ion fragmentation in mass spectrometry 01-16-07f.pdf · ion fragmentation in mass spectrometry ......
TRANSCRIPT
1
S. Barnes - J. Prasain 01/16/07
Ion fragmentation in mass spectrometry
Stephen Barnes and Jeevan [email protected] [email protected]
4-7117 6-2612
S. Barnes - J. Prasain 01/16/07
Lecture goals• Value of fragmentation in determining
structure• How peptides fragment
– Interpreting the tandem mass spectrum• Automating identification of peptides from
their fragment ions– pros and cons
• Controlling fragmentation– Choice of ionization and fragmentation
methods
2
S. Barnes - J. Prasain 01/16/07
Why ion fragmentation provides useful information
• Compounds can have the same empirical formula, i.e., the same molecular weight or m/z, but be different chemically.
• Breaking them into parts (fragmenting them) helps to identify what they are.
• Each of the following gives the same [M+2H]2+ ion– NH2VFAQHLK-COOH NH2VAFQHLK-COOH– NH2VFQHALK-COOH NH2VHLAFQK-COOH
• In proteomics we want to distinguish these peptides
S. Barnes - J. Prasain 01/16/07
What is MS-MS (tandem mass spectrometry)?
Analyzer 1 Analyzer 2Molecular ions selected ion
gas
3
S. Barnes - J. Prasain 01/16/07
Fragmenting a peptidex2
a2R1 O R2
| || |H2N--C--C--N+=C
| | |H H H
R3 O R4
| || |+O=C--HN--C--C--N--C--COOH
| | |H H H
a2x2
http://www.matrixscience.com/help/fragmentation_help.html
NH3+-CHR1-CO-NH-CHR2-CO-NH-CHR3-CO-NH-CHR4-COOH
R1 O R2
| || |H2N--C--C--N=C--C=O+
| | |H H H
R3 O R4
| || |H3N+--C--C--N--C--COOH
| | |H H H
b2 y2
y2
b2
R1 O R2 O| || | ||
H2N--C--C--N=C--C--NH3+
| | |H H H
R3 O R4
| || |+C--C--N--C--COOH| | |H H H
c2 z2
z2
c2
S. Barnes - J. Prasain 01/16/07
bn = [residue masses + 1]yn = [residue masses + H2O + 1]
Calculating expected b- and y-ion fragments
Alanine 71.037 Leucine 113.084Arginine 156.101 Lysine 128.094Asparagine 114.043 Methionine 131.040Aspartic acid 115.027 Phenylalanine 147.068Cysteine 103.009 Proline 97.053Glutamic acid 129.043 Serine 87.032Glutamine 128.058 Threonine 101.048Glycine 57.021 Tryptophan 186.079Histidine 137.059 Tyrosine 163.063Isoleucine 113.084 Valine 99.068
ADGTWLEVRb3 = ADG, 71.04+115.03+57.02+1= 244.09
y3 = EVR, 129.04+99.07+156.10+18+1= 403.21
4
S. Barnes - J. Prasain 01/16/07
Identification of daughter ions and peptides - reading the sequence
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300mass0
100
%
y101100.61
y9987.52
b3375.21
b2262.12
234.13
y1175.12
y2274.19
y8916.49
parent681.45
y4487.32
y5602.32
y6730.43
b8875.42
y111247.69
y7859.45
b4446.23
S. Barnes - J. Prasain 01/16/07
Identification of daughter ions and peptide sequence
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300mass0
100
%
y101100.61
y9987.52
b3375.21
b2262.12
234.13
y1175.12
y2274.19
y8916.49
parent681.45
y4487.32
y5602.32
y6730.43
b8875.42
y111247.69
y7859.45
b4446.23
b ions 262 375 446 503 632 760 875 989 1088 1187 1343 N F L A G E K D N V V R
y ions 1361 1247 1100 987 916 859 730 602 487 373 274 175
5
S. Barnes - J. Prasain 01/16/07
What’s in a peptide MSMS spectrum?
