sign in sign up
<<
Home , Glycobiology, Glycomics

Leading Edge Essay

Glycomics Hits the Big Time

Gerald W. Hart1,* and Ronald J. Copeland1 1Department of Biological Chemistry, School of Medicine, Johns Hopkins University, 725 North Wolfe Street, Baltimore, MD 21205-2185, USA *Correspondence: [email protected] DOI 10.1016/j.cell.2010.11.008

Cells run on . , sequences of carbohydrates conjugated to and , are arguably the most abundant and structurally diverse class of in nature. Recent advances in reveal the scope and scale of their functional roles and their impact on human disease.

By analogy to the , , O-GlcNAc (Hart et al., 2007). Even though Dynamic Structural Complexity or proteome, the ‘‘’’ is the the generic term ‘‘’’ is often Underlies Functions complete set of glycans and glycoconju- used to categorize and lump all glycan provide dynamic struc- gates that are made by a cell or modifications of proteins into one bin, tural diversity to proteins and lipids that under specific conditions. Therefore, side by side with other posttranslational is responsive to cellular phenotype, to ‘‘glycomics’’ refers to studies that attempt modifications such as phosphorylation, metabolic state, and to the developmental to define or quantify the glycome of a cell, acetylation, ubiquitination, or methylation, stage of cells. Complex glycans play crit- tissue, or organism (Bertozzi and Sasise- such a view is not only inaccurate, but ical roles in intercellular and intracellular kharan, 2009). In eukaryotes, also is completely misleading. If one only processes, which are fundamentally glycosylation generally involves the cova- considers the linkage of the first glycan important to the development of multicel- lent attachment of glycans to serine, to the polypeptide in both prokaryotic lularity (Figure 1). Unlike nucleic acids and threonine, or asparagine residues. Glyco- and eukaryotic , there are at proteins, glycan structures are not hard- proteins occur in all cellular compart- least 13 different monosaccharides and wired into the genome, depending upon ments. Glycans are also attached to 8 different amino acids involved in glyco- a template for their synthesis. Rather, lipids, often ceramide, which is comprised protein linkages, with a total of at least the glycan structures that end up on of sphingosine, a hydrocarbon amino 41 different chemical bonds known to be a polypeptide or result from the alcohol and a fatty acid. Complex glycans linking the glycan to the protein (Spiro, concerted actions of highly specific gly- are mainly attached to secreted or cell 2002). Importantly, each one of these cosyltransferases (Lairson et al., 2008), surface proteins, and they do not cycle unique glycan:protein linkages is surely which in turn are dependent upon the on and off of the polypeptide. In contrast, as different in both structure and function concentrations and localization of high- the monosaccharide O-linked N-acetyl- as protein methylation is from acetylation. energy donors, such as glucosamine (O-GlcNAc) cycles rapidly Of course, this modification is not only UDP-N-acetylglucosamine, the endpoint on serine or threonine residues of many about a single linkage. When structural of the hexosamine biosynthetic pathway. nuclear and cytoplasmic proteins. Identi- diversity of the additional oligosaccharide Therefore, the glycoforms of a glycopro- fying the number, structure, and function branches of glycans and the added diver- tein depend upon many factors directly of glycans in cellular biology is a daunting sity of complex terminal saccharides on tied to both expression and cellular task but one that has been made easier in glycans, such as fucose or sialic acids metabolism. recent years by advances in technology (about 50 different sialic acids are known There are at-least 250 glycosyltrans- and by our growing appreciation of how [Schauer, 2009]), are taken into account, ferases in the human genome, and it has integral glycans are to biology (Varki the molecular diversity and varied func- been estimated that about 2% of the et al., 2009). tions of protein-bound glycans rapidly human genome encodes proteins The scope of the glycomics challenge is increase exponentially. Just the ‘‘sia- involved in glycan biosynthesis, degrada- immense. The covalent addition of lome’’ (Cohen and Varki, 2010) rivals or tion, or transport (Schachter and Freeze, glycans to proteins and lipids represents exceeds many other posttranslational 2009). Biosynthesis of the nucleotide not only the most abundant posttransla- modifications in abundance and struc- sugar donors is directly regulated by nu- tional modification (PTM), but also by far tural/functional diversity. In addition, cleic acid, glucose, and energy metabo- the most structurally diverse. Although it chemical modifications, such as phos- lism, and the compartmentalization of is commonly stated that more than 50% phorylation, sulfation, and acetylation, these nucleotide sugar donors is highly of all polypeptides are covalently modified increase the glycan structural/functional regulated by specific transporters. Protein by glycans (Apweiler et al., 1999), even diversity even more. Thus, categorizing glycosylation is therefore controlled by this estimate is far too low because it fails glycosylation as a single type of post- rates of polypeptide translation and to include that myriad nuclear and translational modification is neither useful protein folding, localization of and compe- cytoplasmic proteins are modified by nor at all reflective of reality. tition between glycosyltransferases,

