Human Aldehyde Oxidase 1 Interacts with ATP-binding Cassette Transporter-1 and Modulates its Activity in Hepatocytes
| Author | A. Sigruener1, C. Buechler1, E. Orsó1, A. Hartmann2, P. J. Wild3, L. Terracciano4, M. Roncalli5, S. R. Bornstein6, G. Schmitz1 |
| Affiliation | Affiliation addresses are listed at the end of the article |
Key words
· AOX1 · ABCA1
HepG2 siRNA
· efflux
Abstract
& AOX1, a member of the cytosolic molybdenum hydroxylase family, has been identified by us earlier as an ABCA1-interacting protein. AOX1 is well-described as xenobiotic metabolizing enzyme, which upon oxidation of acetaldehyde and retinaldehyde to acetic acid and retinoic acid generates reactive oxygen species. Here we show that knock-down of AOX1 in HepG2 by small interfering RNA significantly reduced ABCA1- dependent lipid efflux and enhanced phago- cytic uptake of microspheres similar to ABCA1 deficiency, without affecting ABCA1 mRNA and protein levels. ABCA1 and AOX1 are coexpressed in human hepatocytes, kidney proximal tubular epithelial cells, Leydig, and adrenocortical cells.
Expression of ABCA1 and AOX1 was investigated by immunohistochemistry in liver tissue arrays. A strong AOX1 expression was found in normal liver, and in cirrhosis. In contrast, hepatocellu- lar carcinomas showed either a complete loss or reduced expression of AOX1. Significant correla- tions were found between reduced AOX1 expres- sion and tumor stage, or metastatic or regional lymph node states. Deregulation was also observed for ABCA1 expression but to a lesser extent. Our findings show that the interaction of ABCA1 with AOX1 modulates ABCA1-linked cel- lular functions such as lipid efflux and phagocy- tosis in hepatocytes, and the reduced expression of AOX1 in malignant transformed hepatocytes supports the differentiation dependent upregu- lation of AOX1.
Downloaded by: University of Pittsburgh. Copyrighted material.
received 15.05.2007 accepted 15.07.2007
Bibliography DOI 10.1055/s-2007-992129 Horm Metab Res 2007; 39: 781-789 @ Georg Thieme Verlag KG Stuttgart . New York ISSN 0018-5043
Correspondence
Prof.Dr. med. G. Schmitz Institute of Clinical Chemistry Regensburg University Medical Center Franz-Josef-Strauss-Allee 11 93042 Regensburg Germany Tel .: +49/941/944 62 01 Fax: +49/941/944 62 02 gerd.schmitz@klinik.uni- regensburg.de
Introduction
& The ATP binding cassette transporter-1 (ABCA1) plays a key role in cellular lipid efflux and reverse cholesterol transport by generating lipid poor preß-high density lipoprotein (HDL) precursors [1]. Loss of functional ABCA1 results in HDL-defi- ciency syndromes [2], and severely disturbed cellular apolipoprotein A-I (apoA-I)-mediated cholesterol- and phospholipid efflux.
Several proteins have been described that influ- ence ABCA1-linked functions. The PSD-95/Dlg/ ZO-1 domain proteins œ1-syntrophin and ß1-syntrophin enhance lipid efflux by stabilizing ABCA1 [3,4]. ABCA1 was also reported to interact with a ß2-syntrophin/utrophin complex, puta- tively linking it to the actin cytoskeleton [5]. Fur- thermore, ABCA1-mediated lipid efflux is influenced by its interaction with Fas-associated via death domain (FADD) [6] and cdc42 [7-9]. ADP-ribosylation factor-like protein 7 (Arl7) has also been demonstrated to influence lipid efflux, probably in concert with ABCA1 [10]. More
recently, syntaxin 13 was identified to bind ABCA1, and knock-down of syntaxin 13 reduced ABCA1 protein expression and lipid efflux [11]. Screening of a human liver yeast-two-hybrid library with the carboxy-terminal 144 of the amino acids of ABCA1 identified additional puta- tive ABCA1-interaction partners, such as alde- hyde oxidase 1 (AOX1) [5]. AOX1 (EC 1.2.3.1) is a member of the family of cytosolic molybdenum hydroxylases [12] and detectable in the liver and other tissues of humans and other mammalians [13, 14]. The physiological function of AOX1 is not yet fully elucidated, but it is significantly involved in the metabolism of xenobiotics [15]. AOX1 has been suggested to play a role in ethanol-induced liver injury [16] and the generation of reactive oxygen species (ROS) during ethanol metabolism [17,18]. AOX1 is identical to retinal oxidase (EC 1.2.1.36) [19], thus suggesting retinal as physio- logical substrate of AOX1. Indeed, retinal dehy- drogenase and AOX1 are capable of catalyzing the oxidation of retinal to the transcriptional cofactor retinoic acid (RA), but they require dif-
ferent cofactors [20]. Moreover, RA is an important signaling molecule implicated in mesenchymal/epithelial interactions in pancreatic development [21]. AOX1 is highly expressed in nor- mal acinar cells of human pancreas. However, decreased AOX1 expression in chronic pancreatitis, retained expression in tubule formations in the course of acinoductal metaplasia, and a com- plete loss of expression in malignant pancreatic cells were found [22]. Furthermore, thrombin induced NAD(P)H oxidase-medi- ated downregulation of AOX1 in vascular smooth muscle cells (VSMC) was recently described in restenosis and atherosclerosis mouse models [23].
In a yeast-two-hybrid screen of a human liver library, fourteen AOX1 clones (common region: amino acids 972 to 1338 [EMBL- Genbank, Accession No. NP_001150]) were found to interact with the ABCA1 C-terminus [5]. In this work, we analyzed the interaction of AOX1 with ABCA1 in more detail and its potential influence on ABCA1-related cellular functions.
Materials and Methods
&
Materials
Chemicals were obtained from Sigma unless otherwise stated. Mouse monoclonal anti-ABCA1 (AB.H10) and horseradish per- oxidase-conjugated goat anti-glutathione S-transferase (GST) antibodies were purchased from Abcam. Mouse monoclonal IgG against AOX1 was from BD Biosciences and anti-HA monoclonal IgG from Clontech. Apolipoprotein A-I (apoA-I) was purchased from Calbiochem.
