Lipids

, Volume 44, Issue 8, pp 673–683

Docosahexaenoic Acid Activates Some SREBP-2 Targets Independent of Cholesterol and ER Stress in SW620 Colon Cancer Cells

Authors

  • Gro Leite Størvold
    • Department of Laboratory Medicine, Children’s and Women’s HealthNorwegian University of Science and Technology (NTNU)
  • Karianne Giller Fleten
    • Department of Laboratory Medicine, Children’s and Women’s HealthNorwegian University of Science and Technology (NTNU)
  • Cathrine Goberg Olsen
    • Department of Laboratory Medicine, Children’s and Women’s HealthNorwegian University of Science and Technology (NTNU)
  • Turid Follestad
    • Department of Mathematical SciencesNorwegian University of Science and Technology (NTNU)
  • Hans Einar Krokan
    • Department of Cancer Research and Molecular MedicineNorwegian University of Science and Technology (NTNU)
    • Department of Laboratory Medicine, Children’s and Women’s HealthNorwegian University of Science and Technology (NTNU)
Original Article

DOI: 10.1007/s11745-009-3324-4

Cite this article as:
Størvold, G.L., Fleten, K.G., Olsen, C.G. et al. Lipids (2009) 44: 673. doi:10.1007/s11745-009-3324-4

Abstract

The SREBP-2 transcription factor is mainly activated by low cellular cholesterol levels. However, other factors may also cause SREBP-2 activation. We have previously demonstrated activation of SREBP-2 by the polyunsaturated fatty acid docosahexaenoic acid (DHA) in SW620 colon cancer cells. Despite activation of SREBP-2, only a few target genes were induced and cholesterol biosynthesis was reduced. In the present study, gene expression analysis at early time points verified the previously observed SREBP-2 target gene expression pattern. Activation of SREBP-2 using siRNAs targeting Niemann Pick C1 protein (NPC1) led to increased expression of all SREBP target genes examined, indicating that activation of some SREBP-2 target genes is inhibited during DHA-treatment. Cholesterol supplementation during DHA treatment did not abolish SREBP-2 activation. We also demonstrate that activation of SREBP-2 is independent of ER stress and eIF2α phosphorylation, which we have previously observed in DHA-treated cells. Thapsigargin-induced ER stress repressed expression of SREBP-2 target genes, but with a different pattern than observed in DHA-treated cells. Moreover, oleic acid (OA) treatment, which does not induce ER stress in SW620 cells, led to activation of SREBP-2 and induced a target gene expression pattern similar to that of DHA-treated cells. These results indicate that DHA and OA may activate SREBP-2 and inhibit activation of SREBP-2 target genes through a mechanism independent of cholesterol level and ER stress.

Keywords

SREBP-2Cholesterol biosynthesisER stressTranscriptional regulationeIF2α

Abbreviations

ATF6

Activating transcription factor 6

DHA

Docosahexaenoic acid

DnaJA4

DnaJ (Hsp 40) homolog-subfamily A-member 4

ER

Endoplasmic reticulum

eIF2α

Eukaryote translation initiation factor 2 alpha

FA

Fatty acid

FDFT

Farnesyl diphosphate farnesyltransferase

HMGCR

3-Hydroxy-3-methylglutaryl-coenzyme A reductase

IDI

Isopentenyl-diphosphate delta isomerase

Insig

Insulin induced gene

LDLR

Low density lipoprotein receptor

NPC1

Niemann Pick C1

OA

Oleic acid

PUFA

Polyunsaturated fatty acid

SCAP

SREBP cleavage activating protein

siRNA

Small interfering RNA

SRE

Sterol regulatory element

SREBP

SRE binding protein

mSREBP

Mature SREBP

SQS

Squalene synthetase

Tg

Thapsigargin

UFA

Unsaturated fatty acid

Introduction

The sterol regulatory element binding proteins (SREBPs) are transcription factors regulating fatty acid (FA) and cholesterol synthesis. Three family members have been identified to date; SREBP-1a, SREBP-1c and SREBP-2. SREBP-1a and SREBP-1c are transcribed from the same gene using alternate promoters, while SREBP-2 is transcribed from a separate gene [1]. The SREBP-2 and SREBP-1c isoforms are expressed in liver and most other tissues, while in cultured cells SREBP-2 and SREBP-1a are the most dominant isoforms [2]. SREBP-1c mainly regulates genes involved in FA biosynthesis, while SREBP-2 activates genes involved in cholesterol biosynthesis [3]. SREBP-1a, which has a longer transcription activating domain than the SREBP-1c isoform, is able to induce expression of genes involved in both FA and cholesterol biosynthesis [1, 4].

