood cell function. Cell 134, Molina, H., Yang, Y., Ruch, T., Kim, J.W., Mortensen, P., Otto, T., 353–364. Nalli, A., Tang, Q.Q., Lane, M.D., Chaerkady, R., et al. (2009). Kuhlmann, K., Tschapek, A., Wiese, H., Eisenacher, M., Meyer, Temporal profiling of the adipocyte proteome during differen- H.E., Hatt, H.H., Oeljeklaus, S., and Warscheid, B. (2014). The tiation using a five-plex SILAC based strategy. J. Proteome Res. membrane proteome of sensory cilia to the depth of olfactory 8, 48–58. receptors. Mol. Cell. Proteomics 13, 1828–1843. Nguyen, U.T., Guo, Z., Delon, C., Wu, Y., Deraeve, C., Franzel, B., Lau, H.T., Suh, H.W., Golkowski, M., and Ong, S.E. (2014). Compar- Bon, R.S., Blankenfeldt, W., Goody, R.S., Waldmann, H., et al. ing SILAC- and stable isotope dimethyl-labeling approaches for (2009). Analysis of the eukaryotic prenylome by isoprenoid quantitative proteomics. J. Proteome Res. 13, 4164–4174. affinity tagging. Nat. Chem. Biol. 5, 227–235. 698 C. Lindemann et al.: Strategies in quantitative proteomics

Oeljeklaus, S., Meyer, H.E., and Warscheid, B. (2009). New dimen- Stein, M.P., Dong, J., and Wandinger-Ness, A. (2003). Rab proteins sions in the study of protein complexes using quantitative and endocytic trafficking: potential targets for therapeutic mass spectrometry. FEBS Lett. 583, 1674–1683. intervention. Adv. Drug Deliv. Rev. 55, 1421–1437. Oeljeklaus, S., Reinartz, B.S., Wolf, J., Wiese, S., Tonillo, J., Podwo- Sury, M.D., Chen, J.X., and Selbach, M. (2010). The SILAC fly allows jski, K., Kuhlmann, K., Stephan, C., Meyer, H.E., Schliebs, W., for accurate protein quantification in vivo. Mol. Cell. Proteom- et al. (2012). Identification of core components and transient ics 9, 2173–2183. interactors of the peroxisomal importomer by dual-track stable Thompson, A., Schäfer, J., Kuhn, K., Kienle, S., Schwarz, J., Schmidt, isotope labeling with amino acids in cell culture analysis. G., Neumann, T., Johnstone, R., Mohammed, A.K., and Hamon, J. Proteome Res. 11, 2567–2580. C. (2003). Tandem mass tags: a novel quantification strategy Oeljeklaus, S., Schummer, A., Suppanz, I., and Warscheid, B. for comparative analysis of complex protein mixtures by MS/ (2014). SILAC labeling of yeast for the study of membrane pro- MS. Anal. Chem. 75, 1895–1904. tein complexes. Methods Mol. Biol. 1188, 23–46. Triola, G., Waldmann, H., and Hedberg, C. (2012). Chemical biology Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, of lipidated proteins. ACS Chem. Biol. 7, 87–99. H., Pandey, A., and Mann, M. (2002). Stable isotope labeling von Mering, C., Krause, R., Snel, B., Cornell, M., Oliver, S.G., by amino acids in cell culture, SILAC, as a simple and accurate Fields, S., and Bork, P. (2002). Comparative assessment of approach to expression proteomics. Mol. Cell. Proteomics 1, large-scale data sets of protein-protein interactions. Nature 376–386. 417, 399–403. Ong, S.E., Foster, L.J., and Mann, M. (2003a). Mass spectrometric- Wang, X. and Huang, L. (2008). Identifying dynamic interactors of based approaches in quantitative proteomics. Methods 29, protein complexes by quantitative mass spectrometry. Mol. 124–130. Cell. Proteomics 7, 46–57. Ong, S.E., Kratchmarova, I., and Mann, M. (2003b). Properties Washburn, M.P., Wolters, D., and Yates, J.R., 3rd. (2001). Large-scale of 13C-substituted arginine in stable isotope labeling by analysis of the yeast proteome by multidimensional protein amino acids in cell culture (SILAC). J. Proteome Res. 2, identification technology. Nat. Biotechnol. 19, 242–247. 173–181. Wennerberg, K., Rossman, K.L., and Der, C.J. (2005). The Ras super- Ong, S.E., Mittler, G., and Mann, M. (2004). Identifying and quan- family at a glance. J. Cell Sci. 118, 843–846. tifying in vivo methylation sites by heavy methyl SILAC. Nat. Wiener, M.C., Sachs, J.R., Deyanova, E.G., and Yates, N.A. (2004). Methods 1, 119–126. Differential mass spectrometry: a label-free LC-MS method for Rak, A., Pylypenko, O., Niculae, A., Pyatkov, K., Goody, R.S., and finding significant differences in complex peptide and protein Alexandrov, K. (2004). Structure of the Rab7:REP-1 complex: mixtures. Anal. Chem. 76, 6085–6096. insights into the mechanism of Rab prenylation and choroider- Wiese, S., Reidegeld, K.A., Meyer, H.E., and Warscheid, B. (2007). emia disease. Cell 117, 749–760. Protein labeling by iTRAQ: a new tool for quantitative mass Recchi, C. and Seabra, M.C. (2012). Novel functions for Rab GTPases spectrometry in proteome research. Proteomics 7, 340–350. in multiple aspects of tumour progression. Biochem. Soc. Wotske, M., Wu, Y., and Wolters, D.A. (2012). Liquid chromato- Trans. 40, 1398–1403. graphic analysis and mass spectrometric identification of Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., and farnesylated peptides. Anal Chem. 84, 6848–6855. Seraphin, B. (1999). A generic protein purification method for Zhen, Y. and Stenmark, H. (2015). Cellular functions of Rab GTPases protein complex characterization and proteome exploration. at a glance. J. Cell Sci. 128, 3171–3176. Nat. Biotechnol. 17, 1030–1032. Schwanhausser, B., Gossen, M., Dittmar, G., and Selbach, M. (2009). Global analysis of cellular protein translation by pulsed SILAC. Proteomics 9, 205–209. Bionotes Shen, H., An, M., Zou, X., Zhao, X., Wang, Q., Xing, G., and Ji, J. (2015). Evaluation of the accuracy of protein quantifica- Claudia Lindemann tion using isotope TMPP-labeled peptides. Proteomics 15, Ruhr-University Bochum, Medizinisches 2903–2909. Proteom-Center, Universitätsstraße 150, Shenoy, A. and Geiger, T. (2015). Super-SILAC: current trends and D-44801 Bochum, Germany future perspectives. Expert Rev. Proteomics 12, 13–19. Silva, J.C., Gorenstein, M.V., Li, G.Z., Vissers, J.P., and Geromanos, S.J. (2006). Absolute quantification of proteins by LCMSE: a virtue of parallel MS acquisition. Mol. Cell. Proteomics 5, 144–156. Singh, S., Springer, M., Steen, J., Kirschner, M.W., and Steen, H. (2009). FLEXIQuant: a novel tool for the absolute quantification of proteins, and the simultaneous identification and quanti- Claudia Lindemann studied biochemistry at the University of fication of potentially modified peptides. J. Proteome Res. 8, Bayreuth/Germany. She completed her PhD thesis in the lab of Lars 2201–2210. Leichert, working on the thiol-based redox regulation in Escherichia Soufi, B., Kumar, C., Gnad, F., Mann, M., Mijakovic, I., and Macek, coli. She is currently working as post doc in the group of Katrin B. (2010). Stable isotope labeling by amino acids in cell culture Marcus at the Medizinisches Proteom-Center (MPC), Ruhr-University (SILAC) applied to quantitative proteomics of Bacillus subtilis. Bochum. Her research focusses on the MS-based analyses of struc- J. Proteome Res. 9, 3638–3646. tures and post-translational modifications of proteins. C. Lindemann et al.: Strategies in quantitative proteomics 699

