activity-based protein profiling: from enzyme chemistry to ... · posed that the β-lactam nucleus...

ANRV345-BI77-17 ARI 28 April 2008 15:9

Activity-BasedProtein Profiling: FromEnzyme Chemistryto Proteomic ChemistryBenjamin F. Cravatt,1 Aaron T. Wright,1

and John W. Kozarich2

1The Skaggs Institute for Chemical Biology and Department of Chemical Physiology,The Scripps Research Institute, La Jolla, California 92037; email: [email protected] Biosciences, La Jolla, California 92037; email: [email protected]

Annu. Rev. Biochem. 2008. 77:383–414

First published online as a Review in Advance onMarch 26, 2008

The Annual Review of Biochemistry is online atbiochem.annualreviews.org

This article’s doi:10.1146/annurev.biochem.75.101304.124125

Copyright c© 2008 by Annual Reviews.All rights reserved

0066-4154/08/0707-0383$20.00

Key Words

affinity label, mass spectrometry, proteomics

AbstractGenome sequencing projects have provided researchers with a com-plete inventory of the predicted proteins produced by eukaryoticand prokaryotic organisms. Assignment of functions to these pro-teins represents one of the principal challenges for the field of pro-teomics. Activity-based protein profiling (ABPP) has emerged as apowerful chemical proteomic strategy to characterize enzyme func-tion directly in native biological systems on a global scale. Here, wereview the basic technology of ABPP, the enzyme classes address-able by this method, and the biological discoveries attributable to itsapplication.

383

Click here for quick links to Annual Reviews content online, including:

• Other articles in this volume• Top cited articles• Top downloaded articles• Our comprehensive search

FurtherANNUALREVIEWS

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

Contents

INTRODUCTION. . . . . . . . . . . . . . . . . 384THE BASIC TECHNOLOGY

OF ABPP . . . . . . . . . . . . . . . . . . . . . . . . 386General Design Features

for ABPP Probes . . . . . . . . . . . . . . 386Analytical Platforms for ABPP . . . . 386Types of Biological Experiments

that Can Be Performedwith ABPP . . . . . . . . . . . . . . . . . . . . 390

ENZYME CLASSESADDRESSABLE BY ABPP . . . . . . 392Serine Hydrolases . . . . . . . . . . . . . . . . 392Cysteine Proteases . . . . . . . . . . . . . . . 396Metallohydrolases . . . . . . . . . . . . . . . . 398Additional Classes of Hydrolases . . 400Kinases and Nucleotide-Binding

Proteins . . . . . . . . . . . . . . . . . . . . . . 400Glycosidases . . . . . . . . . . . . . . . . . . . . . 405Cytochrome P450s . . . . . . . . . . . . . . . 406Phosphatases . . . . . . . . . . . . . . . . . . . . . 406Nondirected Probes that Target

Multiple Enzyme Classes . . . . . . 407

INTRODUCTION

Explosive advances in genomics have spawnedmultidisciplinary efforts to develop new tech-nologies for proteomics. To date, genome se-quences of 639 organisms have been com-pleted, including three primates—rhesus,chimpanzee, and man. Over 2000 othergenome projects are underway. This embar-rassment of riches for proteomics has put insharp relief the daunting challenge of assign-ing functional, evolutionary, and interactiverelationships to the full complement of pro-teins (the proteome) encoded by genomes.The context and temporal variability of theproteome distinguishes it from the back-ground of a relatively stable genome and lay-ers a manifold of complexity on the problemof developing large-scale methods for proteincharacterization.

Genomic signatures, such as chromoso-mal translocation and gene amplification,have provided some insights into protein ex-pression (1, 2). Other genomic technolo-gies, including transcriptional profiling andRNA-interference-based gene silencing, haveprovided insights into physiological andpathological processes, e.g., tumorigenesis (3,4), bacterial pathogenesis (5, 6), and insulinsignaling (7). However, these methods relyon profiling and manipulation of gene ex-pression to deduce protein function and donot generally capture the myriad of posttrans-lational events that regulate protein activ-ity (8). Sparked by this necessity, proteomicshas introduced several strategies for theglobal analysis of protein expression and func-tion. Liquid chromatography-mass spectrom-etry (LC-MS) platforms for shotgun analysis(9, 10), yeast two-hybrid methods (11), andprotein microarrays (12) have greatly en-riched our understanding of the expressionpatterns, interaction maps, and in vitro func-tional properties of proteins. However, an ac-curate assessment of the functional state ofproteins in cells and tissues requires moredirect methods to assess protein activity.Activity-based protein profiling (ABPP) hasemerged as a key technology in the evolutionof functional proteomics.

ABPP relies on the design of active-site-directed covalent probes to interrogate spe-cific subsets (families) of enzymes in com-plex proteomes and to provide the basis fora quantitative readout of the functional stateof individual enzymes in the family. Proto-typic ABPP probes target a large, but man-ageable, fraction of the enzyme proteome,often defined by shared catalytic features.The size of the enzyme subset, however, canrange from a few proteins to several hundred.ABPP probes utilize a range of chemical scaf-folds, including mechanism-based inhibitors,protein-reactive natural products, and generalelectrophiles. Because the activity state of anenzyme is regulated by a variety of posttrans-lational mechanisms (8, 13), chemical probes

384 Cravatt ·Wright · Kozarich

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

that report on the structure and reactivity ofenzyme active sites in cells and tissues canprovide high-content proteomic informationthat is beyond the reach of standard expres-sion profiling technologies.

Although ABPP derives its enormous po-tential from recent genomic advances, itsroots extend back for nearly a century. ABPPevolved out of the preconvergence of organicchemistry and enzymology (14). Motivated bya desire to glean mechanistic and structuralinformation on the nature of enzymatic catal-ysis, enzyme chemists studied the modifica-tion of purified proteins by reagents rangingfrom simple organic compounds to affinity la-bels to elaborate mechanism-based inhibitors.With the goal of achieving a physical chemist’ssimplicity of experimental design, major ef-fort was focused on the up-front purificationof specific proteins by often heroic protocolsto isolate a few units of active enzyme fordownstream modification and analysis. Theresults were impressive; the work of Balls &Jensen (15–16) and their coworkers (17–19)in the early 1950s on the stoichiometric inhi-bition of the serine hydrolases, chymotrypsin,trypsin, and cholinesterase by diisopropyl flu-orophosphate was relevant to the future de-sign of ABPP probes. Subsequent work onthe irreversible inactivation of serine, cys-teine, and threonine proteases by a wide vari-ety of molecules has been extensively reviewed(20).

There is no doubt that early experiments inthe style of ABPP were performed before theword “proteomics” was invented. A case canarguably be made that the first such experi-ments were reported in the early 1970s usingthe classic serine-modifying antibiotic peni-cillin (21). Tipper & Strominger (22) had pro-posed that the β-lactam nucleus of penicillinwas a structural analog of the D-ala-D-ala ter-minus of bacterial peptidoglycan and that thechemically reactive β-lactam was positionedto react covalently with the catalytic serineof the transpeptidase(s) responsible for cell

wall biosynthesis. Using radiolabeled peni-cillin G, whole cells or membrane fractions,and primitive gel electrophoretic methods bytoday’s standards, Strominger and colleagues(23, 24) established that multiple penicillin-binding proteins (termed PBPs) are presentin a bacterium. The number and functionof these serine hydrolases vary in differentbacteria. Using radiolabeled penicillin G asa probe, they were subsequently able to pro-file the PBP selectivity of other penicillin andcephalosporin analogs and developed covalentaffinity chromatography methods for captur-ing and releasing PBPs for rapid purification(25, 26). Ultimately, this early approach linkedenzyme class function to biology by demon-strating that, by selectively inhibiting or mu-tating single PBPs, one could infer specificroles for PBPs in cell wall biosynthesis andcell shape determination (27). The limitingfactor in these studies was the difficulty in ob-taining protein sequence information and inlinking it to genomics information, which wasnonexistent at that time.

The advances in protein separation tech-nology and mass spectrometry over the past30 years, coupled with genomics, now allowroutine downstream deconvolution of com-plex proteomic samples. ABPP probes play acrucial role in fractionating these samples onthe basis of a key aspect of enzyme function—catalytic activity. In addition, the interplay be-tween new ABPP methods and the down-stream analytical technologies has created asynergistic cycle of advancement of both dis-ciplines. To date, probes have been developedfor more than a dozen enzyme classes, in-cluding proteases, kinases, phosphatases, gly-cosidases, and oxidoreductases. They havecontributed to our understanding of enzymeactivity in specific physiological and patho-logical processes on a proteome-wide scale.Uncharacterized enzymes for which no pre-vious function had been assigned abound inthese profiles, providing new insight into theirbiology.

www.annualreviews.org • Activity-Based Protein Profiling 385

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

Several excellent reviews on ABPP haveappeared (28–33). The purpose of this reviewis to describe the most recent advances andplace them thematically within the field.

THE BASIC TECHNOLOGYOF ABPP

Before describing specific examples of ABPPprobes and their biological applications, webriefly review the general design concepts andanalytical platforms that support the ABPPtechnology.

General Design Featuresfor ABPP Probes

The fundamental building blocks of ABPPare small-molecule probes that covalently la-bel the active site of a given enzyme or en-zymes. Ideal ABPP probes would target alarge, but manageable, number of enzymes(tens to hundreds) to provide researchers witha global view of the functional state of theproteome. This level of target promiscuitymust, however, be counterbalanced by limitedcross-reactivity with other proteins in the pro-teome. Most ABPP probes achieve a desiredlevel of intraclass coverage and minimal extra-class cross-reactivity by displaying a combina-tion of reactive and binding groups that targetconserved mechanistic or structural featuresin enzyme active sites (Figure 1). Reactivegroups can be further divided into two gen-eral classes: (a) electrophilic groups that mod-ify conserved active-site nucleophiles, and(b) photoreactive groups that label proximalresidues in enzyme active sites following UVirradiation. ABPP probes must also possessa third element, a reporter tag, to facilitatetarget characterization (Figure 1). Examplesof reporter tags include fluorophores, biotin,and latent analytical handles such as alkynesor azides, which can be modified by clickchemistry [copper-catalyzed Huisgen’s azide-alkyne cycloaddition (34)] methods to visual-ize protein targets postlabeling (35, 36). As iselaborated in the following section, the iden-

Reactivegroup

Spacer/binding group

NN

S

O

CO2H

N3

ON N

CO2H

a ABPP probe:

b Tags:

Tag

Figure 1(a) Representative structure of an ABPP probe,which contains a reactive group ( green), a spaceror binding group (black), and a reporter tag( purple). (b) A variety of reporter tags can be usedfor enzyme visualization and enrichment,including fluorophores (e.g., rhodamine) andbiotin, as well as “clickable” handles, such as azidesand acetylenes.

tity of the reporter tag dictates experimen-tal options for downstream analysis in ABPPexperiments.

Analytical Platforms for ABPP

With a proficient ABPP probe in hand, re-searchers can investigate proteomes using anumber of analytical platforms. These plat-forms exhibit distinct advantages and lim-itations that must be matched with the


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

Healthy

(kD

a)

ABPP probe

Diseased

150

75

50

37

25

SDS-PAGE

In-gel fluorescencescanning

Figure 2Gel-based activity-based protein profiling (ABPP), where probe-labeled enzymes are visualized andquantified across proteomes by in-gel fluorescence scanning.

specific experimental problem to maximizethe information content acquired in ABPPinvestigations.

Gel electrophoresis platforms for ABPP.Original versions of ABPP utilized gel-basedmethods for the detection of enzyme activi-ties (37–40). In this approach, probe-treatedproteomes are first resolved by one- (1D)or two-dimensional (2D) polyacrylamide gelelectrophoresis (PAGE), and the labeled en-zymes then are visualized by either in-gel flu-orescence scanning (for fluorescent probes)or avidin blotting (for biotinylated probes)(Figure 2). A typical functional proteomic in-vestigation would likely employ both modesof detection, first using fluorescent probesfor rapid screening of proteomes by 1D-sodium dodecyl sulfate (SDS)-PAGE, andthen applying biotinylated probes in com-bination with (strept)avidin chromatography,gel separation, and mass spectrometry analysisto identify probe-labeled enzymes. Althoughhigher-resolution, “gel-free” analytical plat-forms have since emerged for ABPP (see be-low), gel-based methods are still widely em-ployed today owing to their robustness andthroughput. Indeed, hundreds of proteomescan be analyzed per day by an individual re-search group using fluorescent probes and

1D-SDS-PAGE, a level of throughput thatvastly exceeds most other methods. Gel-basedABPP is thus still preferred for research prob-lems that require the rapid comparative anal-ysis of many proteomes in parallel.

