$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The high-density lipoprotein-specific phospholipid efflux (HDL-SPE) assay measures HDL functionality and is intended for use in biomedical research and clinical diagnostic laboratories to (i) gain further insights into the mechanisms underlying HDL-mediated removal of lipids from atherosclerotic lesions, (ii) serve as a new platform to improve clinical CVD risk assessment and, (iii) develop new therapeutics to treat cardiovascular disease.
HDL-cholesterol (HDL-C) is an important determinant of CVD risk1, but its usefulness is compromised insofar as elevated levels of HDL-C (>80 mg/dL) are associated with greater CVD risk2,3,4,5. Elevated levels of HDL-C may represent the accumulation of dysfunctional HDL that has a reduced capacity to remove arterial plaque lipids6.
The cellular cholesterol efflux capacity (CEC) assay was the first in vitro assay developed to measure HDL functionality. In this assay, cultured macrophages (mouse or human), typically labeled with 3H- or fluorescent Bodipy-cholesterol, are incubated with apoB-depleted plasma or serum. The % cholesterol efflux is calculated based on the amount of cell- and medium-associated fluorescence/radioactivity7,8,9. The CEC assay has been shown to be a better predictor of incident CVD than HDL-C7. The CEC assay is not amenable to conversion to a routine clinical diagnostic assay mainly due to its requirements for cell culture and long duration (e.g., days). Furthermore, the CEC assay is limited by its inherent variability (e.g., different types of cells, labeling techniques, and PEG-depletion of β-lipoproteins, which alters HDL composition10).
Several research groups have developed cell-free HDL cholesterol efflux assays for use as tools to gain new insights into HDL function and for potential translation into a clinical diagnostic assay to better assess CVD risk11,12. These efflux assays typically involve multiple steps and use a lipid donor particle labeled with a fluorescent cholesterol, and like the CEC assay, they either require depletion of β-lipoproteins or, alternatively, the capture of apoA-I-containing particles using anti-apoA-I antibodies, to measure plasma HDL cholesterol efflux. Although these alternative assays have been validated to be associated to varying degrees with CVD risk, to date, only one13, which assesses apoA-I exchange rate, has been shown to predict incident CVD risk14. Interestingly, the kinetics of the exchange of apoA-I on and off plasma HDL has been shown to be independently associated with CEC15 and to be markedly reduced in patients with acute coronary syndrome16. In addition, several HDL exchangeable apolipoproteins have been shown to mediate both cholesterol and phospholipid from ABCA1-expressing cells in vitro17.
We have developed an alternative approach to assess HDL functionality18. The simple, rapid, and reliable assay described here requires only two components, namely whole human (or animal) plasma or serum samples (fresh or freeze-thawed), and a fluorescent lipid-coated donor particle. The assay measures the ability of apoA-I and other HDL-associated exchangeable apolipoproteins in whole plasma/serum to remove a non-exchangeable, head group-tagged fluorescent phosphatidylethanolamine (PE) from a lipid-coated donor particle18 (Figure 1).
The donor particles in this assay consist of calcium silicate hydrate crystals (CSH) coated with cholesterol, the phospholipid phosphatidylcholine (PC), and a fluorescent PE. Exchangeable HDL-associated apolipoproteins, mostly apoA-I, released from HDL particles bind to the donor particles and then dissociate from them along with donor particle lipids, including fluorescent PE. In this manner, HDL, but not the other lipoproteins present in whole plasma/serum, specifically acquires fluorescent PE from the donor particles18,19. Whole plasma/serum is incubated with the fluorescent lipid-coated calcium silicate hydrate (LC-CSH) donor particles and saline at 37 °C, with shaking for 1 h. Since LC-CSH is about twice the density of water, the donor particles are readily separated from plasma by low-speed centrifugation.
The % PE efflux measured in clinical studies is typically normalized by dividing by the value of the PE fluorescence acquired by HDL in the supernatant of the reaction mixture of a clinical sample by the value obtained by HDL in a reference control human plasma. Normalization to a reference plasma is performed to correct for variations in reagent preparation. For research applications, % PE efflux can be determined by calculating the supernatant PE fluorescence divided by the total fluorescence (supernatant PE fluorescence + LC-CSH fluorescence). Supernatant fluorescent measurement values are pasted into a template in spreadsheet files to calculate %PE Efflux (Supplementary Table 1 for triplicate measurements and Supplementary Table 2 for duplicate measurements). The instructions for performing these calculations in the spreadsheet are provided as supplementary files (Supplementary File 1, Supplementary File 2, Supplementary File 3, Supplementary File 4, Supplementary File 5, Supplementary File 6, Supplementary File 7, and Supplementary File 8), and an example calculation is provided as Supplementary Table 3.

Figure 1: HDL-SPE Assay. HDL-specific phospholipid efflux is assayed in 96-well plates. Saline, calcium silicate hydrate crystals (gray) coated with lipids (LC-CSH), including non-exchangeable lissaminerhodamine-labeled phosphatidylethanolamine (LRh-PE; red membrane layers), and up to 30 whole plasma/serum samples (yellow; in triplicate) are loaded onto 96-well plates. After mixing for 1 h at 37 °C in a thermomixer, plates are briefly centrifuged to pellet the LC-CSH donor particles. Aliquots of the supernatants, containing HDL specifically labeled with LRh-PE, are transferred to a black plate containing detergent (to solubilize Lh-Rh-PE) and saline. PE fluorescence is then measured with a fluorimeter. Saline (Negative Control; NC), for background correction, and pooled healthy human plasma (Positive Control; PC) for plate normalization, are loaded instead of sample plasma/serum in the last set of wells. Please click here to view a larger version of this figure.