672 Cell 143, November 24, 2010 ª2010 Elsevier Inc. it is estimated that the binding sites of glycan-binding proteins (GBPs), such as antibodies, , receptors, toxins, mi- crobial adhesions, or (Figure 1), can accommodate only up to two to six monosaccharides within a glycan struc- ture (Cummings, 2009). Therefore, the number of specific glycan substructures that bind to biologically important GBPs in a cell may be fewer than 10,000, a number that is within the realm of current analytical and, if targeted, chemi- cal or enzymatic synthetic capabilities. Until recently, the lack of tools and the inherent complexity of glycans have been major barriers preventing most biol- ogists from embracing the importance of glycans in biology. Recent technological advances have significantly lowered these barriers. Indeed, the tools of glycomics and the subfields of glycoproteomics, gly- colipidomics, and proteoglycomics have all progressed substantially in recent years (Krishnamoorthy and Mahal, 2009; Laremore et al., 2010). Major technolog- ical advances, many of which are shared with , have recently allowed Figure 1. Glycans Permeate Cellular Biology semiquantitative profiling of glycans and Complex glycans at the cell surface are targets of microbes and viruses, regulate cell adhesion and devel- (Krishnamoorthy and opment, influence metastasis of cells, and regulate myriad :ligand interactions. Glycans Mahal, 2009; Vanderschaeghe et al., within the secretory pathway regulate protein quality control, turnover, and trafficking of molecules to organelles. Nucleocytoplasmic O-linked N-acetylglucosamine (O-GlcNAc) has extensive crosstalk with 2010). Some of these advances are the phosphorylation to regulate signaling, cytoskeletal functions, and in response to nutrients result of the National Institute of General and stress. Medical Science’s (NIGMS) support of the Consortium for Functional Glycomics cellular concentration and localization of gene expression of glycan-processing (CFG), which has served to focus and nucleotide , the localization of enzymes, by polypeptide structure at all assist more than 500 researchers on glycosidases, and membrane trafficking. levels, and by cellular metabolism. issues related to glycomics (Paulson Thus, individual glycosylation sites on the et al., 2006; Raman et al., 2006). same polypeptide can contain different Technology of Glycomics Kobata and colleagues were among the glycan structures that reflect both the A detailed understanding of cellular first to profile N-glycans, well before the type and status of the cell in which they processes will require a detailed appreci- current concepts of glycomics were are synthesized. For example, the glyco- ation of the glycans modulating proteins conceived. Despite the lack of many forms of the membrane protein Thy-1 are and pathways. Although this ultimate modern methods, their pioneering work very different in lymphocytes than they goal of glycomics is laudable, we are was characterized by a high level of rigor are in brain, despite having the same poly- a very long way from having the tech- in defining the arrays of N-glycan struc- peptide sequence (Rudd and Dwek, nology to completely characterize the gly- tures present in cells and tissues and on 1997). Conversely, even small changes in come of even a simple cell or tissue. Not specific proteins (Endo, 2010). Currently, polypeptide sequence or structure will only is the glycome much more complex a wide variety of high-resolution and alter the types of glycan structures than the genome, transcriptome, or pro- highly sensitive methods are available, attached to a polypeptide. For example, teome, as noted above, it is also much including capillary (CE), histocompatibility antigen polypeptides more dynamic, varying considerably not high-performance liquid with more than 90% sequence homology only with cell type, but also with the (HPLC), and microarrays. contain different N-linked glycan profiles developmental stage and metabolic state Glycans are often profiled after their at individual sites, reflective of their of a cell. Even very conservative esti- release from polypeptides, which results in allelic type, even when they are synthe- mates indicate that there are well over the loss of any information about proteins sized within the same cells (Swiedler a million different glycan structures in and sites to which they were attached. et al., 1985). Thus, site-specific protein a mammalian cell’s glycome. However, Even though it is much more difficult, it is glycosylation is highly regulated by upon considering ‘‘functional glycomics,’’ also much preferable to perform

