Field of the invention

The present invention relates to novel therapy for apolipoprotein CIII (apoCIII)-induced hypertriglyceridemia, linked to over-expression of plasma apoCIII, which arises from the finding that the lipid transporter ABCA1 (ATP-binding cassette sub-family A, member 1 ) is required to promote association of apoCIII with HDL (high density lipoprotein). It has been found that deficiency of ABCA1 leads to accumulation of plasma apoCIII on VLDL (very low density lipoprotein) particles which in turn prevents efficient lipolysis of VLDL-triglycerides by lipoprotein lipase (LpL) and leads to hypertriglyceridemia. This mechanism of hypertriglyceridemia is shown herein with reference to apoCIII, but can be extrapolated to other apolipoproteins which interact with ABCA1 and are linked to hypertriglyceridemia, notably, for example, apoE, apoA-l and apoA-ll. Hence, increasing expression of ABCA1 is now indicated as a new therapeutic approach to treating hypertriglyceridemia, and other conditions of metabolic syndrome such as diabetes and coronary heart disease when linked with over- expression of apolipoprotein and accumulation of triglyceride-rich VLDL with abnormal apolipoprotein composition, especially, for example, apoCIII- and triglyceride-rich VLDL.

Background to the invention

ApoCIII is a 79 amino acid glycoprotein synthesized by the liver and intestine(1). It is an important component of the lipid and lipoprotein transport system and is primarily responsible for modulating plasma triglyceride levels. As further discussed below, excess apoCIII in plasma results in hypertriglyceridemia, one of the components of metabolic syndrome, mainly due to inhibition of plasma LpL activity (16; 17; 31-33).

It is currently believed that apoCIII is secreted in a lipid poor form in the circulation (2) where it subsequently associates with pre-existing very low density (VLDL), low density (LDL), and high density (HDL) lipoproteins (3). Approximately one-half of apoCIII is found in HDL (3; 4), and the remaining in VLDL and chylomicrons (3). Small quantities of apoCIII are also found in intermediate density lipoprotein (IDL) and low density lipoprotein (LDL) particles (5;6). ApoCIII is also found as a component of particles with defined lipid and apolipoprotein composition such as LpB:E:CIII, LpAI:AII:CIII, and other combinations (7;8). In addition, in vitro experiments have established that association of purified lipid-free apoCIII with egg lecithin forms discoidal particles with a minor axis of 4 nm and a major axis of 20 nm (9, 10).

As indicated above, apoCIII plays a major role in plasma triglyceride metabolism. Numerous epidemiological and animal studies have established a direct correlation of plasma apoCIII levels to plasma triglyceride levels, and an inverse relationship to the rate of post-prandial lipoprotein clearance (11-18).

Many in vitro studies have suggested an important role of apoCIII in the catabolism of triglyceride-rich lipoproteins. ApoE-mediated uptake of triglyceride rich emulsions by cultured HepG2 cells and rat hepatocytes was inhibited by apoCIII (19;20). Liver perfusion studies indicated that apoCs, including apoCIII, inhibited the hepatic uptake of chylomicrons and VLDL (1 ;19-28). In addition, apoCIII inhibited the binding of apoB- containing lipoproteins to the LDL-receptor probably by masking the binding domain of apoB, as well as the binding of chylomicrons and VLDL to the lipolysis stimulated receptor (24).

Other studies showed that apoCIII inhibits the activity of lipoprotein lipase, the enzyme that hydrolyzes the triglycerides of chylomicrons and VLDL (29; 30). In patients with apoA-l and apoCIII deficiency, lack of apoCIII facilitates the conversion of triglyceride- rich VLDL to IDL and LDL (31). In hypertriglyceridemic patients, apoCIII was found to be a specific inhibitor of lipoprotein lipase (LpL) (32; 33) while purified apoCIII acted as a competitive inhibitor of apoCII in the hydrolysis of triolein (32). In vitro studies using synthetic peptides suggested that the N-terminal domain of apoCIII is responsible for the inhibitory effect of apoCIII on LpL activity (34) though this result was never confirmed in vivo. Overall, there have been three different mechanisms proposed to explain apoCIII-induced hypertriglyceridemia: a) apoCIII displaces apoCII from VLDL and reduces LpL activation by apoCII (47), b) apoCIII acts as a direct noncompetitive inhibitor of LpL (16-18), and c) apoCIII displaces apolipoprotein E (apoE) from the VLDL thus reducing the apoE- and LDLr- mediated clearance of triglyceride-rich lipoproteins (14; 15).

In addition to the role of apoCIII in the development of hypertriglyceridemia, recent studies suggested a direct link between plasma apoCIII levels and the development of other conditions associated with metabolic syndrome. Specifically, it was found that apoCIII levels in plasma show a direct correlation with body mass index and insulin resistance (35). Furthermore, studies in humans showed that accumulation of apoCIII- containing VLDL and chylomicrons in plasma are strong predictors of coronary heart disease (36-38).

Previous studies by the inventor (50) and others (reviewed in 52) have shown that lipid poor apoE and apoA-l interact functionally with ABCA1 and this interaction is required for the formation of apoE and apoA-l-containing HDL. Previous studies have also shown that under conditions of over-expression, apoE (46; 55), apoA-l (56) and apoA-ll (57) accumulate on TG-rich VLDL and inhibit LpL activity. As detailed further in the exemplification herein, the inventor has now additionally investigated the ability of ABCA1 and apoCIII to promote the cfe novo biogenesis of apoCIII-containing HDL particles in vivo and the role of these particles in apoCIII-induced hypertriglyceridemia.

Adenovirus-mediated gene transfer was used to express the wild-type human apoCIII under the control of the cytomegalovirus (CMV) promoter (39) in apoE ' x apoA-l ^" ' ^" mice or ABCA1 ^" ' ^" mice. A moderate dose of apoCIII-expressing adenovirus was used that does not trigger hypertriglyceridemia in wild-type C57BL/6 mice in order to avoid non- physiologically high levels of apoCIII expression. The data showed that apoCIII promotes the de novo biogenesis of HDL-like particles in apoE ^" ' ^" x apoA-l ^" ' ^" mice that are distinct from classical apoE- and apoA-l-containing HDL. This process requires the lipid transporter ABCA1 , since its deficiency prevented formation of apoCIII-containing HDL. Furthermore, as indicated above, deficiency in ABCA1 promoted the accumulation of plasma apoCIII exclusively on the VLDL, and triggered hypertriglyceridemia in ABCAT ^7" mice that were infected with the recombinant adenovirus. Control experiments showed that plasma apoCIII levels and hepatic mRNA expression of apoCIII were similar in apoE ^"7" x apoA-l ^"7" mice and ABCAT' ^" mice infected with the recombinant adenovirus, confirming that the differences in plasma triglyceride levels between these two mouse groups were not due to differences in apoCIII expression levels.

Taken together, the data supports the conclusion that de novo biogenesis of apoCIII- containing HDL with the participation of ABCA1 is a key process contributing to the prevention of apoCIII-induced hypertriglyceridemia; formation of apoCIII-containing HDL prevents accumulation of excess plasma apoCIII on VLDL and allows for the efficient lipolysis of VLDL-triglycerides by LpL (see Figure 9). Under conditions of apoCIII-overexpression, the ABCA1 -mediated pathway of apoCIII-HDL formation may become saturated; in this case, excess lipid poor apoCIII will become available for binding to VLDL leading to hypertriglyceridemia.

Since lipid poor apoE, apoA-l and apoA-ll also interact functionally with ABCA1 (50; 52; 53) it is further postulated that ABCA1 has a similar role as with apoCIII in promoting association of apoE, apoA-l and apoA-ll with HDL whereby HDL particles provide buffering capacity against accumulation of such apolipoproteins on VLDL and decreased efficiency of lipolysis of VLDL triglycerides.

In humans, mutations in ABCA1 that impair its function cause Tangier's disease (58), an autosomal codominant disorder that is characterised by an extremely marked reduction in plasma HDL cholesterol and mild to moderate hypertriglyceridemia (59; 60). In previous studies, analysis of composition (59;60) and post heparin-lipolytic activities of VLDL (59) from patients with Tangier's disease showed that deficiency in ABCA1 results in abnormal apolipoprotein compositon of VLDL, reduced reactivity of VLDL -triglycerides with plasma LpL, and hypertriglyceridemia (59). The findings presented herein provide a full mechanistic interpretation for the hypertriglyceridemia associated with Tangier's disease (see the minus ABCA1 branch in Figure 9) and at the same time make evident additional use of increasing expression of ABCA1 in combating hypertriglyceridemia beyond that triggered by ABCA1 deficiency.

