Transcript Lecture Slides

Mass spectrometry-based
proteomics
Jeff Johnson
Feb 19, 2014
MS Proteomics in a Nutshell
• Ionization
– Delivering macromolecules to the MS
• Ion Manipulation
– Trapping and ejecting analytes of interest
• Fragmentation
– Breaking apart for more information
• Mass analysis and detection
– Measuring masses and quantifying intensities
MS Proteomics in a Nutshell
• Ionization
– Delivering macromolecules to the MS
• Ion Manipulation
– Trapping and ejecting analytes of interest
• Fragmentation
– Breaking apart for more information
• Mass analysis and detection
– Measuring masses and quantifying intensities
Macromolecular Ionization for MS
• Analyte must be in the gas phase for mass
analysis
• Analyte must be charged in order to be
manipulated by electric and magnetic fields
– All mass analyzers measure mass-to-charge ratios
(m/z)
• Two predominant approaches (shared the Nobel
prize in 2002)
– Matrix assisted laser desorption ionization
– Electrospray ionization
MALDI Ionization
• Sample is spotted in a matrix that
readily absorbs UV/IR light and is
vaporized by a laser
– Common matrix: 2,5dihydroxybenzoic acid (DHB)
• Advantages
– Fast and easy
– Spots can be reanalyzed later
– Most analytes get one +ive charge
makes it easy to deconvolute
• Disadvantages
– Harsh. Often breaks analytes apart
(e.g., breaks phosphorylation)
– Not easily combined with online
HPLC separations
Electrospray Ionization
•
Sample is dissolved in liquid and pushed
through a charged needle and sprayed into
an evaporation chamber
– Droplets pulled into the MS source by electric
potential between the needle and the MS
– Heated ion transfer tube evaporates water
molecules in droplets leaving +ively charged
analytes in the gas phase
•
Advantages
– Compatible with online HPLC separations
– “Soft” ionization maintains label and noncovalent interactions
•
Disadvantages
– Analytes can have different numbers of
charges, can be difficult to deconvolute
without high mass accuracy
– Different samples going through the same
electrospray tip causes carryover problems
•
Especially bad with online HPLCs
Ionization is Nearly Impossible to Predict
• Different molecules
ionize with different
efficiencies and are
very difficult to predict
• MS intensity ratios
between different
molecules do not
reflect ratios in the
sample from which
they were derived
• Most quantification by
MS is relative
2x
A
A
X
B
B
Ionization is Nearly Impossible to Predict
Sample 1
Sample 2
A
A
Sample 1
Sample 2
A
A
• Different molecules
ionize with different
efficiencies and are
very difficult to predict
• MS intensity ratios
between different
molecules do not
reflect ratios in the
sample from which
they were derived
• Most quantification by
MS is relative
* Assumption: MS run 1 = MS run 2
MS Proteomics in a Nutshell
• Ionization
– Delivering macromolecules to the MS
• Ion Manipulation
– Trapping and ejecting analytes of interest
• Fragmentation
– Breaking apart for more information
• Mass analysis and detection
– Measuring masses and quantifying intensities
Ion Manipulation
• We need a way to select only ions of interest
– Most detectors are just electron multipliers that don’t
