-Anti-atherogenic antioxidants regulate the expression and function of proteasome α -type subunits in human endothelial cells

explored. It found that two compounds, BO-653 and probucol, inhibited the expression of three -type proteasome subunits, PMSA2, PMSA3, and PMSA4 in human umbilical vein endothelial cells. Here we report that both BO-653 and probucol caused not only inhibition of the mRNA levels of these three subunits but also inhibition of both the gene expression and protein synthesis of the α type subunit, PMSA1. Other subunit components of the proteasome such as the β -type subunits (PMSB1, PMSB7), the ATP-ase subunit of 19S (PMSC6), the non-ATP-ase subunit of 19S (PMSD1), and PA28 (PMSE2) were not significantly affected by treatment with these compounds. The specific inhibition of α -type subunit expression in response to these antioxidants resulted in functional alterations of the proteasome with suppression of degradation of multi-ubiquitinated proteins and I- κ B α . These results suggest that certain compounds previously classified solely as antioxidants are able to exert potentially important modulatory effects on proteasome function.


Introduction
Genomics and proteomics offer powerful tools for discovering molecular mechanisms underlying complex biological responses. For example, these methods enable comparison of the levels of gene and protein expression between diseased and normal cells or cells treated with pharmacological agents. Atherosclerosis is an interesting example in this respect since it is a complex chronic inflammatory condition resulting from the interaction between modified lipoproteins, monocyte-derived macrophages, and other cellular elements of artery wall in which numerous pharmacological agents have been tested for efficacy (1,2). Indeed, one important class of molecules assessed in these models are generically classified as "antioxidants", and the anti-atherogenic effects of these compounds have contributed to the evolution of the LDL oxidation hypothesis of atherosclerosis. However, the detailed mechanisms of antioxidant compounds, such as probucol, remain complex and elusive with a number of biological responses identified which are not easily reconciled with a unique mechanism depending solely on free radical scavenging. Recently, we have applied the oligonucleotide microarrays analysis (HuGene human FL array, Affimetrix, Inc., Santa combination of n vitro (3,4) and in vivo models of the atherosclerotic processes (5,6).
With the advent of effective lipid lowering therapies probucol's use as an antiatherogenic agent has become limited. However, the mechanisms underlying probucol's action remain important. For example, recent studies have shown that probucol is effective in reducing the incidence of restenosis after percutaneous transluminal coronary angioplasty (PTCA) (7,8). It has been also reported that probucol inhibits expression of adhesion molecules such as vascular cell adhesion molecule-1 (VCAM-1) both in animal models (9,10) and in HUVEC (11,12). Not withstanding the continuing interest in probucol molecular mechanisms remain obscure with several controversial results with probucol observed in different animal models. For example, probucol effectively inhibited development of atherosclerosis in Watanabe heritable hyperlipidemia (WHHL) rabbits (5,6) while it exacerbated the development of atherosclerosis in apolipoprotein E (apoE) knockout mice (13) and LDL receptor knockout mice (14). The responses in apoE knockout mice were particularly interesting with a recent report showing that lesion development was increased by probucol in the aortic root yet decreased in the aortic arch (15). These studies implicate mechanisms compound, BO-653, is such an example and shares the structural motif for scavenging peroxyl radicals, the phenol group, while the rest of the molecule is structurally distinct from probucol ( Figure 1). It was found that BO-653 exerted a potent antioxidant activity against lipid peroxidation (16) and oxidative modification of LDL (17) and also exhibited anti-atherogenic effects in three different animal models (14). Surprisingly, it is about 10 times more effective than probucol as an antioxidant yet still requires similar concentrations to exert anti-atherogenic properties. We hypothesized that properties other than the antioxidant activity of these compounds could contribute to inhibition of atherosclerosis.
The use of genomic analysis to elucidate the mechanisms of pharmacological agents is an important application of this emerging technology. Indeed, a novel and unexpected finding derived from gene chip analysis (18) indicated that regulation of proteasome function could occur on exposure of HUVEC to phenolic antioxidants.
The ubiquitin-proteasome pathway has been shown to be involved in various biologically important processes, such as the cell cycle, cellular metabolism, apoptosis, -7 - (23). The regulatory complex termed PA700 (also called the "19S complex") associates with 20S proteasome to form the 26S proteasome with a molecular mass of ~2500 kDa (24). Moreover, another proteasome activator PA28 (11S REG), consisting of α and β subunits, is known to bind to α-ring of the 20S proteasome (19). In this report we show that the selective inhibition of the α-type subunits of the 20S proteasome in Gene Chip TM analysis. Oligonucleotide microarray analysis was performed by using gene chip, the HuGene human FL array (Affymetrix Inc., Santa Clara, CA) as described (18).

