The raft-promoting property of virion-associated cholesterol, but not the presence of virion-associated Brij 98 rafts, is a determinant of human immunodeficiency virus type 1 infectivity

Abstract

Lipid rafts are enriched in cholesterol and sphingomyelin and are isolated on the basis of insolubility in detergents, such as Brij 98 and Triton X-100. Recent work by Holm et al. has shown that rafts insoluble in Brig 98 can be found in human immunodeficiency virus type 1 (HIV-1) virus-like particles, although it is not known whether raft-like structures are present in authentic HIV-1 and it is unclear whether a virion-associated raft-like structure is required for HIV replication. Independently, it was previously reported that virion-associated cholesterol is critical for HIV-1 infectivity, although the specific requirement of virion cholesterol in HIV-1 was not examined. In the present study, we have demonstrated that infectious wild-type HIV-1 contains Brij 98 rafts but only minimal amounts of Triton X-100 rafts. To directly assess the functional requirement of virion-associated rafts and various features of cholesterol on HIV-1 replication, we replaced virion cholesterol with exogenous cholesterol analogues that have demonstrated either raft-promoting or -inhibiting capacity in model membranes. We observed that variable concentrations of exogenous analogues are required to replace a defined amount of virion-associated cholesterol, showing that structurally diverse cholesterol analogues have various affinities toward HIV-1. We found that replacement of 50% of virion cholesterol with these exogenous cholesterol analogues did not eliminate the presence of Brij 98 rafts in HIV-1. However, the infectivity levels of the lipid-modified HIV-1s directly correlate with the raft-promoting capacities of these cholesterol analogues. Our data provide the first direct assessment of virion-associated Brij 98 rafts in retroviral replication and illustrate the importance of the raft-promoting property of virion-associated cholesterol in HIV-1 replication.

FIG. 1.

WT HIV-1 particles were analyzed for lipid raft-like domains. WT HIV-1 particles purified on OptiPrep gradients were treated with 0.5% Brij 98 or Triton detergent and subjected to sucrose flotation gradient analysis (see Materials and Methods). Fractions 1 to 10 (lanes 1 to 10) were taken from the top of the gradient. 293T cells were used as the positive control for successful Triton raft isolation, with caveolin-containing raft membranes floating to fractions 3 and 4 (n = 5). Viral proteins were precipitated from each fraction and analyzed by SDS-PAGE and immunoblotting for raft-resident caveolin protein and p24 (in the Brij 98 raft gradient) and for viral gp41 envelope proteins (in both the Brij 98 and Triton raft gradients). Data are representative of the results of two experiments.

FIG. 2.

Structures of sterols used in the present report. Variation in the 3β-hydroxyl group (A) or A,B ring structure (B) of cholesterol is shown. Conventional numerical labeling of the carbon atoms and alphabetical labeling of the four rings are indicated for cholesterol.

FIG. 3.

Density gradient profiles of lipid-modified HIV-1 where exogenous cholesterol or analogues with modification in the 3β-hydroxyl group (A to C) or with alterations in the A,B ring structure (D to F) were used. WT HIV-1 was mock-treated, CD-treated (0.5 mM), or CD-treated and incubated with exogenous cholesterol, epicholesterol, or 4-cholestenone. RT activity was detected in density gradients (A to C) of untreated (□), CD-treated (○), and CD-treated HIV-1 that had subsequently been replenished with a low concentration (▵) or a 100 μM concentration (▿) of exogenous sterol. CD-treated HIV-1 was replenished with low concentrations of exogenous sterols as follows: 25 μM cholesterol (A), 25 μM epicholesterol (B), 50 μM 4-cholestenone (C), 25 μM 7-dehydrocholesterol (D), 12 μM dihydrocholesterol (E), and 12 μM coprostanol (F). The duplicate percentage of total RT activity data points varied by <3% (the SE). Data are representative of three experiments. Corresponding densities (broken line) of each fraction (A) were determined by refractivity index measurements, where duplicate measurements varied by 0.001 g/ml (the SE).

FIG. 4.

