Selective Proteolytic Processing of Rat Hepatic Sterol Regulatory Element Binding Protein-1 (SREBP-1) and SREBP–2 During Postnatal Development*

Sterol regulatory element-binding protein-1c (SREBP-1c) plays a major role in hepatic lipogenic gene expression. In adult animals, insulin and oxysterols induce SREBP-1c gene transcription, whereas polyunsaturated fatty acids suppress the nuclear content of SREBP-1c through pre-translational regulatory mechanisms. A decline in nuclear SREBP-1 is associated with suppression of hepatic lipogenesis. In contrast to adult rats, hepatic lipogenesis in preweaned neonatal rats is low. Ingestion of milk fat by the neonate may contribute to low hepatic lipogenesis. In this report, we tested the hypothesis that low lipogenic gene expression prior to weaning correlates with low mRNASREBP-1c, as well as low precursor and nuclear forms of SREBP-1. In contrast to expectations, levels of mRNASREBP-1c and the 125-kDa SREBP-1 precursor in livers of preweaned rats was comparable with adult levels. Despite high levels of SREBP-1 precursor, mature (65 kDa) SREBP-1 was not detected in rat liver nuclei prior to 18 days postpartum. Weaning rats at 21 days postpartum was accompanied by a rise in nuclear SREBP-1 levels as well as increased lipogenic gene expression. In contrast, SREBP-2 was present in rat liver nuclei, and its target gene,HMG-CoA reductase, was expressed above adult levels prior to weaning. These studies indicate that, prior to weaning, SREBP-2 but not SREBP-1 is proteolytically processed to the mature form. As such, SREBP-2-regulated genes are active. Failure of SREBP-1 to be processed to the mature form <18 days postpartum correlates with low hepatic lipogenic gene expression. This mechanism differs from the hormonal and fatty acid-mediated pre-translational control of SREBP-1c in adult liver.

suppress transcription of hepatic lipogenic genes. One of the key transcription factors controlling hepatic de novo lipogenesis is sterol regulatory element-binding protein-1c (SREBP-1c) (3,4). SREBP-1c is a member of a family of basic helix-loophelix leucine zipper transcription factors involved in fatty acid, triglyceride, and cholesterol synthesis.
SREBPs are translated as ϳ125-kDa precursors (pSREBP) attached to the endoplasmic reticulum (ER) (4,5). After proteolytic processing in the Golgi, the nuclear form, nSREBP (ϳ65 kDa), accumulates in nuclei where it binds sterol regulatory elements in promoters of many genes involved in fatty acid, triglyceride, and cholesterol synthesis. The proteolytic processing of SREBP is mediated by at least three proteins, i.e. SREBP-cleavage activating protein (SCAP), site-1 protease (S1P), and site-2 protease (S2P) (4). The escort of SREBP from the ER to the Golgi by SCAP is inhibited by the accumulation of sterols in cells, thus preventing maturation of SREBP to a form regulating gene transcription. Germline modification of mice has shown that SCAP and S1P are important for the processing of both SREBP-1 and SREBP-2 (4, 6, 7).
Unsaturated fatty acids have been reported to suppress nuclear SREBP levels in HepG2 cell, rat primary hepatocytes, and rat liver (16 -21). However, in the liver, nuclear levels of SREBP-1c but not of SREBP-2 are suppressed by PUFAs (18 -21). Feeding rodents diets supplemented with polyunsaturated fatty acids or treating primary hepatocytes with PUFA will suppress mRNA SREBP-1c and lead to a decline in both the precursor and nuclear forms of SREBP-1c (18 -20). Clarke and co-workers (20) have reported that the principal mechanism for the pre-translational control involves PUFA-enhanced mRNA SREBP-1c turnover. Overexpression of nSREBP-1c in pri-mary hepatocytes or in vivo eliminates the PUFA effects on several lipogenic genes, indicating that SREBP-1c is a key target for PUFA suppression of de novo lipogenesis (19,21).
During postnatal development of rats and mice, hepatic lipogenesis is low prior to weaning at 21 days postpartum (22)(23)(24)(25)(26)(27)(28)(29). The activities of key enzymes, as well as their mRNAs, are very low during the suckling phase. Where examined, low lipogenic gene expression is due to low transcription rates (24). Based on studies with adult animals (2), low lipogenic gene transcription in newborns has been attributed to the ingestion of a high fat milk diet (26).
Because PUFA controls nuclear SREBP-1c levels through a pre-translational regulatory mechanism (18 -20), we were interested in determining whether low hepatic lipogenic gene expression in preweaned animals correlates with low mRNA SREBP-1c . Contrary to expectations, our studies indicated that both mRNA SREBP-1 and pSREBP-1c are well expressed in rat liver prior to weaning. Only nSREBP-1c levels are low in neonatal liver, reflecting abrogated maturation of pSREBP-1 to nSREBP-1.

Animals-Female
Sprague-Dawley rats with litters were obtained from Charles River Laboratories (Kalamazoo, MI) and maintained on a Tek-Lad chow diet, ad libitum. Male rats at 15, 18, and 21 days postpartum were used for this analysis and compared with adult male rats (Ն30 days postpartum).

