Yu YL et al / Acta Pharmacol Sin 2003 Sep; 24 (9): 873-877

Protein expressions in macrophage-derived foam cells: comparative analysis by two-dimensional gel electrophoresis¹

YU Yan-Ling^2,3, YANG Peng-Yuan^4,5, FAN Hui-Zhi², HUANG Zhen-Yu², RUI Yao-Cheng^4,6, YANG Peng-Yuan^2,6

²Department of Chemistry, Fudan University; ³Department of Pharmaceutical Analysis,

⁴Department of Pharmacology, School of Pharmacy, Second Military Medical University, Shanghai 200433, China

¹ Project supported by the National 973 Basic Research Project (No 2001CB510202), and the National Natural Science Foundation of China (No 39870870).

⁵ Contribute equally to this work as the first author.

⁶ Correspondence to Prof RUI Yao-Cheng. Phn/Fax 86-21-2507-4471. E-mail Ruiyc@smmu.edu.cn Prof YANG Peng-Yuan. Phn 86-21-6564-2009. Fax 86-21-6564-1740. E-mail pyyang@fudan.edu.cn

Received 2002-07-18 Accepted 2003-03-06

KEY WORDS foam cells; lipoproteins LDL; two-dimensional gel electrophoresis

ABSTRACT

AIM: To study the overall protein expression changes induced by oxidized low-density lipoprotein (ox-LDL) in the U937 foam cells. METHODS: Foam cell model was established by incubating the human monoblastic leukemia (U937) cells with ox-LDL. Each protein samples in U937 control cells and U937 foam cells were separated by two-dimensional gel electrophoresis (2-DE). After the gels were stained by silver staining method, the images were analyzed by PDquest 2D-image-analysis software (Bio-Rad). Some of the spots were available via the Internet with links to the U937 proteomic map provided from the ExPASy Proteomics server. RESULTS: Using 2-DE, the overall protein map was obtained, in which 150 spots were matched with the control gel (match ratio: 75 %). Compared with U937 cells, 37 spots significantly changed in the foam cells (P<0.05), among which the expression levels in 28 spots increased and those in 9 spots decreased. Especially, 8 spots in U937 cells were absent in the foam cells, while 11 spots in the foam cells were absent in the control cells. CONCLUSION: The changed protein profiles induced by ox-LDL in U937 foam cells were established to support the functional studies on the macrophage-derived foam cells in atherosclerotic pathological states.

INTRODUCTION

The term proteome was recently coined to describe all the proteins encoded from a specific genome^[1]. Proteomic analysis involves the qualitative alterations in proteins along with the quantitative changes in protein expression levels that occur in response to a given set of conditions. Despite the fact that proteomics-based research has only recently begun to be used, it is fast approaching a crossroad with respect to future utilization. The use of two dimensional gel electrophoresis (2-DE) is the best technique to separate complex protein mixtures^[2]. Using 2-DE, the investigation of the cellular or molecular "big picture" with the use of proteomics may be one of the best approaches to elucidate the complex and multi-factorial basis for many aspects of cardiovascular biology, especially disease processes.

Atherosclerosis is not merely a disease in its own right, but a process that contributes principally to the pathogenesis of myocardial and cerebral infarction, gangrene and loss of function in the extremities. During the process of atherosclerosis, monocytes seem to play a central role. Once monocytes adhere to the subendothelial space and enter into the intima of the artery, oxidized low-density lipoprotein (ox-LDL) and other substances associated with atherosclerosis may participate in transformation of the monocytes in macrophage. Uptake of ox-LDL by the macrophage through scavenger receptors will lead to foam cell formation^[3,4].

The macrophage-derived foam cells not only result in formation of fatty streaks, which are believed to represent the earliest type of atherosclerotic plaque, but also play roles in the fibroproliferative process by their capacity to form numerous growth factors in particular platelet-derived growth factor (PDGF)^[5], vascular endothelial growth factor (VEGF)^[6,7], interleukin-1 (IL-1)^[8], as well as tumour necrosis factor (TNF)^[9]. Therefore, a further study on the global changes in protein expression accompanying the formation of foam cells in vivo becomes an urgent mission to be investigated.

In our previous studies, we established a macrophage-derived foam cell model through incubating the human monoblastic leukemia (U937) cells with oxLDL^[10,11]. The ICAM-1 and VEGF expression levels have been determined to be the same as the foam cells in the atherosclerotic plaques, which confirm the successful establishment of the pathological model. The aim of this study was to further investigate the global changes in protein expressions between the U937 foam cells and the control U937 cells using 2-DE.

MATERIALS AND METHODS

Reagents CuSO₄ and edetic acid were purchased from Sigma Chemical Co; bisacrylamide (Bis), hydroxymethyl aminomethane (Tris), sodium dodecyl sulfate (SDS), glycine, N,N,N',N'-tetramethylethyl-diamide (TEMED), ammonium persulfate (APS), glycerol, ultrapure urea, and bromophenol blue were purchased from Pharmacia Biotech; acrylamide, dithiothreitol (DTT), and iodoacetamide (IAA) were purchased from Fluka BioChemika; All the water used in the experiments was 18-M deionized from a MilliQ water purification system.

