Search

Membrane-active peptides (MAPs) have several potential therapeutic uses, including as antimicrobial drugs. Many traditional methods used to evaluate the membrane interactions of MAPs have limited applicability. Low-throughput methods, such as microscopy, provides detailed information but often relies on fluorophore-labeled MAPs, and high-throughput assays, such as the calcein release assay, cannot assess the mechanism behind the disruption of vesicular-based lipid membranes. Here we present a flow cytometric assay that provides detailed information about the peptide-lipid membrane interactions on single artificial lipid vesicles while being high-throughput (1000–2000 vesicles/s) and based on label-free MAPs. We synthesized and investigated six MAPs with different modes of action to evaluate the versatility of the assay. The assay is based on the flow cytometric readouts from artificial lipid vesicles, including the fluorescence from membrane-anchored and core-encapsulated fluorophores, and the vesicle concentration. From these parameters, we were able to distinguish between MAPs that induce vesicle solubilization, permeation (pores/membrane distortion), and aggregation or fusion. Our flow cytometry findings have been verified by traditional methods, including the calcein release assay, dynamic light scattering, and fluorescence microscopy on giant unilamellar vesicles. We envision that the presented flow cytometric assay can be used for various types of peptide-lipid membrane studies, e.g. to identify new antibiotics. Moreover, the assay can easily be expanded to derive additional valuable information.

The use of high-throughput flow cytometry to characterize nanoparticles has received increased interest in recent years. However, to fully realize the potential of flow cytometry for the characterization of nanometer-sized objects, suitable calibrators for size estimation must be developed and the sensitivity of conventional flow cytometers has to be advanced. Based on the scattered signal, silica and plastic beads have often been used as flow cytometric size calibrators to evaluate the size of extracellular vesicles and artificial vesicles (liposomes). However, several studies have shown that these beads are unable to accurately correlate scatter intensity to vesicle size. In this work, we present a novel method to estimate the size of individual liposomes in flow cytometry based on liposomal size calibrators prepared by fluorescence-activated cell sorting (FACS), here coined fluorescence-activated nanoparticle sorting (FANS). These calibration liposomes exhibit sizes, structures, and refractive indexes identical to the particles being studied and thus can serve as unique calibrators. First, a sample of polydisperse fluorophore-labeled unilamellar liposomes was prepared and analyzed by flow cytometry. Next, different fractions of the polydisperse liposomes were FANS-sorted according to their fluorescence intensity. Thereafter, we employed nanoparticle tracking analysis (NTA) to evaluate the liposome sizes of the FANS-sorted liposome fractions. Finally, we correlated the flow cytometric readouts (side scatter and fluorescence intensity) of the FANS-sorted liposome fractions with their corresponding size obtained by NTA. This procedure enabled us to translate the liposome fluorescence intensity to the liposome size in nanometers for all detected individual liposomes. We validated the size distribution of our polydisperse liposome sample obtained from flow cytometry in combination with our FANS-calibrators against standard methods for sizing nanoparticles, including NTA and cryo-transmission electron microscopy. This work also highlights the limitation of using the flow cytometric side scattering readout to determine the size of small (30-300 nm) artificial vesicles. © 2019 International Society for Advancement of Cytometry.

A confocal fluorescence microscopy-based assay was used for studying the influence of the preparation route on the supramolecular organization of lipids in a vesicular system. In this work, vesicles composed of cholesterol and CTAB (1/1 mol %) or cholesterol and DOPC (2/8 mol %) and incorporating two membrane dyes were prepared by either a compressed fluid (CF)-based method (DELOS-susp) or a conventional film hydration procedure. They were subsequently immobilized and imaged individually using a confocal fluorescence microscope. Two integrated fluorescence intensities, I(dye1) and I(dye2), were assigned to each tracked vesicle, and their ratio, I(dye1)/I(dye2), was used for quantifying the degree of membrane inhomogeneity between individual vesicles within each sample. A distribution of I(dye1)/I(dye2) values was obtained for all the studied vesicular systems, indicating intrasample heterogeneity. The degree of inhomogeneity (DI) was similar for Chol/DOPC vesicles prepared by both procedures. In contrast, DI was more than double for the hydration method compared to the CF-based method in the case of Chol/CTAB vesicles, which can suffer from lipid demixing during film formation. These findings reveal a more homogeneous vesicle formation path by CFs, which warranted good homogeneity of the vesicular system, independently of the lipid mixture used.

In vivo studies have reported preferential partitioning of Ras GTPases into ordered lipid-protein membrane domains, a process believed to regulate both cellular signaling and protein trafficking.1 However studies in vitro have failed to quantify a preferential partitioning of full length Ras proteins into the liquid ordered phase2,3 and thus a biophysically validated mechanism for in vivo sorting of Ras is still missing. We recently showed that lipidated proteins localize to highly curved membranes in vitro.4 Here we study both in vitro and in vivo whether recruitment by membrane curvature can sort full length lipid-anchored Ras.We employ a single vesicle fluorescence based assay to quantify in vitro the sorting by membrane curvature of full-length Ras proteins. We demonstrate a more than 50 fold increase in protein density on membranes of high curvature as compared to the density on flat membranes. To test for recruitment by membrane curvature in vivo we utilize hypo-osmotic swelling of cells, which flattens curved membrane regions. By measuring the local protein density using FRET,5 we detect a significant reduction in the clustering of Ras and other lipidated proteins upon membrane flattening. This demonstrates that recruitment by membrane curvature can sort Ras and potentially other lipidated proteins in cellular membranes. Furthermore sorting by membrane curvature constitutes the first biophysical sorting mechanism for Ras validated by both in vitro and in vivo measurements.1 Hancock, J. F. Nat. Rev. Mol. Cell Biol.4 (2003).2 Johnson, S. A. et al.Biochim. Biophys. Acta - Biomembranes1798 (2010).3 Nicolini, C. et al.J. Am. Chem. Soc.128 (2006).4 Hatzakis, N. S. et al.Nat. Chem. Biol.5 (2009).5 Kohnke, M. et al.Chem. Biol.19 (2012).