Liquid chromatography-mass spectrome-try (LC-MS) platforms for ABPP. Recog-nizing that the principal limitation of gel-based methods was resolution, researchershave introduced multiple LC-MS strategiesfor analyzing probe-treated proteomes thatvastly improve the information content ac-quired in ABPP experiments. LC-MS ap-proaches for ABPP can be generally dividedinto two categories—those that analyze theprotein targets of probes and those thatspecifically analyze probe-labeled peptidesderived from these targets. The former strat-egy involves using biotinylated probes tolabel proteomes, which are then directlyincubated with (strept)avidin beads to en-rich probe-labeled proteins (41) (Figure 3a).Enriched proteins are then digested on-bead with trypsin and analyzed by mul-tidimensional LC-MS/MS. This approach,which essentially represents a fusion of ABPPand the multidimensional protein identifica-tion technology (ABPP-MudPIT), has provencapable of identifying 50–100+ enzyme


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

Nu NuPeptideidentification

a

ABPP probe (Strept)avidinenrichment

Trypsin LC/MS

Nu NuLabeling siteidentification

b

ABPP probe (Strept)avidinenrichment

Trypsin LC/MS

Nu

Nu

Figure 3Liquid chromatography-mass spectrometry (LC/MS)-based platforms for activity-based proteinprofiling (ABPP). (a) ABPP and multidimensional protein identification technology (41), wherebyprobe-labeled enzymes are first captured on (strept)avidin beads and then digested with trypsin and theresulting peptide mixture analyzed by LC-MS/MS. (b) Active-site peptide profiling (44, 45), wherebyprobe-treated proteomes are digested with trypsin prior to affinity enrichment. Enriched peptides arethen eluted with organic solvents and analyzed by LC-MS/MS.

activities in individual proteomes. Moreover,the relative levels of enzyme activities intwo or more proteomes can be determinedby ABPP-MudPIT using simple, semiquan-titative parameters such as spectral counting(41–43).

One shortcoming of ABPP-MudPIT isthat this method does not offer a straight-forward way to identify the probe-labeledpeptides of enzyme targets. This goal hasbeen achieved using a distinct LC-MS plat-form, termed active-site peptide profiling(ASPP), that involves digesting probe-treatedproteomes with trypsin prior to incubationwith affinity enrichment resins (Figure 3b)(44, 45). Probe-labeled peptides are thus se-lectively enriched using either (strept)avidinor antibody resins (for biotinylated and fluo-rescent probes, respectively) and then elutedwith organic solvents and analyzed by LC-

MS/MS. Modified versions of the SEQUESTsearch algorithm are employed to simultane-ously identify enzyme targets and their spe-cific sites of probe labeling. If ABPP probesthat contain a disulfide linker are utilized, la-beled peptides can be eluted with excess freethiols and analyzed by LC-MS/MS (46).

ABPP-MudPIT and ASPP possess com-plementary advantages and limitations. Be-cause the former method analyzes manytryptic peptides for each probe target, the like-lihood of accurately identifying and quanti-fying proteins improves. In contrast, the siteof labeling data provided by ASPP greatlystrengthens confidence that probes are target-ing active sites to provide a legitimate read-out of the functional state of enzymes. Couldboth sets of information be acquired in asingle ABPP experiment? Speers & Cravatt(47) broached this subject by introducing


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

“Clickable”ABPP probe

Click chemistry

(Strept)avidinenrichment

Trypsin

TEV protease

LC/MSprotein 1D

LC/MS(labelingsite 1D)

TEV-tag

N3

NN

N

NN

N

NN

N

NN

N

m/z

Figure 4Tandem-orthogonal proteolysis (TOP)-activity-based protein profiling (ABPP) for simultaneouscharacterization of protein targets of probes and sites of probe labeling (47). Abbreviations: 1D, onedimensional; LC-MS, liquid chromatography-mass spectrometry; m/z, mass/charge; TEV, Tobacco etchvirus.

a Tobacco etch virus (TEV) protease cleav-age site between the reactive group and bi-otin tag of ABPP probes. Specifically, theresearchers made use of click chemistry toconjugate an azide-TEV cleavage peptide-biotin tag onto enzymes labeled by an alkyne-sulfonate ester ABPP probe (Figure 4). Theimplementation of click chemistry meth-ods (35, 36) circumvented potential prob-lems that the large reporter tag might havecaused for probe activity. Probe-labeled pro-teins were then captured on (strept)avidinbeads and subject to tandem digestions withtrypsin and TEV protease to release bulk (un-labeled) and probe-labeled peptides. Theseprotease digestions were then analyzed in se-quential LC-MS runs. Cross-correlating thedata from each protease digestion greatly im-proved confidence in protein assignments,allowing for the removal of false-positivesignals and the identification of unantici-pated targets of ABPP probes. This plat-form, dubbed tandem-orthogonal proteol-

ysis (TOP)-ABPP, thus allows researchersto simultaneously profile enzyme targets ofABPP probes and their specific sites of probemodification. Recently, chemical strategieshave been introduced for the cleavage ofprobe-modified proteins/peptides from affin-ity resins (48, 49), which should serve as usefulcomplements to the TOP-ABPP method.

It is important to recognize that the su-perior resolution and information contentafforded by LC-MS approaches for ABPPcome with a cost to investigators. First, LC-MS requires much larger quantities of pro-teome compared to gel-based methods (0.5–1.0 mg versus 0.01–0.02 mg, respectively).Although for many studies, including thosethat analyze rodent tissues and cell lines, pro-teome may be in ample supply, other samples,such as primary human biopsies, are of a fi-nite amount, which may hinder their analysisby LC-MS. LC-MS methods for ABPP arealso much slower than 1D-SDS-PAGE, espe-cially when performed to maximize resolution


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

(i.e., multidimensional separations, which re-quire several hours per sample). Thus, eventhough LC-MS would certainly be a pre-ferred method for the in-depth comparisonof a handful of proteomes, the analysis ofdozens or hundreds of samples will likely re-quire other approaches (or access to severalLC-MS instruments).

Emerging platforms for high-throughput,high-resolution ABPP. An ideal ABPP plat-form would combine the throughput andminimal sample requirements of gel-basedanalysis with the resolution afforded by LC-MS. Although no such platform yet exists,substantial progress has been made towardthis goal by combining ABPP with analyt-ical methods that require neither gel norMS readouts. One such strategy exploitscapillary electrophoresis (CE) coupled withlaser-induced fluorescence (LIF) scanning asseparation and detection techniques, respec-tively (45). In this method, proteomes aretreated with fluorescent probes, digested withtrypsin, and the resulting probe-labeled pep-tides enriched with antifluorophore antibodyresins. Enriched peptides are then eluted andanalyzed by CE-LIF, which was shown toprovide vastly superior resolution comparedto 1D-SDS-PAGE, especially for enzymetargets that share similar molecular mass.CE-LIF ABPP also has the advantages of con-suming minimal amounts of sample and of po-tentially achieving high throughput. Indeed,CE run times are very short (15–20 min), andmany samples can be analyzed in parallel on96-channel instruments. One potential draw-back of CE-LIF ABPP is that the identity ofenzyme targets initially remains unclear. Al-though this information can be obtained incomplementary LC-MS experiments, even-tually leading to unequivocal assignment ofindividual CE peaks to specific enzyme tar-gets, this process is slow. CE-LIF ABPP maytherefore be most applicable for the repeti-tive analysis of well-characterized proteomes,as is required when screening large chemicallibraries for target selectivity.

A second emerging platform that holds po-tential for satisfying the requirements of low-sample demand and high-throughput/high-resolution analysis is the ABPP microarray(50). In this method, antibodies that specif-ically recognize enzyme targets of ABPPprobes are arrayed on glass slides and usedas capture tools. Once bound to their cognateantibodies on the microarray, probe-labeledenzymes can be directly detected by fluo-rescence scanning. ABPP microarrays wereshown to exhibit significantly improved sensi-tivity compared to gel-based methods for thedetection of protease activities in proteomes.Additionally, minimal amounts of proteome(<0.01 mg) were required for ABPP microar-ray experiments. The throughput of ABPPmicroarrays is also potentially excellent, asmany antibodies can be arrayed in parallel ona single slide. Finally, ABPP microarrays, byremoving the requirement for a second anti-body detection agent, address one of the maintechnical challenges facing the implementa-tion of standard protein microarrays. Theprincipal limitation that currently hinders theroutine application of ABPP microarrays isa dearth of commercially available antibod-ies that can selectively recognize enzyme tar-gets in this assay format. Alternative microar-ray formats for ABPP have also been intro-duced that couple probes to peptide-nucleicacid tags (51). Probe-labeled enzymes arethen captured by hybridization to arrays bear-ing complementary oligonucleotides. Usingthis technology, Harris and colleagues (52)identified a selective inhibitor of the dustmite protease Der p 1 and provided evi-dence that this protease plays a role in allergyprogression.

Types of Biological Experimentsthat Can Be Performed with ABPP

As the methods for performing ABPP havediversified and matured, multiple types of bi-ological applications for this technology haveemerged. In this section, we briefly summa-rize the most common uses of ABPP. Specific


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

examples of the biological impact of ABPPapplied in these various formats are coveredin the following section on enzyme classes ad-dressable by ABPP.

Comparative ABPP for target discovery.The original, and still most common appli-cation of ABPP is for “target discovery” inbiological systems. A typical target discoveryexperiment would comparatively analyzetwo or more proteomes by ABPP to identifyenzymes with differing levels of activity(Figure 2). If the proteomes under com-parison in turn display distinct biologicalproperties (e.g., healthy versus normal), thenthe altered enzyme activities identified byABPP can be hypothesized to regulate thesephenotypes. The testing of such hypotheses,of course, requires further experimentation;examples are detailed below. In comparisonto more conventional expression-based ge-nomics and proteomics, ABPP has multipleadvantages for target discovery experiments.First, ABPP accounts for myriad posttrans-lational mechanisms that regulate enzymeactivity (but not necessarily expression) inliving systems. Second, because ABPP probeslabel enzymes using conserved active-sitefeatures rather than mere expression level,these reagents provide exceptional access to

low-abundance proteins in samples of highcomplexity.

Competitive ABPP for inhibitor discov-ery. ABPP has been adapted for a secondmajor application—the discovery of enzymeinhibitors. This pursuit involves performingABPP in a competitive mode, during whichinhibitors are identified by their ability toblock probe labeling of enzymes (38, 53, 54)(Figure 5). Competitive ABPP offers sev-eral advantages over conventional inhibitorscreening methods. First, enzymes are testedin native proteomes, alleviating the need forrecombinant expression and purification. Sec-ond, probe labeling acts as a surrogate for sub-strate assays, meaning that novel enzymes thatlack known substrates are amenable to anal-ysis. Finally, because ABPP tests inhibitorsagainst many enzymes in parallel, potency andselectivity factors can be simultaneously as-signed to these compounds. Both reversible(54) and irreversible (53) inhibitors of en-zymes can be characterized by competitiveABPP, although the details of the assay for-mat differ slightly. Reversible inhibitors mustbe screened under kinetically controlled con-ditions where probe labeling of enzyme tar-gets has not yet reached completion. Anal-ysis of irreversible inhibitors, which is morestraightforward and can even be performed in

1 2 3 4

Inhibitor sensitivity

ABPP probeInhibitors SDS-PAGE

Figure 5Inhibitor discovery by competitive activity-based protein profiling (ABPP). The selectivity and potencyof enzyme inhibitors can be determined by initial incubation of a proteome with inhibitors followed byprobe treatment. Inhibitor-bound enzymes are detected by a reduction in probe labeling intensity.Abbreviation: SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis.


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

vivo, involves preincubation with proteomesfor various amounts of time, after which resid-ual enzyme activity is monitored by additionof ABPP probes (43, 55, 56).

Characterization of enzyme active sitesby ABPP. Genome sequencing projects haverevealed that eukaryotic and prokaryoticorganisms possess a remarkable numberof unannotated proteins. ABPP offers apotentially powerful means to gain functionalinsights into uncharacterized enzymes. Forexample, discovery that an uncharacterizedenzyme reacts with an ABPP probe mayfacilitate assignment of the enzyme to aspecific mechanistic class (57, 58). Similarly,characterization of unanticipated sites oflabeling in enzyme active sites can designatepotential catalytic roles for residues of pre-viously unknown function (44). Finally, thereactivity profile of uncharacterized enzymeswith libraries of ABPP probes offers insightsinto active-site recognition, which may inturn lead to hypotheses about the structuresof endogenous substrates and products (59).

ENZYME CLASSESADDRESSABLE BY ABPP

Research efforts over the past decade haveengendered ABPP probes for numerous en-zyme classes. These probes have been used togarner striking insights into enzyme functionin a wide range of biological systems. Here,we summarize the progress made to date inthe development and biological application ofABPP probes for individual enzyme classes.

Serine Hydrolases

Serine hydrolases represent a large and di-verse enzyme class that includes proteases,peptidases, lipases, esterases, and amidases.These enzymes collectively constitute ap-proximately 1% of the predicted pro-tein products encoded by most eukaryoticgenomes. Serine hydrolases are united by acommon catalytic mechanism that involves

activation of a conserved serine nucleophilefor attack on a substrate ester/thioester/amidebond to form an acyl-enzyme intermediate,followed by water-catalyzed hydrolysis of thisintermediate to liberate the product. Thegreatly enhanced nucleophilicity of the cat-alytic serine renders it susceptible to covalentmodification by many types of electrophiles,including fluorophosphonates (FPs) and arylphosphonates, sulfonyl fluorides, and car-bamates. Phosphonate electrophiles haveemerged as a particularly powerful class ofaffinity labels for the design of ABPP probesthat target serine hydrolases (Figure 6).