The SREBP transcription factors are synthesized as inactive precursor proteins localized in the endoplasmic reticulum (ER) membrane [5]. The SREBP precursor protein forms a complex with another ER-localized protein, the SREBP cleavage activating protein (SCAP), which contains a cholesterol sensing domain [6]. Binding of cholesterol when ER cholesterol levels are high induces a conformational change in SCAP. This enables binding of the SCAP-SREBP complex to another ER resident protein; insulin-induced gene 1 (Insig1), which retains the SCAP-SREBP complex in the ER [7, 8]. When cells are depleted of sterols, SCAP dissociates from Insig-1, and the SCAP-SREBP complex is translocated to the Golgi where the SREBP precursor is cleaved by two proteases; site-1 protease (S1P) and site-2 protease (S2P) [5]. This releases the N-terminal fraction of the protein, termed mature SREBP (mSREBP), which contains the transcription factor domain. The active transcription factor is translocated to the nucleus where it induces expression of target genes, including the SREBP transcription factors themselves [3, 5]. SREBP target genes contain a sterol regulatory element (SRE) sequence or modified forms of this sequence in their promoters, while SREBP-1c target genes mainly contain E-boxes or E-box like sequences [9, 10]. The SREBP transcription factors have weak transcriptional domains, and require co-factors like NF-Y or Sp1 to activate target genes [9, 11].

In addition to cholesterol, unsaturated fatty acids (UFAs) have been shown to regulate mSREBP levels. A unified mechanism explaining the regulation of SREBP by UFAs has not yet been established. It has been demonstrated that UFAs may inhibit the proteolytic processing of the precursor protein [1215], reduce SREBP-1 mRNA stability [16], as well as transcription [12, 1719]. Studies performed on hepatic tissue in vivo or in hepatic cell lines have demonstrated that polyunsaturated FAs (PUFAs), but not the monounsaturated FA oleic acid (OA), reduce the levels of mSREBP-1 [15, 16, 19] or both SREBP-1 and -2 [20]. Other studies have demonstrated that OA as well as PUFAs, reduce the levels of both SREBP-1 and -2 [13, 14]. In a study using the non-hepatic 293T fibroblast-like cell line, both OA and arachidonic acid (AA) reduced SREBP-1 levels, while no alterations in SREBP-2 levels were found [12]. A similar response was seen when the CaCo-2 colon adenocarcinoma cell line was treated with PUFAs, while OA had no effect [18].

Regulation of the SREBP transcription factors independent of cholesterol and FA levels has also been reported. Several studies have demonstrated activation of the SREBP transcription factors by cellular stress like hypotonic shock [21], ER stress [2123] or apoptosis [24].

ER stress is induced by disturbances in ER homeostasis and leads to initiation of the unfolded protein response (UPR). The response is mediated by three ER-localized transmembrane proteins; activating transcription factor 6 (ATF6), inositol-requiring enzyme 1 (IRE1) and the eukaryote translation initiation factor 2 alpha (eIF2α) kinase 3 (EIF2AK3/PERK). Activated PERK phosphorylates eIF2α [25], thereby inducing a global inhibition of translation resulting in reduced protein synthesis and depletion of short lived proteins like cyclin D1 [26].

We have previously demonstrated that treatment of SW620 human colon cancer cells with the PUFA docosahexaenoic acid (DHA; 22:6n-3) leads to reduced levels of SREBP-1 [27], while the SREBP-2 transcription factor is activated [28]. Despite increased levels of the activated transcription factor, only a few SREBP-2 target genes were up-regulated at mRNA level, and synthesis of cholesterol from acetate was reduced in DHA-treated cells [28]. In addition, DHA treatment induced accumulation of cholesteryl esters in SW620 cells without a corresponding increase in total cholesterol [27, 28]. These data made us hypothesize that free cholesterol was redistributed from the ER membrane to cholesterol esters, inducing local cholesterol depletion in the ER and activation of SREBP-2. DHA treatment also induced ER stress in SW620 cells [28], and it was hypothesized that depletion of cholesterol in the ER could lead to the observed ER stress and growth inhibition observed in DHA-treated SW620 cells.

In the present study we demonstrate that activation of SREBP-2 by DHA is not caused by cholesterol depletion or ER stress induction. Treatment of SW620 cells with OA, which does not induce cholesteryl ester accumulation [27] or eIF2α phosphorylation [28], also lead to activation of SREBP-2, indicating that UFAs may have a common mechanism for SREBP-2 activation. Further, we observe that in DHA- and OA- treated cells, despite activated SREBP-2, expression of only a few target genes is increased. In cells where SREBP-2 is activated by cholesterol depletion all target genes examined are induced, indicating that activation of some SREBP-2 target genes is inhibited during DHA and OA treatment.

