Proteomic analysis of biological fluids

Transcription

1 Proteomic analysis of biological fluids Anne Gonzalez, CNRS-IPBS, Toulouse Équipe «Analyse Protéomique et Spectrométrie de Masse des Biomolécules» Odile Schiltz-Bernard Monsarrat Workshop «Protéomique et Maladies Rares», 25 septembre 212

2 Biofluid proteomics: objectives and challenges Serum Plasma Cerebrospinal fluid Urine Bile Tears Saliva Biological fluids are in close contact with tissue that may liberate protein components and their protein content may be affected by the disease discovery of new biomarkers for diagnosis and monitoring of therapy outcome Key issues for biomarker identification using proteomic approaches: selection in the discovery phase of a fluid in close proximity with the organ affected by the disease, to increase the probability of finding a biomarker originating from pathological tissue in-depth characterization of the fluid using a proteomic method enabling the detection of a high number of species, even low-abundant ones analysis of a significant number of samples using reproducible quantitative proteomic methods, to yield statistically significant results a validation phase following the discovery phase, using a high-throughput method applicable on a larger cohort of patients

3 Analytical techniques for biofluids proteomics 2D gels Separation of each protein isoform on a 2D gel Spot detection/quantification performed by protein staining or fluorescence A limited number of spots are then characterized by MS shotgun proteomics (nanolc-ms/ms) Trypsin digestion of protein mixtures Systematic sequençing by MS/MS -> protein identification Quantification by analysis of MS signal MRM based quantitative assays Trypsin digestion of protein mixtures A few proteotypic peptides are specifically monitored as signature peptides for a panel of candidate proteins Quantification by analysis of MS signal + internal peptide standards Tryptic peptides MS MS/MS b2 t b3 Nano-LC y4 MS/MS y5 MS analysis (protein identification) AIILAAAPGEK y6 y7 y8 y9 Elution time

4 Analytical techniques for biofluids proteomics 2D gels Separation of each protein isoform on a 2D gel Spot detection/quantification performed by protein staining or fluorescence A limited number of spots are then characterized by MS Dynamic range/ sensitivity Quantitative analysis Gel image analysis shotgun proteomics (nanolc-ms/ms) Trypsin digestion of protein mixtures Systematic sequençing by MS/MS -> protein identification Quantification by analysis of MS signal MRM based quantitative assays Trypsin digestion of protein mixtures A few proteotypic peptides are specifically monitored as signature peptides for a panel of candidate proteins Quantification by analysis of MS signal + internal peptide standards Relative quantification of MS signal for >1 s peptide ions Bioinformatics Analysis of MS signal for a few peptide ions + internal standard

5 Analytical techniques for biofluids proteomics 2D gels Separation of each protein isoform on a 2D gel Spot detection/quantification performed by protein staining or fluorescence A limited number of spots are then characterized by MS shotgun proteomics (nanolc-ms/ms) Trypsin digestion of protein mixtures Systematic sequençing by MS/MS -> protein identification Quantification by analysis of MS signal Non-targeted Discovery MRM based quantitative assays Trypsin digestion of protein mixtures A few proteotypic peptides are specifically monitored as signature peptides for a panel of candidate proteins Quantification by analysis of MS signal + internal peptide standards Targeted Validation

6 In-depth proteomic characterization: dynamic range Dynamic range of protein concentrations spanning 12 orders of magnitude in plasma Anderson, N. L. (22) Mol. Cell. Proteomics

7 How to deal with biofluids dynamic range? Analysis of microparticles or exosomes Miguet et al, J Proteome Res. 29. Proteomic analysis of malignant B-cell derived microparticles reveals CD148 as a potentially useful antigenic biomarker for mantle cell lymphoma diagnosis.

8 How to deal with biofluids dynamic range? Prefractionation / depletion of highly abundant proteins Shi et al, Proteomics 212

9 How to deal with biofluids dynamic range? Prefractionation / depletion of highly abundant proteins Shi et al, Methods 212

10 How to deal with biofluids dynamic range? Reduction of sample dynamic range with ProteoMiner beads LARGE DYNAMIC RANGE REDUCED DYNAMIC RANGE Hexapeptide ligand library (~ 64 millions of different ligands) Flow-Through: Highly abundant proteins Roux-Dalvai et al. Extensive analysis of the cytoplasmic proteome of human erythrocytes using the peptide ligand library technology and advanced mass spectrometry. Mol Cell Proteomics 28 Mouton-Barbosa et al, In-depth exploration of cerebrospinal fluid by combining peptide ligand library treatment and label-free protein quantification. Mol Cell Proteomics.21