measure mass but just detect a thing hitting the
multiplier
– We can manipulate ions to deliver defined mass
ranges to the detector to get a mass spectrum
• Two common tools:
– Ion traps
– Quadrupoles
– Both use electric and magnetic fields to select ions of
a particular m/z range
Ion Trap
• Ions are trapped by 3D electric field by DC and AC applied to
the electrodes
• An ion trap can accumulate ions as they come in from the
source and store them
• Low resolution: +/- 1 Da
Quadrupole
• Can be thought as a mass filter
• DC and AC fields applied that stabilize a trajectory for ions in
a desired mass range, undesired ions are ejected
• Quadrupole operate with a continuous flow of ions
• Low resolution (+/- 1 Da)
MS Proteomics in a Nutshell
• Ionization
– Delivering macromolecules to the MS
• Ion Manipulation
– Trapping and ejecting analytes of interest
• Fragmentation
– Breaking apart for more information
• Mass analysis and detection
– Measuring masses and quantifying intensities
Fragmentation
• Usually measuring the mass of an analyte is not
enough to conclusively identify it
• By fragmenting an analyte and measuring the
masses of the fragments we can obtain further
information to identify the analyte
• There are many types of fragmentation but
collision-induced dissociation (CID) is the most
common
– Fastest and most generally successful for the widest
variety of proteins and peptides
Collision-Induced Dissociation
• Give ions kinetic energy and collide with gas molecules (He)
• Collisions build up potential energy until a fragmentation
event can occur
• Ideally potential energy is strong enough to break a single
peptide bond but not strong enough to fragment further
• Can be done in an ion trap or a quadrupole
Collision Induced Dissociation
A
E
P
T
I
R
H2 O
Fragment (somewhat) randomly along the peptide backbone
Intensity
B-type Ions
A
E
P
A
E
P
A
E
A
201.1
T
298.1
399.2
72.0
M/z
Y-type Ions
R
I
T
Intensity
H2 O
M/z
P
E
A
B-type, A-type, Y-type Ions
R
I
T
Intensity
H2 O
M/z
P
E
A
MS Proteomics in a Nutshell
• Ionization
– Delivering macromolecules to the MS
• Ion Manipulation
– Trapping and ejecting analytes of interest
• Fragmentation
– Breaking apart for more information
• Mass analysis and detection
– Measuring masses and quantifying intensities
Mass Analysis and Detection
• All mass analyzers
achieve the same
thing: physical
separation based
on mass:charge
• Magnetic sector is
the simplest and
one of the earliest
types
Magnetic Sector MS
FT-ICR MS
• FT-ICR = Fourier transform – ion cyclotron resonance
• Ion injected in line with a strong magnetic field that
induces a cyclical motion
• Radius of the cyclotron motion is proportional to m/z
Time-of-flight MS
Medium / High Resolution
Quadrupole and Ion Trap MS
Electron multiplier
• You can use a quadrupoles or ion traps to “scan out” ions
across an entire mass range to a detector by gradually
ramping voltages
• Low resolution but electron multipliers make these very
sensitive
Orbitrap MS
• Characteristic frequencies:
– Frequency of rotation ωφ
– Frequency of radial oscillations ωr
– Frequency of axial oscillations ωz
r
 
