Genomic mapping in outbred mice reveals overlap in genetic susceptibility for HZE ion and γ-ray induced tumors

2016 TAGC IMGS Conference in Orlando, FL. Determining the significance of space radiation exposures: high resolution genomic mapping to determine overlap in susceptibility loci to HZE ion and γ-ray induced tumors

Transcript

1. Elijah Edmondson, DVM DACVP elijah.edmondson@colostate.edu Weil Lab, Colorado State Univeristy Determining the significance of space radiation exposures: high resolution genomic mapping to determine overlap in susceptibility loci to HZE ion and γ-ray induced tumors Chromosome 2
2. NASA: NNJ11ZSA001N The use of genetically diverse populations of mice to represent human heterogeneity that will elucidate the spectrum of tumor types caused by HZE nuclei as well as possible differences compared to γ-rays.
3. Background The use of (2) genetically diverse populations of mice to represent human heterogeneity that will elucidate the spectrum of tumor types caused by (1) HZE nuclei as well as possible differences compared to γ-rays.
4. GCR: • 85% protons • 14% helium • 1% HZE ions
5. HZE ions “Heavy ions” • High (H) atomic number (Z) and energy (E) charged particle – Accelerated by supernova explosions – Atomic nuclei stripped of electrons – Travel at relativistic speeds (up to 85% speed of light) – Contain very high energies (cannot shield) – Ionization is proportional to charge^2 GCR: • 85% protons • 14% helium • 1% HZE ions
6. • Assumption: Terrestrial Radiation (γ-ray) exposures increase the risk of the same tumor types as Celestial Radiation (HZE-ion) exposures. • If these hold, the use of existing epidemiological data, most notably the atomic bomb survivors, of human populations exposed to γ-ray is valid. Hiroshima Nagasaki Current Assumptions
7. Advantages to inbred mice include: • strains produce lower phenotypic variance • thus fewer mice are needed to detect statistical differences Disadvantages of using inbred mice: • strain-specific responses may obscure the variability we expect in a genetically diverse population such as humans Toxicity Studies using Inbred mice
8. •Each marker is informative •Lower mapping resolution •Less phenotypic diversity • High mapping resolution (<3 Mb) • Abundant phenotypic diversity • Large numbers needed • Functional follow-up is challenging in HS *DO has more balanced allele frequencies than HS (multiple funnels used) and allow the ability to utilize CC to test candidate genes QTL Mapping in mice and mouse models of genetic diversity Woods LC, 2013. Physical Genomics. QTL Mapping in outbred populations: successes and challenges. Woods LC, 2013. Physical Genomics. QTL Mapping in outbred populations: successes and challenges.
9. Forward Genetics (unbiased) 1. Define a phenotype 2. Determine gene(s) Reverse Genetics 2. Determine phenotype(s) 1. Alter a gene
10. MegaMUGA (illumina SNP Array: 77,808 SNPs)
11. Each mouse is a genetically unique mosaic of the 8 parental strains. *All of these strains have been completely sequenced. Mouse 20
13. 0.4 Gy HZE ions 240 MeV/n 28Si 600 MeV/n 56Fe 3.0 Gy 137Cs gamma rays Unirradiated Controls Experimental Design 613 mice, ♀ and ♂ 615 mice, ♀ and ♂ 622 mice, ♀ and ♂ • Irradiate at 7-12 weeks old, ♀ / ♂ • Monitor to 800 days of age • Necropsy, all organ systems • Tumor classification (Histo, IHC) • Neurobehavioral & Ocular exams • Map Quantitative Trait Loci (QTL)
14. 1. Generalized linear modeling (link Logit) for mapping binomial incidence • 51 QTL for 11 distinct tumor histotypes • 95% CI: 3.4 Mb (range 0.02 - 7.54) • Biologic effect size: 3.72% (range 0.75 - 7.46%) 2. CoxPH regression modeling for mapping genotype effects on latency • 39 QTL for tumor latency for 9 distinct tumor histotypes • 95% CI: 3.6 Mb (range 0.29 - 7.68) • Biologic effect size: 2.43% (range 1.04 - 5.79%) *All publicly available at https://github.com/elijahedmondson/HZE * Overall picture: many loci each contributing a small proportion to the variance. 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering Overview: QTL Mapping in HS mice
15. • Tumor incidence 1. Thymic PreT LSA — chr 4 2. Thyroid adenoma — chr 2 3. B cell LSA — chr 11 4. AML — chr 2 5. DLBCL LSA — chr 17 6. HCC — chr 2 7. B cell LSA — chr 3 8. HCC — chr 15 9. HCC — chr 8 10.Thyroid adenoma — chr 10 11.Thymic PreT LSA — chr 17 • Tumor Latency 1. B cell LSA — chr 7 2. Thymic PreT LSA — chr 4 3. B cell LSA — chr 1 14 Large Effect QTL (QTL explains > 5% of variance) 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
16. • Tumor incidence 1. Thymic PreT LSA — chr 4 2. Thyroid adenoma — chr 2 3. B cell LSA — chr 11 4. AML — chr 2 5. DLBCL LSA — chr 17 6. HCC — chr 2 7. B cell LSA — chr 3 8. HCC — chr 15 9. HCC — chr 8 10.Thyroid adenoma — chr 10 11.Thymic PreT LSA — chr 17 • Tumor Latency 1. B cell LSA — chr 7 2. Thymic PreT LSA — chr 4 3. B cell LSA — chr 1 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering 14 Large Effect QTL (QTL explains > 5% of variance)
17. Thyroid Follicular Adenoma 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
18. Thyroid Follicular Adenoma 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering Chromosome 2
19. Thyroid Follicular Adenoma: chr 2 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
20. Thyroid Follicular Adenoma: chr 2 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
21. Coat Color Brown Genome Scan Coat Color Albino Genome Scan Visualizing coincident QTL 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
22. Unsupervised Hierarchical Clustering of QTL: Coat Color Black Locus Tyrosinase Locus Dilution factor Locus 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering 99% confidence
23. Clustering occurs based on tumor histotype Clustering occurs based on radiation exposure vs. Unsupervised Hierarchical Clustering of Tumor QTL 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
24. Clustering occurs based on tumor histotype Clustering occurs based on radiation exposure vs. Unsupervised Hierarchical Clustering of Tumor QTL 1. HZE vs γ: are histotypes the same? 2. HZE vs γ: is tumorigenesis the same? 1. Overview 2. Thyroid Adenoma 3. QTL Clustering
25. Conclusions • Susceptibility QTL for tumors often overlaps between HZE and γ-ray irradiated populations—corroborated by clustering procedures *These findings support the current NASA risk model • 14 major effect QTL and 37 moderate effect QTL for 11 neoplasms have been mapped — many of which are novel — with a 95% confidence interval of 3.40 Mb
26. Colorado State University Mike Weil Lab Chrissy Fallgren Elvin Garcia Paula Genik Todd Bass Debra Kamstock Jackson Labs Daniel M. Gatti OHSU Ovidiu Dan Iancu John Belknap Jacob Raber Columbia Norman Kleiman International Mammalian Genome Society Funded by NASA grant NNX12AB54G Acknowledgements Elijah Edmondson, DVM DACVP elijah.edmondson@gmail.com