SYSTEMS BIOLOGY APPROACHES TO IDENTIFY MECHANISMS OF IMMUNE MEDIATED PROTECTION TRANSLATING RESEARCH INTO HEALTH
Novel assays to decipher protective immune responses Decoding the immune response to infectious diseases requires a functional genomics systems approach immune system is reactive and adaptive involves a diverse array of components, regulatory pathways and networks Biomarkers of pathogenesis and correlates of immune mediated protection have yet to be identified for most diseases and vaccines, respectively The use of integrated technologies which allow the analysis of multiple parameters is suited for this purpose VACCINE & GENE THERAPY INSTITUTE
QRT-PCR Clinical Informatics Patients & the Community Animal Models In Vitro Systems iscan GAIIx Tissues Cells DNA RNA Proteins Phosphoproteins Systems Approach FACS CBA Genomics (Sequencing/Transcriptome, Genotyping, Gene Regulation, Whole-Genome Expression) Proteomics (Content, Quantity, Modifications) Bioinformatics Modeling Hit Identification Functional Clustering Gene Ontology Canonical Signaling & Transcriptional Networks Mathematical Models Biomarker Identification Integrated Platform Host Response Correlates of Immune Pathology/Protection Immunity Diagnostics Prognostics Therapeutics Vaccines
STRATEGY Work with very homogeneous groups of subjects : gender, ethnic group, age, disease status Work with very homogeneous populations of cells Work with minimal number of cells Develop assays which are multi-parametric, high throughput and highly sensitive Multiple validation steps including PCR, proteomics and genome wide sirnas VACCINE & GENE THERAPY INSTITUTE
Similar mechanisms to protection: lessons from the Live attenuated virus and Elite contollers TRANSLATING RESEARCH INTO HEALTH
APPROACH Deep deconvolution of CD4+ and CD8+ T cellrecognized epitopes within the SIV proteome and functional characterization of each measurable optimal epitope response Use of microarray and bioinformatic technology to characterize both in situ immune responses in target tissues (either innate or adaptive) and to identify T cell response characteristics that correlate with protection. VACCINE & GENE THERAPY INSTITUTE
PC125 Microarray samples Tissues of RNA Sample PBMC(ABC) BAL(ABC) LN(ABC) Gut(ABC) PBMC(ABC) BAL(ABC) LN(ABC) Gut(ABC) Gut(ABC) PBMC(ABC) BAL(ABC) LN(ABC) Gut(ABC) LN(B) LN(A) LN(AB) Gut(AB) PBMC(AB) BAL(AB) PBMC(AB) BAL(AB) LN(AB) Gut(AB) Gut(AB) PBMC(AB) BAL(AB) PBMC(B) LN(AB) Gut(AB) Time point -6/-7 PID 0PID 9PID 120/126 PID 168/169 PID 211PID 295PID 344PID (-7PCD) 351PID (0PCD) 4PCD 11PCD 14PCD 21PCD LAV infection SHIV-89.6 SIVmac239Δnef SIVmac239Δ3 SIVsmE543Δnef SC SIVmac239 Monkey ID SIVmac239 I.V challenge LAV Group A Group B Group C LAV; live attenuated vaccine SHIV; Simian-Human Immunodeficiency Virus SIV; Simian Immunodeficiency Virus SC SIV; Single Cycle SIV LN; lymph node BAL; Bronchoalveolar lavage PID; post LAV infection day PCD; post challenge day SHIV-89.6 A1=22741, A3=23657 B2=23009, B10=23690 C1=22783, C5=23701 SIVmac239Δnef SIVmac239Δ3 SIVsmE543Δnef SC SIVmac239 A8=24724, A9=24746 B6=23166, B12=23789 C2=23343, C4=23696 A2=23048, A4=23698 B1=23003, B8=23434 C6=23703, C10=23841 A5=23840, A7=23880, A11=23687, A12=23793 B3=23026, B5=23164 C3=23431, C7=23709 A6=23847, A10=24761 B4=23070, B11=23722 C8=23796, C9=23808 Control A13=23278, A14=23605 B7=24347, B9=24465 C11=24062, C12=25430
Viral Load SHIV (89.6) SIVmacE543 (dnef) Viral Copy / ml 2,500 2,000 1,500 1,000 500 0-10 -5-500 0 5 10 15 20 25 Days Rh22741 Rh23657 Viral Copy / ml 50,000,000 40,000,000 30,000,000 20,000,000 10,000,000 0-10 -10,000,000-5 0 5 10 15 20 25 Days Rh23840 Rh23880 Rh23687 Rh23793 SIVmac239 (dnef) SIVmac239 (SC) Viral Copy / ml 350 300 250 200 150 100 50 0-10 -5 0 5 10 15 20 25 Days Rh24724 Rh24746 Viral Copy / ml 60,000,000 50,000,000 40,000,000 30,000,000 20,000,000 10,000,000 0-10 -10,000,000-5 0 5 10 15 20 25 Days Rh23847 Rh24761 SIVmac239 (d3) Unvaccinated Controls Viral Copy / ml 1,200 1,000 800 600 400 200 0-10 -5 0 5 10 15 20 25 Rh23048 Rh23698 Viral Copy / ml 80,000,000 60,000,000 40,000,000 20,000,000 0-10 -20,000,000-5 0 5 10 15 20 25 Rh23605 Rh23278 Days Days
