Direct Experimental Observation of Functional Protein Isoforms by Tandem Mass Spectrometry

1
Direct Experimental Observation of Functional
Protein Isoforms by Tandem Mass Spectrometry

• Nathan Edwards
• Center for Bioinformatics and Computational
  Biology
• University of Maryland, College Park

2
Synopsis

• MS/MS spectra provide evidence for the amino-acid
  sequence of functional proteins.
• Key concepts
• Spectrum acquisition is unbiased
• Direct observation of amino-acid sequence
• Sensitive to small sequence variations

3
Synopsis

• MS/MS spectra provide evidence for the amino-acid
  sequence of functional proteins.
• Applications
• Cancer biomarkers
• Genome annotation

4
Mass Spectrometry for Proteomics

• Measure mass of many (bio)molecules
  simultaneously
• High bandwidth
• Mass is an intrinsic property of all
  (bio)molecules
• No prior knowledge required

5
Mass Spectrometer

• ElectronMultiplier(EM)

• Time-Of-Flight (TOF)
• Quadrapole
• Ion-Trap

• MALDI
• Electro-SprayIonization (ESI)

6
High Bandwidth
7
Mass is fundamental!
8
Mass Spectrometry for Proteomics

• Measure mass of many molecules simultaneously
• ...but not too many, abundance bias
• Mass is an intrinsic property of all
  (bio)molecules
• ...but need a reference to compare to

9
Mass Spectrometry for Proteomics

• Mass spectrometry has been around since the turn
  of the century...
• ...why is MS based Proteomics so new?
• Ionization methods
• MALDI, Electrospray
• Protein chemistry automation
• Chromatography, Gels, Computers
• Protein / genome sequences
• A reference for comparison

10
Sample Preparation for Peptide Identification
11
Single Stage MS
MS
m/z
12
Tandem Mass Spectrometry(MS/MS)
m/z
Precursor selection
m/z
13
Tandem Mass Spectrometry(MS/MS)
Precursor selection collision induced
dissociation (CID)
m/z
MS/MS
m/z
14
Peptide Identification

• For each (likely) peptide sequence
• 1. Compute fragment masses
• 2. Compare with spectrum
• 3. Retain those that match well
• Peptide sequences from (any) sequence database
• Swiss-Prot, IPI, NCBIs nr, ESTs, genomes, ...
• Automated, high-throughput peptide identification
  in complex mixtures

15
Peptide Identification

• ...can provide direct experimental evidence for
  the amino-acid sequence of functional proteins.
• Evidence for
• Functional protein isoforms
• Translation start and frame
• Proteins with short open-reading-frames

16
Why is this useful for ...... genome annotation?

• Evidence for SNPs and alternative splicing stops
  with transcription
• No genomic or transcript evidence for translation
  start-site.
• Conservation doesnt stop at coding bases!
• Statistical gene-finders struggle with
  micro-exons, translation start-site, and short
  ORFs.

17
Why is this useful for ...... cancer biomarkers?

• Alternative splicing is the norm!
• Only 20-25K human genes
• Each gene makes many proteins
• Some splicing is believed to be silencing
• Lots of splicing in cancer
• Proteins have clinical implications
• Statistical biomarker discovery
• Putative malfunctioning proteins

18
What can be observed?

• Known coding SNPs
• Novel coding mutations
• Alternative splicing isoforms
• Microexons ( non-cannonical splice-sites )
• Alternative translation start-sites ( codons )
• Alternative translation frames
• Dark open-reading-frames

19
Splice Isoform

• Human Jurkat leukemia cell-line
• Lipid-raft extraction protocol, targeting T cells
• von Haller, et al. MCP 2003.
• LIME1 gene
• LCK interacting transmembrane adaptor 1
• LCK gene
• Leukocyte-specific protein tyrosine kinase
• Proto-oncogene
• Chromosomal aberration involving LCK in
  leukemias.
• Multiple significant peptide identifications

