Westermeier R., Naven T., H?pker H.-R. Proteomics in Practice: A Guide to Successful Experimental Design

Подождите немного. Документ загружается.

3 Mass Spectrometry230

3.2.4

Quadrupole Time-of-Flight

Initially described in 1996 (Morris et al. 1996) for oligosaccharide ana-

lysis and more recently by Lobada et al. (2000), these quadrupole

time-of-flight (QTOF) instruments rapidly became the instrument

standard for MS/MS applications within proteomics. By combining a

mass filtering quadrupole analyzer and a collision cell (rf-only quad-

rupole) with a non-scanning reflectron TOF analyzer, the user is able

to acquire MS and, most notably MS/MS data with high mass accu-

racy, resolution and sensitivity. The instrument is typically coupled

with HPLC (Bell et al. 2001; Gavin et al. 2002) and is predominantly

used in MS/MS mode (see Figure 3.11).

Fig. 3.11: Schematic of a quadrupole TOF hybrid analyzer.

It is commonly coupled with ESI for proteomics applications.

Morris HR, Paxton T, Dell A,

Langhorne J, Berg M, Bordoli

RS, Hoyes J, Bateman RH.

Rapid Commun Mass Spec-

trom 10 (1996) 889–896.

Shevchenko A, Chemushevich

IV, Ens W, Standing KG,

Thomson B, Wilm M, Mann

M. Rapid Commun Mass Spec-

trom 11 ( 1997) 1015–1024.

Bell AW, Ward MA, Blackstock

WP, Freeman HNM,

Choudhary JS, Lewis AP,

Chotai D, Fazel A, Gushue JN,

Paiement J. J Biol. Chem 276

(2001) 5152–5165.

Gavin et al. Nature 415 (2002)

141–147.

3.2 Ion Separation 231

Later developments have seen a MALDI ion source coupled with a

hybrid instrument enabling similar MS/MS performance to be

attained from singly charged ions produced by MALDI. (Lobada et al.

2000; Shevchenko et al. 2000; Baldwin et al. 2001).

This configuration enables a peptide mass fingerprint and MS/MS

data with high resolution and mass accuracy to be acquired using the

same instrument. A similar approach has been developed with the

TOF/TOF analyzer.

3.2.5

Hybrid Triple Quadrupole Linear Ion Trap

As the name suggests, the hybrid triple quadrupole linear ion trap

(QTRAP) instrument combines the features of a triple quadrupole

and a linear ion trap analyzer. For proteomics applications, this

equates to fast scans with high sensitivity, accurate quantification and

complete product ion, precursor ion and neutral loss tandem mass

spectrometry functionality. The additional scan functionality has

been shown to be useful for post translational modification analysis

(Sandra et al. 2004).

The QTRAP is predominantly used in MS/MS mode.

3.2.6

TOF/TOF Analyzer

A TOF/TOF analyzer coupled with a MALDI ion source (Medzih-

radszky et al. 2000) enables the generation of peptide mass finger-

print data and peptide sequence data in a single instrument. Basi-

cally, two TOF analyzers are separated by a collision cell, with the first

TOF analyzer, used for precursor ion selection. (see Figure 3.12).

High energy collisions occur within the collision cell, and the second

TOF analyzer resolves the ions. The configuration allows for high

sensitivity and high resolution in both MS and MS/MS modes, and is

capable of reducing protein identification to a one tier process. When

coupled with nano-LC using an LC-MALDI spotter, this configuration

is a powerful tool for the proteomics researcher.

Lobada A, Krutchinsky A,

Bromirski M, Ens W, Standing

KG. Rapid Commun Mass

Spectrom 14 (2000)

1047–1057.

Shevchenko A, Loboda A,

Shevchenko A, Ens W,

Standing KG. Ana. Chem 72

(2000) 2132–2141.

Baldwin MA, Medzihradszky

KF, Lock CM, Fisher B, Setti-

neri TA, Burlingame AL. Anal

Chem 73( 2001) 1707–1720.

Sandra K, Devreese B, Van

Beeumen J, Stals I, Claeyssens

M. J Am Soc Mass Spectrom

15 (2004) 413–423.