Cell culture
HepG2 cells were grown in an incubator at 37℃ (95% humidity, 10% CO2) in DMEM (Bio WHITTAKER) containing 10% FCS (v/v), 1% MEM (v/v) (both Gibco BRL) and glutamine until 80% conflu- ence. After trypsination, cells were split 1:12 once a week.
Recombinant expression and purification of ABCA1 C- terminus
ABCA1 C-terminus was cloned from bp 6352 to 6786 (EMBL- Genbank, Accession No. AB055982) in fusion with GST in pGEX- 5X-1 and the fusion protein was purified with glutathione sepharose 4B beads in a batch procedure following the manufac- turer’s instructions (Amersham Pharmacia Biotech).
Pull down assays
In vitro translation was performed according to the manufactur- er’s protocol (Promega). Equal amounts of GST or GST-ABCA1 were incubated with the in vitro translated AOX1 fragment [amino acids 972-1338 (EMBL-Genbank, Accession No. NP_ 001150)] in frame with an N-terminal HA-Tag overnight at 4℃ while shaking in a total volume of 500 ul phosphate buffered saline (PBS) and 0.1% NP40. The reactions were cleaned up with the GST Micro Spin Purification Module (Amersham Pharmacia Biotech) following the manufacturer’s instructions. Washing was carried out in PBS, containing 0.1% NP40. Bound proteins were eluted in Laemmli buffer and analyzed by immunoblot.
RNA isolation
Cells were harvested, washed in PBS, resuspended in buffer RLT and isolated using the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Human tissue RNA-samples were obtained from Clontech.
TaqMan PCR
TaqMan PCR assays were performed on the ABI Prism 7900 HT Sequence Detection System (PerkinElmer Applied Biosystems). Relative quantification was carried out as described earlier [24].
Protein isolation
Cells were harvested, washed in PBS and resuspended in the mentioned lysis buffer followed by incubation for 15 minutes at 4℃ and additional 15 minutes at 25℃. Cell debris was sepa- rated by centrifugation at 14000g at 4℃ for 10 minutes.
Coimmunoprecipitation
HepG2 cells were lysed in PBS with protease inhibitor mixture (Roche Molecular Biochemicals), 0.5% Triton X-100 as described above. Three µg of the supernatant were incubated with magne- tobeads coupled with Protein-A (Dynal) linked to monoclonal anti-AOX1 antibody overnight at 4℃ while shaking. Washing was performed in PBS with protease inhibitor mixture (Roche Molecular Biochemicals), and 0.5% Triton X-100. Bound proteins were eluted in Laemmli buffer and analyzed by immunoblot.
Immunoblot
Proteins were separated by SDS-polyacrylamide gelelectro- phoresis and transferred to PVDF membranes. For the immuno- blot of AOX1 in different human tissues a commercial blot from Imgenex was used according to the manufacturer’s instructions. Incubations were performed in PBS containing 5% nonfat dry milk, and 0.1% Tween. Detection of immune complex was car- ried out using ECL Plus chemiluminescence substrate (Amer- sham Pharmacia Biotech).
Knock-down of AOX1
Nonsilencing control 1 and siRNA duplex targeting AOX1 (EMBL- Genbank, Accession No. NM_001159, bp 312-330, AOX1 siRNA: sense: GCCAAUGCCUGUCUGAUUCtt; antisense: GAAUCAGACAG- GCAUUGGCtg) were ordered from Ambion. HepG2 cells were seeded in a 12-well plate at a density of 106 cells/well. Twenty- four hours after seeding AOX1 or nonsilencing control, siRNA was added using RNAifect (Qiagen) and the method was optimized according to the manufacturer’s instructions. Functional studies were started 24 hours (lipid efflux studies) or 48 hours (phago- cytic activity) after transfection.
Phagocytic activity
Phagocytic capacity was determined by flow cytometry. HepG2 cells were incubated with fluoresbrite yellow green micro- spheres, 0.75 um (Polysciences Europe GmbH) for 1 hour, washed two times with PBS and thereafter fresh medium was added. Cells were incubated for 2 hours, washed with PBS and fluores- cence intensity of cells was determined by a FACSCalibur flow cytometer (BD Biosciences).
Assessment of apoA-I-mediated cholesterol and phospholipid efflux
Cells were labeled with [3H]choline and [14C]cholesterol (Amer- sham Pharmacia Biotech) for 31 hours, and subsequently incu- bated in the presence of bovine serum albumin (BSA) or BSA plus apoA-I for 17 hours and processed as described previously [25].
Immunohistochemistry
Immunohistochemical studies for the expression of ABCA1 and AOX1 utilized an avidin-biotin peroxidase method with a 3,3’- diaminobenzidine (DAB) chromatogen. Four um sections were cut from formalin-fixed and paraffin wax-embedded tissues. After deparaffinization for 10 minutes in xylene, tissue sections were rehydrated in descending ethanol series. After antigen retrieval (microwave oven for 10 minutes at 250 W in citrate buffer [pH 7.1]) IHC was carried out in a NEXES immunostainer (Ventana) following manufacturer’s instructions. The following primary antibodies were used: anti-ABCA1 (dilution 1:10), and anti-AOX1 (dilution 1:50). For target proteins the ChemMate detection kit (DAKO) was used, and finally the slides were coun- terstained with hematoxylin for 1 minute. Incubation with the primary and secondary antibody was performed according to the manufacturer’s protocol. Normal hepatocytes were chosen as internal positive control for ABCA1 and AOX1 IHC. A surgical pathologist (A.H.) performed the evaluation of the slides.
