Acta Chim. Slov. 2008, 55, 85–92
85
Scientific paper
High-fat Medium and Circadian Transcription Factors
(Cryptochrome and Clock) Contribute to the Regulation
of Cholesterogenic Cyp51 and Hmgcr Genes
in Mouse Embryonic Fibroblasts
Martina Fink, Klemen [paninger, Ur{ula Prosenc and Damjana Rozman*
Centre for Functional Genomics and Bio-Chips, Institute of Biochemistry; Faculty of Medicine, University of Ljubljana, Zalo{ka 4, 1000 Ljubljana, Slovenia
* Corresponding author: E-mail: damjana.rozman@mf.uni-lj.si Phone: +386 1 543 7591; Fax: +386 1 543 7588
Received: 02-11-2007
Abstract
The aim of our research was to investigate how cholesterol, unsaturated fatty acids and circadian genes affect the expression of cholesterogenic genes, Cyp51 and Hmgcr, in somatic and in embryonic fibroblast cell lines. We found that in immortal Hepa1–6 cells cholesterol represses the transcription of Hmgcr and Cyp51 for 80%, while unsaturated fatty acids have different effects: Hmgcr was repressed for 50%, but Cyp51 was unaffected by the presence of linoloeic acid. In embryonic fibroblasts the abundance of cholesterol in the media did not repress the expression of Cyp51 and Hmgcr, while the presence of linoleic acid repressed transcription of both genes for 40%. Mutation of the Clock gene activated the basal transcription of Cyp51 and Hmgcr and also reconstituted the cholesterol feedback loop, that was lacking in the wild type embryonic fibroblasts. Deletion of repressor genes Cry1 and Cry2 resulted in activated transcription of chole-sterogenic genes after addition of linoleic acid, while response to cholesterol was unchanged compared to wild type embryonic fibroblasts. Our results indicate that cholesterol, unsaturated fatty acids and the circadian transcription factors participate in the regulation of cholesterogenesis through different molecular mechanisms, presumably using different SREBP transcription factors and their coregulatory proteins.
Keywords: Cholesterol biosynthesis, circadian rhythm, Cyp51, Hmgcr, mouse embryonic fibroblasts
1. Introduction
Cholesterol is a prerequisite for the animal life, because it is a component of membranes. Its biosynthetic pathway is not destined only to the production of the end product, but several other physiologically important molecules, such as vitamin D, heme, isopentenyl-tRNA, dimethylalyl pyrophosphate, ubiquinone, dolichol, farnesyl pyrophosphate, originate from the pathway. The homeostasis of cholesterol is precisely regulated. The disturbances can lead to hypercholesteremia, one of the leading causes of death in the developed countries. On the other hand, hypocholeste-remia is linked to the higher risk for aggression, depression and even suicide1. Cholesterol is important also during the embryonic development,2–4 where the main roles are likely its covalent binding to the Sonic Hedgehog proteins which act as signaling molecules.2,5 Shortage of cholesterol can
Figure 1: Shematic representation of cholesterol biosynthesis and homeostasis.
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
Acta Chim. Slov. 2008, 55, 85–92
be caused by inhibition of enzymes participating in its biosynthesis or by the mutations of cholesterogenic genes.3 Lack of cholesterol during embryogenesis leads to serious developmental abnormalities.6 Cellular cholesterol requirements are met through de novo synthesis and/or uptake of plasma lipoproteins (Fig 1).
Cholesterol is synthesized via isopreonid biosynthetic pathway (Fig 1) starting from acetyl-coenzyme A (Ac– CoA), which is transformed into hydroxymethyl-glutaryl coenzyme A (HMG–CoA). HMG–CoA is reduced into mevalonate with HMG–CoA reductase (HMGCR), the rate-limiting enzyme of the cholesterol biosynthetic pathway. Mevalonate is transformed into squalene, leading to lanosterol and the cholesterol. Lanosterol is reduced into 4,4-dimethylcholesta-8(9),14,-24-trien-3ί-ol (FF-MAS) by the lanosterol 14?-demethylase (CYP51).
