Differential intestinal mucosal protein expression in hypercholesterolemic mice fed a phytosterol-enriched diet

Proteomics 2007, 7, 2659–2666

2659

DOI 10.1002/pmic.200600792

RESEARCH ARTICLE

Differential intestinal mucosal protein expression in hypercholesterolemic mice fed a phytosterol-enriched diet Laura Calpe-Berdiel1, Joan Carles Escolà-Gil1, Josep Julve1, Edgar Zapico-Muñiz1, Francesc Canals2 and Francisco Blanco-Vaca1* 1 2

Servei de Bioquímica i Institut de Recerca, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain Proteomics Laboratory. Medical Oncology Research Program, Vall d’Hebrón University Hospital Research Institute, Barcelona, Spain

The molecular mechanisms involved in the phytosterol-induced decrease in intestinal cholesterol absorption remain unclear. Further, other biological properties such as immunomodulatory activity and protection against cancer have also been ascribed to these plant compounds. To gain insight into the mechanisms underlying phytosterol actions, we conducted a proteomic study in the intestinal mucosa of phytosterol-fed apolipoprotein E-deficient hypercholesterolemic (apoE2/2) mice. With respect to control-fed apoE2/2 mice, nine differentially expressed proteins were identified in whole-enterocyte homogenates using 2-D DIGE and MALDI-TOF MS. These proteins are involved in plasma membrane stabilization, cytoskeleton assembly network, and cholesterol metabolism. Four of these proteins were selected for further study since they showed the highest abundance change or had a potential functional relationship with known effects of phytosterols. Annexin A2 (ANXA2) and b-actin decrease and annexin A4 (ANXA4) and annexin A5 (ANXA5) increase were confirmed by Western blot analysis. Intestinal gene expression of ANXA2 and A5 and b-actin was reduced, whereas that of ANXA4 was unchanged. The main results were retested in normocholesterolemic C57BL/6J mice. ANXA4 and ANXA5 protein upregulation and ANXA2 and b-actin downregulation were reproduced in these animals. However, no changes in gene expression were found in C57BL/6J mice in either of the four proteins selected. ANXA2, A4, and A5 and b-actin are proteins of special interest given their pleiotropic functions that include cholesterol-ester transport from caveolae, apoptosis, and antiinflammatory properties. Therefore, the protein expression changes identified in this study might be involved in the biological effects of phytosterols.

Received: October 17, 2006 Revised: March 15, 2007 Accepted: April 30, 2007

Keywords: Annexin / ApoE-deficient mice / Cancer / Gut / Inflammation

1

Introduction

Correspondence: Dr. Joan Carles Escolà-Gil, Hospital de la Santa Creu i Sant Pau, Servei de Bioquímica, C/Antoni M. Claret 167, 08025, Barcelona. Spain E-mail: [email protected] Fax: 134-93-2919196

Dietary phytosterols reduce the intestinal cholesterol absorption and constitute a recommended therapeutic option to decrease LDL (low-density lipoprotein) cholesterol

Abbreviations: ANX, annexin; apoE2/2, apolipoprotein E-deficient; FPPS, farnesyl pyrophosphate synthase

* Additional corresponding author: Dr. Francisco Blanco-Vaca, E-mail: [email protected]

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com

2660

L. Calpe-Berdiel et al.

in the current guidelines of the National Cholesterol Education Program (NCEP) [1]. One mechanism that could be involved in this hypocholesterolemic action is the physical competition between phytosterols and cholesterol for incorporation into micelles [2]. However, recent studies suggested that plant sterols may exert an unknown molecular action inside enterocytes and hepatocytes, especially considering that phytosterols do not need to be present in the intestinal lumen simultaneously with cholesterol to inhibit its absorption [3, 4]. Additionally, it has recently been described that some plant sterols, with an unsaturation within the side chain, disrupt cholesterol homeostasis since they are endogenous regulators of LXR and SREBP [5, 6]. This suggests the presence of a sterol transport system that regulates enterocyte cholesterol uptake [3, 7]. Previous reports demonstrated that phytosterols reduced both plasma cholesterol and atherosclerosis in hypercholesterolemic apolipoprotein E-deficient (apoE2/2) mice [8, 9]. Although most of the studies have focused on the cholesterol-lowering activity of phytosterols, other biological properties such as immunomodulation and protection against common cancers have also been ascribed to these compounds [10–13]. We recently performed a global gene expression profiling of the intestine of apoE2/2 mice fed a phytosterol-enriched diet [14]. This approach did not elucidate the molecular mechanisms by which phytosterols inhibit the intestinal cholesterol absorption, but did identify changes suggestive of immunomodulatory properties [14]. A known limitation of gene expression analysis is that the correlation between transcript and protein levels is often poor. Therefore, the aim of the current study was to analyze the intestinal mucosa proteome in apoE2/2 mice after a phytosterol-enriched diet. The main findings of the study were retested in C57BL/6J mice.

