Human Acyl-CoA:Cholesterol Acyltransferase-1 Is a Homotetrameric Enzyme in Intact Cells and in Vitro *

Acyl-CoA:cholesterol acyltransferase (ACAT) is a key enzyme in cellular cholesterol homeostasis and in atherosclerosis. ACAT-1 may function as an allosteric enzyme. We took a multifaceted approach to investigate the subunit composition of ACAT-1. When ACAT-1 with two different tags were co-expressed in the same Chinese hamster ovary cells, antibody specific to one tag caused co-immunoprecipitation of both types of ACAT-1 proteins. Radioimmunoprecipitations of cells expressing the untagged ACAT-1 or the 6-histidine-tagged ACAT-1 yielded a single radiolabeled band of predicted size on SDS-polyacrylamide gel electrophoresis. These results show that ACAT-1 exists as homo-oligomers in intact Chinese hamster ovary cells. We solubilized HisACAT-1 with the detergent deoxycholate or CHAPS (3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonic acid), performed gel filtration chromatography and sucrose density gradient centrifugations in H2O and D2O, and determined the Stokes radii and sedimentation coefficients of the HisACAT1-detergent complexes. The estimated molecular mass of HisACAT-1 is 263 kDa, which is 4 times that of the HisACAT-1 monomer (69 kDa). Finally, cross-linking experiments in intact cells and in vitro show that the increase in cross-linker concentrations causes an increase in size of the HisACAT-1-positive signals, forming material(s) 4 times the size of the monomer, supporting the conclusion that ACAT-1 is a homotetrameric enzyme.

Acyl-CoA:cholesterol acyltransferase (ACAT) 1 is an intracellular enzyme present in various eukaryotic cells. This enzyme utilizes long chain fatty acyl-coenzyme A and cholesterol as the two substrates, to catalyze the formation of cholesteryl esters and coenzyme A (for a recent review, see Ref. 1). In various cell types, ACAT is responsible for producing essentially all of the cytoplasmic cholesteryl esters. Cholesterol as a polar lipid partitions in the cell membranes, while cholesteryl ester as a neutral lipid aggregates in the cell cytoplasm and forms lipid droplets. A main function of ACAT is believed to guard against the toxic buildup of unesterified cholesterol in the cell membranes. In specialized cell types such as hepatocytes and intestinal enterocytes, cholesteryl esters along with triacylglycerol constitute the majority of the neutral lipid core in lipoproteins (i.e. very low density lipoproteins and chylomicrons; for recent reviews, see Refs. 2 and 3). In macrophages, the accumulation of cholesteryl esters leads to foam cell formation, which is an early step of atherosclerosis (recently reviewed in Ref. 4). In mammals, two different ACAT genes, ACAT-1 and ACAT-2, have been identified (5)(6)(7)(8). The physiological functions of these two enzymes in various species are currently under investigation in several laboratories. The results from immunodepletion analysis suggest that in humans, ACAT-1 protein is responsible for most of the ACAT activities present in the homogenates of various adult tissues/cells examined, including liver, hepatocytes, HepG2 cells, adrenal gland, macrophages, kidney, but not intestines (9). Whether the intestinal ACAT is mainly due to the presence of ACAT-2 enzyme is under investigation.
A consensus octanucleotide sequence element, designated as the sterol response element, is present within the promoter regions of many sterol-sensitive genes (reviewed in Refs. 10 and 11). Sequence analysis revealed that the sterol response element-like sequence is not present in the ACAT-1 promoters (12). Previous studies have shown that the main mode of regulation of ACAT-1 by sterol in various cell types is at the posttranslational level (reviewed in Ref. 1). Recently, we have purified the recombinant ACAT-1 enzyme to homogeneity with retention of catalytic activity. The cholesterol substrate saturation curves of the purified enzyme, assayed in either mixed micelles or reconstituted vesicles, are both highly sigmoidal. In contrast, the oleoyl-coenzyme A substrate saturation curves of the enzyme assayed under the same conditions are both hyperbolic. These results support the hypothesis that ACAT-1 is an allosteric enzyme regulated by cholesterol, i.e. cholesterol may serve as an activator as well as a substrate for the enzyme (1).
