Cytosolic retinoid dehydrogenases govern ubiquitous metabolism of retinol to retinaldehyde followed by tissue-specific metabolism to retinoic acid.

The ability of vitamin A (retinol) to control growth and development depends upon tissue-specific metabolism of retinol to retinoic acid (RA). RA then functions as a ligand for retinoid receptor signaling. Mouse genetic studies support a role for cytosolic alcohol dehydrogenases (ADH) in the first step (oxidation of retinol to retinaldehyde) and a role for cytosolic retinaldehyde dehydrogenases (RALDH) in the second step (oxidation of retinaldehyde to RA). Mice lacking ADH3 have reduced survival and a growth defect that can be rescued by dietary retinol supplementation, whereas the effect of a loss of ADH1 or ADH4 is noticed only in mice subjected to vitamin A excess or deficiency, respectively. Also, genetic deficiency of both ADH1 and ADH4 does not have additive effects, verifying separate roles for these enzymes in retinoid metabolism. As for the second step of RA synthesis, a null mutation of RALDH2 is embryonic lethal, eliminating most mesodermal RA synthesis, whereas loss of RALDH1 eliminates RA synthesis only in the embryonic dorsal retina with no obvious effect on development. Analysis of RA-rescued RALDH2 mutants has also revealed that RALDH3 and at least one additional enzyme produce RA tissue-specifically in embryos. Collectively, these genetic findings indicate that metabolism of retinol to retinaldehyde is not tissue-restricted as it is catalyzed by ubiquitously-expressed ADH3 (a low activity form) as well as by tissue-specifically expressed ADH1 and ADH4 (high activity forms). In contrast, further metabolism of retinaldehyde to RA is tissue-restricted as all enzymes identified are tissue-specific. An important concept to emerge is that selective expression of enzymes catalyzing the second step is what limits the tissues that can completely metabolize retinol to RA to initiate retinoid signaling.