Hermanson G. Bioconjugate Techniques, Second Edition

Подождите немного. Документ загружается.

660 16. Mass Tags and Isotope Tags

analysis to multiple samples, because of the shear number of peaks that result in the mass spec

separation.

A new type of mass tag extends the beneﬁ ts of stable isotopic labeling of peptides for mass

spec to the analysis of multiple samples and multiple proteins separated simultaneously within

the same mass spec run. An isobaric tag consists of a reactive group for coupling to peptides fol-

lowed by an isotopically labeled mass normalization group, a cleavable linker, and another iso-

topically labeled group, called a mass reporter ( Figure 16.11 ). For every isotopic substitution in

the mass reporter region, the mass normalization group has an inverse mass substitution. Using

this balancing process, the total molecular mass of the entire tag always stays the same–thus the

name “isobaric tag ”. A series of isobaric labels can be created by careful selection of the isotopic

substitutions in the reporter group, which are exactly balanced in the normalization group.

If there are enough potential isotopic substitution sites in the isobaric tag design, a set of

4–10 tags can be created in which each has a different reporter mass, but all of them have the

same total molecular weight. Proteome Sciences, Applied Biosystems, and PerkinElmer/Agilix

each have commercialized isobaric tag sets based on the reporter group/normalization group

blueprint for multiplexed sample analysis. Figure 16.12 shows examples of isobaric tag design,

which were obtained from company advertising, scientiﬁ c publications, or the associated pat-

ents. Most of the tags also contain a tertiary amine group or a guanidino group to aid in the

ionization of the labeled peptide and provide better mass spec signals.

In use, a protein sample ﬁ rst is proteolytically digested and labeled with an individual iso-

baric tag. The most common isobaric tag design contains an amine-reactive NHS ester group

to label each peptide at its N-terminus. Therefore, within a given sample all of the different

peptides after labeling will have a unique isobaric tag modifying them with a characteristic

reporter group mass. Unlike the use of an ICAT tag that targets only cysteine-containing pep-

tides, modiﬁ cation of the N-terminus of every peptide assures 100 percent coverage of the pro-

teome. If multiple samples are being analyzed, then each peptide sample is separately labeled

Figure 16.9 The O-ECAT reagent structure contains a DOTA chelating group and a terminal aminoxy group

for coupling to aldehyde and ketone sites of oxidation within biological molecules.

3. Isobaric Tags 661

Figure 16.10 The O-ECAT mass tag can covalently link to any oxidized proteins containing aldehydes, forming

an oxime bond.

with a different isobaric tag in the series. After labeling, all the samples then are combined and

separated chromatographically to reduce the total sample complexity.

Since each isobaric tag in a set is structurally identical except for its isotopic substitution

pattern, all of them perform identically with regard to peptide modiﬁ cation and chromato-

graphic separation. Multiple peptide samples then can be labeled separately with different tags

in a series, the samples combined, subjected to fractionation typically done by multidimen-

sional protein identiﬁ cation technology (MudPIT; Liu et al., 2004), and injected into a mass

spectrometer for analysis. As each peak comes off the MudPIT separation system and goes

into the mass spec, it contains the same sequences of labeled peptides from each sample that

happen to elute under the instant conditions of salt strength and solvent addition being used at

that moment. The only difference in the peptide mixture within a given chromatographic peak,

662 16. Mass Tags and Isotope Tags

then, is the type of isobaric label attached to them, which is indicative of the sample it came

from. Thus, the peptides representing a particular sample will all be labeled with a tag having a

characteristic reporter group mass. Another set of peptides with the same amino acid sequence

but coming from another sample will be labeled with a different isobaric tag having another

reporter group mass signature.

It is only upon tandem mass spec analysis that the isobaric labels can be distinguished and

the peptides identiﬁ ed. For instance, if four samples were each labeled with a different isobaric

tag, in the ﬁ rst dimension of a MS separation a given peptide from each sample will appear in

the same peak, because the peptides will all have the same sequences and the isobaric tags labe-

ling them will all have the same mass signatures. Therefore, their mass/charge ratios will all be

the same and they will be indistinguishable at this point. However, upon MS/MS separation

wherein additional energy is used to promote fragmentation of the labeled peptides within the

peak (usually by collision-induced dissociation (CID) or electron capture dissociation (ECD)),

the peptides will break down into their respective charged amino acids and the isobaric tags

will be cleaved to release their reporter groups. From the peaks that result from this second

stage MS separation both the peptide sequence and the sample from which it came can be iden-

tiﬁ ed from the amino acids and the mass of the respective reporter groups.

