Mechanisms of Lipid Elevations Associated With the Treatment of Patients With HIV Infection

Peter J. Piliero, MD

In This Article

Potential Mechanisms of PI-Related Dyslipidemia

Whether plasma lipid alterations are a direct consequence of PI therapy or reflect an HIV-associated metabolic defect that is exacerbated by the drug remains unclear. Although lipodystrophy and dyslipidemia appear to be related metabolically, the pathophysiologic nature of the association is unknown, as they are not always present simultaneously. A variety of potential mechanistic metabolic alterations have been described, with alterations in retinoic acid metabolism being the most prominent ( Table 2 ). Nonetheless, it is likely that the adverse consequences of PI therapy are due to the drug's effect on multiple metabolic pathways.

Retinoic acid plays a central role in gene expression. Two families of nuclear retinoid receptors have been described: retinoic acid receptors (RARs) and retinoid X receptors (RXRs).[32] Both families have alpha, beta, and gamma isoforms. Following activation by retinoic acid in the cytosol, RARs form heterodimers with RXRs, while activated RXRs can also form heterodimers with a number of other nuclear receptors (eg, peroxisome proliferator-activated receptors [PPARs], glucocorticoid receptor, thyroid hormone receptor). The heterodimers then translocate to the nucleus, where they bind to retinoic acid response elements (RAREs) located in the regulatory regions of target genes. By modulating gene transcription, the heterodimers influence cellular activity both directly and indirectly.

Cytoplasmic retinoic acid-binding protein type 1. Cytoplasmic retinoic acid-binding protein type 1 (CRABP-1) is a ubiquitous protein.[33] It binds virtually all intracellular retinoic acid and presents it to cytochrome P450 (CYP) 3A isoforms for conversion to cis-9-retinoic acid, the sole ligand of the RXRs. In adipocytes, binding of cis-9-retinoic acid to an RXR activates the receptor, which then forms heterodimers with the gamma subtype of PPAR. Binding of the heterodimers to adipocyte RAREs subsequently inhibits apoptosis and stimulates differentiation and proliferation of fat cells (Figure 1).

The role of CRABP-1 in adipocyte differentiation. CIS-9-RA, cis-9-retinoic acid; CYP 3A, cytochrome P-450 3A isoform; CRABP-1, cytoplasmic retinoic acid-binding protein type 1; PPAR-gamma, gamma subtype of peroxisome proliferator-activated receptor; RA, retinoic acid; RARE, retinoic acid response element; RXR, retinoid X receptor

PIs have a high affinity for the catalytic site of HIV-1 protease.[33] The 12-amino acid sequence in this HIV-1 catalytic site has a 58% homology with a C-terminal region of CRABP-1. Binding of a PI to this or other retinoic acid-binding residues of CRABP-1 can inhibit binding of retinoic acid to the transport protein. It has been proposed that this inhibition initiates a cascade of events in fat cells, including decreased conversion of retinoids to cis-9-retinoic acid, reduced RXR activation, decreased formation of RXR/PPAR-gamma heterodimers, aberrant activation of target genes, reduced adipocyte differentiation and proliferation, and increased lipocyte apoptosis. These events, by reducing TG storage and increasing lipid release into the circulation, produce hyperlipidemia.[21]

CYP3A isoform. HIV PIs are potent inhibitors of CYP3A, the enzyme responsible for conversion of retinoic acid to cis-9-retinoic acid.[21] PI-induced inhibition of CYP3A decreases the formation of cis-9-retinoic acid. Consequently, with less available ligand, there is decreased formation of RXR and RXR/PPAR-gamma heterodimers. This results in the effects described above. The importance of this mechanism as a factor in dyslipidemia is supported by studies of ritonavir, the most potent CYP3A inhibitor of the current PIs. Ritonavir administered in a dose-dependent fashion is the most likely PI to cause significant dyslipidemia in patients treated with current PI-containing regimens.[25]

Several additional mechanisms potentially contribute to PI-associated dyslipidemia. The common element in these proposed mechanisms is impaired clearance of lipids from the bloodstream.

LDL receptor-related protein. LDL receptor-related protein (LRP) is a hepatic cell surface receptor that plays an important role in the clearance of chylomicrons.[34] It is also coexpressed with lipoprotein lipase (LPL) on capillary endothelial cells.[35] The LRP/LPL complex releases free fatty acids from circulating TGs for later storage in adipocytes. The HIV protease catalytic site shares a 63% homology with a region of the lipid-binding domain of LRP.[33] Binding of PIs to the lipid-binding domain of LRP would, therefore, be expected to inhibit the breakdown of circulating chylomicrons and triglycerides.