Cell 143, November 24, 2010 ª2010 Elsevier Inc. 673 glycopeptide profiling (glycoproteomics) to development of purified recombinant lec- custom-made DNA microarrays that first identify attachment sites prior to tins, and better definition of the specific- represent glycosyltransferases and detailed profiling or structural analysis of ities of many lectins (Gupta et al., 2010). glycan-binding proteins. The CFG also the glycans present on a polypeptide. The Both matrix-assisted laser desorption has developed databases that present ultimate goal of glycoproteomics, which is ionization (MALDI) and electrospray phenotypic and biochemical data on gly- to define all of the molecular species (glyco- have played a key cosyltransferase knockout mice. Even forms) of glycoproteins in a cell or tissue, role in glycan profiling and in glycoproteo- though knocking out a single glycosyl- has not yet been realized for any glycopro- mics (An et al., 2009; North et al., 2010; transferase gene often affects hundreds tein with more than one glycan attachment Zaia, 2010). For discovery, of glycoconjugates and myriad biological site. N-glycans are generally released from affinity enrichment approaches, based processes, these mutant mice have proteins by peptide-N-glycosidase F upon chemical modification and solid- proven valuable in revealing the funda- (PNGase F), which cleaves most, but not phase extraction of N-linked glycopro- mental biological importance of glycans. all, N-glycans. Unfortunately, no such teins, have proven useful in profiling The microarrays and the databases broadly specific exists for N-linked sites from serum- produced by the CFG member community O-glycans, which are generally released or even from paraffin-embedded tissues at large are publically available on the CFG by chemical methods, such as alkali- (Tian et al., 2009). Recently, using lectin website (http://www.functionalglycomics. induced b elimination, or by hydrazinolysis. binding combined with advanced mass org) and have resulted in a profound However, for relatively pure glycoproteins, spectrometric methods, thousands of increase in our understanding of the so called ‘‘top-down’’ mass spectrometric N-glycan attachment sites have been binding specificities of GBPs, including methods, which do not involve prior release mapped, a prerequisite for understanding lectins key to inflammation and immunity, of the glycans, may eventually prove useful, their functions (Zielinska et al., 2010). and on infectious microbes or viruses. as instrumentation and methods improve Given the structural diversity of However, a major barrier preventing (Reid et al., 2002). glycans, all of these glycomic approaches glycan biology from being incorporated Due to the small sample sizes involved, generate vast amounts of data. Glycan bi- more into the mainstream is the continued most CE or HPLC separation methods oinformatics has made great strides failure by the community to adopt a univer- require chemical modification of released within recent years with major efforts sally standard glycan structural format glycans with fluorescent compounds. CE from several laboratories (Aoki-Kinoshita, and database that are easily accessed and HPLC methods provide high-resolu- 2008). At least four major publicly worldwide. Most importantly, glycan data- tion separation of glycans, and when available databases (Glyco- bases must eventually be incorporated combined with laser-induced fluorescent sciences.de, KEGG GLYCAN, Euro- into standard interactive databases that detection (LIF), tagged glycans can be de- carbDB, and CFG) are now maintained, are supported by public agencies (such tected in the low femtomole range. High and efforts to structure them in a uniform as NCBI or EMBL) before glycan biology pH anion-exchange chromatography format have been in progress for quite can be fully integrated into the wider (HPAEC) with pulsed-amperometric some time. In addition, the Carbohy- research community. detection separates glycans with high drate-Active EnZyme database (CAZy) resolution and detects them with high has played a key role in providing a global From Glycomics to Biology sensitivity without chemical modification, understanding of carbohydrate active Glycans are directly involved in almost but the high alkalinity employed can be enzymes, documenting their evolutionary every biological process and certainly problematic for some labile structures. relationships, providing a framework for play a major role in nearly every human Lectins, which are defined as carbohy- elucidating common mechanisms, and disease (Figure 1). Genetic studies in drate-binding proteins that are neither establishing the relationship between gly- tissue culture cells indicate that specific antibodies nor enzymes, have a wide cogenomics and expressed by complex glycan structures are generally range of glycan binding specificities, suit- cells (Cantarel et al., 2009). Moreover, not essential to a cell growing in culture, able for partial characterization of a gly- recent advances in bioinformatic analysis indicating that most of the functions of come. Lectin microarrays use methods tools for complex glycomic mass spec- complex glycans are at the multicellular and equipment similar to that employed trometry data sets have allowed complex level. In contrast, the cycling monosac- for nucleic acid arrays. Given the large data to be presented in formats useful to charide, O-GlcNAc, on nuclear and cyto- number of different lectins available, lectin nonexperts in all fields of biology (Ceroni plasmic proteins, is essential even at the microarrays can provide information et al., 2008; Goldberg et al., 2005). single cell level in mammals (Hart et al., about the glycome in a high-throughput Perhaps one of the most important 2007). fashion, which is particularly useful in contributions to the field of functional The critical roles of glycans in mammals profiling glycans produced by infectious glycomics has been the development of are now well established not only by the organisms (Hsu et al., 2006). In the future, well-defined glycan microarrays, which dearth of mutations in glycan biosynthetic it is highly likely that glycomics will play currently display more than 500 different enzymes that survive development, but a central role in combating infectious glycan structures (Smith et al., 2010). also by the severe phenotypes generated disease. However, many technical issues The NIGMS-supported Consortium for when such mutations are not lethal. remain to be resolved, such as standard- Functional Glycomics (CFG) has gener- These severe phenotypes are clearly illus- ization required for clinical use, the ated and made publicly available trated by the congenital disorders of