• In most cases, some, but rarely all, of the theoretic b- and y-ions are observed
• Besides b- and y-ions, other types of fragmentation can occur to form an and xnions, as well as also losing CO, NH3 and H2O groups
• Internal cleavage reactions can occur at acidic (Asp - Glu) residue sites
S. Barnes - J. Prasain 01/16/07
Immonium and Related Ions84.08 70.07
87.06 120.08 86.10 --- --- 102.05 101.11 88.04 87.06 72.08 72.08 87.09129.10 100.09
112.09N-terminal ionsa-NH3 ions --- 217.10 330.18 401.22 458.24 587.28 715.38 830.40 944.45 1043.52 1142.58 ---a ions --- 234.12 347.21 418.24 475.27 604.31 732.40 847.43 961.47 1060.54 1159.61 ---b-NH3 ions --- 245.09 358.18 429.21 486.23 615.28 743.37 858.40 972.44 1071.51 1170.58 ---b-H2O ions --- --- --- --- --- 614.29 742.39 857.42 971.46 1070.53 1169.59 ---b ions --- 262.12 375.20 446.24 503.26 632.30 760.40 875.43 989.47 1088.54 1187.61 ---
1 2 3 4 5 6 7 8 9 10 11 12H - N F L A G E K D N V V R
y ions --- 1247.67 1100.61 987.52 916.48 859.46 730.42 602.33 487.30 373.26 274.19 175.12y-NH3 ions --- 1230.65 1083.58 970.50 899.46 842.44 713.39 585.30 470.27 356.23 257.16 158.09
y-H2O ions --- 1229.66 1082.60 969.51 898.47 841.45 712.41 584.32- --- --- -- ---
Other ions observed in peptide fragmentation
6
S. Barnes - J. Prasain 01/16/07
Towards automated MSMS sequencing
• The 2D-LC-ESI-MSMS method (MuDPIT) generates 50,000+ MSMS spectra for each sample
• If it takes 15 min to hand interpret one MS-MS spectrum, then it would take 12,500 hours to complete the analysis. For someone working 8 hours/day and a five-day week, this would be about 6 years!
• Using SEQUEST and MASCOT, methods were developed to use computer-driven approaches to analyze MSMS data
S. Barnes - J. Prasain 01/16/07
Issues in MS-MS experiment• At any one moment, several peptides may
be co-eluting• Data-dependent operation:
– The most intense peptide molecular ion is selected first (must exceed an initial threshold value)
– A 2-3 Da window is used (to maximize the signal)
– The ion must be in 2+ or 3+ state– Since the ion trap scan of the fragment
ions takes ~ 1 sec, only the most intense ions will be measured
– However, can use an exclusion list on a subsequent run to study minor ions
threshold
7
S. Barnes - J. Prasain 01/16/07
The SEQUEST approach• Each observed MSMS spectrum has a corresponding molecular ion
[M+nH]n+. For ion trap data, ions are selected from the known or virtualproteome that are within 1 Da. These are then “fragmented” in silico to produce b- and y-ions and less abundant fragment ions.
• The cross correlations of the observed MSMS spectrum to each of the virtual MSMS spectra are calculated. The peptides are scored and the one having the highest score is deemed to be identified.
S. Barnes - J. Prasain 01/16/07
What has SEQUEST provided to proteomics?
• Initially, it seemed an awful lot! Typically, the “identified” proteins covered most of known biochemistry, so they satisfied everybody
• But the method obviously has limitations. There is redundancy - each protein yields multiple peptides
• The number of unique proteins was much less than the observed peptides
• Critically, it was missing controls
8
S. Barnes - J. Prasain 01/16/07
SEQUEST sequencing
• Use of SEQUEST requires considerable computing power - if there are 500 possible peptides to compare, then examination of 50,000+ spectra would require 25 million correlations
• Data analysis is typically carried out using computer clusters to accelerate the analysis
S. Barnes - J. Prasain 01/16/07
More haste, less speed?
