Autism (ASD) Genetics to Mechanism

We have brought together genome wide associate data sets (GWAS) and ClinVar analysis of genes tied to autism development highlighting lack of mechanistic knowledge of the system and an overwhelming number of Variants of Uncertain Significance (VUS).

Analysis of ASD genetics using GWAS and ClinVar. A) Breakdown of “Autism” associated loci (from GWAS) for 420 lead SNPs resulting in 10,029 LD SNPs, with multiple predicted to be functional using RegulomeDB (red). B) Breakdown of the coding variants from “Autism” GWAS LD blocks with damaging predictions from PolyPhen2, Provean, and SIFT. C) ClinVar analysis for variants within 84 SFARI top genes for ASD showing clinical interpretation breakdown of variants. D) Analysis of SNPs in C analyzed with PolyPhen2, Provean, and SIFT binned into clinical annotation. E-F) ClinVar variants found for “Autism” (E) and “Epilepsy” (F) binned into clinical interpretation. G) STRING network for genes identified in the genetic analysis of A-F.
Analysis of ASD genetics using GWAS and ClinVar. A) Breakdown of “Autism” associated loci (from GWAS) for 420 lead SNPs resulting in 10,029 LD SNPs, with multiple predicted to be functional using RegulomeDB (red). B) Breakdown of the coding variants from “Autism” GWAS LD blocks with damaging predictions from PolyPhen2, Provean, and SIFT. C) ClinVar analysis for variants within 84 SFARI top genes for ASD showing clinical interpretation breakdown of variants. D) Analysis of SNPs in C analyzed with PolyPhen2, Provean, and SIFT binned into clinical annotation. E-F) ClinVar variants found for “Autism” (E) and “Epilepsy” (F) binned into clinical interpretation. G) STRING network for genes identified in the genetic analysis of A-F.

Using Computational tools to identifying functional regions on ASD associated genes

Using our computer strategies we have take multiple genes involved in autism and developed deep codon usage evolutionary information for each.

Genes analyzed through our deep codon evolution as preliminary data for grant. Each gene was an alignment of ORFs followed by codon selection analysis and amino acid conservation with the scores from each amino acid placed on a 21 codon sliding window (such that score is any site plus 10 up and downstream of the site).
Genes analyzed through our deep codon evolution as preliminary data for grant. Each gene was an alignment of ORFs followed by codon selection analysis and amino acid conservation with the scores from each amino acid placed on a 21 codon sliding window (such that score is any site plus 10 up and downstream of the site).

Detailing ASD VUS

Then we take a deep dive into all known ClinVar variants for the gene, identifying VUS that are likely functional and developing a working hypothesis for variant mechanism that can be tested.

TSC1 and TSC2 genomic variants. A-B) Conservation analysis of TSC1 (A) and TSC2 (B) amino acids. C-D) Variant impact scores for TSC1 (C) and TSC2 (D) for protein coding SNPs identified in gnomAD (black), ClinVar (red) and COSMIC (blue). E-F) Predicted functional outcomes for variants of the 3 databases (colors same as C-D) in TSC1 (E) and TSC2 (F) using PolyPhen2 (left), SIFT (middle), or our conservation score (right). G-H) Pathogenic variants from ClinVar for TSC1 (G) and TSC2 (H) shown for our variant impact scores. I-J) VUS from TSC1 (I) and TSC2 (J) showing variant impact scores and listing the top-ranking variants that will be studied in this grant.
TSC1 and TSC2 genomic variants. A-B) Conservation analysis of TSC1 (A) and TSC2 (B) amino acids. C-D) Variant impact scores for TSC1 (C) and TSC2 (D) for protein coding SNPs identified in gnomAD (black), ClinVar (red) and COSMIC (blue). E-F) Predicted functional outcomes for variants of the 3 databases (colors same as C-D) in TSC1 (E) and TSC2 (F) using PolyPhen2 (left), SIFT (middle), or our conservation score (right). G-H) Pathogenic variants from ClinVar for TSC1 (G) and TSC2 (H) shown for our variant impact scores. I-J) VUS from TSC1 (I) and TSC2 (J) showing variant impact scores and listing the top-ranking variants that will be studied in this grant.

Neural Organoids and mixed neural cultures

To test the role of genes we use neural progenitor cells (NPCs) and iPSCs that we can manipulate with CRISPR/Cas9 technology followed by differentiation to brain organoids or mixed neural cultures.

iPSC generated neural organoids
iPSC generated neural organoids
Mixed Neural Cultures from male and female NPCs
Mixed Neural Cultures from male and female NPCs

Studying the transcriptional control of male elevated ASD risk

Finally, we study the complex synergistic regulation of the male specific gene, SRY, and the Androgen Receptor, AR, that regulate male specific transcriptional regulation.

AR and SRY/SOX3 coregulation. A) Model of AR and SRY DNA recruitment. B) Cell culture luciferase promoter assay with co-transfections with SRY, AR, and mutated SRY (P to T) in charcoal stripped sera media without (gray) or with (red) testosterone. C) Animal blood pressure following SRY delivery into the kidney of rats. Olmesartan (AT1 inhibitor) blocked the blood pressure role of SRY. D) Diagram for crossing the SHR/y hypertensive rat with the testicular feminized male (tfm, AR mutation) that blocks the blood pressure elevation in males
AR and SRY/SOX3 coregulation. A) Model of AR and SRY DNA recruitment. B) Cell culture luciferase promoter assay with co-transfections with SRY, AR, and mutated SRY (P to T) in charcoal stripped sera media without (gray) or with (red) testosterone. C) Animal blood pressure following SRY delivery into the kidney of rats. Olmesartan (AT1 inhibitor) blocked the blood pressure role of SRY. D) Diagram for crossing the SHR/y hypertensive rat with the testicular feminized male (tfm, AR mutation) that blocks the blood pressure elevation in males