To further examine the molecular mechanism that causes ACAT-1 to be regulated by cholesterol, it is essential to dissect the enzyme structure at the biochemical level. Human ACAT-1 is an integral membrane protein; when expressed in Chinese hamster ovary (CHO) cells, it spans the endoplasmic reticulum membranes seven times (13). The subunit composition of the enzyme in intact cells and in vitro remains unknown. In this report, by using several independent methods, we have determined that ACAT-1 in intact cells and in vitro exists as a homotetrameric enzyme.

EXPERIMENTAL PROCEDURES
Materials-CHAPS, taurocholate, fetal bovine serum (FBS), various protein standards used for gel filtration chromatography and for sucrose gradient centrifugation, mouse and rabbit IgGs, Sephacryl S-400HR, and Sepharose CL-6B resins were from Sigma. The chemical cross-linker dithiobis-succinimidyl propionate (DSP) was from Pierce. Sucrose (ultracentrifugation grade) was from Fisher. Anti-HA mouse monoclonal antibody (HA11) was from Babco; anti-T7 monoclonal antibody was from Novagen. The various constructs employed for transfection experiments were previously described (13,14). The F-12 medium was from Cellgro; Grace's medium and Opti-MEM serum-free medium were from Life Technologies, Inc. Tran 35 S-label was from ICN.
Other reagents were from the same source as described previously (5,15).
Cell Culture-CHO cells were routinely maintained in medium A (F-12 medium containing 10% FBS) as described previously (16). When indicated, CHO cells were cultured in the cholesterol-free medium (medium D; F-12 medium containing 35 M oleic acid and 10% delipidated FBS) or in the cholesterol starvation medium (medium S; medium D with 50 M mevinolin and 230 M mevalonate) (17). Insect Sf9 cells and High Five™ cells (from Invitrogen) were maintained in Grace's medium with 5% FBS unless specified in the text. Procedures for transfecting CHO cells and for infecting insect cells have been described previously (13,14).
Transfection and Co-immunoprecipitation-Two cell lines derived from an ACAT-deficient CHO cell line AC29 (18) were used: the HisACAT-1 cell, stably expressing the N-terminal 6-His-tagged and the T7-tagged hACAT-1 protein (19), and the ACAT1-HA1 cell, stably expressing the HA-tagged ACAT-1 protein (13). Both cell lines express catalytically active ACAT-1 protein. The HisACAT-1 cells were transfected with the construct ACAT1-HA1 (13), while the ACAT1-HA1 cells were transfected with the construct HisACAT-1 (19). Forty-eight h after transfection, the cells were solubilized with 1 ml of 1% DOC in Buffer A (50 mM Tris, and 1 mM EDTA, pH.8.3). The lysates were cleared by passing through a 21-gauge needle 10 times. When indicated, lysate mixtures were prepared by mixing 0.5 ml of lysate from each cell type. Unless indicated, all subsequent procedures were carried out at 4°C. Lysates were centrifuged at 100,000 ϫ g for 30 min. The supernatant was precleared by mixing with 20 g of irrelevant mouse IgG for 1 h and then with 30 l of protein A-Sepharose CL-4B in Buffer A (1:1 in volume) for 1 h. The mixing was with an orbital shaker (Bellco). The supernatants were isolated by centrifugation at 3,000 rpm for 2 min, after which 2 g of anti-HA mouse monoclonal antibody IgG (Babco) or 3 g of anti-T7 mouse monoclonal antibody IgG (Novagen) was added to the supernatants, respectively. The samples were mixed for 2 h; 30 l of protein A-Sepharose CL-4B were added; and the samples were further mixed for 1 h, then pelleted by centrifugation at 3000 rpm for 2 min, and washed three times with Buffer A. Each wash included a 10-min incubation at 37°C, with samples gently reverted by hand two or three times during the incubation, and a 30-min mixing at 4°C by gentle rotation on orbital shaker. After the washes, the bound ACAT protein was dissociated by adding 40 l of 5ϫ sample loading buffer (10% SDS, 250 mM dithiothreitol, 50% glycerol, 0.25% bromphenol blue, 50 mM Tris, pH 6.8) and incubated at 37°C for 10 min. After pelleting by rapid centrifugation in a microcentrifuge, all of the recovered supernatant was used for analysis by 10% SDS-PAGE. 35 S Metabolic Labeling and Radioimmunoprecipitation-This was based on a procedure previously published (15) with modifications. Cells were seeded in medium A and grown in 25-cm 2 flasks to approximately 70% confluence, washed with sterile phosphate-buffered saline, and fed 1.5 ml/flask methionine-deficient Dulbecco's modified Eagle's medium supplemented with 35 mg/liter L-proline and 10% dialyzed FBS for 1 h. 10 l/flask (at 10.5 Ci/l) Tran 35 S-label was added, and incubated for 12 h. Cell lysates were prepared and precleared in the same manner as described in the previous section. To the precleared supernatant, 5 l/tube anti-ACAT-1 antibodies DM10 (at 0.3 g/l) was added and the mixture was gently rotated at 4°C for 2 h. Then 50 l/tube protein A-Sepharose beads were added, and the mixture was rotated for another 60 min. The immune complex was pelleted, washed, solubilized, and analyzed by 10% SDS-PAGE as described in the previous section.