Medzihradszky KF, Campbell

JM, Baldwin MA, Falick AM,

Juhasz P, Vestal ML, Burlin-

game AL. Anal Chem 72

(2000) 552–558.

3 Mass Spectrometry232

Fig. 3.12: Schematic of a TOF/TOF analyzer. Coupled with MALDI

to perform peptide mass fingerprinting and peptide sequence analysis

by MS/MS. Often coupled with a nanoLC spotter in LC-MALDI.

3.2.7

Fourier Transform Ion Cyclotron

Similarly to an ion trap, a Fourier transform ion cyclotron (FT-ICR)

instrument is capable of trapping and storing ions (see Figure 3.13).

The ICR cell resides in a strong magnetic field and consists of three

parallel plates arranged in a cube. In the cell, ions of a given m/z ratio

have a given cyclotron frequency of a given orbit radius. On applying

an rf voltage at the same frequency as the cyclotron frequency, the

respective ions absorb energy and are accelerated to a larger orbit

radius. When the rf voltage is removed the energized, accelerated

ions still rotate at a constant radius. Ions which have a different cyclo-

tron frequency remain unexcited and hence ions of differing mass

can be separated. As the cyclotron frequency of an ion is determined

by its mass to charge ratio, a Fourier transform can be performed on

the signal to determine the mass of the ion. FT-MS affords high sen-

sitivity and exceptional resolution and thus mass accuracy.

Fig. 3.13: Schematic of the fundamental operation of an FT ICR mass

spectrometer. It is commonly coupled with ESI for proteomics applications.

3.3 Generating MS Data for Protein Identification 233

3.2.8

Orbitrap

The Orbitrap analyzer is the most recently of developed analyzers

(commercially available from Thermo Scientific since 2005). It is not

a conventional ion trap, in that there is no RF or magnet to hold the

ions in the cell. The analyzer traps ions in an electrostatic field, with

the attraction towards a central electrode. The electrode confines the

ions so that they move or orbit in complex spiral patterns (Makarov

2002, and Hardman and Makarov, 2003). A Fourier Transform is

used to extract the masses accurately from the oscillation frequencies

This type of analyzer affords high resolution, high mass accuracy and

an increased dynamic range, similar in performance to an FT-ICR

instrument and high sensitivity (Olsen et al. 2005). This type of mass

spectrometer can be used in both MS and MS/MS modes.

The Orbitrap has recently been coupled with a linear ion trap

(Finnigan LTQ-Orbitrap, Thermo) (see Figure 3.14). This config-

uration is a hybrid consisting of the two mass analyzers, both capable

of detecting ions and recording spectra: the two analyzers can be

used independently or in conjunction. The instrument can be used

for both bottom up (peptide based approach to protein ID) and top

down (protein based approach to protein ID) approaches (see Section

3.6.2).

Fig. 3.14: Schematic of the hybrid LTQ-Orbitrap analyzer.

3.3

Generating MS Data for Protein Identification

Generally, two different approaches, or combinations thereof, can be

used for protein identification (ID): peptide mass fingerprinting and

tandem mass spectrometry (MS/MS). In the first instance, the mass/

charge ratio (m/z) of the analytes are measured and the calculated

masses are indicative of the composition of the individual analytes.

Hardman M, Makarov A. Anal

Chem 75 (2003) 1699–1705.

Makarov A. Anal Chem 72

(2000) 1156–1162.

Olsen JV, de Godoy LMF, Li G,

Macek B, Mortensen P, Pesch

R, Makarov A, Lange O,

Horning S, Mann M. Mol Cell

Proteomics 4 (2005) 2010–

2021.

3 Mass Spectrometry234

In the second instance, tandem mass spectrometry (MS/MS) is used

to acquire peptide sequence information, which is used to identify

and characterize proteins with high specificity (see Figure 3.15).

Fig. 3.15: Schematic of the methods of protein identification

using a mass spectrometer: Peptide mass fingerprinting is performed

in MS mode and peptide sequence analysis in MS/MS mode.

3.3.1

Peptide Mass Fingerprint

The technique originally described in 1993 comprises protein diges-

tion, MALDI TOF analysis and sequence database search algorithms.

(Pappin et al. 1993; Henzel WJ et al. 1993; Mann et al. 1993; Yates et

al. 1993; James et al. 1993).