Summary of the invention

On the basis of the data provided herein, as hereinbefore indicated there is now provided a new therapeutic approach for treating apoCIII-induced hypertriglyceridemia linked to over-expression of apoCIII, which can be extended to hypertriglyceridemia associated with other apolipoproteins which interact functionally with ABCA1 , such as apoE, apoA-l, and apoA-ll, and more generally to other disease aspects of metabolic syndrome when linked with higher than normal plasma level of an apolipoprotein and accumulation of triglyceride-rich VLDL with abnormal apolipoprotein composition, especially apoCIII-and triglyceride-rich VLDL. More particularly, in one aspect the present invention provides an agent which increases availability in vivo of the lipid transporter ABCA1 , or a functional variant thereof, for use in treating one or more disease aspects of metabolic syndrome when linked to over-expression of one or more apolipoproteins leading to accumulation of triglyceride-rich VLDL with abnormal apolipoprotein composition, particularly for example apoCIII-induced hypertriglyceridemia associated with apoCIII- and triglyceride-rich VLDL. The administration may be to a human or non-human mammal. In this context, it will be understood that 'functional variant' extends to any variant of wild-type ABCA1 , which, like wild-type ABCA1 , will promote association of one or more apolipoproteins of concern, e.g apoCIII, with HDL. Thus, where the administration is to a human, the availability of human ABCA1 or any suitable natural or mutant version thereof may be increased in vivo such that association of one or more apolipoproteins of concern, e.g. apoCIII, or possibly additionally or alternatively, for example, any of apoE, apoA-l or apoA-ll, is favoured with HDL particles over VLDL particles. In this way, formation of the disease-associated triglyceride-rich VLDL is reduced. It is envisaged that functional variants of a wild-type ABCA1 protein suitable for provision in vivo in accordance with the invention may include truncated and/or mutated versions provided there is retention of the required activity to promote association of apolipoprotein with HDL particles as discussed above.

As indicated above, in addition to hypertriglyceridemia, treatment in accordance with the invention, may also assist other aspects of metabolic syndrome such as insulin resistance, resulting in type Il diabetes, and coronary heart disease when linked with over-expression of one or more apolipoproteins leading to accumulation of triglyceride- rich VLDL with abnormal apolipoprotein composition, such as apoCIII- and triglyceride rich VLDL. Accumulation of such VLDL particles may arise through accumulation of excess plasma apoCIII whereby the normal route of biogenesis of apoCIII-containing HDL particles requiring functional ABCA1 is compromised. In this case, an agent which provides ABCA1 (or a functional variant thereof as noted above) can be expected to promote formation of apoCIII-rich HDL and thereby increase the ratio of apoCIII associated with HDL to apoCIII associated with VLDL towards the normal for non-disease controls.

It is extrapolated that in similar fashion aspects of metabolic syndrome, e.g. hypertriglyceridemia, may additionally or alternatively be triggered by over-expression of one or more further apolipoproteins, especially, for example, any of apoE, apoA-l and apoA-ll. Again, the anticipated consequence mirrors that to be expected from reduced ABCA1 activity-accumulation of triglyceride-rich VLDL particles with abnormal apolipoprotein composition linked with inhibition of LpL activity on VLDL triglycerides. The therapeutic approach of the invention is applicable to all such situations leading to disease development, e.g. apolipoprotein-induced hypertriglyceridemia. From another perspective, there is also provided use of an agent as defined above for the manufacture of a medicament for use in treating one or more disease aspects of metabolic syndrome when linked to over-expression of one or more apolipoproteins leading to accumulation of triglyceride-rich VLDL with abnormal apolipoprotein composition, such as apoCIII-induced hypertriglyceridemia triggered by over- expression of apoCIII and higher than normal apoCIII content of VLDL.

From an additional perspective, there is provided a method of treating one or more disease aspects of metabolic syndrome when linked to over-expression of one or more apolipoproteins as discussed above, especially for example apoCIII-induced hypertriglyceridemia triggered by over-expression of apoCIII, which comprises administering an agent which increases availability in vivo of ABCA1 or a functional variant thereof as defined herein.

Detailed description of the invention

An agent for therapeutic use in accordance with the invention may preferably comprise a polynucleotide capable of expressing the lipid transporter ABCA1 or a suitable functional variant thereof in target liver cells. ABCA1 mRNA has been cloned and the cDNA sequence has been deposited in Genbank with accession number NM_005502. A convenient cloned source of human native ABCA1 coding sequence is OriGene Technologies, lnc (Rockville, MD 20850, USA) from which human ABCA1 cDNA is commercially available (cat# SC 127939). However, as indicated above, variants of native ABCA1 (such as encoded by truncated and /or mutant forms of human ABCA1 cDNA) may also be employed provided they promote desired association of apolipoprotein with HDL, especially association of apoCIII, and possibly also other apolipoproteins such as apoE, apoA-l and apoA-ll, with HDL, Such variants might be readily tested in vivo, using an animal model deficient in ABCA1 , as described in the exemplification. Suitable variants may, for example, comprise or consist of a sequence having at least 90% identity to wild-type human ABCA1 protein and which retain the desired function of the wild-type protein. Such a variant may have one or more additions and/or deletions and/or substitutions, e.g. one or more conservative substitutions.

An agent for gene therapy as above may most desirably be a recombinant viral vector. Such a vector may, for example, be a recombinant adenovirus, gutless adenovirus, lentivirus, herpes virus, adeno-associated virus or retrovirus. Bassol et al. have described the use of a recombinant adenovirus to express human ABCA1 in vivo in the liver of mice (62). It is envisaged that in similar fashion, human ABCA1 , or a functional variant thereof, may be delivered in human liver for the purpose of treatment in accordance with the invention, although it will be appreciated that other types of viral vector may be employed provided they will direct expression of human ABCA1 or a suitable variant thereof in target hepatocytes. Administration will generally desirably be by intravenous injection. Alternatively, naked DNA administration via the circulation, e.g. by direct intravenous injection into the portal vein, might be considered. Such gene delivery methodologies are particularly suitable for the purpose of treatment according to the invention since the natural trophism of the vectors after intravenous administration is the liver - when injected in the circulation a high proportion of the vector will enter the liver and ABCA1 or ABCA1 variant expression will take place in the liver, which is the major site of normal endogenous ABCA1 expression.

A polynucleotide for therapeutic use in accordance with the invention may alternatively be packaged in the form of liposomes. Many liposomal formulations are now known for delivery of therapeutic oliognucleotides and genes in vivo, including oligonucleotide-carrying liposomes designed to improve targeting to the liver, and might be readily adapted for use in accordance with the invention by inclusion of a coding sequence for the ABCA1 protein or a variant thereof functionally linked to a suitable promoter. Again, administration may desirably be by intravenous injection.

In a further aspect, the invention provides an agent as above in the form of a pharmaceutical composition together with a pharmaceutically acceptable carrier or diluent. Such a composition will generally be a buffered composition suitable for systemic delivery, e.g. a buffered liposomal formulation or buffered suspension of a viral vector. Provision of such compositions is well within the capacity of those skilled in the gene therapy art

The invention is illustrated in the following exemplification with reference to the following figures. Brief description of the figures

FIGURE 1. Western-blot analysis of culture medium of HTB-13 cells infected with different multiplicities of infection (m.o.i) of the human apoCIM-expressing adenovirus.

Cultures of HTB-13 cells that do not express endogenous apoCIII were grown to confluence in 6-well plates and then infected with a m.o.i of 0, 3, 6, 12, and 24 of the

AdGFP-CIII _g adenovirus. Following infection for 24 hours, cultures were washed twice with PBS and fresh serum-free medium was added for an additional twenty four hours.

Then, twenty μl of culture medium were isolated and analyzed by western blot analysis using an anti human-apoCIII antibody. The molecular weight of the markers and the position of apoCIII are indicated.

FIGURE 2. Cholesterol and triglyceride levels, and apoCIII distribution among lipoproteins in plasma of WT C57BL/6 mice infected with the adenovirus expressing human apoCIII or the control adenovirus AdGFP. (Panel A) shows plasma cholesterol levels, and (Panel B) shows plasma triglyceride levels of C57BL/6 mice infected with a moderate dose of 8x10 ⁸ pfu (n= 6) or a high dose of 2x10 ⁹ pfu of AdGFP-CIII ₉ (n=6) or the control AdGFP adenovirus (n=4) The dotted line indicates the upper physiological plasma triglyceride level of 150 mg/dl. (Panels C and D) show western-blot analyses using an anti-human apoCIII specific antibody of plasma lipoprotein fractions isolated by density gradient ultracentrifugation of plasma from C57BL/6 mice infected with 8x10 ⁸ pfu (Panel C) or 2x10 ⁹ pfu (Panel D) of AdGFP-CIIIg. The densities and the lipoprotein classes they correspond to are indicated. The statistical significance of the observed differences among groups at each time point is as indicated ( ^* corresponds to p<0.05 and ^** corresponds to p<0.005).

FIGURE 3. Cholesterol and triglyceride levels of apoE ^" ' ^" x apoA-l ^" ' ^" mice and ABCA1 ^" ' ^" mice at different days post-infection with 8x10 ⁸ pfu of the adenovirus expressing the human apoCIII or the control AdGFP virus. Panels A and C show plasma cholesterol levels of apoE ' ^' x apoA-l ^"7' mice (n=6) (Panel A) and ABCA1 ^" ' ^' mice (n=8) (Panel C) at different days post-infection. Panels B and D show plasma triglyceride levels of apoE ^" ' ^" x apoA-l ^" ' ^" mice (n=6) (Panel B) and ABCA1 ^" ' ^' mice (n=8) (Panel D) at different days post-infection. Panel D also shows plasma triglyceride levels of ABCA1 ^" ' ^" mice infected with a mixture of 8x10 ⁸ pfu of the apoCIII-expressing adenovirus and 5x10 ⁸ pfu of the adenovirus expressing the human lipoprotein lipase (LpL) (n=6). The statistical significance of the observed differences among groups at each time point is as indicated (Corresponds to p<0.05 and ^** corresponds to p<0.005). FIGURE 4. Representative mRNA analysis of mice infected with 8x10 ⁸ pfu of the recombinant adenovirus expressing the human apoCIII. Total RNA was isolated from livers of infected mice on day 5 post-infection and analyzed by Northern blotting for human apoCIII and GAPDH mRNA levels. Panel A shows representative autoradiograms of Northern blot analysis of RNA from the livers of C57BU6, apoE ^" ' ^" x apoA-l ^" ' ^" mice, and

ABCA1 ^" ' ^" mice infected with 8x10 ⁸ pfu of AdGFP-CIII _g . Panel B shows the relative human apoCIII mRNA expression normalized for GAPDH expression in individual mice. Panel C shows the average relative apoCIII mRNA expression normalized for GAPDH mRNA levels, for each mouse strain (ns indicates statistically non-significant differences).

