A Missense Mutation in the Loss-of-Function Intolerant Actr2 Gene Suppresses Thrombosis Via Multiple Mechanisms

Susceptibility to venous thromboembolism (VTE) is increased by inheritance of Factor V Leiden (F5^L). However, only ~10% of people inheriting F5^L develop VTE, illuminating the need to identify genes that modify VTE risk. We recently reported a dominant, p.R258G missense mutation in the Actr2 gene (Actr2^+/G) as a major thrombosuppressor of perinatal lethal thrombosis in mice homozygous for F5^L (F5^L/L) and hemizygous for tissue factor pathway inhibitor (Tfpi^+/-). The Actr2 gene encodes the Actin-Related Protein 2 (ARP2), an essential member of the Arp2/3 complex. Arp2/3 is ubiquitously expressed and is necessary for actin cytoskeletal regulation in all eukaryotic cells, including platelets.

We recently attempted to generate independent Actr2^Gand Actr2^- (knockout)alleles in mice using CRISPR/Cas9 genome editing. The failure to produce any mice carrying the Actr2^- allele suggested that even hemizygosity (Actr2^+/-) is incompatible with embryonic survival. To further investigate this unexpected finding, ES cells (Clone: EPD0727_2_H12) containing the targeted Actr2^{tm1a(KOMP)Wtsi} "Knockout First" allele were obtained. Transfer of 79 injected blastocysts into foster mothers yielded five males with 20-90% chimerism. Successful germline transmission occurred from only the 20% chimera. Actr2 locus targeting was confirmed by 5' and 3' long range PCR from tail DNA in the four non-transmitting chimeras. However, long range PCR analysis of the 20% chimera and its progeny revealed only mistargeting outside of the Actr2 locus. This demonstrates that the KOMP ES cells were heterogeneous for correct/incorrect Actr2 targeting and confirms that a 50% reduction of Actr2 function is incompatible with survival.

To characterize the Actr2^G mice, heterozygous Actr2^+/G mice were backcrossed several generations to the wildtype C57BL6/J background and Actr2^+/G x Actr2^+/G crosses were performed. Analysis of 303 progeny revealed significant departures from expected Mendelian frequencies as only 36 Actr2^G/G mice were obtained (p=0.005). Also, all surviving Actr2^G/G mice developed progressive kyphosis and had significantly reduced (~six month) lifespan (N=303; p <0.0001). In contrast, Actr2^+/G mice were born at the expected frequency and were morphologically indistinguishable from wildtype, demonstrating that the Actr2^G allele retains sufficient functionality for normal survival of Actr2^+/G mice.

To investigate the ability of Actr2^G to alter coagulation factor expression, we performed RT-qPCR on whole liver samples derived from Actr2^+/G and Actr2^+/+ mice (N=6). While we found no significant alteration in F5, Tfpi or Actr2 expression, we observed significant downregulation in F10 and dramatic upregulation in the expression of Actb and the thrombin inhibitor Serpine2 relative to wildtype. Liver ARP2 protein levels were unaffected. Interestingly, when investigating whether these effects were the result of loss of ARP2 function, we found that treatment of Neuro-2a cells with the Arp2/3 complex inhibitor CK-666 resulted in significant downregulation in Serpine2 and Actb expression without affecting Actr2 expression or ARP2 protein levels. Analysis of platelet ARP2 levels revealed a significant 50% reduction in comparison to wildtype. Functional analysis of Actr2^+/+ (n=7) and Actr2^+/G (n=10) platelets revealed a decrease in the ability of the mutant platelets to aggregate upon collagen stimulation in whole blood (p=0.0431) with no significant alteration in platelet number between genotypes. Further, we found that Actr2^+/G whole blood had a significantly decreased ability to form a thrombus on collagen at flow rates of 1500s^-1 with occlusion speeds of 22.5 ± 2.7 and 7.7 ± 3.0 kPA/min for Actr2^+/+ and Actr2^+/G mice respectively (Actr2^+/+= Actr2^+/G=3; p=0.0215).

In summary, our attempts to generate Actr2^- mice illustrate the importance of meticulously screening for correct/incorrect vector incorporation in gene targeting experiments. This is especially relevant for loss-of-function intolerant genes such as Actr2, because targeting these genes results in negative selection. Our analyses of coagulation factor expression and platelet function suggest that Actr2^G operates as a hypomorphic and/or gain-of-function mutation to suppress thrombosis via multiple mechanisms. These findings illustrate the particular importance of Actr2^G in maintaining hemostatic balance.