Tissue microarray (TMA)
ABCA1 and AOX1 protein expression in hepatocellular carcino- mas was assessed using TMAs. A commercially available liver tumor TMA (LV801) was obtained from BioCat (Heidelberg, Ger- many). Additionally, a hepatocellular carcinoma TMA from the Institute of Pathology, University of Basel, Switzerland was used, containing a consecutive series of 233 nonselected, formalin- fixed, paraffin-embedded primary hepatocellular carcinomas, 119 specimens with liver cirrhosis, and 18 normal liver samples. Before construction of the TMA, an experienced surgical pathol- ogist (L.T.) with expertise in liver pathology evaluated H & E- stained slides of all specimens to identify representative areas. In the process of histological review all carcinomas were graded using a three-tired nuclear grading system (G1-3) according to WHO criteria (2000). Of the 388 samples on the TMA, represent- ative tumor tissues for AOX1 and ABCA1 IHC were available in 75.5 (289/388) and 69.3% (269/388) of cases, respectively.
Statistical analysis
Statistical analyses were completed using SPSS version 13.0 (SPSS, Chicago, IL). Differences were considered statistically sig- nificant when p values were <0.05. Contingency table analysis and two-sided Fisher’s exact tests were used to study the statis- tical association between clinicopathologic and immunohisto- chemical parameters. In case of continuous variables, the Student’s t-test was used.
Results
&
AOX1 is an ABCA1 interacting protein
We confirmed our previous yeast-two-hybrid data [5] by a pull- down assay using a recombinant GST-fusion protein of the ABCA1-C-terminus (GST-ABCA1) and the in vitro translated human AOX1-fragment found common in the yeast-two-hybrid screen. Equal amounts of purified GST-ABCA1 or GST were immobilized on glutathione sepharose beads and incubated with HA-tagged AOX1. Binding of AOX1 was investigated by immunoblot using a monoclonal anti-HA-Tag antibody. As shown in . Fig. 1A GST-ABCA1 binds significant amounts of AOX1, in contrast to GST alone, indicating a direct interaction of ABCA1 and AOX1 in vitro.
A
GST-ABCA1
-C-Term
GST
AOX1
3
MCF-7
BeWo
PC3
Caco-2
HT-29
T84
HepG2
HeLa
AOX1
C
IP AOX1
Control
ABCA1
AOX1
To further confirm the interaction of ABCA1 and AOX1 in whole cells, different human adenocarcinoma cell lines (MCF-7 breast, PC3 lung, CaCo-2 colon, HT-29 colon), carcinoma cells (T84 colorectal, HepG2 hepatocellular, HeLa cervix) and placental choriocarcinoma BeWo cell line were analyzed by immunoblot for the expression of AOX1. As shown in . Fig. 1B only HepG2 cells showed clearly detectable levels of AOX1 protein, and therefore were chosen for further analysis of the AOX1/ABCA1 interaction. Using these cells ABCA1 could be precipitated from HepG2 cell lysates using a monoclonal anti-AOX1-anti- body, indicating that ABCA1 also interacts with AOX1 in vivo (· Fig. 1C).
Knock-down of AOX1 reduces ABCA1-related cellular functions
The physiological role of the AOX1-ABCA1 interaction was fur- ther elucidated in HepG2 cells by a siRNA approach. AOX1 siRNA significantly decreased AOX1 mRNA and protein levels at 100 nM (· Fig. 2A,B). Higher concentrations of siRNA (up to 300 nM) resulted in no further decrease of protein (data not shown). The decrease of AOX1 protein was detectable between 24 and 72 hours after siRNA transfection by immunoblots. ABCA1 mRNA and protein levels were unaffected in response to AOX1 siRNA treatment (· Fig. 2A,B). Furthermore, knock-down of ABCA1 in HepG2 did not change AOX1 levels (data not shown).
Regarding the crucial role of ABCA1 in apoA-I-dependent cellu- lar cholinephospholipid and cholesterol efflux we determined
Downloaded by: University of Pittsburgh. Copyrighted material.
A
SİRNA
nonsilencing
AOX1 siRNA
ABCA1
AOX1
ß-Actin
B
mRNA Expression
140
relative mRNA expression (%)
120
100
1
王
80
*
60
40
20
0
AOX1
ABCA1
the effect of AOX1 knock-down on lipid efflux. HepG2 cells treated with AOX1 siRNA showed significantly reduced efflux for cholinephospholipids (66±27%) and cholesterol (51 +31%) com- pared to controls ( Fig. 3A), while no significant differences were detectable in cellular choline- and cholesterol content under basal conditions (data not shown). These results indicate that AOX1 not only interacts with ABCA1 but also enhances the functional activity of ABCA1 by a yet unknown mechanism. Recent findings linked ABCA1 function to the phagocytic com- partment [11] and hepatocytes have been described to be able to engulf apoptotic bodies [26]. Therefore, we analyzed the effect of AOX1 knock-down on phagocytic activity of HepG2 cells. Cells treated with AOX1 siRNA show enhanced uptake of fluores- cence-labeled phagobeads (50±6%, · Fig. 3B) compared to cells treated with nonsilencing siRNA. These data further emphasize the influence of AOX1 on ABCA1 function.
Expression pattern of AOX1 in human tissues
As the next step, we analyzed the mRNA expression of AOX1 together with the expression of ABCA1 in human tissues by Taq- Man real-time reverse transcription-PCR. AOX1 mRNA was sig- nificantly expressed in liver, kidney, testis, ovary, and adrenal gland. All other tissues analyzed showed relatively low AOX1 mRNA levels when compared to adrenal gland (Table 1). ABCA1 is more ubiquitously expressed with highest expression levels in adrenal gland, placenta, kidney, lung, and spinal cord.
We also determined AOX1 protein in different human tissues and the highest expression was found in the liver, followed by kidney, ovary, and testis. Weak expression was detected in the heart,
A
ApoA-I dependent lipid efflux
140
relative effflux efficiency (%)
120
100
*
80
60
**
40
20
0
Choline
Cholesterol
B
Phagocytic activity
180
160
**
relative phagocytic activity (%)
140
120
100
I
80
60
40
20
0
nonsilencing SİRNA
AOX1 siRNA
lung, brain, stomach, and spleen, while no AOX1 protein could be observed in skeletal muscle and small intestine (· Fig. 4).