FIGURE 5. Cholesterol {Panels A, D), phospholipid (Panels B ₁ E) and triglyceride (Panels C, F), levels of lipoprotein fractions isolated by density gradient ultracentrifugation analysis of plasma samples of apoE ^* ' ^" x apoA-l ^" ' ^" mice (Panels A-C) and ABCA1 ^" ' ^" mice (Panels D-F) infected with 8x10 ⁸ pfu of the human apoC Ill-expressing adenovirus or the control adenovirus AdGFP, five days post-infection.

FIGURE 6. Total, free and esterified cholesterol levels of lipoprotein fractions isolated following density gradient ultracentrifugation analysis of plasma samples of apoE ^" ' ^" x apoA-l ^"A mice (Panels A, B) and ABCA1 ^" ' ^" mice (Panels C, D) infected with 8x10 ⁸ pfu of the control adenovirus AdGFP (Panels A, C) or the human apoCIII-expressing adenovirus (Panels B, D), five days post-infection.

FIGURE 7. ApoCIII distribution in different lipoprotein fractions isolated by density gradient ultracentrifugation of plasma from apoE ^"Λ x apoA-l ^" ' ^" mice and ABCA1 ^" ' ^" mice infected with the AdGFP-CIII ₉ adenovirus for five days. Plasma samples from apoE ' ^" x apoA-l ^7" mice (Panel A) and ABCA1 ^" ' ^" mice (Panel B) infected with the 8x10 ⁸ pfu of the AdGFP-CIIIg adenovirus were isolated on day 5 post-infection and fractionated by density gradient ultracentrifugation. Different density fractions were resolved on 15% SDS-PAGE and then analyzed by western blot analysis for human apoCIII using a specific anti-human apoCIII antibody. The densities of each fraction and the lipoprotein classes that they correspond to are indicated.

FIGURE 8. Electron microscopy analysis of HDL density fractions isolated from apoE ^" ' ^" x apoA-l ^"7" mice (Panels A, B) and ABCA1 ^"7' mice (Panels C, D) infected with 8x10 ⁸ pfu of the control adenovirus AdGFP (Panels A, C) or the AdGFP-CIII ₉ adenovirus (Panels B, D) five days post-infection. FIGURE 9. Schematic representation summarizing the role of apoCIII and ABCA1 in the biogenesis of apoCIII-containing HDL and the development of hypertriglyceridemia. ApoCIII is synthesized mainly in the liver and secreted in the form of lipid-poor apolipoprotein in the circulation. In the presence of ABCA 1 (+ABCA1 branch), apoCIII acquires cholesterol and promotes the cfe novo biogenesis of apoCIII-containing HDL particles that are distinct from the classical apoA-l- and apoE-containing HDL. The activation of this pathway results in the sequestration of apoCIII mainly in the HDL, thus limiting the amount of apoCIII available for association with VLDL. This results in normal LpL activity and lipolysis of VLDL triglycerides and the efficient removal of lipoprotein remnants from the circulation. In contrast, in the absence of ABCA1 (- ABCA1 branch), no apoCIII-containing HDL is formed, and the vast majority of lipid- poor plasma apoCIII accumulates on VLDL. This results in inhibition of LpL activity, impaired clearance of triglyceride-rich VLDL from the circulation and hypertriglyceridemia.

Example

MATERIALS AND METHODS

Construction of the recombinant adenovirus expressing the wild-type human apoCIII. The construction of pBMT3X-apoClll _g plasmid containing the genomic apoCIII sequence has been described previously (40). Subcloning of the genomic apoClll _g DNA from pBMT3X-apoClll _g plasmid into the Xho I sites of pBlueskript (pBK) vector created the pBK-apoCIM _g -sense and pBK-apoClll _g -antisense plasmids. In the pBK- Clll _g -antisense vector the 5' end of the apoClllg gene is flanked by a unique Kpn I restriction site and the 3' end is flanked by a unique Xba I site of the pBK vector. Subcloning of the Kpnl / Xbal fragment of the pBK-apoClll _g -antisense vector into the Kpn I and Xba I sites in the polylinker of the pAdTrack-CMV vector (39) generated the pAdTrack-CMV-apoClllg vector where the genomic apoCIII sequence is under the control of the CMV promoter.

The recombinant adenovirus was constructed using the AdEasy-1 system where the complete adenovirus genome is generated by homologous recombination in bacteria

BJ-5183 cells (39). Specifically, the recombinant vector pAdTrack-CMV-apoCIM _g was linearized by Pme I digest, and then electroporated into BJ-5183 E. coli cells along with the pAdEasy-1 helper vector. pAdEasy-1 contains the viral genome and the long terminal repeats of the adenovirus and allows for the formation by homologous recombination of the recombinant virus containing the gene of interest. Recombinant bacterial clones resistant to kanamycin were selected and screened for the presence of the gene of interest by restriction endonuclease analysis and DNA sequencing. The resulting recombinant vector also contained the green fluorescence protein (GFP) gene under the independent control of a second CMV promoter, which enables detection of the infection in cells and tissues by their green fluorescence. The recombinant adenovirus expressing the WT human apoCIII was designated as AdGFP-CIII ₉ . Correct clones were propagated in RecA DH5α cells. The recombinant vector was linearized with Pad and used to infect 91 1 cells (41). The subsequent steps involved in the generation and expansion of recombinant adenoviruses were plaque identification/isolation followed by infection and expansion in 911 cells (41). These steps were followed by a purification process involving CsCI ultracentrifugation performed twice, followed by dialysis and titration of the virus. Usually, titers ranging from 2x10 ¹⁰ to 5x10 ¹⁰ pfu/ml were obtained. A control virus expressing only GFP (AdGFP) (42) was also used in this study in order to correct for non-specific effects of the infection process.

Cell culture studies. HTB13 cells (SW1783, human astrocytoma) were grown to confluence in medium containing 10% fetal calf serum (FCS). Confluent cultures were washed twice with Phosphate buffered saline (PBS), switched to culture medium containing 2% heat-inactivated horse serum, and then infected with AdGFP-CIII ₉ at multiplicity of infection (moi) 3, 6, 12, and 24, as indicated. Twenty-four hours postinfection, cells were washed twice with PBS, and fresh serum-free medium was added. Following 24h of incubation, medium was collected and analyzed by enzyme linked immunoabsorbent assay (ELISA) and western blot analysis for apoCIII expression.

Animal studies. Mice were purchased from Jackson Labs (www.iax.org). ABCA1 ' ^" mice were generated by crossing ABCA1 ^+/" mice (43). Female apoE ^" ' ^" x apoA-l ^"7" mice (generated in our lab by crossing apoE ^" ' ^" mice (44) to apoA-l ^" ' ^" mice (45) and ABCA1- deficient mice (43) 8-10 weeks old were used in these studies. Groups were formed after determining the fasting cholesterol and triglyceride levels of the individual mice, to ensure similar average cholesterol and triglyceride levels among groups. For the adenovirus infections, groups of 8-10 mice were injected intravenously through the tail vein with doses of 8x10 ⁸ pfu of the apoClll _g -expressing adenovirus or the control adenovirus AdGFP. Blood was obtained daily following a 4h fasting period, 3 to 5 days post-injection. Aliquots of plasma were stored at 4° and -20 ⁰ C. RNA analysis. To assess the expression of hepatic human apoCIII mRNA, three mice from each group were sacrificed 4 days post-infection. Livers were collected from individual animals, frozen in liquid nitrogen, and stored at -80 ⁰ C. Total RNA was isolated from the livers and analyzed for apoCIII mRNA and GAPDH expression by Northern blotting followed by phosphorimaging (42).

Plasma lipid determination. Plasma triglycerides, total cholesterol, free cholesterol, and phospholipids were determined spectrophotometrically using the GPO-Trinder Kit (Sigma, cat# TR0100-1 KT), the Infinity Cholesterol kit (Thermo Electron Corporation, cat# TR13521), the Free Cholesterol E Kit (Wako, cat# 435-35801), and the Phospholipid C determination kit (Wako, cat# 433-36201) respectively, according to the manufacturer's instructions and as described previously (42).

Fractionation of plasma lipoproteins by density gradient ultracentrifugation. To assess the ability of human apoCIII to associate with different lipoproteins, 0.35 ml of pools of plasma from WT C57BL/6 or apoE ^" ' ^" x apoA-l ^" ' ^" or ABCA1 ^" ' ^" mice infected with the adenovirus expressing the WT human apoCIII were fractionated by density gradient ultracentrifugation and different density fractions were analyzed for human apoCIII protein levels by western blotting, as described below. The lipid composition of each fraction was also determined as described above.