Simply, every protein in the protein database (or in the genome

under investigation) is theoretically digested with the cleavage

reagent used in the digestion reaction, generating many hundreds of

thousands of theoretical peptides. The experimental peptide masses

derived from the MS spectrum, the peptide mass fingerprint (PMF)

is then compared to the theoretical peptide masses and a score is cal-

culated and assigned. The score reflects the match between the theo-

retically and experimentally determined masses; the protein identi-

fied as the most probable is the one that gives the best match between

the experimental and the theoretical peptides. The number of pep-

tides observed in the PMF and the accuracy to which they are mea-

sured determines the confidence of the protein identification.

Pappin DJC, H jrup P, Bleasby

A. Curr Biol 3 (1993) 327–332.

Henzel WJ, Billeci TM, Stults

JT, Wong SC, Grimley C, Wata-

nabe C. Proc Natl Acad Sci

USA 90 (1993) 5011–5015.

Mann M, Hojrup P, Roepstorff

P. Biol Mass Spectrom 22

(1993) 338–345.

Yates JR III, Speicher S, Griffin

PR, Hunkapiller T. Anal

Biochem 214 (1993) 397–408.

James P, Quadroni M, Carafoli

E, Gonnet G. Biochem Biophys

Res Commun 195 (1993)

58–64.

3.3 Generating MS Data for Protein Identification 235

The incorporation of reflectron technology and delayed extraction

(timed delay ion extraction) into MALDI-TOF instrumentation has

enhanced the performance of peptide mass fingerprinting consider-

ably. The

C isotope of the peptide isotopic envelope, the mass used

in a PMF search, can be unambiguously assigned across the peptide

mass range of interest, significantly enhancing mass measurement

of this peak (sub 50 ppm mass accuracy is routine). The improved

mass accuracy further constrains the database search, reducing the

potential for ambiguous protein identifications (Jensen et al. 1996;

Clauser et al. 1999).

Several programs are available to perform this type of search, vary-

ing in the execution of the task (including MASCOT at www.matrix-

science.com; and Phenyx at http://phenyx.vital-it.ch/pwi/login/

choice. Regardless which program is used, four user variables are

important for a PMF search:

Peptide mass list.

Specification of the cleavage agent.

Error tolerance. The accuracy of mass measure-

ment is determined by the calibration, the

higher the mass accuracy the greater the specifi-

city.

Knowledge of peptide modifications i.e. methio-

nine oxidation.

For this approach to give an unambiguous result, a significant num-

ber of experimentally determined peptide masses have to match the

experimental masses, the peptides have to be a result of the correct

cleavage specificity of the cleavage agent used and the protein in

question has to exist in the protein database. As a result, the PMF

approach is highly suited to highly characterized genomes and has

been a core proteomics technique. Shevchenko et al. (1996) demon-

strated that up to 90% of proteins selected from a 2D gel of an E coli

lysate were identified by peptide mass fingerprinting.

Although an unambiguous match may be returned from a database

search, not all the peptides within the peptide mass fingerprint may

be identified. It is important to identify the remaining unidentified

peptide peaks, which may arise from:

The presence of other digested proteins within

the sample (in the case of 2D gel, comigration to

the same spot);

Peptides arising from the simultaneous diges-

tion of contaminant proteins such as keratin;

Jensen ON, Podtelejnikov A,

Mann M. Rapid Commun

Mass Spectrom 10 (1996)

1371–1378.

Clauser KR, Baker P, Burlin-

game AL. Anal Chem 71

(1999) 2871–2882.

Shevchenko A, Jensen ON,

Podtelejnikov A, Sagliocco F,

Wilm M, Vorm O, Mortensen

P, Shevchenko A, Boucherie H,

Mann M. Proc Natl Acad Sci

USA 93 (1996) 1440–1445.

Though many keratin peaks are

commonly known and these

peptides, if present, can be

subsequently omitted from the

search.

3 Mass Spectrometry236

Incorrect cleavage, protease acting in a non-pre-

dicted fashion;

Post-translationally modified peptides, altering

the mass of the native peptide rendering it

unrecognizable in the database search;

The identified protein may be (highly) homolo-

gous to the protein under study, but not exactly

the same and thus unmatched peptides will

exist.