Phosphonate probes for serine hydrolases.Reporter-tagged FPs (Figure 6a) exhibit re-markably broad reactivity with enzymes fromthe serine hydrolase class (37–39). To date,more than 80 distinct serine hydrolases havebeen identified in the literature as targets ofFP probes in human and mouse proteomes(41, 43, 45, 60, 61). A provocative number ofthese enzymes are uncharacterized in terms oftheir endogenous substrates and products, re-flecting that less than half of the members ofthe mammalian serine hydrolase family have adefined metabolic function. Importantly, FPshave been shown to serve as bona fide activity-based probes for serine hydrolases, reactingwith active enzymes but not with their in-active precursor (i.e., zymogen) or inhibitor-bound forms (38, 60). This factor has enabledresearchers to identify serine hydrolases thatshow altered activity, but not expression, in bi-ological systems (60, 61). The activity-basednature of FP labeling has also bolstered confi-dence in the functionality of serine hydrolasesthat lack known substrates, such as the pre-dicted peptidase PREPL, which is deleted inhypotonia-cystinuria syndrome (62). Finally,high-resolution MS platforms have been de-veloped to map the precise sites of FP probelabeling (45), facilitating the identification ofnovel, sequence-unrelated members of theserine hydrolase class (58).

Although the selectivity of FP probescan be partially tuned by varying the linker


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

a b

PHN

O

O

O

HO2C

O

N

O

NH

OHO

OHN

O

NH

TAG5

P

OHN O

OO

NH2

NH

TAG

H2NO

O

PF

ONH

O

Linker TAG

O

PF

OO

OO

OHN

O

Linker TAG

Figure 6Electrophilic phosphonates as prototypic activity-based protein profiling probes for serine hydrolases.(a) Representative fluorophosphonate (FP) probes (37, 38). (b) Representative aryl phosphonate probes(63, 64).

unit that connects this reactive group to thereporter tag [i.e., switching from an alkyl toa pegylated chain (38) (Figure 6a)], theseprobes, in general, show little dependence onadditional binding groups for reactions withmost serine hydrolases. There are exceptions,however, including certain serine proteasesthat display restricted substrate selectivitiesthat reduce their rates of labeling with genericFP probes. To address this problem, peptidicarylphosphonate ABPP probes have beenintroduced (Figure 6b) that show enhancedrates of labeling (63) and, in certain cases,high specificity for individual serine proteases(64). Collectively, these studies suggest thatelectrophilic phosphonates should offer auniversal strategy to profile the activity of es-sentially all members of the serine hydrolasesuperfamily.

Biological applications of ABPP probes forserine hydrolases. ABPP has been used toprofile the activity of serine hydrolases in nu-merous biological settings, including cancer

(41, 60, 61, 65), atherosclerosis (66), immunecell activation (64), and nervous system signal-ing (56, 67). In certain instances, these studieshave facilitated the assignment of metabolicand cellular functions to previously uncharac-terized members of the SH family (68).

Profiling serine hydrolase activities in can-cer. Many serine hydrolases, including pro-teases, lipases, and esterases, have been pos-tulated to play important roles in cancer.These enzymes, by controlling extracellu-lar matrix structure, growth factor activation,and the metabolism of small-molecule sig-nals, may contribute to the malignant be-havior of aggressive cancer cells. To explorethis premise further, Jessani and colleagues(60) globally profiled serine hydrolase activ-ities across a panel of human breast carci-noma and melanoma cell lines that differedin pathogenic properties. The researchersfirst used fluorescent FP probes to gener-ate gel-based profiles for the quantitativecomparison of enzyme activities and then


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

O OH

O

O

HO

O OH

OH

O O

OH

P

O

OOH

N

O O

OH

P

O

OHOH

2-Acetyl MAGE

MAGE

Alkyl-LPC

Alkyl-LPA

KIAA1363

Figure 7The uncharacterized serine hydrolase KIAA1363, which was found byactivity-based protein profiling to be elevated in aggressive cancer cells,regulates an ether lipid signaling network that includes monoalkylglycerolethers (MAGEs), alkyl-lysophosphatidylcholine (LPC), and alkyl-lysophosphatidic acid (LPA) (68).

biotinyated FP probes coupled with avidinchromatography and LC-MS analysis to en-rich and identify these enzymes. Hierarchi-cal clustering analysis identified enzyme ac-tivity signatures that segregated cancer cellsinto subgroups on the basis of tumor of originor state of invasiveness. These enzyme activi-ties included known markers of cancer malig-nancy, such as the serine protease urokinaseplaminogen activator (uPA) and uncharacter-ized proteins, such as the integral membraneprotein KIAA1363, for which no prior associ-ation with cancer had been made. Subsequentstudies using ABPP-MudPIT confirmed thatKIAA1363 activity is highly elevated in ag-gressive estrogen receptor-negative [ER(−)]primary human breast tumors relative to lessaggressive ER(+) tumors or normal breast tis-sue (41).

Intrigued by the strong correlation be-tween KIAA1363 activity and the pathogenicstate of cancer cells, Chiang and colleagues(68) pursued the functional characterizationof this enzyme. A potent and selectivecarbamate inhibitor of KIAA1363 was de-veloped using competitive ABPP methodsand applied to aggressive cancer cell linesthat contain high levels of the enzyme.Analysis of the inhibitor-treated cells us-ing global metabolite profiling methods(69) revealed a selective decrease in anunusual class of lipids—the monoalkylglyc-erol ethers (MAGEs). Biochemical studiesconfirmed that KIAA1363 regulates MAGElevels in cancer cells by hydrolyzing theprecursor lipid 2-acetyl MAGE (Figure 7).Interestingly, MAGEs were further convertedby cancer cells into alkyl-lysophospholipids,including lysophosphatidic acid (LPA)(Figure 7), a bioactive lipid known to pro-mote the pathogenic properties of cancer cells(70). Disruption of KIAA1363 expression byRNA interference impaired this ether lipidnetwork, leading to reductions in cancer cellmigration and tumor growth. These studiesthus indicate that the KIAA1363-ether lipidnetwork supports the aggressive behaviorof cancer cells. Concurrent with these stud-ies, Nomura and colleagues (67) used FPprobes to identify KIAA1363 as a principaldegradative enzyme for organophosphorusnerve toxins in rodent brains, indicating thatthe enzyme may play important functions inxenobiotic metabolism in normal tissues.

In addition to analyzing human cancer celllines in culture, ABPP has been used to pro-file these cells grown in vivo as xenograft tu-mors in immune-deficient mice (61). Thesestudies enabled identification of both tumor-and host-derived enzyme activities that werealtered during tumor progression. This in-vestigation also uncovered a hyperaggressivevariant of the human breast cancer line MDA-MB-231, which was isolated from xenografttumors grown in the mammary fat pad (mfp)and found to display greatly enhanced tumor-forming and metastatic potential in vivo.


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

ABPP of these “231mfp” cells revealed thatthey possess dramatically elevated levels ofmultiple secreted serine protease activities, in-cluding uPA and tissue-plasminogen activator(tPA). Interestingly, transcript levels of uPAand tPA were essentially unaltered comparedto parental MDA-MB-231 cells, thus pro-viding a compelling case wherein ABPP gar-nered functional proteomic information thatwas not reflected in gene expression profiles.

Profiling serine protease activities in im-mune cell activation. Serine proteases ofthe granzyme subclass have been postulatedto play major roles in cytotoxic lymphocyte[natural killer (NK) cell]-mediated cell death.However, the relative contributions made byindividual granzymes to NK cell function re-main unclear owing in large part to a lackof selective chemical inhibitors for these en-zymes. Mahrus & Craik (64) set out to addressthis problem by first screening five humangranzymes for their ability to cleave mem-bers of a combinatorial library of peptidesubstrates. Distinct substrate selectivity pro-files were observed for each granzyme, andthis information was used to create substrate-mimetic diphenylphosphonate inhibitors thatselectively target granzyme A and B. Bi-otinylated versions of the phosphonates con-firmed their target selectivity in NK cellproteomes. Interestingly, the granzyme B in-hibitor blocked NK cell-mediated lysis of tar-get cells by more than 75%; in contrast, thegranzyme A inhibitor displayed little efficacy.These data thus argue that granzyme B is thedominant protease involved in effecting targetcell lysis by NK cells.

Determining the proteome-wide selectiv-ity of serine hydrolase-directed inhibitors.Competitive ABPP has been used to eval-uate the selectivity of several serine hydro-lase inhibitors. In an early example of thevalue of this approach, Leung and colleagues(54) tested inhibitors of the integral mem-brane enzyme fatty acid amide hydrolase(FAAH) for their selectivity in mouse tissue

proteomes. FAAH is responsible for degrad-ing the fatty acid amide family of signalinglipids, which includes the endocannabinoidanandamide (71). Genetic (72) or pharma-cological (73) inactivation of FAAH leads toanalgesic, anti-inflammatory, anxiolytic, andantidepressant effects, suggesting that this en-zyme may represent a novel target for sev-eral nervous system disorders. All inhibitorsof FAAH reported to date possess an elec-trophilic carbonyl element that engages theconserved serine nucleophile in the enzyme’sactive site. This mode of inhibition may targetother members of the SH superfamily. Indeed,competitive ABPP using reporter-tagged FPprobes identified several “off targets” for first-generation FAAH inhibitors, including otherlipases, as well as enzymes of uncharacter-ized function (54). Interestingly, none of theseenzymes shared any sequence similarity withFAAH, indicating that sequence homology isnot necessarily a good predictor of active-siterelatedness in enzyme superfamilies. Com-petitive ABPP has since become an inte-gral component of nearly all FAAH inhibitordevelopment efforts, permitting the concur-rent optimization of potency and selectivity.These efforts have culminated in the gen-eration of multiple classes of reversible (74)and irreversible (75) FAAH inhibitors that arehighly specific for this enzyme. Conversely,inhibitors that show broad activity againstmultiple SHs have also been exposed (43). Ir-reversible FAAH inhibitors have themselvesbeen converted into ABPP probes by incor-poration of a “clickable” alkyne group intotheir structures (56). These agents have beenused to identify the direct targets of FAAHinhibitors in vivo.

Inhibitor development programs for otherserine hydrolases have also benefited fromcompetitive ABPP. For example, Nomanbhoyand colleagues (76) examined the proteome-wide selectivity of a panel of inhibitors ofdipeptidylpeptidase IV (DPP4). DPP4 regu-lates the levels of insulin-promoting hormoneglucagon-like peptide 1 (GLP1) and, there-fore, has become an attractive therapeutic


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

target for type II diabetes. DPP4 inhibitorsmust, however, selectively target this enzymeover several closely related peptidase ho-mologs. Competitive ABPP has provided auseful platform to rapidly assess the selectivityof DPP4 inhibitors in native proteomes andhas facilitated the development of selective in-hibitors for other members of the DPP class(e.g., DPP7) (77).

Cysteine Proteases

Like serine hydrolases (SHs), cysteineproteases are a huge enzyme class whosemembers perform myriad critical functions inprokaryotic and eukaryotic organisms. Thedistinct catalytic mechanisms employed bySHs and cysteine proteases, however, makethem susceptible to inactivation by differ-ent electrophilic chemotypes (i.e., cysteineproteases are not labeled by a FP probe).Multiple reactive groups, including epoxides,vinyl sulfones, diazomethyl ketones, α-haloketones, and acyloxymethyl ketones (78),have been incorporated into ABPP probesdeveloped for cysteine proteases. Here,we highlight the most mature versions ofthese probes, the specific subsets of cysteineproteases that they target, and examples oftheir biological applications.

Epoxide probes that target the papain fam-ily of cysteine proteases. Papains are mem-bers of the CA clan of cysteine proteases thatinclude the cathepsins, which play key roles

in a variety of physiological and pathologi-cal processes (78). Cathepsins are nearly uni-versally inactivated by the natural productE-64 (79) (Figure 8a), which contains an acti-vated epoxide that reacts with the cysteine nu-cleophile of these enzymes. E-64 shows verylimited cross-reactivity with other cysteineproteases, which has made it an excellentpharmacological tool for biologists interestedin investigating the function of cathepsins.Bogyo and colleagues (80) have appended arange of reporter tags onto E-64, including ra-dioisotopes, fluorophores (40), and biotin (80)(Figure 8b), to create a series of ABPP probeswith broad utility for the functional proteomicanalysis of cathepsins. More recently, the re-searchers have shown that cathepsins can alsobe targeted by peptidic acyloxymethyl ke-tone probes that contain hydrophobic P1 sub-stituents (81).

