Mechanics of lipid bilayer junctions affecting the size of a connecting lipid nanotube
 Roger Karlsson^{1},
 Michael Kurczy^{2},
 Richards Grzhibovskis^{3},
 Kelly L Adams^{4, 1},
 Andrew G Ewing^{1},
 AnnSofie Cans^{2} and
 Marina V Voinova^{5}Email author
DOI: 10.1186/1556276X6421
© Karlsson et al; licensee Springer. 2011
Received: 22 December 2010
Accepted: 14 June 2011
Published: 14 June 2011
Abstract
In this study we report a physical analysis of the membrane mechanics affecting the size of the highly curved region of a lipid nanotube (LNT) that is either connected between a lipid bilayer vesicle and the tip of a glass microinjection pipette (tubeonly) or between a lipid bilayer vesicle and a vesicle that is attached to the tip of a glass microinjection pipette (twovesicle). For the tubeonly configuration (TOC), a micropipette is used to pull a LNT into the interior of a surfaceimmobilized vesicle, where the length of the tube L is determined by the distance of the micropipette to the vesicle wall. For the twovesicle configuration (TVC), a small vesicle is inflated at the tip of the micropipette tip and the length of the tube L is in this case determined by the distance between the two interconnected vesicles. An electrochemical method monitoring diffusion of electroactive molecules through the nanotube has been used to determine the radius of the nanotube R as a function of nanotube length L for the two configurations. The data show that the LNT connected in the TVC constricts to a smaller radius in comparison to the tubeonly mode and that tube radius shrinks at shorter tube lengths. To explain these electrochemical data, we developed a theoretical model taking into account the free energy of the membrane regions of the vesicles, the LNT and the high curvature junctions. In particular, this model allows us to estimate the surface tension coefficients from R(L) measurements.
Background
Membrane tethers have been studied extensively over the past 40 years [1–11]. These structures, also called membrane nanotubes, were observed during fluid shear deformation of live cells attached to a substrate. As these cells were dislodged, membranous tethers remained attached to the surface displaying both the fluid and the elastic properties of the membrane [1, 2]. Following this work many naturally forming membrane nanotubes have been identified [7–10]. For example, membrane nanotubes have been shown to exist within the cell, notably in the trans golgi network [10]. Here, lipid and protein cargo destined for various destinations throughout the cell are sorted and pinched off from the tubular membrane of the network. It has also been reported that cells have the ability to use membrane nanotubes for the exchange of organelles [7], and this exchange has interestingly even been recognized between different cell types [8]. Thus, these tethers, which were first observed following a dramatic manipulation, have been shown to be a common occurrence in biology.
Additionally, by inflating a small ("daughter") vesicle at the tip of the micropipette, the diameter of a nanotube placed in between the inner vesicle and membrane of the outer vesicle can be examined in a configuration here called twovesicle configuration (TVC) (see Figure 1A). The measurements show that there is a reduction in tube diameter at shorter length, and the effect appears to be more pronounced in the TVC. In this work we suggest a geometrical model based on direct minimization of the Helfrich's functional for the system of lipid vesicles linked to a LNT via junctions of specific geometry. This new model presents a unified quantitative analysis of TOC and TVC and explains why the length of the LNT in the TVC is twice as high as in TOC for a given radius. Furthermore, the model has just two parameters, which can be chosen to fit the experimental data on monitoring of the size of the LNT. This allows for identifying the contribution of the surface tension to the free elastic energy of the system. This lowtension term has been neglected in the related publication [11], where a phenomenological description of the system was suggested and only a qualitative consistency with experimental data was obtained.
Experimentals
Materials and methods
Surfaceimmobilized giant unilamellar soybean liposomes (SBL) were made from soybean polar lipid extract (Avanti Polar Lipids, Alabaster, AL), as previously described [5, 6, 11, 15]. An injection pipette pulled with a commercial pipette puller (Model PE21, Narishige Inc., London, UK) and was backfilled with a 50 mM catechol solution. The pipette was then electroinserted into the unilamellar liposome with the aid of a voltage pulse generated relative to a 5 μ m counter electrode (ProCFE from Dagan Corp, Minneapolis, MN), which was placed on the opposite side of the liposome from the injection pipette. Carbon fiber working electrodes were fabricated in house and have been described elsewhere [11]. Working electrodes were held at +800 mV versus a silver/silver chloride reference electrode (Scanbur, Sweden). All measurements were made using an Axon 200B potentiostat (Molecular Devices, Sunnyvale, CA).
Nanotube radius measurements and calculations
where R is the radius of the nanotube of a given length L, Δi is the change in measured current with respect to the background, n is the number of moles of electrons transferred per mole of redox species (for catechol, this is equal to 2), F ≈ 96 485.34 C/mol is Faraday's constant, D = 7.0 × 10^{  } ^{6} cm^{2}/s is the diffusion coefficient of the selected redox species (catechol). ΔC is the change in concentration of catechol over the nanotube length and is equal to the concentration of electroactive species in the pipette assuming that the concentration at the electrode surface is zero (in our experiments ΔC = 50 mM).