In order to assess the cellular distribution of AOX1 and ABCA1 proteins, immunohistochemistry for both proteins was per- formed on human tissue sections from organs with the highest AOX1 expression as detected by immunoblots. ABCA1 expres- sion showed a broad cellular distribution in liver, kidney, adre- nal, testicular, and ovarian tissues. Expression of AOX1 was confined to hepatocytes, proximal tubular epithelial cells of the kidney, cortical but not medullary cells of the adrenal gland, tes- ticular Leydig cells, and theca externa cells in ovaries (· Fig. 5).
Reduced expression of AOX1 in hepatocellular carcinoma cells
Expression of ABCA1 and AOX1 was investigated by immunohis- tochemistry using a commercial liver TMA. In normal human liver 62% (24/39) of controls were found positive for ABCA1 and
Downloaded by: University of Pittsburgh. Copyrighted material.
| Tissue | ABCA1 | AOX1 |
|---|---|---|
| brain, whole | • | • |
| brain, cerebellum | • | • |
| spinal cord | .... | • |
| heart | • | • |
| skeletal muscle | • | • |
| lung | .... | • |
| kidney | .... | ... |
| liver | .. | .. |
| stomach | .. | • |
| small intestine | .. | • |
| colon | .. | • |
| salivary gland | • | • |
| thyroid | ... | • |
| pancreas | • | • |
| adrenal gland | ..... | ..... |
| testis | .. | ·· |
| ovary | .. | .. |
| prostate | ·· | • |
| mammary | • | • |
| uterus | ... | • |
| placenta | .... | • |
| thymus | ... | • |
| spleen | ... | • |
| bone marrow | • | • |
Expression was measured by TaqMan real-time reverse transcription-PCR. Experiments were performed in triplicates and results were converted to a linear scale (· · · · ·: 81-100% of highest expression; … .: 61-80%; …: 41-60%; · ·: 21-40%; ·: >0-20%)
100% (39/39) for AOX1. In contrast 68 (28/41) and 59% (24/41) of the samples from hepatocellular carcinomas (HCC) were ABCA1 and AOX1 positive, indicating reduced AOX1 and enhanced ABCA1 expression, respectively. This finding was supported by a second independent TMA with HCC, normal and cirrhotic liver specimens (Table 2). AOX1 and ABCA1 protein expression of any intensity (score 1+-3+) was detected in 77.5% (134/173) and 89.0% (146/164) of hepatocellular carcinomas, respectively. Again, AOX1 expression was significantly reduced from normal liver (100% expression) to cirrhosis (92.5%) to hepatocellular carcinomas (77.5%). Accordingly, expression of ABCA1 signifi- cantly increased from normal liver (56.3%) to cirrhosis (87.3%) to hepatocellular carcinomas (89.0%). AOX1 and ABCA1 immu- noreactivity (intensity score 1+-3+) coassociated significantly in 69.5% of cases (107/154), whereas simultaneous negative AOX1 and ABCA1 expression (score 0) was found only in 6.8% (8/154) of cases (p=0.032).
For descriptive data analysis clinico-pathologic characteristics of hepatocellular carcinomas were associated with AOX1 and ABCA1 IHC (Table 2). Loss of AOX1 staining was significantly associated with higher tumor stages (p=0.008), positive nodal status (p=0.005), and occurrence of distant metastases (p=0.037). Loss of ABCA1 expression was associated with positive nodal status (p=0.021) and occurrence of distant metastases (p=0.014). In general, loss of AOX1 and ABCA1 expression was associated with negative pathologic parameters in hepatocellular carcinomas.
Discussion
&
In this work, AOX1 was confirmed to associate with and influ- ence ABCA1-mediated functions. Knock-down of AOX1 expres-
Sm. Intestine
Brain
Heart
Kidney
Sk. Muscle
Liver
Lung
Stomach
Spleen
Ovary
Testis
5 min
-
AOX1
10 sec
AOX1
sion leads to decreased lipid efflux and enhanced phagocytic activity in HepG2 cells, supporting the involvement of ABCA1 in the modulation of phagocytic uptake in type I- and type II phago- cytosis [11]. These cellular phenotypes are comparable to the phenotype of ABCA1-deficient monocytes and/or fibroblasts, that also show enhanced phagocytic activity and defective apoA- I-dependent cholesterol efflux [27,28].
The physiological function of AOX1 is not yet fully characterized. Although AOX1 is well-described as a xenobiotic metabolizing enzyme, its physiological substrates and functions are poorly understood. AOX1 derived ROS are directly implicated in free radical damage in liver and brain during ethanol metabolism [17,29-32]. Alcohol dehydrogenase produces acetaldehyde and NADH from ethanol. Both are substrates for AOX1 and could lead to the formation of ROS in certain tissues [33-35]. Alcohol-medi- ated ROS injury of the liver was shown to operate directly through the combined activities of AOX1 and its close homo- logue xanthine oxidoreductase (XOR) [17,30,31,34] by an iron- dependent process, suggesting involvement of hydroxyl radical in tissue damage [30,32,34]. Recently AOX1 was shown to be regulated in a NAD(P)H oxidase-mediated manner in VSMC and this regulation was induced by thrombin [23]. NAD(P)H oxidase generated ROS plays a role for several signal transduction path- ways in VSMC [36], during host defense in professional pha- gocytes [37], and contributes to M-CSF induced monocyte/ macrophage survival via regulation of Akt and p38 MAPK [38]. Akt is a kinase downstream from PI3 K, and PI3K/Akt pathways are involved in the PDGF-dependent suppression of ABCA1 expression in VSMC [39]. Akt activity is reduced by ABCA1 [40] and HDL-mediated activation of the PI3K/Akt pathways results in eNOS mediated NO production [41]. These data support a complex regulatory network of AOX1, ABCA1, HDL, NAD(P)H oxi- dase, PI3K/Akt, and ROS formation that might influence various biological processes including atherosclerosis and phagocytosis. Moderate alcohol consumption is associated with a decreased risk of cardiovascular diseases [42] and more than 50% of this effect is due to an increase in HDL-cholesterol [43,44]. More- over, moderate alcohol consumption increases the capacity of serum to induce cholesterol efflux from J774 mouse macro-
Liver
ABCA1
AOX1
Kidney
ABCA1
AOX1
Adrenal Gland
ABCA1
AOX1
Testis
ABCA1
AOX1
Ovary
ABCA1
AOX1
phages, which might be mediated by ABCA1 [45]. Taken together, these data and our recent results suggest that AOX1 itself by direct interaction, and/or AOX1-derived metabolites, like ROS, could influence ABCA1 activity.
