UC Berkeley

Abstract

Dysfunction of retinoid metabolism affects development of obesity and metabolic disorders.

By

Di Yang

Doctor of Philosophy in Metabolic Biology

University of California, Berkeley

Professor Joseph L. Napoli. Chair

All-trans-retinoic acid (atRA) controls adiposity by affecting adipogenesis, lipolysis, fatty acid oxidation and non-shivering thermogenesis. The availability of atRA is controlled precisely by biosynthesis as well as degradation. Multiple retinol dehydrogenases (Rdh) catalyze the first, regulated and rate-limiting step of atRA biosynthesis—conversion of retinol into retinal. Enzymes involved in this process are, but not limited to, Rdh10, Rdh1, Dhrs9 and Rdhe2. Increasing evidence suggests each enzyme may support different biological functions. RDH10Rdh10, the most highly expressed retinol dehydrogenase in both stem cells and many adult tissues, has been reported as a key atRA synthesizing enzyme during embryogenesis. However, the role of RDH10Rdh10 in adult tissues, especially in the context of metabolic disorders has not been defined. Our data suggests that a heterozygous hypomorphic RDH10Rdh10 mutation (HYPO) in mouse embryonic fibroblasts (MEF) decreases the amount of atRA synthesized and leads to increased adipogenesis. This phenotype can be reversed by atRA supplementation in adipocyte induction medium of MEF. HYPO MEF demonstrates an impaired response to retinol compared to WT when treated with graded amounts of retinol as a result of reduced atRA synthesis efficiency. Rdh10+/- mice fed a high-fat diet display an adiposity phenotype along with metabolic disorders such as glucose intolerance, hyperinsulimia, insulin resistance, liver steatosis (males only) and bone marrow adipocyte proliferation (females only). More interestingly, even though both male and female Rdh10+/- fed HFD have increased body weight, they demonstrate different phenotypes in different organs. Rdh10+/- males have increased liver steatosis compared to WT, while no significant differences were detected in females. Female Rdh10+/- mice have increased numbers of bone marrow adipocytes compared to WT, whether fed a HFD or a low-fat diet (LFD). RA supplementation with 5 mg-90 day release pellets (~56 µg/day) by subcutaneous implantation reversed the obesity phenotype in males. Taken together, these data indicate that Rdh10 is an important atRA generating enzyme that controls adipose tissue development and function.

The second step of atRA biosynthesis converts retinaldehyde into atRA by retinaldehyde dehydrogenases such as RALDH1, RALDH2 and RALDH3. RALDH1, the most highly-expressed retinaldehyde dehydrogenase in MEF and adipose tissue, affects adipogenesis and adiposity in an opposite way as RDH10. RALDH1 KO mice are resistant to diet induced obesity (DIO) and Pparγ-induced osteoporosis. A hypothesis by the Plutzky group claims that resistance to DIO of RALDH1 KO was the result of accumulation of retinaldehyde in tissues (Ziouzenkova et al., 2007a). However, tissue retinaldehyde measurements done by our group by a specific and sensitive LC/MS/MS assay showed no differences in retinaldehyde levels in 6-7 week old mice, with significant differences in body weight at that age. Retinaldehyde levels were lower in 26-week-old HFD fed Raldh1 KO mice compared to WT, i.e. after the phenotype was established, which indicated it was the result rather than the cause of the phenotype. Also, treating Raldh1 KO MEFs with RAR or PPARβ/δ antagonists or agonists did not change the phenotype, further suggesting resistance to DIO of Raldh1 KO mice is independent of retinoid function. Further study is needed to elucidate the mechanism by which Raldh1 KO mice resist DIO.