Western blot analysis of apoCIII. Western blot analysis for apoCIII was performed as described previously (42;46-49), using a goat anti-human apoCIII antibody (Biodesign International, cat# K74140G) as primary, and a rabbit anti-goat antibody (Santa-Cruz, cat # sc-2768) as secondary.

Quantification of human apoCIII by ELISA. Serum human apoCIII concentrations were measured by sandwich ELISA (42; 46-49). A polyclonal goat anti-human apoCIII antibody (Biodesign International, cat # K74140G) was used for coating microtiter plates, and a polyclonal goat anti-human apoCIII antibody coupled to horse radish peroxidase (Biodesign International, cat# K74140P), was used as secondary antibody. The immunoperoxidase procedure was employed for the colorimetric detection of apoCIII at 450 nm, using tetramethylbenzidine as substrate. Different dilutions of plasma from healthy human subjects with apoCIII level of 10 mg/dl were used as standards. Rate of hepatic VLDL triglyceride production in mice infected with AdGFP-CIII ₉ . To compare the effects of human apoCIII expression on hepatic VLDL triglyceride secretion in apoE ^"7" x apoA-l ^'7" and ABCA1 ^" ' ^" mice, 6 mice for each group were infected with a dose of 8x10 ⁸ pfu of AdGFP-CIII ₉ adenovirus. Four days post-infection, mice were fasted for 4 hours and then injected with Triton-WR1339 at a dose of 500 mg/kg of body weight, using a 15% solution (w/v) in 0.9% NaCI (Triton-WR 1339 has been shown to completely inhibit VLDL catabolism (13), as described previously (42; 46-49). Briefly, serum samples were isolated for 5 min, 10 min, 20 min, 40 min, 60 min, and 90 min after injection with Triton WR 1339. As control, serum samples were isolated 1 min after injection with the detergent. Plasma triglyceride levels at each time-point were determined as described above and linear graphs of triglyceride concentration vs. time were generated. The rate of VLDL-triglyceride secretion expressed in mg/dl/min was calculated from the slope of the linear graphs for each individual mouse. Then, slopes were grouped together and reported as mean ± standard error of the mean in the form of a bar graph.

Electron microscopy analysis of HDL. Pools of the HDL fractions isolated from equilibrium density gradient ultracentrifugation were dialyzed against ammonium acetate and carbonate buffer, and then stained with sodium phosphotungstate. Particles were visualized in a Phillips CM-120 electron microscope (Phillips Electron Optics, Eindhoven, Netherlands), and photographed as described previously (49; 50). The photomicrographs were taken at x75,000 magnification and enlarged three times. Size measurements of particles (n>200 from each population) were taken directly from image negatives (x75,000) with a 7X measuring magnifier with metric scale of 0.1 mm divisions. Measured values were classified into 1.3 nm intervals and the diameter of each value within an interval was represented by the midpoint of that interval. Descriptive statistics of a measured population of at least 200 particles were prepared using the Analysis Tool Pack in Microsoft Excel 2002.

Statistical analysis. Comparison of data from two groups of mice was performed using the Student t-test. Data are reported as mean ± standard error of the mean. * indicates p<0.05, ^** indicates p<0.005, ns indicates statistically non-significant differences, and n indicates the number of animals tested in each experiment. RESULTS

Expression and secretion of apoCIII by cultures of HTB-13 cell infected with the recombinant adenovirus expressing the human apoCIII. Human astrocytoma HTB-13 cells that do not synthesize endogenous apoCIII were infected with the recombinant adenovirus AdGFP-CIIIg, at a multiplicity of infection of 0, 3, 6, 12, and 24. Western blot analysis of the medium from the infected cultures showed that human apoCIII is secreted efficiently in the culture medium 24 hours post-infection (Figure 1). Analysis of the culture medium by sandwich ELISA confirmed that apoCIII is secreted in the media of the infected cultures at concentrations ranging from 10 to 50 μg of apoCIII per ml of culture medium, 24h after infection.

Expression of apoCIII in wild-type C57BU6 mice infected with the recombinant adenovirus containing the human apoCIH _g . To confirm that infection of mice with the AdGFP-CIII _g adenovirus results in the efficient production and secretion of human apoCIII in the plasma of these mice, wild-type C57BL/6 mice were injected through the tail-vein with a moderate dose of 8x10 ⁸ pfu or a high dose of 2x10 ⁹ pfu of the AdGFP- CIII _g adenovirus and plasma samples were collected on days 3, 4 and 5 post-infection. Then, plasma levels of human apoCIII, cholesterol, and triglycerides were determined as a function of time.

As shown in Figure 2, infection with 2x10 ⁹ pfu of AdGFP-CIII ₉ triggered severe hypertriglyceridemia and a significant increase (p<0.05) in plasma cholesterol levels of the infected mice (Figure 2A, B), a finding consistent with previous results on apoCIII overexpression in vivo (16-18) In contrast, infection of mice with the lower dose of 8x10 ⁸ pfu of AdGFP-CIIIg resulted in only a modest increase (p<0.05) in their plasma triglyceride levels which however remained within normal levels (≤150 mg/dl) throughout the length of the experiment (Figure 2B). This modest increase in plasma triglycerides was accompanied by a modest increase in plasma cholesterol levels of these mice (Figure 2A). As control, infection of C57BL/6 mice with 2x10 ⁹ pfu of the empty AdGFP virus had no effects on the plasma triglyceride or cholesterol levels of these mice confirming that there are no non-specific effects on plasma lipid levels due to the infection (Figure 2A, B). ELISA analysis of plasma samples isolated on day 5 post-infection showed that steady-state human plasma apoCIII levels were in the range of 50 to 80 mg/dl in C57BL/6 mice infected with 2x10 ⁹ pfu of AdGFP-CIII _g and 15 to 25 mg/dl in C57BL/6 mice infected with 8x10 ⁸ pfu of AdGFP-CIII _g . Fractionation of plasma isolated from the infected mice by density gradient ultracentrifugation followed by western blotting for human apoCIII showed that, in C57BL/6 mice infected with 8x10 ⁸ pfu of AdGFP-CIIIg, apoCIII was mainly found in HDL and to a lesser extend in VLDL/IDL (Figure 2C). However, in C57BL/6 mice infected with 2x10 ⁹ pfu of AdGFP-CIIIg, apoCIII levels increased in both VLDL/IDL and HDL while a significant amount of apoCIII was also found on the LDL fractions (Figure 2D), in agreement with previously published studies (3-6).

Therefore, infection of C57BL/6 mice with AdGFP-CIIIg results in the efficient expression and secretion of human apoCIII in the plasma of the infected mice. In addition, the recombinant adenovirus AdGFP-CIII ₉ provides a versatile in vivo expression system that depending on the dose produces different phenotypes that are similar to those previously reported in apoCIII transgenic mouse models.

Plasma lipid and apoCIII levels, and hepatic apoCIII mRNA expression in apoE' ^~ x apoA-r ^A , and ABCAT ^A mice following adenovirus infection. In the following studies, it was selected to use the moderate dose of 8x10 ⁸ pfu of AdGFP-CIII ₉ adenovirus that does not trigger hypertriglyceridemia when administered to wild-type C57BL/6 mice.

Six to eight apoE ^" ' ^" x apoA-l ' ^" or ABCA1 ^"Λ mice were infected with 8x10 ⁸ pfu of the human apoCIII-expressing adenovirus or the control AdGFP adenovirus and fasting plasma cholesterol and triglyceride levels were determined 4 and 5 days post-infection.

In apoE ^"7' x apoA-l ^" ' ^" mice, expression of human apoCIII resulted in a modest elevation of their plasma cholesterol levels on days 4 and 5 post-infection (p<0.05) (Figure 3A). However, plasma triglyceride levels remained normal (<150 mg/dl) during the course of the experiment despite a modest increase on day 5 post-infection (p<0.05) (Figure 3B). In contrast, expression of human apoCIII in ABCA1 ^" ' ^" mice had no significant effects on plasma cholesterol levels of these mice (p>0.05) (Figure 3C) although it triggered severe hypertriglyceridemia on days 4 and 5 post-infection (p<0.005) (Figure 3D).

Treatment of apoE ' x apoA-l ^" ' ^' or ABCA1 ^" ' ^" mice with the control adenovirus expressing the green fluorescence protein (AdGFP) did not change significantly the plasma cholesterol and triglyceride levels of these mice (Figure 3A-D). Potential liver damage following adenovirus infection was assessed by measuring serum transaminases using the Reflotron Plus system (Roche) and normal serum transaminase levels were found in mice infected with 8x10 ⁸ pfu of the recombinant adenoviruses.