674 Cell 143, November 24, 2010 ª2010 Elsevier Inc. landscape, the pharmaceutical industry and the US Food and Drug Administration are rapidly realizing the critical importance, in terms of both bioactivity and safety, of carefully defining the glycoforms of any therapeutics derived from glycoconju- gates.

Glycoproteomics, Glycolipidomics, and Biomarkers Clinical cancer diagnostic markers are often glycoproteins, but most current diagnostic tests only measure the expres- sion of the polypeptide. Clearly, given the Figure 2. Glycomic Complexity Reflects Cellular Complexity long known alterations in glycans associ- Given that glycan structures are regulated by metabolism and glyco-enzyme expression and glycans ated with cancer, it is highly likely that modify both proteins and lipids, functional glycomics also requires the tools of , proteomics, lip- cancer markers that detect specific glyco- idomics, and (modified after Packer et al., 2008). forms of a protein will have much higher sensitivity and specificity for early glycosylation (CDGs) (Schachter and organization of receptors on the cell detection of cancer (Packer et al., 2008; Freeze, 2009), which are associated with surface and play important roles in immu- Taniguchi, 2008). Thus, the convergence severe mental and developmental abnor- nity, infections, development, and inflam- of glycomics and glycoproteomics is key malities. Also, the severe muscular mation (Lajoie et al., 2009). to the discovery of biomarkers for the early dystrophy that results from defective and glycosaminoglycans play a key role in detection of cancer (Taylor et al., 2009). O-glycosylation of a-dystroglycan (Yosh- the regulation of growth factors, in micro- Recently, the Food and Drug Administra- ida-Moriguchi et al., 2010) further bial binding, in tissue morphogenesis, and tion has approved fucosylated a-fetopro- illustrates how a mutation in a glycan in the etiology of cardiovascular disease. tein as a diagnostic marker of primary biosynthetic enzyme results in a devas- Proteoglycans are perhaps the most hepatocarcinoma. In addition, fucosy- tating disease. The interplay between complicated and information-rich mole- lated haptoglobin may be a much better O-GlcNAcylation and phosphorylation on cules in biology, and progress in proteo- marker of pancreatic cancer than simply nuclear and cytoplasmic proteins plays glycomics has begun to accelerate monitoring the expression of the hapto- a key role in the etiology of diabetes, (Ly et al., 2010). Nearly all microbes and globin polypeptide. Indeed, The National neurodegenerative disease, and cancer viruses that infect humans bind to cells Cancer Institute has begun an initiative to (Hart et al., 2007; Zeidan and Hart, 2010). by attaching to specific cell surface discover, develop, and clinically validate It has long been appreciated that alter- glycans. Glycomics and glycan arrays glycan biomarkers for cancer (http:// ations in cell surface glycans contribute to will have a substantial impact upon future glycomics.cancer.gov/). System biology the metastatic and neoplastic properties research toward both diagnosing and analyses of the glycome to identify of tumor cells (Taniguchi, 2008). The func- preventing infectious disease. biomarkers of human disease will, by tions of many receptors are modulated by Some of the most important drugs on the necessity, also employ many of the same their glycans, such as modulation of market are already the result of glycomics. methods used by genomics, proteomics, Notch receptors by the action of specific The anti-flu virus drugs Relenza and Tami- metabolomics, and (Figure 2) glycosyltransferases (Moloney et al., flu are structural analogs of sialic acids that (Packer et al., 2008). Due to the critical 2000), which regulate Notch’s activation inhibit the flu virus neuraminidase and the roles of glycans in cardiovascular disease by its ligands, affecting many develop- transmission of the virus. Natural , and lung disease and in the functions of mental events. Selectins, which specifi- a sulfated glycosaminoglycan, and chemi- blood cells, the National Heart Lung and cally bind to a subset of fucosylated and cally defined synthetic heparin oligosac- Blood Institute (NHLBI) has recognized sialylated glycans, play a critical role in charides have long been widely used in an acute need to train more researchers leukocyte homing to sites of inflamma- the clinic as anticoagulants and for many in the area of glycosciences by creating tion. Indeed, a selectin inhibitor is other clinical uses. Hyaluronic acid, a non- a ‘‘Program of Excellence in Glycoscien- currently in phase two clinical trials for sulfated glycosaminoglycan, is used in the ces,’’ which will not only support collabo- vaso-occlusive sickle cell disease (Chang treatment of arthritis. Many recombinant rative research, but will also provide et al., 2010). Siglecs, which are a family of pharmaceuticals, including therapeutic hands-on laboratory training in the cell surface sialic acid-binding lectins, monoclonal antibodies, are glycoproteins, methods of glycosciences to fellows. play a fundamental role in regulating and their specific glycoforms are key to Thus, though our knowledge about the lymphocyte functions and activation. their bioactivity and half lives in circulation biology of glycans and glycomics Recent studies on galectins, a family of and to their possible induction of delete- continues to lag behind more mainstream b-galactoside-binding lectins, have rious immune responses when they do fields of genomics and proteomics, tech- shown that they play a critical role in the not contain the correct glycans. Given this nological advances in glycomics in the