Electrophilic ketone probes that target thecaspase family of cysteine proteases. ClanCD is another large group of cysteine pro-teases that includes the caspases, which playkey roles in apoptosis-mediated cell death.Active-site-directed probes have a rich historyof application in the field of caspase biology,including serving as key tools for the discoveryof the first caspase (β-interleukin-convertingenzyme) (82). Caspases have since been tar-geted by a number of ABPP probes, includingpeptidic α-halomethyl and acyloxymethyl ke-tones (81, 83, 84) (Figure 9). These probes

HNH2N

NH2

HN

O

NH

O

O

OH

O

HN

O

NH

O

O

O

O

HO125I

O

NH

HN

O

H2N

O

TAG

5

a E-64 b E-64-based probes

Figure 8(a) Structures of the broad-spectrum cathepsin inhibitor E-64 and (b) activity-based protein profilingprobes derived from this natural product (53, 80).


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

NH

HN

O

TAG NH

O

OH

O

O

X NH

O HN

O

OH

NH

O TAGHN

O

O

O

O

F3C

CF3CO2H

X = Cl, F

a b

Figure 9Representative electrophilic ketone probes that target the caspase family of cysteine proteases.(a) Peptidic α-halomethyl ketones (84). (b) Peptidic acyloxymethyl ketones (81, 83).

achieve selectivity for caspases over othercysteine proteases by incorporating a nega-tively charged substituent in the P1 position.Recently, acyloxymethyl ketone probes wereused to provide evidence for an activated full-length form of caspase-7 that occurs early dur-ing the apoptotic cascade (85), suggesting thatcertain caspases can be converted to func-tional proteases without requiring zymogenprocessing.

Electrophilic ubiquitin probes that tar-get the deubiqutinating family of cysteineproteases. Multiple subfamilies of cysteineproteases are involved in cleaving ubiqui-tin and ubiquitin-like modifications. Theseproteases are often referred to as ubiquitin-specific proteases (USPs) or deubiquitinatingenzymes (DUBs) (86). The wide sequencedivergence among these proteases has in-spired the development of functional pro-teomic methods for their identification andcharacterization. Borodovsky and colleagues(57) have addressed this problem by gen-erating electrophilic ubiquitin derivatives asABPP probes for DUBs (Figure 10). Takingadvantage of intein-based chemical ligationmethods, the researchers synthesized a set ofubiquitin probes containing various Michaelacceptors and alkyl halides and showed thateach probe targets a distinct set of DUBs inthe proteome. These probes have been used toidentify novel classes of DUBs, including theovarian-tumor like domain family (57) and the

UL36 gene product of herpes simplex virustype 1 (87).

Biological applications of ABPP probesthat target cysteine proteases. The afore-mentioned sets of ABPP probes have gainedwidespread use for the functional characteri-zation of cysteine proteases in biological sys-tems. Here, we highlight two prominent ex-amples in the areas of parasitology and cancer.

Identification of cysteine proteases thatcontribute to the host cell invasion by themalaria parasite Plasmodium falciparum.Proteases have long been suspected to playkey roles at various stages of the Plasmod-ium falciparum life cycle. Nonetheless, thedaunting number of proteases (>100)

UbNH

HN

O

SOO

UbNH

HN

O

X

X = Br, Cl

UbNH

HN

O

Br

UbNH

HN

O

OCH3

O

UbNH

HN

O

SOO

UbNH

HN

O

CN

Figure 10Representative electrophilic ubiquitin derivatives that serve as activity-based protein profiling probes for deubiquitinating enzymes (57, 86).Abbreviation: Ub, ubiquitin.


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

encoded by the P. falciparum genome presentsa major challenge for researchers interestedin discerning the functions of individualproteases in this parasite system. Greenbaumand colleagues (88) utilized E-64-basedprobes to profile these papain activities acrossthe parasite life cycle, revealing a selectiveelevation in falcipain 1 activity at the time ofhost invasion. Although falcipain 1 had beenof interest to Malaria researchers for severalyears, the protease has proven difficult tocharacterize in recombinant form. Takingadvantage of competitive ABPP methods,which can be applied to natively expressedenzymes, Greenbaum and colleagues devel-oped a selective inhibitor of falcipain 1 andshowed that this agent blocked red blood cellinvasion by merozoite-stage parasites. Theresearchers have since applied competitiveABPP to generate selective inhibitors forother cathepsins, including cathepsin B (40)and legumain (89).

Identification of cysteine proteases thatcontribute to angiogenic switching and tu-mor growth in pancreatic cancer. Inter-ested in understanding the molecular path-ways that support tumorigenesis, Joyce andcolleagues (55) initially performed globalgene expression analyses of progressive stagesof tumor development in the RIP1-Tag2transgenic mouse model of pancreatic can-cer. Among the gene products that showedelevated expression in tumors were severalcathepsins. ABPP probes were then usedto confirm that specific cathepsins, such ascathepsins Z, B/L, and C, were heightenedin activity in tumors relative to normal or an-giogenic islets. The researchers then applied afluorescent ABPP probe to localize cathepsinactivities within tumors and found that theseproteases resided predominantly in infiltrat-ing immune cells as well as in tumor cells at in-vasive fronts of carcinomas. Broad-spectrumcathepsin inhibitors (with E-64 as a basis)affected multiple stages of tumor develop-ment, including angiogenic switching, tumorvascularity, and invasive growth/proliferation.

These data collectively point to cathepsins askey proteases involved in tumorigenesis anda potentially new set of drug targets for thetreatment of cancer.

Metallohydrolases

A third major class of hydrolytic enzymes isthe metallohydrolases, which includes pro-teases, peptidases, and deacetylases. Unlikeserine or cysteine hydrolases, which useenzyme-bound nucleophiles for catalysis, themetallohydrolases accomplish substrate hy-drolysis by a zinc-activated water molecule.This difference in mechanism complicatesthe design of electrophilic ABPP probes formetallohydrolases. As an alternative strategy,photoreactive variants of reversible inhibitorsof metallohydrolases have been introduced(42, 90, 91). These photoreactive ABPPprobes thus achieve target selectivity throughbinding affinity, rather than “mechanism-based” reactivity. Covalent labeling is ac-complished by exposure to UV light, whichinduces the photoreactive group to mod-ify proteins in close spatial proximity tothe probes. Next, we review the develop-ment and biological application of photore-active ABPP probes for two major classesof metallohydrolases—the metalloproteases(MPs) and histone deacetylases (HDACs).

Photoreactive probes that target MPs.MPs are regulated by multiple posttrans-lational mechanisms in vivo, including zy-mogen activation and inhibition by endoge-nous protein-binding partners (e.g., TIMPs).These factors complicate the analysis of MPsusing conventional genomic or proteomicmethods. MPs and the matrix metallopro-teinases (MMPs), in particular, have beenthe subject of intense study by the pharma-ceutical industry for more than a decade.These efforts have produced several broad-spectrum, tight-binding reversible inhibitorsof MMPs, most of which contain a hydrox-amic acid (hydroxamate) group that coordi-nates the conserved active-site zinc atom in


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

a bidentate manner (92) (Figure 11a). Thehydroxamate-based inhibitor GM6001 hasserved as a scaffold for first-generation ABPPprobes of MPs. Specifically, Saghatelian andcolleagues (90) replaced the hydrophobic P2group of GM6001 with a photoreactive ben-zophenone and appended a rhodamine re-porter tag onto the molecule. The resultingprobe, dubbed HxBP-Rh (Figure 11b), wasfound to selectively label active, but not zy-mogen or inhibitor-bound, forms of MMPs.Interestingly, the probe, as well as its par-ent inhibitor GM6001, also targeted multi-ple MPs outside the MMP class, includingthe metallopeptidases neprilysin and leucineaminopeptidases. These peptidases share verylow sequence homology with MMPs. Thus,as was observed for SHs, MPs that are dis-tantly related by sequence have been found todisplay significant overlap in their inhibitorsensitivity profiles. More recently, advancedversions of HxBP probes have been devel-oped that contain an alkyne group in placeof the rhodamine reporter tag (42). Theseprobes were found to display higher-labelingefficiencies than the original HxBP-Rh probefor several MMPs and have been used to iden-tify peptide deformylase as a potential targetmediating the antibacterial effects of MP in-hibitors in Chlamydia trachomatis (93). Addi-tional classes of MP-directed ABPP probeshave been introduced with a direct peptidescaffold as a basis and used to generate active-site fingerprints for a panel of yeast MPs (91).

Photoreactive probes that target HDACs.The reversible acetylation of lysine residueson histone proteins plays a critical role intranscriptional activation and repression. Ly-sine deacetylation is catalyzed by a fam-ily of enzymes, the HDACs, which func-tion as parts of larger protein complexes.Class I and II HDACs are zinc-dependentmetallohydrolases. Inhibitors of class I/IIHDACs have been shown to induce differ-entiation in cancer cell lines and reduce tu-mor volume in vivo (94, 95). Recently, oneof these inhibitors, suberoylanilide hydrox-

NH

OOH

HN

ONH

O

NH

HO

O

O

HN

NH

O

NH

NH

HO

O

O

HN

NH

O

OH

O

N

NNNH

TAG

Marimastat GM6001

HxBP-Rh

a

b

Figure 11(a) Structures of the broad-spectrum matrix metalloproteinase inhibitorsmarimastat and GM6001 and (b) the derived activity-based proteinprofiling probe, HxBP-Rh (90).

amic acid (SAHA) (Figure 12a), was ap-proved by the FDA for the treatment of cu-taneous T cell lymphoma (95), and severalother HDAC inhibitors are in clinical de-velopment. Salisbury & Cravatt (96) soughtto develop an ABPP probe for class I/II

b SAHA-BPyne

HO

HN

O

NH

O

O

NH

O

HO

HN

O

NH

Oa SAHA

Figure 12(a) Structures of the broad-spectrum histone deacetylase (HDAC) inhibitorsuberoylanilide hydroxamic acid (SAHA) and (b) the derived activity-basedprotein profiling probe, SAHA-BPyne (96).


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

HDACs by introducing benzophenone andalkyne groups into the SAHA structure toeffect (UV irradiation-induced) covalent la-beling and (click chemistry-based) taggingand enrichment, respectively. The resultingprobe, dubbed SAHA-BPyne (Figure 12b),was found to target multiple class I and IIHDACs in proteomes. Interestingly, severalHDAC-associated proteins were also labeledby SAHA-BPyne, suggesting that these pro-teins are in close proximity to HDAC activesites, where they could presumably regulatesubstrate recognition and activity. These find-ings thus indicate that ABPP probes have thepotential to profile not only the activity stateof enzymes, but also the binding proteins thatregulate their function.

Additional Classes of Hydrolases

ABPP probes have also been developed fortwo other major classes of hydrolases—theaspartyl proteases and the proteasome. As-partyl proteases are a small clade of proteasesthat include the integral membrane proteinsβ- and γ-secretase, which regulate process-ing of the amyloid precursor protein to theAβ amyloidogenic peptides causally linked toAlzheimer’s disease. To facilitate the molec-ular characterization of γ-secretase, Li andcolleagues (97) developed biotinylated, ben-zophenone analogues of reversible inhibitorsof this protease. These probes, which con-tained an α-hydroxy-amino acid motif thatserved as a transition state analogue for as-partyl protease inhibition, were used to label,affinity purify, and identify presenilin 1 and2 as γ-secretases. Interestingly, presenilin 1and 2 are multipass integral membrane pro-teins that lack sequence homology with classi-cal aspartyl proteases, possibly explaining whymore conventional sequence-guided searchesto identify γ-secretases met with little success.

The proteasome is composed of multi-ple proteolytic subunits that utilize an N-terminal threonine nucleophile for catalysis.Nazif and Bogyo (98) have generated ABPPprobes for proteasomal subunits that con-

tain a peptide-binding group and vinyl sul-fone reactive group. By varying the pep-tide portion of the probes, the researchersgained insights into the substrate recognitionproperties of specific proteasomal subunits,culminating in the development of Z-subunit-specific inhibitors that were used to identifythis subunit as the principal trypsin-like ac-tivity of the proteasome. More recently, Ovaaand colleagues (99) have created azide ver-sions of vinyl sulfone probes and shown thatthese agents can be used to profile proteaso-mal activities in living cells (with detectionaccomplished by Staudinger ligation with aphosphine reporter tag).

Kinases and Nucleotide-Binding Proteins

The nucleotide-binding proteome is arguablythe largest and richest potential area for thedevelopment of ABPP probes. Well over halfof the human proteome binds some form ofnucleotide or nucleotide-containing cofactor,and these proteins span most cellular pro-cesses. It is not surprising that the key roleof ATP in the cell ensures that the lion’s shareof nucleotide-binding space is devoted to ATPbinding. Approximately 60 ATP-binding pro-tein families, each employing structurally dis-tinct binding motifs, have been estimated inthe proteome (100). The protein kinases con-stitute one ATP-binding family.

The human kinome has been determinedto contain at least 518 kinases clustered into17 groups and 134 families (101). Despite thesequence and structural diversity within thekinome, protein kinases all recognize a com-mon substrate, ATP, the key phosphate donorin the phosphoryl transfer to protein or pep-tide substrates. Not unexpectedly, the ATP-binding site is the region with the highestsequence homology among kinases and con-tains two consensus motifs that serve as signa-ture sequences for nearly all protein kinases.