RESULTS AND DISCUSSION
Post-translational Processing of SREBP-1c Regulates Nuclear SREBP-1c Levels during Postnatal Development-In this report, we tested the hypothesis that low levels of hepatic lipogenesis prior to weaning correlate with low mRNA SREBP-1c , which leads to a decline in pSREBP-1 and nSREBP-1. Levels of mRNA encoding SREBP-1 and the lipogenesis-associated protein S14 were measured in 15-and 42-day-old rats and 42-day old rats, respectively, fed a high carbohydrate-fat free diet for 5 days (Fig. 1). S14 was used because SREBP-1c induces S14 gene transcription (19,31). The high carbohydrate fat-free diet was used to illustrate the effect of a fat-free diet on hepatic SREBP-1 and S14 mRNA levels. Hepatic mRNA S14 is essentially absent at 15 days postpartum ( Fig. 1). At this age, S14 gene transcription is not detectable (24). In contrast to mRNA S14 , hepatic mRNA SREBP-1 in 15 day old rats is ϳ1.5-fold higher than that seen in 42-day-old chow-fed rats and is comparable with the level seen in rats fed a high carbohydrate fat-free diet for 5 days. RT-PCR analysis using SREBP-1a-and SREBP-1c-specific primers indicated that SREBP-1c represents Ͼ90% of the SREBP-1 expressed in rat liver at 15 days postpartum (not shown). The ratio of SREBP-1c to SREBP-1a was not different at 15 days postpartum and in adults. Thus, SREBP-1c is the predominant SREBP-1 transcript expressed in neonatal and adult liver.
Insulin-mediated induction of S14 and SREBP-1 gene transcription accounts for the elevated levels of S14 and SREBP-1c mRNAs in livers of adult animals fed high carbohydrate fatfree diets (12-14, 31, 32). However, high SREBP-1c mRNA levels prior to weaning cannot be ascribed to elevated insulin, because blood insulin levels are typically low, and the liver displays elevated ketogenesis (26,28,29). Thus, factors controlling hepatic SREBP-1c mRNA levels in the suckling animal differ from that seen in the adult rat.
Because low lipogenic gene expression cannot be explained by pre-translational suppression of mRNA SREBP-1c , we examined hepatic precursor and nuclear SREBP-1 and SREBP-2 FIG. 1. Expression of S14 and SREBP-1 mRNA in rat liver. RNA was extracted from five suckling rats (15 days postpartum), two chowfed rats, and three rats fed a high carbohydrate (HiCHO) fat-free diet for 7 days. The chow and high carbohydrate fat-free fed animals were 42 days postpartum. RNAs were separated electrophoretically, transferred to nitrocellulose, and probed with [ 32 P]cDNAs for S14 and SREBP-1c. Levels of expression were quantified by PhosphorImager analysis. The results are representative of at least two separate studies. levels in 15-and 30-day-old rats (Fig. 2). Precursor levels of SREBP-1 and SREBP-2 were ϳ1.5and 2-fold higher in livers of 15-day old rats when compared with adults. As a control, the microsomal monooxygenase CYP4A was ϳ2-fold higher in livers derived from 30-day-old animals than in those from 15-dayold animals. Although nSREBP-2 was present in nuclear extracts obtained from both 15-and 30-day-old animals, nSREBP-1 was not detected in hepatic nuclei isolated from 15-day-old rats. However, nSREBP-1 was present in nuclear extracts from 30-day-old animals. HNF-4␣ levels were ϳ2-fold higher in 30-versus 15-day-old rats.
These studies indicate that SREBP-2 matures to the nuclear form, but SREBP-1 maturation is abrogated, leading to little or no nSREBP-1 accumulation in hepatic nuclei of 15-day-old rats. The absence of nSREBP-1c in nuclei correlates with low hepatic de novo lipogenesis in the neonate. Thus, the principal mechanism accounting for low lipogenic gene expression in hepatic nuclei prior to weaning is due, at least in part, to low nSREBP-1 but not low mRNA SREBP-1 or pSREBP-1. The mechanism controlling hepatic nSREBP-1 levels in neonatal and adult liver is clearly different.
Developmental Regulation of SREBP-1c Maturation Correlates with the Induction of Hepatic Lipogenic Gene Expression-Lipogenic gene expression, i.e. S14 and FAS, increases dramatically when rats are weaned (22)(23)(24)(25). To determine whether SREBP-1 maturation followed this same time line, we examined SREBP-1 protein levels in hepatic microsomes and nuclei of animals at 15, 18, 22, and 30 days postpartum. Although these animals are normally weaned at 21 days of age, they begin to ingest solid food between 18 and 21 days postpartum. Microsomal pSREBP-1 remained unchanged over the 15-, 18-, 22-, and 30-day-old period (Fig. 3), a finding that correlates with the modest changes in mRNA SREBP-1c in 15day-old and adult animals (Fig. 1). Hepatic nuclear nSREBP-1 was not detected at 15 days postpartum (Figs. 2 and 3) but was detected at 18 days postpartum. By 22 days of age, nSREBP-1 levels are comparable with adult levels. The maturation of SREBP-1c parallels the dietary switch from a high fat milk diet to a control chow diet.
