Reverse cholesterol transport out of macrophage foam cells via ABCA1/ABCG1 to HDL and apoA-I — measuring HDL function as a cardiovascular risk biomarker
The cholesterol efflux assay begins not with cells but with plasma chemistry: separating the HDL/apoA-I acceptor pool from the LDL and VLDL particles that would otherwise confound the readout. Because LDL and VLDL are not physiological cholesterol acceptors and can themselves donate cholesterol to cells, they must be removed before the acceptor fraction is added to labeled macrophages. This apoB-depletion step, standardized by Rader, Rothblat and de la Llera-Moya in the 2000s, converts a routine blood draw into a functional reagent that captures the entire spectrum of extracellular cholesterol acceptors circulating in that individual.
Polyethylene glycol (PEG-8000) is a volume-exclusion precipitant: at 20% w/v it dehydrates and aggregates large, apoB-100-containing lipoprotein particles (LDL, IDL, VLDL, Lp(a)) while leaving smaller, more soluble apoA-I-containing particles (HDL2, HDL3, lipid-poor/lipid-free apoA-I, pre-β1 HDL) in solution.
Protocol: • Mix serum 1:1 with 20% PEG-8000 in glycine buffer (pH 10) • Incubate 20 min at room temperature, then centrifuge 20 min at 10,000×g, 4°C • Collect supernatant = "apoB-depleted serum" (the acceptor reagent) • Confirm depletion: apoB ELISA on supernatant should show >95% reduction vs whole serum • Aliquot and store at −80°C; avoid freeze-thaw >2× (degrades pre-β1 HDL, the most active acceptor subspecies)
Quality control: • Total cholesterol and apoA-I measured on every apoB-depleted batch to normalize acceptor dose across subjects • A pooled reference serum (large single batch, hundreds of aliquots) is run on every assay plate to correct for day-to-day and cell-passage variability — this is what ultimately allows CEC values from different assay days, months, or sites to be compared on a common SD-unit scale.
Early efflux assays used ultracentrifugally isolated HDL (d=1.063–1.21 g/mL) as the acceptor. This is now understood to discard biologically critical information:
• Ultracentrifugation exposes HDL to high salt density and shear stress for 18–48h, altering surface apoA-I conformation and stripping labile surface lipids • Lipid-poor/lipid-free apoA-I and pre-β1 HDL — the preferred, highest-affinity ligands for ABCA1 — are lost or under-represented in the ultracentrifuged d>1.21 fraction • ApoB-depletion by PEG precipitation preserves the entire physiological acceptor continuum in one step, at native concentration, reflecting what macrophages in an artery wall actually encounter
de la Llera-Moya et al. (Arterioscler Thromb Vasc Biol, 2010) formally validated apoB-depleted serum/plasma against ultracentrifuged HDL and cell-based efflux to define the now-standard clinical CEC protocol used across large cohort studies.
To measure how well a person's serum removes cholesterol, you first need a standardized cell that is loaded with cholesterol and ready to give it up. J774a.1 murine macrophage-like cells (or human THP-1 monocyte-derived macrophages) are the workhorse model: they express the full complement of efflux transporters found in arterial-wall macrophages, take up modified LDL through scavenger receptors exactly as foam cells do in an atherosclerotic plaque, and can be radiolabeled with reproducible, quantifiable precision.
Native LDL does not efficiently load macrophages — its receptor (LDLR) is downregulated by intracellular cholesterol. Foam-cell formation instead requires modified LDL:
• Acetylated LDL (AcLDL): lysine residues on apoB-100 are chemically acetylated, abolishing LDLR recognition and creating a ligand for scavenger receptor SR-A (MSR1) and CD36 • Scavenger receptors are NOT downregulated by cholesterol loading — uptake continues unchecked, exactly reproducing the pathological, unregulated cholesterol accumulation seen in arterial-wall macrophages • Internalized cholesteryl esters are hydrolyzed by lysosomal acid lipase to free cholesterol, re-esterified by ACAT1 into cytoplasmic lipid droplets, and hydrolyzed again by neutral cholesteryl ester hydrolase (NCEH1) — a continuous "futile cycle" that maintains a pool of mobilizable free cholesterol at the plasma membrane, the actual substrate for ABCA1/ABCG1 efflux
An ACAT inhibitor is included during the overnight equilibration step specifically to prevent newly-mobilized free cholesterol from being re-esterified and trapped in droplets, ensuring the radiolabel pool reflects membrane-accessible cholesterol available for transporter-mediated export.
³H-cholesterol (specific activity ~40–60 Ci/mmol) remains the gold-standard label because it tracks the true sterol molecule through uptake, esterification, hydrolysis, and efflux without altering its biophysical behavior.
Alternative approaches increasingly used for higher-throughput or non-radioactive workflows: • BODIPY-cholesterol: a fluorescent cholesterol analog readable by fluorometry/flow cytometry; correlates well (r>0.8) with ³H-based CEC but the bulky fluorophore can perturb membrane packing • Mass spectrometry-based efflux: LC-MS/MS quantification of unlabeled cholesterol mass transferred to acceptor, avoiding radioactivity entirely but requiring higher analytical sensitivity • Cell lines used: J774a.1 (mouse, most common, robust ABCA1/ABCG1 expression), human THP-1 macrophages (better models human receptor biology), and CHO or BHK cells stably transfected with human ABCA1 (isolates the ABCA1-specific component from total efflux)
Reverse cholesterol transport out of a macrophage is not a single reaction but the sum of at least three biophysically distinct pathways operating in parallel, each with its own acceptor preference, kinetics, and clinical significance. Cyclic-AMP stimulation with cpt-cAMP for 16h upregulates the dominant, rate-limiting one — ABCA1 — roughly 4–6-fold via the PKA→LXR/RXR transcriptional axis, converting the assay into a specific readout of ABCA1-mediated efflux capacity.