To confirm that the differences in plasma cholesterol and triglyceride levels between apoE ^" ' ^" x apoA-l ^" ' ^' and ABCA1 ^" ' ^" mice infected with 8x10 ⁸ pfu of AdGFP-CIII _g were not due to different levels of apoCIII expression, we determined plasma apoCIII levels and hepatic mRNA expression in these mice. Analysis of steady-state plasma apoCIII levels on days 4 and 5 post-infection by ELISA showed that apoE ^' ' ^" x apoA-l ' ^" mice infected with 8x10 ⁸ pfu of AdGFP-CIIIg had higher steady-state plasma human apoCIII levels compared to ABCA V' ^* mice infected with the same dose of the human apoCIII- expressing adenovirus. Specifically, plasma levels of human apoCIII in AdGFP-CIIIg- infected apoE ^" ' ^" x apoA-l ^"Λ mice were 18.22±2.32 mg/dl on day 4, and 19.14±2.69 mg/dl on day 5 while plasma levels of human apoCIII in AdGFP-CIM _g -infected ABCA1 ^" ' ^" mice were 16.31 ±2.31 mg/dl on day 4, and 13.85±1.75 mg/dl on day 5.

Furthermore, northern blot analysis of RNA from livers of the infected mice five days post-infection showed similar average hepatic human apoCIII mRNA expression in C57BL/6, apoE ^" ' ^" x apoA-l ^" ' ^" and ABCA1 ^" ' ^" mice infected with 8x10 ⁸ pfu of the AdGFP- CIII _g (Figure 4A-C). However, only infected ABCA1 ^" ' ^" mice developed severe hypertriglyceridemia (Figure 3D) while infected C57BL/6 and apoE ^" ' ^" x apoA-l ^" ' ^" mice had normal triglyceride levels (Figures 2B and 3B).

The hypercholesterolemia of apoE' ^' x apoA-l ^'A mice expressing human apoCIII is due to increased accumulation of cholesterol in the LDL and HDL while the hypertriglyceridemia of ABCAI ^' ' ^' mice expressing apoCIII is due to accumulation of triglyceride-rich VLDL in plasma. To determine the distribution of phospholipids, triglycerides, and total, free, and esterified cholesterol among lipoproteins of apoE ^" ' ^" x apoA-l ^" ' ^" and ABCA1 ^" ' ^" mice infected with 8x10 ⁸ pfu of AdGFP-CIII _g1 plasma samples were isolated 5 days post-infection and fractionated by density gradient ultracentrifugation, as described in Materials and Methods. Then, different density fractions were collected and analyzed for total, free, and esterified cholesterol, triglyceride, and phospholipid levels.

This analysis showed that expression of apoCIII in apoE ^" ' ^' x apoA-l ^"7' mice resulted in an increase in plasma cholesterol levels of the LDL and HDL density fractions while VLDL cholesterol remained unchanged, compared to mice infected with the control AdGFP adenovirus (Figure 5A). The increase in LDL and HDL cholesterol was accompanied by an increase in free cholesterol levels (compare figures 6A to 6B) ₁ consistent with the role of apoCIII as an inhibitor of the plasma enzyme lecithin:cholesterol-acyl transferase (LCAT) (51). Consistent with the increased cholesterol levels, phospholipid levels were also significantly elevated mainly in LDUIDL and HDL fractions of these mice while there was only a modest increase in their VLDL/IDL fraction (Figure 5B).

In contrast, ABCA1 ^" ' ^" mice infected with 8x10 ⁸ pfu of the AdGFP-CIIIg or the control adenovirus AdGFP, had similar but very low levels of total cholesterol. All cholesterol was found only in the VLDL/IDL density fraction (Figure 5D) and was esterified (Figure

6C, D). No cholesterol was found in the LDL or HDL density fractions of both mouse groups. However, there was a significant increase in phospholipid levels of the

VLDL/IDL fraction of the ABCA1 ^~ ' ^~ mice expressing apoCIII, compared to ABCAI ^" ' ^' mice infected with the control AdGFP virus (Figure 5E).

Expression of human apoCIII in apoE ^" ' ^" x apoA-l ^" ' ^" mice resulted in a modest increase in the triglyceride levels of their VLDL/IDL fraction and to a lesser extend of their LDL fractions, compared to mice infected with the control AdGFP adenovirus which had very low levels of triglycerides present only in the VLDL/IDL density fractions (Figure 5C).

However, ABCA1 ^" ' ^" mice infected with AdGFP-CIII ₉ showed that the increase in total plasma triglyceride levels of these mice was due to a dramatic increase in the triglyceride content of their VLDL/IDL fraction and to a much lesser extend in LDL/IDL fraction, compared to the control AdGFP infected ABCA1 ^" ' ^" mice (Figure 5F).

The increased VLDL-triglyceride content of the ABCA1 mice infected with 8x1 Cf pfu of AdGFP-CIIIg correlates with the exclusive accumulation of human apoCIII on the VLDL of these mice. To determine the distribution of human apoCIII among different lipoprotein classes, plasma samples were isolated from apoE ' x apoA-l ^~ ' ^~ and ABCAT' ^" mice on day 5 post-infection with 8x10 ⁸ pfu of AdGFP-CIII ₉ . Then, lipoproteins were separated by equilibrium density ultracentrifugation and different lipoprotein fractions were analyzed by western blotting using an anti-human apoCIII specific antibody, as described in Materials and Methods. This analysis showed that in apoE ^"7" x apoA-l ^"7" mice infected with AdGFP-CIII _g , apoCIII was distributed in the VLDUIDL, LDL, and HDL density fractions (Figure 7A) consistent with the distribution of total cholesterol and phospholipids among these fractions (Figure 5A, B). In contrast, in ABCAT ^7" mice infected with AdGFP-CIIIg, apoCIII accumulated exclusively on the VLDL/IDL fraction (Figure 7B), consistent with the accumulation of total cholesterol, phospholipids, and triglycerides in this fraction (Figure 5D-F). ELISA analysis showed that VLDL of apoE ^"7' x apoA-l ^"7" mice infected with AdGFP-CIII _g had a human apoCIII content of 3.32±0.37 mg/dl, while VLDL of ABCAT ^7" mice infected with the same virus had a human apoCIII content of 11.78±0.23 mg/dl.

Hepatic VLDL-TG secretion assay did not reveal any significant differences in the secretion rates of VLDL-triglycerides between apoE ' x apoA-l ^" ' ^" mice and ABCAT' ^' mice infected with 8x10 ⁸ pfu of the human apoCIII-expressing adenovirus (secretion rates were 4.3±0.1 mg/dl/min for apoE ^"7" x apoA-l ^" ' ^" mice vs. 4.5±0.3 mg/dl/min for ABCAT ^7" mice). However, co-infection of ABCAT ^7" mice with 8x10 ⁸ pfu of AdGFP-CIII ₉ and 5x10 ⁸ pfu of an adenovirus expressing the human lipoprotein lipase (LpL)(49) ameliorated the apoCIII-induced hypertriglyceridemia (Figure 3D) suggesting that accumulation of apoCIII on the VLDL of ABCAT' ^" mice resulted in inhibition of LpL-mediated lipolysis of VLDL-triglycerides and triggered hypertriglyceridemia.

ApoCIII promotes the biogenesis of HDL particles in apoE' ^~ x apoA-f ^' mice and this process requires the participation of ABCA1. Since infection of apoE ' x apoA-l ^" ' ^" mice with AdGFP-CIII _g led to an increased accumulation of human apoCIII in the HDL, in the next set of experiments it was sought to determine whether expression of apoCIII in these mice promotes the biogenesis of HDL particles. Negative staining electron microscopy analysis of a pool of HDL fractions isolated from apoE ^"7' x apoA-l ^" ' ^" mice infected with 8 x 10 ⁸ pfu of the control AdGFP virus for 5 days, revealed the presence of only very small and very few particles and no HDL formation (Figure 8A). In contrast, similar analysis revealed the formation of a mixture of discoidal and spherical HDL particles in apoE ^"7" x apoA-l ^" ' ^" mice following infection with 8 x 10 ⁸ pfu of the human apoCIII-expressing adenovirus for 5 days (Figure 8B).

To test the involvement of ABCA1 in the biogenesis of apoCIII-containing HDL, we performed a similar electron microscopy analysis on pools of HDL fractions isolated from plasma of ABCAT ^7" mice infected with 8x10 ⁸ pfu of either the control AdGFP or the AdGFP-CIIIg adenovirus, five days post-infection. As expected ABCA1 ^'Λ mice infected with the control AdGFP adenovirus, had few very small (non-HDL sized) particles (Figure 8C). A similar pattern was also obtained in ABCA1 ^" ' ^" mice infected with AdGFP-CIII _g for 5 days (compare Figure 8D to Figure 8C), consistent with the absence of apoCIII from the HDL density fractions of these mice (Figure 7B).

DISCUSSION

As noted above, ApoCIII is an important component of the lipid and lipoprotein transport system and is primarily responsible for modulating plasma triglyceride levels. Excess apoCIII in plasma results in hypertriglyceridemia, one of the components of the metabolic syndrome, mainly due to inhibition of plasma LpL activity (16; 17;31-33). In the above study, a moderate dose (8x10 ⁸ pfu) of the apoCIII-expressing adenovirus that does not trigger hypertriglyceridemia in WT C57BL/6 mice was used. This moderate level of human apoCIII expression permitted careful assessment of the sensitivity of ABCA1 ^" ' ^" mice to hypertriglyceridemia in response to apoCIII expression and the contribution of ABCA1 to the formation of apoCIII-containing HDL. The data establish that deficiency in ABCA1 prevents the formation of apoCIII-containing HDL, results in excess accumulation of apoCIII on VLDL and increases sensitivity towards apoCIII-induced hypertriglyceridemia.

