New New diagnostic opportunities for Primary Immune Deficiencies; development of a NextGen PID array. Dr M.E. van Gijn. Clinical Laboratory Geneticist

Transcription

1 New New diagnostic opportunities for Primary Immune Deficiencies; development of a NextGen PID array Dr M.E. van Gijn Clinical Laboratory Geneticist Laboratory for DNA diagnostics Department of Medical Genetics

2 WKZ diagnostic expertise center Primary Immune Deficiencies

3 Primary Immune Deficiencies (PID) Classification: B cell, combined B+T cell, NK cell, granulocyte, complement, syndromatic PID, well defined PID, auto inflammatory diseases Diagnosis made on clinical and laboratory grounds, e.g.: Immunoglobubulin levels Lymphocyte phenotyping Functional testing of lymphocytes. Genetic confirmation whenever possible.

4 PID diagnosis complicated: Mutations in >180 genes associated with PID Genotype-phenotype correlations can be low: one genetic defect may cause a variety of symptoms one immunological disease may be caused by mutations in up to 20 different genes Currently diagnostic Sanger sequencing available for ~50% of associated genes

5 Severe combined immunodeficiencies T -, B + SCID IL-2Rg deficiency* JAK3 deficiency* IL-7Ra deficiency CD45 deficiency CD3d/e/z deficiency Coronin-1A deficiency T -, B - SCID Adenosine deaminase (ADA) deficiency RAG1/2 deficiency Artemis deficiency (DCLRE1C) DNA PKcs deficiency Reticular dysgenesis T+/-, B+ DNA ligase IV deficiency Cernunnos deficiency CD40 ligand deficiency CD40 deficiency Purine PNP deficiency CD3g deficiency CD8 deficiency ZAP-70 deficiency Ca++ channel deficiency HLA class I/II deficiency Winged helix deficiency IL-2Ra deficiency STAT5b deficiency Itk deficiency DOCK8 deficiency Bubble boy disease

6 Importance of efficient genetic diagnosis Initiation of the optimal treatment modality Delay in right treatment: organ damage Delay in referral for stem cell transplantation: decreased survival organ damage auto immune phenomena Genetic counseling Identifying family members at risk Prenatal testing Disease prognosis

7 Next Generation Sequencing Advantages Mutation detection in large, complex genes (e.g. DOCK 8; 48 exons, bp) Parallel analysis of large number of genes Uniform Lab Protocol for all requests (Automation possibilities) Challenges Efficient Enrichment Standardized Data analysis Data interpretation Validation for diagnostic use

8 Laboratory process of PID NGS Library preperation, barcoding individual DNA s Equimolar Pooling of barcoded DNA samples (n=10-35) Enrichment for 180 PID genes (IUIS classification) Agilent SurePrint G3 1M Custom CGH Microarray (1 MB footprint) Custom SureSelectXT design (1.8MB footprint) SOLiD 5500XL for sequencing

9 Mapping and Variant calling via in house bioinformatic pipeline Mapping to Ensembl reference genome (GRCh37/hg19) using Burrows-Wheeler Aligner 3 unique reads 4x in the first 25 bases of a read raw coverage 15x base quality 25 variant percentage 20%.

10 Coverage statistics targeted resequencing Enrichment Median coverage (SD) Uncovered (SD) Covered >20x (SD) Called Genotype (SD) Array 338 (99) 2.87% (0,93) 93,77% (2,47) n.d. SureSelect 192(54) 2,67% (0,13) 91,78% (1,38) 91% (1,05)

11 Comparison to Illumina HumanOmni1- Quad v1.0 genotyping Illumina OmniQuad 1M array: >1 x10 6 SNP s Enrichment Array (n=4) In Solution (n=8) Genotype in both 96,2% (n~1130) 92,4% (n~2100) Concordance FN FP 99,47% 0,48% 0,05% 99,61% 0,34% 0,05% Sensitivity: > 99,5% (FN) Specificity: > 99,9% (FP)

12 Gene specific coverage analysis

13 Quality Highly reliable genotype call made in: 88% to 93 % of target sequence Low genotype call percentage (<40% of bases) for: CARD9, C4A, C4B, CFD, ELANE, FCGR1A,FCGR2B, IKBKG and NCF1 15 % (on-array) / 20% (in solution): genes with low genotype call percentage in one or several exons (core genes supplement with Sanger sequencing)

14 Retrospective analysis for known pathogenic mutations (on-array) Gene Mutation Effect Mutation type Detected BTK c.333del p.tyr112fs deletion yes BTK c.803_804del p.tyr268fs deletion yes BTK c.[1875a>g;1876c>a] p.asp521gly missense yes BTK c.763c>t p.arg255* nonsense yes CD40LG c.494_498del p.gln166fs deletion yes DOCK8 c g>t # p.? splice yes DOCK8 c.850_851del p.leu284fs deletion yes FAS c.563_566del Val188fs deletion no IL2RG c.328g>t p.glu110* nonsense yes JAK3 c.1765g>a # p.gly589ser missense yes STXBP2 c.1621g>a p.gly541ser missense yes STXBP2 c.1172delc p.pro391fs deletion yes STX11 c.369_376delinstgg p.val124fs indel no SPINK5 c.2487_2490dup p.asn831fs insertion yes SPINK5 c.1476dela p.lys493fs deletion yes TNFRSF13B c.310t>c p.cys104arg missense yes UNC13D c.551g>a p.trp184* nonsense yes UNC13D c.2695c>t # p.arg899* nonsense yes WAS c g>a p.? splice yes