MALDI-TOF PMF is fast and simple method of protein identification,

but the success of the method can be compromised in a number of

ways:

Insufficient peptides are observed in the peptide

mass fingerprint to submit to a database search,

i.e. insufficient information to identify the pro-

tein.

The sample maybe a mixture of proteins.

Although it is possible to analyze simple mix-

tures (Axelsson et al. 2001) there may not be a

significant number of peptides from each pro-

tein component in the mixture to yield a success-

ful identification. As several peptides are

required to give a statistical match, this approach

is not readily suited to complex mixtures

Excessive post-translational modifications of the

protein can result in masses, which are not pre-

dicted in the theoretical digestion of the proteins

in the database as well as precluding peptides

from the fingerprint.

Very little homology can be found with another

protein in the database or the protein in question

may not actually exist in the protein database.

If the above is true, peptide mass fingerprinting

cannot be used with confidence for searching

against the EST databases (which is the next

stage of protein identification if a search of the

protein database is unsuccessful). The probabil-

ity that a significant number of peptides from

the peptide mass fingerprint will find matches

with a single EST yielding an unique protein

identification is small.

If any of these situations arise, then more specific information is

required to identify the protein.

Common for small molecular

weight proteins, basic proteins

if digested with trypsin low

abundant proteins.

Axelsson J, Boren M, Naven

TJP, Fenyç D. Proceedings of

the 49

ASMS conference on

mass spectrometry and allied

topics, Chicago ( 2001).

3.3 Generating MS Data for Protein Identification 237

3.3.2

Peptide Mass Fingerprint Combined With Composition Information

Historically, it has been possible to combine PMF information with

compositional information. This was a common technique prior to

the introduction of delayed extraction and reflectron technology. The

composition information provided orthogonal information which

provided additional specificity for a database search by enabling

further constraints on that search (Fenyç et al. 1998). For instance,

this publication demonstrated that with the knowledge of the pre-

sence of a cysteine residue in a tryptic peptide of mass 2000 Da (mea-

sured to 0.5 Da mass accuracy) the number of matching proteins in S

cerevisiae was reduced by a factor of five. A number of approaches

have been employed to gain compositional information. In one exam-

ple reported by Pappin, a PMF is acquired and the peptide masses

noted. A subsequent methyl esterification reaction is performed on a

small aliquot of the sample esterifying the carboxyl side-chain of the

acidic residues, glutamic and aspartic acid, and the carboxylic group

of the C-terminal residue present in each peptide of the digest. A

PMF spectrum is re-acquired. The subsequent mass changes between

corresponding peptides are indicative of the number of acidic groups

each peptide contains. This information was combined with the origi-

nal PMF using a MOWSE composition search potentially increasing

search discrimination by orders of magnitude (Pappin et al. 1995).

The number of exchangeable hydrogens within a peptide offers

composition information. Again, this procedure requires the acquisi-

tion of a PMF, subsequent labeling of the mixture with a deuterium

solution (D

O) and finally re-acquisition of a second spectrum. The

mass increase of each peptide by the number of exchangeable hydro-

gens is indicative of amino acid composition (Sepetov et al. 1993;

James et al. 1994).

Moreover Goodlett et al. (2000) demonstrated composition informa-

tion and high mass accuracy can be very specific information for

unambiguously identifying proteins. By labeling cysteine residues

with a specific isotopic label, IDEnT (the tag contained chlorine,

which has a specific isotopic profile owing to the relative abundance

Cl and

Cl) the distinctive isotopic pattern of the labeled peptide

could be simply recognized. With this specific composition informa-

tion and mass accuracy measured to within 1 ppm using an FT-ICR

MS, the mass of a single peptide was sufficient to unambiguously

identify a protein from the whole yeast database (1 peptide from a

possible in ~345,000 peptides).

Fenyç D, Qin J, Chait BT. Elec-

trophoresis 19 (1998) 998–

1005.

Pappin D, Rahman D, Hansen

HF, Bartlet-Jones M, Jeffrey W,

Bleasby AJ. Methods in Biolo-

gical Sciences (1996) 135–150.