Experimental Procedures

Cell Culture

The human colon adenocarcinoma cell line SW620 was obtained from ATCC (Rockville, MD). Cells were cultured in Leibowitz’s L-15 medium (Cambrex, BioWhittaker, Walkersville, MD) supplemented with L-glutamine (2 mM), FBS (10%) and gentamicin (45 mg/l) (complete growth medium) and maintained in a humidified atmosphere of 5% CO2: 95% air at 37 °C.

Fatty Acid, Cholesterol and Thapsigargin Treatment

Stock solutions of DHA and OA in ethanol (Cayman Chemical, Ann Arbor, MI) were stored at −20 °C and diluted in complete growth medium before experiments (final concentration of ethanol <0.01% v/v). Cholesterol and Tg (both Sigma, St. Louis, MO) were dissolved in ethanol. Stock solutions were kept at −20 °C and diluted in complete growth medium before experiments. Complete growth medium with an equal volume of ethanol as for the treatment media was used as the control media in the experiments. For all treatments cells were plated at a density of 3.5 × 106 cells /175 cm2 flask and allowed to attach over night, further FA, cholesterol or Tg was added and cells were treated for the indicated time periods. For DHA and cholesterol treatments, cells were also plated in triplicates in 6-well trays treated in parallel with the 175 cm2 flasks to assay cell growth. Cells were harvested by trypsination and counted using a Coulter Counter (Beckman Coulter, Fullerton, CA).

Immunoblot Analysis

Preparation of total protein extracts was performed as described previously [27]. An equal amount of protein was separated on 10% precast denaturing NuPAGE gels (Invitrogen, Carlsbad, CA) and transferred to PVDF membranes (Millipore, Billerica, MA). Membranes were blocked in PBS supplied with 5% nonfat dry milk and 0.1% Tween®-20 (BioRad, Herkules, CA) and further incubated over night at 4 °C with the indicated primary antibodies. The following primary antibodies were used to probe the membranes; LDLR (Progen, Germany), FDFT1/SQS (Abgent, San Diego, CA), Insig-1 (Santa Cruz, Santa Cruz, CA), SREBP-2 (BD Biosciences, San Jose, CA), IDI1 and β-actin (both AbCam, UK). Membranes were further incubated with HRP-conjugated secondary antibodies (DAKO, Carpinteria, CA) and detected by chemiluminescence using Super Signal West Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL) and visualized by Kodak Image Station 4000R (Eastman Kodak Company, Rochester, NY). Quantification was performed using Kodak Molecular Imaging Software (version 4.0.1).

siRNA Transfection

siRNAs targeting Niemann Pick C1 protein (NPC1) (NCBI accession number NM_000271) and eukaryote translation initiation factor 2 alpha (eIF2α) (NCBI accession number NM_004094) were purchased from Dharmacon (Lafayette, CO) (NPC1si#2, eIF2αsi#14) and Ambion (Austin, TX) (NPC1si#7). A non-targeting siRNA (Dharmacon, Lafayette, CO) was used as a control. siRNAs were dissolved in 1× siRNA buffer (Dharmacon, Lafayette, CO) and stock solutions were aliquoted and stored at −80 °C. For siRNA transfection experiments, 8.17 × 106 cells were plated in 175 cm2 flasks in complete growth media without gentamicin. Cells were transfected 24 h after plating using siRNA (50 nM) and Dharmafect 1 transfection reagent (Dharmacon). The media was changed 24 h after transfection, and the cells were allowed to recover for 6 h before they were sub-cultured at a density of 3.5 × 106 cells/175 cm2 flask for further growth or DHA treatment. Cells were harvested by scraping in ice cold PBS, and total cell extracts were prepared as described above.

RNA Isolation

For RNA isolation 3.5 × 106 cells were seeded in 175 cm2 flasks and allowed to attach over night. The following morning, the growth medium was replaced with media supplemented with DHA (70 μM) or an equal volume of ethanol (control). After 6 h, cells were harvested by scraping in ice cold PBS and stored at −80 °C. Total RNA was isolated using the High Pure RNA Isolation Kit (Roche, Mannheim, Germany) according to the instruction manual. RNase inhibitor rRNasin (40U/μl, 1 μl) (Promega, Madison, WI) was added and RNA was up-concentrated on a speed vac and resuspended in RNAse free distilled H2O. RNA concentration and quality were determined using the NanoDrop1000 (NanoDrop Technologies, Wilmington, Delaware) and agarose gel electrophoresis.