z
z
2
 Rm 

 1
2  R 
2
r   z
 Rm 

 2
 R 
k
z 
m/ z
φ
U (r , z ) 


k 2
 z  r 2 / 2  Rm2  ln( r / Rm )
2
Power of Fourier Transforms
• FTs convert from
time domain to freq
domain
• Instead of a single
measurement the
m/z is measured
over a period of time
and the FT
essentially averages
all those
measurements
• Resulting data is very
high resolution
Chromatography to Simplify Complexity
Complex
Sample
MS
• Complexity hurts sensitivity
• A constant, defined number of ions can be analyzed in
each MS scan
• Sensitivity is constant (around 1 fmol)
• A scan with fewer ions is more sensitive than a scan with
many
Chromatography to Simplify Complexity
HPLC
Complex Sample
MS
C18 RP column
ACN gradient
A
C
B
A
B
D
C
D
Chromatography to Simplify Complexity
Very
Complex
Sample
Offline HPLC
(e.g., SCX)
SCX Fractions
Injected individually
Online HPLC (RP)
SCX Fractions
MS
Acquiring MS Data
• Data acquisition depends on experimental
goals
– Data-dependent acquisition
• MS attempts to acquire data to allow you to identify a
maximum number of unknowns
• Commonly used for analyses where you don’t know
what you’re looking for
– Targeted acquisition
• MS only acquires data for what you tell it to acquire
• Much more sensitive than data-dependent, but also
more limited in scope
Data-Dependent Acquisition
Data-Dependent Acquisition
1
3
2
High resolution survey scan
(<5 ppm mass accuracy)
Data-Dependent Acquisition
Low resolution MS/MS scan 1
Data-Dependent Acquisition
Low resolution MS/MS scan 2
Data-Dependent Acquisition
Low resolution MS/MS scan 3
Peptide Identification
AA sequence DB
(Species UniProt)
Peptide Identification
AA sequence DB
(Species UniProt)
1
2
3
AA DBs
restricted
by parent
ion mass
measured
in survey
scan
Peptide Identification
AA sequence DB
(Species UniProt)
MS/MS 1
1
MS/MS 2
2
MS/MS 3
3
AA DBs
restricted
by parent
ion mass
measured
in survey
scan
# of Matches
Probabilistic Matching (X!Tandem)
Second
Best
Best Hit
by-Score= Sum of intensities of peaks matching
B-type or Y-type ions
Hyper Score
HyperScore=
( by-Score) × N ! × N !
y
b
Model spectrum comparisons
Pattern Matching (Sequest)
Sequest XCorr
Correlation Score
Cross Correlation
(direct comparison)
Auto Correlation
(background)
Offset (AMU)
XCorr =
CrossCorr
avg AutoCorr offset=-75 to 75
(
)
Targeted Acquisition with a QQQ
SRM Assay
Peptide
Fragment
Q1
Q3
RT
Rel Int
YGFIEGHVVIPR
y8_1
693.88
906.52
22.9
100
YGFIEGHVVIPR
y7_1
693.88
777.47
22.9
91
YGFIEGHVVIPR
b3_1
693.88
368.16
22.9
77
YGFIEGHVVIPR
y4_1
693.88
484.33
22.9
53
YGFIEGHVVIPR
y3_1
693.88
385.26
22.9
42
A priori knowledge required:
SRM assay development for a list of proteins/peptides of interest
 Information derived from label-free unbiased proteomic
analysis
“Sensitivity”
• Sensitivity of a MS is well defined, but the ability
to identify something is a very different concept
– Ability to detect depends on:
• Sample complexity
• MS sensitivity
• MS speed
– A faster MS can collect go deeper in each survey scan
– Think “top 10” vs. “top 50”
• MS mass accuracy
– Better mass accuracy improves the ability to identify peptides but
sacrifices speed and MS sensitivity
– Especially important for variable modifications
– The “best” method is very dependent on the
experimental goals
Database Searching
Ion trap
+/- 1 Da
Orbitrap
+/- 0.002 Da
Database
“search space”
Database Searching
Ion trap
+/- 1 Da
Database
“search space”
+S/T/Y phosphorylation
Orbitrap
+/- 0.002 Da
Database Searching
Ion trap
+/- 1 Da
Database
“search space”
+S/T/Y phosphorylation
Orbitrap
+/- 0.002 Da
Protein Quantification
• A mass spectrometer is an inherently quantitative
device but the ionization source is not
– Different peptides/proteins are ionized with drastically
different efficiencies
– Absolute abundances in a mass spectrometer are not
precisely indicative of abundance in a sample
• Solution: stable isotope labeling
– Compare samples that have been labeled with stable
isotopes (13C, 15N, 2H)
– ‘Heavy’ isotopes behave chemically identically to their
‘light’ counterparts but are separated in the MS
Isotope Coded Affinity Tag (ICAT)
Stable Isotope Labeling of Amino Acids
in Culture (SILAC)
• Grow cells in media
supplemented with stable
isotope-labeled amino
acids
• Combine samples at the
level of cells and process
as one sample
• Minimize variability
between samples for
lysis and digestion
• Different samples
separated by mass in the
MS
Absolute Quantification (AQUA)
Absolute Quantification (AQUA)