To determine whether the change in nSREBP-1 correlated with the onset of lipogenic gene expression, we measured the expression of three SREBP-1 target genes (S14, FAS, and GPAT) as well as the SREBP-2 regulated gene HMG-CoA reductase (Fig. 4A). The mRNAs encoding S14 and FAS increased progressively from very low levels at 15 days of age to high levels at 30 days of age. These changes correlated with increased nSREBP-1 levels (Fig. 3). In contrast, glycerophos- phate acyl transferase (GPAT) mRNA levels were above adult levels at 15-22 days postpartum. Unlike de novo lipogenesis, GPAT is required for the synthesis of phospholipids as well as triglycerides. The growing liver likely requires GPAT expression throughout all phases of development. Expression of the SREBP-2-regulated transcript, HMG-CoA reductase, was also above adult values during the 15-22 day old period, a finding consistent with abundant nSREBP-2 (Fig. 2) and elevated cholesterol synthesis in neonatal liver (28).
The mRNA encoding mtHMG-CoA synthase, a PPAR␣regulated gene, is high prior to weaning (Fig. 4B). High levels of mtHMG-CoA synthase are consistent with elevated ketogenesis associated with suckling rats (26,28). In contrast, another PPAR␣-regulated gene, CYP4A, is low prior to weaning. This apparent paradox is explained by the fact that low blood levels of insulin prior to weaning promote ketogenesis (26,28,29). Longchain PUFAs are PPAR␣ ligands and induce CYP4A gene transcription (33,34). Milk fats are enriched in short to medium chain saturated fatty acids. Based on structural studies, these fatty acids are likely not good ligands for PPAR␣ (23). The LXRregulated transcripts (35), CYP7A (bile acid synthesis), and ABCG5 (cholesterol efflux) are well expressed throughout the 15-30-day postpartum period. Thus, the expression of genes encoding proteins involved in cholesterol synthesis (HMG-CoAreductase), bile acid synthesis (CYP7A), and cholesterol efflux (ABCG5) are well expressed in liver prior to weaning.
Conclusion-Our studies provide the first in vivo evidence for differential regulation of hepatic SREBP-1 and SREBP-2 proteolytic processing. SREBP-2 is expressed as both pSREBP-2 and nSREBP-2 in livers derived from 15-day-old preweaned animals. Moreover, its target gene, HMG-CoA reductase, is well expressed in the liver prior to weaning. In contrast, hepatic lipogenic gene expression (FAS and S14) is very low prior to weaning. Low lipogenic gene expression correlates with low nSREBP-1 levels. The mechanism limiting nSREBP-1 in preweaned animals involves abrogated conversion of pSREBP-1 to nSREBP-1 (Figs. 2 and 3) and not pretranslational suppression of mRNA SREBP-1 as seen in the adult (18 -20). The physiological consequence of selective processing of SREBP-1 and SREBP-2 is a shift of hepatic metabolism toward cholesterol synthesis, i.e. HMG-CoA reductase, and away from de novo lipogenesis, i.e. S14 and FAS (Figs. 2-4).
Because SCAP and S1P are required for both SREBP-1 and SREBP-2 processing (4, 6, 7), it is unlikely that the selective blockade of SREBP-1 maturation is due to deficient SCAP or S1P. Alternatively, two other mechanisms might contribute to selective SREBP-1 and SREBP-2 proteolytic processing. Worgall et al. (17) reported that unsaturated fatty acid inhibition of SREBP processing was linked to sphingolipid metabolism and ceramide generation. This mechanism inhibited both SREBP-1 and SREBP-2 proteolytic processing without effects on SREBP-1 or SREBP-2 mRNAs.
An alternative mechanism involves the insulin regulated ERassociated proteins INSIG-1 and INSIG-2. INSIG-1 mRNA was originally isolated by Taub and co-workers (36) and found to be induced by insulin in Rueber H35 hepatoma cells. Yang et al. (37) and Yabe et al. (38) reported that INSIG-1 and INSIG-2 bind SCAP in the ER and block SREBP export to the Golgi for proteolytic processing. Janowski (10) recently reported that LXR agonists attenuate INSIG-1 gene expression (10). The LXRregulated transcripts, ABCG5 and CYP7A1, are well expressed in neonatal liver implicating elevated hepatic oxysterols and active LXR (Fig. 4B). This observation, coupled with the fact that neonatal rats have low blood insulin levels and elevated hepatic ketogenesis (Refs. 28 and 29, and Fig. 4B), suggests that INSIG-1 or INSIG-2 might be low in the neonatal liver.
Clearly, more studies will be required to establish the roles sphingolipid metabolism, INSIG-1, and INSIG-2 play in the control of hepatic SREBP-1 and SREBP-2 proteolytic processing. We anticipate that definition of the molecular basis of differential processing of SREBP-1 and SREBP-2 in neonatal rat liver will provide important clues to selective control hepatic cholesterol and fatty acid synthesis.