Gene Expression Profiling

A 5-μg amount of total RNA was used for cDNA- and cRNA synthesis according to the eukaryote expression manual (Affymetrix, Santa Clara, CA). cRNA was hybridized to the Human Genome U133 Plus 2.0 Arrays (Affymetrix). Washing and staining were performed using the Fluidics Station 450 and the arrays were scanned using an Affymetrix GeneChip Scanner 3000 with autoloader (Affymetrix, USA). Expression profiling was performed in triplicate at all time points using RNA from independent biological replicates, except for the 1 h time point when only two control experiments were used. All experiments were submitted to ArrayExpress with accession number E-MEXP-2010. The statistical analysis was performed based on summary expression measures of the probe sets of the GeneChips, using the raw data (CEL) files and a linear statistical model for background-corrected, quantile normalized and log-transformed PM values, performed by the robust multiarray average (RMA) method [29, 30]. Data from all time points (0.5, 1, 3, and 6 h) were used in the statistical analysis.

For each transcript, a linear regression model including parameters representing treatment and time effects and treatment-time interactions was fitted to the RMA expression measures. Based on the estimated effects, tests for significant differential expression due to DHA treatment and time were performed using moderated t tests, in which gene-specific variance estimates are replaced by variance estimates found by borrowing strength from data on the remaining genes [31]. To account for multiple testing, adjusted P values controlling the false discovery rate (FDR) were calculated [32]. Differentially expressed genes were selected based on a threshold of 0.05 on the adjusted P-values.

The statistical analysis was performed in R (http://www.r-project.org), using the packages Limma and affy from Bioconductor [33]. Differentially expressed genes were annotated using the NetAffx Analysis Centre (http://www.Affymetrix.com) and NMC Annotation Tool/eGOn V2.0 (http://www.GeneTools.no).

Results

Expression of mSREBP-2 Target Genes at mRNA and Protein Level in DHA-Treated SW620 Cells

We have previously found that only some SREBP-2 target genes are up-regulated in SW620 cells after 12–48 h of DHA treatment despite activation of SREBP-2 [28]. In the present study, gene expression analysis was performed on SW620 cells treated with DHA for short time periods using arrays covering the whole human genome. No SREBP-2 target genes were found differentially expressed at mRNA level after 0.5, 1 and 3 h treatment with DHA. However, after 6 h of DHA treatment up-regulation of some, but not all, SREBP-2 target genes was observed (Table 1), in agreement with our previous findings [28]. In addition, two SREBP-2 target genes that were not present on the Focus arrays previously used were found up-regulated at mRNA level; farnesyl-diphosphate farnesyltransferase 1 (FDFT1), encoding for squalene synthetase (SQS), an enzyme in the cholesterol biosynthetic pathway, and DnaJ (Hsp40) homolog, subfamily A, member 4 (DnaJA4), recently demonstrated to be a SREBP-2 target gene (Table 1) [34].
Table 1

Significant differentially expressed transcripts of SREBP target genes in SW620 cells treated with DHA (70 μM) for 6 h

Gene symbol

Affymetrix ID

Refseq NCBI ID

Description

SW620 fold change

6 h

Chaperones/protein folding

DnaJA4

1554333_at

NM_18602.3

DnaJ (Hsp40) homolog, subfamily A, member 4

2.39

1554334_a_at

  

2.18

225061_at

  

2.45

Cholesterol biosynthesis, uptake, metabolism and transport

LDLR

202068_s_at

NM_000527

Low density lipoprotein receptor

2.28

202067_s_at

  

1.55

NPC1

202679_at

NM_000271

Niemann-Pick disease, type C1

1.58

SC4MOL

209146_at

NM_001017369

Sterol-C4-methyl oxidase-like

2.2

 

NM_006745

  

FDFT1

243658_at

NM_004462

Farnesyl-diphosphate farnesyltransferase 1

2.89

241954_at

  

2.08

238666_at

  

1.93

HMGCS1

221750_at

NM_002130

3-hydroxy-3-methylglutaryl-coenzyme A synthase 1 (soluble)

1.5

 

NM_001098272

  

VLDLR

209822_s_at

NM_001018056

Very low density lipoprotein receptor

3.23

 

NM_003383

  

IDI1

204615_x_at

NM_004508

Isopentenyl-diphosphate delta isomerase 1

NC

208881_x_at

   

233014_at

   

242065_x_at

   

*NC no change

Several of the SREBP-2 target genes were also analyzed at protein level by Western blot. Increased protein levels of low density lipoprotein receptor (LDLR), SQS/FDFT1 and DnaJA4 were observed after 12 h of DHA treatment, and remained elevated up to 48 h of treatment (Fig. 1a). The level of the SREBP target gene isopentenyl-diphosphate delta isomerase 1 (IDI1), also an enzyme in the cholesterol biosynthetic pathway, was not changed by DHA treatment neither at mRNA nor protein level (Table 1, Fig. 1a).
https://static-content.springer.com/image/art%3A10.1007%2Fs11745-009-3324-4/MediaObjects/11745_2009_3324_Fig1_HTML.gif
Fig. 1