Gel Filtration Chromatography-The experiments were conducted at 4°C. A 0.8-ml sample was applied onto a Sephacryl-400HR or a Sepharose CL-6B column (1.5 ϫ 35 cm), pre-equilibrated with 1% DOC/Buffer A, or 0.5% CHAPS/1 M KCl/Buffer A as indicated. The sample was eluted using the same buffer at 20 ml/h; 0.5-ml fractions were collected. Proteins of known Stokes radii as indicated were used as standards, and were detected by measuring the absorbance of each collected fraction at 280 nm. To monitor the ACAT-1 elution profile, fractions were subjected to Western blot analysis using anti-ACAT-1 antibodies DM10.
Sucrose Density Gradient Centrifugation-The experiments were conducted at 4°C. The centrifugations were carried out using a Beckman SW 60 Ti rotor in a Beckman model L8 -55M ultracentrifuge. Hydrodynamic Parameters of ACAT-1-ACAT-1-detergent complex (M c ) was calculated from the Stokes radius (R s ), partial specific volume ( c ), and sedimentation coefficient (s 20,w ) using Equation 1 (20).
To calculate the Stokes radius (R s ) of the ACAT-1-detergent complex, gel filtration chromatography experiments were conducted, and the distribution coefficient, K av , was calculated from the equation where V e is the elution volume of each reference protein or the ACAT-1 detergent complex. The Stokes radius of the ACAT-1detergent complex was determined by interpolation using a calibration curve, which was constructed by plotting Stokes radii of reference proteins versus (Ϫlog K av ) 1/2 according to the relationship (Ϫlog K av ) 1/2 ϭ ␣ (␤ ϩ R s ) (21). The partial specific volume of the ACAT-1-detergent complex was calculated by performing sucrose density gradient centrifugation experiments in H 2 O and in D 2 O. The calculation was according to Equation 2 (Equation 9 in Ref. 22).
A and B are the partial specific volumes of the ACAT-1-detergent protein complex (A) and a reference protein Aldolase, ␤-amylase, and catalase were used as the reference proteins. The weight fraction of ACAT-1 in the protein-detergent complex (X p ) was calculated using the equation X p ϭ ( c Ϫ d )/( p Ϫ d ) (23). p is the partial specific volume of HisACAT-1, calculated from its amino acid composition (24). d is partial specific volume of the detergent. For DOC, d ϭ 0.78 cm 3 /g (25); for CHAPS, d ϭ 0.81 cm 3 /g (26).
Chemical Cross-linking in Mammalian Cells-Experiments were conducted at 37°C. HisACAT-1 cells were seeded in six-well dishes and grown overnight in F-12 medium with 10% FBS, to approximately 80% confluence. One h before cross-linking, cells were given a fresh medium change. DSP was added directly into the 1-ml medium from a 250 mM DSP stock (freshly prepared in Me 2 SO). The final Me 2 SO concentration in the medium was adjusted to 1%. Cells were incubated at 37°C for 30 min. The reaction was stopped by adding 100 l of 1 M Tris-HCl (pH 7.4); the mixture was incubated for 30 min at room temperature. Cells were washed with cold phosphate-buffered saline twice, and lysed with 500 l of 10% SDS. A 100-l amount of 5ϫ SDS-PAGE sample loading buffer (without reducing agent) was added to 300 l of lysate, and incubated at 37°C for 15 min. A 150-l amount of the final mixture was subjected to a non-reducing SDS-PAGE (6% separating gel, and 3% stacking gel) and Western blot analysis with anti-ACAT-1 antibodies DM10. To perform cross-linking using cells under cholesterol starvation, HisACAT-1 cells were seeded in medium A for one day, grown in medium D for 2 days, then grown in medium S for 8 h.