11 About Cerebro-Spinal Fluid Blood vessel Fenestrated blood capillary Brain CSF epithelial cells with tight junction Physical protection of the brain and metabolic function A small amount of CSF originates from the extracellular space of the brain potential biomarkers for neurological diseases ependyma with gap junctions Ventricular cavity Bottlenecks for proteomic studies on CSF: Very large dynamic range of protein concentrations : 1 1, albumin = 45% of total protein Low protein concentration :.4 mg/ml (2x less than serum) Low available volume : Lumbar puncture = 1 to 2 ml

12 ProteoMiner treatment of CSF NH2 Library COOH Library Albumin (45%) Number of identified proteins by LTQ-Orbitrap analysis 1146

13 Label-free quantification of ProteoMiner-treated CSF using MFPaQ and an identification database Number of quantified proteins after single run analysis Distribution of Coefficients of Variation for peptides intensities (4 replicates) after equalization on 5µL or 2µL beads µl 5 µl CV Reduction of sample dynamic range through sample treatment (Proteominer beads) Use of a label-free quantitative approach, easy to implement on biological fluids (MFPaQ software) Increase the number of quantified proteins through the use of a protein identification database PROTELL: Biomarkers identification in CNS relapse of diffuse large B cell lymphomas Coordinator: Catherine Thieblemont, hôpital Saint-Louis, Paris

14 Discovery of candidate urinary biomarkers of congenital unilateral ureteropelvic junction (UPJ) obstruction (Chrystelle Lacroix, collaboration équipe Joost Schanstra, INSERM U858, Toulouse) scintigraphies Spontaneous resolution Spontaneous resolution Intermediate obstruction? Relief surgery (pyeloplasty) pyeloplasty Ureteropelvic junction (UPJ) obstruction birth 1 to 2 years Identify early urinary biomarkers indicative for surgery Better understand the pathophysiology of UPJ obstruction

15 Samples (n=5/group) for discovery Healthy Mild obstruction No surgery Severe obstruction Need surgery spontaneous healthy bladder pelvis resolution Sample processing by FASP nanolc-ms/ms analysis on a LTQ-Orbitrap Velos,2h gradients Quantitative analysis with MFPaQ surgery

16 -Log1(pvalue Student) Label-free quantitative analysis of urine samples using MFPaQ Around 8 urinary proteins quantified Volcano plot showing t test p-values versus protein ratio -Log1(pvalue Student) Log2(Severe Pelvis/Healthy Ratio) -Log1(pvalue Student) Log2(Severe Bladder/Healthy ratio) Log2(Mild/Healthy ratio)

17 Anti-ARG1 signal ARG1 immunodetection in urinary samples LC-MS data Western-blot data Healthy Mild Severe Severe Bladder Pelvis n= 11 n= 19 n= 11 n= 5

18 Validation by MRM Healthy sample Intensity, cps ,63 Peptide ,63 16,5 17, 17,5 18, 18,5 19, Time, min / / / Intensity, cps ,55 Peptide , ,5 13, 13,5 14, / / / Pool of patients , 16,5 17, 17,5 18, 18, Time, min 1,6e4 1,4e4 1,2e4 1,e4 8, 6, 4, 2,, 13,48 1,3e4 1,e4 5e3 13,48, 12,5 13, 13,5 14, Time, min Confirmed expression where Western blot failed Perspectives : Detection on the smallest volume of urine to increase patient number during validation o Optimisation of the sample preparation o Relative quantification of the proteins for validation on a cohort of 2 new samples per group

19 Monitoring biomarkers by MRM in large patient cohorts Hüttenhain et al, Reproducible quantification of cancer-associated proteins in body fluids using targeted proteomics. Sci Transl Med. 212 Jul Generation of a library of MRM assays for more than 1 proposed candidate biomarker proteins, previously associated with cancer Detectability in biofluids: 182 proteins detected in depleted plasma and 48 in urine Monitoring of 34 biomarkers candidates across 83 patient plasma samples

20 Remerciements Proteomics and Mass Spectrometry of Biomolecules Florence Dalvai Emmanuelle Mouton-Barbosa David Bouyssie Roxana Martinez Chrystelle Lacroix Alexandre Stella Odile Schiltz Bernard Monsarrat