Sepetov NF, Issakova OL, Lebi

M, Swiderek K, Stahl DC, Lee

TD. Rapid Commun Mass

Spectrom. 7 (1993) 58–62.

James P, Quadroni M, Carafoli

E, Gonnet G. Protein Sci 3

(1994) 1347–1350.

Goodlett DR, Bruce JE,

Anderson GA, Rist B, Pasa-

Tolic L, Fiehn O, Smith RD,

Aebersold R. Anal Chem 72

(2000) 1112–1118.

Cysteine was chosen as the

labeled residue because it is one

of the rarest amino acid resi-

dues, allowing constrained

database searching.

3 Mass Spectrometry238

The isotopical labeling of cysteine residues has also been exploited

for extra composition information. Sechi and Chait (1998) reported a

method for the modification of the cysteinyl thiol group with isotopi-

cally labeled acrylamide.

Additionally, composition information can be inferred from immo-

nium ions contained within a MALDI PSD spectrum, this is particu-

larly discriminating information. Immonium ions present in a

MALDI PSD spectrum indicate the presence of certain amino acid

residues; again this can be included in a composition search as

described above. (Clauser et al. 1999). This information is also avail-

able in a product ion MS/MS spectrum acquired with a MALDI

QTOF or TOF/TOF instrument.

However, with the ubiquitous MS/MS functionality available in

many laboratories, these approaches are seldom used today.

3.3.3

Peptide Mass Fingerprint Combined With Partial Sequence Information

Again, historically, supplementing PMF data with some partial

sequence has been an attractive approach; PMF and sequence infor-

mation can be used in the same search using the sequence query

function with the MASCOT search engine. An unsuccessful search of

the protein database with the peptide mass fingerprint data may be

overturned with some partial sequence if the protein exists in the pro-

tein database.

The data required for this type of database search can be obtained

with a MALDI-TOF instrument capable of post-source decay. Post-

source decay (PSD) is a technique performed on a MALDI TOF

instrument equipped with a reflectron, though not genuine MS/MS,

it can generate stretches of peptides sequence which can be used in a

search engine querry of a protein database. A worked example is

demonstrated below whereby a PMF is acquired prior to the chemical

derivatization of the peptide digest with chemically assisted fragmen-

tation (CAF; see Section 3.3.4.4) reagent and subsequent CAF-PSD is

performed.

An unsuccessful result was obtained from the PMF search (Figure

3.16). Following chemical derivatization, PSD was performed on two

of the derivatized peptides. The spectra of each peptide include sev-

eral sequence specific C-terminal fragment ions (Figures 3.17a,

3.18a), which is indicative of partial amino acid sequence can be

inferred for both peptides from. Each sequence is combined with the

peptide mass fingerprint in a sequence query search using MASCOT.

Sechi S, Chait BT. Anal Chem

70 (1998) 5150–5158.

Clauser KR, Baker P, Burlin-

game AL. Anal Chem 71

(1999) 2871–2882.

3.3 Generating MS Data for Protein Identification 239

Combining the interpreted sequence derived from the PSD spectrum

of peptide m/z 1569.79 with the original PMF in a database sequence

query search, yields an unambiguous match (Figure 3.18b). Similarly,

an unambiguous match is returned when the determined sequence

from the second peptide (peptide m/z 1157.57) is combined with the

PMF in a sequence query search (Figure 3.18c). When both sets of

sequence data are combined with the peptide mass fingerprint the

confidence in the match returned from the database search increases

dramatically (Figure 3.18c). Close inspection of the PSD spectra

demonstrated that the sequence can readily determined using the

CAF chemistry. The CAF reagent generates only C-terminal fragment

ions (see Section 3.3.4.4), which makes the sequence relatively simple

to interpret. In the case of the second peptide, there is some ambigu-

ity whether the residue is a glutamine or a glutamic acid (experimen-

tally measured difference 128.7 Da) in part of the determined

sequence. However, the MASCOT search query allows both amino

acid residues to be inputted into the search query (see Step 15 on

page 407). As the figures demonstrate a short stretch of determined

sequence combined with the PMF is a powerful approach for re-

ducing ambiguous protein identifications; particularly if genuine

MS/MS analysis is not an option.

Fig. 3.16: An unsuccessful PMF search of the protein database.