In somatic cells the homeostasis of cholesterol is regulated via several signaling pathways, such are cholesterol feedback loop, the cAMP-dependent signaling pathway, etc,7–11 the cholesterol feedback loop being seemingly most important and best understood. Cholesterogenesis is also under the circadian control, although the underlying mechanisms are poorly understood.12,13
Dietary polyunsaturated fatty acids and cholesterol suppresses transcription of hepatic genes involved in the metabolism of fatty acids and cholesterol through the action of sterol responsive element binding proteins (SREBP) transcription factors. Newly synthesized SREBPs are embedded into nuclear and endoplasmic reti-culum membranes and are transcriptionally inactive in sterol/fatty acids loaded conditions. In lipid deprived physiological conditions, the N-terminal part of SREBPs are released from membranes and transferred to the nuc-lei.14 SREBP-1 isoforms (SREBP-1a and SREBP-1c) primarily regulate fatty acid synthesis, while SREBP-2 is mainly involved in the regulation of cholesterol homeo-stasis.15 Contrary to the animal tissues where SREBP-1c and SREBP-2 are highly expressed, SREBP-1a is mainly expressed in immortal cell lines. SREBP-1 seems to be mainly involved in energy metabolism including fatty acid and glucose/insulin metabolism, whereas SREBP-2 is specific to cholesterol synthesis,16 although all types of SREBPs can modulate both cholesterol and fatty acids metabolism in vitro. The proteolytic activation of SREBP-1 and SREBP-2 can be regulated individually. For example, in rodents statins upregulate SREBP-2 activity and reduce SREBP-1 activity, while polyunsaturated fatty acids inhibit SREBP-1, but contrary to the oxysterols they have no effect on SREBP-2 maturation and activity, indicating that SREBP-1c is a key target for PUFA-dependent suppression of de novo lipogenesis.17–19
Circadian rhythm is a biological rhythm with the period of approximately 24 h. It is daily entrained by environmental cues (e.g. light, temperature), and it is generated by the transcription-translation negative feedback loop composed of a set of circadian genes. Circadian proteins
are transcriptional activators and repressors binding to several promoter-regulatory elements such as E-box, cAMP-responsive elements, DBP/E4BP4 and REV-ERB? /ROR. Among the most investigated mammalian circadian genes are also activators Clock and Bmal and repressors of the chryptochrome (Cry1, Cry2) and period (Per1, Per2, Per3) families.20–22 Molecular mechanisms of the circadian regulation are explained elsewhere.23–27
The goal of our work was to investigate for the first time the effects of lipids (cholesterol and fatty acids) on the expression of model cholesterogenic genes (Hmgcr and Cyp51) in mouse embryonic fibroblasts and in immortal liver Hepa 1–6 cells. The role of mutation/deletion of cir-cadian genes Clock and Cryptochrome 1 and 2 to the process of cholesterogenesis has also been investigated.
2. Methods and Materials
2. 1. Cell culture Conditions
We used human choricarcinoma (Jeg3) cells, mouse hepatocarcinoma cells (Hepa 1–6), wild type mouse embryonic fibroblast cells (Mef3), mouse embryonic fi-broblast cells with the mutated Cry1 and Cry2 genes (Cry1, 2 -/-) and mouse embryonic fibroblast cells with the mutated Clock gene (Clock -/-). Jeg3 and Hepa 1–6 cells were grown in DMEM media (Sigma, Taufkirchen, Germany) containing 5% bovine calf serum and 1% L-glutamine in a 5% CO2 incubator at 37oC, while the three mouse embryonic fibroblast cell lines were cultured in similar conditions, only that 5% serum bovine calf serum was exchanged with 10% fetal calf serum. All used mouse embryonic fibroblast cells were obtained from P. Sassone-Corsi laboratory and maintained by previously published protocols.28 Mice with deleted Cry1, 2 genes were prepared as published,29 while Clock mutated animals originate from the laboratory of J. Tahakashi.30
In experiments the following media were used: (1) COPUFA – lipid rich medium (DMEM with 1% bovine serum albumin (BSA), cholesterol (10 ΅g/ml), 25-hydroxycholesterol (1 ΅g/ml), linoleic acid (0.15 mM), (2) CO – COPUFA without linoleic, (3) PUFA – COPU-FA media without cholesterol and 25-hydroxycholesterol and normal medium (NM). Cells were grown to 90% confluency in T75 tissue culture flasks.
For RNA isolation cells (Hepa 1–6, Mef3, Cry1,2 -/-, Clock -/-) were split 1:3 24 h before the beginning of the experiment and were put into 6-well microtiter plates into normal media. At the beginning of the experiment, media was changed into normal, COPUFA, CO or PUFA, as indicated. RNA was isolated 24 h after media have been changed. For the transfections only Jeg3 cells were used. After splitting cells were put directly into experimental medium (normal, COPUFA, CO, PUFA). At the beginning of the experiment, medium was replaced with fresh and was changed every 24 h.
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
Acta Chim. Slov. 2008, 55, 85–92
2. 2. RNA Isolation
Total cellular RNA was isolated form Hepa 1–6, Mef3, Cry 1,2 -/- and Clock -/- cell lines. After 24 h the media were aspirated. Cells were washed twice with RNAse-free PBS buffer and 0,5 ml of TRI reagent (Sigma, Taufkirchen, Germany) was added directly to the cells. After 5 min the solution of TRI reagent with lysed cells was transferred into RNAse-free microcentrifuge tubes and total RNA was isolated. RNA concentration and quality were determined by RNA 6000 Nano Assay with Agilent 2100 Bioanalyzer (Agilent Technologies). At least two different samples of each RNA sample have been isolated and investigated for each experimental condition.