15 Retrospective analysis for known pathogenic mutations (in solution) Gene Mutation Effect Mutation type Detected ADA c.736c>t p.gln246* nonsense yes ATM c.788del p.leu263fs deletion yes ATM c.1561_1562del p.arg521fs deletion yes CD3G c.80-1g>c # p.? splice yes CD19 c.969_970insa # p.arg325fs insertion yes CD81 c.561+1g>a p.? Splice yes IGHM c.1406c>a p.thr469asn missense yes IKBKG c.505g>c p.ala169pro missense no Il2RG c.670c>t p.arg224trp missense yes IL7RA c.696_697inst p.asn232fs deletion yes IL7RA c.876+1g>a p.? splice yes IL7RA c.898_902del p.pro300fs deletion yes JAK3 c.1153g>t p.ala385ser missense yes LIG4 c.1297_1299del p.gln433del inframe deletion yes RAG1 c.1516c>t p.leu506phe missense yes RAG1 c.2333g>a p.arg778gln missense yes RAG2 c.913c>g p.pro305ala missense yes RAG2 c.1357t>a p.trp453arg missense yes ZAP70 c.t121g # p.leu337arg missense yes

16 Validation results 33 samples with 38 known mutations in 24 different genes. 21/22 SNV s detected. The two variants that were missed were located in known poorly covered genes; IKBKG. 9 small deletions, 3 insertions/duplications and 1 indel mutation. Of those variants 11/13 were detected. Missed because of mapping problems: c.563_566del mutation (4bp deletion) in the FAS gene located in a T-stretch c.369_376delinstgg mutation in STX11

17 STX11 c.369_376delinstgg Raw sequence data mapping problem..

18 Detection of exonic deletions Hyper IgE syndrome (DOCK8): DOCK8: 48 exons, well covered with PID NGS Mutation spectrum: mostly multi-exonic deletions PID NGS suitable? In house bioinformatic algoritm (Z-scores) detects exonic deletions 8/8 deletions retrospectively detected in DOCK8

19 Variant analysis and classification Genotype generated after bioinformatic analysis: ~1500 variants/patient Cartagenia BENCHlab NGS module for analysis and classification of variants

20 Managed Variant List filtering Alamut mutation database uploaded as Managed Variant List (MVL) Currently validating analysis pipeline using MVL Goal: General Pipeline based primarily on MVL and HGMDpro Additional pipelines needed?

21 Variant analysis filtering Presence in our in house mutation database. n=1500 Classify Benign, VUS, pathogenic Presence in HGMD database. Classify (possible) Pathogenic >5 % incidence in the normal population (dbsnp, EVS, 1000Genomes) Silent variants, variants that do not alter splice site Classify Benign consensus sequences Conservation and impact on protein sequence Polyphen2, SIFT, GERP, Grantham, Alamut (V2.2). Nonsense, frameshift, and canonical splice site n=15-25 Classify VUS, pathogenic variants

22 Add phenotype to sample

23 30-May Rank variants by phenotype involvement

24 30-May Hyperlink to literature

25 Prospective analysis in PID patients 26 PID patients without genetic diagnosis 5 patients: genetic diagnosis 3/5 patients reclassification: WHIM (WARTS, HYPOGAMMAGLOBULINEMIA, INFECTIONS, AND MYELOKATHEXIS) syndrome without warts Hyper IgE syndrome with normal IgE levels PAPA (PYOGENIC STERILE ARTHRITIS, PYODERMA GANGRENOSUM, AND ACNE ) syndrome without arthritis 6 Netherton patients (SPINK5, 33 exons, 3745 bp) Autosomal recessive 5 patients genetic diagnosis 1 patient 1 mutation detected

26 NGS request/clinical information NGS result -Known mutation in relevant gene -Only SNPs report lab specialist genetics, lab specilaist immunology, clinical genetisist, clinical immunologist, VUS in relevant gene Known mutation in clinically unsuspected gene VUS in unsuspected gene Discuss functional tests, clinic and literature Discuss functional tests, clinic and literature Discuss functional tests, clinic and literature Additional (functional) tests? Additional (functional) tests? report, possibly recommendation additional (functional) tests Report, Additional (functional) tests? report mention/do not mention VUS?

27 Conclusions Our data indicate that this new NGS-based approach is an efficient strategy for rapid mutation and exonic deletion detection in PID. The diagnostics yield of such a strategy remains to be evaluated, and is likely to be highest in cases with large genotype phenotype variability or in diseases where defects in multiple genes may cause one specific phenotype. Moreover our approach will give more insight in genotype phenotype correlations for the different PID disorders enabling earlier diagnosis and better treatment.

28 Acknowledgements (Team effort) Dept. of Medical Genetics UMCU Ies Nijman Stef van Lieshout Patrick van Zon Esther Janson Marc van Tuil Maartje Vogel Marlous Hoogstraat Bert van der Zwaag Martin Elferink Hans Kristian Ploos van Amstel Edwin Cuppen Funding: Fonds NutsOhra Dept. of Pediatric Immunology UMCU Joris van Montfrans Marianne Boes Suzanne Pasmans Dept. of Immunology UMCU Lisette van de Corput Dept. of Immunology, ErasmusMC Mirjam vd Burg University Children's Hospital, Munich Ellen Renner Dept. of Medical Genetics AMC Marielle Alders