Analysis of protein levels of mSREBP-2 and selected SREBP target genes after DHA and OA treatment. a Western blot analysis of mSREBP-2 target gene protein levels in total protein extracts from SW620 cells treated with DHA (70 μM) for the indicated time periods (h). Controls were harvested at all time points, only 24 h control (C) is shown. b Western blot analysis of mSREBP-2 and SREBP target gene protein levels in total protein extracts from SW620 cells treated with OA (70 μM) for 24 h. β-actin was used as a control for equal protein loading. The blots were quantified and protein band intensities normalized relative to β-actin. For results in Fig. 1a, the β-actin adjusted band intensities from the DHA and control membranes were further normalized relative to the 24 h control band, present at all membranes, to adjust for differences in signal intensities between the membranes. The numbers under the blots represent mean fold change (with standard deviation(SD)) of DHA- or OA- treated samples relative to control at indicted time points for three independent experiments. *Significantly different from control (Student’s t test, P < 0.05)

Treatment of SW620 cells with OA, resulted in a similar protein expression pattern as for DHA-treated cells. Protein levels of mSREBP-2 and SQS were increased after 24 h of OA treatment, while IDI1 was unchanged (Fig. 1b).

Activation of SREBP-2 by NPC1 Knockdown Induces Expression of all Target Genes Examined in SW620 Cells

Activated SREBP-2 has been demonstrated to generally induce expression of all target genes [3537]. In order to investigate whether activation of SREBP-2 by NPC1 knockdown would induce expression of target genes not activated during DHA treatment, SW620 cells were transfected with siRNAs targeting NPC1. The NPC1 protein is involved in intracellular cholesterol transport of LDLR-derived cholesterol from the lysosomes to the interior of the cell, and inhibition of NPC1 function leads to cholesterol depletion in the ER [38]. Significantly reduced NPC1 protein levels were observed in cells transfected with NPC1 siRNA, as well as increased levels of mSREBP-2 (Fig. 2a). Further, induction of both SQS and IDI1 was observed (Fig. 2a), in contrast to DHA-treated cells where only SQS was induced. These results demonstrate that SW620 cells are able to increase expression all target genes examined after activation of SREBP-2 by cholesterol depletion induced by NPC1 knockdown.
https://static-content.springer.com/image/art%3A10.1007%2Fs11745-009-3324-4/MediaObjects/11745_2009_3324_Fig2_HTML.gif
Fig. 2

Analysis of protein levels of mSREBP-2 and selected SREBP target genes after induction of cholesterol depletion or cholesterol supplementation. a Western blot analysis of protein levels of NPC1, mSREBP-2 and mSREBP-2 targets in total protein extracts from SW620 cells transfected with NPC1 siRNAs. Cells were transfected with two siRNAs targeting independent sequences in the NPC1 transcript (NPCsi#2 and NPCsi#7), while control cells were transfected with a non-targeting siRNA. Cells were harvested for analysis 72 h after transfection. The blots were quantified and protein band intensities normalized relative to β-actin loading control. The numbers under the blots represent mean fold change (with SD) of NPC1 siRNA transfected samples relative to control at indicted time points (n = 3). b Western blot analysis of mSREBP-2 levels in total protein extracts from SW620 cells after DHA treatment and cholesterol supplementation. Cells were incubated for 24 h in control media (C), DHA (70 μM), 30 μg/ml cholesterol (Chol), or both cholesterol (30 μg/ml) and DHA (70 μM); (Chol + DHA). The blots were quantified and intensities normalized relative to β-actin loading control. The numbers under the blots represent mean fold change (and SD) relative to control for three independent experiments. *Significantly different from control (Student’s t test, P < 0.05)

Cholesterol Supplementation During DHA Treatment Does Not Inhibit Activation of SREBP-2 or Relieve DHA-Induced Cell Cycle Arrest

We have previously observed increased cholesterol esterification in addition to reduced de novo cholesterol synthesis in SW620 cells treated with DHA [27, 28]. This may lead to depletion of cholesterol in the ER, which could lead to the observed activation of SREBP-2 [28] and possibly cause growth inhibition [39]. To further investigate this, we incubated SW620 cells with cholesterol alone or in combination with DHA. Cholesterol treatment alone reduced the level of mSREBP-2 compared to control, indicating uptake of sterols in ER and inhibition the proteolytic activation of the SREBP-2 transcription factor (Fig. 2b). When cells were treated with DHA in the presence of cholesterol, a significant increase in mSREBP-2 levels relative to control was observed (1.45 ± 0.16 relative to control with cholesterol only, P < 0.05), demonstrating activation of SREBP-2 by DHA in the presence of excess cholesterol. Further, the cell cycle arrest observed in SW620 cells after DHA treatment [27] was not rescued by co-incubation with cholesterol. Cholesterol treatment alone did not affect cell growth (results not shown). The data indicates that activation of mSREBP-2 and the growth arrest observed in SW620 cells after DHA treatment are not caused by depletion of cholesterol.