Protein kinases, as regulators of many cel-lular pathways, have emerged as major drugtargets for oncology and other therapeutic


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

indications. Thus far, seven kinase inhibitorshave been approved as drugs. Recent esti-mates suggest that clinically relevant kinasetarget space consists of less than 40 proteins,but this is certainly destined to increase as ourbiological understanding of the kinome grows(102, 103). Understandably, the developmentof tools to interrogate the kinome has been animportant focus.

The kinases present a considerable chal-lenge to the design of ABPP probes. In thecase of the serine and cysteine hydrolasesdiscussed above, the eponymous amino acidresidues are hypernucleophilic and are cat-alytically essential to the formation of covalentenzyme-bound intermediates leading to prod-ucts. Such nucleophilic residues are ideallysuited for specific covalent modification byelectrophilic probes. Kinases catalyze phos-phate transfer from ATP to substrate by a di-rect transfer mechanism that does not involvecovalent enzyme intermediates. Thus, the ki-nase active site does not contain any unusuallynucleophilic residues. Conserved active-sitelysines do, however, present a possible targetfor ABPP probes. Sequence comparisons haveshown that virtually all protein kinases have atleast one conserved lysine residue within theiractive sites (101, 104). One lysine residue, inthe ATP-binding loop region of the kinaseprimary sequence, is conserved in all typicalprotein kinases with few exceptions. A sec-ond lysine, which resides two residues to theC terminus of the catalytic aspartic acid, isconserved in the majority of serine/threoninekinases. Cocrystal structures of protein ki-nase catalytic domains bound to ATP revealthat these lysine residues are positioned inclose proximity to the β- and γ-phosphatesof bound ATP, suggesting that the positivelycharged ε-amino groups of these lysines pro-vide electrostatic interactions with the phos-phate backbone of ATP (105). The possibil-ity that an appropriately placed electrophilecould efficiently react with the equilibratingdeprotonated lysine through elevated effec-tive concentration thus presented itself as aviable strategy.

Kinase probes based on 5′-p-fluorosulfo-nylbenzoyl adenosine. The strategic under-pinning for the design of kinase probes use-ful for proteomic analysis lies in the extensivework that has focused on the development ofnucleotide-based affinity labels. Roberta Col-man (106–108) and coworkers have been in-strumental over the past 30 years in creatinga wide array of chemical tools for the anal-ysis of nucleotide-binding proteins. The re-active groups incorporated into these probesare usually alkylating, acylating (sulfonylat-ing), or photolytic groups. The literature isextensive and has been the subject of a numberof reviews (106–108). The nucleotide-basedaffinity labels that have been reported maybe roughly divided into three categories re-flecting where the labeling group is incorpo-rated: modifications on the nucleoside base,modifications of the 5′-hydroxyl of the ribose,and other modifications of the ribose sugar.In general, incorporation of the labeling moi-ety on the 5′-hydroxyl has offered the bestpossibility of creating a reasonably close fac-simile of ATP without structural perturbationof the adenosine moiety that would affect itsbroad recognition by kinases and other ATP-binding proteins.

The prototypic nucleotide probe thathas emerged from these efforts is 5′-p-fluorosulfonylbenzoyl adenosine (5′-FSBA)(106). The fluorosulfonyl moiety of 5′-FSBA,reminiscent of the fluorophos(pho)nates dis-cussed above as serine hydrolase probes, is anelectrophile that can react with nucleophilicamino acid residues with elimination of thefluoride to form a covalent adduct and thatoccupies, in principle, the position of the γ-phosphate of ATP (Figure 13). Thus, a spe-cific reaction with the lysine amino groupsin protein kinases is possible. 5′-FSBA re-acts with a variety of nucleotide-binding pro-teins by sulfonylation of predominantly ly-sine amino groups and tyrosine hydroxyls.The reaction is not broadly based and fa-vors certain proteins, thereby lacking thegenerality of an ideal ABPP process. Sev-eral unfavorable characteristics appear to


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

N

NN

NO

NH2

OHHO

O

O

S

O

O

F

FSBA

Figure 13Structure of 5′-p-fluorosulfonylbenzoyl adenosine(FSBA), an orignal affinity label for ATP-bindingenzymes (106).

contribute to this. 5′-FSBA is uncharged,making it a marginal ATP analog. It exhibitsrelatively low affinity for nucleotide-bindingsites in proteins, related no doubt to the lackof charged phosphates. Finally, the moleculedoes not possess a straightforward site for ap-pending a reporter tag, such as biotin, to per-mit target enrichment and identification fromproteomes.

Despite these shortcomings, some recentwork has appeared on the application ofLC/MS-based methods for the developmentof 5′-FSBA as a proteomic tool for kinase pro-filing. Thus far, these studies have been lim-ited to model systems of purified kinases, suchas ALK5 and CDK2, in order to map peptidelabeling sites (109–111).

Kinase probes with ATP as a basis. A re-cent report by Patricelli and colleagues (112)on the use of nucleotide acyl phosphatesas ABPP probes for kinases and nucleotide-binding space appears to circumvent the prob-lems of 5′-FSBA and promises to providebroad proteome coverage, simplified analysis,and high sensitivity. The finding is based onthe idea that ATP itself would serve as an obvi-ous choice for a scaffold. In addition, no mod-ifications of the adenosine moiety should beused to ensure universal recognition of the nu-cleoside. Therefore, placing an electrophileat the terminal phosphate of the nucleotidewas considered. Cocrystal structures of pro-

tein kinase catalytic domains bound to ATPreveal that, in addition to the close proxim-ity of the conserved lysine residues to the β-and γ-phosphates of bound ATP, the termi-nal phosphate can accommodate an additionalsubstitutent owing to its solvent accessibility.Thus, by using ATP or ADP as a scaffold,the probe could bind to all protein kinasesin an orientation that would place a portionof the probe in the proximity of a lysineamine. The attachment of an electrophilicacyl group directly to the terminal phosphateof the nucleotide was explored as a possible so-lution. Among the number of chemically re-active phosphates, acyl phosphates have themost appropriate reactivity with amines and,at the same time, sufficient stability in an aque-ous environment. Acyl phosphates have beenknown for more than 50 years; the parent con-gener acetyl phosphate was extensively stud-ied by Di Sabato & Jencks (113, 114) and acylphosphates of amino acids and AMP are well-studied intermediates in the aminoacylationof tRNAs and other biological acyl transferreactions (115). Most notably, acetyl phos-phate is particularly reactive toward primaryamines via addition of the amine nitrogen tothe carbonyl group and facile elimination ofthe phosphate by C–O bond cleavage, result-ing in the formation of the corresponding sta-ble acetamide.

Remarkably little research has been re-ported for acyl phosphates of di- andtriphosphates given the importance of thecorresponding monophosphates. Kluger &Huang (116) synthesized acyl pyrophos-phates as mechanism-based inhibitors ofpyrophosphate-binding enzymes. They de-monstrated that isopentenoyl pyrophosphatebut not acetyl pyrophosphate efficiently in-activated farnesyl synthetase, presumably byacyl transfer to an active-site nucleophile. Theacyl phosphates of nucleotide di- and triphos-phates have likewise been virtually unexploredsince their first synthetic report in 1976 (117).Huynh-Dinh and colleagues (118–120) havepublished the only other reports of these com-pounds. They synthesized acyl phosphates


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

with a fatty acid side chain, such as a myris-toyl or cholesteryl group, in an attempt toenhance the lipophilicity and cell perme-ability of nucleotides of therapeutic interest(120–122). Although they demonstrated thatcholesteryloxycarbonyl-ATP could transport∼6% of the ATP across a model lipid bi-layer (122), the rapid aminolysis of theseacyl phosphates precluded any demonstra-tion of the phenomenon in cell culture(121).

The ABPP method of Patricelli andcolleagues (112) relies on acyl phosphate-containing nucleotides, prepared from a bi-otin derivative and ATP or ADP (Figure 14a).The acyl phosphate probes have been shownto bind selectively and react covalently at theATP-binding sites of at least 75% (∼400)of the known human protein kinases in celllysates. As predicted, the acyl phosphate re-active group on these probes showed selec-tive reactivity with one of the two conservedlysine amines in kinase active sites. The el-evated effective concentration of the active-site lysine(s) accelerated reaction with the ac-tivated acyl phosphate to form a stable amidebond with the biotin tag, releasing ATP orADP. The biotinylated kinases could then besubjected to proteolytic digestion with trypsinand purification of the probe-labeled pep-tides, accomplished with streptavidin-agarosebeads. The biotinylated peptides were thenanalyzed by LC-MS/MS to determine theidentity of the labeled protein as well asthe site of labeling (45). This platform canthus be used to profile the functional stateof many kinases in parallel directly in na-tive proteomes. Competitive ABPP studies,using broad (e.g., staurosporine) and selec-tive kinase inhibitors, can also be performedto determine inhibitor potency and selectiv-ity against native protein kinases as well asagainst hundreds of other ATPases. The abil-ity to broadly profile kinase activities in na-tive proteomes is an exciting prospect forboth target discovery and inhibitor selectivityprofiling.

a Acyl phosphate ATP probe

b Wortmannin probes

c Hypothemycin

d fmk-pa

TAG = biotin, alkyne

N

NN

N

NH2

O

OH OH

OP

O

OOH

P

O

OH

PO

O

OH

O

TAG

O

O

O

OMe

Me

O

H

OTAGO

O

O

OOH

OO

HO

OH

Enz-SH O

O

OOH

OO

HO

OH

S-Enz

N

N N O

F

NH

TAG

NH2

Figure 14Structures of different classes of kinase-directed activity-based proteinprofiling probes. (a) An acyl phosphate ATP probe (112). (b) Wortmannin-derived probes (125, 126). (c) Hypothemycin, a resorcylic acid lactonepolyketide, which reacts with protein kinases via Michael addition (128).(d ) The Rsk kinase-selective probe, fmk-pa (131, 132).

Kinase probes with covalent inhibitors asa basis. Covalent inhibitors of kinases repre-sent a complementary, targeted approach tothe more general ATP probe strategy. Wort-mannin, a steroidal metabolite of the fungusPenicillium funiculosum, is a potent inhibitorof members of the phosphoinositide 3-kinase


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

(PI 3-kinase) and the PI 3-kinase-related ki-nase (PIKK) families and has been shown toinactivate PI 3-kinase by covalent modifica-tion of a lysine in the active site (123). This ly-sine is conserved in both families. X-ray crys-tallographic studies have provided a detailedpicture of the wortmannin/PI 3-kinase in-teraction (124). On the basis of this cova-lent modification, Yee et al. (125) synthe-sized biotin- and fluorophore-based probes ofwortmannin and studied the scope of their re-action (Figure 14b). They found that wort-mannin probes inactivated PI 3-kinase andPIKK family members, consistent with thepresence of the conserved active-site lysine.The biotinylated probe permitted isolationof the sensitive kinases. They also demon-strated that, in nuclear extracts, the probe,as expected, pulled down three importantmembers of the PIKK family—ATM, ATR,and DNA-PKcs. The fluorescent probes wereuseful in quantifying activity and expressionpatterns of relevant kinases. The BODIPYderivative was also shown to be cell perme-able. Using a related fluorophore-based wort-mannin probe, Liu and colleagues (126) unex-pectedly found that polo-like kinase 1 (PLK1)is potently inhibited in a number of breast can-cer cell lines. The PLKs are conserved regula-tors of several stages of cell cycle progressionand are unrelated to the PI 3-K and PIKKfamilies. The in vitro IC50 for wortmanninof 24 nM is 1000-fold lower than scytone-min, a known PLK1 inhibitor. Subsequentwork (127) established that wortmannin in-hibits both PLK1 and PLK3 by covalent mod-ification of a conserved active-site lysine. Inhi-bition in live cells occurs at wortmannin con-centrations commonly used to inhibit the PI3-kinases. Inhibition of PLK3 by wortmanninin GM00637 cells significantly impaired theupregulation of p53 serine phosphorylation,thereby demonstrating a downstream effect.The wortmannin probe studies highlight thevalue of an ABPP probe that targets a selectnumber of enzymes from a much larger super-family. The high sensitivity and ability to re-solve proteins in complex proteomes by sim-

ple gel-based methods allow accurate quanti-tation of known protein targets, as well as theidentification of new targets.

Resorcylic acid lactones (RALs), polyke-tide natural products (Figure 14c), have re-cently shown potential as covalent modifiersof a specific fraction of the kinome. Schirmerand colleagues (128) demonstrated that RALsare susceptible to Michael addition by a cys-teine thiolate, found on some protein kinases,which results in time-dependent inactivationof the kinase that is competitive with ATP.Bioinformatics analysis revealed that 46 pro-tein kinases (∼9% of the kinome) contain aconserved cysteine necessary for this reac-tion. The RAL-sensitive kinases were well dis-tributed throughout the kinome. In an in vitropanel of 124 kinases containing 19 putativeRAL-sensitive kinases, all but one, GSK3a,showed the predicted RAL inhibition. Threekinases, cSRC, TRKA, and TRKB, which lackthe target cysteine residue, were also inhib-ited. Subsequent analysis revealed that thesethree kinases were inhibited reversibly, con-sistent with a lack of the cysteine requiredfor covalent modification. Furthermore, celllines dependent on the activation of tyro-sine kinase mitogen receptor targets of theRALs were unusually sensitive to the com-pounds and showed the predicted inhibitionof kinase phosphorylation owing to inhibi-tion of the mitogen receptors. The RALand wortmannin studies suggest that naturalproducts that irreversibly inactivate kinasesin predictable ways can be useful as ABPPprobes to interrogate well-defined fractions ofthe kinome and possibly identify unexpectedtargets.