2. 3. Quantitative RT-PCR
For Q-RT-PCR analysis 1 ΅g of total RNA was converted into cDNA in a 20 ΅l reaction mixture using a SuperScript II Reverse transcriptase (Invitrogen, Carlsbad, CI) with random primers (Promega, Madison, WI). Reaction mixture was treated with DNase I (Sigma, Taufkirchen, Germany) to remove the contaminating DNA. The quality of each cDNA was tested in PCR reaction with ί-actin primers. Q-RT-PCR analysis was preformed using SybrGreen technology on an Applied Biosystems Prism 7900 HT sequence detection system. Specific primers for Cyp51 and Hmgcr were published previously.31,32 Primers were validated by the analysis of template titration and dissociation curves. PCR products were analyzed by melting curve analysis. All reactions were performed in triplicate, control 18S rRNA or investigated Cyp51 and Hmgcr primers were amplified as follows: after incubation at 50 °C for 2 min and denaturating at 95 °C for 10 min, 40 cycles were performed at 95 °C for 15 s, and 60 °C for 1 min. The relative amounts of all mRNAs were calculated by the comparative Ct (cycle number at threshold) method (User Bulletin No.2, Applied Biosystems, PE) using 18S rRNA as the internal control. Relative mRNA levels were determined by expressing the mRNA amount of cells grown in normal media comparable to the expression of investigated genes in cells grown in lipid rich media (COPUFA, CO, PUFA). Expression level in embryonic fibroblast lines with deleted / mutated Cry 1, 2 and Clock genes was normalized to the expression in wild type Mef3 cells grown in the normal media. Data were processed with 2-??CT method as described previously.33 The average value, SEM and p values from two-tailed t test were calculated with the Excel program (Microsoft Corp., Redmond, WA).
2. 4. Transient Transfection Assays and
Determination of Reporter Gene Activity
One day prior transfections Jeg3 cells were plated into 12-well plates. On the day of transfection, the media was replaced with fresh media and transfection was performed with Ca2+ precipitate method with the firefly luciferase hu-
man CYP51 D7 reporter (–334/+316). Cells were trans-fected with 500 ng of luciferase reporter, 125 ng of pSV ί-galactosidase and 100 ng of pCMV SREBP-2, pCMV BMAl1, pCMV CLOCK, pCMV Per1 and pCMV CRY1 expression plasmids. In the Clock/Bmal1 titration experiment quantities of expression plasmid are indicated in the Fig 4B. pCAT basic plasmid was used as the carrier DNA to 1 ΅g. ί-galactosidase was used for normalization of the transfection efficiency. Media was changed every 24 h. 48 h after transfection cells were harvested with Promega passive lysis buffer (Promega, Madison, WI). The assay for determining the ί-galactosidase activity was performed as described.34 Activity of the firefly luciferase was analyzed with the commercial kit for luciferase (Promega, Madison, WI) and Turner TD-20/20 luminometer. All transfec-tion experiments were performed at least three times in duplicates for each experimental condition. In all transfec-tion experiments, reporter activity has been calculated by the formula: reporter activity = (normalized reporter gene activity) / (reporter activity in non-treated cells). Reporter activity in non-treated cells represents unit 1 and is shown in column 1 of diagram. The average value, SEM and p values from two-tailed t test were calculated with the Excel program (Microsoft Corp., Redmond, WA). Preparation of human CYP51 luciferase reporter construct (–334/+316) was published previously.11
SREBP- 2 expression plasmid originates from the lab of J. Horton (University of Texas Southwestern Medical Center, Dallas, TX, USA) while Bmal1, Clock, Per1 and Cry1 expression plasmids were prepared in P. Sassone-Corsi group (IGBC, Strasbourg, France).
3. Results
Cholesterol and unsaturated fatty acids affect expression of Cyp51 and Hmgcr in mouse hepatoma and in embryonic fibroblasts.
Cholesterol biosynthesis is a housekeeping process, which is in completely differentiated somatic cells primarily regulated by the negative feedback loop, where transcription factors of SREBP family are used as major mediators. SREBPs in such physiological conditions activate the transcription of cholesterogenic genes in lipid deprived conditions, but in lipid loaded conditions SREBPs are not present in cell nuclei.8,35,36 In embryonic cells the cholesterol feedback loop is not functional probably due to the increased need for cholesterol.5 We wanted to reveal whether the same molecular mechanisms participate in the regulation of the two model cholesterogenic genes (Hmgcr and Cyp51) in somatic cells and embryonic fi-broblasts. Figures 2A and 2B show that lipid rich medium (COPUFA) represses the transcription of cholesterogenic Cyp51 and Hmgcr in mouse Hepa 1–6 cells. The repressor effect of linoleic acid and cholesterol is additive in the case of Hmgcr (Fig 2B). Presence of linoleic acid alone in
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
88
Acta Chim. Slov. 2008, 55, 85–92
the media repressed Hmgcr expression for 50%, but Cyp51 transcription was almost unaffected, repression was only 10% (Fig 2A). Cyp51 expression was downregu-lated only with cholesterol and 25-hydroxycholesterol (Fig 2A white and grey bars). Cyp51 and Hmgcr expression in wild type mouse embryonic fibroblasts (Mef3) was unaffected by the addition of cholesterol and 25–hydroxycholesterol (CO) to the media (Fig 2A, B), while polyunsaturated fatty acids (PUFA) have the repression effect (the striped bars in Mef3, Fig 2A, B).