Cell 143, November 24, 2010 ª2010 Elsevier Inc. 675 last 5 years have begun to accelerate the Chang, J., Patton, J.T., Sarkar, A., Ernst, B., Mag- Raman, R., Venkataraman, M., Ramakrishnan, S., integration of into the other nani, J.L., and Frenette, P.S. (2010). Blood 116, Lang, W., Raguram, S., and Sasisekharan, R. 16 major fields of biomedical research. A 1779–1786. (2006). Glycobiology , 82R–90R. complete mechanistic understanding of Cohen, M., and Varki, A. (2010). 14, Reid, G.E., Stephenson, J.L., Jr., and McLuckey, 74 the etiology of almost any disease will 455–464. S.A. (2002). Anal. Chem. , 577–583. 5 depend upon the elucidation of the func- Cummings, R.D. (2009). Mol. Biosyst. , 1087– Rudd, P.M., and Dwek, R.A. (1997). Crit. Rev. Bio- 1104. 32 tions of all posttranslational modifications chem. Mol. Biol. , 1–100. 147 but will especially depend upon our Endo, T. (2010). J. Biochem. , 9–17. Schachter, H., and Freeze, H.H. (2009). Biochim. 1792 understanding the many roles of glycans, Goldberg, D., Sutton-Smith, M., Paulson, J., and Biophys. Acta , 925–930. 5 the most abundant and structurally Dell, A. (2005). Proteomics , 865–875. Schauer, R. (2009). Curr. Opin. Struct. Biol. 19, diverse type of posttranslational modifi- Gupta, G., Surolia, A., and Sampathkumar, S.G. 507–514. (2010). OMICS 14, 419–436. cation. Smith, D.F., Song, X., and Cummings, R.D. (2010). Hart, G.W., Housley, M.P., and Slawson, C. (2007). Methods Enzymol. 480, 417–444. Nature 446, 1017–1022. Spiro, R.G. (2002). Glycobiology 12, 43R–56R. ACKNOWLEDGMENTS Hsu, K.L., Pilobello, K.T., and Mahal, L.K. (2006). Swiedler, S.J., Freed, J.H., Tarentino, A.L., Nat. Chem. Biol. 2, 153–157. We thank Dr. Chad Slawson for helpful sugges- Plummer, T.H., Jr., and Hart, G.W. (1985). J. Biol. Krishnamoorthy, L., and Mahal, L.K. (2009). ACS 260 tions. Original research in the author’s laboratory Chem. , 4046–4054. Chem. Biol. 4, 715–732. was supported by NIH grants R01CA42486, R01 Taniguchi, N. (2008). Mol. Cell. Proteomics 7, DK61671, and R24 DK084949. Lairson, L.L., Henrissat, B., Davies, G.J., and 626–627. Withers, S.G. (2008). Annu. Rev. Biochem. 