Recent studies have reported on the de-sign of synthetic, irreversible kinase in-hibitors that selectively target subfamiliesof the kinome. Klutchko and colleagues(129) have extensively explored the structure-activity relationship (SAR) of alkynamidesof 4-anilinoquinazolines and 4-anilinopyrido-[3,4-d]pyrimidines as irreversible inhibitorsor the erbB family of tyrosine kinase re-ceptors. Previously, Fry and colleagues (130)


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

had shown that 6- or 7-acrylamido-4-anilinoquinazolines were potent, irreversibleinhibitors of epidermal growth factor recep-tor (erbB1) and erbB2 kinase activities. Theacrylamido moiety acted as a Michael acceptorto covalently react with a cysteine conservedin the ATP-binding site of both kinases. Ef-ficient covalent modification required thatthe acrylamido moiety be unsubstituted, andthere was a kinetic preference for the 6-acrylamido position. Klutchko further refinedand expanded the SAR of these kinase in-hibitors (129). It was found that analogs bear-ing a 5-dialkylamino-2-pentynamide class ofMichael acceptor had the best potency aspan-erbB inhibitors. This was likely dueto the ability of the dialkylamino substitu-tent to function as an intramolecular gen-eral acid/base to assist in the cysteine thioladdition to the alkynamide. Although theerbB inhibitor work was purely directed to-ward the design of inhibitors as therapeu-tics, the compounds may have value as war-heads for selective kinase-targeting ABPPprobes.

Expanding on the exploitation of specificcysteines in kinases for irreversible inhibi-tion, Cohen and coworkers (131) have takenan intriguing bioinformatics-based approachto identify potentially related kinase subsets.Starting with the known selectivity filter inthe ATP-binding site, the “gatekeeper,” theyused a kinome-wide sequence alignment tosearch for kinases that had the simultaneousoccurrence of a threonine as gatekeeper (topermit access to a large hydrophobic bind-ing pocket) and a solvent-exposed cysteineon a nearby conserved glycine-rich loop thatcould form a covalent bond with an inhibitor.Three kinases, RSK1, RSK2 and RSK4, mem-bers of the p90 ribosomal protein S6 ki-nase family, fulfilled the criteria. RSK3 andeight other kinases had the cysteine but hadlarger amino acids at the gatekeeper posi-tion, thereby limiting access to the bindingpocket. Guided by crystallographic informa-tion, a series of p-tolyl-pyrazolopyrimidinesthat bore a halomethyl ketone at the C-8 po-

sition were synthesized as adenine analogs.The p-tolyl moiety provided the threoninegatekeeper selectivity. The fluoromethyl ke-tone derivative and its biotinylated analog(Figure 14d ) were found to covalently mod-ify the predicted kinases by reaction with theconserved cysteine with high specificity. Ex-change of the biotin group with an alkynehas created clickable versions of these kinaseprobes (Figure 14d ) that permit profiling ofRSK activities in living cells (132). Becauseit is estimated that 20% (∼100) of the pro-tein kinases have a solvent-exposed cysteinein the ATP pocket, the prospect of identify-ing other exploitable motifs for covalent bondformation in the kinome is encouraging. Ad-ditionally, cysteines can be engineered intokinase active sites to create orthogonal kinase-ABPP probe pairs for chemical genetics stud-ies (133).

Glycosidases

Glycosidases are a large class of enzymesfound in all branches of life that play key rolesin cellular metabolism and in the regulationof carbohydrate modifications to lipids andproteins. Several experimental strategies havebeen put forth for the functional proteomicanalysis of glycosidases. Tsai and colleagues(134) synthesized candidate ABPP probes forglycosidases that utilized a latent quinonemethide electrophile to promote labeling.Although effective with purified glycosidases,this approach led to cross-reactivity withother nonglycosidase proteins in complexmixtures owing to the diffusible nature of thereleased quinone methide group. Vocadlo& Bertozzi (135) created an azide-modifiedfluorosugar as an ABPP probe for retainingβ-glycosidases (Figure 15a). Fluorosug-ars act as mechanism-based inhibitors ofretaining β-glycosidases by trapping thecovalent glycosyl-enzyme intermediate(Figure 15a). The azide group permitteddetection of probe-labeled glycosidases byStaudinger ligation with reporter-taggedphosphines. Most recently, Hekmat and


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

a

b

OO

O

OH

XF

HO

N3

O O

H

O O

OO

O

OH

FHO

N3

X = F, p-nitrophenyl

OO

FHO

NO2

OO

OHHO

NH

O

SSO

HN

NO2

TAG

Figure 15Activity-based protein profiling (ABPP) probes for β-retainingglycosidases. (a) Representative ABPP probes for exo-glycosidases (135).The active-site nucleophile of the glycosidase reacts to form a covalentprobe-enzyme adduct. (b) Structure of a cleavable probe for β-retainingglycosidases that permits identification of probe labeling sites.

colleagues generated an advanced 1-(2,4-dinitrophenyl)-2-fluorosugar probe with abiotin tag separated by a cleavable disulfidelinker (Figure 15b). This probe allowedthe authors to profile both retaining β-glycosidases and their specific sites oflabeling, resulting in the discovery of a newβ-1,4-glycanase from the soil bacteriumCellulomonas fimi.

Cytochrome P450s

The cytochrome P450 (P450) monooxy-genase enzyme family metabolizes a largenumber of endogenous signaling molecules,xenobiotics, and drugs. The human genomeencodes 57 distinct P450 enzymes (136),many of which remain uncharacterized. Theactivity of P450 enzymes is regulated bymultiple factors in vivo: membrane com-

position and localization; protein-bindingpartners, e.g., Dap1/PGRMC1 (137); post-translational modification (138); endogenousconcentrations of cofactors (e.g., NAPDH);and regulatory enzymes (e.g., NADPH-P450reductases). Wright & Cravatt (139) have de-veloped ABPP probes for P450 enzymes onthe basis of mechanism-based inhibitors ofthe aryl acetylene class. The inactivation ofP450 enzymes by aryl acetylenes, such as 2-ethynylnaphthalene (2EN) (Figure 16a), hasbeen well studied (140). The hydrophobicnaphthalene group directs inhibitor bindingtoward P450 active sites, wherein enzyme-catalyzed oxidation of the conjugated 2-acetylene generates a highly reactive ketene(Figure 16b). The ketene intermediate canthen acylate nucleophilic residues within theP450 active site. Modification of 2EN withan alkyl (unconjugated) acetylene group con-verted this inhibitor into an ABPP probe(2EN-ABP) (Figure 16a) capable of labelingmultiple P450 enzymes in vitro and in vivo.Labeling of P450 enzymes was NADPH de-pendent and blocked by P450 inhibitors. No-tably, the 2EN-based probe was used to moni-tor induction and inhibition of P450 enzymesin living mice, thus providing a potentiallypowerful assay to characterize P450-drug in-teractions in vivo.

Phosphatases

Protein phosphorylation is a reversible post-translational modification. Two major classesof phosphatases, the serine/threonine andtyrosine phosphatases, are responsible forremoving phosphate groups from proteins.Serine/threonine phosphatases are metal-dependent enzymes that function as partof multisubunit complexes. Several naturalproducts have been identified that potentlyinhibit serine/threonine phosphatases, in-cluding the cyanobacterial hepatotoxin mi-crocystin, which covalently modifies a non-catalytic cysteine residue in the activesites of these enzymes (141). Shreder and


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

NH

O

2EN-ABP

P450

O

CC

O

P450oxidation

Ketene

P450nucleophile

2EN

a

b

Figure 16Activity-based protein profiling probes for cytochrome P450 enzymes. (a) Structures of the broad-spectrum cytochrome P450 (P450) inhibitor 2-ethynylnaphthalene (2EN) (140) and the derived probe,2EN-ABP (139). (b) Mechanism of inactivation of P450s by aryl acetylyne inhibitors.

colleagues (142) synthesized a rhodamine-tagged derivative of microcystin and used thisagent to profile serine/threonine phosphataseactivities in proteomes.

Tyrosine phosphatases utilize a cysteinenucleophile for catalysis, which initially mightsuggest that these enzymes would representfacile targets for ABPP. However, the cys-teine nucleophile of tyrosine phosphatases ap-pears to be less reactive toward electrophilesthan the nucleophile of cysteine proteases.In this regard, mechanistic analysis of ty-rosine phosphatases indicates that these en-zymes operate by a predominantly dissociativemechanism whereby proton transfer to theleaving group oxygen precedes nucleophilicattack by the catalytic cysteine (143). Thesefactors have complicated the developmentof ABPP probes for tyrosine phosphatases.Nonetheless, some progress has been made.Zhang and colleagues (144) have gener-ated reporter-tagged bromobenzylphospho-nate probes that label several purified tyrosinephosphatases in an active-site-directed man-ner and that are currently under examinationfor their ability to target these enzymes inproteomes.

Nondirected Probes that TargetMultiple Enzyme Classes

Each of the aforementioned enzyme classeswas targeted by ABPP probes that incorpo-rated known binding groups and/or affin-ity labels. These chemotypes were thus usedto direct probe labeling to specific enzymeclasses. Ongoing directed ABPP efforts con-tinue to provide probes for new enzymeclasses, including, for example, arginine deim-inases (145) and methione aminopeptidases(146). However, many enzymes lack cog-nate inhibitors or affinity labels; to extendABPP to these proteins, a nondirected strat-egy for probe discovery has been intro-duced (147, 148). Nondirected ABPP involvesthe synthesis and proteomic screening of li-braries of probes that bear a common re-active group and a variable binding group.Specific protein targets are identified in pro-teomes on the basis of activity-dependent(e.g., heat-sensitive) labeling and/or selectivereactivity with a subset of the probe library.Several probe libraries have been synthe-sized, including those that bear sulfonate ester(147, 148) (Figure 17a), α-chloroacetamide


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

SO

O

OR

HN

O

TAG

NH

X

OHN

O

NH

O

HNTAG

HN

O

Cl

R2

NH

X

OHN

O

NH

OR1

HNTAG

HN

O

Cl

a

b

c

R = binding group

X = H or biotinR1, R2 = binding groups

OO

O

O

HN

O

R

R = binding groups

Figure 17Representativenondirectedlibraries of ABPPprobes. (a) Alibrary of sulfonateester probes (147,148). (b) A libraryof α-chloro-acetamide probes,where R1 and R2

correspond tonatural andunnatural aminoacid side chains(149). (c) A libraryof spiroepoxideprobes made inclickable format forcell-basedscreening (150).

(149) (Figure 17b), and spiroepoxide (150)(Figure 17c) reactive groups. Enzyme tar-gets of these probes include aldehyde/alcoholdehydrogenases (36, 44, 47, 147, 148),enoyl-CoA hydratases (148), glutathione S-transferases (148, 149), nitrilases (149), and

transglutaminases (151). A clickable spiro-epoxide library was recently screened for an-tiproliferative effects in breast cancer cells,resulting in the identification of a bioactiveprobe that targeted the glycolytic enzymephosphoglycerate mutase (150).

SUMMARY POINTSThe field of ABPP has quickly reached an inflection point toward making importantcontributions to the overarching goals of systems biology. The ability of ABPP to in-tegrate the classical disciplines of organic synthesis and mechanistic enzymology withcontemporary analytical methods sets this technology apart as a highly interdisciplinaryendeavor. Indeed, by interrogating fractions of the proteome on the basis of shared func-tional properties, rather than mere abundance, ABPP analyzes portions of biomolecularspace that are inaccessible to other large-scale profiling methods.

In conjunction with novel cell and animal models of disease, ABPP has alreadyestablished a track record of identifying enzyme activities associated with a range of dis-eases, including cancer (60, 61), malaria (88), and metabolic disorders (149). From a phar-maceutical perspective, ABPP is also contributing to the design and evaluation of selective


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

inhibitors for disease-linked enzymes, including enzymes of uncharacterized function(68, 152). This duality of ABPP as a discovery tool to identify enzymes associated withspecific physiological and/or pathological processes and as an evaluative tool for thedesign of selective inhibitors, and potentially drugs, that target these enzymes is thehallmark of a robust technology that can serve as both a hypothesis generator and ahypothesis tester.