20
Splice Isoform
21
Novel Splice Isoform
22
Novel Mutation

• HUPO Plasma Proteome Project
• Pooled samples from 10 male 10 female healthy
  Chinese subjects
• Plasma/EDTA sample protocol
• Li, et al. Proteomics 2005. (Lab 29)
• TTR gene
• Transthyretin (pre-albumin)
• Defects in TTR are a cause of amyloidosis.
• Familial amyloidotic polyneuropathy
• late-onset, dominant inheritance

23
Novel Mutation
Ala2?Pro associated with familial amyloid
polyneuropathy
24
Novel Mutation
25
Translation Start-Site

• Human erythroleukemia K562 cell-line
• Depth of coverage study
• Resing et al. Anal. Chem. 2004.
• THOC2 gene
• Part of the heteromultimeric THO/TREX complex.
• Initially believed to be a novel ORF
• RefSeq mRNA in Jun 2007, no RefSeq protein
• TrEMBL entry Feb 2005, no SwissProt entry
• Genbank mRNA in May 2002 (complete CDS)
• Plenty of EST support
• 100,000 bases upstream of other isoforms

26
Translation Start-Site
27
Translation Start-Site
28
Translation Start-Site
29
Translation Start-Site
30
Easily distinguish minor sequence variations

• Two B. anthracis Sterne a/ß SASP annotations
• RefSeq/Gb MVMARN... (7441 Da)
• CMR MARN... (7211 Da)
• Intact proteins differ by 230 Da
• 7441 Da vs 7211 Da
• N-terminal tryptic peptides
• MVMAR (606.3 Da), MVMARNR (876.4 Da), vs
• MARNR (646.3 Da)
• Very different MS/MS spectra

31
Bacterial Gene-Finding

• Find all the open-reading-frames...

TAGAAAAATGGCTCTTTAGATAAATTTCATGAAAAATATTGA
Stopcodon
Stopcodon
...courtesy of Art Delcher
32
Bacterial Gene-Finding

• Find all the open-reading-frames......but
  they overlap which ones are correct?

Reversestrand
Stopcodon
ATCTTTTTACCGAGAAATCTATTTAAAGTACTTTTTATAACT
TAGAAAAATGGCTCTTTAGATAAATTTCATGAAAAATATTGA
Stopcodon
Stopcodon
ShiftedStop
...courtesy of Art Delcher
33
Coding-Sequence Score
...courtesy of Art Delcher
34
Glimmer3 Performance

• Glimmer3 trained compared to RefSeq genes with
  annotated function
• Correct STOP
• 99.6
• Correct START
• 84.3
• Not all the genomes necessarily have
  carefully/accurately annotated start sites, so
  the results for number of correct starts may be
  suspect.

35
N-terminal peptides

• (Protein) N-terminal peptides establish
• start-site of known unexpected ORFs
• Use
• Directly to annotate genomes
• Evaluate and improve algorithms
• Map cross-species

36
N-terminal peptide workflows

• Typical proteomics workflows sample peptides from
  the proteome randomly
• Caulobacter crescentus (70)
• 3733 Proteins (RefSeq Genome annot.)
• 66K tryptic peptides (600 Da to 3000 Da)
• 2085 N-terminal tryptic peptides (3)

37
N-terminal peptide workflow

• Protect protein N-terminus
• Digest to peptides
• Chemically modify free peptide N-term
• Use chem. mod. to capture unwanted peptides

Nat Biotech, Vol. 21, pp. 566-569, 2003.
38
Increasing N-terminal peptide coverage

• Multiple (digest) enzymes
• trypsin-R 60 (80)
• acid lys-C trypsin 85 (94)
• Repeated LC-MS/MS
• Precursor Exclusion / Inclusion lists
• MALDI / ESI
• Protein separation and/or orthogonal
  fractionation

Anal Chem, Vol. 76, pp. 4193-4201, 2004.
39
Proteomics Informatics

• Search spectra against
• Entire bacterial genome
• All Met initiated peptides or
• Statistically likely Met initiated peptides.
• Easily consider initial Met loss PTM, too
• Off-the-shelf MS/MS search engines (Mascot /
  X!Tandem / OMSSA)

40
Other Practical Issues

• Suitable for commonly available instrumentation
• Only the sample prep. is (somewhat) novel.
• Need living organism
• Stage of life-cycle?
• Bang for buck?
• N-terminal peptides /
• In discussions with JCVI (ex TIGR)
• Possible pilot project?