DHA Regulates mSREBP-2 and Target Genes Independent of ER Stress and eIF2α Function

We have previously observed induction of ER stress and phosphorylation of eIF2α in SW620 cells during DHA treatment [28]. Previous reports have demonstrated that ER stress activates SREBP-2 [22, 23], but may inhibit SREBP-2 target gene expression [40]. To investigate whether ER stress related mechanisms are causing the activation of mSREBP-2 and repression of target genes in our system, we used three different approaches; (1) induction of ER stress by thapsigargin (Tg), (2) measuring the level of the negative regulatory protein Insig-1 and (3) siRNAs targeting eIF2α.

Thapsigargin Treatment Regulates mSREBP-2 and Target Genes

Tg induces ER stress by perturbing ER calcium homeostasis. Activation of SREBP-2 was observed when SW620 cells were treated with Tg for 6 h (Fig. 3a), but the level of mSREBP-2 was higher than in cells treated with DHA for the same time period [28]. The level of mSREBP-2 declined after 24 h of Tg treatment, but was still significantly higher than control (Fig. 3a). These results demonstrate that ER stress may lead to activation of SREBP-2 in SW620 cells, but that the response differs from the response observed after DHA treatment, where the level of mSREBP-2 remained elevated up to 48 h of treatment [28]. The same response was also observed when using a tenfold lower concentration of Tg, indicating that the difference in response is not caused by high concentrations of Tg with a corresponding stronger ER stress response. The protein levels of selected SREBP-2 target genes were assayed after Tg-treatment of SW620. Unexpectedly, SQS and DnaJA4, both induced by DHA treatment, were down-regulated in Tg-treated cells (Fig. 3a). Furthermore, expression of IDI1, which was unchanged in DHA treated cells, was increased relative to control after 24 h of Tg-treatment (Fig. 3a).
https://static-content.springer.com/image/art%3A10.1007%2Fs11745-009-3324-4/MediaObjects/11745_2009_3324_Fig3_HTML.gif
Fig. 3

Regulation of protein levels of mSREBP-2 and SREBP target genes by ER stress and eIF2α knockdown. a Western blot analysis of the protein levels of mSREBP-2 and mSREBP-2 target genes in total protein extracts from SW620 cells treated with 0.2 μM thapsigargin (Tg) for the indicated time periods (h). The blots were quantified and protein band intensities normalized relative to β-actin loading control. The numbers under the blots represent mean fold change (with SD) of Tg-treated samples relative to control at indicted time points for three independent experiments. b Western blot analysis of Insig-1 protein levels in total protein extracts from SW620 cells treated with DHA (70 μM) for the indicated time periods (h). Controls were harvested at all time points; only 24 h control (C) is shown. The blots were quantified and protein band intensities normalized relative to β-actin. The β-actin adjusted band intensities from the DHA and control membranes were further normalized relative to the 24 h control band, present at all membranes, to adjust for differences in signal intensities between the membranes. The numbers under the blots represent mean fold change (with SD) of DHA treated samples relative to control at indicted time points for three independent experiments. c Western blot analysis of the protein levels of eIF2α, cyclin D1 and mSREBP-2 in total protein extracts from SW620 cells transfected with a siRNA targeting eIF2α. Cells were harvested for analysis 48 h after transfection. d Western blot analysis of the protein levels of eIF2α and mSREBP-2 in total protein extracts from SW620 cells transfected with a siRNA targeting eIF2α and further treated with DHA for 24 h. For all eIF2α experiments control cells were transfected with a non-targeting siRNA. The blots were quantified and protein band intensities normalized relative to β-actin loading control. The numbers under the blots represent mean fold change (with SD) of siRNA transfected cells relative to the corresponding control at indicted time points for three independent experiments. *Significantly different from control (Student’s t test, P < 0.05)

mSREBP-2 Activation is not Caused by Depletion of the Regulatory Protein Insig-1

It has been suggested that activation of mSREBP-2 upon cellular stress may be caused by depletion of the negative regulatory protein Insig-1 due to eIF2α phosphorylation and inhibition of protein translation [21]. However, the level of Insig-1 was not altered upon Tg nor DHA treatment in SW620 cells (Fig. 3a, b) demonstrating that SREBP-2 is not activated by Insig-1 depletion in these settings.

