Comment on 'Orthogonal lipid sensors identify transbilayer asymmetry of plasma membrane cholesterol'

Abstract

The plasma membrane in mammalian cells is rich in cholesterol, but how the cholesterol is partitioned between the two leaflets of the plasma membrane remains a matter of debate. Recently, Liu et al. used domain 4 (D4) of perfringolysin O as a cholesterol sensor to argue that cholesterol is mostly in the exofacial leaflet (Liu et al., 2017). This conclusion was made by interpreting D4 binding in live cells using in vitro calibrations with liposomes. However, liposomes may be unfaithful in mimicking the plasma membrane, as we demonstrate here. Also, D4 binding is highly sensitive to the presence of cytosolic proteins. In addition, we find that a D4 variant, which requires >35 mol% cholesterol to bind to liposomes in vitro, does in fact bind to the cytoplasmic leaflet of the plasma membrane in a cholesterol-dependent manner. Thus, we believe, based on the current evidence, that it is unlikely that there is a significantly higher proportion of cholesterol in the exofacial leaflet of the plasma membrane compared to the cytosolic leaflet.

Keywords: cell biology; cholesterol; none; orthogonal lipid sensors; plasma membrane; transbilayer asymmetry.

Figure 1.. D4 binding is influenced by phospholipid composition and is subject to competition from proteins.

(A) Purified DAN-D4 (0.5 µM) was incubated with 100 µM large unilamellar vesicles (LUVs) composed of POPC/egg SM/cholesterol (36:24:40) and POPC/POPE/POPS/soy PI/cholesterol (18:18:18:6:40). The change in fluorescence emission (ΔF) at 450 nm is used to approximate cholesterol-dependent liposome binding and is corrected for non-specific binding to a cholesterol-free liposome. The results were normalized to the maximal ΔF (ΔFmax). (B) DAN-D4 (0.5 µM) binding to increasing concentrations of phosphatidylcholine/cholesterol (60:40) LUVs with various phosphatidylcholine acyl chain saturation. (C) DAN-D4 (0.5 µM) binding to 100 µM DOPC/cholesterol (60:40) LUVs in the presence of increasing concentrations of rat liver cytosol. The change in fluorescence was determined relative to cholesterol-free liposomes at 450 nm and then normalized to the control (ΔF/F). All data were acquired with a PTI scanning spectrofluorometer (ex. 380 nm and em. 420–560 nm). Each experiment was repeated at least three times and error bars represent standard error of the mean.

Figure 2.. D4 variants can bind to the cytoplasmic leaflet of the PM in a cholesterol-dependent manner.

(A) CHO cells transiently transfected with mCherry-D4^D434A and D4^{D434A, A463W} and the plasma membrane marker, Pleckstrin homology domain of phospholipase C δ (PH-PLC δ) were examined using spinning-disc confocal microscopy. (B) Live-cell images were acquired of cells expressing the same probes as in (A) following incubation with 10 mM methyl-β-cyclodextrin (mβCD) for 20 min to extract plasmalemmal cholesterol. Scale bar, 10 µm. (C) Quantitation of the plasmalemmal enrichment of the mCherry signal seen in (A) and (B). means ± std. dev. n = 20.