Activation Thermodynamics of Poly(Ethylene Glycol)-Mediated Model Membrane Fusion Support Mechanistic Models of Stalk and Pore Formation

Abstract

Membrane fusion, essential to eukaryotic life, is broadly envisioned as a three-step process proceeding from contacting bilayers through two semistable, nonlamellar lipidic intermediate states to a fusion pore. Here, we introduced a new, to our knowledge, experimental approach to gain insight into the nature of the transition states between initial, intermediate, and final states. Recorded time courses of lipid-mixing, content-mixing, and content-leakage associated with fusion of 23 nm vesicles in the presence of poly(ethylene glycol) at multiple temperatures were fitted globally to a three-step sequential model to yield rate constants and thereby activation thermodynamics for each step of the process, as well as probabilities of occurrence of lipid-mixing, content-mixing, or content-leakage in each state. Experiments with membranes containing hexadecane, known to reduce interstice energy in nonlamellar structures, provided additional insight into the nature of fusion intermediates and transition states. The results support a hypothesis for the mechanism of stalk formation (step-1) that involves acyl chain protrusions into the interbilayer contact region, a hypothesis for a step-2 mechanism involving continuous interconversion of semistable nonlamellar intermediates, and a hypothesis for step-3 (pore formation) mechanism involving correlated movement of whole lipid molecules into hydrophobic spaces created by geometry mismatch between intermediate structures.

A Novel Assay for Detecting Fusion Pore Formation: Implications for the Fusion Mechanism

“…It is immediately obvious from Figures 3A and 3B(ii) that these two types of assays for content mixing report very different events occurring on different time scales. The three time courses recorded in Figures 3B(i-iii) all required description by double exponential curves, but together allowed us to analyze the data obtained with the Tb 3+ /DPA content mixing/leakage assay in terms of our ensemble kinetic treatment as described in Methods (9, 10). This treatment determines a rate constant k 1 and for converting the aggregated state (A) to the initial intermediate state (I), which then converts to a final fusion pore (FP state) with a rate constant k 3 (9, 10, 22).…”

“…Since the diameter of an SUV is ~ 23-25 nm, it is unlikely that more than one such large pore could form between SUVs at one time and one point of contact. A pore of this size would require lipid rearrangements substantially larger than the correlated movements of only a few lipids that has been suggested to constitute the transition state to pore formation (10). Such substantial lipid rearrangements are reported to occur at rates comparable to that of final pore formation.…”

“…The answer lies in the ensemble nature of our experiments and analysis (9, 10, 20, 26) and in the similar ensemble nature of dwell-time analysis of single events (8). The modified stalk hypothesis (3), as applied to fusion of a single pair of highly curved vesicles, predicts that the free energy of intermediate structures from the initial stalk (I 1 ) to the final pre-fusion intermediate (I 2 ) evolves continuously as the geometry and structure of the pair of hemi-fused vesicles evolves (36).…”

“…We analyzed Tb 3+ /DPA content mixing and leakage along with BODIPY500-PC and BODIPY530-PE lipid-mixing data in terms of a sequential, 3-state (two step) fusion process (see following Diagram as described in detail elsewhere (10). Each “state” represents a thermodynamic ensemble of similar molecular arrangements (“microstructures”) (9), each of which is related to the structures proposed by Siegel and others (3), although it is envisioned that the range of molecular microstructures contributing to each thermodynamic state is quite broad.…”

“…The time course was best described by a single exponential equation, y = y 0 + a (1− e − bx ), with the resulting parameters tabulated in Table 1. (B) Traces of (i) lipid mixing, (ii) content mixing and (iii) content leakage for the PC/PE/SM/CH SUVs at pH 7.4 using conventional assays (10) in absence (red) and in presence of gp41 fusion peptide (green) at a peptide-to-lipid ratio1:200. Membrane fusion was induced with 5% PEG at 23 °C.…”