77, Taylor, A.D., Hancock, W.S., Hincapie, M., Tanigu- 521–555. chi, N., and Hanash, S.M. (2009). Genome Med 1, REFERENCES Lajoie, P., Goetz, J.G., Dennis, J.W., and Nabi, I.R. 57. (2009). J. Cell Biol. 185, 381–385. Tian, Y., Gurley, K., Meany, D.L., Kemp, C.J., and An, H.J., Froehlich, J.W., and Lebrilla, C.B. (2009). Laremore, T.N., Leach, F.E., III, Solakyildirim, K., 13 Zhang, H. (2009). J. Proteome Res. 8, 1657–1662. Curr. Opin. Chem. Biol. , 421–426. Amster, I.J., and Linhardt, R.J. (2010). Methods Aoki-Kinoshita, K.F. (2008). PLoS Comput. Biol. 4, Enzymol. 478, 79–108. Vanderschaeghe, D., Festjens, N., Delanghe, J., and Callewaert, N. (2010). Biol. Chem. 391, e1000075. Ly, M., Laremore, T.N., and Linhardt, R.J. (2010). 149–161. Apweiler, R., Hermjakob, H., and Sharon, N. OMICS 14, 389–399. Varki, A., Cummings, R.D., Esko, J.D., Freeze, (1999). Biochim. Biophys. Acta 1473, 4–8. Moloney, D.J., Panin, V.M., Johnston, S.H., Chen, H.H., Stanley, P., Bertozzi, C.R., Hart, G.W., and J., Shao, L., Wilson, R., Wang, Y., Stanley, P., Bertozzi, C.R., and Sasisekharan, R. (2009). Glyco- Etzler, M.E. (2009). Essentials of Glycobiology, Irvine, K.D., Haltiwanger, R.S., and Vogt, T.F. mics. In Essentials of Glycobiology, Second Second Edition (Cold Spring Harbor, NY: Cold (2000). Nature 406, 369–375. Edition, A. Varki, R.D. Cummings, J.D. Esko, H.H. Spring Harbor Laboratory Press). Freeze, P. Stanley, C.R. Bertozzi, G.W. Hart, and North, S.J., Jang-Lee, J., Harrison, R., Canis, K., Yoshida-Moriguchi, T., Yu, L., Stalnaker, S.H., M.E. Etzler, eds. (Cold Spring Harbor, NY: Cold Ismail, M.N., Trollope, A., Antonopoulos, A., Davis, S., Kunz, S., Madson, M., Oldstone, M.B., Spring Harbor Laboratory Press). Pang, P.C., Grassi, P., Al-Chalabi, S., et al. Schachter, H., Wells, L., and Campbell, K.P. (2010). Methods Enzymol. 478, 27–77. Cantarel, B.L., Coutinho, P.M., Rancurel, C., (2010). Science 327, 88–92. Bernard, T., Lombard, V., and Henrissat, B. Packer, N.H., von der Lieth, C.W., Aoki-Kinoshita, Zaia, J. (2010). OMICS 14, 401–418. (2009). Nucleic Acids Res. 37(Database issue), K.F., Lebrilla, C.B., Paulson, J.C., Raman, R., D233–D238. Rudd, P., Sasisekharan, R., Taniguchi, N., and Zeidan, Q., and Hart, G.W. (2010). J. Cell Sci. 123, 8 Ceroni, A., Maass, K., Geyer, H., Geyer, R., Dell, A., York, W.S. (2008). Proteomics , 8–20. 13–22. and Haslam, S.M. (2008). J. Proteome Res. 7, Paulson, J.C., Blixt, O., and Collins, B.E. (2006). Zielinska, D.F., Gnad, F., Wisniewski, J.R., and 1650–1659. Nat. Chem. Biol. 2, 238–248. Mann, M. (2010). Cell 141, 897–907.

676 Cell 143, November 24, 2010 ª2010 Elsevier Inc.