There are many publications describing the development and use of isobaric tags. For

instance, Thompson et al. (2003) describes the development of tandem mass tags based on an

isotopically labeled peptide design. A later iteration of this concept uses commercially available

isotopes of alanine to form the reporter group, followed by a piperazine ring (as the charge-

carrying group), a second alanine used as the mass normalization group, and a proline residue,

which functions as the electrospray cleavable linker (Thompson et al., 2007). Ross et al. (2004)

used a set of four isobaric tags designed around an N-methyl piperazine group to study the glo-

bal protein expression of a wild-type yeast strain versus strains defective in certain pathways.

These tags are part of the iTRAQ (isotope tags for relative and absolute quantitation) system

from Applied Biosystems. Figure 16.12 shows some of the major structural characteristics of

commercial isobaric tags. Figure 16.13 illustrates the design of a multiplex isobaric set that

uses a piperidine ring as the reporter group, showing all of the various isotopic modiﬁ cations

done to balance the reporter and normalization group to create six unique tags (TMT system:

Tandem Mass Tags from Proteome Sciences).

Figure 16.11 The general structure of an isobaric mass tag reagent. The reactive group facilitates coupling to

discrete sites on peptides, such as amines. The reporter group creates a unique mass signal in MS

analysis and

its total mass is exactly balanced by the balance group by changing the stable isotopic labels to provide the

opposite mass differential as that of the reporter group. The result is that all isobaric tags have the same initial

molecular mass, but upon fragmentation in MS

, the reporter group is released and it provides the unique mass

signal to identify the sample being analyzed.

3. Isobaric Tags 663

Figure 16.12 A number of isobaric tags have been developed having different structural motifs. (a and b):

Isobaric tags used as part of the iTRAC reagents (Applied Biosystems). (c) An early design of isobaric tags using

a core of amino acid derivatives (Proteome Sciences). (d) The TMT tag design as commercialized by Proteome

Sciences. All of these isobaric tags contain a sensitization group to enhance mass spec detection, which is repre-

sented either by a tertiary amine that can be protonated to carry a positive charge or a guanidino group.

664 16. Mass Tags and Isotope Tags

Isobaric labels thus permit quantitative information regarding protein expression levels in

multiple samples analyzed simultaneously by MS. The multiplexed capability of these reagents

allows the measurement of peptides and proteins in diseased samples, treated samples, and nor-

mal samples all in the same experiment. In addition, since all peptides from a given protein get

labeled at their N-termini, the MS analysis generates more than one peptide signal, which can be

used to conﬁ rm protein identity with greater conﬁ dence than using a cysteine label, like ICAT.

The protocol for using isobaric tags differs from that described previously for the ICAT or

ECAT type reagents. In the following method, the proteins are denatured and the disulﬁ des

reduced and then alkylated to block them permanently. This eliminates disulﬁ de re-association

and also prevents the isobaric tags from forming thioester modiﬁ cation with cysteine thiols.

Next, the proteins are digested with trypsin and then modiﬁ ed with an isobaric tag. Each sam-

ple is labeled with a different isobaric compound so that the samples can be differentiated upon

MS/MS analysis.

Figure 16.13 Isobaric tags allow multiplexed analysis of different samples by changing the mass of the reporter

group and balancing that change by the opposite change in the balance group. This ﬁ gure illustrates the 6-plex

tag system from Proteome Sciences, which contains six different reporter groups (boxed areas). All of the rea-

gents contain amine-reactive NHS esters for modifying lysine side chains in proteins and peptides.

The following protocol illustrates the modiﬁ cation reaction and the handling of the protein

samples, but it is not meant to be instructive of mass spec techniques.