Nikolas Thomanek Helmut E. Meyer Ruhr-University Bochum, Medizinisches Ruhr-University Bochum, Medizinisches Proteom-Center, Universitätsstraße 150, Proteom-Center, Universitätsstraße 150, D-44801 Bochum, Germany D-44801 Bochum, Germany

Nikolas Thomanek studied biology and biotechnology at the Ruhr-University of Bochum/Germany. He completed his PhD thesis Helmut E. Meyer studied biochemistry at the University of composed at the Medizinisches Proteom-Center (MPC), Ruhr-Uni- Tübingen. He completed his PhD thesis in the biochemistry versity Bochum. He is currently working as post doc in the group of department, Prof. Pfleiderer, of the Ruhr-University of Bochum on Katrin Marcus. His research focusses on the regulation of lipopoly- the primary structure of porcine LDH. He is a renowned protein saccharide biosynthesis in Escherichia coli and quantitative mass chemist and was working in this field for 40 years. He founded the spectrometry using SILAC and super-SILAC. Medizinisches Proteom-Center and is retired as full-professor in March 2014. Franziska Hundt Ruhr-University Bochum, Biomolecular Dirk Wolters Mass Spectrometry, Universitätsstraße 150, Ruhr-University Bochum, Biomolecular D-44801 Bochum, Germany Mass Spectrometry, Universitätsstraße 150, D-44801 Bochum, Germany

Franziska Hundt studied Biology at the Ruhr-University Bochum Dirk Wolters obtained his PhD in 1998 from the Ruhr University in a bachelor’s and master’s-program. In 2016 she successfully Bochum. He pursued his research interest in biomolecular mass completed her doctorate in the research group of Dirk Wolters. spectrometry with John R. Yates at the University of Washington, She focusses on establishing new enrichment strategies and mass Seattle, USA and different pharmaceutical companies (Novartis, spectrometric quantitation methods for studying Rab GTPases in San Diego, USA; Hoffmann-La Roche, Basel, Switzerland) before he different human cell lines. accepted a position as head of the Biomolecular Mass Spectrometry group of the Analytical Chemistry department at the Ruhr-University Thilo Lerari Bochum. Ruhr-University Bochum, Medizinisches Proteom-Center, Universitätsstraße 150, Katrin Marcus D-44801 Bochum, Germany Ruhr-University Bochum, Medizinisches Proteom-Center, Universitätsstraße 150, D-44801 Bochum, Germany, [email protected]

Thilo Lerari studied biology and biotechnology at the Ruhr-Univer- Katrin Marcus is director of the Medizinisches Proteom-Center, sity of Bochum/Germany. He completed his PhD thesis composed at Ruhr-University Bochum since 2014. Her expertise is in the field the Medizinisches Proteom-Center (MPC), Ruhr-University Bochum. of proteomics and mass spectrometry with a special focus on He is currently working as research assistant in the group of Katrin neurodegenerative and neuromuscular diseases. Biomarkers are Marcus. His research focusses on quantitative phosphoproteomics investigated by the application of state-of-the-art proteomics and the optimization and establishment of the applied strategies for methods and the development of new high-performance technolo- enrichment and detection. gies proteins involved in the pathogenesis of those diseases as well as diagnosis.