by COPUFA in Mef3 cells with deleted cryptochrome genes (Fig 3A, white bars in Cry 1, 2 -/- column), while addition of cholesterol alone repressed transcription of Cyp51 for 50% (Fig 3A) but did not affected the expression of Hmgcr (Fig 2B). Growing mouse embryonic cells with deleted Cry1, 2 genes in the medium with linoleic acid activated the expression of Cyp51 and Hmgcr genes 2,5 fold (Fig 3A, B).
Mutation of Clock activated transcription of Cyp51 for 6-fold, while Hmgcr was induced 2-fold (Fig 3A, B. Addi-
Figure 2: Role of lipid rich media on the Cyp51 and Hmgcr expression in Hepa 1–6 cells and wild type Mef3 cells. A; Cyp51. B; Hmgcr. Total RNA was isolated from cells grown in indicated media for 24 h. Normal media (NM) – black bars, COPUFA (10 ΅g/ml cholesterol, 1 ΅g/ml 25-hydroxyc-holesterol, 0,15 mM linoleic acid) – white bars, CO – (10 ΅g/ml cholesterol, 1 ΅g/ml 25-hydroxycholestero) – grey bars or PUFA (0.15 mM linole-ic acid ) – striped bars. * – significantly modulated expression (p < 0.05) compared to the expression in normal media (black bars).
The second question was how the deletion/mutation of three circadian genes (Clock, Cry1, Cry2) changes the response of Hmgcr and Cyp51 to cholesterol and unsaturated fatty acids in mouse embryonic fibroblasts. Deletion of both cryptochrome (Cry1 and Cry2) genes led to a weak repression (30–35%) of Hmgcr compared to expression in the wild type mouse embryonic fibroblasts (compare black bars in Fig 2 B and 3 B) in normal media, while basal expression of Cyp51 was unaffected under the same (normal) conditions (Fig 3B). Media change into the lipid rich medium (COPUFA) led to the diminished repression
tion of linoleic acid to the media did not cause changes in the expression of Cyp51 and Hmgcr in mouse embryonic cells with mutated Clock (Fig 3A, B, column Clock -/-, striped bars). Loading media with cholesterol and 25-hydroxycholesterol repressed expression of Cyp51 and Hmgcr for 90% and 70% compared to level in normal media (Fig 3A, B, column Clock -/-, grey bars).
To check if the proximal promoter is responsible for the effects of mutation/deletion circadian genes seen in Fig 3A to the Cyp51 transcription, the transfections of human CYP51 luciferase reporter construct (–334/+316) in-
Figure 3: Role of lipid rich media in the Cyp51 and Hmgcr gene expression in Mef3 cells with deleted/ mutated Cry1, 2 and Clock genes. A; Cyp51 expression, B; Hmgcr expression. Total RNA was isolated from cells grown in indicated media for 24 h. Relative expression unit 1 is expression in wild type Mef3 cells in normal media (Fig 2, black bars in Mef3 column). Normal media (NM) – black bars, COPUFA (10 ΅g/ml cholesterol, 1 ΅g/ml 25-hydroxycholesterol, 0.15 mM linoleic acid ) – white bars, CO – (10 ΅g/ml cholesterol, 1 ΅g/ml 25-hydroxycholesterol) – grey bars or PU-FA (0.15 mM linoleic acid ) – striped bars. * – significantly modulated expression (p < 0.05) compared to the expression in normal media (black bars), # significantly modulated expression (p < 0.05) compared to the expression in wild type Mef3 (black bars in figure 2).
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
Acta Chim. Slov. 2008, 55, 85–92
89
Figure 4: Circadian transcription factors Clock and Bmal1 do not transactivate the proximal promoter of human CYP51 gene in Jeg3 cells. A; Transient transfections of Jeg3 with human CYP51 proximal promoter/reporter construct with the overexpression of SREBP-2 and circadian transcription factors. Overexpression of 1; Clock + Bmal1, 2; Cry1 + Per1, 3; SREBP-2. B; Transfections of human CYP51 proximal promoter lucifera-se reporter with different amouns of Clock and Bmal1 overexpression vectors. * – significantly modulated expression (p < 0.05).
to Jeg3 cells were performed. Figure 4 shows that in ex vivo conditions the proximal promoter of human CYP51 gene is mainly responsive to the cholesterol feedback loop and transcription factor SREBP-2 (Fig 4A, column 3) and is almost unresponsive to the overexpression of circadian transactivators CLOCK and BMA1, irrespectively of their quantity (Fig 4A column 1 and Fig 4B) or repressors PER1 and CRY1 (Fig 4A, column 2).