2

Materials and methods

2.1 Mice and diets All animal procedures were considered and approved by the Institutional Animal Care Committee of the Hospital de la Santa Creu i Sant Pau. Use of apoE2/2 mice with a C57BL/ 6J background has been described previously [15]. Eight female apoE2/2 mice were maintained in a temperaturecontrolled (207C) room with a 12-h light/dark cycle and food and water were provided ad libitum. Female mice were used in this study since they show increased intestinal cholesterol absorption [16]. Mice were randomized in two groups (n = 4) and fed for 1 month either a control Western-type diet (200 g/kg fat, polyunsaturated/saturated = 0.07, 0.8 g/kg cholesterol, 170 g/kg protein, 105 g/kg fiber; Mucedola srl, Settimo Milanese, Italy) or a 2% phytosterol-enriched Western-type diet w/w for 4 wk [14]. The phytosterol product was composed of 20% campesterol, 22% stigmasterol, and 41% b-sitosterol (Lipofoods, S. L., Gavà, Spain). Eight female wildtype C57BL/6J mice were obtained from Jackson Labora© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2007, 7, 2659–2666

tories (Bar Harbour, ME) and were fed either the control Western-type diet or the 2% phytosterol-enriched diet for 4 wk (n = 4 in each group). 2.2 Plasma lipid and lipoprotein analyses The methods used for plasma lipid and lipoprotein analyses had been described in detail elsewhere [15]. 2.3 Fluorescence 2-D DIGE Mice fed the two different diets were euthanized at the end of the study and the small intestines were cut, washed with PBS containing a proteinase inhibitor mixture (PMSF, pepstatin A, leupeptin, aprotinin, and EDTA), and opened longitudinally [17]. Mucosa was collected using a sharp blade to avoid contamination from the muscle layer and stored at 2807C until required [17–19]. Remaining tissue preparations were examined microscopically to confirm that mucosal cells had been removed. The mucosal scraping technique affords an enriched preparation of cells of the absorptive villi and immature crypt cells and, consequently, enzymes and transporters involved in lipid enterohepatic circulation [17–20]. Intestinal mucosa protein samples from four apoE2/2 animals of each experimental group were prepared. Protein extraction was performed by the addition of 400 mL of lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 30 mM Tris, pH 8.5) with 2 mM DTT to 100 mg of tissue. The mixture was sonicated on ice and centrifuged at 15 0006g at 47C for 5 min. Protein solutions were further purified by a modified TCA-acetone precipitation (2-D-CleanUp kit, Amersham Biosciences, Munich, Germany) and, finally, resuspended in DIGE labeling buffer (8 M urea, 4% w/v CHAPS, 30 mM Tris at pH 8.0). Protein concentration was determined using the BioRad RCDC Protein Assay (BioRad, Madrid, Spain) and this was adjusted to 2 mg/mL by the addition of DIGE labeling buffer. The protein concentration obtained was similar in phytosterol-fed and control-fed mice (2.4 6 0.1 vs. 2.4 6 0.3 mg/mg of tissue). A pool consisting of equal amounts of each animal sample was prepared to be used as internal standard for quantitative comparisons [21]. Inclusion of an the internal standard in 2-D gels significantly eliminated errors resulting from artefacts of electrophoresis and allowed accurate quantitative analyses [22]. Four 2-D DIGE analyses were performed. Each contained the internal standard, one sample of control-fed animal, and one sample of phytosterol-fed animals. Half of the samples from each experimental group were labeled either with Cy3 or Cy5 dyes, whereas the internal standard pooled sample was labeled with Cy2 (400 pmol of Cy dye in 1 mL of anhydrous DMF per 50 mg of protein) [21, 22]. An additional reverse-labeling experiment with four 2-D DIGE analyses was conducted to allow the labeling of each sample with Cy3 and Cy5. After 30 min of incubation on ice and in the dark, the reaction was quenched with 10 mM lysine and incubated again for 10 min. A mixture of 50 mg of protein per Cy dye per gel was www.proteomics-journal.com