Liposomes, or vesicles, have been studied extensively both as models of biological membranes and as drug delivery vehicles. Typically it is assumed that all liposomes within the same preparation are identical. Here by employing pairs of fluorescently labeled lipids we demonstrated an up to 10-fold variation in the relative lipid composition of individual liposomes with diameters between 50 nm and 15 μm. Since the physicochemical properties of liposomes are directly linked to their composition, a direct consequence of compositional inhomogeneities is a polydispersity in the properties of the individual liposomes in an ensemble.

A critical feature of neurotransmission is that only a subset of vesicles react at a given Ca++ pulse. To achieve such a fractional response the synapse utilises a number of complex spatio-temporal and biochemical mechanisms that are still under debate but collectively appear to create distinct vesicle pools. We currently ignore whether differentiation of seemingly identical vesicles exists also for membrane trafficking vesicles.We recently established that vesicles within an ensemble display large compositional inhomogeneities(1). Annexins, as well as other Ca2+-dependent membrane binding proteins, are known to bind specific phospholipid components in cellular membranes(2). Here we investigated how the unique capability of Ca2+-regulated membrane-binding proteins to bind certain lipids enables them to discriminate between single vesicles based on their lipid compositional heterogeneity. We employed human Annexin A5 (AnxA5), which is widely used as a model system for the membrane binding properties of the Annexin superfamily(2), and our previously described single vesicle assay(3, 4). We demonstrate that heterogeneities in the lipid composition of individual vesicles within a population, in conjunction with Ca2+ concentration, regulate the fraction of vesicles recruiting Annexins with digital precision. This is a recurrent phenomenon observed in multiple vesicle samples and for different Ca2+ sensing proteins. Thus, we anticipate that the cell utilize this feature to translate analog Ca2+ influxes into a selective digital targeting of vesicle sub-populations within the ensemble for specific functions.1. J. Am. Chem. Soc. 133: 10685-10687.2. Nat Rev Mol Cell Biol. 6: 449-461.3. The EMBO Journal. 28: 3303-3314.4. Nat. Chem. Biol. 5: 835-841.

Sorting and trafficking of membrane-anchored Ras GTPases is critical for signaling and is believed to rely on their preferential portioning in ordered lipid-protein membrane domains (1). However studies in vitro have failed to quantify the preferential partitioning of full length Ras proteins into the liquid ordered phase(2), indicating that a physical principle underlying sorting of Ras is missing. We recently showed that lipidated proteins localize to highly curved membranes in vitro(3, 4). Here we provide a mechanistic insight on how membrane curvature can drive N-Ras sorting.Combining the results of our in vitro assays, measurements on single vesicles, with in vivo studies, hypo-osmotic swelling of cells that flattens curved membrane regions, revealed that : a) N-Ras is preferentially recruited in areas of high membrane curvature and b) membrane curvature is the enabling factor underlying the selective partitioning of NRas in ordered domains. The combined readout of mean field theory calculations and coarse grain simulations provided a mechanistic insight on preferential partitioning in highly curved areas, via the changes in lateral pressure of the outer monolayer when curving an ordered versus a disordered membrane. In addition to providing the first biophysical sorting mechanism for Ras validated by both in vitro and in vivo measurements, our data indicate that membrane curvature may act as a generic cue underlying trafficking and sorting of multiple lipidated proteins.References1. J. F. Hancock, Nat. Rev. Mol. Cell Biol. 4, 373 (2003).2. S. A. Johnson Biochim. Biophys. Acta - Biomembranes 1798, 1427 (2010).3. N. S. Hatzakis et al., Nat. Chem. Biol. 5, 835 (2009).4. V. K. Bhatia, et al Semin. Cell Dev. Biol. 21, 381 (2010).

Ras proteins are small GTPases that are post-translationally modified by the attachment of lipid moieties (1). This modification is essential for the correct trafficking and sorting of Ras proteins through the vesicular pathway from the Golgi to the plasma membrane (2).Traditionally the sorting of Ras is primarily discussed in the context of membrane domains in flat membranes, neglecting the influence of membrane shape. Recently we have demonstrated, utilizing our single liposome curvature (SLiC) assay that the minimal anchoring motif of N-Ras (tN-Ras) up-concentrates in areas of high membrane curvature (not published), suggesting that curvature might act as a cue for the spatial localization of Ras proteins.In the SLiC assay, curvature-sensing molecules are added from aqueous solution to vesicles of different curvatures, but the vesicles are not in diffusive contact (3, 4). In vivo Ras proteins are anchored to membranes and laterally sorts between curved and planar membranes, which are in diffusive contact. To study the curvature-sensing ability of tN-Ras in a setup mimicking the in vivo scenario we developed a membrane tube based assay in which the tubes are in diffusive contact with a lipid bilayer.Membrane tubes are formed by heating a confined lipid bilayer (5). The tubes eventually adsorb to the flat membrane, which enable imaging by confocal fluorescence microscopy. After addition of tN-Ras we observed a preferential sorting into curved tubes rather than the flat bilayer. This observation further implies a pivoting role of membrane shape as a regulator of Ras-protein localization.1. Prior & Hancock, Semin Cell Dev Biol 23:145(2012).2. Choy et al., Cell 98:69(1999).3. Kunding et al., Biophysical Journal 95:1176(2008).4. Hatzakis et al., Nat Chem Biol 5:835(2009).5. Weirich & Fygenson, Poster Abstract 2731, Biophys. Soc. 2011.