Chapter 8. Morphology of vesicles

U. Seifert
Institut für Festkörperforschung, Forschungszentrum Jülich,
52425 Jülich, Germany

R. Lipowsky
Max-Plank-Institut für Kolloid- und Grenzflächenforschung,
Kantstr. 55, 14153 Teltow-Seehof, Germany

1. Introduction: the shapes of vesicles

All biological cells contain complex mixtures of macromolecules not present in their environment. The 'container' of these cells is provided by the plasma membrane. This membrane forms a closed bag which is essentially impermeable for the macromolecules. Likewise, many organelles or compartments within the cells are surrounded by additional membranes which ensure that different compartments can have a different composition of molecules. The biomembranes of all present-day cells are composed of lipids and amphiphilic proteins. In fact, the basic and universal building block of all biomembranes seems to be a lipid bilayer which is 'decorated' by proteins.

Biomembranes exhibit a rather complex morphology [1]. The surface formed by intracellular organelles often have a very complicated topology such as those of the endoplasmatic reticulum or the Golgi apparatus. Likewise, the plasma membrane can develop small buds as in exo- or endocytosis, microvilli, i.e. finger-like protrusions, and pseudopods. These dynamic shape transformations are used in cell locomotion. Some aspects of these morphological transformations can be studied with lipid vesicles which are closed bags formed spontaneously from lipid bilayers in aqueous solutions. These vesicles provide the simplest model system for the formation of distinct compartments in biological systems. They can conveniently be studied with video microscopy.

The current theoretical understanding of the morphology and morphology transformations of vesicles is based on the important notion of bending elasticity introduced some twenty years ago by Canham [2], Helfrich [3, 4] and Evans [5]. However, a systematic analysis of these so-called curvature models was performed only recently [6-9]. These studies revealed a large variety of shapes which minimize the energy for certain physical parameters such as the enclosed volume and the area of the vesicle. These shapes are then organized in so-called 'phase diagrams' in which trajectories predict how the shape transforms as, e.g., the temperature is varied.

As an example for this approach, see fig. 1, where the most prominent shape transformation, the budding transition, is shown as it has been observed by video microscopy [7, 10, 11]. An increase in temperature transforms a quasi-spherical vesicle via thermal expansion of the bilayer to a prolate shape and then to a pear. Finally, a small bud is expelled from the vesicle. Upon further increase of temperature several additional buds can emerge. Usually, the buds remain connected to the mother vesicle via narrow constrictions or necks. Another shape transformation, the discocyteÐstomatocyte transition, is shown in fig. 2. This transition resembles a shape transformation which can be induced with red-blood cells by depletion of cholesterol.

Shape transformations are also predicted to arise in vesicles consisting of a bi-layer with several components due to a different mechanisms. If the two components form domains, the line energy associated with these domain boundaries can be reduced by budding of such a domain [12]. Even if the membrane is in the one-phase region, a temperature induced budding process should lead to curvature-induced phase segregation since, in general, the two components couple differently to the local curvature [13]. So far, these effects are not yet verified experimentally even though budding and fission have been observed in vesicles consisting of lipid mixtures [14].

Qualitative new behavior has been predicted for vesicles with non-spherical topology, i.e. vesicles with holes or handles. For tori, i.e. vesicles with one hole or handle, the shape of lowest energy is non-axisymmetric for a large range of parameters [15, 16]. For shapes with at least two holes or handles, one even predicts that the shape of lowest energy is no longer unique. This degeneracy leads to the phenomenon of 'conformal diffusion' in shape space [17]. The simplest examples of artificial vesicles with non-trivial topology are shown in fig. 3 [18, 19].

The interaction of vesicles with substrates (or other vesicles) also effects their shape. For weak adhesion, an adhesion transition driven by the competion between bending elasticity and adhesion energy is predicted [20, 21]. For strong adhesion, the notion of an effective contact angle becomes applicable [20]. Experimentally, the adhesion of vesicles can be studied either with the micro-pipet aspiration technique [22] or by reflection interference microscopy [23, 24]. The latter technique allows very precise length measurements and, thus, to deduce the shape of adhering vesicles. In this way, a flat pancake can be distinguished from a quasi-spherical shape which is only slightly distorted by the adhesion to the substrate.

In this chapter, we describe the present understanding of these phenomena with an emphasis on the theoretical aspects. However, we make an attempt to link the theory with the experiments where this is possible. Various aspects of the work presented here have also been treated in recent reviews, see refs. [25-27].
 
 

Fig. 1. Budding transition induced by temperature which is raised from 31.4 o C (left) to 35.8 o C (right) in this sequence. The theoretical shapes have been obtained within the bilayer-couple model assuming an asymmetric thermal expansion of the two monolayers [7].

Fig. 2. Discocyte-Stomatocyte transition induced by temperature which is raised from 43.8 o C (left) to 44.1 o C (right) in this sequence [7].

Fig. 3. Vesicles of non-spherical topology. (a) a non-axisymmetric torus, (b) an axisymmetric torus, and (c) a 'button' shape [18, 19].

 
  [Full text] (PDF 1977 Kbytes)