Radiol Oncol 1998; 32(1): 119-124
On mechanisms of cell plasma membrane vesiculation
Veronika Kralj-Iglič1, Urška Batista1, Henry Hagerstrand2, Aleš Iglic1'3, Janja Majhenc1, Miha Sok4
1Institute of Biophysics, Medical Faculty, Ljubljana, Slovenia, 2Department of Biology, Abo Akademy University, Abo/Turku, Finland, 3Faculty ofElectrical Engineering, Ljubljana, Slovenia, 4Department ofThoracal Surgery, Clinical Center, Ljubljana, Slovenia
Vesiculation ofthe celi membrane is studied. It is proposed that the shape ofthe membrane segment free of the cytoskeleton is driven towards the shape of its maximal possible difference between the two membrane layer areas by the rearrangement of the laterally mobile membrane constituents. It is shown that the shapes corresponding to the extrema of the area difference can be spherical, planar and cylindrical, depending on the enforced constraints. Correspondingly, the spherical vesicles and the cylindrical protrusions observed in vesiculating MCF7 cancer (human breast adenocarcinoma) cells are shown. The proposed mechanism ofvesiculation also provides an explanation far different relative content of some substances in the membrane of the released vesicles than in the membrane of the residual cells.
Key words: budding; membrane bilayer; vesiculation
Introduction
Membranes of some cells can form during the budding process small protrusions which are eventually released from the membrane as vesicles. The amount of the involved membrane varies from relatively large portions with or without enclosed elements of the cytoplasmic material to very small fragments filled only with the cytosol. It is a common feature that the disruption of the cytoskele-ton or its detachment from the membrane bilayer occurs prior to vesiculation. It was
Correspondence to: dr. Veronika Kralj-Iglič, Institute of Biophysics, Medical Faculty, Lipičeva 2, SI-1000 Ljubljana, Slovenia; Tel.: +386-61-314-127; Fax.: +386-61131-51-27; E-mail: vera.kralj-iglic@biofiz.mf.uni-lj.si.
also observed that the membrane of the released vesicles differs from the membrane of the residual cell in the relative content of some membrane constituents, indicating that the rearrangement of the membrane constituents occurs during the budding. The vesiculation is therefore a mechanism through which the cell membrane looses some substances. In cancer cells the budding and vesiculation process occurs spontaneously thereby causing a continuous loss of some important substances from the cell membrane and leading to the alteration of the cell function.1,2,3
A system that is due to its simplicity convenient to study the general features of the membrane budding and vesiculation is the
120
Kralj-Iglič Vet al.
mammalian erythrocyte. In erythrocytes, it was found that a wide variety of treatments and conditions such as incubation with dimyristoyl phosphatidyl choline vesicles,4'5 incubation with various amphiphiles,6'7 ATP depletion8 and extreme pH in the suspension9'10 may influence the vesiculation process. As the features observed in vesicula-tion of the erythrocytes are relevant also in general11, there is a possibility to manipulate the cancer cells in such a way as to stabilize the membrane and prevent the loss of the important membrane constituents from the membrane. It is therefore of interest to understand the mechanisms taking place in membrane budding and vesiculation. In this work we focus on the features involving the membrane segments that are already detached from the cytoskeleton.
Material and methods
The proposed mechanism of budding and vesiculation
The proposed mechanism is schematically represented in Figure l.
In the description of the membrane segment the membrane is taken to be a two dimensional liquid composed of phospholipid molecules, in which various other molecules such as the membrane proteins are embedded. The embedded molecules are more or less free to move laterally over the membrane surface. The connections of the membrane with the intracellular and the extracellular matrix are of importance, since they may impose limits and obstacles for the lateral motion of the membrane constituents. The disruption or detachment of the cytoskeleton from the membrane thus increases the pool of the laterally mobile molecules.
It is proposed that following the detachment of the cytoskeleton, the development of
membrane
A
embedded molecule
cytoskeleton
o
Fi^^e l. A scheme of the proposed mechanism of the budding process. The membrane with embedded molecules that favour Iarge membrane curvature is shown. First, the membrane becomes Iocally detached from the cytoskeleton thereby increasing the pool of Iaterally mobile embedded molecules. The membrane segment free of the cytoskeleton then forms a protrusion as this is energetically favourable. The Iaterally mobile membrane embedded molecules accumulate in the region of favourable curvature so that the area density of the number of the embedded molecules is higher in the membrane of the vesicles than in the membrane of the residual celi.
the bud can be described as a local event involving the cytoskeleton free membrane segment. Further, it is proposed that the exchange of the laterally mobile membrane constituents between the membrane segment and the membrane of the residual cell provides the driving mechanism for the bud development, ending with the formation of the vesicle.
As the cytoskeleton free membrane segment is very small comparing to the membrane of the residual cell, the membrane of the residual cell can be treated as a reservoir for the laterally mobile membrane con-
On mechanisms ojcell plasma membrane vesiculation
121
stituents. The energy of the embedded molecule at a given site in the membrane depends on the local membrane curvature,12 so that the laterally mobile molecules tend to accumulate in the regions at which the membrane curvature is energetically more favourable while the membrane segment attains the shape with larger regions of the favourable curvature.12 As a result, a protrusion of the membrane exhibiting higher curvature than the residual cell is formed and the laterally mobile membrane constituents that favour large curvature accumulate in the protrusion.
