W.M. Keck Foundation Biotechnology Resource Laboratories. Erol Gulcicek Christopher Colangelo Terence Wu

Transcription

1 W.M. Keck Foundation Biotechnology Resource Laboratories Erol Gulcicek Christopher Colangelo Terence Wu

2 - Site is currently up with basic information - Overview of profiling technologies has recently been added since the url was first distributed - Ideas for future features are very welcome

3 Yale/NIDA Neuroproteomics Center Fact Sheet Title of Center Grant: Theme of Center: Principal Investigator: Co-Principal Investigator: #Grants Awarded: Total Funding Awarded to Yale: Yale/NIDA Neuroproteomics Research Center "Proteomics of Altered Signaling in Addiction" Kenneth R. Williams, Ph.D. Director, W.M. Keck Foundation Biotechnology Resource Laboratory Professor (Adj,) Research Dept. of Molecular Biophysics & Biochemistry Yale University School of Medicine Angus C. Nairn, Ph.D. Professor, Department of Psychiatry Yale University School of Medicine 2 Nationally $7.7 million Total Budget Period: 9/23/2004 5/31/2009 Center Cores: Center Investigators: Administrative, Bioinformatics and Biostatistics, Protein Identification, Protein Microarray; and Protein and Lipid Separation and Profiling 14 from Yale University, The Veterans Administration Medical Center (West Haven) Rockefeller University

4 Overall Goal: To substantially increase understanding of the biochemical mechanisms that underlie substance abuse and its treatment. Specific Aims: Bring together Yale faculty working at the forefront of such key neuroscience areas as Signal transduction, Plasticity, Neuronal morphogenesis, Lipid metabolism in neuronal signaling and Synaptic function Synaptic response to psychotropic drugs with experts in Proteomics, Biostatistics, Bioinformatics.

5 Proteomics of Altered Signaling in Addiction Receptor-Mediated Signaling Transcriptional Regulation Endocytosis and Exocytosis Regulation of the Cytoskeleton DIGE Neuronal Development Phosphoproteome Profiling ICAT Neuronal Plasticity 2D-LC Beckman Coulter System Cell Death Lipid Profiling MudPIT Translational Research Identification and Characterization of Signaling Proteins Whose Expression is Altered by Drug Addiction Serum Biomarkers Targeted Proteomics Clinical Studies of Drug Addiction

6 Yale/NIDA Neuroproteomics Program Initial Stage: - Meet individual PI s to: - Better understand the PI s research area - Jointly determine best to implement proteomics technologies available - Timelines from the Cores. - Review of current technologies - Development of new technologies - Some already know Core personnel and have already initiated studies in which case please continue working with these staff and additionally, I would be very glad to talk with you also if you have questions about other technologies. Erol E. Gulcicek erol.gulcicek@yale.edu

7 Commonly Used Protein Profiling Workflows in Mass Spectrometry Primary Biological Samples Organ, Tissue, cell extracts, biological fluids, etc. Proteins Peptides LC, MS and Tandem-MS Protein Sample Preparation Sub cellular components, fractionation, affinity enrichment, Protein complexes, etc. Protein Labeling Chemistries ICAT, PhIAT, SILAC, Protein Sample Simplification (or Purification): 18 O proteolysis, etc. 1D-SDS 2D-SDS Protein LC DIGE IEF PAGE PAGE (Q, S, RP, SEX, etc. FFE Digestion Digestion Digestion Digestion Enrichment step and/or Peptide Labeling Chemistries (SCX, IMAC, itraq TM etc.) (Non LC) MS/MS Top Down MS or MudPit MS/MS RP LC MS/MS Peptide Mass Fingerprint Protein MS/MS HT Protein ID & Protein sequence ID, (PMF) sequence ID, Protein Sequencing, Quantitation PTM and Quantitation Protein ID PTM and PTM SELDI or Predominantly Predominantly MALDI Predominantly Predominantly ESI QTOF& IT ESI QTOF & IT (3D and 2D), MALDI TOF/TOF MS. TOF MS MALDI TOF MS ESI FTMS (3D and 2D) Also FTMS and Triple-Q (incl. Quad. Trap) Biomarker patterns

8 Protein Profiling Overview Christopher M. Colangelo, PhD Director of Protein Profiling 300 George Street Room G006 (203) (W)

9 Brief Overview of the Keck Laboratory Founded in 1980: 40 Full time staff 100 Major instrument systems purchased at a cost of >$11 million dollars 25,000 ft 2 of space 12 individual Resources that provide a wide range of proteomic and genomic biotechnologies Completes 270,000 protein and nucleic acid syntheses and analyses annually for 260 Yale and 640 non-yale investigators at 300 institutions & companies in 27 countries NHLBI Proteomics Center (2002) NIH High End Instrumentation Grant (2002): FTICR-MS National Biodefense Center of Excellence: Proteomics Core (2003) NIH High End Instrumentation Grant: Center for High Performance Computation In Biomedicine (2004) NIDA Neuroproteomics Center (2004) (