Chemical Cross-linking in Insect Cells-Experiments were performed at room temperature. Insect High Five cells were cultured in six-well dishes in Grace's medium with or without 12.5% FBS. Cells were infected with high titer HisACAT-1 virus for 36 -48 h. Infected cells were harvested by gentle pipetting, washed with Hanks' buffer twice, then resuspended in 1 ml of Hanks' buffer. Cross-linking with DSP and analysis by SDS-PAGE were performed as described above.
Chemical Cross-linking in Vitro-Experiments were performed at 37°C. 20 l of HisACAT-1 purified to near homogeneity (see below), containing approximately 0.1 g/tube, was employed. The enzyme was diluted to 1 ml/tube in buffer A containing 0.5% CHAPS and 1 M KCl.
The cross-linkings with DSP at indicated concentrations were performed for 20 min as described above. The purified ACAT-1 protein samples were prepared from crude extracts of HisACAT-1 cells, based on the enzyme purification procedure published previously (19). To avoid the presence of 0.5 M imidazole (which interferes with the crosslinking reaction) in the final sample, it was necessary to modify the original purification scheme by employing the nickel affinity column chromatography step before the ACAT-1 monoclonal antibody affinity column chromatography step. Its purity is approximately 60 -70%, based on analysis by SDS-PAGE and silver staining.

ACAT-1 Expressed in Intact Cells Forms Homo-oligomeric
Complex-The recombinant ACAT-1 protein containing the 6-histidine tag and the T7 tag at the N terminus (designated as HisACAT-1; Ref. 19), or the HA tag inserted before amino acid residue 25 (designated as ACAT1-HA1; Ref. 13), could be expressed in mutant CHO cells deficient in ACAT. Transfectant clones stably expressing the HisACAT-1 protein or the ACAT1-HA1 protein (designated as the HisACAT-1 cells or the ACAT1-HA1 cells) have been isolated (13,19). To examine whether the ACAT1-HA1 protein can associate with the HisACAT-1 protein within the same cell, we expressed the ACAT1-HA1 protein in the HisACAT-1 cells by transient transfection using the construct ACAT1-HA1 (13), then performed immunoprecipitation experiments using the anti-HA antibody. The results show that the anti-HA antibody co-immunoprecipitated both the ACAT1-HA1 protein and the HisT7ACAT-1 protein (Fig. 1A, lane 2). Control experiments show that the anti-HA antibody failed to immunoprecipitate any HisACAT-1 protein from extracts of HisACAT-1 cells (Fig. 1A, lane 3); the same antibody efficiently immunoprecipitated the ACAT1-HA1 protein from extracts of ACAT1-HA1 cells (Fig. 1A, lane 4). When we mixed the extracts of HisACAT-1 cells and the extracts of ACAT1-HA1 cells in vitro, then performed the immunoprecipitation using the anti-HA antibody, we found that the anti-HA antibody could only immunoprecipitate the ACAT1-HA1 protein but not the HisACAT-1 protein (Fig. 1A, lane 5). In additional experiments, we expressed the HisACAT-1 protein in the ACAT1-HA1 cells by transient transfection using the construct HisACAT-1, then performed various immunoprecipitation experiments using anti-T7 monoclonal antibody. The results (shown in Fig. 1B) fully corroborated the results shown in Fig. 1A. Together, these results demonstrate that the ACAT-1 protein can form homooligomers in intact CHO cells but not in vitro. Additional evidence showed that the ACAT-1 protein expressed in insect cells (14) also exists as homo-oligomers (data not shown).
To examine the subunit composition of the ACAT-1 enzyme, we labeled cellular proteins by performing 35 S metabolic labeling in cells stably expressing untagged ACAT-1 protein (designated as ACAT-1 cells), and the HisACAT-1 cells. We then performed radioimmunoprecipitation experiments. When extracts of ACAT-1 cells were used, the antibodies immunoprecipitated a major labeled protein band identical in size to that of the untagged ACAT-1 protein (50 kDa; Fig. 2, left lane). 2 Parallel experiment using extracts of metabolically 35 S-labeled AC29 cells, showed that no discrete band could be detected (Fig. 2, right lane). This result raised two alternative interpretations; either the ACAT-1 polypeptide forms a homo-oligomer with itself, or the ACAT-1 polypeptide forms a hetero-oligomer with a different polypeptide that exhibits the same apparent molecular weight as the ACAT-1 polypeptide on SDS-PAGE. To distinguish between these two possibilities, we performed an additional radioimmunoprecipitation experiment, using extracts of cells stably expressing HisACAT-1. The result (Fig. 2, middle lane) showed that a single radiolabeled band identical in size to that of the HisACAT-1 protein (54 kDa) was detected; specifically, no discrete band at the 50-kDa region was detectable. This result indicates that the functional ACAT-1 enzyme in intact CHO cells is composed of the ACAT-1 polypeptide only.