As also noted above, previous studies by the inventor (5) and others (reviewed in 52) have shown that lipid poor apoE and apoA-l interact functionally with ABCA1 and this interaction is required for the formation of apoE and apoA-l-containing HDL. The above study looked at the ability of apoCIII to promote the de novo biogenesis of HDL using apoE ^" ' ^" x apoA-l ^" ' ^" mice that do not form classical apoE- or apoA-l- containing HDL particles. It is noted that in apoE ^" ' ^" x apoA-l ^"Λ mice infected with 8x10 ⁸ pfu of AdGFP- CIIIg, a significant amount of human apoCIII was found in HDL, while the rest was distributed among VLDL/IDL, and LDL fractions. In agreement with these findings, further analysis of the HDL fraction by electron microscopy revealed the presence of a mixture of spherical and discoidal particles.

It is common belief that lipid poor apoCIII associates randomly in the circulation with existing classical apoE- and apoA-l-containing HDL (2). However, the data now presented establishes that the accumulation of apoCIII on HDL in vivo is not simply the result of a random association of apoCIII with pre-existing HDL since infection of apoE ^" ' ^" x apoA-l ^" ' ^" mice (that do not have endogenous apoE- or apoA-l-containing HDL) with AdGFP-CIIIg results in the accumulation of apoCIII in HDL density fractions and the formation of a mixture of discoidal and spherical HDL-like particles. Furthermore, this process depends on the action of the lipid transporter ABCA1.

The requirement of ABCA 1 in the formation of apoCIII-containing HDL was tested by classical biochemical methods and electron microscopy (EM) analysis of HDL density fractions. EM analysis showed that infection of the ABCA1 ^" ' ^" mice with the human apoCIII-expressing adenovirus did not result in either discoidal or spherical particles. In agreement with these findings, expression of human apoCIII in ABCA1 ^" ' ^" mice resulted in the exclusive accumulation of apoCIII on VLDL, while only trace amounts of apoCIII were detectable in the other lipoprotein fractions. It is possible that the presence of trace amounts of apoCIII in HDL or LDL may be an artifact of the fractionation process during density gradient ultracentrifugation. Thus, the findings establish that functional interactions between ABCA1 and apoCIII are essential for the formation of apoCIII- containing HDL-like particles, which are distinct from classical HDL containing apoE or apoA-l. Furthermore, these findings are in full agreement with the work of Fitzgerald and coworkers (53) who showed that in cultures of 293 cells transfected with wild-type ABCA1 cDNA, purified apoCIII interacted with ABCA1 , stimulated lipid efflux and inhibited cross-linking of wild-type apoA-l to ABCA1.

Increased accumulation of apoCIII on the VLDL of ABCA1 ^" ' ^" mice correlated with increased VLDL triglyceride and phospholipid content while cholesterol content remained unchanged, as compared to ABCA1 ^7" mice infected with the control adenovirus. The increase in VLDL-phospholipid content is consistent with increased VLDL size due to defective lipolysis and increased accumulation of triglycerides on VLDL.

In the above described studies, moderate levels of apoCIII expression that do not trigger hypertriglyceridemia in WT C57BL/6 or apoE ^'7' x apoA-l ^" ' ^" mice, resulted in significant hypertriglyceridemia in ABCAT' ^" mice. The increased sensitivity of ABCA1 ^" ' ^" mice towards apoCIII-induced hypertriglyceridemia correlated with accumulation of excess human apoCIII on the VLDL of these mice. A hepatic VLDL-TG secretion assay did not reveal any significant differences in the rates of hepatic triglyceride secretion between ABCA T' ^' and apoE ^" ' ^" x apoA-l ^" ' ^" mice infected with 8x10 ⁸ pfu of AdGFP-CIII ₉ .

However, co-infection of ABCA1 ^" ' mice with the human apoCIII-expressing adenovirus and an adenovirus expressing LpL, ameliorated the apoCIII-induced hypertriglyceridemia. These findings suggest that in the absence of ABCA1 , inhibition of LpL activity by excess accumulation of apoCIII on the VLDL of ABCA T' ^~ mice is responsible for the development of hypertriglyceridemia and the diminished conversion of VLDL into LDL in these mice.

Expression of apoCIII in apoE ^" ' ^" x apoA-l ^" ' ^" mice was also accompanied by an increase in the free cholesterol content of mainly their HDL and LDL lipoproteins. This finding is consistent with previous reports that have identified apoCIII as an inhibitor of LCAT activity in in vitro LCAT assays (51). Under conditions of increased human apoCIII expression in the apoE ^" ' ^" x apoA-l ^" ' ^" mice, LCAT activity is inhibited resulting in slower esterification of the free cholesterol of discoidal HDL, and a reduced rate of conversion of discoidal HDL particles into spherical. Similarly, in a separate experiment expression of human apoCIII in wild-type C57BL/6 mice, which normally contain classical spherical HDL, resulted in a mixture of discoidal and spherical HDL particles (data not shown), further confirming the in vitro data on the role of apoCIII as an inhibitor of LCAT(51).

It has been suggested that increased plasma apoCIII levels lead to an accumulation of apoB-and triglyceride-rich particles (LpB:apoCIII) in the circulation, and results in a decrease in the relative HDL to VLDL apoCIII ratio (54). As indicated above, it is now postulated that, under conditions of apoCIII-overexpression, the ABCA1 -mediated pathway of apoCI M-HDL formation becomes saturated and the remaining excess lipid- poor apoCIII becomes available for binding to VLDL, leading to hypertriglyceridemia. Conditions that promote the accumulation of apoCIII on HDL ₁ such as increased ABCA1 activity, may lead to the preferential accumulation of apoCIII on HDL, reduction in the apoCIII content of VLDL, and prevention of apoCIII-induced hypertriglyceridemia.

As noted above, previous studies also showed that under conditions of overexpression, apoE (46; 55), apoA-l (56), and apoA-ll (57) accumulate on TG-rich VLDL and inhibit LpL activity, thus leading to hypertriglyceridemia. It is thus now extrapolated that, as with apoCIII, ABCA1 -deficiency may also increase the sensitivity of mice towards apoE-, or apoA-l-, or apoA-ll-induced hypertriglyceridemia by preventing the formation of apoE- or apoA-l-, or apoA-ll-containing HDL and promoting the accumulation of excess of these apolipoproteins on triglyceride-rich VLDL. Applying the data herein to development of the hypertriglyceridemia of Tangier's disease further supports the suggestion that, in addition to its other established properties, HDL acts as a buffer that prevents accumulation of excess plasma apolipoproteins (such as apoCIII and possibly other apolipoproteins) on VLDL. In the absence of ABCA1 , HDL buffering capacity for apolipoproteins is thought to be eliminated, resulting in the abnormal apolipoprotein composition of VLDL and the hypertriglyceridemia which are recognised characteristics of Tangier's disease.

In summary, the above noted findings show that in the presence of ABCA1 , apoCIII expression promotes the formation of apoCIII-containing HDL, and limits the amount of lipid poor apoCIII that is available for association with VLDL (Fig. 9, +ABCA1 branch).

This results in normal lipolysis of VLDL triglycerides in plasma, and the efficient removal of lipoprotein remnants from the circulation. In contrast, deficiency in ABCA1 , prevents the de novo biogenesis of apoCIII-containing HDL, and promotes accumulation of the vast majority of lipid-poor plasma apoCIII on VLDL (Fig. 9, -ABCA1 branch). This results in reduced LpL-mediated lipolysis of VLDL-triglycerides and the development of hypertriglyceridemia. It is extrapolated that, in a similar fashion,

ABCA1 -deficiency also increases the sensitivity of mice towards apoE- or apoA-l-, or apoA-ll-induced hypertriglyceridemia by promoting the accumulation of excess plasma apoE or apoA-l, or apoA-ll on VLDL.

Overall, the studies now presented identify apoCIII-containing HDL and the lipid transporter ABCA1 as important contributors towards prevention of apoCIII-induced hypertriglyceridemia. Since plasma apoCIII levels correlate with increased body mass index and the development of insulin resistance (35), it is possible that apoCIII- containing HDL and the lipid transporter ABCA 1 may also play an important role in the prevention of insulin resistance and type Il diabetes.

REFERENCES

1. Shulman, R. S., Herbert, P. N., Fredrickson, D. S., Wehrly, K., and Brewer, H. B., Jr. (1974) Isolation and alignment of the tryptic peptides of alanine apolipoprotein, an apolipoprotein from human plasma very low density lipoproteins, J. Biol. Chem. 249, 4969-4974.

2. McKeone, B. J., Massey, J. B., Knapp, R. D., and Pownall, H. J. (1988) Apolipoproteins C-I, C-Il, and C-III: kinetics of association with model membranes and intermembrane transfer, Biochemistry 27, 4500-4505. 3. Herbert, P. N., Assmann, G., Grotton, J. A. M., and Fredrickson, D. S. (1999) in

Metabolic basis of inherited disease (Stanbury, J. B., Wyngaarden, J. B., and Fredrickson, D. S., Eds.) pp 589-651 , McGraw-Hill, New York.

4. Onat, A., Hergenc, G., Sansoy, V., Fobker, M., Ceyhan, K., Toprak, S., and Assmann, G. (2003) Apolipoprotein C-III, a strong discriminant of coronary risk in men and a determinant of the metabolic syndrome in both genders,

Atherosclerosis 168, 81-89.

