Hepatic lipogenesis combined with elevated endoplasmic reticulum (ER) stress is central to non-alcoholic steatohepatitis (NASH). However, the therapeutic targeting of key molecules is considerably less accomplished. Adeno-associated virus (AAV)-mediated gene therapies offer a new solution for various human ailments. Comprehensive bio-functional validation studies are essential to assess the impact of AAVs in the target organ for developing both preclinical and clinical gene therapy programs. Here, we have established a robust and efficient protocol for high-titer AAV production to enable detailed Selective ORgan Targeting (SORT) of AAV1, 5, 7, and 8 in vivo. Our results for in vivo SORT showed single organ (liver) targeting by AAV8, no organ targeting by AAV1, and dual organ transduction (liver-brain and liver-VAT) by AAV5 and AAV7. Using a human dataset and preclinical murine models of NASH, we identified an inverse correlation between ER stress-triggered CRELD2 and the de novo lipogenesis driver FASN. Furthermore, liver-specific silencing of CRELD2 via AAV8-shCreld2 strongly supports the contribution of CRELD2 to de novo lipogenesis through FASN regulation. Thus, our study demonstrates a robust method for producing clinically translatable AAVs that could be readily adapted for liver and/or liver-VAT or liver-brain targeted gene therapy.