Inhibition of Protein Translation Using siRNAs Targeting eIF2α

To investigate whether inhibition of protein translation would lead to activation of SREBP-2 through a mechanism other than reduced level of Insig-1, siRNAs targeting eIF2α were used to deplete the protein from the cells. Transfection with siRNA significantly reduced the level of eIF2α, and a significant reduction in cyclin D1 level was also observed, indicating inhibition of protein translation (Fig. 3c). Knockdown of eIF2α did not result in increased levels of mSREBP-2, indicating that inhibition of protein translation does not lead to mSREBP-2 activation in SW620 cells. Instead decreased levels mSREBP-2 was observed (Fig. 3c). When eIF2α siRNA transfected cells were treated with DHA for 24 h, increased level of mSREBP-2 relative to cells not treated with DHA was observed (Fig. 3d). These results indicate that activation of mSREBP-2 by DHA is independent of ER stress and eIF2α function.

Discussion

Activation of the SREBP-2 transcription factor and induction of cholesterol biosynthesis are mainly regulated by cellular cholesterol levels [1]. However, cellular stress and ER stress have also been reported to activate SREBP-2, independent of cholesterol levels [21, 22]. We have previously demonstrated that DHA treatment leads to activation of the SREBP-2 transcription factor in the human colon cancer cell line SW620, but that the activated transcription factor only induces expression of some SREBP target genes [28]. In the present study we demonstrate that activation of SREBP-2 by DHA occurs independent of cholesterol levels and ER stress induction, and that activation of some SREBP-2 target genes is inhibited during DHA treatment.

In our previous study, activation of the SREBP-2 transcription factor was observed as early as after 3 h of DHA treatment [28], and in the present study we demonstrate by gene expression analysis that increased mRNA levels of SREBP target genes can be observed after 6 h of DHA treatment. The gene expression analysis also verified the previously observed expression pattern in DHA-treated SW620 cells where activation of only a few SREBP-2 target genes was observed despite the activated transcription factor. In SW620 cells treated with OA, we observe a similar effect as for DHA on mSREBP-2 and target gene protein levels.

Previous studies have demonstrated that when SREBP-2 is activated by cholesterol depletion or by over-expression of the active transcription factor, expression of all known target genes is generally induced [3537]. In line with this we demonstrate that SREBP-2 activated by knockdown of NPC1 is able to induce all target genes examined, unlike the response observed in DHA-treated cells. These results demonstrate that during DHA treatment activated SREBP-2 is not able to increase expression of all target genes.

Previously we have demonstrated increased protein levels of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) and NPC1 in SW620 cells treated with DHA [28], and in the present study we demonstrate increased protein levels of LDLR. Together with the observed activation of mSREBP-2 [28], these results suggest an increased demand for intracellular cholesterol. However, our data demonstrate that activation of SREBP-2 is not inhibited by cholesterol supplementation to the cell culture media during DHA treatment. It has been suggested that cholesterol dissolved in ethanol added to cell culture media will not be efficiently absorbed by cells [41], and that observed effects could be caused by oxysterol impurities that are more easily absorbed. Thus, reduced levels of mSREBP-2 observed after cholesterol supplementation to SW620 cells could be caused by oxysterols present in the cholesterol. However, oxysterols are not able to substitute for the cellular functions of cholesterol [5, 41]. Also, others have demonstrated by using the same approach, that cholesterol supplementation may restore growth in cholesterol depleted cells [39, 42]. These results strongly indicate that SREBP-2 activation in DHA-treated cells is not caused by cholesterol depletion, despite the accumulation of cholesteryl esters and reduced cholesterol biosynthesis observed previously [27, 28]. Furthermore, OA treatment of SW620 cells does not induce cholesteryl ester accumulation [27], but we observe activation of SREBP-2 (this work). The data demonstrates that both DHA and OA may activate SREBP-2 independent of cellular cholesterol levels, possibly through a mechanism common for UFAs.

We further demonstrate that activation of SREBP-2 by DHA treatment is most likely not caused by ER stress or inhibition of protein translation, and that it is independent of Insig-1 protein levels and eIF2α function. In SW620 cells treated with OA we observe activation of SREBP-2, but not induction of ER stress and eIF2α phosphorylation [28]. These data further support a common mechanism for UFAs for activation of SREBP-2 independent of both cholesterol levels and ER stress.

Activation of SREBP-2 by induction of apoptosis has also been reported [24], but we have previously not detected induction of apoptosis or activation of caspases in DHA-treated SW620 cells [27, 28].