5. Herbert, P. N., Assmann, G., Gotto, A. M., Jr., and Fredrickson, D. S. (1982) in The Metabolic Basis of Inherited Disease (Stanbury, J. B., Wyngaarden, J. B., Fredrickson, D. S., Goldstein, J. L., and Brown, M. S., Eds.) pp 589-651 , McGraw-Hill, New York.

6. Zannis, V. I. and Breslow, J. L. (1980) in Electrophoresis (Radola, B. J., Ed.) pp 437-473, Walter de Gruyter, Berlin.

7. Alaupovic, P., Tavella, M., Bard, J. M., Wang, C. S., Attman, P. O., Koren, E., Corder, C, Knight-Gibson, C, and Downs, D. (1988) Lipoprotein particles in hypertriglyceridemic states, Adv. Exp. Med. Biol. 243, 289-297.

8. Fruchart, J. C. and Bard, J. M. (1991) Lipoprotein particle measurement: An alternative approach to classification of lipid disorders, Curr. Opin. Lipidol. 2, 362-366.

9. Laggner, P., Gotto, A. M., Jr., and Morrisett, J. D. (1979) Structure of the dimyristoylphosphatidylcholine vesicle and the complex formed by its interaction with apolipoprotein C-III: X-ray small-angle scattering studies, Biochemistry 18, 164-171.

10. Forte, T., Gong, E., and Nichols, A. V. (1974) Interaction by sonication of C- apolipoproteins with lipid: an electron microscopic study, Biochim. Biophys. Acta 337, 169-183.

11. Windier, E. E., Kovanen, P. T., Chao, Y. S., Brown, M. S., Havel, R. J., and Goldstein, J. L. (1980) The estradiol-stimulated lipoprotein receptor of rat liver. A binding site that membrane mediates the uptake of rat lipoproteins containing apoproteins B and E, J. Biol. Chem. 255, 10464-10471. 12. Sehayek, E. and Eisenberg, S. (1991) Mechanisms of inhibition by apolipoprotein C of apolipoprotein E-dependent cellular metabolism of human triglyceride-rich lipoproteins through the low density lipoprotein receptor pathway, J. Biol. Chem. 266, 18259-18267. 13. Aalto-Setala, K., Fisher, E. A., Chen, X., Chajek-Shaul, T., Hayek, T., Zechner, R., Walsh, A., Ramakrishnan, R., Ginsberg, H. N., and Breslow, J. L (1992) Mechanism of hypertriglyceridemia in human apolipoprotein (apo) CIII transgenic mice. Diminished very low density lipoprotein fractional catabolic rate associated with increased apo CIII and reduced apo E on the particles, J. CHn.

Invest 90, 1889-1900.

14. Aalto-Setala, K., Weinstock, P. H., Bisgaier, C. L., Wu, L., Smith, J. D., and Breslow, J. L. (1996) Further characterization of the metabolic properties of triglyceride-rich lipoproteins from human and mouse apoC-lll transgenic mice, J. Lipid Res. 37, 1802-1811.

15. de Silva, H. V., Lauer, S. J., Wang, J., Simonet, W. S., Weisgraber, K. H., Mahley, R. W., and Taylor, J. M. (1994) Overexpression of human apolipoprotein C-III in transgenic mice results in an accumulation of apolipoprotein B48 remnants that is corrected by excess apolipoprotein E, J. Biol. Chem. 269, 2324-2335.

16. Maeda, N., Li, H., Lee, D., Oliver, P., Quarfordt, S. H., and Osada, J. (1994) Targeted disruption of the apolipoprotein C-III gene in mice results in hypotriglyceridemia and protection from postprandial hypertriglyceridemia, J. Biol. Chem. 269, 23610-23616. 17. Ebara, T., Ramakrishnan, R., Steiner, G., and Shachter, N. S. (1997)

Chylomicronemia due to apolipoprotein CIII overexpression in apolipoprotein E- null mice. Apolipoprotein Clll-induced hypertriglyceridemia is not mediated by effects on apolipoprotein E ₁ J. CHn. Invest 99, 2672-2681.

18. Jong, M. C, Dahlmans, V. E., Hofker, M. H., and Havekes, L M. (1997) Nascent very-low-density lipoprotein triacylglycerol hydrolysis by lipoprotein lipase is inhibited by apolipoprotein E in a dose-dependent manner, Biochem. J. 328 ( Pt 3), 745-750.

19. Quarfordt, S. H., Michalopoulos, G., and Schirmer, B. (1982) The effect of human C apolipoproteins on the in vitro hepatic metabolism of triglyceride emulsions in the rat, J. Biol. Chem. 257, 14642-14647.

20. Oswald, B. and Quarfordt, S. (1987) Effect of apoE on triglyceride emulsion interaction with hepatocyte and hepatoma G2 cells, J. Lipid Res. 28, 798-809.

21. Windier, E., Chao, Y., and Havel, R. J. (1980) Determinants of hepatic uptake of triglyceride-rich lipoproteins and their remnants in the rat, J. Biol. Chem. 255, 5475-5480.

22. Agnani, G., Bard, J. M., Candelier, L., Delattre, S., Fruchart, J. C ₁ and Clavey, V. (1991) Interaction of LpB, LpB:E, LpB:C-lll, and LpB:C-lll:E lipoproteins with the low density lipoprotein receptor of HeLa cells, Arterioscler. Thromb. 11, 1021-1029. 23. Clavey, V., Lestavel-Delattre, S., Copin, C, Bard, J. M., and Fruchart, J. C.

(1995) Modulation of lipoprotein B binding to the LDL receptor by exogenous lipids and apolipoproteins Cl, CII, CIII, and E, Arterioscler. Thromb. Vase. Biol. 15, 963-971. 24. Mann, C. J., Troussard, A. A., Yen, F. T., Hannouche, N., Najib, J., Fruchart, J. C, Lotteau, V., Andre, P., and Bihain, B. E. (1997) Inhibitory effects of specific apolipoprotein C-III isoforms on the binding of triglyceride-rich lipoproteins to the lipolysis-stimulated receptor, J. Biol. Chem. 272, 31348-31354. 25. Windier, E., Chao, Y., and Havel, R. J. (1980) Regulation of the hepatic uptake of triglyceride-rich lipoproteins in the rat. Opposing effects of homologous apolipoprotein E and individual C apoproteins, J. Biol. Chem. 255, 8303-8307.

26. Windier, E. and Havel, R. J. (1985) Inhibitory effects of C apolipoproteins from rats and humans on the uptake of triglyceride-rich lipoproteins and their remnants by the perfused rat liver, J. Lipid Res. 26, 556-565.

27. Shelbume, F., Hanks, J., Meyers, W., and Quarfordt, S. (1980) Effect of apoproteins on hepatic uptake of triglyceride emulsions in the rat, J. Clin. Invest 65, 652-658.

28. Kortz, W. J., Schirmer, B. D., Mansbach, C. M., Shelburne, F., Toglia, M. R., and Quarfordt, S. H. (1984) Hepatic uptake of chylomicrons and triglyceride emulsions in rats fed diets of differing fat content, J. Lipid Res. 25, 799-804.

29. Brown, W. V. and Baginsky, M. L. (1972) Inhibition of lipoprotein lipase by an apoprotein of human very low density lipoprotein, Biochem. Biophys. Res. Commun. 46, 375-382. 30. Krauss, R. M., Herbert, P. N., Levy, R. I., and Fredrickson, D. S. (1973) Further observations on the activation and inhibition of lipoprotein lipase by apolipoproteins, Circ. Res. 33, 403-411.

31. Ginsberg, H. N., Le, N. A., Goldberg, I. J., Gibson, J. C ₁ Rubinstein, A., Wang- Iverson, P., Norum, R., and Brown, W. V. (1986) Apolipoprotein B metabolism in subjects with deficiency of apolipoproteins CIII and Al. Evidence that apolipoprotein CIII inhibits catabolism of triglyceride-rich lipoproteins by lipoprotein lipase in vivo, J. CHn. Invest 78, 1287-1295.

32. Wang, C. S., McConathy, W. J., Kloer, H. U., and Alaupovic, P. (1985) Modulation of lipoprotein lipase activity by apolipoproteins. Effect of apolipoprotein C-III, J. CHn. Invest 75, 384-390.

33. Lambert, D. A., Catapano, A. L., Smith, L. C, Sparrow, J. T., and Gotto, A. M., Jr. (1996) Effect of the apolipoprotein C-II/C-III1 ratio on the capacity of purified milk lipoprotein lipase to hydrolyse triglycerides in monolayer vesicles, Atherosclerosis 127, 205-212. 34. McConathy, W. J., Gesquiere, J. C ₁ Bass, H., Tartar, A., Fruchart, J. C, and

Wang, C. S. (1992) Inhibition of lipoprotein lipase activity by synthetic peptides of apolipoprotein C-III ₁ J. Lipid Res. 33, 995-1003.

