Systems Approaches to Blood Based C Cancer Biomarkers Bi k S Hanash S. Hanash shanash@mdanderson.org
Why so few biomarkers to date? y f - Developing biomarkers shares some of the same challenges as developing drugs, yet the investment in biomarkers is far, far less! - Requires a road map from discovery to validation for specific/intended clinical applications specific/intended clinical applications - Need to establish a mechanistic link to the disease process beyond statistical associations
Targeted cancers
If we screen for it and catch it early does it save lives? does it save lives?
Challenges with CT based screening Challenges with CT based screening • High percentage of false positives. 96.4% of pulmonary nodules identified by LDCT p y y are non-malignant unnecessary work- ups ups. • If all smokers and former smokers do three CT scan per NLST protocol, the C S health care system can’t afford it. y
LUNG CANCER Number of lives that could be saved each year with CT screening h h 8 100 8,100 Estimated today’s costs per life saved: $ 100,000 ‐ 240,000 $ 100,000 240,000 Goulart et al J Natl Compr Canc Netw 2012;10:267 ‐ 275
CONCEPTUAL FRAMEWORK CONCEPTUAL FRAMEWORK TUMOR The The INFILTRATING INFILTRATING MICROENVIRONMENT CELLS TUMOR CELLS ANGIOGENESIS In tumor initiation, development and d l d FIBROBLASTS progression S Systemic host response t i h t
Mouse models Genomics Genomics Glycomics Glycomics y Proteomics Proteomics Biomarker panels Biomarker panels Bi Bi k k l l Cancer cells Human studies ♦ plasma/serum l / ♦ whole cell extracts h l ll ♦ tissues ♦ secretome/exosome ♦ surface proteins ♦ nuclear proteins Immunomics Immunomics Immunomics Immunomics Metabolomics Metabolomics Metabolomics Metabolomics
blood From the tumor….to ….. blood to From the tumor
Mass Spectrometry Capability: 30,000 LC/MS runs 3,000 proteins 3,000 proteins 8 ‐ 10,000 proteins 8 10,000 proteins 4 ‐ 6,000 proteins 4 6,000 proteins
Proteomic signatures Chemical Chemical Modifications eg altered glycosylation Protein Cleavages eg Protein Cleavages eg Alternative shed receptors and Splicing adhesion molecules Isoforms Altered dynamics of Formation of Formation of protein sorting eg i i complexes eg release of immune complexes chaperone proteins Translational Translational Translational Translational Implications Implications
+/- various treatments O Light Heavy Cells In Culture H 2 N LYS LYS OH NH 2 2 mix heavy and light CELL SURFACE SECRETOME Cell surface Protein biotinylation Cell culture media TOTAL EXTRACT concentration Lysis and affinity Gl copeptide Glycopeptide capture t Capture Cell Lysis Elution NUCLEAR PROTEINS PHOSPHOPROTEINS
Source of conditioned media proteins Conditioned media Proteins localization Conditioned media Proteins localization • Pre- • Pre- using Uniprot Database filtration filtration 19374 2% 29% Cytoplasm • Post- • Post- 42% Nuclear filtration filtration 7% 2480 Surface 20% Extracellular space Unknown Hi h V l High Value targets t t 176 7% 61% 32% Ratio <1 Ratio S/M >2 Ratio: 1- 2 1 913 DDR1 protein: epithelial discoidin domain-containing receptor 1 Cell line Compartment MS2 counts H1993 Media 65 Cytoplasmic domain T Transmembrane domain b d i H1373 Media 51 Extracellular domain HPC9 Media 37 Signal Peptide H1993 Surface 103 H1373 Surface 73 HPC9 Surface 78 I. Babel
Function of cytosolic and nuclear proteins released into conditioned media A A B B Biological Process (Gene Number of p ‐ value ontology Term) proteins Bonferroni Macromolecule metabolic 48 1.88E ‐ 02 process Cellular component organization p g 30 30 2.84E ‐ 02 2 84E 02 at cellular level Cellular localization 22 9.99E ‐ 03 Establishment of localization in 21 5.58E ‐ 03 cell Protein transport 19 1.72E ‐ 03 Protein localization P i l li i 19 19 4 95E 02 4.95E ‐ 02 Intracellular transport 18 1.99E ‐ 03 Intracellular protein transport 17 9.94E ‐ 07 Cellular protein localization 17 3.05E ‐ 05 Carbohydrate metabolic process 15 2.73E ‐ 02 Oli Oligosaccharide metabolic h id b li 8 2.79E ‐ 05 process Polysaccharide catabolic process 5 3.88E ‐ 02 Oligosaccharide catabolic process 3 8.64E ‐ 03 Glycosphingolipid catabolic 3 4.28E ‐ 02 process p Macromolecule metabolic process Lysosomal enzymes Cellular localization Nuclear localization Component organization p g