Brij 98 raft analysis of CD-treated HIV-1 replenished with raft-promoting and raft-disrupting sterol analogues. Concentrated HIV-1 particles were treated with CD and replenished with exogenous sterol analogues in the following concentrations: 25 μM 7-dehydrocholesterol, 12 μM dihydrocholesterol, 50 μM 4-cholestenone, and 12 μM coprostanol. Modified HIV-1 particles were purified by OptiPrep gradients, treated with 0.5% Brij 98 detergent, and subjected to sucrose flotation gradient analysis (see Materials and Methods). Fractions 1 to 10 (lanes 1 to 10) were taken from the top of the gradient. CD-treated HIV-1 was included as a control for Brij 98 detergent resistance. Viral proteins were precipitated from each fraction and analyzed by SDS-PAGE and immunoblotting for viral gp41 envelope proteins (n = 2). Quantification of protein bands was performed by using Image Gauge version 3.3 software and is represented in fractions 2 to 4 (the floating membrane region) as a percentage of the total protein loaded on the gradient (n = 2; SE, 6%).

FIG. 5.

Viral infectivity of lipid-modified HIV-1. HIV-1 containing the luciferase reporter gene was mock treated, CD treated (0.5 mM), or CD treated and incubated with exogenous cholesterol (A), 7-dehydrocholesterol (B), epicholesterol (C), dihydrocholesterol (D), 4-cholestenone (E), or coprostanol (F). Lipid envelope-modified and untreated HIV-1 were normalized for p24 capsid protein content and used to infect MT2 cells. The levels of luciferase activity were used as the indicator of infectivity levels, which are reported as percentages (means ± SE of duplicate samples) of the untreated control. Infectivity levels of untreated (a), CD-treated (b), and CD-treated HIV-1 that has subsequently been incubated with a 12 (c), 25 (d), 50 (e), or 100 (f) μM concentration of exogenous sterol are shown. Filled bars represent replenished virions that exhibited WT particle density profiles (same as the low replenishment described in legends for Fig. 2 and 3). P values were 0.01, 0.2, 0.1, <0.01, and 0.01, respectively, for filled bars in panels B, C, D, E, and F; n = 4. Data are representative of three experiments. *, epicholesterol-replenished particles of WT density exhibited an infectivity range of 45 to 80%.

FIG. 6.

Density gradient profiles and infectivity levels of HIV-1 supplemented with sterol analogues in the absence of CD. WT HIV-1 was mock-treated or incubated with exogenous sterol analogues in the absence of CD (see Fig. 2). RT activity detected in density gradients (A) of untreated HIV-1 (□) or HIV-1 incubated in 100 μM concentrations of epicholesterol (○), 4-cholestenone (▵), coprostanol (▿), dihydrocholesterol (◊), or 7-dehydrocholesterol (+). Duplicate percentages for total RT activity data points varied by <3% (the SE). HIV-1 containing the luciferase reporter gene was mock treated (B, bar a) or incubated with 100 μM concentrations of epicholesterol (bar b), 4-cholestenone (bar c), coprostanol (bar d), dihydrocholesterol (bar e), or 7-dehydrocholesterol (bar f). All samples were normalized for p24 capsid protein content and used to infect MT2 cells. The levels of luciferase activity were used as indicators of infectivity levels, which are reported as percentages (means ± SE of duplicate samples) of the untreated control. Data are representative of two to three experiments.

FIG. 7.

Immunoblot analysis of HIV-1 gp120 envelope proteins and p24 CA proteins present in lipid-modified HIV-1. Concentrated HIV-1 particles were treated with CD and replenished with exogenous sterol analogues. Microvesicle contamination was removed by purification on OptiPrep gradients as described in Materials and Methods. Sample input was standardized by p24 CA content. Lane 1, untreated HIV-1 control; lane 2, CD-treated HIV-1; lane 3, CD-treated HIV-1 replenished with cholesterol; lane 4, CD-treated HIV-1 replenished with epicholesterol; lane 5, CD-treated HIV-1 replenished with 4-cholestenone; lane 6, CD-treated HIV-1 replenished with 7-dehydrocholesterol; lane 7, CD-treated HIV-1 replenished with dihydrocholesterol; lane 8, CD-treated HIV-1 replenished with coprostanol (n = 2).