CRISPR/Cas9 Inducible Gene Knockout in Brown Adipose Tissue

Abstract

The ability to study adipose tissue in vivo has centered around the use of transgenic mice, which require significant time, resources, and investment to generate. With nearly 40% of adults in the United States considered obese, efficient strategies to improve the way in which adipose tissue is investigated are paramount. Recently, viral or nonviral delivery of CRISPR/Cas9 for inducible gene knockout in somatic tissues has been used to efficiently model disease and study gene function. Its use in adipose tissue, however, has not been attempted owing to an inability to target adipocytes in vivo. My dissertation work sought to identify novel delivery strategies that would enable CRISPR/Cas9 inducible gene knockout in adipose tissue. To accomplish this, I explored the use of adeno-associated viruses (AAVs) and lipid nanoparticles (LNPs) for targeting adipose tissue in vivo. I demonstrated that vector and route of administration are the two key factors that govern adipocyte transduction efficiency. Local delivery of AAV serotype 8 was shown to maximize adipocyte transduction specifically in brown adipose tissue (BAT). I thus applied this delivery strategy to attempt CRISPR/Cas9 inducible gene knockout in BAT of adult mice. I developed the Brown Adipocyte CRISPR (BAd-CRISPR) method, in which AAV8 is used to deliver a single guide RNA (sgRNA) to mice expressing Cas9. I generated a brown adipocyte Cas9-expressing mouse line and developed a cloning pipeline that enabled sgRNAs targeting any gene to be packaged into AAV8 and locally injected to BAT. BAd-CRISPR allowed for rapid interrogation of one or multiple genes and was used to knockout adiponectin (ADIPOQ), adipose triglyceride lipase (ATGL), fatty acid synthase (FASN), perilipin 1 (PLIN1), and stearoyl-CoA desaturase 1 (SCD1) specifically in BAT of adult mice. Importantly, BAd-CRISPR did not result in accumulation of substantial off-target mutations. These studies demonstrated that the BAd-CRISPR method can be used to inducibly knockout genes in BAT and that transgenic mice can be generated in 1-2 months. BAd-CRISPR was then used to generate the first inducible uncoupling protein 1 (UCP1) knockout mouse and assess the effects of UCP1 loss on adaptive thermogenesis in adult mice. Inducible UCP1 knockout mice defend core body temperature at 20-21ºC and at 5ºC despite complete loss of UCP1 expression. Interestingly, inducible UCP1 knockout led to an upregulation of genes involved in peroxisomal lipid oxidation and protein synthesis and turnover, and a decrease in genes involved in mitochondrial metabolism, suggesting a compensatory mechanism to maintain adaptive thermogenesis. Ultimately, my doctoral work produced an efficient method to streamline the paths to discovery in BAT. This work demonstrated that CRISPR/Cas9 can be harnessed to generate BAT specific gene knockout models that require less financial investment and substantially reduced time compared to traditional transgenic approaches. Given the utility of this method, it is expected that with further optimization, BAd-CRISPR will be applicable not only to white and marrow adipose tissues, but also adapted for use in other tissues as well.