Animal Proteomics

Proteomics 2007, 7, 2659–2666

combined, diluted two-fold with IEF sample buffer (8 M urea, 4% w/v CHAPS, 2% DTT, 2% Pharmalyte pH 3–10) and subjected to 2-D DIGE using GE-Healthcare reagents and equipment (Amersham Biosciences) as previously described [22]. Image analysis of 2-D DIGE and quantification of relative protein abundances were performed using DeCyder V. 5.0 software (Amersham Biosciences). During spot detection, the estimated number of spots were set at 2000, the exclude filter was set at slope .1.7, area ,100, peak height ,100, and volume ,10 000. Dividing each Cy3 or Cy5 spot volume by the corresponding Cy2 (internal standard) spot volume within each gel gave a normalized abundance, thereby correcting intragel variations [21]. A mean normalized abundance value from the two replicates was calculated for each spot in each mouse. Mann–Whitney U test was used to compare the abundance of all the matched spots (n = 4 in each group). 2.4 Protein identification by MALDI-TOF mass spectrometry Protein spots of interest were excised from the gel using an automated Spot Picker (GE healthcare Amersham Biosciences) [22]. In-gel trypsin digestion was performed as described [23], using autolysis-stabilized trypsin (Promega, Madrid, Spain). Tryptic digests were purified using ZipTip microtiter plates (Millipore, Madrid, Spain). MALDI (matrixassisted-laser desorption/ionization)-TOF analysis of tryptic peptides was performed on an Ultraflex TOF–TOF Instrument (Bruker, Wissembourg, France). Samples were prepared using CHCA as the matrix on anchor-chip targets (Bruker, Bremen, Germany). The spectra were processed using Flex Analysis 2.2 software (Bruker Daltonics). The resulting final peak list was used for the identification of the proteins by peptide mass fingerprint. MASCOT 2.0 program (Matrix Science, London UK) was used to search the MSDB database, 20050929 release (Imperial College, London) limiting the search to mouse proteins (73733 sequences). Criteria for positive identification were a significant MASCOT probability score (score .61, p ,0.05). A minimum score of 71 with a difference in score to the second ranked nonhomologous match greater than 30, was obtained for all the differentially expressed proteins identified.

2661

human annexin A2 (ANXA2) (with crossreactivity to mouse ANXA2) [24], mouse annexin A5 (ANXA5) (ABCAM, Cambridge, UK) and mouse b-actin (Sigma-Aldrich, St. Louis, Missouri, USA) and with mouse mAb against mouse annexin A4 (ANXA4) (BD Biosciences, San Jose, CA, USA) for 2 h. This was followed by incubation with horseradish peroxidase-conjugated antirabbit IgG (1:10 000 dilution) (Sigma-Aldrich) for 2 h. Protein bands were visualized using a chemiluminescent substrate (BioRad) and analyzed to give a quantitative estimation of intensity changes using the Quantity One® Software adapted to a Chemidoc XRS densitometer (BioRad). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to further confirm similar protein load in each well. High molecular weight markers (HMW, Amersham Biosciences) were used to estimate the relative molecular weight of each protein. 2.6 Quantitative real-time RT-PCR Total intestinal mucosa RNA was isolated from apoE2/2 mice used in 2-D DIGE analyses and C57BL/6J mice. Total RNA samples were isolated using the trizol RNA isolation method (Gibco/BRL), repurified, checked for integrity by agarose gel electrophoresis, and reverse-transcribed with Oligo(dT)23 using M-MLV Reverse Transcriptase, RNase H Minus, Point Mutant to generate cDNA [14]. Predesigned validated primers (Assays-on-Demand, Applied Biosystems, Foster City, CA, USA) were used with Taqman probes. PCR assays were performed on an Applied Biosystems Prism 7000 sequence detection system (Applied Biosystems, Foster City, CA, USA) as described [25]. All analyses were performed in duplicate and relative RNA levels were determined using PGK1 as the internal control. Use of GAPDH as an alternative internal control did not change results compared with those obtained with PGK1. 2.7 Statistical analyses Mann–Whitney U test was used to compare data obtained from phytosterol-fed and control-fed mice. GraphPad Prism 4.0 software (GraphPad, San Diego, CA) was used to perform all statistical analyses. A p value ,0.05 was considered statistically significant.