4. Discussion
Cholesterol is a molecule with the diverse and indispensable functions during prenatal and postnatal life of animals. Different molecular mechanisms are involved in regulation of cholesterol homeostasis in the cells, tissues and whole organism in different physiological and pat-hophysiological conditions. The major pathway regulating cellular cholesterol level in normal physiological conditions in adult somatic cells is the SREBP pathway, which represses transcription of genes involved in the biosynthesis and uptake of cholesterol and fatty acids in the conditions of abundance of lipids. However, when cells lack lipids, the SREBP signaling pathway is activated and genes are actively transcribed leading to a higher level of biosynthesis and uptake of lipids.8,35,36
Hepa 1–6 and Jeg3 immortal cell lines are frequently used as ex vivo models for research of cholesterol homeo-stasis in completely differentiated somatic tissues. Our earlier research of CYP51 was done principally in Jeg3 cells and it was determined that the COPUFA medium represses CYP51 transcription for 70%–90% in 24 h.31 In the present study we investigated how loading of cells with unsaturated fatty acids (linoleic acid) and/or cholesterol affect the transcription of two model cholesteroge-nic genes Hmgcr and Cyp51 in different physiological conditions when the circadian activators or repressors are depleted from the experimental system.
In some tissues, such as germ cells and embryonic tissues, the cholesterol feedback loop does not work pro-
perly and cholesterol is synthesized although there is abundance of lipids. The role for such regulation is not well understood, but it is proposed that the cholesterol and intermediates are important for the development of fetus and mature germ cells.5,37,38 The differences in responsiveness of sterol synthesis rates between adult and fetal tissues are at least partially due to the constitutive processing of SREBP-2.5 Both model cholesterogenic genes Hmgcr and Cyp51 are in Hepa 1–6 cells 85–90% down-regulated by the presence of cholesterol. On the other hand, the addition of linoleic acid to the media repressed for 50% only transcription of Hmgcr, while Cyp51 expression was reduced for only 10% (Fig 2A, B). This indicates that Hmgcr and Cyp51 might be differently regulated by SREBP-1 and SREBP-2, although they are both a part of the same cholesterol biosynthetic pathway. We propose that Cyp51 is regulated mostly by SREBP-2 in completely developed cells and that unsaturated fatty acids acting through SREBP-1 are not important for Cyp51 transcription. Hmgcr, the rate limiting enzyme, of the early phase of the cholesterol biosynthesis, which is not destined only for the cholesterol synthesis, is regulated much more precisely by the participation and interactions of several other signaling pathways. One of these is probably mediated by SREBP-1 and fatty acids.
We also investigated how lipids affect expression of Hmgcr and Cyp51 in embryonic fibroblasts (Mef3). The presence of cholesterol and 25-hydroxycholesterol in the media did not repressed Cyp51 and Hmgcr genes (Fig 2A, B). Lower expression of Cyp51 and Hmgcr in COPUFA media is only due to the repression by linoleic acid (Fig 2A, B). Experiments with animals with deleted Srebp genes showed the urgent presence of SREBP-2 as a principal activator of cholesterogenesis during embryonic development, because deletion of Srebp-2 is 100% embryonic lethal, while Srebp-1 knock-out is in 15–50% of mice compatible with life.39 The repression of Cyp51 and Hmgcr by linoleic acid in wild type Mef3 cells may be explained by the suggestion of Woollett et al, that polyunsaturated fatty acids (PUFA) alter lipid metabolism within fetus via
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
90
Acta Chim. Slov. 2008, 55, 85–92
SREBP-1 and PPAR signaling pathways.40 Embryonic uptake of exogenous maternal cholesterol was previously demonstrated to rescue the drug induced Smith-Lemli-Opizt syndrome,40 suggesting a dispensable action of cholesterol feedback loop in the embryo. It was proposed that in fetal tissues, the ability to suppress and activate sterol synthesis may not be advantageous since the need for cholesterol is high and constant. This apparent dysregulation of sterol synthesis in fetal and other rapidly growing tissues might be the result of the changed regulation of SREBP-2 activity, which can occur at different levels, such constitutive processing of SREBP-2 and its decreased degradation.5
Cholesterol homeostasis is also a circadian process. Mutation of the Clock gene is involved in the development of metabolic syndrome, characterized by obesity, hyperli-pidemia, hyperglycemia and hypoinsulinemia in the mouse C57BL/6J strain, while the ICR strain is almost unaf-fected.41–43 In our experiments mutation of Clock activated the basal expression of Cyp51 and Hmgcr genes (Fig 3, black bars) indicating a key role of the Clock protein even in early embryonic development. This observation can be explained by the report of Doi,44 that CLOCK acts as a hi-stone acetyltransferase and in this way regulates expression of target genes. Addition of cholesterol repressed the transcription of Cyp51 in both investigated embryonic fi-broblast cell lines with mutations/deletions of circadian genes for 50% (Cry1,2 -/-) and 90% (Clock -/-) (Fig 3A), similarly to the effect observed in