Molecular Membrane Biology, 1995, 12, 89-91
Cellular plasma membrane domains
Michael P. Sheetz
Department of Cell Biology, Box 3709, Duke University
Medical Center, Durham, NC 27710, USA
Summary
The plasma membranes of migrating cells differentiate into at
least three distinct domains as definedby the laser tweezers and
the motile behaviour of particles bound to specific membrane
glycoproteins. These domains are important for steps in the cell
migration process. First, there is a domain at the leading edge
of the lamellipodium where preferential attachment of cross-
linked glycoproteins to the cytoskeleton occurs. The second
domain at the rear of the cell is differentiatedfor releasing from
Mol Membr Biol Downloaded from informahealthcare.com by Mcgill University on 11/03/14
substrata and shows decreased support of the membrane by the
cytoskeleton. The third domain is the highly curved region(s) of
the plasma membrane wherein certain membrane glycoproteins
concentrate and is a site for controlling extension and attach-
ment. Using single particle tracking and video analysis we find
that the quantitative differences between plasma membrane
domains are in the 4-20-fold range at any given time. These
values are consistent with the rapid fluctuations seen in cell
migration rates and directions. Over a longer time-scale, cells
can possibly integrate these selective advantagesto give a much
higher overall fidelity for cell chemotaxis and neuronal path
finding.
For personal use only.
Keywords: Cell migration, adhesion, single particle tracking,
integrims. membrane curvature.
Introduction
A multitude of different domains could form within biological
membranes and many of the possible ones are discussed
in this volume. Of particular importance are the domains that
nature has chosen to use in cellular functions. This paper
discusses the domains that we have observed in migrating
fibroblasts. Figure 1 describes the characteristic morphology
of such cells, and shows the major features of the endoplasm
(rich in microtubules, intermediate filaments and vesicular
organelles), ectoplasm (rich in actin and supports the plasma
membrane), leading lamellipodium, filopodia and retraction
fibres at the back of the cell. The process of cell migration
has been reviewed recently (Cooper 1991, Pollard et a/. 1991,
Fechheimer and Zigmond 1993, Sheetz 1994). Cell migration
can be described as a five-step process that involves: (1) the
extension of the leading lamellipodium,(2) attachment to the
substrate, (3) contraction to bring the body of the cell forward,
(4) release from attachment sites, and (5) recycling of
membrane components. Of these, the recycling step is the
least understood and whether or not substrate attachment
proteins do recycle through an endocytic mechanism
(Bretscher 1989) or by surface movement (Regen and Horwitz
1992) is still not established. We have used the optical
tweezers to probe the different regions of the cell with small
latex spheres of 0-1-0-5 pm diameter and have employed
video tracking methods to follow particle movements with a
precision of 5-10 nm per frame (Gelles eta/. 1988). The
advantage of the laser optical tweezers is that it allows one
to apply a known force to a specified region of the cell. The
tracking algorithms can give apparant diffusion coefficients
0968-7688/95 510.00 0 1995 Taylor a Francis
Figure 1. A typical fibroblastic cell in which the regions of the cell
that are discussed in this article are defined.
of beads and the mode of behaviour of the beads (Sheetz
eta/. 1989, Qian et a/. 1991, Schmidt et a/. 1993). While the
particle is still in the tweezers, tracking can give the force
applied to the particle from its position within the trap (Kuo
and Sheetz 1993).
Attachment at the leading edge
A classic observation in fibroblast migrationwas the rearward
transport of carbon particles from the leading edge of the cell
to the nuclear region (Abercrombie et a/. 1970). This rearward
transport process is constant in all motile fibroblasts even
though in some cases it is not coupled with migration. The
fact that all of the surfacebound particles move rearward has
led to the suggestion that the membrane was moving
rearward (Bretscher 1984); however, recent studies have
shown definitively that the membrane is passive and the
particles moving rearward are attached to the cytoskeleton
(see reviews by Heath and Holifield 1991, She& 1994)). It
is reasonable to postulate that in rearward transport the cell
is pulling rearward on external attachments to move itself
forward (Kucik et a/. 1991, Sheetz 1994). If this is true as we
believe, then the maximal migration benefit would come from
pulling on attachments formed at the leading edge. In the
three systems we have studied there is a much greater
attachment of crosslinked glycoproteins to the cytoskeleton
at the leading edge than at the rear or even a couple of
microns back from the leading edge (Kucik et a/. 1991,
Schmidt eta/.,1993, Schmidt et a/.,1995). In the keratocyte
(fish scale cells that migrate up to 40 pmlmin or 10-40 times
faster than a typical fibroblast), attachment is clearly rstricted
to a very small region within 1 w of the actual leading edge.