2.5 Western blot analyses Mucosal scrapings of the apoE2/2 mice used in 2-D DIGE analyses and C57BL/6J mice were resuspended and sonicated in radioimmunoprecipitation assay (RIPA) lysis buffer (50 mM Tris, 150 mM NaCl, 1% NP40, 0.1% SDS, 0.5% Nadeoxycholate, pH 8.0). Proteins were measured with the Bradford method as indicated by the supplier (BioRad) and 25 mg applied to each well to be subjected to electrophoresis on 15% SDS-polyacrylamide gels and transferred to PVDF membranes (BioRad). The membrane blots were incubated with rabbit polyclonal antibodies (1:1000 dilution) against © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

3

Results

3.1 Hypocholesterolemic effects of phytosterols apoE2/2 mice given 2% phytosterols for 4 wk exhibited a marked decrease in plasma total cholesterol (11.6 6 0.8 vs. 16.4 6 1.0 mmol/L in control mice, p ,0.05). This 30% reduction in plasma cholesterol was slightly lower than other previously described reductions that ranged from 37 to 50% [8, 14]. FPLC profiles showed a marked reduction in apoBwww.proteomics-journal.com

2662

L. Calpe-Berdiel et al.

Proteomics 2007, 7, 2659–2666

shown) although they did induce a decrease in the intestinal cholesterol absorption [14]. 3.2 Proteome analysis

Figure 1. Lipoprotein profile of phytosterol-fed and control-fed apoE2/2 mice. Cholesterol concentration was measured in 0.2 mL of lipoprotein fractions isolated by FPLC from pooled plasma of each group. Positions of elution of the VLDL, IDL (intermediate-density lipoprotein) 1 LDL, and HDL are represented by horizontal lines.

containing lipoproteins in phytosterol-treated mice, mainly in very-low-density lipoprotein (VLDL) (Fig. 1). Consistent with previous observations, phytosterols produced no effect on plasma cholesterol levels in C57BL/6J mice (data not

Figure 2 shows a representative 2-D gel of an intestinal mucosa protein extract from a pool of all studied samples of phytosterol-fed mice and control-fed apoE2/2 mice stained with Cy2. Analysis of the eight 2-D DIGE performed (each containing an internal standard, one control-fed mouse sample, and one phytosterol-fed mouse sample) showed the existence of nine significantly differentially expressed spots (see Fig. 2; p,0.05). All the detected differentially expressed spots were greater than 1.15-fold in abundance ratio (Table 1 and Fig. 3). Identification of these spots was resolved in all cases by MALDI-TOF MS. The proteins whose abundance changed can be seen in the gel shown in Fig. 2 and in Table 1. Among nine differentially expressed proteins found in intestinal mucosa of phytosterol-fed animals, five corresponded to upregulated proteins and four to downregulated proteins. Among nine proteins whose levels changed with dietary phytosterols, four corresponded to the cytoplasmic compartment, three to membrane proteins, one to ER, and one to a secreted protein (Table 1 and Fig. 3).

Figure 2. 2-D gel containing intestinal proteins from a pool of all analyzed intestinal mucosa samples stained with Cy2. Migration of proteins with expression level changes greater than 1.2-fold (p,0.05) detected by 2-D DIGE analyses were identified in the gel (see also Table 1).

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com

Animal Proteomics

Proteomics 2007, 7, 2659–2666

2663

Table 1. Proteins identified by MS changed in mice fed with phytosterols

Abundance Protein de- Protein ratioa) phyto- signation description sterols/control

Accession no.b)

pI/MWc)

Peptides Cell location (Seq, %)d)

11.7

ANXA4

Annexin A4

Q7TMN7

4.83/35773

20 (60)

11.65

ALB1

Albumin 1

Q8C7C7

5.49/67013

23 (43)

11.37

ACTN4

Actinin a-4

BC013616

5.25/105368 13 (20)

11.34

PDIA3

5.08/72640

22 (35)

11.25

ANXA5

Disulfide isomerase B34930 ERp72 Annexin A5 Q99LA1

4.83/35773

20 (60)

-1.17

FPPS

Farnesyl pyrophosphate synthase

Q920E5

5.49/40556

11 (39)

-1.19

ANXA2

Annexin A2

P07356

7.53/38806

9 (41)

-1.22

EC 2.8.2.9

Tyrosine-ester Q9R2C2 sulfotransferase

5.9/36860

26 (65)

-1.51

ACTB

b-Actin

5.78/39446

15 (38)

a) b) c) d)