liver cells. Expression of Hmgcr in cells with deleted Cry1 and Cry2 genes was not changed in cholesterol loaded conditions, but was repressed for 70% in cells with mutated Clock gene, indicating that the mutation of Clock might reconstitute the cholesterol feedback of embryonic fibroblasts missing in wild type embryonic fibroblast cells. The effect of linoleic acid was opposite to the effect of cholesterol. Addition of lino-leic acid to the media lead to a 2-fold induction of Cyp51 and Hmgcr expression in Cry 1,2 -/- cells while mutation of Clock did not affect transcription of cholesterogenic genes (Fig 2A, B, striped rows in Cry1,2 -/- and Clock -/-). Proposed explanation is that deficiency of circadian activator Clock leads to the loss of circadian expression pattern of cholesterogenic genes and consequently to the activation of cholesterogenic gene transcription. This is in accordance with our unpublished results with mice with the deleted Crem gene, where the expression of most cho-lesterogenic genes is upregulated (A~imovi~ et al, unpublished) and with the observations in Clock mutant mice that develop metabolic syndrome with hypoinsulinemia and hyperlipidemia.41 We propose that there is a possibility of development of hyperlipidemia in this mouse strain as a consequence of the obesity and disturbed energy metabolism through the activation of SREBP-2 system and inhibition of SREBP-1 proteins.
Deletion of cryptochrome genes had almost no effect on the transcription of Cyp51 and Hmgcr mediated by
cholesterol (Fig 3, grey lines), while addition of linoleic acid surprisingly activated expression of both cholestero-genic genes, indicating a novel role of repressor cryptochrome proteins in the regulation of cholesterogenic genes through the SREBP-1 system. The effects of CRY proteins might be indirect, because these proteins do not interact directly with DNA but through protein – protein interactions. We propose that in the case of cholesterol ho-meostasis the role of activating CLOCK protein is greater because the lack of Clock leads also to the depletion of cryptochrome repressors. In our research mutation of Clock completely changed the response of embryonic fi-broblast to the addition of lipids by inhibition of the SREBP-1 pathway and activation of the SREBP-2 pathway. On the other hand, deletion of cryptochrome genes left the SREBP-2 system unchanged and repressed only the SREBP-1 pathway.
The results shown in Fig 3 clearly demonstrate that the expression of Hmgcr and Cyp51 is closely connected with the circadian proteins. The target DNA regulatory element of circadian activators CLOCK and BMAl1 is E-box with SRE as a variant and we wanted to know if the proximal promoter region, which is responsible for SREBP-depen-dent expression of CYP51, is also responsible for the cir-cadian regulation. Transfections of proximal promoter lu-ciferase reporter construct of human CYP51 gene suggested that the circadian transcription factors likely act through distal promoter regulatory elements or they participate in the regulation of cholesterogenic genes through interactions with other signaling pathways.
Further investigations are needed for elucidation of the proposed hypothesis. Our results are in accordance with the results of Oishi et al. in the ICR mice strain, reporting that diet with high cholesterol significantly decreased expression of Hmgcr, Ldl and Cyp7a1 in livers of Clock mutant mice.42 Repressed transcription of cholesterogenic genes prevented the development of the metabolic syndrome. Although the overexpression of circadian transcription factors CLOCK, BMAl1, PER1 and CRY1 did not affect the expression of human CYP51 proximal promoter/reporter construct (Fig 4) in ex vivo conditions (also hamster HMGCR – data not shown), bioinformatic analysis of human and mouse CYP51 distal promoter regions showed the presence of several circadian regulatory elements (E-Box, CRE, ROR-REW) in the distal promoters of both mammalian CYP51 genes ([paninger, unpublished). This suggests that the circadian regulation of chole-sterogenic genes might be mediated through distal and not proximal promoters.
5. Conclusion
Lipid metabolism is an important factor during development and is also significant for maintaining the body energy homeostasis. Lipid homeostasis is regulated by se-
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
Acta Chim. Slov. 2008, 55, 85–92
91
veral signaling pathways although the molecular mechanisms are not fully understood. We showed for the first time, that 1. Cholesterol homeostasis is regulated in a different manner in embryonic cells and in completely differentiated cells. 2. Circadian genes are directly involved in the regulation of cholesterol homeostasis: mutation of Clock gene activates the expression of cholesterogenic Cyp51 and Hmgcr and reconstitutes the cholesterol feedback loop while deletion of cryptochromes extinguishes the repression effect of unsaturated fatty acids to the cholesterol biosynthesis. Deletion/mutation of circadian activators and repressors has different effects to the expression of cholesterogenic Hmgcr and Cyp51 probably due to their different response to SREBP-1 and SREBP-2 transcription factors. Further research is needed to elucidate the role of individual signaling pathway and the interactions between different factors in different physiologi-cal/pathophysiological conditions.
