Jamie Boehmer

1
Quantitative Mass Spectrometric Multiple Reaction
Monitoring Assays for Major Plasma Proteins.

• Leigh Anderson and Christie L. Hunter
• The Plasma Proteome Institute, Washington DC
• Applied Biosystems, California

2
The Plasma Proteome

• Growing body of evidence that plasma proteins are
  potential candidates for clinical biomarkers.
• Quantitation of serum and plasma proteins very
  difficult due to complexity and extreme dynamic
  range.
• Survey approach decreased sensitivity for the
  detection of low abundance proteins.
• Candidate-based approach- increased sensitivity
  but only for the detection of selected proteins
  hinders discovery of novel proteins.

3
Selected Reaction Monitoring

• SRM - Mass spectrometry approach to the
  quantitative detection of select proteins and
  peptides in a complex biological sample.
• Intact analyte selected by MS1 and specific
  fragments of analyte selected in MS2 following
  collision induced dissociation.
• MS approach provides structural specificity for
  selected analytes, is very sensitive, and also
  allows for quantitation with the use of stable
  isotope-labeled internal standards.
• Currently how we screen tissue, fluid, and feed
  samples for drug metabolites.
• Can be multiplexed to perform multiple assays in
  one run.

4
Multiple Reaction Monitoring

• Multiplexed SRM used to identify specific
  peptides in complex tryptic digests of plasma.
• One peptide chosen as a representative of the
  cleaved protein and quantitated against a stable
  isotope standard.
• Hybrid methods developed SISCAPA Stable
  Isotope Standards with Capture by Anti-Peptide
  Antibodies.
• Couples MRM assay with enrichment strategy
• Immobilized anti-peptide antibodies co-enrich for
  specific peptides and spiked stable-isotope-labele
  d (SIS) internal standards of the same sequence.
   

5
Specific Aims

• How many plasma proteins can be measured in a
  tryptic plasma digest by quantitating their
  peptides?
• How precise can these measurements be?
• Can this type of measurement strategy be adapted
  for existing high throughput LC-MS/MS platforms?

6
The Methods Generating MRM Criteria

• In Silico methods 177 candidate markers of
  cardiovascular disease previously assembled from
  literature by author.
• 62 proteins out of the 177 were selected based on
  availability of normal plasma concentration data.
• Predicted tryptic digests for each protein were
  generated.
• Determined likelihood of experimental detection
  based on previously reported data
• hits for given peptide/ hits for most
  frequently detected peptide from same protein.
• From subset of 62 proteins, 30 proteins were
  selected.

7
Stable Isotope Standards

• Stable Isotope Standards (SIS) produced by
  transcription and translation of a synthetic gene
  coding for 30 concatenated tryptic peptide
  sequences in the presence of U-13C6, U-15N2-
  labeled lysine (produces an 8 amu mass shift).
• C-terminus fragments (y-ions) will carry label
  and N-terminus fragments (b-ions) will have same
  mass as natural peptide.

8
In Silico Results

• All 30 detected by MRM in tryptic digest of SIS
  peptides.
• 19 detected in SIS spiked digest of whole human
  plasma.

9
The Methods Generating MRM Criteria

• Two Direct Proteomic Survey Experiments.
• Carried out LC-MS/MS analysis of plasma digests.
• Major ions observed subjected to MS/MS using
  ion-trap (Instrument 4000 Q TRAP).
• Peptides with best signal intensity and
  chromatographic peak for parents proteins were
  selected.
• Used the Global Proteome Machine database.
• Peptides were selected from target proteins
  frequently detected by database users.
• Identified 54 plasma proteins ranging from
  albumin to fibronectin.

10
The Methods Generating MRM Criteria

• Adapted MRM-initiated Detection and Sequencing
  (MIDAS) approach.
• Selected protein sequences were digested in
  silico
• y-ions predicted and theoretical MRMs generated
  for all the peptides in a given size range.
• The theoretical MRMs used in a survey scan of
  plasma to detect specific peptide peaks.
• When the specified MRM peak was detected, it was
  then subjected to full scan MS/MS to obtain
  sequence information for the hypothesized
  peptide.
• Peptide specificity was then verified by BLASTP
  exact match searches.
• 12 proteins examined, and 9 produced at least one
  good MRM.

11
Unwin et al, 2005
12
Selecting the MRMs

• Applied all the approaches and pooled all the
  results.
• A set of optimized MRMs was assembled that
  included at total of 60 peptides that represented
  53 high to medium abundance human plasma
  proteins.
• For all but one, 2 fragments were measured per
  peptide, yielding 119 MRMs.
• MRMs for 18 stable isotope-labeled internal
  standard versions of the target peptides were
  included making a total of 137 MRMs.

13
Methods

• Plasma depleted using Multiple Affinity Removal
  System (MARS) to remove 6 most abundant plasma
  proteins (albumin, IgG, IgA, transferrin,
  haptoglobin, and anti-trypsin).
• Both depleted and undepleted plasma was digested
  with trypsin.

14
Methods

• Plasma digests with and without added SIS
  peptides were analyzed by electrospray LC-MS/MS.
• Instrument 4000 Q TRAP hybrid triple
  quadrupole/linear ion trap with a NanoSpray
  source.

B, E high peptide loading A, C, and D low peptide
loading 110ng of total peptides
15
The MRMs
16
Full Scan MS/MS - Fibronectin
17
Manual Inspection to Determine Match to Specific
Peptide
18
Results

• For depleted sampled, greater than 60 of the
  MRMs show within-run critical values less than
  10.
• Almost 50 of the CVs in the depleted samples
  were below 5.
• MRMs for apolipoportein E, heparin cofactor II,
  plasminogen, prothrombin and others had CVs of
  3-4 which is equivalent to good clinical
  immunoassays.
• Analyses of undepleted plasma samples showed
  higher CVs.
• 20-50 of MRMs had CVs greater than 10.

19
Effects of Plasma Depletion
fibrinogen
a2-macroglobulin
complement C3
low abundance proteins
20
Results

• Substantial reductions in albumin, transferrin,
  and haptoglobin confirmed, but a1-antitrypsin not
  detected at all. IgG and IgA not measured.
• Ratios between natural peptides and and SIS
  peptides were similar in all experiments and
  maintained a linear relationship.
• L-selectin and fibronectin, two low abundance
  plasma proteins, were detected among the MRMs
  tested.
• Report that L-selectin was measured at 0.67mg/ml
  and that this concentration reflects normal
  plasma concentrations, but never measured
  concentration in whole plasma samples.

21
Results

• Six of the 53 target proteins were not reliably
  observed.
• Due to poor peptide selection or low abundance.
• Of the 47 target proteins that were reliably
  observed 78 had within-run CVs of 10 or less.
• Depletion and tryptic digests were highly
  reproducible.

22
Conclusions

• Peptide MRM measurements in plasma digests could
  provide a high throughput specific assay for
  biomarker validation.
• MRM assays could be extended to low abundance
  plasma proteins if coupled with enrichment
  techniques such as SISCAPA.
• Existing high throughput LC-MS/MS platforms can
  be used to identify plasma proteins using MRM
  assays.

23
(No Transcript)