CAA27396

Function/pathway

Plasma Calcium-dependent phospholipid binding. membrane Possible modulator of plasma-membrane Cl2 channels Secreted Lipid binding. Regulation of the colloidal osmotic pressure of blood Cytoplasm F-actin crosslinking protein that anchors actin to a variety of intracellular structures ER lumen Catalyzes the arrangement of –S–S– bonds in proteins Plasma Calcium-regulated membrane-binding membrane protein. Stabilizes certain plasma-membrane structures. Anticoagulant that acts as an indirect inhibitor of the thromboplastin-specific complex Cytoplasm Cholesterol biosynthesis. Catalyzes the sequential condensation of isopentenyl pyrophosphate with the allylic pyrophosphates, dimethylallyl pyrophosphate, and then with the resultant geranylpyrophosphate to the ultimate product farnesyl pyrophosphate Plasma Calcium-regulated membrane-binding membrane protein. ANXA2 and (CAV1) seem to be components of an intestinal sterol transport complex Cytoplasm This family includes a range of sulfotransferase proteins responsible for the transfer of sulfate groups to specific compounds Cytoplasm Structural protein

Abundance ratio averaged from the different gels of the DIGE experiment. Accession numbers of proteins were derived from Swiss-Prot database. pI and theoretical molecular mass calculated from the amino acid sequence and mature protein, respectively. Number of peptides matched and percentage of sequence coverage.

3.3 Western blot confirmation A selection of four proteins that appeared to be modified by the treatment was studied by Western blot in apoE2/2 mice and C57BL/6J mice. ANXA4 and b-actin were selected because of their highest change in expression increase and decrease, respectively, and ANXA2 and ANXA5 owing to the interest in the annexin (ANX) family in relation to intestinal cholesterol absorption. ANXA4 and ANXA5 protein upregulation and ANXA2 and b-actin downregulation were confirmed by this analysis in both apoE2/2 and C57BL/6J mice (Figs. 4A and 5A).

Figure 3. Variability in the protein expression of the different control-fed and phytosterol-fed apoE2/2 mice samples. Ratio-tospot volumes in control mice (normalized spot volume) for proteins for which statistically significant differences in expression levels were observed. Average values plus SEM are plotted.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

3.4 Intestinal mucosa gene expression analyses Quantification of intestinal mRNA by real-time RT-PCR revealed that ANXA5, ANXA2, and b-actin were downregulated 2.2-, 1.8-, and 3.6-fold, respectively, in phytosterolwww.proteomics-journal.com

2664

L. Calpe-Berdiel et al.

Proteomics 2007, 7, 2659–2666

Figure 4. (A) Western blot analyses of ANXA4, ANXA5, ANXA2, and b-actin in the small intestinal mucosa of control-fed mice and phytosterol-fed apoE2/2 mice. Protein extracts were prepared individually and immunoblotted with antibodies as described in Section 2. To permit comparisons between groups, relative units of selected proteins were calculated taking 100% that of the control-fed value for each protein. A representative Western blot for each protein of each experimental group is shown. Relative molecular weights were 36, 38, 38, and 42 kDa for ANXA4, ANXA5, ANXA2, and b-actin, respectively. (B) Relative intestinal mRNA levels. PGK1 was used as an internal control for these studies, and mRNA expression in the control-fed apoE2/2 mice was set to a normalized value of 100 arbitrary units. Results are expressed as mean 6 SEM of individual animals (n = 4). *p,0.05 compared with the control.

Figure 5. (A) Western blot analyses of ANXA4, ANXA5, ANXA2, and b-actin in the small intestinal mucosa of control-fed mice and phytosterol-fed C57BL/6J mice. Protein extracts were prepared individually and immunoblotted. To permit comparisons between groups, relative units of selected proteins were calculated taking 100% that of the control-fed value for each protein. A representative Western blot for each protein of each experimental group is shown. Relative molecular weights are shown in Fig. 4. (B) Relative intestinal mRNA levels. PGK1 was used as an internal control and mRNA expression in the control-fed C57BL/6J mice was used as a normalized value of 100 arbitrary units. Results are expressed as mean 6 SEM of individual animals (n = 4). *p,0.05 compared with the control.

treated apoE2/2 mice (Fig. 4B). ANXA4 intestinal gene expression did not change in these mice with dietary phytosterol (Fig. 4B). In contrast, no expression alteration was found when the same four intestinal mRNA levels were assessed in C57BL/6J mice (Fig. 5B).