6. Acknowledgments
The authors are grateful to P. Sassone-Corsi (Irvine University, Irvine, CA, USA) for the gift of Mef3 cells and Clock, Bmal1, Per1 and Cry1 expression vectors. We would like also to thank J. Takahashi (Northwestern University, Evanston, IL, USA) and G. van der Horst lab (Erasmus University, Rotterdam, The Netherlands) who prepared Clock -/- and Cry1,2 -/- animals. This work was supported by the Agency of Science of Republic Slovenia (ARRS) grants J1-6713, J1-9438 and P1-104. K [panin-ger is supported by the ARRS graduate fellowship.
7. Abbreviations
BSA           bovine serum albumin
CO             lipid rich medium containing cholesterol and
25-hydroxycholesterol COPUFA    lipid rich medium containing cholesterol,
25-hydroxycholesterol and linoleic acid CRE           cAMP response element
Cry            chryptochrome
Cyp7a1       cytochrome P450 7a1 Cyp51        lanosterol 14?-demethylase Hmgcr        3-hydroxy-3-methylglutaryl coenzyme A reductase Mef            mouse embryonic fibroblast
P450          cytochrome P450
Per             period
PUFA         lipid rich media with linoleic acid, SREBP, sterol
regulatory element-binding protein SRE           sterol regulatory element
8. References
1. J. Brunner, K. G. Parhofer, P. Schwandt, T. Bronish, Phar-macopsychiatry 2002, 35, 1–5.
2. R. V. J. Farese, J. Herz, TIG 1998, 14, 115–120.
3. G. Wolf, J. Nutr. Biochem. 1999, 10, 188–192.
4. H. R. Waterham, R. J. Wanders, Biochim. Biophys. Acta 2000, 1529, 340–356.
5. L. Yao, K. Jenkins, P. S. Horn, M. H. Lichtenberg, L. A. Woollett, Biochim. Biophys. Acta 2007, 1771, 1372–1379.
6. H. R. Waterham, FEBS Lett. 2006, 580, 5442–5449.
7. R. B. Rawson, Nat. Rev. Mol. Cell. Biol. 2003, 4, 631–640.
8. M. S. Brown, J. L. Goldstein, Proc. Natl. Acad. Sci. USA
1999, 96, 11041–11048.
9. M. K. Bennett, T. F. Osborne, Proc. Natl. Acad. Sci. USA
2000, 97, 6340–6344.
10. K. A. Dooley, M. K. Bennett, T. F. Osborne, J. Biol. Chem. 1999, 274, 5285–5291.
11. S. Halder, M. Fink, M. R. Waterman, D. Rozman, Mol. Endocrinol. 2002, 16, 1853–1863.
12. M. Brewer, D. Lange, R. Baler, A. Anzulovich, J. Biol. Rhythms 2005, 20, 195–205.
13. B. Desvergne, L. Michalik, W. Wahli, Physiol. Rev. 2006, 86, 465–514.
14. M. S. Brown, J. Ye, R. B. Rawson, J. L. Goldstein, Cell 2000, 100, 391–398.
15. J. D. Horton, I. Shimomura, M. S. Brown, R. E. Hammer, J. L. Goldstein, H. Shimano, J. Clin. Invest. 1998, 101, 2331– 2339.
16. H. Shimano, Prog. Lipid. Res. 2001, 40, 439–452.
17. J. Xu, M. T. Nakamura, H. P. Cho, S. D. Clarke, J Biol Chem 1999, 274, 23577–23583.
18. D. Botolin, D. B. Jump, J. Biol. Chem. 2003, 278, 6959– 6962.
19. L. W. Weber, M. Boll, A. Stampfl, World J Gastroenterol
2004, 10, 3081–3087.
20. A. Balsalobre, Cell Tissue Res 2002, 309, 193–199.
21. S. A. Brown, U. Schibler, Curr Opin Genet Dev 1999, 9, 588–594.
22. D. Bell-Pedersen, V. M. Cassone, D. J. Earnest, S. S. Golden, P. E. Hardin, T. L. Thomas, M. J. Zoran, Nat Rev Genet
2005, 6, 544–556.
23. M. C. Antle, R. Silver, Trends Neurosci. 2005, 28, 145–151.
24. A. J. Davidson, O. Castanon-Cervantes, F. K. Stephan, Liver Int. 2004, 24, 179–186.
25. A. Kohsaka, J. Bass, Trends Endocrinol. Metab. 2007, 18, 4–11.
26. H. R. Ueda, S. Hayashi, W. Chen, M. Sano, M. Machida, Y. Shigeyoshi, M. Iino, S. Hashimoto, Nat. Genet. 2005, 37, 187–192.