Cell culture LDL (d=1.019 to 1.063 kg/L, purchased from Sigma Co, USA) was sterilized by filtration through 0.45-µm Millipore membranes, and stored at 4 ºC as described previously^[12]. After edetic acid was removed by dialysis, LDL was oxidized by incubation in CuSO₄ 10 µmol/L at 37 ºC for 16 h, and then dialyzed in phosphate buffer saline (PBS) containing edetic acid 0.1 mmol/L at 4 ºC for 24 h.

The human monocyte line U937 was obtained from Cell Bank in Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences. U937 cells were cultured in RPMI-1640 containing 10 % fetal bovine serum. Then the cells were incubated with ox-LDL 80 mg/L for 48 h, and the control group was the U937 cells which were not treated with ox-LDL.

Protein extraction and quantification The cells were disrupted by ultrasonication (Sonics & Materials Inc, Vibra Cell^TM) twice, each for 20 s, on ice in extraction solution which containing urea 8 mol/L, 4 % 3[3 (cholamidopropyl) dimethylammonio]-1-propane-sulfonate (CHAP), and Tris 40 mmol/L. The resulting homogenate was centrifuged at 12 000×g for 10 min. The supernatant was fractioned in aliquots and stored at -20 ºC till further analyzed.

Protein quantitation was performed similarly to the Bradford method with some modifications. Briefly, using bovine serum albumin (BSA) in Tris solution as standard protein solution, furthermore, a small quantity of HCl (0.1 mol/L) was added in the reaction system with the reagent blank experiment to calibrate. The standard and samples were prepared in duplicates to minimize some occasional error during the course.

Two-dimensional gel electrophoresis Each sample of 40 µg protein was added to a rehydration solution containing urea 8 mol/L, 4 % CHAPs, Tris40 mmol/L, DTT 50 mmol/L, 0.2 % ampholyte (pH=3-10), and trace bromophenol blue. Immobilized dry strips (Amersham Pharmacia), 7 cm, linearly covering a pH range of 3-10, were allowed to rehydrate in 130 µL the above solution for 12 h under mineral oil. The following voltage/time profile was used: increasing voltage by gradient from 0-250 V for 30 min, increasing voltage by step-n-hold procedure from 250-4000 V for 3 h, and a final phase of 4000 V for 8000 Vh.

After the first dimensional run, the individual strip was equilibrated, and then put on the top of the second dimensional gel (12 %). Each gel was run at 10 mA at the beginning and 15 mA per gel was used till all the proteins being transferred from strip into gel. The run was completed until the bromophenol blue front reaching the bottom of the gel.

Silver staining and image analysis The gels were stained by silver staining method similarly to the method described^[13] with modifications. Images were captured by scanning the silver-stained gels using Molecular Imager Fx (Bio-Rad), and then processed by PDquest 2D-image-analysis software (Bio-Rad).

Protein bioinformatic analysis In addition to the PDquest software packages for analyzing the electrophoretic separation, bioinformatic tools were used. Some of the spots were available via the Internet with links to the U937 proteomic map provided from the ExPASy Proteomics server (http://us.expasy.org).

Statistics All results were shown as mean±SD. Statistical analysis was performed with t-test in the PDquest software.

RESULTS

Two-dimensional protein maps of the control U937 cells and foam cells All the gels were normalized against one in the U937 cell group as the control. A typical analytical 2-D gel from U937 cell line was shown in Fig 1A. Through image processing by PDquest, more than 1000 spots were detected, among which 200 spots were accorded with Gaussian distribution. Most of the spots were centralized in the area of acidic region (pI: 4-7). Through the 2-DE protein database of the U937 in the ExPASy Proteomics server, 6 spots were identified.

The foam cells were incubated four times in the same condition (n=4). The 2-D patterns obtained after 2-DE separation of 40 µg protein sample from U937 foam cells (Fig 1B). Through the PDquest software, 150 spots were matched with the spots in the control gel (match ratio: 75 %).

Fig 1. The two-dimensional protein maps of the U937 cells and foam cells. Protein extractions in U937 cells were separated by 2-DE and analyzed by the PDquest software (n=4). The annotated protein spots were provided from ExPASy Proteomics server. A) U937 control cells; B) U937 foam cells. : the protein spots only found in U937 control cells; : the spots only found in U937 foam cells; : the expression levels increased in foam cells compared with control cells; : the levels decreased in foam cells compared with control.

Proteomic comparison analysis on the U937 foam cells Compared with control U937 cells, 37 spots were found significantly changed in the foam cells (P<0.05, Tab 1). Among them, 8 spots were absent in the U937 foam cells compared with the normal U937 cells. However, 11 spots found only in the U937 foam cells showed no expression in control U937 cells.

Tab 1. The significant changed protein spots in the U937 foam cells compared with U937 control cells analyzed by 2-DE. n=4. Mean±SD. ^bP<0.05 vs U937control cells.