1) ABCA1 (ATP-binding cassette A1): an active, ATP-dependent lipid floppase that translocates cellular cholesterol and phospholipid to lipid-poor/lipid-free apoA-I, generating nascent discoidal pre-β HDL de novo. This is the rate-limiting, cAMP-inducible pathway probed by the standard clinical CEC assay and the one lost in Tangier disease.
2) ABCG1: transfers cholesterol preferentially to mature, spherical HDL2/HDL3 particles rather than lipid-poor apoA-I; works cooperatively downstream of ABCA1, moving cholesterol that ABCA1-generated nascent HDL has already begun to acquire.
3) SR-BI (scavenger receptor class B type I): a bidirectional facilitator that equilibrates free cholesterol between the plasma membrane and HDL along its concentration gradient — non-saturable, does not require ATP, and can also mediate net influx if the gradient reverses.
4) Aqueous diffusion: unassisted, non-saturable desorption of cholesterol monomers from the membrane into the aqueous phase followed by re-uptake by any nearby acceptor — the pathway measured as "background" in serum-free control wells and subtracted from total efflux.
Tangier disease, caused by loss-of-function mutations in ABCA1 (Bodzioch et al., Nat Genet 1999), produces near-absent HDL-C (<5 mg/dL vs. a normal 40–60 mg/dL) and orange, cholesteryl-ester-engorged tonsils — direct human proof that ABCA1-mediated efflux, not merely HDL-C concentration, is the pathway that protects against cholesterol accumulation and premature atherosclerosis.
After the 4-hour efflux reaction, the biology has to become a number. Media and cell fractions are separated, solubilized, and counted for beta-emission from the ³H label; the ratio between them, corrected against a serum-free background well, produces the percent-efflux value that underlies every downstream statistic in the assay.
Step 1 — Fraction collection: media supernatant is pipetted off each well into scintillation vials; the remaining cell monolayer is lysed in 0.1N NaOH and neutralized, then also transferred to vials.
Step 2 — Scintillation cocktail: an organic scintillant (e.g. Ultima Gold) is added to each vial; beta particles emitted by ³H decay excite the scintillant, which fluoresces, and photons are counted by the instrument's photomultiplier tubes as counts per minute (cpm).
Step 3 — Calculation: % efflux = [media cpm ÷ (media cpm + cell cpm)] × 100
Step 4 — Background subtraction: CEC (%) = % efflux(patient serum well) − % efflux(serum-free control well) This isolates the acceptor-dependent, transporter-mediated component from passive aqueous diffusion.
Step 5 — Batch normalization: every plate also carries the pooled reference serum; a patient's CEC is expressed relative to that plate's reference value (ratio or SD-unit), which is what allows CEC measured on different days, different cell passages, or in different laboratories to be statistically pooled in cohort studies.
The entire point of the assay is captured in this final stage: does the number predict disease? Across multiple large, independent cohorts, cholesterol efflux capacity predicts incident cardiovascular events and prevalent atherosclerosis better than — and independently of — HDL cholesterol concentration itself, resolving why raising HDL-C pharmacologically has repeatedly failed to reduce cardiovascular events.
Khera et al. (NEJM, 2011) first showed in the Dallas Heart Study (n=2,924) that cholesterol efflux capacity was inversely associated with carotid intima-media thickness and angiographically-defined coronary artery disease — and that this association was independent of, and stronger than, HDL-C or apoA-I levels alone.
Subsequent prospective cohorts extended this to hard clinical endpoints: • Rohatgi et al. (NEJM, 2014), Dallas Heart Study prospective follow-up: each 1-SD increase in CEC associated with a ~33% reduction in incident cardiovascular events over a median 9.4-year follow-up, after adjustment for traditional risk factors and HDL-C • Saleheen et al. (Lancet Diabetes Endocrinol, 2015), EPIC-Norfolk case-control cohort (n≈2,600): CEC inversely associated with incident coronary heart disease, again independent of HDL-C • Li et al. and multiple subsequent meta-analyses have replicated the core finding: CEC, not HDL-C mass, tracks with cardiovascular protection
For two decades, raising HDL-C was pursued as a therapeutic strategy on the assumption that more HDL cholesterol equals more protection. CETP (cholesteryl ester transfer protein) inhibitors — torcetrapib, dalcetrapib, evacetrapib — raised HDL-C by 30–130% in large randomized trials, yet none reduced cardiovascular events (torcetrapib was withdrawn after excess mortality in ILLUMINATE, 2007).
The efflux capacity framework explains this apparent paradox: HDL-C is a static mass measurement of cholesterol carried inside HDL particles, but says nothing about whether those particles are functionally competent to accept more cholesterol from peripheral cells. CETP inhibition inflates HDL particle size and cholesterol cargo without necessarily improving — and in some cases while impairing — the particles' functional capacity to drive efflux from macrophages.
This is why cholesterol efflux capacity is increasingly described as a functional assay of "HDL quality," positioned as a candidate risk-stratification biomarker and pharmacodynamic endpoint for the next generation of HDL-directed and reverse-cholesterol-transport-directed therapeutics.
The clinical lesson of CEC research is that reverse cholesterol transport is a process, not a number on a lipid panel. A patient with low HDL-C but high efflux capacity may be at lower cardiovascular risk than a patient with high HDL-C but dysfunctional, low-efflux-capacity particles — which is why efflux assays, not cholesterol mass, are now used to interrogate HDL function in mechanistic and therapeutic cardiovascular research.