Protocol

1. Grow cells to 70–80 percent conﬂ uence and harvest by scraping the cells into 5 ml of

0.1 M sodium borate, pH 7.5. Avoid the use of amine-containing buffers, as these will react

with the NHS esters on the isobaric tags. Aliquot cell counts of approximately 2.5–5  10

for processing. Lyse cells using a detergent lysis buffer (e.g., Poppers, Thermo Fisher) or by

mechanical means. Centrifuge and discard the cellular debris. Measure the total protein

concentration using the BCA assay (Thermo Fisher) and adjust the protein concentration

to 1 mg/ml using 0.1 M sodium borate, 0.1 percent SDS, pH 7.5. A total of 100 g of each

protein sample can be used in this protocol. Separately process test samples and a control

sample made up of different cell populations. Simultaneous measurement can be done on a

total number of samples equal to the number of different isobaric tags available.

2. Reduce disulﬁ des in the protein sample by the addition of 2 l of 50 mM TCEP (Thermo

Fisher) to each 100 l aliquot of protein solution (ﬁ nal concentration 1 mM). Cover and

boil the samples for 10 minutes in a water bath to completely denature and reduce the

proteins. Alternatively, reduction may be done at 60°C for 1 hour. Avoid the use of thiol-

containing reductants, such as DTT, as these will react with the thiol blocking agent used

in the next step.

3. Add 6 l of iodoacetamide to each sample solution and react with mixing for 30 minutes

at room temperature.

4. Prepare a solution of TCPK-trypsin in 0.1 M sodium borate, pH 7.5, at a concentration

of 400 ng/ l. Add 10 l of the trypsin solution to each sample and incubate at 37°C for

12–16 hours or overnight with mixing.

5. Dissolve the isobaric tagging reagents in acetonitrile or ethanol at a concentration of

50 mM (or according to the manufacturer ’s recommendations). Use a fume hood to han-

dle organic solvents.

6. Add a quantity of the appropriate isobaric tag solution to each sample to provide a ﬁ nal

concentration of 10–20 mM. This quantity of reagent will assure a large molar excess

of reagent over the concentration of peptides present in order to modify completely all

peptides at their N-terminus. Note that the -amino groups of lysine residues also will be

modiﬁ ed by this procedure. React for 1 hour at room temperature.

7. To eliminate acylation products at tyrosine residues, add 1 l of 15 M hydroxy-

lamine solution in water to each protein sample and incubate for 30 minutes at 37°C

(Zappacosta et al., 2006).

8. Combine the contents of each labeled sample into a single tube and mix by vortexing,

then centrifuge.

9. Before LC–MS/MS analysis, the sample must be cleaned up to remove excess salts, SDS,

reducing agent, and cell lysis buffer components. This typically is done by using a strong

cation exchange matrix. Protocols may be found for this procedure in the instruction

booklets related to the use of isobaric tagging reagents (e.g., Proteome Sciences, Applied

Biosystems, or Perkin Elmer/Agilix).

3. Isobaric Tags 665

666

The many dozens of reactions that are available for bioconjugation purposes generally are

designed to work with biological molecules and the functionalities they contain. The main goal

of most bioconjugate techniques is to use the functional groups on biomolecules to label with

another type of biomolecule or link to synthetic probes. However, it is often desirable to cou-

ple one molecule to another without the potential for cross-reactivity with biomolecules. The

term “chemoselective ligation ” has been coined to describe the coupling of one reactive group

speciﬁ cally with another reactive group without side reactions in aqueous solution or in the

presence of biological material (Lemieux and Bertozzi, 1998).

Unfortunately, the incredible diversity of biomolecules in cells and organisms presents prob-

lems for this goal of total chemoselectivity and bioorthogonality, as the number and variety of

reactive sites on the molecules of life is extraordinary. Even the best crosslinkers or labeling

reagents designed to be somewhat site-speciﬁ c in their reactions often display cross-reactivity

with functional groups on biomolecules other than the ones intended for coupling. For instance,

N-hydroxysuccinimide (NHS) esters usually are considered amine reactive, but they also react

with cysteine, histidine, serine, threonine, and tyrosine side-chain groups. Maleimide groups too

are touted as being thiol speciﬁ c, but amines also can react with maleimides given the right

conditions.

Bioorthogonal reagents ideally should contain a reactive group that only will react with

another speciﬁ c reactive group without any potential for cross-reactivity with biomolecule func-

tionalities. In other words, a bioorthogonal reactive group could be added to a complex mixture

of biological molecules in aqueous solution without reacting with any of them. Moreover, the

ideal bioorthogonal system should be immune to instability in aqueous solutions, such as the

tendency to hydrolyze or easily oxidize. True bioorthogonal reagents of this type will link to each

other and only each other in the presence of intracellular environments or in cell lysates or in

deﬁ ned biomolecule solutions.

