In insulin resistant animals, hyperinsulinemia is associated with the elevated hepatic gluconeogenesis and lipogenesis, which should have been respectively suppressed and stimulated by insulin 
. Our lab has shown that retinoids regulate the expression levels of genes involved in hepatic glucose and fatty acid metabolism. These regulations contributed to the modulation of the insulin-mediated expression of these genes in primary hepatocytes 
. The implication of these observations is that the dynamic regulation of RA production may be responsible for the hepatic insulin resistance.
To confirm that the treatments of ROL and RAL can regulate gene expression probably through the production of RA, we analyzed the expression levels of Cyp26a1
, an RA responsive gene 
, in primary hepatocytes (). Retinoids induced the expression of Cyp26a1
in a dose- and time-dependent manner. This induction can be blocked in the presence of RAR antagonist, indicating the requirement of RAR activation. This result and the fact that RAL induced the Cyp26a1
expression indicate that RA production occurs.
To find the cultured cell lines that can be used as tools for further studies of the retinoid catabolism, we have analyzed the retinoid-mediated gene expression in HL1C rat hepatoma cells (). ROL, RAL, and RAL also induced the expression levels of Cyp26a1 mRNA in HL1C cells in a dose- and time-dependent manner, which is similar to the results obtained in primary hepatocytes. This induction is mediated by the activation of RAR and RXR. The induction of Cyp26a1 mRNA results in the elevated levels of CYP26A1 proteins in HL1C cells. However, the elevated levels of CYP26A1 protein are not as high as that of Cyp26a1 mRNA at 12 and 24 hours (), indicating the potential roles of protein translation and/or stability in the determination of the CYP26A1 protein levels in HL1C cells. The maximal induction folds by reintoids are higher in primary hepatocytes than in HL1C cells. ROL and RAL respectively induce the expression of Cyp26a1 mRNA at 2 and 0.02 µM in primary hepatocytes and at 20 and 2 µM in HL1C cells. These results indicate that primary hepatocytes seem to be more readily to convert ROL to RAL and then to RA. Alternatively, HL1C cells are more readily to convert RAL to ROL so that less RA is generated. Whether it is true or not deserves further investigation. It also indicates that HL1C cells can be used as a tool to study the metabolism of retinoids and its effects on the regulation of the expression of RA responsive genes.
It has been shown that the short chain dehydrogenase/reductase family 16C member 5 (SDR16C5), RDH2, RDH10, RALDH1-4 are expressed in the liver 
. Since Raldh4
mRNA (encoding RALDH4) is for the production of 9-cis
, we focused on Rdh2
. Their mRNA levels in both freshly isolated and cultured primary hepatocytes of ZL and ZF rats were compared. The mRNA level of Raldh1
, but not that of Rdh2
, is higher in ZF than that in ZL rat hepatocytes. In addition, the hepatic RALDH1 protein level is also higher in ZF rats than in ZL rats (). The elevated expression levels of Raldh1
mRNA and RALDH1 protein levels may be resposible for the altered hepatic lipogenesis in ZF rats. To test the hypothesis, we made recombinant adenovirus Ad-Raldh1 to over-express RALDH1 protein in INS-1 rat insulinoma cells. We have shown previously 
and here () that RA, but not RAL, induced Srebp-1c
expression in INS-1 cells. It seems that INS-1 cells lack the enzymatic activities converting RAL to RA. This makes INS-1 cells an ideal tool to test our hypothesis that the elevated expression of RALDH1 in ZF rats causes the increase of RA production, and in turn, the induction of RA responsive gene expression. After the over-expression of Raldh1
mRNA and RALDH1 protein in INS-1 cells, RAL started to induce the Srebp-1c
expression as RA did. This indicates that RAL was oxidized to RA by RALDH1 before the induction of Srebp-1c
. The elevated levels of hepatic Raldh1
mRNA and RALDH1 protein in ZF rats may contribute to the regulation of hepatic lipogenesis.
It has been shown that the mRNA level of Raldh1
is elevated in the kidney of db/db
. Both db/db
mice and ZF rats develop obesity due to the mutation of leptin receptor 
. It has been shown that ICR mice fed high cholesterol diet have the elevated expression levels of Raldh1
. This is attributed to the oxysterol-induced expression of SREBP-1c which directly binds to the sterol regulatory response elements (SREs) at the proximal promoter regions of Raldh1
, and induces their expression 
. Since the expression of Srebp-1c
is induced by insulin 
and RA 
in primary hepatocytes, the regulation of Raldh1
expression becomes a converge point for a feed-forward mechanism by which VA status regulates lipogenesis in the liver. It is reasonable to hypothesiz that in hepatocytes of ZL rats, RA derived from ROL synergizes with insulin to induce the expression of Srebp-1c
mRNA (). The mature SREBP-1c protein supports the expression of RALDH1, which promotes the production of RA to maintain the homeostasis of Srebp-1c
expression in the liver of ZL rats (). ZF rats are hyperphagic due to the defect of leptin receptor, which leads to the development of obesity, insulin resistance, hyperinsulinemia and hyperlipidemia 
. We have shown the ZF rat hepatocytes have elevated expression levels of Srebp-1c
. We think that the excessive supply of dietary VA caused by hyperphagia and hyperinsulinemia in ZF rats probably work together to trigger an elevation of RA production and an induction of Srebp-1c
expression in their hepatocytes, respectively. The elevation of the mature SREBP-1c protein induces the expression of Raldh1
mRNA, which leads to more RA production, and further enhances the expression levels of Srebp-1c
and its down-stream lipogenic genes. This creates a feed-forward mechanism by which the Srebp-1c
expression is maintained at a higher level in the liver of ZF rats (). Whether the proposed feed-forward mechanism is true or not, and what mechanism is responsible for the up-regulation of Raldh1
expression in the ZF rat liver deserve further investigation.
The hypothesized role of the RA production in the feed-forward induction of the expression of Srebp-1c and its downstream lipogenic genes in the liver of ZL (A) and ZF (B) rats.
In summary, the results shown here suggest that ROL and RAL are metabolized into RA to regulate gene expression in rat primary hepatocytes and hepatoma cells. In primary rat hepatocytes, the responsible enzymes are most likely RDH2 and RALDH1. The expression of Raldh1 mRNA is higher in primary hepatocytes from ZF rats than that from ZL rats, which leads to the elevated RALDH1 protein levels in the liver of ZF rats. Over-expression of RALDH1 introduces the RAL-mediated induction of Srebp-1c in INS-1 cells. Thus, we hypothesize that the change of RA production from the over-supply of dietary VA due to the hyperphagia of ZF rats results in higher Srebp-1c expression in ZF hepatocytes. The elevated SREBP-1c expression can further induce Raldh1 expression to create a feed-forward mechanism that could be one of the reasons responsible for the increased lipogenesis in the liver of ZF rats.