4

Discussion

The data presented here provide, to the best of our knowledge, the first proteomic analysis characterizing the intestinal mucosa protein expression changes in response to a phytosterol-enriched diet. Mucosal scrapings of the entire small intestine were used in our study to analyze the whole duodenal-ileal axis and include all potential proteins involved © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

in the enterohepatic circulation of lipids [17–19]. Cholesterol absorption is thought to be mainly controlled by distal intestine by jejunal and ileal ABCG5/ABCG8 transporters and NPC1L1 [26, 27]. More than 40 phytosterols, which are highly abundant in vegetable oils, seeds, and cereals, have been identified. The ones used in this study (campesterol, stigmasterol, and b-sitosterol) are more than 95% present in vegetable extracts [28, 29]. These plant sterols were previously chosen to show that, at appropriate doses, they can influence critical steps in cellular cholesterol homeostasis [5, 6, 30]. Our study demonstrated expression changes in the multifunctional ANX family, such as ANXA2, ANXA4, and ANXA5, in the intestinal mucosa of apoE2/2 mice fed with phytosterols. These ANXs have a common ability to interact

www.proteomics-journal.com

Proteomics 2007, 7, 2659–2666

with cellular membranes in a reversible and Ca21-regulated manner. This interaction may reversibly modify membrane properties such as fluidity and permeability, anchoring of cytoskeletal elements, vesicle aggregation, and ion conductance regulation [31, 32]. ANXA2 and ANXA4 have recently been demonstrated to be a part of lipid rafts of the intestinal brush border [33], whereas ANXA5 is expressed specifically by murine M cells [34]. Phytosterols may regulate ANXA4, ANXA5, and ANXA2 through post-transcriptional mechanisms since in the present study this effect at protein level coexisted with unchanged or decreased intestinal expression of these genes in apoE2/2 mice and C57BL/6J mice. The mentioned protein changes may reflect the participation of these ANXs in several unrelated events regulated by phytosterols. ANXA2 has been shown to form a lipid– protein complex with caveolin 1 (CAV1) and cholesteryl esters, which may be involved in the internalization/endocytic transport of cholesteryl esters from caveolae to internal membranes [31, 35, 36]. Pharmacological treatment with the selective hypocholesterolemic agent ezetimibe disrupted this complex in hypercholesterolemic mice [36]. Thus, the downregulation of ANXA2 by phytosterols may decrease complex formation and cholesterol processing, probably inhibiting the cholesterol transport [36]. However, the functional significance of this process is unclear since rabbit’s small intestine does not contain the ANXA2–CAV1 complex [37] and CAV1-deficient mice show normal fractional cholesterol absorption [38]. The inhibitor of intestinal cholesterol absorption, ezetimibe, seems to act upon Niemann-Pick C1like protein [19] and/or aminopeptidase N (CD13) [39]. No changes in the intestinal expression of these two proteins were detected by our analysis, which could have been due to different mechanisms of action of phytosterols and ezetimibe [40] or, alternatively, to the fact that their effect does not depend on changes in the intestinal protein expression [19]. Strong evidence indicates that upon activation of cell death mechanisms, ANXA5 binds with high affinity to the external cell membrane phosphatidylserine on apoptotic tumor cells [41, 42]. Increased ANXA4 levels have also been found in proapoptotic liver cells [43]. ANXA5 and ANXA4 might therefore be involved in tumor cell apoptosis. Thus, phytosterol-mediated regulation of ANXA4 and ANXA5 could help explain the potential protective role of phytosterols in some types of cancer [12, 13]. Further, ANXA4 inhibits Gram-positive bacteria attachment to human macrophages [44] and this could, therefore, be related to the inflammatory modulation properties of dietary phytosterols. It is possible that the increased levels of actinin a-4 and reduced levels of b-actin after phytosterol treatment may, in part, be a consequence of the functional interaction between these cytoskeletal proteins and several types of ANXs [31, 32]. The decrease in the intestinal protein expression of b-actin found in the intestinal mucosa of apoE2/2 mice was corroborated in C57BL/6J mice. However, b-actin intestinal mRNA expression was decreased in apoE2/2 but not in C57BL/6J-phytosterol-fed mice. © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Animal Proteomics