LDL receptor. Binding of LDL to the LDL receptor (LDL-R) is the primary mechanism for clearing the blood of LDLs. As measured by flow cytometry, LDL-R expression on peripheral blood monocytes has been found to be significantly lower in HAART-treated patients than in controls (P = .009).[36] Whereas no differences in LDL-R expression have been observed between PI-treated and PI-naive patients, the expression has been found to be significantly lower in patients with lipodystrophy (P = .002).[36] This suggests that downregulation of the LDL-R may contribute to the lipid abnormalities of HIV disease.

Interferon-alpha. Interferon-alpha therapy in patients with chronic hepatitis C is associated with significant lipid abnormalities. TG levels increase significantly during therapy and return to baseline after treatment is stopped.[37] This dyslipidemia appears to be the result of an interferon-alpha-induced downregulation of the activities of LPL, hepatic TG lipase, and cholesterol ester transfer protein.[38] The increase in endogenous interferon-alpha in patients with AIDS is strongly correlated with TG clearance time.[39] This suggests that in PI-treated patients with advanced disease, cytokines may contribute to the etiology of dyslipidemia.

Scavenger receptor class B, type I. Scavenger receptor class B, type I (SR-BI), mediates the transfer of cholesterol from HDL to cells. This SR-BI-mediated selective cholesterol uptake can promote the formation of lipid-laden macrophages or foam cells. Both saquinavir and amprenavir increase the expression of SR-BI mRNA, selective cholesterol uptake, and the amount of cell-associated HDL.[39] This has the potential to interfere with HDL-mediated cholesterol transfer and promote atherogenesis.

Alterations in the number and size of fat cells influence adipose tissue mass and distribution, with resultant changes in lipid and carbohydrate metabolism.[40] Consequently, PIs that block adipogenesis and increase lipolysis might also contribute to dyslipidemia. Incubation of stem cells with nelfinavir, saquinavir, and ritonavir has been shown to increase acute lipolysis and to decrease TG accumulation, lipogenesis, and expression of adipose markers aP2 and LPL.[41] Stem cell preparations stained for neutral fat with oil red-O show decreased cytoplasmic fat. This effect appears to be PI dependent, although it has not been seen with amprenavir and indinavir and is apparently not due to an RXR-alpha- or PPAR-gamma-related mechanism. Thus, reduced adipogenesis and increased lipolysis have the potential to contribute to PI-associated dyslipidemia and metabolic abnormalities.

PIs stimulate hepatic TG synthesis.[42] For example, in fed mice, neither nelfinavir nor ritonavir affected serum cholesterol, but both increased TG and fatty acids, by 57% to 108%. The effects of the 2 drugs appear to be differentially regulated, since the activity of nelfinavir disappears in fasting animals. The mechanism of this effect is not clear, but may be PI-induced alterations in the level of sterol regulatory element-binding proteins (SREBPs). These proteins function as transcriptional factors that play a central role in adipogenesis, fatty acid biosynthesis, and the metabolic effects of insulin.[43] Three different SREBPs have been described. SREBP-1 has a relatively greater effect on genes involved in fatty acid biosynthesis than do the other sterol regulatory elements, whereas the effects of SREBP-2 favor cholesterol biosynthesis.[44] In a murine model, ritonavir was shown to produce a 2-fold increase in plasma TG levels and secretion of very-low-density lipoprotein (VLDL).[45] This hyperlipidemia appears to be secondary to a ritonavir-mediated interference with proteasomal degradation of activated nuclear SREBP-1. Increased SREBP-1 activity then differentially upregulates TG biosynthesis over that of cholesterol.

PI-induced dyslipidemia may also have a genetic component. Apo C-III is a major component of VLDL and modulates the rate of clearance of TG-rich lipoprotein remnant particles. In a study of apo C-III polymorphisms, differing interactions between the polymorphic variants and PIs appeared to contribute to dyslipidemia.[46] In this study, plasma TG levels increased according to the number of polymorphic variant apo C-III alleles that were present. These data indicate that some patients may have a genetic predisposition to PI-induced dyslipidemia.