Sensitized mutagenesis screen in Factor V Leiden mice identifies thrombosis suppressor loci Randal J. Westricka,b,c,1,2, Kärt Tombergc,d,1, Amy E. Sieberta,1, Guojing Zhuc, Mary E. Winne, Sarah L. Dobiesc, Sara L. Manningc, Marisa A. Brakea, Audrey C. Cleurenc, Linzi M. Hobbsa, Lena M. Mishacka, Alexander J. Johnstona, Emilee Kotnikc, David R. Siemieniakf, Jishu Xud, Jun Z. Lid, Thomas L. Saundersg, and David Ginsburgc,d,f,h,i,2 aDepartment of Biological Sciences, Oakland University, Rochester, MI 48309; bCenter for Data Science and Big Data Analysis, Oakland University, Rochester, MI 48309; cLife Sciences Institute, University of Michigan, Ann Arbor, MI 48109; dDepartment of Human Genetics, University of Michigan, Ann Arbor, MI 48109; eBioinformatics and Biostatistics Core, Van Andel Research Institute, Grand Rapids, MI 49503; fHoward Hughes Medical Institute, University of Michigan, Ann Arbor, MI 48109; gTransgenic Animal Model Core, University of Michigan, Ann Arbor, MI 48109; hDepartment of Internal Medicine, Ann Arbor, MI 48109; and iDepartment of Pediatrics, University of Michigan, Ann Arbor, MI 48109 Contributed by David Ginsburg, July 24, 2017 (sent for review April 7, 2017; reviewed by Monica J. Justice and Joost Meijers) L Factor V Leiden (F5L) is a common genetic risk factor for venous 13). However, <2% of F5 heterozygotes would be expected to thromboembolism in humans. We conducted a sensitized N-ethyl-N- coinherit a mutation at one or more of these loci, suggesting that a nitrosourea (ENU) mutagenesis screen for dominant thrombosuppres- large number of additional genetic risk factors for VTE and/or L sor genes based on perinatal lethal thrombosis in mice homozygous modifiers of F5 remain to be identified (3, 10). L L/L L for F5 (F5 ) and haploinsufficient for tissue factor pathway inhibitor Mice carrying the orthologous F5 mutation exhibit a mild to +/− L/L +/− (Tfpi ). F8 deficiency enhanced the survival of F5 Tfpi mice, moderate prothrombotic phenotype closely mimicking the human L/L +/− demonstrating that F5 Tfpi lethality is genetically suppressible. disorder (14). We previously reported a synthetic lethal interaction L/L L/+ +/− L L/L ENU-mutagenized F5 males and F5 Tfpi females were crossed between F5 homozygosity (F5 ) and hemizygosity for tissue factor L/L +/− +/− to generate 6,729 progeny, with 98 F5 Tfpi offspring surviving pathway inhibitor (Tfpi ) (15). Nearly all mice with this lethal L/L +/− until weaning. Sixteen lines, referred to as “modifier of Factor 5 Leiden genotype combination (F5 Tfpi ) succumb to widespread, sys- (MF5L1–16),” exhibited transmission of a putative thrombosuppressor temic thrombosis in the immediate perinatal period (15). GENETICS to subsequent generations. Linkage analysis in MF5L6 identified a N-ethyl-N-nitrosourea (ENU) mutagenesis in mice has been used chromosome 3 locus containing the tissue factor gene (F3). Although effectively to identify novel genes involved in a number of biological no ENU-induced F3 mutation was identified, haploinsufficiency for +/− L/L +/− processes (16, 17). ENU-induced germline mutations transmitted F3 (F3 ) suppressed F5 Tfpi lethality. Whole-exome sequenc- from a mutagenized male mouse (G0) occur at ∼1.5 mutations per ing in MF5L12 identified an Actr2 gene point mutation (p.R258G) as megabase, at least 50-fold higher than the endogenous background the sole candidate. Inheritance of this variant is associated with sup- mutation rate (18). Several previous reports have successfully ap- F5L/L Tfpi+/− P = × −6 pression of lethality ( 1.7 10 ), suggesting that plied an existing phenotype as a sensitizer to identify modifier genes. Actr2p.R258G is thrombosuppressive. CRISPR/Cas9 experiments to A dominant suppressor screen in Mecp2-deficient mice (Rett syn- Actr2 generate an independent knockin/knockout demonstrated drome) identified a mutationinsqualeneepoxidase(Sqle)asa Actr2 that haploinsufficiency is lethal, supporting a hypomorphic heritable suppressor, resulting in prolonged survival and ameliora- Actr2p.R258G or gain-of-function mechanism of action for . Our findings tion of neurologic manifestations (19). Other successful sensitized identify F8 and the Tfpi/F3 axis as key regulators in determining L screens include analysis of mouse mutants predisposed to dia- thrombosis balance in the setting of F5 and also suggest a role for Sox10 Actr2 betic nephropathy (20), a screen in haploinsufficient in this process. mice identifying the Gli3 gene as a modifier of neurocristopathy venous thromboembolism | Factor V Leiden | ENU mutagenesis | tissue factor pathway inhibitor | genetic screen Significance Venous thromboembolism (VTE) is a common disease characterized enous thromboembolism (VTE) is a common disease that af- by the formation of inappropriate blood clots. Inheritance of specific fects 1–3 per 1,000 individuals each year (1). VTE susceptibility V