The results for the tube radius deduced from the simultaneous measurement of electrochemical current and the tube length by using formula (1) are presented in this study. In comparison with our previous publication [11], a wider range of the length L of the tube is presented for the TOC configuration.
Theoretical approach
The system under consideration
where f _{0} is the force needed to hold the tube of radius R _{0} at a fixed position [12]. However, it was shown that for lipid vesicles interconnected with LNTs, either pulled outward from the vesicle wall [5, 15] or inward into the vesicle interior [11, 17], the neck elements (the junctions between the lipid tube and the vesicle body) also contribute to the total free energy of the membrane. Below we consider a theoretical model based on the Helfrich functional to find the equilibrium shape of the membrane accounting for the junctions of the specific geometry. By comparing the results of numerical computations with experimental data, we are able to determine the tension in the LNT after fitting the experimental data with the geometrical model described below.
The geometrical model
When the inner vesicle or the junction between the micropipette and the nanotube is subjected to the translation movement along the LNT axis, the length of the tube is changed (increasing or decreasing its value in a controlled way, which can be monitored under the microscope). During these manipulations the radius of the tube adapts to minimize the Helfrich's free energy (2) with C _{0} = 0, as we neglect any contribution from spontaneous curvature.
The toroidal part of the surface can be parametrised by (5) due to translation invariance of the energy functional (4). The multiplier ν assumes the value 1 for TOC and 2 for TVC to represent both junctions.
An important feature of the proposed model is the asymptote (compared to (3)), which is present in all three relations (7), (8), and (9). As we increase L, the radius R grows and the energy of the cylindrical part of the surface becomes dominant over the energy of the toroidal junctions. Thus, in the limit case L → ∞, we obtain the junction free equilibrium value of R given by (3).
Fitting the parameters
The latter method is preferable when the relative measurement error for R is greater than the one for L.
Results and discussion
Assuming the well established value of bending modulus k = 10^{12} erg [16], the recalculated coefficients of the surface tension are found in the interval of σ ~ 0.010.02 dyn/cm. These tension values are much smaller comparing to the lipid molecular compressibility (100 dyn/cm) [18] but much larger than the critical surface tension for the instability of the membrane cylinder and "pearling" (10^{  } ^{5} dyn/cm for (DGDG/DMPC membrane) LNTs of radius R ~ 0.3  5 μ m found in [19] work) while comparable with the magnitude of the lateral tension (higher limit) for mutual adhesion of lecithine membranes ~10^{  } ^{4} erg/cm^{2} [20].
The small value of the junction radius corresponds to the strongly deformed state of the membrane. These small values should be considered as order estimates, since they are attributes of the assumed toroidal geometry of junctions. The real shape of these junctions is probably more complex and, thus, cannot be described by just two scalar valued parameters. Although freezefracture electron microscopy does not reveal bilayers with curvature less than 20 nm, the value r ~ 1.5 nm which is found from the model is similar to the radius curvature of small inverted pores (for example, it is known that phospholipids spontaneously form inverted membrane structures with the radius varying between 0.5 and 5 nm, and smallest fusion pores have a calculated diameter less than 2.5 nm) [21, 22].
Conclusions
We propose a simple geometrical model for the quantitative explanation of the experimental results on equilibrium geometrical shape and LNTs parameters, R(L), in the different configurations. The experimental observations show that the nanotube diameter is reduced at shorter lengths and also that the diameter is consistently smaller for the TVC as compared to the TOC for a given length. The observed effect is ascribed to originate from the elastic junctions, since the phenomenon is accentuated in a system containing two necks connected to a vesicle membrane. We approximate the shape of these junctions by simple geometrical shapes and express the free elastic energy of the membrane in terms of the length of the LNT, its radius, the radius of the junction and the tension of the membrane. Variation of the energy with respect to the nanotube radius yields an explicit relation between the radius and the length. The relation is in agreement with observed values. The model enables estimations of the current surface tension coefficient and the curvature at junction regions. The estimated values of the surface tension are of order 10^{  } ^{2} dyn/cm and the curvature value at junctions are comparable to ones at fusion pores. Furthermore, the proposed model offers a clear explanation of the difference in measurements for TVC and TOC: in contrast to TOC, the TVC features two junction regions, thus, the length of the LNT in this configuration must be twice as long to achieve the same value of the radius.
Abbreviations
 DGDG:

digalactosyldiacylglycerol
 DMPC:

dimyristoylphosphatidylcholine
 LNT:

lipid nanotube
 SBL:

soybean liposomes
 TOC:

tubeonly configuration
 TVC:

twovesicle configuration.
Declarations
Acknowledgements
The authors are grateful to Prof. Sergei Rjasanow for the helpful discussion of the geometrical model. Part of this study supported by the German Academic Exchange Service (Deutscher Akademischer Austausch Dienst). ASC acknowledges support from the Swedish Research Council (VR) and the Knut and Alice Wallenberg Foundation. AGE acknowledges support from the European Research Council, VR and the USA National Institutes of Health.
Authors’ Affiliations
References
 Hochmuth R, Mohandas N, Blackshearjr P: Measurement of the elastic modulus for red cell membrane using a fluid mechanical technique. Biophys J 1973, 13: 747–762. 10.1016/S00063495(73)860217View ArticleGoogle Scholar
 Evans E: New membrane concept applied to the analysis of fluid shear and micropipettedeformed red blood cells. Biophys J 1973, 13: 941–954. 10.1016/S00063495(73)860369View ArticleGoogle Scholar
 Waugh R: Surface viscosity measurements from large bilayer vesicle tether formation. II. Experiments. Biophys J 1982, 38: 29–37. 10.1016/S00063495(82)84527XView ArticleGoogle Scholar
 Evans E, Bowman H, Leung A, Needham D, Tirrell D: Biomembrane templates for nanoscale conduits and networks. Science 1996, 273: 933–935. 10.1126/science.273.5277.933View ArticleGoogle Scholar
 Karlsson A, Karlsson R, Karlsson M, Cans A, Strömberg A, Ryttsén F, Orwar O: Networks of nanotubes and containers. Nature 2001, 409: 150–152. 10.1038/35051656View ArticleGoogle Scholar
 Cans A, Wittenberg N, Karlsson R, Sombers L, Karlsson M, Orwar O, Ewing A: Artificial cells: unique insights into exocytosis using liposomes and lipid nanotubes. Proc Natl Acad Sci USA 2003, 100: 400–404. 10.1073/pnas.232702599View ArticleGoogle Scholar
 Onfelt B, Davis D: Can membrane nanotubes facilitate communication between immune cells? Biochem Soc Trans 2004, 32: 676–678. 10.1042/BST0320676View ArticleGoogle Scholar
 Rustom A, Saffrich R, Markovic I, Walther P, Gerdes H: Nanotubular highways for intercellular organelle transport. Science 2004, 303: 1007–1010. 10.1126/science.1093133View ArticleGoogle Scholar
 Onfelt B, Nedvetzki S, Benninger R, Purbhoo M, Sowinski S, Hume A, Seabra M, Neil M, French P, Davis D: Structurally distinct membrane nanotubes between human macrophages support longdistance vesicular traffic or surfing of bacteria. J Immunol 2006, 177: 8476–8483.View ArticleGoogle Scholar
 de Matteis M, Luini A: Exiting the Golgi complex. Nat Rev Mol Cell Biol 2008, 9: 273–284. 10.1038/nrm2378View ArticleGoogle Scholar
 Adams K, Engelbrektsson J, Voinova M, Zhang B, Eves D, Karlsson R, Heien M, Cans A, Ewing A: Steadystate electrochemical determination of lipidic nanotube diameter utilizing an artificial cell model. Anal Chem 2010, 82: 1020–1026. 10.1021/ac902282dView ArticleGoogle Scholar
 Derényi I, Jülicher F, Prost J: Formation and interaction of membrane tubes. Phys Rev Lett 2002, 88: 238101.View ArticleGoogle Scholar
 Bo L, Waugh R: Determination of bilayer membrane bending stiffness by tether formation from giant, thinwalled vesicles. Biophys J 1989, 55: 509–517. 10.1016/S00063495(89)828449View ArticleGoogle Scholar
 Cuvelier D, Derényi I, Bassereau P, Nassoy P: Coalescence of membrane tethers: experiments, theory, and applications. Biophys J 2005, 88: 2714–2726. 10.1529/biophysj.104.056473View ArticleGoogle Scholar
 Karlsson M, Sott K, Cans A, Karlsson A, Karlsson R, Orwar O: Micropipetteassisted formation of microscopic networks of unilamellar lipid bilayer nanotubes and containers. Langmuir 2001, 17: 6754–6758. 10.1021/la0108611View ArticleGoogle Scholar
 Helfrich W: Elastic properties of lipid bilayers: theory and possible experiments. Z Naturforsch C: Biochem Biophys Biol Virol 1973, 28: 693–703.Google Scholar
 Michalet X, Bensimon D, Fourcade B: Fluctuating vesicles of nonspherical topology. Phys Rev Lett 1994, 72: 168–171. 10.1103/PhysRevLett.72.168View ArticleGoogle Scholar
 Evans E, Rawicz W: Entropydriven tension and bending elasticity in condensedfluid membranes. Phys Rev Lett 1990, 64: 2094–2097. 10.1103/PhysRevLett.64.2094View ArticleGoogle Scholar
 BarZiv R, Moses E: Instability and "pearling" states produced in tubular membranes by competition of curvature and tension. Phys Rev Lett 1994, 73: 1392–1395. 10.1103/PhysRevLett.73.1392View ArticleGoogle Scholar
 Servuss R, Helfrich W: Mutual adhesion of lecithin membranes at ultralow tensions. J Phys 1989, 50: 809–827.View ArticleGoogle Scholar
 Nanavati C, Markin V, Oberhauser A, Fernandez J: The exocytotic fusion pore modeled as a lipidic pore. Biophys J 1992, 63: 1118–1132. 10.1016/S00063495(92)81679XView ArticleGoogle Scholar
 Sackmann E, Merkel R: Lehrbuch der Biophysik. Berlin: WileyVCH; 2010.Google Scholar
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.