The reality is that there are few options in this category of bioconjugate reagents and each of

the systems or reactant pairs that have been developed for bioorthogonal applications have dif-

fering degrees of how well they perform in this task. The following sections describe the major

chemoselective ligation reactions, which can be considered to have a degree of bioorthogonal

characteristics. In each system, the chemoselective pair of reactants can be separately linked

or built into the design of crosslinkers or labeling reagents and used to modify biomolecules,

Chemoselective Ligation: Bioorthogonal Reagents

surfaces, particles, or organic compounds. Subsequently, these labeled components can be

brought together, even in complex solutions, to facilitate conjugation between the two

bioorthogonal reacting species.

In addition, for several reactant strategies in chemoselective ligation, one of the reagent pairs

can be designed into a biological monomer that can be utilized by cellular processes to become

incorporated into biopolymers. This advantage provides a unique in vivo labeling capability

through the feeding of monomer analogs to cells or organisms, such as modiﬁ ed amino acids

or sugar derivatives, which then get selectively added into proteins, carbohydrates, or lipids.

Thus, the ultimate application of bioorthogonal chemoselective ligation is to label speciﬁ cally

the molecules of life directly within living systems and without cross-reactions with other bio-

logical functional groups. For a review on the use of chemoselective reactions in living systems,

see Prescher and Bertozzi, (2005).

1. Diels–Alder Reagent Pairs

The Diels–Alder reaction has long been a staple for forming carbon–carbon bonds in organic

synthesis (Smith and March, 2007). The typical reaction proceeds through the 2  4 cycloaddi-

tion of a double bond (alkene) and a diene to give a 6-member ring product. The double bond

reactant is often called a dienophile, and electron-withdrawing constituents next to the alkene

are used to accelerate the reaction (such as COOH, CHO, and COR groups, among others).

Conversely, electron-donating groups on the diene are important for increasing reaction rates

(Figure 17.1 ).

Some reports indicated that the Diels–Alder reaction could be done in aqueous environ-

ments with a potential for accelerated reaction rates under the right conditions (Rideout and

Breslow, 1980; Blokzijl and Engberts, 1992; Pai and Smith, 1995; Otto et al., 1996; Wijnen

and Engberts, 1997), and the addition of InCl

was determined to act as a catalyst in aqueous

environments (Loh et al., 1996). For a review of organic reactions that can be done in aqueous

media, see Li (2005).

Figure 17.1 A general Diels–Alder reaction consists of a 4  2 cycloaddition between a diene and an alkene,

often called a dienophile. The reaction rate and yield increase if the diene contains an electron-donating group

and the alkene contains an electron-withdrawing group.

1. Diels–Alder Reagent Pairs 667

668 17. Chemoselective Ligation: Bioorthogonal Reagents

In addition, it has been discovered that there are naturally occurring enzymes that facilitate

Diels–Alder type reactions within certain metabolic pathways and that enzymes are also instru-

mental in forming polyketides, isoprenoids, phenylpropanoids, and alkaloids (de Araujo et al .,

2006). Agresti et al. (2005) identiﬁ ed ribozymes from RNA oligo libraries that catalyzed multiple-

turnover Diels–Alder cycloaddition reactions.

In 2001, Hill et al. extended the use of aqueous phase Diels–Alder reactions for the bio-

conjugation of diene-modiﬁ ed oligonucleotides. The dienophile that was used consisted of a

simple maleimide derivative, which is present on a broad range of commercially available bio-

conjugation reagents. Modiﬁ ed oligonucleotides were prepared by solid phase synthesis using

a 3,5-hexadiene phosphoramidite derivative, which could be incorporated into the oligo at the

5 end. Maleimide compounds investigated for oligo bioconjugation include N -ethylmaleimide,

biotin–BMCC, ﬂ uorescein–maleimide, coumarin–maleimide, maleimide–PEG

–biotin, and an

mPEG–maleimide. Figure 17.2 shows the reaction of maleimide–PEG

–biotin with the diene-

modiﬁ ed oligo to yield the cycloaddition product.

