Assembly formation of minor dihydrosphingomyelin in sphingomyelin-rich ordered membrane domains

Abstract

The lipidome of mammalian cells not only contain sphingomyelin (SM) but also, as a minor component, dihydrosphongomyelin (DHSM), in which the double bond at C4-C5 in the sphingosine base is reduced to a single-bond linkage. It has been indicated that DHSM forms ordered domains more effectively than SM due to its greater potential to induce intermolecular hydrogen bonds. However, direct information on partition and dynamic behaviors of DHSM in raft-like liquid-ordered (L_o) and non-raft-like liquid-disordered (L_d) phase-segregated membranes has been lacking. In the present study, we prepared fluorescent derivatives of DHSM and compared their behaviors to those of fluorescent SM and phosphatidylcholine (PC) derivatives. Fluorescence microscopy showed that DHSM is more preferentially localized to the L_o domains in the L_o/L_d phase-segregated giant unilamellar vesicles than SM and PC. Most importantly, diffusion coefficient measurements indicated that DHSM molecules form DHSM-condensed assembly inside the SM-rich L_o domain of the SM/dioleoylphosphatidylcholine/cholesterol system even when DHSM accounts for 1-3.3 mol% of total lipids. Such heterogeneous distribution of DHSM in the SM-rich L_o domains was further confirmed by inter-lipid FRET experiments. This study provides new insights into the biological functions and significance of minor component DHSM in lipid rafts.

Figure 1

Structures of dipalmitoylphosphatidylcholine (DPPC), palmitoylsphingomyelin (pSM), palmitoyldihydrosphingomielin (DHpSM), and their fluorescent derivatives, 488neg- and 594neg-DPPCs (neg-DPPCs), 488neg- and 594neg-pSMs (neg-pSMs), and 488neg- and 594neg-DHpSMs (neg-DHpSMs). While fluorescently labeled stearoyl-SMs (C18:0) were previously reported, the lengths of the lipid acyl chains in this study were unified as palmitoylate (C16:0).

Figure 2

Preparation of 488neg- and 594neg-DHpSM.

Figure 3

Fluorescence micrographs of ternary-component GUVs consisting of (left) pSM/DOPC/chol, (center) DHpSM/DOPC/chol, and (right) DPPC/DOPC/chol at a molar ratio of 1:1:1. These samples contain 0.2 mol% neg-pSMs, neg-DHpSM, or neg-DPPC. The L_d phase was labeled by 0.2 mol% 488neg-DOPC or TXred-DPPE. All of these samples underwent phase separation between the L_o and L_d domains. A bar indicates 10 μm. The brightness and contrast have been enhanced for clarity.

Figure 4

Schematic drawing of the formation of DHpSM-condensed assembly in the L_o domain. The blue and red heads correspond to pSM and DHpSM, respectively. Orange membrane regions indicate the DHpSM-condensed assembly (see text for details).

Figure 5

Fluorescent spectra of (a) pSM/chol (70:30 by moles) and (b) pSM/DHpSM/chol (66.5:3.5:30 by moles) multi-lamellar vesicles. These samples contain 488neg-pSM/594neg-pSM FRET pair (neg-pSMs; black curve) or 488neg-DHpSM/594neg-DHpSM FRET pair (neg-DHpSMs; red curve). The content of each fluorescent SM is 0.4 mol%. Fluorescent measurements were repeated five and four times for the pSM/chol and pSM/DHpSM/chol samples, respectively, and averaged spectra are shown. Error bars show standard errors. The vertical axes indicate the normalized intensity, in which emission peak intensity at 525 nm of the acceptor-free sample was set to 1. Therefore, (1 − normalized peak intensity at 525 nm) corresponds to the FRET efficiency. Panel (a) shows that the FRET efficiency for 488neg-pSM/594neg-pSM pair in pSM/chol is ca. 50%. On the other hand, applying Refs.^, using the Förster-radius (5.7 nm) and acceptor molar ratio (0.4 mol%), the FRET efficiency between randomly distributed donor and acceptor is calculated to be ca. 60%. This is roughly consistent with the observed FRET efficiency for 488neg-pSM/594neg-pSM (ca, 50%), thus supporting their random distribution in pSM-chol membrane.

Figure 6

Fluorescent micrographs of (top) 488neg-pSM and (bottom) 488neg-DHpSM-labeled erythrocyte ghosts membranes (left) before and (right) after Triton X-100 treatment. The intensity ratio of after Triton X-100 treatment (I_+TX) to before treatment (I_control) was described at the left-side of the corresponding micrographs. The insertion of each panel shows magnification of the region indicated by the dashed square.