On the other hand, alcohol and retinol were supposed to be metabolized in part by the same enzymes [46], and AOX1 was suggested to be identical to retinal oxidase [19]. In rabbit liver two pathways for RA biosynthesis exist. One of them involves the cytosolic NAD+-dependent retinal dehydrogenase and the other the oxygen-dependent retinal oxidase (= AOX1). Both cat- alyze the same reaction with an activity ratio of 59 and 41% under experimental conditions [20]. While the metabolic sig- nificance is unclear, it is possible that NAD+/NADH tissue levels affect the rate of retinal oxidation. Retinal dehydrogenase may
be more effective under conditions with high NAD+ levels while retinal oxidase may maintain basic RA level in liver and circula- tion [20]. The involvement of AOX1 in retinoid metabolism is further supported by chicken AOX1, which also possesses reti- naldehyde oxidase activity [47]. 9-cis RA and all-trans RA are both transcriptional modulators for retinoic acid receptor (RAR)/ retinoid X receptor (RXR) target genes. Recent data from our laboratory [48] demonstrate the effect of retinoids on lipid efflux in macrophages. A strong upregulation was observed for genes involved in cellular cholesterol homeostasis, including ABCA1 which is under strong transcriptional control through liver X receptor (LXR)/RXR. Considering that AOX1 might be involved in retinoid metabolism and ABCA1 is a retinoid responsive gene, we expected that AOX1 might influence ABCA1 on the transcrip-
Downloaded by: University of Pittsburgh. Copyrighted material.
| Variable | Categorization | AOX1 expression | pt | ABCA1 expression | pt | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n analyzable | negative | score 1 + | score 2+ | score 3+ | n analyzable | negative | score 1 + | score 2+ | score 3 + | ||||
| clinico-pathologic data | |||||||||||||
| histologic tumor type | |||||||||||||
| hepatocellular | 173 | 39 | 15 | 36 | 83 | 0.001 | 164 | 18 | 45 | 80 | 21 | 0.018 | |
| carcinoma | |||||||||||||
| cirrhosis | 80 | 6 | 2 | 17 | 55 | 71 | 9 | 23 | 31 | 8 | |||
| normal liver | 18 | 0 | 0 | 2 | 16 | 16 | 7 | 6 | 2 | 1 | |||
| age at diagnosis | |||||||||||||
| ≤70 years | 94 | 19 | 11 | 18 | 46 | 0.576 | 94 | 8 | 28 | 44 | 14 | 0.947 | |
| >70 years | 51 | 12 | 3 | 13 | 23 | 45 | 5 | 12 | 22 | 6 | |||
| gender | |||||||||||||
| female | 24 | 6 | 2 | 4 | 12 | 0.694 | 24 | 3 | 6 | 12 | 3 | 0.950 | |
| male | 87 | 15 | 9 | 23 | 40 | 83 | 7 | 22 | 43 | 11 | |||
| tumor stage | |||||||||||||
| pT1 | 37 | 3 | 4 | 7 | 23 | 0.008 | 37 | 3 | 10 | 22 | 2 | 0.675 | |
| pT2 | 14 | 0 | 1 | 4 | 9 | 12 | 2 | 2 | 6 | 2 | |||
| pT3 | 17 | 8 | 0 | 3 | 6 | 18 | 4 | 6 | 7 | 1 | |||
| pT4 | 2 | 1 | 0 | 1 | 0 | 3 | 0 | 1 | 2 | 0 | |||
| regional lymph node status | |||||||||||||
| pN0 | 29 | 3 | 1 | 6 | 19 | 0.005 | 28 | 1 | 5 | 20 | 2 | 0.021 | |
| PN1 | 11 | 6 | 0 | 3 | 2 | 12 | 4 | 2 | 4 | 2 | |||
| metastatic status | |||||||||||||
| M0 | 22 | 3 | 1 | 3 | 15 | 0.037 | 20 | 2 | 2 | 16 | 0 | 0.014 | |
| M1 | 17 | 6 | 0 | 6 | 5 | 19 | 2 | 6 | 7 | 4 | |||
| histologic grade | |||||||||||||
| G1 | 6 | 0 | 1 | 1 | 4 | 0.299 | 5 | 0 | 2 | 3 | 0 | 0.056 | |
| G2 | 112 | 24 | 8 | 23 | 57 | 100 | 11 | 22 | 57 | 10 | |||
| G3 | 46 | 13 | 6 | 11 | 16 | 51 | 7 | 17 | 16 | 11 | |||
| immunohistochemistry (IHC) | |||||||||||||
| ABCA1 | |||||||||||||
| negative | 18 | 8 | 3 | 5 | 2 | 0.032 | |||||||
| score 1 + | 43 | 11 | 6 | 4 | 22 | ||||||||
| score 2 + | 74 | 14 | 4 | 15 | 41 | ||||||||
| score 3 + | 19 | 4 | 1 | 6 | 8 | ||||||||
`Only hepatocellular carcinomas were included
Fisher’s exact test (2-sided); bold face representing significant data
Pearson Chi-Square test (2-sided); bold face representing significant data
tional level. Surprisingly our findings indicate that the effect of AOX1 knock-down on ABCA1 activity is not due to changes in mRNA or protein expression of ABCA1.
AOX1 was also reported to catalyze the predominant reductive pathway of ziprasidone (Geodon, Zeldox), an atypical antipsy- chotic agent for the treatment of schizophrenia [49]. Interest- ingly ziprasidone was reported to have favorable effects on total cholesterol, LDL, and HDL levels in patients [50], which, in con- text with our results, may be due to altered ABCA1/AOX1 inter- action related modulation of HDL-cholesterol levels.