LAV analysis No Take No Protection Take with Protection SIV239 6 3 2 1 SIV239dNEF 6 scsiv239 SIV239d3 3 SHIV89.6 3 Control 6 SIVE543dNEF 3 3 2
Summary of Results In general, LAVs elicit Very broad SIV-specific CD4+ and CD8+ T cell responses of modest to moderate magnitude, Medium to high titre SIV-responses which neutralize tissue culture adapted SIV, but not the challenge virus Protection ranging from complete to none Neither nab titers, nor the magnitude and breadth of peripheral blood or effector site (BAL) T cell responses accurately predict challenge outcome Protection is, however, closely linked to LAV replication LAV replication is highest in LN, and protection is strongly correlated with both the degree of cell-associated LAV in LN and the magnitude of the CD4+ and CD8+ LN T cell response (CD4 > CD8) VACCINE & GENE THERAPY INSTITUTE
QUESTIONS Distinct pathways in controller and non controller macaques at baseline, prior to challenge ( 9 PID and -7 PCD ) and upon challenge Distinct pathways in different lymphoid tissues in controller and non controller macaques at baseline, prior to challenge and upon challenge Which pathways can predict protection VACCINE & GENE THERAPY INSTITUTE
Time Point Approach For each time point and operational category, the number of differentially expressed genes relative to the baseline are presented as well as the number of genes with a differential time response between operational categories. Primary focus became the LN «No take» versus «No protection» contrast
Two-Way cluster analysis discriminates controllers from non-controllers in BAL Non-cont. Day 4 Controllers Controllers Day 14 Non-cont.
Hypothesis Generation / GSEA Objective: To find the most relevant co-regulated gene networks (or pathways) that characterize the samples identified by Challenge Outcome Status. Strategy: (1) We use gene set enrichment analysis (GSEA), an annotation-driven algorithm that determines the statistical difference between the Challenge Outcome States based on the frequency of labels such as function, pathway, or other descriptive biological annotation.
No_take_LN_4PCD.VS.No_Protection_LN_4PCD
Differential gene expression No take relative to no protection: LN, 4PCD, baseline normalized Canonical pathway: Protein Ubiquitination
Pathways_C2_No_take_LN_14PCD.VS.No_ Protection_LN_14PCD
POST-CHALLENGE ANALYSIS SUMMARY Systems biology and transcriptional profiling allows the identification of unexpected pathways differentially regulated in controller and non controller macaques Genes upregulated in the LN at 4 days post-challenge in «notake» macaques relative to non-protected animals include HSP and protein ubiquitination genes. These gene responses resolve by day 14. At 14+ days post-challenge, «no-take» macaques lack the integrated antiviral IFN response gene expression present in non-protected animals in the LN but T and B cell receptor signaling pathways are upregulated. Inflammation genes expressed in the gut of non-protected animals at 21 DPC.
PRE-CHALLENE ANALYSIS PREDICTORS OF PROTECTION TRANSLATING RESEARCH INTO HEALTH
PRE-CHALLENGE ANALYSIS Hypothesis Generation / Class Prediction Objective: To find identify pre-challenge correlates of immune protection or features (genes) of expression data that are predictive of the biological outcome for each tissue. Strategy: (1) We used Linear Discriminant Analysis (LDA), a data-driven method that looks for a linear combination of variables that specifically models the differences between outcome classes. (2) We data mine these lists to generate new hypotheses.
