Membrane Lipid Reshaping Underlies Oxidative Stress Sensing by the Mitochondrial Proteins UCP1 and ANT1

Abstract

Oxidative stress and ROS are important players in the pathogenesis of numerous diseases. In addition to directly altering proteins, ROS also affects lipids with negative intrinsic curvature such as phosphatidylethanolamine (PE), producing PE adducts and lysolipids. The formation of PE adducts potentiates the protonophoric activity of mitochondrial uncoupling proteins, but the molecular mechanism remains unclear. Here, we linked the ROS-mediated change in lipid shape to the mechanical properties of the membrane and the function of uncoupling protein 1 (UCP1) and adenine nucleotide translocase 1 (ANT1). We show that the increase in the protonophoric activity of both proteins occurs due to the decrease in bending modulus in lipid bilayers in the presence of lysophosphatidylcholines (OPC and MPC) and PE adducts. Moreover, MD simulations showed that modified PEs and lysolipids change the lateral pressure profile of the membrane in the same direction and by the similar amplitude, indicating that modified PEs act as lipids with positive intrinsic curvature. Both results indicate that oxidative stress decreases stored curvature elastic stress (SCES) in the lipid bilayer membrane. We demonstrated that UCP1 and ANT1 sense SCES and proposed a novel regulatory mechanism for the function of these proteins. The new findings should draw the attention of the scientific community to this important and unexplored area of redox biochemistry.

Keywords: bending moduli; lateral pressure profile; lipid shape; lipid–protein interaction; mitochondrial membrane protein; protonophoric function; reactive aldehydes; stored curvature elastic stress.

Figure 1

Determination of the elastic parameters of the lipid bilayer membrane based on nanotube (NT) pulling. (A) Scheme of NTs pulled from a bilayer lipid membrane (grey) and held by a patch-pipette (white). A voltage applied to the ends of the NTs (U) induces an ion current (I_NT) flowing through the NT interior. (B) A representative measurement shows the dependence of measured membrane conductance (G) (black circles) on the NT length change (ΔL), required for calculation of the NT radius (r_NT) (see Section 2). (C) Dependence r_NT on U, obtained for the membrane lipid compositions DOPC:CL 90:10 (black) and DOPC:DOPE:CL 45:45:10 (red). (D) Box plot and distribution of the bending modulus, k, depending on the molar concentration of the phosphatidylethanolamine (DOPE). The buffer solution contained 100 mM KCl, 10 mM HEPES, and 1 mM EDTA at pH 7.0 and T = 295 K. Data points represent the mean and standard deviation from more than 10 independent experiments. *** p < 0.001, t-test.

Figure 2

Elastic properties of the lipid bilayer membrane in the presence of reactive aldehydes (RAs). (A,B) Relative bending rigidity k/k₀ and relative lateral tension σ/σ₀ for lipid membranes composed of DOPC:DOPE:CL 45:45:10 in (A) and DOPC:CL 90:10 in (B), incubated with the RAs 4-hydroxy-2-hexenal (HHE), 4-hydroxy-2-nonenal (HNE), and 4-oxo-2-nonenal (ONE). k₀ and σ₀ are bending modulus and lateral tension in the absence of RAs, respectively. RAs were added in a concentration range of 0.5–0.7 mM. Buffer composition: 100 mM KCl, 10 mM HEPES, 1 mM EDTA pH = 7.0, T = 295 K. Data points represent mean and standard deviation from more than 10 independent experiments. * p < 0.05; ** p < 0.01; *** p < 0.001, t-test.

Figure 3

Impact of PE and OPC on the elastic parameters k/k₀ (A) and σ/σ₀ (B). k₀ and σ₀ are bending modulus and lateral tension in the absence of PE, respectively. Buffer solution contained 100 mM KCl, 10 mM HEPES, 1mM EDTA pH = 7.0, T = 295 K. Data points represent mean and standard deviation from more than 10 independent experiments. * p < 0.05; ** p < 0.01; *** p < 0.001, t-test.

Figure 4

Lipid shape affects UCP1- and ANT1-mediated total membrane conductance (G_m). (A,B) Dependence of G_m measured in membranes reconstituted with recombinant UCP1 on the concentration of lysolipids MPC (A) and OPC (B). (C). Dependence of G_m measured in membranes reconstituted with recombinant ANT1 on the concentration of OPC. The lipid composition in control experiments was DOPC:DOPE:CL (45:45:10). The specified lysolipid amount (mol%) was used instead of DOPC and DOPE. The concentrations of UCP1 and ANT1 were 4–5 µg/(mg lipid). The concentrations of lipid and AA were 1.5 mg/mL and 15 mol%, respectively. The buffer solution contained 50 mM Na₂SO₄, 10 mM MES, 10 mM Tris, and 0.6 mM EGTA, at pH = 7.32 and T = 305 K. Data points represent means and standard deviation from 3–5 independent experiments.

Figure 5

MD simulations of LPP dependency on lipid shape. (A–C) Impact of lipids with pronounced SC on the LPP, p. (D–F) Comparison of areas below the pressure profiles in the headgroup (head), water–lipid interface (w/l), and acyl chains (acyl) and for the whole bilayer profile (all) for p shown in (A–C). The lipid ratio in bi-component membranes was 50:50 mol%. Color labels: DOPC:DOPE (red), DOPC (grey), DOPC:ONE-DOPE (cyan), and DOPC:OPC (blue).

Figure 6

Proposed mechanism for the SCES sensing by mitochondrial proteins ANT1 and UCP1. (A) Schematic representation of the lipid membrane leaflet and change in lipid shapes under oxidative stress in mitochondria. (B) Effect of PE adducts, ONE-PE and HNE-PE, on the UCP1-mediated G_m in the presence of AA. The relative membrane conductance G_m/G_m,DOPC was calculated based on data from [37]. G_m,DOPC, and G_m are specific membrane conductance in the absence or presence of PE or RA-PEs, respectively. (C) Schematic representation of the LPP redistribution in the lipid bilayer membrane caused by a change in the lipid shape. A comparison of the LPP (left) and protein structure (right) suggests that the decreased lateral pressure (dark blue), which appeared in the area of the FA binding site (*) and in the protein cavity region (**), promotes FA translocation. Most likely, the transformed lipid environment (i) increases the probability that anionic FA reaches a protein binding site and (ii) facilitates a protein conformational change, which in turn supports a faster translocation of FA to the opposite leaflet. (D) The increase in the UCP1- and ANT1- mediated G_m correlates with a decrease in both the lateral pressure, p, and bending rigidity, k, in the lipid bilayer membrane. If the change in p and k goes in the opposite direction, the protein-mediated G_m is not affected, as demonstrated here for the DOPE-containing membrane.