genetic variants, such as the Factor V Leiden polymorphism, increases exhibits a complex etiology involving contributions of both genes and ∼ ≈ VTE susceptibility. However, only 10% of people inheriting Factor V environment. Genetic risk factors explain 60% of the overall risk Leiden develop VTE, suggesting the involvement of other genes that for VTE (2). Recent large-scale genome-wide association studies ABO F2 F5 F11 FGG PROCR are currently unknown. By inducing random genetic mutations into (GWAS) confirmed , , , ,and as mice with a genetic predisposition to VTE, we identified two genomic thrombosis susceptibility genes, with only two additional novel loci, regions that reduce VTE susceptibility. The first includes the gene for TSPAN15 and SLC44A2, identified (3–6), leaving the major com- blood coagulation, Factor 3, and its role was confirmed by analyzing ponent of VTE genetic risk still unexplained. L mice with an independent mutation in this gene. The second contains The Factor V Leiden variant (F5 )isacommoninheritedrisk a mutation in the Actr2 gene. These findings identify critical genes for factor for VTE with an average allele frequency of 3.5% in the Eu- the regulation of blood-clotting risk. L ropean population (7–9). F5 is estimated to account for up to 25% of the genetically attributable thrombosis risk in this population (7). Author contributions: R.J.W., K.T., A.E.S., and D.G. designed research; R.J.W., K.T., A.E.S., L However, penetrance is incomplete, with only ∼10% of F5 hetero- G.Z., M.E.W., S.L.D., S.L.M., M.A.B., A.C.C., L.M.H., L.M.M., A.J.J., E.K., D.R.S., and T.L.S. performed research; R.J.W., K.T., A.E.S., G.Z., M.E.W., S.L.D., A.C.C., E.K., D.R.S., J.X., J.Z.L., zygotes developing thrombosis in their lifetimes. The severity of T.L.S., and D.G. analyzed data; and R.J.W., K.T., A.E.S., and D.G. wrote the paper. thrombosis also varies widely among affected individuals (10), limiting Reviewers: M.J.J., The Hospital for Sick Children; and J.M., Sanquin Research. F5L the clinical utility of genotyping in the management of VTE (11). The authors declare no conflict of interest. The incomplete penetrance and variable expressivity of throm- 1 F5L R.J.W., K.T., and A.E.S. contributed equally to this work. bosis among patients can at least partially be explained by ge- 2 F5L To whom correspondence may be addressed. Email: [email protected] or rjwestrick@ netic interactions between and other known thrombotic risk oakland.edu. factors such as hemizygosity for antithrombin III or proteins C or S, This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. as well as the common prothrombin 20210 polymorphism (10, 12, 1073/pnas.1705762114/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1705762114 PNAS Early Edition | 1of6 Downloaded by guest on October 1, 2021 L/L (21), and identification of a mutation in the c-Myb gene as a depicted in Fig. 1B. ENU-mutagenized G0 F5 males were crossed L/+ +/− dominant modifier for platelet count in Mpl-deficient mice to F5 Tfpi females to generate G1 mice, which were screened L +/− (congenital thrombocytopenia) (22). We now report the results by genotyping at weaning for F5 and Tfpi . Previously described of a dominant, sensitized ENU mutagenesis screen for suppressors visible dominant mutant phenotypes (24), including belly spotting L/L +/− of F5 Tfpi -dependent lethal thrombosis. and skeletal abnormalities, were observed in ≈5.9% of G1 offspring, similar to the ∼4.2% rate of observable mutants in previous studies Results and Discussion (24). This is consistent with the ∼20–30 functionally significant L/L +/− F8 Deficiency Suppresses F5 Tfpi Lethality. X-linked hemophilia mutations per G1 mouse expected with this ENU mutagenesis A results in a moderate-to-severe bleeding disorder in humans and protocol (25). Although 25% of G1 embryos from this cross are L/L L/L +/− is caused by mutations in the F8 gene. To test whether the F5 expected to carry the synthetic lethal F5 Tfpi genotype, most +/− Tfpi lethal thrombotic phenotype is suppressible by hemophilia A are lost at birth. Given a total of 6,631 G1s for the other three F5L/+ Tfpi+/− F8+/− genotypes observed at weaning (∼1/3 for each genotype), a similar in mice, triple-heterozygous female mice were L/L +/− L/L number of F5 Tfpi G1 conceptuses, ∼2,210 (6,631 ÷ 3), would generated and crossed to F5 male mice (Fig. 1A). One quarter of L/L +/− L/L +/− F5 Tfpi conceptuses are expected to carry the F5 Tfpi genotype, with have been expected. The 98 live mice (45 females, 53 males) thus represented 4.4% of the number expected with this half of the total expected male conceptuses completely F8 deficient L/L +/− − F5 Tfpi (F8 ). Thus, 1/16th of the overall offspring from this mating are genotype. Survival data were collected for 57 of the L/L +/− − G1 mice, 34 of which lived past 70 d of age; precise dates of death expected to be F5 Tfpi F8 males.