2665

In most of the cases, changes in ANXs studied and bactin protein abundance were found to be higher in Western blot analysis than in 2-D DIGE/MALDI-TOF MS. This could be due to nonlinear chemiluminescent Western blot signal or to differences in specificity against particular epitopes of antibodies used. Phytosterols reduced the abundance of intestinal proteins related to cholesterol biosynthesis such as farnesyl pyrophosphate synthase (FPPS). This change might be the primary response to a decreased cholesterol influx through the enterocyte, which could lead to a decrease in LXR-mediated activation of sterol regulatory element-binding protein-1 (SREBP-1) [45]. In a previous study using microarray-based technology in livers of phytosterol-treated apoE2/2 mice, a two-fold increase was observed in FPPS, which may be indicative of an inverse regulation in liver and intestine, as is the case of ABCG8 [14], or be due to inverse transcriptional and post-transcriptional regulation as observed for ANXA5 in apoE2/2 mice. Of note, disulfide isomerase Erp72 was found to be upregulated after phytosterol treatment. This change might attempt to compensate for the reduced intestinal cholesterol by promoting apoB folding, chylomicron formation, and secretion [46]. The protein and gene expression changes identified in the current study were undetected in the previous global intestinal gene expression analysis using microarrays or suppression subtractive hybridization [14, 47]. By using two alternative internal controls (PGK1 and GAPDH), we ruled out the possibility that the use of b-actin as an internal control in selected RT-PCR analysis [14] accounted for the discrepancies between these studies. Therefore, the discrepancies between our previous microarray study [14] and the present analysis could be due, in part, to the fact that different samples were analyzed (whole intestine vs. intestinal mucosa) [14, 47]. In conclusion, our proteomic analysis of intestinal mucosa in phytosterol-treated apoE2/2 mice permitted the identification of nine differentially expressed proteins involved in processes of plasma membrane structure stabilization, cytoskeleton assembly and networking, and cholesterol metabolism. As the protein expression changes identified in this study might be involved in the biological effects induced by phytosterols, these observations provide a framework for an improved understanding of phytosterol actions. However, further work will be required to distinguish between primary and compensatory changes as well as to define other phytosterol-induced changes in protein structure and function that would not be detected by the proteomic approach used in this study.

We are grateful to Christine O’Hara for the editorial assistance and Pilar Navarro for kindly providing the ANXA2 antibody. This work was funded by FIS grants 03/1058, 05/ 1921, and 05/1267. J. C. E.-G. is a Ramón y Cajal researcher www.proteomics-journal.com

2666

L. Calpe-Berdiel et al.

funded by the Ministerio de Educación y Ciencia and E. Z.-M. a Fondo de Investigaciones Sanitarias researcher (CM 04/ 00029).

5

References

Proteomics 2007, 7, 2659–2666 [23] Shevchenko, A., Wilm, M., Vorm, O., Mann, M., Anal. Chem. 1996, 68, 850–858. [24] Aguilar, S., Corominas, J. M., Malats, N., Pereira, J. A. et al., Am. J. Pathol. 2004, 165, 1129–1139. [25] Ribas, V., Palomer, X., Roglans, N., Rotllan, N. et al., Biochim. Biophys. Acta 2005, 1737, 130–137.

[1] Ncep, E. P., JAMA 2001, 285, 2486–2497.

[26] Duan, L. P., Wang, H. H., Wang, D. Q., J. Lipid. Res. 2004, 45, 1312–1323.

[2] de Jong, A., Plat, J., Mensink, R. P., J. Nutr. Biochem. 2003, 14, 362–369.

[27] van der Veen, J. N., Kruit, J. K., Havinga, R., Baller, J. F. et al., J. Lipid. Res. 2005, 46, 526–534.

[3] Plat, J., van Onselen, E. N., van Heugten, M. M., Mensink, R. P., Eur. J. Clin. Nutr. 2000, 54, 671–677.

[28] Weihrauch, J. L., Gardner, J. M., J. Am. Diet. Assoc. 1978, 73, 39–47.

[4] Kruit, J. K., Groen, A. K., van Berkel, T. J., Kuipers, F., World J. Gastroenterol. 2006, 12, 6429–6439. [5] Yang, C., McDonald, J. G., Patel, A., Zhang, Y. et al., J. Biol. Chem. 2006. [6] Yang, C., Yu, L., Li, W., Xu, F. et al., J. Clin. Invest. 2004, 114, 813–822. [7] Lammert, F., Wang, D. Q., Gastroenterology 2005, 129, 718– 734. [8] Moghadasian, M. H., McManus, B. M., Pritchard, P. H., Frohlich, J. J., Arterioscler Thromb. Vasc. Biol. 1997, 17, 119–126. [9] Moghadasian, M. H., McManus, B. M., Godin, D. V., Rodrigues, B., Frohlich, J. J., Circulation 1999, 99, 1733–1739. [10] Bouic, P. J., Curr. Opin. Clin. Nutr. Metab. Care 2001, 4, 471– 475. [11] Nashed, B., Yeganeh, B., HayGlass, K. T., Moghadasian, M. H., J. Nutr. 2005, 135, 2438–2444. [12] Awad, A. B., Fink, C. S., J. Nutr. 2000, 130, 2127–2130. [13] Awad, A. B., Roy, R., Fink, C. S., Oncol. Rep. 2003, 10, 497– 500. [14] Calpe-Berdiel, L., Escola-Gil, J. C., Ribas, V., Navarro-Sastre, A. et al., Atherosclerosis 2005, 181, 75–85. [15] Escola-Gil, J. C., Julve, J., Marzal-Casacuberta, A., OrdonezLlanos, J. et al., J. Lipid Res. 2000, 41, 1328–1338.