• Post analysis, the masses of the peptides triggering MS-MS are used to create a set of virtual peptides with masses within + 1 Da
• Predicted MS-MS are compared to the observed and the best fit is reported as a hit
• The abundance of these hits are plotted in the figure as closed circles
• Post analysis, the masses of the peptides triggering MS-MS are used to create a set of virtual peptides with masses within + 1 Da
• Predicted MS-MS are compared to the observed and the best fit is reported as a hit
• The abundance of these hits are plotted in the figure as closed circles
However, if the sequences of the peptides within + 1 Da are reversed in silico and their predicted MS-MS compared to the observed spectra, a similar histogram is obtained (open circles), but without the right side tail
A forced fit to a set of data will always come up with a match, but not necessarily the truth
A forced fit to a set of data will always come up with a match, but not necessarily the truth
Resing et al. (2004)
9
S. Barnes - J. Prasain 01/16/07
Normal sequence
Reversed sequence
b ions 262 375 446 503 632 760 875 989 1088 1187 1343N F L A G E K D N V V R
y ions 1361 1247 1100 987 916 859 730 602 487 373 274 175
b ions 256 355 469 584 712 841 898 969 1082 1229 1343R V V N D K E G A L F N
y ions 1361 1205 1106 1007 893 778 650 521 464 393 280 133
So, b2corr = y2reverse - 18 and y2corr = b2reverse + 18
A reversed sequence has its problems as a control
S. Barnes - J. Prasain 01/16/07
How to improve MUDPIT• Reproducible column engineering
– Tandem columns, each built to separate, but high specifications
– Columns on a chip• More careful selection of the parent ion
– Accurate measurement of the peptide’s mass will eliminate many false peptides
– Accurate measurement of peptide fragments’masses
• Greater stringency in assessing score cutoff
10
S. Barnes - J. Prasain 01/16/07
Let’s take a closer look at fragmentation
b3 ion - oxazaloneno b1 ion
Wysocki et al. 2005
S. Barnes - J. Prasain 01/16/07
Other amino acid fragment ions
Wysocki et al. 2005
11
S. Barnes - J. Prasain 01/16/07
Detecting posttranslational modifications by MS
• A key issue is that the energy of ionization or the collisional process should not exceed the dissociational energy of the PTM
• MALDI-TOF MS with a N2 laser causes fragmentation of a nitrated tyrosine residue– Use ESI to make the molecular ion– Go to another wavelength
• O-glucosyl groups fragment more easily than the peptide to which they are attached– Use electron capture dissociation
S. Barnes - J. Prasain 01/16/07
Fragmentation of nitrated peptides in MALDI-TOF experiment
12
S. Barnes - J. Prasain 01/16/07
ESI-tandem MS of a nitrated peptide
S. Barnes - J. Prasain 01/16/07
m/z400 500 600 700 800 900 1000 1100 1200 1300 1400
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
[Abs. Int. * 1000000]c V S S A N M
514.26180c 6
613.33054c 7
903.44322c 8
990.47346c 9
1061.51275c 10
1175.55852c 11
1306.59534c 12
c8 fragment with sugar attached
[M+2H]2+
Sequencing O-GlcNAc peptides by ECD FT-ICR-MSCasein kinase II - AGGSTPVSSANMMSG
H3N CH
C
R1 O
NH
CH
R2
C NH
O
CH
COOH
R3
c ion cleavageb ion cleavage
13
S. Barnes - J. Prasain 01/16/07
Fragment ions of a small 5-mer peptide
Homework - write down the masses of the b and y ions
A Q Y E K
100 150 200 250 300 350 400 450 500 550 600 650
m/z
36
27
18
9
0
Rel
. Int
(%)
129
147
183
200
240
258
276363
403
439
492549
567
637
bn = [residue masses + 1]
yn = [residue masses + H2O + 1]
S. Barnes - J. Prasain 01/16/07
What is tandem mass spectrometry?
• Tandem mass spectrometry, abbreviated MS/MS, is any general method involving at least two stages of mass analyzers (one to pre-select an ion and the second to analyze fragments induced. This dual analysis can be dual in space, or dual in time. The most commonly used tandem mass spectrometry is the triple quadrupole.
Q1 Q2 Q3
14
S. Barnes - J. Prasain 01/16/07
Metabolomics is an emerging “omics” science that determines the physiological status of a sample or tissue by comparing the concentration of small molecules (biochemicals) in a tissue or sample with a similar measurement in a control sample. Small molecules are a diverse group of natural and synthetic substances that generally have a low molecular weight (f. w. 150-3000).
Metabolomics and small molecules
Genistein(a plant secondary metabolite)
Triglycerides Taxol
S. Barnes - J. Prasain 01/16/07
Nomenclature: the main names and acronyms used in MS/MS
• Molecular ion: Ion formed by addition or the removal of one or several electrons to or from the sample molecules-Electron Impact (EI-MS). M + e- → M+• + 2e-
• Adduct Ion: Ion formed through interaction of two species and containing all the atoms of one of them plus one or several atoms of them (e.g. alkali, ammonium).
15
S. Barnes - J. Prasain 01/16/07
Contd..
• Pseudomolecular ion: Ion originating from the analytemolecule by abstraction of a proton [M-H]- or addition of proton [M+H]+
• Precursor ion/parent ion: Ions undergoing fragmentation.