41
Other Research Projects

• Improving peptide identification by MS/MS
• Spectral matching using HMMs
• Combining search engine results
• Spectral matching for detection and quantitation
• Microorganism identification using MS
• Live public web-site and database
• (Inexact) uniqueness guarantees
• Primer/Probe oligo design
• Pathogen detection (DNA Peptide)
• Significant false-positive peptide identifications

42
Spectral Matching

• Detection vs. identification
• Increased sensitivity
• No novel peptides
• NIST GC/MS Spectral Library
• Identifies small molecules,
• 100,000s of (consensus) spectra
• Bundled/Sold with many instruments
• Dot-product spectral comparison
• Current project Peptide MS/MS

43
Peptide DLATVYVDVLK
44
Peptide DLATVYVDVLK
45
Hidden Markov Models for Spectral Matching

• Capture statistical variation and consensus in
  peak intensity
• Capture semantics of peaks
• Extrapolate model to other peptides
• Good specificity with superior sensitivity for
  peptide detection
• Assign 1000s of additional spectra (w/ p-value lt
  10-5)

46
www.RMIDb.org
47
www.RMIDb.org

• Statistics
• 16.7 x 106 (6.4 x 106) protein sequences
• 40,000 organisms, 19,700 species
• 557 (415) complete genomes
• Sources
• TIGRs CMR, SwissProt, TrEMBL, Genbank Proteins,
  RefSeq Proteins Genomes
• Inclusive Glimmer3 predictions on Genomes
• Pfam and GO assignments using BOINC grid

48
www.RMIDb.org
Accessed from all over the world...
49
Uniqueness guarantees

• 20-mer oligo signatures for B. anthracis
• In all available strains as exact match
• No (inexact) match to other Bacillus species

50
Uniqueness guarantees

• Human genome primer design problem
• 4-unique DNA 20-mers
• Edit-distance 5 to any non-specific
  hybridization site
• No such valid loci on Chr. 22!
• Currently analyzing entire genome
• 3-unique DNA 20-mers
• Initial experiments suggest 0.01 valid
• Approx. 1 valid oligo every 10,000 bases

51
Future Research Plans

• Cancer biomarkers
• Optimize proteomics workflow for protein sequence
  coverage
• Improve informatics infrastructure to make
  interpretation easier
• Identify splice variants in cancer cell-lines
  (MCF-7) and clinical brain tumor samples

52
Future Research Plans

• Genome Annotation
• Collect evidence for functional alternative
  splicing in public datasets into dbPEP.
• Conduct pilot project for bacterial genome
  annotation with JCVI.
• Improve informatics infrastructure to make
  interpretation easier.

53
Future Research Plans

• Peptide Identification
• Expand library of HMM models for high-confidence
  spectral matching
• Spectral matching for biomarkers and quantitation
  (with Calibrant).
• Specificity metric for peptides identified using
  MS/MS

54
Future Research Plans

• Microorganism identification by mass
  spectrometry
• Specificity of tandem mass spectra
• Revamp RMIDb prototype
• Incorporate spectral matching, top-down.

55
Future Research Plans

• Oligonucleotide Design
• Uniqueness oracle for inexact match in human
• Integration with Primer3
• Tiling, multiplexing, pooling, tag arrays

56
Acknowledgements

• Catherine Fenselau, Steve Swatkoski
• UMCP Biochemistry
• Chau-Wen Tseng, Xue Wu
• UMCP Computer Science
• Cheng Lee, Brian Balgley
• Calibrant Biosystems
• PeptideAtlas, HUPO PPP, X!Tandem
• Funding NIH/NCI, USDA/ARS