[29] Law, M., BMJ 2000, 320, 861–864. [30] Field, F. J., Born, E., Mathur, S. N., J. Lipid Res. 1997, 38, 348– 360. [31] Gerke, V., Creutz, C. E., Moss, S. E., Nat. Rev. Mol. Cell. Biol. 2005, 6, 449–461. [32] Gerke, V., Moss, S. E., Physiol. Rev. 2002, 82, 331–371. [33] Nguyen, H. T., Amine, A. B., Lafitte, D., Waheed, A. A. et al., Biochem. Biophys. Res. Commun. 2006, 342, 236–244. [34] Verbrugghe, P., Waelput, W., Dieriks, B., Waeytens, A. et al., J. Pathol. 2006, 209, 240–249. [35] Uittenbogaard, A., Everson, W. V., Matveev, S. V., Smart, E. J., J. Biol. Chem. 2002, 277, 4925–4931. [36] Smart, E. J., De Rose, R. A., Farber, S. A., Proc. Natl. Acad. Sci. USA 2004, 101, 3450–3455. [37] Kramer, W., Corsiero, D., Girbig, F., Jahne, G., Biochim. Biophys. Acta 2006, 1758, 45–54. [38] Valasek, M. A., Weng, J., Shaul, P. W., Anderson, R. G., Repa, J. J., J. Biol. Chem. 2005, 280, 28103–28109. [39] Kramer, W., Girbig, F., Corsiero, D., Pfenninger, A. et al., J. Biol. Chem. 2005, 280, 1306–1320. [40] Batta, A. K., Xu, G., Bollineni, J. S., Shefer, S., Salen, G., Metabolism 2005, 54, 38–48.

[16] Schwarz, M., Russell, D. W., Dietschy, J. M., Turley, S. D., J. Lipid Res. 2001, 42, 1594–1603.

[41] Boersma, H. H., Kietselaer, B. L., Stolk, L. M., Bennaghmouch, A. et al., J. Nucl. Med. 2005, 46, 2035–2050.

[17] Minowa, T., Ohtsuka, S., Sasai, H., Kamada, M., Electrophoresis 2000, 21, 1782–1786.

[42] Corsten, M. F., Hofstra, L., Narula, J., Reutelingsperger, C. P., Cancer Res. 2006, 66, 1255–1260.

[18] Altmann, S. W., Davis, H. R., Jr., Zhu, L. J., Yao, X. et al., Science 2004, 303, 1201–1204.

[43] Rodriguez-Ariza, A., Lopez-Sanchez, L. M., Gonzalez, R., Corrales, F. J. et al., Liver Int. 2005, 25, 1259–1269.

[19] Garcia-Calvo, M., Lisnock, J., Bull, H. G., Hawes, B. E. et al., Proc. Natl. Acad. Sci. USA 2005, 102, 8132–8137.

[44] Gotoh, M., Takamoto, Y., Kurosaka, K., Masuda, J. et al., Immunol. Lett. 2005, 98, 297–302.

[20] Mahmood, A., Yamagishi, F., Eliakim, R., DeSchryver-Kecskemeti, K. et al., J. Clin. Invest. 1994, 93, 70–80.

[45] Horton, J. D., Shah, N. A., Warrington, J. A., Anderson, N. N. et al., Proc. Natl. Acad. Sci. USA 2003, 100, 12027–12032.

[21] Alban, A., David, S. O., Bjorkesten, L., Andersson, C. et al., Proteomics 2003, 3, 36–44.

[46] Zhang, J., Herscovitz, H., J. Biol. Chem. 2003, 278, 7459– 7468.

[22] Canals, F., Colome, N., Ferrer, C., Plaza-Calonge Mdel, C., Rodriguez-Manzaneque, J. C., Proteomics 2006, 6, S28–S35.

[47] Jolodar, A., Hourihane, S., Moghadasian, M. H., Nutr. Res. 2005, 25, 847–858.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com