In turn, the molecule that favours large curvature may be expected to occupy larger portion of the outer membrane layer than the inner one. A presence of such a molecule in the membrane segment increases the area of the outer layer of the segment with respect to the area of the inner one. If the cytoskeleton free membrane segment forms a bud, the laterally mobile molecules that favour large membrane curvature flow to the bud from the membrane of the residual cell and cause the area of the outer layer of the segment to increase relative to the inner one. The process may proceed up to a limit imposed by the geometrical constraints where the shape of the maximal possible area difference is reached.
Theory
In order to obtain the shapes of the bilayer segment of an extreme area difference M at a given area of the bilayer neutral surface A a variational problem is stated by constructing a functional
G = M-Ja (fdA-A) ,	(1)
where for thin bilayers
M = hf (Cj + C2)dA,	(2)
is the Lagrange multiplier, Cj and C2 are
the principal membrane curvatures, M is the area element and h is the distance between the neutral surfaces of the two bilayer monolayers in the direction perpendicular to the membrane surface. The analysis is restricted to axisymmetric shapes. It is chosen that the symmetry axis of the body coincides with the x axis, so that the shape is given by the rotation of the function y(x) around the x axis. In this case the principal curvatures are expressed by y(x) and its derivatives with respect to x; y'=<9y/ax and y"= a2y/ax2, as C1=2/y(l+y'2j112 and C2 = -y"/(2+y'2)3/2 while the area element is M=2n (2+y'2)3/2 y dx. The sign of the principal curvatures is taken to be positive for a sphere. Inserting the above expressions for Cv C2 and M into (1) and rearranging, the functional normalized with respect to 2jth becomes
G = fg(x,y,y, y"\.x
(3)
where
g(x,y,y',y") = 1-yyy (2+y'2)V2, (4)
A^Apj/fc. The variation b G = O is performed by solving the Poisson - Euler equation
dg dgj + dL ( di)
dy dx ( dy'j + dx2\ \.uj
0 .
(5)
Obtaining the necessary differentiations of (4), the Poisson - Euler equation is expressed as
2y"/(î+y'2)2 + ((2+y'2j-V2 -yy"(l+y'2j-3/2j = O.(6)
If the area of the segment is fixed there is an analytical solution of (6), given by a circle of the radius rrir : y = - x2)1^2^ This solution represents spheres of a radius 2/rcir=}.,N and a segment of a plane 2/rcir=0. If the area A is not fixed i.e if = O the possi-
122
Kralj-Iglič Vet al.
ble analytical solution of the equation (6) is a constant y = const, representing a cylinder.
Experiment
Cells: The cells MCF7 (human breast adenocarcinoma) were grown in Eagle MEM, supplemented with 1 percent nonessential amino acids, 10 percent fetal calf serum (FCS), penicillin (100 U/ml) and streptomycin (100 Mg/ml) at 37°C in a CO2 incubator.
Induction of membrane vesiculation: Exponentially growing MCF7 cells were detached by 0.25 percent trypsin solution. The cells were resuspended in Eagle MEM without FCS and put on ice (4°C). After one hour the cell suspension was placed on 37°C for two hours.13 The cells were then observed by the phase contrast microscope (Obj. Ph 3, 100X, NA 1.2).
area of the membrane segment while the cylindrical shape is unconstrained, but confined at its radius.
The theoretical predictions are compared to the phenomena observed in the experiment. In a vesiculating cancer cell, spherical vesicles as well as cylindrical protrusions can be observed (Figs. 2A and 2B, respectively). The vesiculation was promoted as described in the Material and methods ensuring that the integrity of the cytoskeleton was destroyed.13 The observed shapes of the vesicles and protrusions correspond well to the theoretically predicted ones.
It was indicated by experiments that the properties of the membrane constituents strongly influence the nature of the protrusions and vesicles. In erythrocyte suspension, incubated with the exogenously added amphiphiles the budding and vesiculation of the erythrocyte membrane was observed.6'7
Results and discussion
It was shown above that the shapes of the bilayer segment corresponding to the extreme difference between the two monolayer areas are spherical, planar and cylindrical. The spherical shape and a planar circular segment are characterized by one parameter, respectively. The respective parameter can be determined from the constraint requiring a fixed area. Therefore the sphere and the planar segment can be established as the shapes of the extreme area difference.14 If the sphere is involved, the extremum is a maximum, as calculated by the minimization of the membrane bending energy for a sequence of shapes describing the formation of a spherical vesicle.15 In order to establish a cylindrical shape as a shape of the extreme area difference, a boundary condition should be stated, such as a requirement for a fixed radius of the cylinder. It can be concluded that the spherical and the planar shapes of the maximal area difference are connected to the fixed
HHRm^HHIfH^HNtahTHII
^^HHHR11
IfgaMEi.
illiiiMWBlgi ^11 IMBBBBBÎIHBBB
Fi^e 2. A vesiculating MCF7 cancer cell exhibiting a cylindrical protrusion (A) and a spherical vesicle (B), (bar=lO^m).