FUTURE ISSUESA number of challenges lie ahead. Most obvious from the perspective of this review isthe expansion of the proteome coverage of ABPP. Thus far, the field has taken a narrow,albeit pragmatic, focus on protein activity as specified by enzymatic catalysis. Giventhe natural affinity of ABPP to enzyme and medicinal chemistry as a source of small-molecule probes, this is understandable, and clearly, there is much more to be donehere. However, enzymes need not be the only component of the proteome addressed byABPP. Future directions may include the development of probes to interrogate receptors(e.g., G protein–coupled receptors), ion channels, and structural proteins. In these cases,chemical and biological tools to interrogate specific proteins do exist; conversion of theseagents into ABPP probes to capture specific fractions of proteomic space remains to beimplemented. This orthogonal expansion beyond enzymes, in conjunction with advancesin the sensitivity, resolution, and throughput of analytical platforms, would completeand extend the transition from enzyme chemistry to a new, multifaceted proteomicchemistry.

We view as an ultimate goal the integration of ABPP with other large-scale profil-ing technologies and targeted experimental approaches to expand our understanding ofthe biochemical mechanisms of health and disease. For instance, ABPP is already be-ing united with complementary systems biology methods, such as metabolomics (68), aportion of biomolecular space that constitutes the major biochemical output of enzymeactivity in vivo. By perturbing enzyme activity in living systems and then profiling themetabolic consequences, researchers may succeed in integrating both known and unchar-acterized enzymes into the higher-order biochemical networks of cells and tissues. Inthis manner, a new experimental platform could be established for rapidly moving fromthe discovery of enzyme activities associated with biological processes to elucidation ofthe mechanistic basis and functional significance of these relationships. This network-ing of systems biology platforms united and enriched by an ever-expanding repertoireof proteomic chemistry would lead a higher-order understanding of the biology of lifeand the pathology of disease. The next generation of novel therapeutics should certainlybenefit from such advances.

DISCLOSURE STATEMENT

Dr. Kozarich is Chairman and President of Activx Biosciences, a company that specializesin the use of activity-based proteomics technologies for drug discovery, and Dr. Cravatt is aconsultant for Activx Biosciences.


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

ACKNOWLEDGMENTS

We gratefully acknowledge the support of the National Institutes of Health (CA087660,CA118696, DA017259), the Helen L. Dorris Child and Adolescent Neuropsychiatric DisorderInstitute, the California Breast Cancer Research Program (A.T.W.), and the Skaggs Institutefor Chemical Biology (A.T.W. and B.F.C.).

LITERATURE CITED

1. Hanahan D, Weinberg RA. 2000. Cell 100:57–702. Vogelstein B, Kinzler KW. 2004. Nat. Med. 10:789–993. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, et al. 2003. Cancer Cell 3:537–494. Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, et al. 2005. Nature 436:518–245. Daily J, Le Roch K, Sarr O, Fang X, Zhou Y, et al. 2004. Malar. J. 3:30–376. Young JA, Fivelman QL, Blair PL, de la Vega P, Le Roch KG, et al. 2005. Mol. Biochem.

Parasitol. 143:67–797. Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, et al. 2003. Nat. Genet.

34:267–738. Kobe B, Kemp BE. 1999. Nature 402:373–769. Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold R. 1999. Nat. Biotechnol.

17:994–9910. Washburn MP, Wolters D, Yates JR 3rd. 2001. Nat. Biotechnol. 19:242–4711. Ito T, Ota K, Kubota H, Yamaguchi Y, Chiba T, et al. 2002. Mol. Cell. Proteomics 1:561–6612. MacBeath G. 2002. Nat. Genet. 32(Suppl.):526–3213. Khan AR, James MN. 1998. Protein Sci. 7:815–3614. Means GE, Feeney RE. 1990. Bioconjug. Chem. 1:2–1215. Balls AK, Jansen EF. 1952. Adv. Enzymol. Relat. Subj. Biochem. 13:321–4316. Balls AK, Jansen EF. 1952. Annu. Rev. Biochem. 21:1–2817. Jansen EF, Curl AL, Balls AK. 1951. J. Biol. Chem. 190:557–6118. Jansen EF, Jang R, Balls AK. 1952. J. Biol. Chem. 196:247–5319. Jansen EF, Balls AK. 1952. J. Biol. Chem. 194:721–2720. Powers JC, Asgian JL, Ekici OD, James KE. 2002. Chem. Rev. 102:4639–75021. Kresge N, Simoni RD, Hill RL. 2007. J. Biol. Chem. 282:e2522. Tipper DJ, Strominger JL. 1965. Proc. Natl. Acad. Sci. USA 54:1133–4123. Blumberg PM, Strominger JL. 1972. J. Biol. Chem. 247:8107–1324. Suginaka H, Blumberg PM, Strominger JL. 1972. J. Biol. Chem. 247:5279–8825. Blumberg PM, Strominger JL. 1974. Methods Enzymol. 34:401–526. Kozarich JW, Strominger JL. 1978. J. Biol. Chem. 253:1272–7827. Iwaya M, Goldman R, Tipper DJ, Feingold B, Strominger JL. 1978. J. Bacteriol. 136:1143–

5828. Adam GC, Sorensen EJ, Cravatt BF. 2002. Mol. Cell. Proteomics 1:781–9029. Kozarich JW. 2003. Curr. Opin. Chem. Biol. 7:78–8330. Speers AE, Cravatt BF. 2004. ChemBioChem 5:41–4731. Jessani N, Cravatt BF. 2004. Curr. Opin. Chem. Biol. 8:54–5932. Berger AB, Vitorino PM, Bogyo M. 2004. Am. J. Pharmacogenomics 4:371–8133. Evans MJ, Cravatt BF 2006. Chem. Rev. 106:3279–30134. Kolb HC, Sharpless KB. 2003. Drug Discov. Today 8:1128–3735. Speers AE, Adam GC, Cravatt BF. 2003. J. Am. Chem. Soc. 125:4686–8736. Speers AE, Cravatt BF. 2004. Chem. Biol. 11:535–46


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

37. Liu Y, Patricelli MP, Cravatt BF. 1999. Proc. Natl. Acad. Sci. USA 96:14694–9938. Kidd D, Liu Y, Cravatt BF. 2001. Biochemistry 40:4005–1539. Patricelli MP, Giang DK, Stamp LM, Burbaum JJ. 2001. Proteomics 1:1067–7140. Greenbaum D, Baruch A, Hayrapetian L, Darula Z, Burlingame A, et al. 2002. Mol. Cell.

Proteomics 1:60–6841. Jessani N, Niessen S, Wei BQ, Nicolau M, Humphrey M, et al. 2005. Nat. Methods 2:691–

9742. Sieber SA, Niessen S, Hoover HS, Cravatt BF. 2006. Nat. Chem. Biol. 2:274–8143. Alexander JP, Cravatt BF. 2006. J. Am. Chem. Soc. 128:9699–70444. Adam GC, Burbaum JJ, Kozarich JW, Patricelli MP, Cravatt BF. 2004. J. Am. Chem. Soc.

126:1363–6845. Okerberg ES, Wu J, Zhang B, Samii B, Blackford K, et al. 2005. Proc. Natl. Acad. Sci. USA

102:4996–500146. Hekmat O, Kim YW, Williams SJ, He S, Withers SG. 2005. J. Biol. Chem. 280:35126–3547. Speers AE, Cravatt BF. 2005. J. Am. Chem. Soc. 127:10018–1948. Verhelst SH, Fonovic M, Bogyo M. 2007. Angew. Chem. Int. Ed. Engl. 46:1284–8649. Everley PA, Gartner CA, Haas W, Saghatelian A, Elias JE, et al. 2007. Mol. Cell. Proteomics

6:1771–7750. Sieber SA, Mondala TS, Head SR, Cravatt BF. 2004. J. Am. Chem. Soc. 126:15640–4151. Winssinger N, Ficarro S, Schultz PG, Harris JL. 2002. Proc. Natl. Acad. Sci. USA 99:11139–

4452. Harris J, Mason DE, Li J, Burdick KW, Backes BJ, et al. 2004. Chem. Biol. 11:1361–7253. Greenbaum DC, Arnold WD, Lu F, Hayrapetian L, Baruch A, et al. 2002. Chem. Biol.

9:1085–9454. Leung D, Hardouin C, Boger DL, Cravatt BF. 2003. Nat. Biotechnol. 21:687–9155. Joyce JA, Baruch A, Chehade K, Meyer-Morse N, Giraudo E, et al. 2004. Cancer Cell

5:443–5356. Alexander JP, Cravatt BF. 2005. Chem. Biol. 12:1179–8757. Borodovsky A, Ovaa H, Kolli N, Gan-Erdene T, Wilkinson KD, et al. 2002. Chem. Biol.

9:1149–5958. Jessani N, Young JA, Diaz SL, Patricelli MP, Varki A, Cravatt BF. 2005. Angew. Chem.

Int. Ed. Engl. 44:2400–359. Barglow KT, Cravatt BF. 2006. Angew. Chem. Int. Ed. Engl. 45:7408–1160. Jessani N, Liu Y, Humphrey M, Cravatt BF. 2002. Proc. Natl. Acad. Sci. USA 99:10335–4061. Jessani N, Humphrey M, McDonald WH, Niessen S, Masuda K, et al. 2004. Proc. Natl.

Acad. Sci. USA 101:13756–6162. Jaeken J, Martens K, Francois I, Eyskens F, Lecointre C, et al. 2006. Am. J. Hum. Genet.

78:38–5163. Pan Z, Jeffery DA, Chehade K, Beltman J, Clark JM, et al. 2006. Bioorg. Med. Chem. Lett.

16:2882–8564. Mahrus S, Craik CS. 2005. Chem. Biol. 12:567–7765. Madsen MA, Deryugina EI, Niessen S, Cravatt BF, Quigley JP. 2006. J. Biol. Chem.

281:15997–600566. Milner J, Kevorkian L, Young D, Jones D, Wait R, et al. 2006. Arthritis Res. Ther. 8:R2367. Nomura DK, Leung D, Chiang KP, Quistad GB, Cravatt BF, Casida JE. 2005. Proc. Natl.

Acad. Sci. USA 102:6195–20068. Chiang KP, Niessen S, Saghatelian A, Cravatt BF. 2006. Chem. Biol. 13:1041–5069. Saghatelian A, Trauger SA, Want EJ, Hawkins EG, Siuzdak G, Cravatt BF. 2004. Bio-

chemistry 43:14332–39


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

70. Lu J, Xiao Y-j, Baudhuin LM, Hong G, Xu Y. 2002. J. Lipid Res. 43:463–7671. McKinney MK, Cravatt BF. 2005. Annu. Rev. Biochem. 74:411–3272. Cravatt BF, Demarest K, Patricelli MP, Bracey MH, Giang DK, et al. 2001. Proc. Natl.

Acad. Sci. USA 98:9371–7673. Kathuria S, Gaetani S, Fegley D, Valino F, Duranti A, et al. 2003. Nat. Med. 9:76–8174. Lichtman AH, Leung D, Shelton CC, Saghatelian A, Hardouin C, et al. 2004. J. Pharmacol.

Exp. Ther. 311:441–4875. Ahn K, Johnson DS, Fitzgerald LR, Liimatta M, Arendse A, et al. 2007. Biochemistry

46:13019–3076. Nomanbhoy T, Rosenblum J, Aban A, Burbaum JJ. 2003. Assay Drug Dev. Technol. 1:137–

4677. Shreder KR, Wong MS, Corral S, Yu Z, Winn DT, et al. 2005. Bioorg. Med. Chem. Lett.

15:4256–6078. Fonovic M, Bogyo M. 2007. Curr. Pharm. Design 13:253–6179. Barrett AJ, Kembhavi AA, Brown MA, Kirschke H, Knight CG, et al. 1982. Biochem. J.

201:189–9880. Greenbaum D, Medzihradszky KF, Burlingame A, Bogyo M. 2000. Chem. Biol. 7:569–8181. Kato D, Boatright KM, Berger AB, Nazif T, Blum G, et al. 2005. Nat. Chem. Biol. 1:33–3882. Thornberry NA, Bull HG, Calaycay JR, Chapman KT, Howard AD, et al. 1992. Nature

356:768–7483. Thornberry NA, Peterson EP, Zhao JJ, Howard AD, Griffin PR, Chapman KT. 1994.

Biochemistry 33:3934–4084. Faleiro L, Kobayashi R, Fearnhead H, Lazebnik Y. 1997. EMBO J. 16:2271–8185. Berger AB, Witte MD, Denault JB, Sadaghiani AM, Sexton KM, et al. 2006. Mol. Cell

23:509–2186. Hemelaar J, Galardy PJ, Borodovsky A, Kessler BM, Ploegh HL, Ovaa H. 2004. J. Proteome

Res. 3:268–7687. Kattenhorn LM, Korbel GA, Kessler BM, Spooner E, Ploegh HL. 2005. Mol. Cell 19:547–

5788. Greenbaum DC, Baruch A, Grainger M, Bozdech Z, Medzihradszky KF, et al. 2002.

Science 298:2002–689. Sexton KB, Witte MD, Blum G, Bogyo M. 2007. Bioorg. Med. Chem. Lett. 17:649–5390. Saghatelian A, Jessani N, Joseph A, Humphrey M, Cravatt BF. 2004. Proc. Natl. Acad. Sci.