Inhibition of transcription of SREBP-2 target genes has previously been associated with ER stress-induced activation and cleavage of ATF6 [40]. Tg treatment activates both ATF6 [40] and mSREBP-2, as demonstrated in our study. The reduced protein levels of SQS and DnaJA4 observed after Tg treatment are most likely caused by ATF6 interaction with SREBP-2 transcription, since reduced levels of SQS after ATF6 activation have been demonstrated previously [40]. Unexpectedly, proteins that were repressed in Tg-treated cells were the same proteins found induced upon SREBP-2 activation in DHA treated cells. ATF6 activation was not assayed in this work, but it could be speculated that the ER stress response or activation of ATF6 in DHA-treated cells may be weaker than in Tg-treated cells, or that all the branches of the ER stress response are not equally activated upon DHA treatment [35]. In OA-treated cells, where no ER stress has been observed, the SREBP targets SQS and IDI1 display the same response as observed in DHA-treated cells. Together the data indicates that lack of activation of some SREBP-2 target genes during DHA or OA treatment is not caused by activation of ER stress and ATF6.

Regulation of the SREBP transcription factors by UFAs have been shown in several studies [12, 13, 17], but our study is the first to demonstrate increased levels of mSREBP-2 after UFA treatment. Why UFA treatment would cause activation of SREBP-2 in the SW620 colon cancer cell line is not known and remains to be investigated in our system. However, activation of SREBP-2 by DHA does not result in activation of all target genes and increased cholesterol biosynthesis. Instead, activation of some SREBP-2 target genes in DHA- and OA-treated cells seems to be inhibited, and a significant reduction in de novo cholesterol biosynthesis has previously been observed after DHA treatment [28].

Protein modification of the SREBP-2 transcription factor by phosphorylation, acetylation, sumoylation or ubiquitination may affect the binding affinity or activation of target promoters [4345]. UFAs could also possibly affect co-factors or regulate repressors that are involved in SREBP-2 target gene transcription [46, 47]. The SREBP-1a transcription factor has also been demonstrated to be involved in the regulation of genes involved in cholesterol biosynthesis [10, 36, 37]. We have previously observed reduced levels of mSREBP-1 in DHA-treated SW620 cells [27]. SREBP-1 levels were not altered in NPC1 siRNA transfected cells, where expression of all SREBP target genes examined were induced (results not shown). Whether the observed SREBP-2 target gene expression pattern induced by DHA and OA could be caused by reduced SREBP-1 levels, or if it is caused by other factors, remains to be investigated.

The liver is the main site for cholesterol biosynthesis in the body, but extra-hepatic tissue also synthesizes a significant amount of cholesterol (reviewed in [48]). In addition cancer cells and tumors display increased rates of cholesterol biosynthesis [49]. In spite of this, most studies investigating the effect of dietary fatty acids on cholesterol metabolism have been performed in fibroblasts or hepatic cell lines or tissues [1416, 19, 20, 50]; only a few studies have addressed regulation of SREBPs or cholesterol biosynthesis in other tissues and cell types [18, 51]. The results from our study have revealed that UFAs may affect cholesterol homeostasis and reduce cholesterol biosynthesis in a colon cancer-derived cell line by interfering with induction of SREBP-2 transcription. This is in line with previous observations where UFA treatment reduced transcriptional levels of cholesterogenic enzymes or reduced cholesterol biosynthesis in both hepatic tissue and cells, as well as glioma cells [19, 20, 50, 51]. Altered levels of cholesterol biosynthesis in the liver may contribute to altered plasma cholesterol levels [48], and it has been suggested that the reduction in cholesterogenesis induced by UFAs may be a mechanism explaining the plasma cholesterol-lowering effect that has been observed in some studies after UFA treatment [20, 50]. In addition reduced levels of cholesterol biosynthesis in cancer cells may contribute to reduced cancer cell growth [52, 53]. However, further studies are needed to confirm our observations and investigate possible implications in other cancer cell lines.

Together the data presented in this report demonstrate that DHA and OA activate SREBP-2 independent of cholesterol levels and ER stress, possibly through a mechanism common for UFAs. Further, we show that activation of SREBP-2 by these FAs only increases expression of some SREBP target genes. The mechanisms by which DHA and OA increase the level of mSREBP-2, but inhibit induction of some SREBP target genes in SW620 human colon cancer cells remain to be investigated.

Acknowledgments

Technical assistance from Caroline Hild Hakvaag Pettersen, Mari Sæther and Jens Erik Slagsvold is highly appreciated. The project was financed by The Faculty of Medicine, NTNU, The Cancer Research Fund, Trondheim University Hospital and The Research Council of Norway through grants from the Functional Genomics Program (FUGE). Microarray experiments were performed at the microarray core facility at the Norwegian Microarray Consortium (NMC), Trondheim, which is supported by FUGE.

Copyright information

© AOCS 2009