Nuclear proteins are released by lung cancer cells in exosomes H920 H920 A A B B Nucleus Media Exo TCE FT XPO1 XPOT TNPO1 Exosome marker ALIX 200 nm Note: from 300mL of media: 20x more exosome fraction from cancer cell lines than transformed control cell
Blood Based biomarkers Newly Diagnosed and post Rx for predictive markers 3+ yrs pre diagnostic 3+ yrs pre-diagnostic 0 2 yrs pre diagnostic 0-2 yrs pre-diagnostic For risk markers for early detection
Blood Based biomarkers Newly Diagnosed and post Rx for predictive markers 3+ yrs pre diagnostic 3+ yrs pre-diagnostic 0 2 yrs pre diagnostic 0-2 yrs pre-diagnostic For risk markers for early detection
Intact Protein Analysis System (IPAS) Case Control Ig bound fraction: MS I m m unodepletion I m m unodepletion bound fraction: MS (Top six proteins) (Top six proteins) Concentration, buffer Concentration, buffer , , exchange and exchange and labeling labeling Reduction w ith DTT Reduction w ith DTT SAMPLE A SAMPLE A SAMPLE B SAMPLE B and Alkylation and Alkylation Light Acrylam ide Light Acrylam ide Heavy Acrylam ide Heavy Acrylam ide SAMPLES MIXED SAMPLES MIXED ANI ON EXCHANGE ANI ON EXCHANGE CHROMATOGRAPHY CHROMATOGRAPHY REVERSE-PHASE REVERSE-PHASE CHROMATOGRAPHY CHROMATOGRAPHY Shotgun LC/ MS/ MS Shotgun LC/ MS/ MS 96 fractions 96 fractions
Mouse models ♦ Lung cancer EGFR: TetO-EGFR L858R /CCSP-rtTA ( H. Varmus/K. Politi ) K Kras: TetO-Kras4b G12D /CCSP-rtTA ( H. Varmus/K. Politi ) T O K 4b G12D /CCSP TA ( H V /K P li i ) Urethane: introperitoneal injection of urethane ( C. Kemp ) SCLC: Trp53 F2-10/F2-10 ; Rb1 F19/F19 ( J. Sage ) ♦ Breast cancer ♦ Breast cancer HER2: MMTV -rtTA/TetO-NeuNT ( L. Chodosh ) PyMT 0.5cm: Tg(MMTV-PyMT) 634Mul (C. Kemp ) PyMT 1.0cm: Tg(MMTV-PyMT) 634Mul ♦ Pancreas cancer PanIN: Pdx1-Cre; LSL-Kras G12D ; Ink4a/Arf lox/lox ( R. DePinho, N. Bardeesy ) PDAC: Pdx1-Cre; LSL-Kras G12D ; Ink4a/Arf lox/lox ♦ Colon cancer: Apc Δ 580/+ ( R. Kucherlapati, K. Hung ) C Kras model ( R. DePinho ) Mlh1 and Msh2 mutant models ( R. Kucherlapati ) ♦ Ovarian cancer: LSL-Kras G12D/+ ; Pten loxP/loxP ( D, Dinulescu, T. Jacks ) G12D/+ Pt loxP/loxP ( D Di ♦ Ovarian cancer: LSL K l T J k ) ♦ Prostate ca (Strain comparison): Pten pc-/- , Pten pc -/- ; Smad4 pc-/- ( R. DePinho ) ♦ Inflammation Acute: Carrageenan-sponge implantation ( C. Kemp ) A t ( C K C i l t ti ) Chronic: intradermal injection of type II collagen ( C. Kemp )
Plasma signature for lung adenocarcinoma driven in part by the master regulator NKX2 ‐ 1
Validation in humans of mouse lung adenocarcinoma blood markers Newly diagnosed set y g 1.0 8 0.8 eFraction New ly diagnosed set 0.6 A ssay y Norm al Cancer T-test Mann W hitney y test 0 e Positive EGFR 1.00 ± 0.067 0.77 ± 0.041 0.0094 0.004 SFTPB 1.00 ± 0.135 1.43 ± 0.205 0.0708 0.0332 0.4 W FDC2 1.00 ± 0.233 4.70 ± 1.145 0.0005 < 0.0001 A NGPTL3 1.00 ± 0.073 1.53 ± 0.205 0.008 0.0038 True 0.2 Com bin ed (AU C= 0.882 ) EG FR (AU C= 0.708 ) SFT PB (AU C= 0.654 ) W FDC2 (AU C= 0.864 ) 0 0.0 AN G PT L3 (AU C= 0.709 ) 0.0 0.2 0.4 0.6 0.8 1.0 False Positive Fraction a se os t e act o
Is there evidence that blood markers could be useful? Validation in the pan Canadian Lung Ca screening study Red: Base model”|: AUC= .642 Blue: Base model + 1 marker (AUC = 736 Blue: Base model + 1 marker (AUC .736 Difference in AUCs p = .0002 Sin et al JCO in press
Harnessing the immune response to tumor antigens for lung cancer early detection • Humoral immune response to tumor antigens occurs early during tumor development g p • Immune response is not limited to mutated or otherwise altered proteins lt d t i • Implementation of a proteomic strategy to identify proteins Implementation of a proteomic strategy to identify proteins in lung cancer that induce an autoantibody response
Search for antigens that induce an antibody response in lung cancer • Recombinant protein arrays • Natural protein arrays • Mass spectrometry of circulating antigen ‐ M t t f i l ti ti antibody complexes • Peptide arrays ( in collaboration with Roche )
Search for antigens that induce an antibody response in cancer The complete repertoire of peptides encoded in The complete repertoire of peptides encoded in the human genome on a chip + (Mutant peptides and pathogen peptides) (Mutant peptides and pathogen peptides)
Recommend
More recommend