On the other hand, AOX1 was shown to be a dioxin inducible gene [51]. This effect may be due to AHR-mediated downregula- tion of PPARx [52], as AOX1 mRNA and protein expression were recently shown do be suppressed by PPAR& agonists [53]. The observed hyperlipidemia in TCDD patients and the inverse regu- lation of ABCA1 and AOX1 by PPAR« [53,54] may follow differ- ential regulation since our data indicate a positive influence of AOX1 on ABCA1 activity which is also supported by the effects of ziprasidone on plasma HDL levels. This might be due to differ- ences in target tissue sensitivity or differential transcriptional cofactor production for these two orphan receptors. Further experiments are necessary to clarify this point.
Immunohistochemical data show a cell-specific expression of AOX1 in different human tissues (liver, kidney, adrenal gland, testis, ovary), which differ at least in part from previous results [13]. Besides the ovary, the cell types that express high amounts of AOX1 also express high levels of ABCA1. This further under- lines the physiological association of ABCA1 and AOX1 at the cellular level. Our expression studies indicate that beyond the liver, where hepatocytes express both AOX1 and ABCA1, in other organs only specific cells show highly restricted coexpression. Notably AOX1 and ABCA1 are both highly expressed in ster- oidogenic tissues. Data from ABCA1-/- mice showed impaired fertility due to reduced sperm count, reduced intratesticular tes- tosterone levels and reduction of Leydig cells lipid droplets in these animals [55]. In males, testosterone is primarily secreted by Leydig cells and to a lesser extent by the adrenal cortex, both expressing high levels of AOX1 that is coexpressed with ABCA1. Taken together, these data may imply involvement of the AOX1/ ABCA1-complex in sterol and/or steroid hormone metabolism. Analysis of the expression of AOX1 and ABCA1 in normal liver compared to HCC indicates a differential expression of AOX1 in cancer cells, which in part overlaps with ABCA1. Our results in HCC show a direct correlation of AOX1 and ABCA1 expression, and it is possible that in HCC the interaction of AOX1 and ABCA1 is disrupted. In context with the possible involvement of AOX1 in retinoid-metabolism this could influence cell differentiation/ proliferation. Intriguingly both AOX1 and ABCA1 are significantly correlated with positive nodal status (AOX1: p=0.005; ABCA1: p=0.021), and metastatic status (AOX1: p=0.037; ABCA1: p=0.014). These findings indicate that AOX1, perhaps in combi- nation with ABCA1, might be useful as a histopathologic tumor- marker. The changed expression of AOX1 and ABCA1 in HCC is intriguing and requires further investigation. Our findings are consistent with recent observations of Crnogorac-Jurcevic et al. [22], who reported strong expression of AOX1 in pancreatic aci- nar cells, reduced expression in chronic pancreatitis, and loss of AOX1 in malignant pancreatic adenocarcinoma cells.
In conclusion, we have demonstrated that AOX1 associates with and modulates ABCA1 function. The precise molecular mecha- nism and the indicated complex regulatory network of AOX1, ABCA1, HDL, NAD(P)H oxidase, PI3K/Akt, ROS formation and
retinoids has still to be clarified and is the subject of our ongoing work.
Acknowledgments
&
We thank Rudolph Jung for excellent technical assistance in immunohistochemistry. This work was supported by the Deut- sche Forschungsgemeinschaft projects Li 923/2-1 and the Tran- sregional Collaborative Research Centre 13: “Membrane microdomains and their role in human disease” and by the Fraunhofer Project Group - Disease and toxicoproteomics of aging disorders.
Affiliation
1 Institute of Clinical Chemistry and Laboratory Medicine, Regensburg University Medical Center, Regensburg, Germany
2 Institute of Pathology, Regensburg University Medical Center, Regensburg, Germany
3 Institute of Pathology, University Hospital Zürich, Zürich, Switzerland
4 Institute of Pathology, University Hospital, Basel, Switzerland
5 Department of Pathology, Humanitas Clinical Institute of Rozzano, University of Milan, Rozzano, Milano, Italy
6 Medical Clinic III, Medical Faculty Carl Gustav Carus, University of Technology, Dresden, Germany
References
1 Wang N, Silver DL, Costet P, Tall AR. Specific binding of ApoA-I, enhanced cholesterol efflux, and altered plasma membrane morphol- ogy in cells expressing ABC1. J Biol Chem 2000; 275: 33053-33058
2 Bodzioch M, Orso E, Klucken J, Langmann T, Boettcher A, Diederich W, Drobnik W, Barlage S, Buechler C, Porsch-Ozcurumez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATP-binding cassette transporter 1 is mutated in Tan- gier disease. Nat Genet 1999; 22: 347-351
3 Munehira Y, Ohnishi T, Kawamoto S, Furuya A, Shitara K, Imamura M, Yokota T, Takeda S, Amachi T, Matsuo M, Kioka N, Ueda K. Alpha1- syntrophin modulates turnover of ABCA1. J Biol Chem 2004; 279: 15091-15095
4 Okuhira K, Fitzgerald ML, Sarracino DA, Manning JJ, Bell SA, Goss JL, Freeman MW. Purification of ATP-binding cassette transporter A1 and associated binding proteins reveals the importance of beta1-syn- trophin in cholesterol efflux. J Biol Chem 2005; 280: 39653-39664