• Product ion/daughter ion: Ions resulting from parent/precursor ions.
• Neutral loss: Fragments lost as neutral molecules
• In positive ionization mode, a trace of formic acid is often added to aid protonation of the sample molecules; in negative ionization mode a trace of ammonia solution or a volatile amine is added to aid deprotonation of the sample molecules. Proteins and peptides are usually analysed under positive ionization conditions and polyphenols and acids under negative ionization conditions. In all cases, the m/z scale must be calibrated.
S. Barnes - J. Prasain 01/16/07
Ways to approach predicting MS/MS
• Likely sites of protonation or deprotonation.• Likely leaving group.• Mobility of protons• Literature study
Fragmentation always follows the basic rules of chemistry
O
OOCH3
HO
OH
Biochanin A
Where are the sites of Deprotonation/protonation?What is the most likely leaving Group in this molecule?
16
S. Barnes - J. Prasain 01/16/07
Applications of MS/MS
• Identification, characterization and authentificationof chemical components in a crude mixture
• Substructure analysis of unknown components
• Quantification of analytes in biological samples
S. Barnes - J. Prasain 01/16/07
Profiling taxoids metabolites in T.Wallichiana extract
Prasain et al. Anal Chem, 2001
17
S. Barnes - J. Prasain 01/16/07
ESI-MS/MS spectra of taxoids (1-3). Peaks m/z 194 and 210represent the intact alkaloid side chain.
S. Barnes - J. Prasain 01/16/07
MS/MS precursor-scan spectra of typical alkaloid side chains to identify the basic taxoids compounds in an ethyl acetate
extract of T. wallichiana.
18
S. Barnes - J. Prasain 01/16/07
600 620 640 660 680 700 720 740 760 780 800 820 840 860 880
605622
714
742
758 800
612 628 668 710
742
806.2 848 862
668684
726
744
758770 786
686 710
728
744770 786
800
A
B
C
D
Scaffold (m/z 309)
Side chain (m/z 210)
Side chain (m/z 252)
Side chain (m/z 194)
m/z
Inte
nsity
Comparison of precursor scan spectra obtained from thescaffold m/z 309 and side chain m/z 194, 210 and 252
Taxoids with scaffold m/z 309 and alkaloid side chains are shown by dashed lines
S. Barnes - J. Prasain 01/16/07
O
OH
O
O
O
OHOH
CH2OH
OH
O
OH
HO
O
OHOCH2OH
Puerarin; mol. wt. 416 Daidzin; mol. wt. 416
OHHO
MS/MS may help distinguish isomer structures
The Kudzu as a source of isoflavones
19
S. Barnes - J. Prasain 01/16/07
O
OH
HO
O
OHO
HO
CH2OHOH
O
OH
O
O
O
OHOH
CH2OH
OH
PuerarinM/z 415
O- and C-glucosides fragment differently in ESI-MS/MS
Prasain et al. J. Agric. Food Chem. 2003
220 280 300 320 360 380 m/z0
%
0
100
%
255.050
256.057
297.043
267.037
268.041281.051
307.065321.046
363.046335.061351.044 381.055
100
240 260 340
[A]
[B]
-162 Da
Yo+
-120 Da
Daidzin m/z 415
S. Barnes - J. Prasain 01/16/07
Possible product ions of puerarin in ESI-MS/MSin negative ion mode
O
O-
HO
O
OH
O
O-
HO
O
O
O-
HO
O
HO
HO
O
O-
HO
O
OHOCH 2OH
OHHO
m/z 415
m/z 325
-90 Dam/z 295
-120 Da
m/z 267
-28 Da
Prasain et al. J Agric Food Chem. 51, 4213, 2003.