The reaction kinetics between a maleimide derivative and a 3,5-hexadiene derivative var-

ies depending on the maleimide compound being reacted. Cycloaddition yields of greater than

80 percent and often as much as 90–95 percent can be expected within 1–18 hours at room

temperature or slightly elevated reaction conditions (e.g., 30 °C).

The conjugation of oligonucleotides with peptides also can be done using Diels–Alder cyclo-

additions in water (Tona and Häner, 2005). Marchan et al. (2006) used the same 3,5-hexadiene

phosphoramidite derivative as Hill et al. (2001), but in this case used a maleimide-modiﬁ ed

Figure 17.2 Maleimide groups provide good dienophiles for a Diels–Alder reaction. Biotin–PEG

–maleimide

can react with an oligo-diene molecule to form a covalent cycloaddition product, which adds the biotin tag to

the oligo.

peptide sequence. A mild, aqueous Diels–Alder reaction between them resulted in the forma-

tion of the cycloaddition product.

The Diels–Alder reaction for bioconjugation also has been used for the chemoselective liga-

tion of peptides and proteins in aqueous solution (de Araujo et al., 2006). Peptides modiﬁ ed

using a 2,4-hexadienyl ester were derivatized to contain a diene and were found to be reac-

tive toward other peptides containing an N-terminal maleimide group to give the cycloaddition

product in high yield. The hexadienyl group also was attached to biotin and allowed to interact

with streptavidin, which then could be conjugated with peptides containing a maleimide group.

Diels–Alder cycloaddition reactions also have been used to link covalently carbohydrates

to proteins (Pozsgay et al., 2002) as well as for the immobilization of oligonucleotides on

glass surfaces to create arrays (Latham-Timmons et al., 2003). In an application that used two

chemoselective ligation reactions, Sun et al. (2006) employed sequential Diels–Alder and azide–

alkyne (click chemistry) cycloaddition reactions to immobilize protein, biotin, or carbohydrate

ligands on solid surfaces. In this case, glass slides containing maleimidocaproyl groups were

used as the dienophiles and a PEG

spacer containing an alkyne on one end and a cyclopen-

tadiene at the other end was the reactive linker. A Diels–Alder reaction coupled the maleimide

groups to the cyclopentadiene groups on the spacer, while the alkyne groups at the other end

were used in a click chemistry reaction to attach azide-containing ligands ( Figure 17.3 ). The

cycloaddition reaction between the maleimide groups on the slides and the cyclopentadiene

group on the spacer was done in a 1:1 solution of water: tert-BuOH at room temperature for

12 hours. After washing with the same water/solvent mixture, the slides contained hydrophilic

spacers terminating in alkyne groups, which then could be coupled with the azide-containing

ligands using click chemistry (see Section 4, this chapter).

Chemoselective ligation reactions using the Diels–Alder cycloaddition process offer another

bioconjugation route using a reactive component available commercially (maleimide-containing

reagents). However, their use as true bioorthogonal reactants is limited due to the cross-reactivity

of maleimides toward thiols. For instance, in the modiﬁ cation of a Rab protein, a cysteine resi-

due had to be protected prior to cycloaddition using a maleimide compound (de Araujo et al .,

2006); otherwise, the maleimide would have coupled to the sulfhydryl, too.

2 . Hydrazine–Aldehyde Reagent Pairs

The reaction between an aldehyde or ketone and a hydrazide or hydrazine derivative to form a

hydrazone bond has been frequently used for bioconjugation purposes (Chapter 2, Section 5.1).

The reaction is appealing from a bioorthogonal perspective, because natural biopolymers don ’t

normally contain these reactive groups. Although aldehydes may form temporary Schiff base

interactions with amines on proteins and other biomolecules, in aqueous solution they are fully

reversible and will rapidly exchange with a hydrazide or hydrazine, if present. This also holds

true of aminoxy (hydroxylamine) derivatives, which form stable oxime bonds with aldehydes,

although the number of reagents available with an aminoxy functionality is limited.

The only potential problem of cross-reactivity for this chemoselective reaction pair with

molecules of biological origin might occur from a hydrazine reacting with aldehyde or ketone

containing metabolic intermediates, reducing sugars, or similar small organic molecules present

within cells or cell lysates. However, if the hydrazine reagent is added in sufﬁ cient excess,

the desired coupling reaction still will occur in such environments, as evidenced by the many

2. Hydrazine–Aldehyde Reagent Pairs 669