5 Buechler C, Boettcher A, Bared SM, Probst MC, Schmitz G. The carboxy- terminus of the ATP-binding cassette transporter A1 interacts with a beta2-syntrophin/utrophin complex. Biochem Biophys Res Commun 2002; 293: 759-765
6 Buechler C, Bared SM, Aslanidis C, Ritter M, Drobnik W, Schmitz G. Molecular and functional interaction of the ATP-binding cassette transporter A1 with Fas-associated death domain protein. J Biol Chem 2002; 277: 41307-41310
7 Nofer JR, Remaley AT, Feuerborn R, Wolinnska I, Engel T, Eckardstein A von, Assmann G. Apolipoprotein A-I activates Cdc42 signaling through the ABCA1 transporter. J Lipid Res 2006; 47: 794-803
8 Hirano K, Matsuura F, Tsukamoto K, Zhang Z, Matsuyama A, Takaishi K, Komuro R, Suehiro T, Yamashita S, Takai Y, Matsuzawa Y. Decreased expression of a member of the Rho GTPase family, Cdc42Hs, in cells from Tangier disease - the small G protein may play a role in cholesterol efflux. FEBS Lett 2000; 484: 275-279
9 Nofer JR, Feuerborn R, Levkau B, Sokoll A, Seedorf U, Assmann G. Involve- ment of Cdc42 signaling in apoA-I-induced cholesterol efflux. J Biol Chem 2003; 278: 53055-53062
10 Engel T, Lueken A, Bode G, Hobohm U, Lorkowski S, Schlueter B, Rust S, Cullen P, Pech M, Assmann G, Seedorf U. ADP-ribosylation factor (ARF)- like 7 (ARL7) is induced by cholesterol loading and participates in apolipoprotein AI-dependent cholesterol export. FEBS Lett 2004; 566: 241-246
11 Bared SM, Buechler C, Boettcher A, Dayoub R, Sigruener A, Grandl M, Rudolph C, Dada A, Schmitz G. Association of ABCA1 with syntaxin 13 and flotillin-1 and enhanced phagocytosis in tangier cells. Mol Biol Cell 2004; 15: 5399-5407
12 Garattini E, Mendel R, Romao MJ, Wright R, Terao M. Mammalian molybdo-flavoenzymes, an expanding family of proteins: structure,
genetics, regulation, function and pathophysiology. Biochem J 2003; 372: 15-32
13 Moriwaki Y, Yamamoto T, Takahashi S, Tsutsumi Z, Hada T. Widespread cellular distribution of aldehyde oxidase in human tissues found by immunohistochemistry staining. Histol Histopathol 2001; 16: 745- 753
14 Obach RS. Potent inhibition of human liver aldehyde oxidase by raloxifene. Drug Metab Dispos 2004; 32: 89-97
15 Obach RS, Huynh P, Allen MC, Beedham C. Human liver aldehyde oxi- dase: inhibition by 239 drugs. J Clin Pharmacol 2004; 44: 7-19
16 Moriwaki Y, Yamamoto T, Higashino K. Distribution and pathophysi- ologic role of molybdenum-containing enzymes. Histol Histopathol 1997; 12: 513-524
17 Shaw S, Jayatilleke E. The role of aldehyde oxidase in ethanol-induced hepatic lipid peroxidation in the rat. Biochem J 1990; 268: 579-583
18 Wright RM, MacManaman JL, Repine JE. Alcohol-induced breast cancer: a proposed mechanism. Free Radic Biol Med 1999; 26: 348-354
19 Tomita S, Tsujita M, Ichikawa Y. Retinal oxidase is identical to aldehyde oxidase. FEBS Lett 1993; 336: 272-274
20 Huang DY, Ichikawa Y. Two different enzymes are primarily respon- sible for retinoic acid synthesis in rabbit liver cytosol. Biochem Bio- phys Res Commun 1994; 205: 1278-1283
21 Colakoglu N, Kukner A, Oner J, Sonmez MF, Oner H, Ozan E. Effects of high dose retinoic acid on TGF-beta2 expression during pancreatic organogenesis. J Mol Histol 2005; 36: 413-418
22 Crnogorac-Jurcevic T, Gangeswaran R, Bhakta V, Capurso G, Lattimore S, Akada M, Sunamura M, Prime W, Campbell F, Brentnall TA, Costello E, Neoptolemos J, Lemoine NR. Proteomic analysis of chronic pancreatitis and pancreatic adenocarcinoma. Gastroenterology 2005; 129: 1454-1463
23 Vendrov AE, Madamanchi NR, Hakim ZS, Rojas M, Runge MS. Thrombin and NAD(P)H oxidase-mediated regulation of CD44 and BMP4-Id pathway in VSMC, restenosis, and atherosclerosis. Circ Res 2006; 98: 1254-1263
24 Langmann T, Mauerer R, Zahn A, Moehle C, Probst M, Stremmel W, Schmitz G. Real-time reverse transcription-PCR expression profiling of the complete human ATP-binding cassette transporter superfamily in various tissues. Clin Chem 2003; 49: 230-238
25 Drobnik W, Borsukova H, Boettcher A, Pfeiffer A, Liebisch G, Schutz GJ, Schindler H, Schmitz G. Apo AI/ABCA1-dependent and HDL3-mediated lipid efflux from compositionally distinct cholesterol-based microdo- mains. Traffic 2002; 3: 268-278
26 Dini L, Lentini A, Diez GD, Rocha M, Falasca L, Serafino L, Vidal- Vanaclocha F. Phagocytosis of apoptotic bodies by liver endothelial cells. J Cell Sci 1995; 108 (Pt 3): 967-973
27 Zha X, Genest Jr J, MacPherson R. Endocytosis is enhanced in Tangier fibroblasts: possible role of ATP-binding cassette protein A1 in endo- somal vesicular transport. J Biol Chem 2001; 276: 39476-39483
28 Zha X, Gauthier A, Genest J, MacPherson R. Secretory vesicular transport from the Golgi is altered during ATP-binding cassette protein A1 (ABCA1)-mediated cholesterol efflux. J Biol Chem 2003; 278: 10002-10005
29 Bondy SC, Orozco J. Effects of ethanol treatment upon sources of reac- tive oxygen species in brain and liver. Alcohol Alcohol 1994; 29: 375-383
30 Shaw S, Jayatilleke E. The role of cellular oxidases and catalytic iron in the pathogenesis of ethanol-induced liver injury. Life Sci 1992; 50: 2045-2052