20
S. Barnes - J. Prasain 01/16/07
m/z
100
50
0
50 100 150 200 250m/z
100
50
0
Rel
. Int
. (%
) 65 107
117
151
225
269
6391
107
133
159 180196
224
240
269O
OH
HO
OOH
Genistein
OOH
HO
OOH
Apigenin
O
O-
O-
149
Isomers like genistein and apigenin are readily separated by tandem mass spectrometry
S. Barnes - J. Prasain 01/16/07
M/z
100
50
0
50 100 150 200 250 300
100
50
0
63 107 139
167
183
195211 239
267
303
139
167
195
223251
287-36
-28
-28-28
-28
-36
-28
-28
O
OO_
HO
Cl
O
OO_
HO
Cl
OH
HCl loss (36 Da) is diagnostic for 3’-chloro derivative of genistein and daidzein in ESI-MS/MS
21
S. Barnes - J. Prasain 01/16/07
100
50
0
Rel
. Int
. (%
)
50 100 150 200 250m/z
100
50
0
Rel
. Int
. (%
)
123
152
170213 229242
269
285
118 152169 197
214225
242
270
285O
OOH
H3CO
HO
O
OOCH 3
HO
OH
H+
H+
Biochanin A
Glycitein
[A]
[B]
Biochanin A and glycitein can be distinguished byESI-MS/MS in positive ion mode
S. Barnes - J. Prasain 01/16/07
Intensity of product ions indicates their stability -ions bearing aromatic ring are more intense
O
OH
HO
OHO
O
OH
OH
OH
OH
OH44
22
S. Barnes - J. Prasain 01/16/07
Metal ion adducts are common in the positive ion mode
TIC
EMS (positive survey scan in trap)
EPI (Enhanced Product Ion Scan of m/z 579 in Trap)
Dimer [M+H]+Dimer [M+Na]+
Trimer
S. Barnes - J. Prasain 01/16/07
Comparison of product ions help elucidate the unknown structures
O
OOH
HO
O
HO
OH
OH
H2COH
+OH
150 200 250 300 350 400
100
%
255268
282
311
-120 431
Monohydroxylatedpuerarin
200 250 300 350 400
100
50
0
267 295
\
415
-120
Puerarin
23
S. Barnes - J. Prasain 01/16/07m/z
100
50
0
100 200 300 400m/z
100
50
0.0
Rel
. Int
. (%
)253
109
161
224
253
271
433
433
180 Da
162 Da
Tandem mass spectrometry can provide crucialinformation on structure of unknown glycosides
O
O-
HO
O
O
OHO
OH
HOCH2OH
O
O-
O
O
OH
OHO
HO
HO CH2OH
162 Da
Possiblestructure
Possiblestructure
S. Barnes - J. Prasain 01/16/07m/z
0
300 400 500
100
50
0
433
447449
463
259291
433
449
100
50
479
Ions obtained from neutral loss scan of 162 [A] and 132 [B] In the EtOAc extract of cranberry
fresh fruits
24
S. Barnes - J. Prasain 01/16/07
100 200 300 400m/z
100
50
0
Rel
ativ
e In
tens
ity (%
)
5985
113133 181 224
269
Glucuronide loss (176 Da)Da)
O
OHHO
OH
COOH
445
Product ion spectrum of genistein glucuronidein ESI-MS/MS
Glucosides/glucuronides conjugates are easily cleaved off by higher potential at orifice
S. Barnes - J. Prasain 01/16/07
Change in mass
Methylation 14Demethylation -14Hydroxylation 16Acetylation 42Epoxidation 16Desulfuration -32Decarboxylation -44Hydration 18Dehydration -18
Metabolic rxn
Change in mass is associated with possiblemetabolic reactions
25
S. Barnes - J. Prasain 01/16/07
Ionization mo Scan
Glucuronides pos/neg NL 176 amuHexose sugar pos/neg NL 162 amuPentose sugar pos/neg NL 132 amuPhenolic sulph pos NL 80 amuPhosphate neg Precursor of m/z 79 Aryl-GSH pos NL 275 amuAliphatic-GSH pos NL 129taurines ps Precursor of m/z 126N-acetylcysteneg NL 129 amu
NL = neutral loss.
Conjugat
Characteristic fragmentation of drug conjugates by MS/MS
S. Barnes - J. Prasain 01/16/07
References1. Electrospray Ionization Mass Spectrometry by Richard B.
Cole.
2. Stefanowicz P, Prasain JK, Yeboah KF, Konishi Y. Detection and partial structure elucidation of basic taxoids from Taxus wallichiana by electrospray ionization tandem mass spectrometry. Anal Chem. 2001;73:3583-9.
3. Prasain JK, Patel R, Kirk M, Wilson L, Botting N, Darley-Usmar VM, Barnes S. Mass spectrometric methods for the analysis of chlorinated and nitrated isoflavonoids: a novel class of biological metabolites. J Mass Spectrom. 2003;38:764-71.
26
S. Barnes - J. Prasain 01/16/07
Question-?What structural information do you get from the MS/MS Spectra shown below.?