Previous work had demonstrated that P-407 treatment significantly decreased
cholesterol efflux from macrophages by down-regulating ABCA1 gene expression and that P-407 did
not abrogate the action of a known LXRα agonist to stimulate apoA1-mediated cholesterol efflux from macrophages (Johnston et al 2006
). In the present study, we investigated whether P-407 modulated
PPARα activity with the goal of extending our work to the PPAR-LXR-ABCA1 signaling pathway and to further elucidate the mechanism(s) responsible for reduced cellular cholesterol efflux along this pathway, as well as the dyslipidemia observed in this murine model of atherogenesis (Johnston 2004
Using a transactivation assay, we determined that P-407 was unable to activate the transcriptional activity of either mouse or human PPARα. Furthermore, P-407 did not inhibit a known PPARα agonist from activating human PPARα. To ascertain whether PPARα played a role in P-407-mediated dyslipidemia, we compared the effects of P-407 treatment in wild-type (B6) and PPARα-deficient mice. It was critical to evaluate P-407 for its potential to mediate dyslipidemia in PPARα-deficient mice, since a compound’s ability to modulate PPARα is not always predicted from the results of an in vitro
transactivation assay (Peters et al 1996
). After taking into consideration the increase in plasma HDL-cholesterol and, consequently, total cholesterol, seen in PPARα-deficient mice, there were no significant differences in P-407-induced plasma concentration-time profiles of total cholesterol, HDL-cholesterol, non-HDL-cholesterol (calculated), and TG. Therefore, these data strongly support the conclusion that neither P-407, nor some intermediate which results from the administration of P-407, is functioning as a PPARα agonist in vivo
, similar to the action of fibrate drugs described below.
Assessment of anti-atherogenic activity of PPARα agonists in rodent models of atherosclerosis is hampered by the fact that a) rodents develop a proinflammatory peroxisome proliferative response in the liver, and b) classical animal models, such as the LDL-receptor or apoE-deficient mice, display an aberrant hyperlipidemic response to the fibrate class of hypolipidemic drugs (Marx et al 2004
). As an example of the latter, several studies have evaluated the use of fibrates (PPARα agonists) in several rodent models. Duez et al.
reported that even though administration of fenofibrate to apoE-/-
mice at a dose of 100 mg/kg/d for 8 weeks resulted in a significant increase in both plasma total cholesterol and triglycerides, there was a reduction in atherosclerosis (Duez et al 2002
). Similarly, Fu et al.
(Fu et al 2003
) reported that ciprofibrate markedly increased the plasma levels of atherogenic lipoproteins in apoE-/-
mice; however, in contrast to Duez et al.
, this investigator reported a significant enhancement in the development of atherosclerosis. Clearly, the ability of fibrate drugs to increase plasma lipids in mice is well established, although their capacity to inhibit atherosclerosis is still controversial. Notably, the effects of fibrates in rodents are just opposite to that observed clinically. Activation of PPARα with fibrate drugs in humans increases fatty acid oxidation, decreases very-low-density lipoprotein (VLDL), increases lipolysis, and increases HDL-cholesterol and reverse cholesterol transport. These actions result in a decrease in plasma TG, a decrease in plasma small, dense, atherogenic LDL-cholesterol, and an increase in HDL-cholesterol (Marx et al 2004
Finally, if PPARα had been involved in the dyslipidemic response induced by administration of P-407, then plasma lipid concentration-time profiles for mice in Group 3 (PPARα-knockouts administered P-407), after correcting for increases in HDL, and, in turn, total cholesterol (due to the gene knockout), would not have been similar to corresponding profiles determined for mice in Group 1 (B6 mice administered P-407). Furthermore, as described above, P-407, or some intermediate formed following its administration, is not functioning as a PPARα agonist, similar to the fibrate drugs. If P-407 had functioned as a PPARα agonist, then its administration to B6 mice would have been expected to increase plasma lipid concentrations similar to the studies cited above for fibrates; however, P-407 would have no capacity to increase plasma lipids to the same extent in PPARα-deficient mice (Group 3 mice). Therefore, we conclude that P-407 does not modulate PPARα activity (e.g.
, by functioning as either a PPARα agonist or antagonist) following its administration to B6 mice. Additionally, our data support the conclusion that the dyslipidemic response to P-407 in mice is not mediated by its effects at the level of PPARα, which strongly corroborates our findings using the transactivation assay. Instead, acute P-407 administration to mice (i.e.
, a single injection) causes hypertriglyceridemia, in part, by inhibiting the activity of capillary-bound lipoprotein lipase (Johnston & Palmer 1993) and hypercholesterolemia, in part, by inhibiting the activity of cholesterol 7α-hydroxylase (the rate-limiting enzyme for elimination of cholesterol into bile) (Johnston et al 2001), while simultaneously increasing both the activity and protein expression of 3-hydroxy-3-methylglutaryl-CoA reductase and down-regulating low-density lipoprotein receptor expression (Leon et al 2006
; Johnston & Palmer 1997
ApoA1, the major cholesterol acceptor of HDL, is induced by fibrates in humans, but is decreased by fibrates in rodents (Duez et al 2005
). This is because HDL metabolism is regulated in an opposite fashion by fibrates in rodents (decrease) and man (increase) due to sequence divergences in the respective apoA1 promoters (Ngoc et al 1998
). Consequently, in rats, fibrates, such as fenofibrate and clofibrate, significantly decrease plasma apoA1, apoA2, and apoE concentrations (Staels et al 1992
), whereas, in mice, fibrates cause a decrease in the plasma concentration of apoA1, but an increase in apoA2 (Duez et al 2005
). In the present study, P-407 caused a reduction in the plasma concentration of apoA1 in treated mice. Since it is primarily apoA1 that confers the atheroprotective effects of HDL, this may help explain why mice develop aortic atherosclerotic lesions after 16 weeks of P-407 administration (Johnston 2004
; Palmer et al 1998
). Simultaneously, there was a slight increase in the plasma concentration of apoE. Interestingly, we previously demonstrated that HDL, isolated from the plasma of P-407-treated mice, contained more apoE, relative to apoA1 (Johnston et al 1999
The decrease in plasma apoA1 following P-407 administration is not likely to be mediated by PPARα, since a) P-407 was unable to directly modulate the activity of mouse PPARα using a transactivation assay, and b) the atherogenic dyslipidemia that occurs following administration of P-407 to mice was not mediated through PPARα. Therefore, as it relates to apolipoproteins, even though P-407 treatment caused a reduction in the plasma concentration of apoA1, similar to the fibrates, we conclude that P-407 did not function as a PPARα agonist. Instead, we suggest that P-407, by significantly reducing ABCA1 gene expression, contributes, in part, to induction of an atherogenic plasma lipid profile. Indeed, Joyce et al.
(Joyce et al 2002
) demonstrated that overexpression of ABCA1 in C57BL/6 mice resulted in a unique anti-atherogenic profile characterized by decreased plasma cholesterol, cholesteryl esters, free cholesterol, non-HDL cholesterol, and apoB, but markedly increased HDL, apoA1, and apoE. In contrast, administration of P-407 to C57BL/6 mice results in an atherogenic profile characterized by increased plasma total cholesterol, triglycerides, and non-HDL cholesterol, but markedly decreased HDL-cholesterol and apoA1. The fact that P-407 induced this same type of atherogenic profile in PPARα-deficient mice provides further evidence that the observed reduction in plasma apoA1 levels, which resulted from the administration of P-407, was not mediated through PPARα, and that P-407 was not functioning as a PPARα agonist.