35. Hernandez-Mijares, A., Jarabo-Bueno, M. M., Lopez-Ruiz, A., Sola-lzquierdo,

E., Morillas-Arino, C, and Martinez-Triguero, M. L. (2007) Levels of C3 in patients with severe, morbid and extreme obesity: its relationship to insulin resistance and different cardiovascular risk factors, Int. J. Obes. (Lond) 31, 927-

932. 36. Alaupovic, P., Mack, W. J., Knight-Gibson, C, and Hodis, H. N. (1997) The role of triglyceride-rich lipoprotein families in the progression of atherosclerotic lesions as determined by sequential coronary angiography from a controlled clinical trial, Arterioscler. Thromb. Vase. Biol. 17, 715-722. 37. Sacks, F. M., Alaupovic, P., Moye, L. A., Cole, T. G., Sussex, B., Stampfer, M.

J., Pfeffer, M. A., and Braunwald, E. (2000) VLDL, apolipoproteins B, CIII, and E, and risk of recurrent coronary events in the Cholesterol and Recurrent Events (CARE) trial, Circulation 102, 1886-1892.

38. Lee, S. J., Campos, H., Moye, L. A., and Sacks, F. M. (2003) LDL containing apolipoprotein CIII is an independent risk factor for coronary events in diabetic patients, Arterioscler. Thromb. Vase. Biol. 23, 853-858.

39. He ₁ T. C, Zhou, S., da Costa, L. T., Yu, J., Kinzler, K. W., and Vogelstein, B. (1998) A simplified system for generating recombinant adenoviruses, Proc. Natl. Acad. ScL U. S. A 95, 2509-2514. 40. Roghani, A. and Zannis, V. I. (1988) Mutagenesis of the glycosylation site of human ApoCIII. O-linked glycosylation is not required for ApoCIII secretion and lipid binding, J. Biol. Chem. 263, 17925-17932.

41. Fallaux, F. J., Kranenburg, O., Cramer, S. J., Houweling, A., Van Ormondt, H., Hoeben, R. C, and van der Eb, A. J. (1996) Characterization of 911 : a new helper cell line for the titration and propagation of early region 1 -deleted adenoviral vectors, Hum. Gene Then 7, 215-222.

42. Kypreos, K. E., Van Dijk, K. W., van Der, Z. A., Havekes, L. M., and Zannis, V. I. (2001) Domains of apolipoprotein E contributing to triglyceride and cholesterol homeostasis in vivo. Carboxyl-terminal region 203-299 promotes hepatic very low density lipoprotein-triglyceride secretion, J. Biol. Chem. 276, 19778-19786.

43. McNeish, J., Aiello, R. J., Guyot, D., Turi, T., Gabel, C, Aldinger, C, Hoppe, K. L., Roach, M. L., Royer, L. J., de Wet, J., Broccardo, C ₁ Chimini, G., and Francone, O. L. (2000) High density lipoprotein deficiency and foam cell accumulation in mice with targeted disruption of ATP-binding cassette transporter-1 , Proc. Natl. Acad. ScL U. S. A 97, 4245-4250.

44. Zhang, S. H., Reddick, R. L., Piedrahita, J. A., and Maeda, N. (1992) Spontaneous hypercholesterolemia and arterial lesions in mice lacking apolipoprotein E, Science 258, 468-471.

45. Williamson, R., Lee, D., Hagaman, J., and Maeda, N. (1992) Marked reduction of high density lipoprotein cholesterol in mice genetically modified to lack apolipoprotein A-I ₁ Proc. Natl. Acad. ScL U. S. A 89, 7134-7138.

46. Kypreos, K. E., Teusink, B., Van Dijk, K. W., Havekes, L. M., and Zannis, V. I. (2001) Analysis of the structure and function relationship of the human apolipoprotein E in vivo, using adenovirus-mediated gene transfer, FASEB J. 15, 1598-1600.

47. Kypreos, K. E., Morani, P., Van Dijk, K. W., Havekes, L. M., and Zannis, V. I. (2001) The amino-terminal 1-185 domain of apoE promotes the clearance of lipoprotein remnants in vivo. The carboxy-terminal domain is required for induction of hyperlipidemia in normal and apoE-deficient mice, Biochemistry 40, 6027-6035.

48. Kypreos, K. E., Li ₁ X., Van Dijk, K. W., Havekes, L. M., and Zannis, V. I. (2003) Molecular mechanisms of type III hyperlipoproteinemia: The contribution of the carboxy-terminal domain of ApoE can account for the dyslipidemia that is associated with the E2/E2 phenotype, Biochemistry 42, 9841-9853.

49. Kypreos, K. E., Van Dijk, K. W., Havekes, L M., and Zannis, V. I. (2005) Generation of a recombinant apolipoprotein E variant with improved biological functions: hydrophobic residues (LEU-261 , TRP-264, PHE-265, LEU-268, VAL- 269) of apoE can account for the apoE-induced hypertriglyceridemia, J. Biol.

Chem. 280, 6276-6284.

50. Kypreos, K. E. and Zannis, V. I. (2005) Novel pathway of biogenesis of apoE- containing HDL with the participation of ABCA1 and LCAT: Implications for dyslipidemias and atherogenesis, Circulation 16693. 51. Nishida, H. I., Nakanishi, T., Yen, E. A., Arai, H., Yen, F. T., and Nishida, T.

(1986) Nature of the enhancement of lecithin-cholesterol acyltransferase reaction by various apolipoproteins, J. Biol. Chem. 261, 12028-12035.

52. Zannis, V. I., Chroni, A., Kypreos, K. E., Kan, H. Y., Cesar, T. B., Zanni, E. E., and Kardassis, D. (2004) Probing the pathways of chylomicron and HDL metabolism using adenovirus-mediated gene transfer, Curr Opin Lipidol. 15,

151-166.

53. Fitzgerald, M. L., Morris, A. L, Chroni, A., Mendez, A. J., Zannis, V. I., and Freeman, M. W. (2004) ABCA1 and amphipathic apolipoproteins form high- affinity molecular complexes required for cholesterol efflux, J. Lipid Res. 45, 287-294.

54. Chivot, L., Mainard, F., Bigot, E., Bard, J. M., Auget, J. L, Madec, Y., and Fruchart, J. C. (1990) Logistic discriminant analysis of lipids and apolipoproteins in a population of coronary bypass patients and the significance of apolipoproteins C-III and E, Atherosclerosis 82, 205-211. 55. Havel, R. J., Kotite, L, Vigne, J. L, Kane, J. P., Tun, P., Phillips, N., and Chen,

G. C. (1980) Radioimmunoassay of human arginine-rich apolipoprotein, apoprotein E. Concentration in blood plasma and lipoproteins as affected by apoprotein E-3 deficiency, J. Clin. Invest 66, 1351-1362.

56. Chroni, A., Kan, H. Y., Kypreos, K. E., Gorshkova, I. N., Shkodrani, A., and Zannis, V. I. (2004) Substitutions of glutamate 110 and 111 in the middle helix 4 of human apolipoprotein A-I (apoA-l) by alanine affect the structure and in vitro functions of apoA-l and induce severe hypertriglyceridemia in apoA-l-deficient mice, Biochemistry 43, 10442-10457.

57. Boisfer, E., Lambert, G., Atger, V., Tran, N. Q., Pastier, D., Benetollo, C, Trottier, J. F., Beaucamps, I., Antonucci, M., Laplaud, M., Griglio, S., Chambaz,

J., and Kalopissis, A. D. (1999) Overexpression of human apolipoprotein A-Il in mice induces hypertriglyceridemia due to defective very low density lipoprotein hydrolysis, J. Biol. Chem. 274, 11564-1 1572. 58. Brooks-Wilson, A., Marcil, M., Clee, S. M., Zhang, L. H., Roomp, K., van Dam, M., Yu, L., Brewer, C, Collins, J. A., Molhuizen, H. O., Loubser, O., Ouelette, B. F., Fichter, K., Ashboume-Excoffon, K. J., Sensen, C. W., Scherer, S., Mott, S., Denis, M., Martindale, D., Frohlich, J., Morgan, K., Koop, B., Pimstone, S., Kastelein, J. J., and Hayden, M. R. (1999) Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency, Nat. Genet. 22, 336- 345.

59. Wang, C. S., Alaupovic, P., Gregg, R. E., and Brewer, H. B., Jr. (1987) Studies on the mechanism of hypertriglyceridemia in Tangier disease. Determination of plasma lipolytic activities, k1 values and apolipoprotein composition of the major lipoprotein density classes, Biochim. Biophys. Acta 920, 9-19.

60. Kolovou, G., Daskalova, D., Anagnostopoulou, K., Hoursalas, I., Voudris, V., Mikhailidis, D. P., and Cokkinos, D. V. (2003) Postprandial hypertriglyceridaemia in patients with Tangier disease, J. Clin. Pathol. 56, 937- 941.

61. Heinen, R. J., Herbert, P. N., and Fredrickson, D. S. (1978) Properties of the plasma very low and low density lipoproteins in Tangier disease, J. CHn. Invest 61, 120-132.

62. Federica Basso, Lita Freeman, Catherine L. Knapper, Alan Remaley, John Stonik, Edward B. Neufeld, Terese Tansey, Marcelo J. A. Amar, Jamila

Fruchart-Najib, Nicholas Duverger, Silvia Santamarina-Fojo and H. Bryan Brewer, Jr. (2003) Role of the hepatic ABCA1 transporter in modulating intrahepatic cholesterol and plasma HDL cholesterol concentrations. J Lipid Res. 44, 296-302.