Hydrodynamic Properties of HisACAT-1 Solubilized in Detergent DOC-To determine the molecular weight of the ACAT-1 enzyme in solution, we used DOC-solubilized extracts of insect cells expressing HisACAT-1 as the enzyme source. We first performed gel filtration experiments. A typical elution profile is shown in Fig. 3A. HisACAT-1/DOC complex exhibited a single peak with a calculated Stokes radius of 8.2 nm Ϯ 0.2 nm (n ϭ 3) (Fig. 3B). We next performed sucrose density gradient centrifugation experiments in H 2 O and D 2 O. Western blot analyses of centrifugation fractions showed that HisACAT-1 sedimented as a sharp peak in H 2 O and in D 2 O 2 It could be seen from the left lane of Fig. 2 that the antibodies also immunoprecipitated a minor radiolabeled protein band 30 -35 kDa in size. The 30 -35-kDa band had previously been observed in cells overexpressing human ACAT-1 protein (15), and was shown to be derived from the 50-kDa band through intracellular protein degradation.  (Fig. 4, A and B). Based on these results, The hydrodynamic properties of HisACAT-1 in DOC are derived and are summarized in Table I. Based on its amino acid composition, the partial specific volume of the HisACAT-1 protein is 0.75 cm 3 /g. Assuming the same amount of DOC binds to HisACAT-1 in H 2 O and in D 2 O, DOC contributes to approximately 23% by mass in the ACAT-1/DOC complex. The molecular mass of HisACAT-1 protein is estimated to be 280 kDa. Since the predicted monomer size (based on amino acid composition) of the HisACAT-1 polypeptide is 69.4 kDa, this result suggests that HisACAT-1 functions as a homotetramer in DOC. In additional experiments, we have obtained essentially the same results used DOC-solubilized extracts of CHO cells that express HisACAT-1 as the enzyme source (data not shown).
Hydrodynamic Properties of HisACAT-1 in the Detergent CHAPS-In addition to DOC, the HisACAT-1 enzyme expressed in CHO cells or in insect cells can also be solubilized with the zwitterionic detergent CHAPS. The presence of 1 M KCl in the buffer has been found to protect ACAT-1 against inactivation by the detergent CHAPS. Therefore, 1 M KCl has been included when CHAPS is used as the detergent. HisACAT-1 solubilized in CHAPS and 1 M KCl has been purified to homogeneity with retention of enzyme activity (19). To investigate the subunit composition of the HisACAT-1 enzyme solubilized in CHAPS, we performed additional gel filtration chromatography and sucrose density gradient centrifugation experiments. For gel filtration experiments, we used extracts of HisACAT-1 cells as the enzyme source. A typical elution profile is shown in Fig. 5A. 3 The major peak of HisACAT-1/CHAPS complex exhibits a Stokes radius of 7.2 nm (Fig. 5B). For sucrose gradient centrifugation experiments, we used HisACAT-1 purified to homogeneity as the enzyme source. Typical sedimentation profiles in H 2 O and in D 2 O are shown in Fig. 6 (A and B). The hydrodynamic properties of HisACAT-1 in CHAPS are summarized in Table II. In the HisACAT-1/CHAPS complex, CHAPS contributes to approximately 15% by mass. The molecular mass of HisACAT-1 protein is 263 kDa. In  additional experiments, we have obtained essentially the same results by using extracts of insect cells expressing HisACAT-1 as the enzyme source (data not shown).