Northern blot analysis (RNA Blot Hybridization).
Total RNA was extracted by the method of Chomczynski et al. (26). Samples of 5 µg of total RNA were denatured and separated by electrophoresis on a 1 % agarose gel containing 2.2 % formaldehyde. Total RNA was then transferred to Hybond-XL nylon membrane (Amersham pharmacia bioteck, Buckinghamshire, UK) and hybridized with 32 P-labeled probes as described previously (27). For the Northern blot analysis of proteasomes, the cDNAs for the subunits PMSA1 (HC2), PMSA2 (HC3), PMSA3 (HC8) and PMSA4 (HC9) of human proteasomes were constructed according to the literature (28) and the cDNA of glyceraldehyde-3-phospate dehydrogenase (GAPDH) was used as a standard. These probes were labeled with a random primer labeling kit, the BCA BEST TM Labeling kit (Takara, Shizuoka, Japan), using [α-32 P] dCTP. The membranes were washed as described previously (29) and autoradiographed with a BAS-1800 (Fuji Photo Film Co., Tokyo, Japan). To measure the stability of proteasome subunit mRNA, cells were treated with Actinomycin D (5 µg / ml), an inhibitor of transcription of mRNA at the same time as antioxidant treatment as described (30).

Results
Phenolic Antioxidant-dependent modulation of the level of mRNAs encoding proteasomal α-type subunits: Gene chip analysis revealed a decrease in the levels of mRNAs encoding proteasome α-type subunits, PMSA2(HC3), PMSA3(HC8) and PMSA4(HC9) in HUVEC exposed to the phenolic antioxidants, probucol (50 µM) and BO-653 (50 µM) for 6 hours (Table1). These concentrations are equivalent to those achieved in both human and animal studies (14,25). The gene chip analysis can result in false positives, and thus the responses were carefully validated and extended in the first series of experiment using Northern blot analysis. It is clear that treatment of HUVEC with either BO-653 or probucol results in a decreased expression of these three genes at rate of mRNA degradation as shown for PMSA2 (Fig. 2 F). The decreased mRNA for PMSA1 on exposure to the antioxidants was also found to result in decreased protein levels of PMSA1 after a 12 hour exposure, with levels making a partial recovery to control values at 24 hours (Fig. 3) Table 2).

Specificity of regulation of proteasome subunits by BO-653 and probucol:
From the data obtained from the gene analysis (Table 1) it was evident that only the α-type proteasome subunits were down-regulated significantly in HUVEC treated with BO-653 or probucol. To further assess the specificity of this response, we investigated changes in the β-type subunits (PMSB1 (HC5) and PMSB7 (Z)) in response to BO-653 and probucol by means of Northern and Western blotting. Neither subunit exhibited significant changes. The data of mRNA and protein levels for PMSB7 are shown after exposure to either BO-653 or probucol for 6 hours and 12 hours, respectively (Fig. 4 A   and B). In addition, the effects of BO-653 and probucol on protein levels of PMSC6 (ATP-ase subunit of 19S), PMSD1 (non-ATP-ase subunit of 19S), and PMSE2 (PA28) were determined and found to be unchanged by treatment of cells with these antioxidants over a period of 12 hours (Fig. 4 C, D and E). These results are consistent with those of the gene chip experiment (Table 1). These data support the hypothesis that both BO-653 and probucol show selectivity in modulating the expression of the α-type subunits.
Accumulation of the proteasome target multi-ubiquitinated proteins and I-κBα in HUVEC treated with antioxidants: These data suggest that BO-653 and probucol have a functional impact on proteasome activity. We investigated this further by determining the accumulation of multi-ubiquitinated proteins and I-κBα, well-known target proteins of the proteasome. HUVECs were pre-incubated with the protein synthesis inhibitor cycloheximide for 2 hours, followed by additional incubation with or without either BO-653 or probucol to determine if the stability of ubiquitinated proteins and I-κBα was increased. Incubation of HUVEC with cycloheximide alone significantly decreased levels of both multi-ubiquitinated proteins and I-κBα compared to nontreated cells (Fig. 5, lane 2 vs lane 1). This decrease is thought to be due to degradation of these proteins in the absence of new protein synthesis. In contrast, co-incubation of the cells with either BO-653 or probucol in the presence of cycloheximide maintained the levels of multi-ubiquitinated proteins and I-κBα at control levels. (Fig, 5, lane 3 and 4 vs lane 1). This increase is consistent with inhibition of degradation of these proteins by these antioxidants, but cannot be due to increased protein synthesis.

Discussion
Taken together these data provide a novel and interesting perspective on the biological properties of phenolic lipid peroxyl radical scavengers. The two compounds selected for study have both been shown to be anti-atherogenic in animal models, yet differ markedly in radical scavenging ability, with probucol being about 10 times less effective than BO-653 (4,14,16). The unforeseen property revealed by genomic analysis was an apparent modulatory effect on the transcriptional regulation of subunits of the proteasome. Such an effect would be particularly significant in atherosclerosis, since nuclear factor κB (NF-κB) expression is increased in human atherosclerotic lesions (31), and is thought to make a significant contribution to the development of the inflammatory process. The ubiquitin-proteasome pathway degrades the inhibitory binding protein, I-κBα, which ordinarily keeps NF-κB sequestered in the cytoplasm.       The numbers (mean +/-SD ) show a ratio against control (DMSO-treatment).
mRNA and protein levels of PMSA1