DISCUSSION

Foam cells are the characteristic pathological cells in the atherosclerotic lesions, which represent the inflammatory reactions in atherosclerosis. The U937 foam cells were well established as a macrophage-derived foam cell model, which showed that U937 cells internalized ox-LDL through scavenger receptor and became foam formation.

2-DE is considered to be the key separation technique in proteome analysis due to its advantages of simultaneously separation of thousands of proteins at a time and excellent reproducibility. In the procedure of 2-DE analysis, we used silver staining method proposed by Andrej S et al in 1996 because of its high sensitivity together with its compatibility with mass spectrometric identification further. During the staining course as described previously^[13], we found that the removal of acetic acid used for fixing proteins was very important to reduce the background and increase the contrast between the stained protein spots and grounding. So we increased washing times for the acid removal, but decreased the period of each washing turn to balance the overall time used.

The annotated proteins and 2-DE databases are the bioinformatics core of proteome research. The information from these database is invaluable to both researchers, novice, and expert alike, wishing to obtain indirect protein identification in a comparable 2D image of validation of a protein previously identified through analytical analysis. However, the available spots provided from the ExPASy Proteomics server are not significant changed during the formation of the foam cells. Since most of the annotated proteins in the 2-DE database are structure proteins, ox-LDL may have few effects on the expressions of the structure proteins during the foam cell formation. Subsequent studies should succeed in identifying additional proteins, thereby enhancing the list of already characterized proteins found in respective database.

The importance of 2-deprotein database has led to a major initiative in our laboratory to establish several databases to support several ongoing functional studies. The present studies provided a 2-DE protein database of the foam cells, which gave a further analysis on the complete complement of proteins expressed by a biological system in response to atherosclerotic pathophysiological conditions. Therefore, this finding develops a novel way to elucidate the cellular and molecular mechanisms associated with atherosclerotic pathological states.

Proteomic method has been proven to be an efficient way to investigate different pathological processes resulted from different conditions. Analysis by 2-DE just gave us the primary screening to find variant expressions during the formation of foam cells. The target proteins (Tab 1) remain to be identified with mass spectrometry (MS) techniques. Further studies are focusing on both the target protein identifications and their corresponding functional characteristics.

REFERENCES

• 1 Wilkins MR, Scanchez JC, Gooley AA, Appel RD, humphery SI, Hochstrasser DF, et al. Progress with protome projects: why all proteins expressed by a genome should be identified and how to do it. Biotechnol Genet Eng Rev 1996; 13: 19-50.
• 2 Macri J, Rapundalo ST. Application of proteomics to the study of cardiovascular biology. Trends Cardiovasc Med 2001; 11: 66-75.
• 3 Berliner JA, Heinecke JW. The role of oxidized lipoproteins in atherogenesis. Free Radic Biol Med 1996; 20: 707-27.
• 4 Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 1993; 362: 801-9.
• 5 Ross R, Masada J, Raines EW, Gown AM, Katsuda S, Sasahara M, et al. Location of PDGF- protein in macrophage on all phases of atherogenesis. Science 1990; 248:1009-12.
• 6 Inoue M, Itoh H, Ueda M, Naruko T, Kojima A, Komatsu R. Vascular endothelial growth factor (VEGF) expression in human coronary atherosclerotic lesions: possible pathophysiological significance of VEGF in progression of atherosclerosis. Circulation 1998; 98: 2008-16.
• 7 Ramos MA, Kuzuya M, Esaki T, Miura S, Satake S, Asai T, et al. Induction of macrophage VEGF in response to oxidized LDL and VEGF accumulation in human atherosclerotic lesions. Arterioscler Thromb 1998; 18: 1188-96.
• 8 Libby P, Friedman GB, Salomon RN. Cytokines as modulators of cell proliferation in fibrotic diseases. Am Rev Respir Dis 1989; 243: 393-6.
• 9 Libby P, Ordovas JM, Auger KR, Robbins AH, Birinyi LK, Dinarello CA. Endotoxin and tumour necrosis factor induce interleukin-1 gene expression in adult human vascular endothelial cells. Am J Pathol 1986; 124: 179-85.
• 10 Yang PY, Rui YC, Zhang L, Li TJ, Qiu Y, Wang JS, et al. Expression of vascular endothelial growth factor in U937 foam cells and the inhibitory effect of drugs. Acta Pharm Sin 2002; 37: 86-9.
• 11 Yang PY, Rui YC, Li K, Huang XH, Jiang JM, Yu L. Expression of intercellular adhesion molecule-1 in U937 foam cells and the inhibitory effect of imperatorin. Acta Pharmacol Sin 2002; 23: 327-30.
• 12 Kuzuya M, Naito M, Funaki C, Hayashi T, Asai K, Kuzuya F. Lipid peroxide and transition-metals are required for the toxocoty of oxidized low-density-lipoprotein to cultured endothelial-cells. Biochim Biophys Acta 1991; 1096: 155-61.
• 13 Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Anal Chem 1996; 68: 850-8.