31 Shaw S, Jayatilleke E. Cimetidine as a scavenger of ethanol-induced free radicals. Alcohol 1992; 9: 363-367
32 Shaw S, Eng J, Jayatilleke E. Ethanol-induced free radical injury to the hepatocyte glucagon receptor. Alcohol 1995; 12: 273-277
33 Wright RM, Riley MG, Weigel LK, Ginger LA, Costantino DA, MacManaman JL. Activation of the human aldehyde oxidase (hAOX1) promoter by tandem cooperative Sp1/Sp3 binding sites: identification of complex architecture in the hAOX upstream DNA that includes a proximal promoter, distal activation sites, and a silencer element. DNA Cell Biol 2000; 19: 459-474
34 Mira L, Maia L, Barreira L, Manso CF. Evidence for free radical genera- tion due to NADH oxidation by aldehyde oxidase during ethanol metabolism. Arch Biochem Biophys 1995; 318: 53-58
35 Wright RM, Weigel LK, Repine JE. Aldehyde oxidase generates deoxyribonucleic acid single strand nicks in vitro. Redox Rep 1995; 1: 313-321
36 Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res 2000; 86: 494-501
37 Babior BM. NADPH oxidase. Curr Opin Immunol 2004; 16: 42-47
38 Wang Y, Zeigler MM, Lam GK, Hunter MG, Eubank TD, Khramtsov VV, Tridandapani SV, Sen CK, Marsh CB. The Role of the NADPH Oxidase Complex, p38 MAPK and Akt in Regulating Human Monocyte/Macro- phage Survival. Am J Respir Cell Mol Biol 2007; 36: 68-77
39 Nagao S, Murao K, Imachi H, Cao WM, Yu X, Li J, Matsumoto K, Nishiuchi T, Ahmed RA, Wong NC, Ueda K, Ishida T. Platelet derived growth factor regulates ABCA1 expression in vascular smooth muscle cells. FEBS Lett 2006; 580: 4371-4376
40 Landry YD, Denis M, Nandi S, Bell S, Vaughan AM, Zha X. ATP-binding cassette transporter A1 expression disrupts raft membrane microdo- mains through its ATPase-related functions. J Biol Chem 2006; 281: 36091-36101
41 Nofer JR, Giet M van der, Tolle M, Wolinska I, Wnuck LK von, Baba HA, Tietge UJ, Godecke A, Ishii I, Kleuser B, Schafers M, Fobker M, Zidek W, Assmann G, Chun J, Levkau B. HDL induces NO-dependent vasorelaxa- tion via the lysophospholipid receptor S. 1P3. J Clin Invest 2004; 113: 569-581
42 Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ. Moderate alco- hol intake and lower risk of coronary heart disease: meta-analysis of effects on lipids and haemostatic factors. BMJ 1999; 319: 1523-1528
43 Castelli WP, Doyle JT, Gordon T, Hames CG, Hjortland MC, Hulley SB, Kagan A, Zukel WJ. Alcohol and blood lipids. The cooperative lipopro- tein phenotyping study. Lancet 1977; 2: 153-155
44 Criqui MH, Cowan LD, Tyroler HA, Bangdiwala S, Heiss G, Wallace RB, Cohn R. Lipoproteins as mediators for the effects of alcohol consump- tion and cigarette smoking on cardiovascular mortality: results form the Lipid Research Clinics Follow-up Study. Am J Epidemiol 1987; 126: 629-637
45 Beulens JW, Sierksma A, Tol A van, Fournier N, Gent T van, Paul JL, Hendriks HF. Moderate alcohol consumption increases cholesterol efflux mediated by ABCA1. J Lipid Res 2004; 45: 1716-1723
46 Wang XD. Alcohol, vitamin A, and cancer. Alcohol 2005; 35: 251-258
47 Terao M, Kurosaki M, Barzago MM, Varasano E, Boldetti A, Bastone A, Fratelli M, Garattini E. Avian and canine aldehyde oxidases: novel insights into the biology and evolution of molybdo-flavoenzymes. J Biol Chem 2006; 281: 19748-19761
48 Langmann T, Liebisch G, Moehle C, Schifferer R, Dayoub R, Heiduczek S, Grandl M, Dada A, Schmitz G. Gene expression profiling identifies retinoids as potent inducers of macrophage lipid efflux. Biochim Bio- phys Acta 2005; 1740: 155-161
49 Beedham C, Miceli JJ, Obach RS. Ziprasidone metabolism, aldehyde oxidase, and clinical implications. J Clin Psychopharmacol 2003; 23: 229-232
50 Brown RR, Estoup MW. Comparison of the metabolic effects observed in patients treated with ziprasidone versus olanzapine. Int Clin Psy- chopharmacol 2005; 20: 105-112
51 Yang YM, Huang DY, Liu GF, Zhong JC, Du K, Li YF, Song XH. Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on vitamin A metabolism in mice. J Biochem Mol Toxicol 2005; 19: 327-335
52 Shaban Z, El-Shazly S, Abdelhady S, Fattouh I, Muzandu K, Ishizuka M, Kimura K, Kazusaka A, Fujita S. Down regulation of hepatic PPARalpha function by AhR ligand. J Vet Med Sci 2004; 66: 1377-1386
53 Neumeier M, Weigert J, Schaffler A, Weiss TS, Schmidl C, Buttner R, Bollheimer C, Aslanidis C, Scholmerich J, Buechler C. Aldehyde oxidase 1 is highly abundant in hepatic steatosis and is downregulated by adiponectin and fenofibric acid in hepatocytes in vitro. Biochem Bio- phys Res Commun 2006; 350: 731-735
54 Chinetti G, Lestavel S, Bocher V, Remaley AT, Neve B, Torra IP, Teissier E, Minnich A, Jaye M, Duverger N, Brewer HB, Fruchart JC, Clavey V, Staels B. PPAR-alpha and PPAR-gamma activators induce cholesterol removal from human macrophage foam cells through stimulation of the ABCA1 pathway. Nat Med 2001; 7: 53-58
55 Selva DM, Hirsch-Reinshagen V, Burgess B, Zhou S, Chan J, MacIsaac S, Hayden MR, Hammond GL, Vogl AW, Wellington CL. The ATP-binding cassette transporter 1 mediates lipid efflux from Sertoli cells and influences male fertility. J Lipid Res 2004; 45: 1040-1050
Downloaded by: University of Pittsburgh. Copyrighted material.