On mechanisms of cell plasma membrane vesiculation
123
The released vesicles were spherical and cylindrical, depending on the species of the added amphiphiles. Besides the properties of the membrane constituents, the manner of the disruption or the detachment of the skeleton may also influence the character of the vesicles by determining the amount of the membrane segment available to the bud, and also by determining the amount of the laterally mobile membrane embedded molecules. In deciding whether the protrusion will lead to a spherical vesicle or to a cylindrical protrusion, it should therefore be established which of the two processes is possible. If both of them are possible it should be distinguished, which of them would be energetically more favourable. In this case, the free energy of the segment under consideration should be minimized, taking into account the local composition of the segment. This is however beyond the scope of this work.
The laterally mobile molecules that favour large membrane curvature are accumulated in the buds which develop into vesicles and are released from the membrane. Thereby the relative content of these molecules in the residual cell membrane is diminished which may be unfavourable regarding the cell function. In this context, adding to the cell membrane the substances that decrease the area difference would tend to inhibit the budding and vesiculation. Indeed, chlorpro-mazine which was proved to intercalate into the inner membrane layer of erythrocytes16'17 was also shown to have therapeutic effects in cancer treatment.18 These promissing results encourage further studies of the mechanisms taking place in the membrane during budding and vesiculation.
References
cellular medium by human breast carcinoma cells carry tumor - associated surface antigens. Clin Exp Metastasis 1995; 13: 277-86.
2.	Hakomori S. Tumor malignancy defined by aberrant glycosylation and sphingo(glyco)lipid metabolism. Cancer Res 1996; 56: 5309 - 18.
3.	Black RA Rausch CT, Kozlosky CJ Peschon JJ, Slack JL, Wolfson MF et al. A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 1997; 385: 729 - 33.
4.	Ott P, Hope MJ, Verkleij AJ, Roelofsen B, Brodbeck U, Van Deenen LLM. Effects of dymiristoylphos-phatilylcholine on intact erythrocytes. Release of spectrin - free vesicles without ATP depletion. Biochim Biophys Acta 1981; 641: 79 - 87.
5.	Yamaguchi T Yamada S, Kimoto E. Effects of cross - linking of membrane proteins on vesiculation induced by dimyristoyl phosphatidylcholine in human erythrocytes. J Biochem 1994, 115: 659-63.
6.	Hägerstrand H, Isomaa B. Vesiculation induced by amphiphiles in erythrocytes. Biochim Biophys Acta 1989; 982: 179 - 86.
7.	Hägerstrand H, Isomaa B. Morphological characterization of exovesicles and endovesicles released from human erythrocytes following treatment with amphiphiles. Biochim Biophys Acta 1992; 1190: 409 -15.
8.	Lutz HU, Liu SC, Palek J. Release of spectrin-free vesicles from human erythrocyte during ATP depletion. J Cell Biol 1977; 73: 548 - 60.
9.	Bobrowska-Hägerstrand M, Iglic A, Hägerstrand H. Erythrocytes vesiculate at high pH. Cell Mol Biol Lett 1997; 2: 9 -13.
10.	Gros M, Vrhovec S, Brumen M, Svetina S, Zeks B. Low pH induced shape changes and vesiculation of human erythrocytes. Gen Physiol Biophys 1996; 15: 145 - 63.
11.	Wagner GM, Chiu DTY, Yee MC Lubin BH. Red celi vesiculation - a common membrane physiological event. J Lab Clin Invest 1986; 108: 315 - 24.
12.	Kralj-Iglic V, Svetina S, Zeks B. Shapes of bilayer vesicles with membrane embedded molecules. Eur Biophys J 1996; 24: 311 -21.
13.	Liepins A, Hillman AJ. Shedding of tumor celi surface membranes. Cell Biol Inter Rep 1981; 5: 15 -26.
l. Dolo V, Adobati E, Canevari S, Picone MA Vit-torelli ML. Membrane vesicles shed into the extra-
14. Elsgolc LE. Calculus of variations (1963) Pergamon Press, Oxford, pp. 79.
124
Kralj-Iglič Vet al.
15.	Zeks B, Iglic A, Svetina S. Bilayer membrane models and a theoretical analysis of the vesiculation process. Suppl Minerva Biotechnologica 1990; 2: 47.
16.	Deuticke B. Transformation and restoration of the biconcave shape of human erythrocytes induced by agents and change of ionic environment. Biochim Biophys Acta 1968; 163: 494 - 500.
17.	Sheetz MP, Singer SJ. Biological membranes as bilayer couples. A molecular mechanism of drug-erythrocyte interactions. Proc Natl Acad Sci 1974; 71: 4457 - 61.
18.	Levij IS, Polliack A. Inhibition of chemical carcinogenesis in the hamster cheek pouch by topical chlorpromazine. Nature 1970; 223: 1096.