10 Disease Based Proteomics Receptor-Mediated Signaling DIGE Phosphoproteome Profiling Transcriptional Regulation ICAT Endocytosis and Exocytosis 2D-LC Beckman Coulter System Lipid Profiling Regulation of the Cytoskeleton MudPIT Translational Research Identification and Characterization of Signaling Proteins Whose Expression is Up/Down Regulated Serum Biomarkers Targeted Proteomics Clinical Studies of Biomarker Efficacy

11 Keck MALDI-MS Based Disease Biomarker Discovery Sera from >50 normal patients Sera from >50 disease patients Automatically desalt each sample (submitted in 96 well plate format) using a Micromass MassPrep robot and C-18 Zip-Tips. Automated MALDI-MS on Micromass MALDI-L/R instrument: 130,000 m/z vs intensity datapoints acquired from 700 to 28,000 Da. Linearly normalize baseline-corrected intensities in each spectrum to the overall median for all normal + disease spectra in the training set. Filter out noise and all datapoints which fail to pass peak and unique peptde ion criteria. Use a Random Forest Algorithm customized by Keck Biostatistics Resource (Wu et, al 2003) to identify peptide markers whose intensities can be used to maximally discriminate all normal from disease samples. Optimize marker selection by increasing the size of the training set until the ability to correctly classify a testing set is maximized

12 Median MALDI-MS Spectra for Sera from 47 Ovarian Cancer (top) vs 42 Control (middle) Patients and the Resulting Difference Spectrum (bottom)

13 m/z Versus Intensity Distribution for 129 Samples Around the 2084 Ovarian Cancer Marker (relative importance = 2.9) Marker Position Median Cancer Median Control

14 Impact of Training Set Size & Number of Biomarkers on Classification of 34 "Unknown" Ovarian Cancer Sera #Biomarkers = 25 Training Set Size: 102 Sera 80% 80% Correctly Classified Test Sera 75% 70% 65% 60% 55% %Correctly Classified Test Sera 75% 70% 65% 50% % #Samples in Training Set Number of Biomarkers

15 Genomics The blueprint of life DNA Central Dogma PCR Mullis (1985) Human Genome Sequence Watson and Crick (1953) Nature 171, Crick (1970) Nature 227, Crick (1958) Symp. Soc. Exp. Biol. 138 The International Human Genome Mapping Consortium (2001) Nature 409, Venter et. al. (2001) Science Feb :

16 What is proteomics? Genomics provides us with the words, but doesn t provide us with any meaning Proteomics - provides us with the definitions for the genome The identification, characterization and quantification of all proteins involved in a particular pathway, organelle, cell, tissue, organ or organism that can be studied in concert to provide accurate and comprehensive data about that system. Protein Profiling provide us with the grammar and syntax to form a meaningful language Understanding how proteins interact with each other, their environment, and function with outside molecules Challenges - proteomics and protein profiling are dynamic languages

17 How do we bridge the gap from genomics to proteomics? Genomics Proteomics

18 Gene Microarrays First step towards bridging the gap between human genome and function Keck Human 4.6K cdna Array Cluster Analysis

19 mrna vs. Protein correlation (R=0.66) Comparing protein abundance and mrna expression levels on a genomic scale Greenbaum D, Colangelo C, Williams K, Gerstein M Genome Biol. 2003; 4(9): 117. Epub 2003 Aug 29.

20 Reality check for proteomics Avogadro s number - 6 x molecules/mole For MS proteomics we want 100 fmoles (10-14 moles) = 6.02 x molecules cell copy number cells needed for 100 fmoles 10 6 billion (6.02 x 10 9 ) million (6.02 x 10 8 ) million (6.02 x 10 7 ) 10,000 6 million (6.02 x 10 6 ) 100, million (6.02 x 10 5 ) 10cm culture dish = 10 5 cells 100,000 copy number limit 1g of tissue = 10 8 cells 1000 copy number limit

21 Levels of proteins in plasma Molecular & Cellular Proteomics 2003, Anderson and Anderson 2 (1): 50

22 Protein Profiling

23 Disease Based Proteomics Receptor-Mediated Signaling DIGE Phosphoproteome Profiling Transcriptional Regulation ICAT Endocytosis and Exocytosis 2D-LC Beckman Coulter System Lipid Profiling Regulation of the Cytoskeleton MudPIT Translational Research Identification and Characterization of Signaling Proteins Whose Expression is Up/Down Regulated Serum Biomarkers Targeted Proteomics Clinical Studies of Biomarker Efficacy