THE EARLY IMMUNE RESPONSE OF MACAQUES THAT WILL SHOW PROTECTION IS DIFFERENT TRANSLATING RESEARCH INTO HEALTH
EARLY POST-IIMUNISATION PREDICTORS OF PROTECTION Ingenuity analysis was performed with significant genes at 9PID (protection vs. no protection): 121 Immune related genes Data mining to look for direct interactions between the 121 molecules
PBMC THE INTERFERON PATHWAY IS SUSTAINED IN MACAQUES THAT WILL SHOW PROTECTION B CELL SIGNATURE IS ABSENT IN PROTECTED MACAQUES WHILE EXPANDED IN NON PROTECTED MACAQUES
INNATE IMMUNE RESPONSE IS SUSTAINED IN PROTECTED MACAQUES 9 DAYS POST-IMMUNISATION
SUSTAINED INNATE IMMUNE RESPONSE IN CONTROLLER MACAQUES EARLY AFTER IMMUNISATION Two major observations are observed at 9 days post-immunization that correlate with protection a year later in immunized animals: Interferon induction is much higher in protected macaques B cells appears to be down regulated in the blood of protected macaques and appear to be expanded in non protected macaques
Linear Discriminant Analysis (LDA) Overview A B C D Tissue Number of Discriminating Genes BAL 180 Gut 52 LN 108 PBMC 360 A putative list of genes were generated for each tissue
Ingenuity Canonical Pathways Linear Discriminant Analysis identified a putative list of genes that segregated each outcome class for each tissue type, using samples pooled by outcome class. the LDA and hypothesis-driven results were used to select canonical pathways that might further differentiate SIVE543dNef (mixed outcome) vaccinated macaques from SIV239dNEF (all no-take)
Differential gene expression SIVE543dNef vs. control: LN, Pre-challenge Canonical pathway: Protein Ubiquitination
Differential gene expression SIVE543dNef vs. control: LN, Pre-challenge Canonical pathway: TCR signaling
Summary Linear Discriminant Analysis identified a putative list of genes that segregated each outcome class for each tissue type. Protein ubiquitination pathway components were upregulated in the LN of SIV239dNEF (no-take) vaccinated macaques pre-challenge relative to SIVE543dNef (mixed outcome), HSPs were found to be generally downregulated prior to challenge relative to controls. On the other hand, the LDA gene lists implicated the involvement of TCR signaling and Antigen Presentation pathways these two pathways were found to be downregulated in SIVE543dNef (mixed outcome) vaccinated macaques pre-challenge relative to SIV239dNEF (all no-take)
HYPOTHESIS Is the same pathway upregulated in Elite contolers
Actual/Pre Error Acute Chronic EC ST dicted Rate Acute 7 0 0 0 0.00 Chronic 2 4 0 0 0.33 EC 0 0 8 0 0.00 ST 0 1 5 2 0.75 Overall error rate 0.266 Rows represent actual class membership Columns represent what the classifier using the 135 genes predicted for the samples. Heat map of 135 genes capable of discriminating between EC, ST, Chronic and acute patients.
HSP and DNAJ family of chaperones Specific gene signature discriminates HIV specific T cells in Elite controllers Chronic Elite Elite Chronic HIV HIV
HSP expression on CD8 T cell subsets V = IDR194 ST = HTM332 EC = LTNP007 HSF1 V ST EC Bip1 V ST EC HSP90 V ST EC CM TM EM Naive TEMRA
CM HSP expression on CD8 T cell subsets HSP70 V ST EC HSP40 (DNAJ) V ST EC HSP60 V ST EC V = IDR194 ST = HTM332 EC = LTNP007 TM EM Naive TEMRA
Network analysis shows increase expression in TNF-α pathway in EC
Increase survival and memory function in HIV specific T cells in EC could be mediated by HSP and mtor networks mtor
CONCLUSIONS Systems biology and transcriptional profiling allows the identification of unexpected pathways differentially regulated in controller and non controller macaques Non controller macaques show specific and temporal induction of inflammatory chemokines in lymph nodes Non controller macaques show sustained upregulation of type I interferons in all tissues VACCINE & GENE THERAPY INSTITUTE
CONCLUSIONS The heat shock and unfolded protein pathways are upregulated in LN from controler macaques This same pathway is also upregulated in T cells from Elite controlers `These pathways are strongly associated to survival and inhibition of apoptosis These pathways are downstream of TNF and other survival cytokines These pathways intercept with mtor, target of rapamycin This pathway should be targeted by futures vaccines and adjuvants VACCINE & GENE THERAPY INSTITUTE
Funders
ACKNOWLEDGEMENTS VGTI Oregon Louis Picker Yoshi Fukazawa VGTI Florida Peter Wilkinson Mark Cameron John Schatzle Li Pan Oleg Yegorov Andrew Smith Angela Duque U de Montreal Jean Philippe Goulet Ali Filali Francois Lefebvre Elias Haddad