Chemical Cross-linking of ACAT-1-To further substantiate the notion that ACAT-1 is a homotetramer, we performed chemical cross-linking experiments in intact cells. We chose to use the homo-bifunctional, primary amine-reactive cross-linking agent DSP, because of its high selectivity, moderate spacer length, and membrane permeability. The experiments were first performed in insect cells expressing HisACAT-1. Fig. 7A shows a typical result; the increase in DSP concentrations added in the medium causes an increase in size of the HisACAT-1-positive signals, and leads to the formation of material(s) approximately 4 times the size of the monomer. We next performed cross-linking experiments in CHO cells stably expressing HisACAT-1. Cells were grown in medium containing lipoprotein-bound cholesterol, or in cholesterol starvation medium. The results (Fig. 7B) show that upon cross-linking by DSP, ACAT-1 forms oligomeric material 4 times the size of the ACAT-1 monomer; at intermediate concentration of DSP, additional oligomeric forms, 2 or 3 times the size of the ACAT-1 monomer, can also be detected. These results also show that the ACAT-1 cross-linking pattern is unaltered by changing cholesterol content of the cells. We performed additional crosslinking experiments, using human melanoma cells (15) grown in cholesterol-rich medium or in cholesterol-poor medium, and have obtained essentially the same result as shown in Fig. 7B (data not shown). We next performed chemical cross-linking studies in vitro, using the highly purified HisACAT-1 protein as the source. As shown in Fig. 7C, without DSP, ACAT-1 migrates mainly as a monomer of 54 kDa; with low concentrations of DSP (5-25 M), the ACAT forms an oligomer of approximately 4 times its monomer size (200 kDa). This result shows that ACAT-1 in highly purified state maintains as a homotetramer. DISCUSSION ACAT-1 is an integral membrane protein in the endoplasmic reticulum. Determining the oligomeric state of an integral membrane protein requires multifaceted approaches. Previously, the oligomeric state of ACAT has been studied by several laboratories; by monitoring the kinetics of decay in enzyme activity upon radiation exposure, earlier studies showed that the minimal molecular mass of ACAT in rat liver microsome ranged from 170 to 224 kDa (27,28). At present it is not clear whether the ACAT activity in rat liver microsome is mainly derived from ACAT-1, from ACAT-2, or from both. More recently, by using antibodies specific to ACAT-1 to detect ACAT-1 signal in chemical cross-linking studies, Kawasaki et al. (29) showed that ACAT-1 in rat adrenal gland microsome exists as homo-and/or hetero-oligomeric complexes in vitro. In this study, we expressed the hACAT-1 in two different cell types (mutant CHO cells and the insect cells), both devoid of endogenous ACAT. In intact cells, using co-immunoprecipitation, FIG. 5. A, gel filtration chromatography of HisACAT-1/CHAPS complex, using Sepharose-CL 6B. Details were described under "Experimental Procedures." CHAPS-solubilized extracts of HisACAT-1 cells were used as the enzyme source. B, the Stokes radius (R s ) of HisACAT-1/CHAPS complex. K av for the major peak of the HisACAT-1/CHAPS complex is 0.44, which corresponds to a R s of 7.2 Ϯ 0.1 nm (n ϭ 3).  radioimmunoprecipitation, and chemical cross-linking, our results show that hACAT-1 exists as a homotetramer. In addition, using gel filtration chromatography and sucrose density gradient centrifugation in H 2 O and D 2 O, as well as chemical cross-linking in vitro, our results show that hACAT-1 present either in crude cell extracts or in purified form also mainly exist as a homotetramer. Our results do not exclude the possibility that in intact cells, a minor portion of the ACAT-1 protein may interact with certain non-ACAT-1 polypeptide(s) and form hetero-oligomeric complex(es). Further experiments are needed to address this issue.
We have previously demonstrated that ACAT-1 present in vesicles or in mixed micelles respond to cholesterol as its substrate in sigmoidal manner, implying that ACAT-1 may be an allosteric enzyme regulated by its own substrate cholesterol. The original theory of allostery is based on the idea that cooperative binding of substrates may arise in proteins with two or more structures in equilibrium (reviewed in Ref. 30; see also Ref. 31). To our knowledge, the only other enzyme reported in literature that responds to a membrane lipid molecule in sigmoidal manner is protein kinase C (which uses phosphatidylserine as an activator (32)(33)(34). ACAT-1 spans the endoplasmic reticulum membrane seven times (13), and contains four identical subunits as demonstrated in this report. In contrast, protein kinase C does not contain any membrane-spanning domain and is a monomeric protein (for a review on protein kinase C structure, see Ref. 35). Therefore, the apparent allosterism demonstrated by ACAT-1 has no obvious precedence in literature. The molecular mechanism that controls the ACAT-1 allosterism thus represents a major challenge to investigators. In the future, it will be important to identify the structural motif(s) that govern the oligomerization state of ACAT-1, and to examine whether altering the oligomerization state may affect the catalysis, and/or the apparent allosteric property of the enzyme.