24 Cleavable ICAT Reagents Allows selection and concentration of cysteinecontaining peptides Removes biotin which improves MS/MS fragmentation Provide 9 Da which provides ratio of heavy/light peptide conc. Reacts with free cysteines Applied Biosystems

25 Differential expression analysis on human leukemic precursor cells vs. differentiated monocytes/neutrophils 16 SCX fractions collected 100ug of protein is labeled with acid cleavable ICAT TM reagent/trypsinized Cys-containing peptides affinity purified/ cleaved with TFA Online LC-MS/MS analysis on QStar XL System

26 ICAT-Determined Expression Ratio of Zinc Finger Protein 9 in Two Human Cell Lines Overall Average Expression Ratio of Protein 9 in Two Human Cell Lines: 3.4 fold decrease (average H/L = )

27 Web-based query results from Yale Protein Expression Database (YPED)

28 Disease Based Proteomics Receptor-Mediated Signaling DIGE Phosphoproteome Profiling Transcriptional Regulation ICAT Endocytosis and Exocytosis 2D-LC Beckman Coulter System Lipid Profiling Regulation of the Cytoskeleton MudPIT Translational Research Identification and Characterization of Signaling Proteins Whose Expression is Up/Down Regulated Serum Biomarkers Targeted Proteomics Clinical Studies of Biomarker Efficacy

29 Automated 2D Comparative LC Protein Profiling (Beckman-Coulter PF2D Platform) Partial pi/uv map of a colon cancer cell line before and after drug treatment. The RP-HPLC profiles illustrate differences in the pi fraction which are shown also by the color coded band depictions of these RP-HPLC profiles (from Beckman-Coulter).

30 Two-dimensional Comparison Plot of NB4 (Green) and NB4 + RA (Red) (Neutrophil) from Beckman PF2D System 5 mg of total protein Retention Time ph The 2D overlay plot shows the relative differential expression for each of the 21 ph fractions based on overlay comparisons of 214 nm reversed phase chromatograms Fraction

31 Overlay of NB4 and NB4+RA Reversed Phase Chromatograms from Fraction 7 The differential peaks in the UV chromatogram represent intact proteins that are expressed or modified between control and treated samples. Absorbance (211 nm) Retention Time

32 APEX FTMS APEX-Q Industrial Design Life Science Tools is our Business

33 Typical Project Milestones Weeks DIGE AAA results Differential Gel Preparative gel Mass Spectrometry Data Analysis ICAT AAA results ICAT Labeling Cationexchange Mass Spectrometry Data Analysis PF2D AAA results 2D Chromatography ICAT Labeling Mass Spectrometry Data Analysis

34 Disease Based Proteomics Receptor-Mediated Signaling DIGE Phosphoproteome Profiling Transcriptional Regulation ICAT Endocytosis and Exocytosis 2D-LC Beckman Coulter System Lipid Profiling Regulation of the Cytoskeleton MudPIT Translational Research Identification and Characterization of Signaling Proteins Whose Expression is Up/Down Regulated Serum Biomarkers Targeted Proteomics Clinical Studies of Biomarker Efficacy

35 2D Electrophoresis 2D gel electrophoresis a Gateway proteomic tool for the separation of complex protein mixtures from many different biological sample types Pros: -High resolution separation by pi and MW -Possible to resolve several thousand proteins on a single gel -IPG strips feature high reproducibility and resolution (single ph unit) -PTMs, such as phosphorylation and glycosylation are clearly visualized Cons: - Membrane/ hydrophobic proteins have solubility problems a) can be addressed by detergents, e.g., ASB-14, Triton X-100 b) some proteins are prone to precipitation at their IE point c) try an alternative technology, such as ICAT -Gel-to-gel reproducibility and therefore Quantitation is a problem

36 DIFFERENTIAL FLUORESCENCE GEL ELECTROPHORESIS PROTEOME PROFILING (DIGE)

37 DIGE Detection Limits # of cells protein copies/cell total # of proteins moles of protein ng of 50kd 1 x pmole 8,304 (~1 g of tissue) pmole 830 MS DIGE pmole fmole fmole fmol 0.08

38 DIGE method is highly reproducible but only with good sample preparation! -use of label compatible buffers 7M Urea, 2M Thiourea, 4% CHAPS, in 30 mm TRIS adjusted to ph 8.5 -ensure that the sample preparation is homogenous - clean -up of sample via precipitation, dialysis, nuclease -low abundance issues are best addressed at the sample preparation step a)subcellular fractionation b)directed biochemical enrichments/pre-fractionation c)microdissection/punches of specific tissue regions d)removal of high abundance species -quantitate samples prior to analysis to normalize, AAA is best -ensure there is enough sample for downstream studies!

39 Albumin and IgG depletion from serum

40 cy 3 image

41 cy 5 image

42 spots with a greater or = to 3X increase (Blue) or Decrease (Red) In cy5/cy3 spot volume ratio

43