USA 101:10000–591. Chan EW, Chattopadhaya S, Panicker RC, Huang X, Yao SQ. 2004. J. Am. Chem. Soc.

126:14435–4692. Whittaker M, Floyd CD, Brown P, Gearing AJH. 1999. Chem. Rev. 99:2735–7693. Balakrishnan A, Patel B, Sieber SA, Chen D, Pachikara N, et al. 2006. J. Biol. Chem.

281:16691–9994. Bolden JE, Peart MJ, Johnstone RW. 2006. Nat. Rev. Drug Discov. 5:769–8495. Marks PA, Breslow R. 2007. Nat. Biotechnol. 25:84–9096. Salisbury CM, Cravatt BF. 2007. Proc. Natl. Acad. Sci. USA 104:1171–7697. Li YM, Xu M, Lai MT, Huang Q, Castro JL, et al. 2000. Nature 405:689–9498. Nazif T, Bogyo M. 2001. Proc. Natl. Acad. Sci. USA 98:2967–7299. Ovaa H, Van Swieten PF, Kessler BM, Leeuwenburgh MA, Fiebiger E, et al. 2003. Angew.

Chem. Int. Ed. Engl. 42:3626–29100. Kuttner YY, Sobolev V, Raskind A, Edelman M. 2003. Proteins 52:400–11101. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. 2002. Science 298:1912–34


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

102. Fedorov O, Sundstrom M, Marsden B, Knapp S. 2007. Drug Discov. Today 12:365–72103. Vieth M, Sutherland JJ, Robertson DH, Campbell RM. 2005. Drug Discov. Today 10:839–

46104. Hanks SK, Hunter T. 1995. FASEB J. 9:576–96105. Zheng J, Knighton DR, ten Eyck LF, Karlsson R, Xuong N, et al. 1993. Biochemistry

32:2154–61106. Colman RF. 1983. Annu. Rev. Biochem. 52:67–91107. Colman RF. 1990. In The Enzymes, ed. DS Sigman, PD Boyer, pp. 283–321. New York:

Academic108. Colman RF. 2006. Lett. Drug Des. Discov. 3:462–80109. Khandekar SS, Feng B, Yi T, Chen S, Laping N, Bramson N. 2005. J. Biomol. Screen.

10:447–55110. Ratcliffe SJ, Yi T, Khandekar SS. 2007. J. Biomol. Screen. 12:126–32111. Renzone G, Salzano AM, Arena S, D’Ambrosio C, Scaloni A. 2006. J. Proteome Res.

5:2019–24112. Patricelli MP, Szardenings AK, Liyanage M, Nomanbhoy TK, Wu M, et al. 2007. Bio-

chemistry 46:350–58113. Di Sabato G, Jencks WP. 1961. J. Am. Chem. Soc. 83:4393–400114. Di Sabato G, Jencks WP. 1961. J. Am. Chem. Soc. 83:4400–5115. Demoss JA, Genuth SM, Novelli GD. 1956. Proc. Natl. Acad. Sci. USA 42:325–32116. Kluger R, Huang Z. 1991. J. Am. Chem. Soc. 113:5124–25117. Shumyantzeva VV, Sokolova NI, Shabarova ZA. 1976. Nucleic Acids Res. 3:903–16118. Bonnaffe D, Dupraz B, Ughetto-Monfrin J, Namane A, Huynh Dinh T. 1995. Nucleosides

Nucleotides 14:783–87119. Bonnaffe D, Dupraz B, Ughetto-Monfrin J, Namane A, Huynh Dinh T. 1995. Tetrahedron

Lett. 36:531–34120. Kreimeyer A, Andre F, Bluzat A, Gouyette C, Huynh Dinh T. 1999. Nucleosides Nucleotides

18:995–99121. Bonnaffe D, Dupraz B, Ughetto-Monfrin J, Namane A, Henin Y, Huynh Dinh T. 1996.

J. Org. Chem. 61:895–902122. Kreimeyer A, Andre F, Gouyette C, Huynh-Dinh T. 1998. Angew. Chem. Int. Ed. Engl.

37:2853–55123. Wymann MP, Bulgarelli-Leva G, Zvelebil MJ, Pirola L, Vanhaesebroeck B, et al. 1996.

Mol. Cell. Biol. 16:1722–33124. Walker EH, Pacold ME, Perisic O, Stephens L, Hawkins PT, et al. 2000. Mol. Cell 6:909–

19125. Yee MC, Fas SC, Stohlmeyer MM, Wandless TJ, Cimprich KA. 2005. J. Biol. Chem.

280:29053–59126. Liu Y, Shreder KR, Gai W, Corral S, Ferris DK, Rosenblum JS. 2005. Chem. Biol. 12:99–

107127. Liu Y, Jiang N, Wu J, Dai W, Rosenblum JS. 2007. J. Biol. Chem. 282:2505–11128. Schirmer A, Kennedy J, Murli S, Reid R, Santi DV. 2006. Proc. Natl. Acad. Sci. USA

103:4234–39129. Klutchko SR, Zhou H, Winters RT, Tran TP, Bridges AJ, et al. 2006. J. Med. Chem.

49:1475–85130. Fry DW, Bridges AJ, Denny WA, Doherty A, Greis KD, et al. 1998. Proc. Natl. Acad. Sci.

USA 95:12022–27131. Cohen MS, Zhang C, Shokat KM, Taunton J. 2005. Science 308:1318–21


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

ANRV345-BI77-17 ARI 28 April 2008 15:9

132. Cohen MS, Hadjivassiliou H, Taunton J. 2007. Nat. Chem. Biol. 3:156–60133. Blair JA, Rauh D, Kung C, Yun C-H, Fan Q-W, et al. 2007. Nat. Chem. Biol. 3:229–38134. Tsai CS, Li YK, Lo LC. 2002. Org. Lett. 4:3607–10135. Vocadlo DJ, Bertozzi CR. 2004. Angew. Chem. Int. Ed. Engl. 43:5338–42136. Guengerich FP, Wu ZL, Bartleson CJ. 2005. Biochem. Biophys. Res. Commun. 338:465–69137. Hughes AL, Powell DW, Bard M, Eckstein J, Barbuch R, et al. 2007. Cell Metab. 5:143–49138. Aguiar M, Masse R, Gibbs BF. 2005. Drug Metab. Rev. 37:379–404139. Wright AT, Cravatt BF. 2007. Chem. Biol. 14:1043–51140. Kent UM, Juschyshyn MI, Hollenberg PF. 2001. Curr. Drug Metab. 2:215–43141. Goldberg J, Huang HB, Kwon YG, Greengard P, Nairn AC, Kuriyan J. 1995. Nature

376:745–53142. Shreder KR, Liu Y, Nomanhboy T, Fuller SR, Wong MS, et al. 2004. Bioconjug. Chem.

15:790–98143. Asthagiri D, Dillet V, Liu T, Noodleman L, Van Etten RL, Bashford D. 2002. J. Am.

Chem. Soc. 124:10225–35144. Kumar S, Zhou B, Liang F, Wang WQ, Huang Z, Zhang ZY. 2004. Proc. Natl. Acad. Sci.

USA 101:7943–48145. Luo Y, Knuckley B, Bhatia M, Pellechia PJ, Thompson PR. 2006. J. Am. Chem. Soc.

128:14468–69146. Qiu WW, Xu J, Li JY, Li J, Nan FJ. 2007. ChemBioChem 8:1351–58147. Adam GC, Cravatt BF, Sorensen EJ. 2001. Chem. Biol. 8:81–95148. Adam GC, Sorensen EJ, Cravatt BF. 2002. Nat. Biotechnol. 20:805–9149. Barglow KT, Cravatt BF. 2004. Chem. Biol. 11:1523–31150. Evans MJ, Saghatelian A, Sorensen EJ, Cravatt BF. 2005. Nat. Biotechnol. 23:1303–7151. Adam GC, Sorensen EJ, Cravatt BF. 2002. Mol. Cell. Proteomics 1:828–35152. Li W, Blankman JL, Cravatt BF. 2007. J. Am. Chem. Soc. 129:9594–95


Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Annual Review ofBiochemistry

Volume 77, 2008Contents

Prefatory Chapters

Discovery of G Protein SignalingZvi Selinger � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1

Moments of DiscoveryPaul Berg � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 14

Single-Molecule Theme

In singulo Biochemistry: When Less Is MoreCarlos Bustamante � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 45

Advances in Single-Molecule Fluorescence Methodsfor Molecular BiologyChirlmin Joo, Hamza Balci, Yuji Ishitsuka, Chittanon Buranachai,and Taekjip Ha � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 51

How RNA Unfolds and RefoldsPan T.X. Li, Jeffrey Vieregg, and Ignacio Tinoco, Jr. � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 77

Single-Molecule Studies of Protein FoldingAlessandro Borgia, Philip M. Williams, and Jane Clarke � � � � � � � � � � � � � � � � � � � � � � � � � � � � �101

Structure and Mechanics of Membrane ProteinsAndreas Engel and Hermann E. Gaub � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �127

Single-Molecule Studies of RNA Polymerase: Motoring AlongKristina M. Herbert, William J. Greenleaf, and Steven M. Block � � � � � � � � � � � � � � � � � � � �149

Translation at the Single-Molecule LevelR. Andrew Marshall, Colin Echeverría Aitken, Magdalena Dorywalska,and Joseph D. Puglisi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �177

Recent Advances in Optical TweezersJeffrey R. Moffitt, Yann R. Chemla, Steven B. Smith, and Carlos Bustamante � � � � � �205

Recent Advances in Biochemistry

Mechanism of Eukaryotic Homologous RecombinationJoseph San Filippo, Patrick Sung, and Hannah Klein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �229

v

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Structural and Functional Relationships of the XPF/MUS81Family of ProteinsAlberto Ciccia, Neil McDonald, and Stephen C. West � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �259

Fat and Beyond: The Diverse Biology of PPARγ

Peter Tontonoz and Bruce M. Spiegelman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �289

Eukaryotic DNA Ligases: Structural and Functional InsightsTom Ellenberger and Alan E. Tomkinson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �313

Structure and Energetics of the Hydrogen-Bonded Backbonein Protein FoldingD. Wayne Bolen and George D. Rose � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �339

Macromolecular Modeling with RosettaRhiju Das and David Baker � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �363

Activity-Based Protein Profiling: From Enzyme Chemistryto Proteomic ChemistryBenjamin F. Cravatt, Aaron T. Wright, and John W. Kozarich � � � � � � � � � � � � � � � � � � � � � �383

Analyzing Protein Interaction Networks Using Structural InformationChristina Kiel, Pedro Beltrao, and Luis Serrano � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �415

Integrating Diverse Data for Structure Determinationof Macromolecular AssembliesFrank Alber, Friedrich Förster, Dmitry Korkin, Maya Topf, and Andrej Sali � � � � � � � �443

From the Determination of Complex Reaction Mechanismsto Systems BiologyJohn Ross � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �479

Biochemistry and Physiology of Mammalian SecretedPhospholipases A2

Gerard Lambeau and Michael H. Gelb � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �495

Glycosyltransferases: Structures, Functions, and MechanismsL.L. Lairson, B. Henrissat, G.J. Davies, and S.G. Withers � � � � � � � � � � � � � � � � � � � � � � � � � � �521

Structural Biology of the Tumor Suppressor p53Andreas C. Joerger and Alan R. Fersht � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �557

Toward a Biomechanical Understanding of Whole Bacterial CellsDylan M. Morris and Grant J. Jensen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �583

How Does Synaptotagmin Trigger Neurotransmitter Release?Edwin R. Chapman � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �615

Protein Translocation Across the Bacterial Cytoplasmic MembraneArnold J.M. Driessen and Nico Nouwen � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �643

vi Contents

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

AR345-FM ARI 7 May 2008 14:43

Maturation of Iron-Sulfur Proteins in Eukaryotes: Mechanisms,Connected Processes, and DiseasesRoland Lill and Ulrich Mühlenhoff � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �669

CFTR Function and Prospects for TherapyJohn R. Riordan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �701

Aging and Survival: The Genetics of Life Span Extensionby Dietary RestrictionWilliam Mair and Andrew Dillin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �727

Cellular Defenses against Superoxide and Hydrogen PeroxideJames A. Imlay � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �755

Toward a Control Theory Analysis of AgingMichael P. Murphy and Linda Partridge � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �777

Indexes

Cumulative Index of Contributing Authors, Volumes 73–77 � � � � � � � � � � � � � � � � � � � � � � � �799

Cumulative Index of Chapter Titles, Volumes 73–77 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �803

Errata

An online log of corrections to Annual Review of Biochemistry articles may be foundat http://biochem.annualreviews.org/errata.shtml

Contents vii

Ann

u. R

ev. B

ioch

em. 2

008.

77:3

83-4

14. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by S

crip

ps R

esea

rch

Inst

itute

- K

RE

SGE

LIB

RA

RY

on

10/0

2/08

. For

per

sona

l use

onl

y.

activity-based protein profiling: from enzyme chemistry to ... · posed that the β-lactam nucleus...

Documents