27. H. Wijnen, M. W. Young, Annu. Rev. Genet. 2006, 40, 409– 448.
28. M. P. Pando, D. Morse, N. Cermakian, P. Sassone-Corsi, Cell 2002, 110, 107–117.
29. G. T. van der Horst, M. Muijtjens, K. Kobayashi, R. Takano, S. Kanno, M. Takao, J. de Wit, A. Verkerk, A. P. Eker, D. van Leenen, R. Buijs, D. Bootsma, J. H. Hoeijmakers, A. Yasui, Nature 1999, 398, 627–630.
30. N. Gekakis, D. Staknis, H. B. Nguyen, F. C. Davis, L. D. Wilsbacher, D. P. King, J. S. Takahashi, C. J. Weitz, Science
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...
92
Acta Chim. Slov. 2008, 55, 85–92
1998, 280, 1564–1569.
31. M. Fink, J. Acimovic, T. Rezen, N. Tansek, D. Rozman, Endocrinology 2005, 146, 5321–5331.
32. J. Yang, J. L. Goldstein, R. E. Hammer, Y. A. Moon, M. S. Brown, J. D. Horton, Proc. Natl. Acad. Sci. U S A 2001, 98, 13607–13612.
33. K. J. Livak, T. D. Schmittgen, Methods 2001, 25, 402–408.
34. J. Sambrook, E. F. Fritsch, T. M. Maniatis, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, NY, New York, 1989.
35. J. D. Horton, J. L. Goldstein, M. S. Brown, J. Clin. Invest. 2002, 109, 1125–1131.
36. M. S. Brown, J. L. Goldstein, Nutrition Reviews 1998, 56, S1–S3.
37. H. Wang, F. Liu, C. F. Millette, D. L. Kilpatrick, Mol. Cell. Biol. 2002, 22, 8478–8490.
38. K. Fon Tacer, S. Kalanj-Bognar, M. R. Waterman, D. Rozman, J. Steroid. Biochem. Mol. Biol. 2003, 85, 429–438.
39. H. Shimano, I. Shimomura, R. E. Hamer, J. Herz, J. L. Goldstein, M. S. Brown, J. Clin. Invest. 1997, 100, 2115– 2124.
40. L. A. Woollett, Curr. Opin. Lipidol. 2001, 12, 305–312.
41. F. W. Turek, C. Joshu, A. Kohsaka, E. Lin, G. Ivanova, E. McDearmon, A. Laposky, S. Losee-Olson, A. Easton, D. R. Jensen, R. H. Eckel, J. S. Takahashi, J. Bass, Science 2005, 308, 1043–1045.
42. K. Oishi, G. Atsumi, S. Sugiyama, I. Kodomari, M. Kasama-tsu, K. Machida, N. Ishida, FEBS Lett. 2006, 580, 127–130.
43. T. Kudo, K. Horikawa, S. Shibata, J. Pharmacol. Sci. 2007, 103, 139–143.
44. M. Doi, J. Hirayama, P. Sassone-Corsi, Cell 2006, 125, 497–508.
Povzetek
Namen na{ega dela je {tudij vpliva holesterola, nenasi~enih ma{~obnih kislin in cirkadialnih genov na izra`anje holeste-rogenih genov Cyp51 in Hmgcr v nesmrtnih celi~nih linijah tkivnih celic in zarodnih fibroblastov. Ugotovili smo, da v nesmrtni celi~ni liniji Hepa1-6 prisotnost holesterola za 80 % zavre prepisovanje genov Cyp51 in Hmgcr, medtem ko je dodatek nenasi~enih ma{~obnih kislin privedel do druga~nega odziva obeh genov. Prisotnost linolenske kisline v mediju ni vplivala na izra`anje Cyp51, prepisovanje Hmgcr pa je bilo zavrto za pribli`no 50 %. V mi{jih zarodnih fibrobla-stih (Mef3) izobilje holesterola ni vplivalo na izra`anje Cyp51 in Hmgcr, medtem ko je dodatek linolenske kisline zavrl izra`anje obeh genov za pribli`no 40 %. Mutacija cirkadialnega gena Clock je aktivirala osnovno prepisovanje Cyp51 in Hmgcr ter ponovno vzpostavila negativno povratno zanko holesterola, ki je v nespremenjenih celicah zarodnih fibrob-lastov nismo opazili. Delecija genov zaviralnih transkripcijskih faktorjev CRY1 in CRY2 je aktivirala prepisovanje ho-lesterogenih genov ob dodatku linolenske kisline, medtem ko je bil odziv celic na izobilje holesterola enak kot v zarod-nih fibroblastih divjega tipa. Na{i rezultati nakazujejo, da holesterol, nenasi~ene ma{~obne kisline in cirkadialni ritem sodeljujejo pri uravnavanju homeostaze holesterola preko razli~nih molekularnih mehanizmov, ki vklju~ujejo tudi kombinacije transkripcijskih faktorjev dru`ine SREBP in njihovih koregulatornih proteinov.
Fink et al.:  High-fat Medium and Circadian Transcription Factors (Cryptochrome and Clock) ...