Extension of the leading edge requires actin filament
assembly, which from biochemical studies also occurs at
the leading edge (Symons and Mitchison 1991). Thus, both
the processes of extension and attachment appear to be
localized to the leading edge and could be biochemically
co-ordinated.
Control of both extension and attachment appears to
depend upon the crosslinkingof membrane glycoproteins by
specific extracellular matrix ligands. Substrata that both inhibit
and stimulate migration have been described (Lauffenburger
 90 M. P. Sheetz
1991). Specific receptors which are present at the leading
edge of the cell have been involved in making specific
attachments to those substrata. lntegrins have been
implicated in controlling motility and the combination of the
( ~ 1P1
, integrins is thought to interact directly with laminin in
stimulating the forward migration of cells. Recently, the
attachment of crosslinked 131 integrins to the cytoskeleton at
the front of the cell was shown to depend upon the
cytoplasmic domain of pl, and further that modification of
a single amino acid to mimic a phosphorylation in the
cytoplasmic domain could block preferential attachment at
the front (Schmidt etal. 1993). The difference between front
and back was only seen in migrating cells. Thus, the cell
knows its front from its back in very significant biochemical
and physical ways, and to study the processes of attachment
Mol Membr Biol Downloaded from informahealthcare.com by Mcgill University on 11/03/14
and extension requires a position-specificprobing of the cell.
Release at the rear of the cell
In their analysis of cell migration, DiMilla and colleagues have
shown that the strength of attachment is correlated in a
bimodal way with the rate of migration (DiMilla et al. 1993).
Very weak attachments do not allow the cell to spread and
to migrate across the substratum, but very strong attachments
block mibration by not allowing the cell to release from that
substratum. Here the mechanisms whereby cells are released
For personal use only.
from the substrate attachments in the rear are considered.
There are two basic processes to effect release, one is to
reduce the affinity of the membrane receptor for ligand and
the second is to physically separate the cell from the
attachment site. Pulling on submicron membrane attach-
ments results in the formation of thin membrane tubes or
tethers that are commonly called retraction fibres and
normally appear at the rear of migrating cells (Figure 1). We
have observed that the focal application of force to the rear
of the cell is much more likely to pull a membranous tether
than at the front of the cell (Schmidt e t a / . 1993). The optical
tweezers provide a rapid and reliable way of applying a focal
force to different regions of the cell using latex spheres that
are attached either by non-specific or by specific ligands to
the cell surface. Thus it should be possible to probe this
important membrane property to understand the molecular
basis of regional changes in the membrane cytoskeleton that
lead to the greater ease of release at the rear of the cell.
Concentration at curved edges
The final process that will be described is the concentration
of membrane glycoproteins at curved edges. Although this
phenomenon may be most relevant at the front of the cell
to concentrate receptors for environmental cues where they
will be first encountered, it can be generally employed
wherever glycoproteins need to be concentrated or depleted.
This phenomenon was first noted in studies of the diffusion
behaviour of the early membrane antigen, EMA (Baumrind
etal. 1992) on cortical neuron growth cones. Antibody-coated
gold particles were concentrated at the edge of the growth
cone as was the antigen, and, surprisingly, the particles were
diffusing rapidly along the edge (Sheetz e t a / . 1990). Over
a 30-s period, particles would briefly diffuse away from the
edge but would often return to diffuse along the edge. Both
the average distribution of particles and the distribution of
single particles over time indicated that the antigen preferred
the edge 10-20-fold over the flat regions of the growth cone.
Upon analysing the movements of the particles along the
edge, the apparent diffusion coefficient ( - l o - ’ cm2/sec)
was unaltered from that of particles diffusing over flat regions.
Thus, there is no fixed binding site that holds them at the
edge such as the synaptic receptor complexes. Alternatively,
the membrane protein may well have a lower energy state
in a curved region of membrane. In the case of EMA there
was a 10-20-fold concentration of that antigen at the edge
meaning that there was roughly a 2-4 kT energy difference
for that protein between the flat region of the lamellipodium
and the curved edge (Figure 2). This is not a great differernce
in energy for a relatively large protein and there could well
be a simple structural characteristic of such proteins that
would favour their being placed within curved regions. We
have found a similar behaviour for the p1 integrin, however
concentration in edge regions was not dependent upon the
cytoplasmic tail suggesting that the (Y integrin or the external
or membrane regions of the integrin dimer were responsible
for edge concentration. Specific antibodies attached to gold
or small latex particles and the methodologyof single particle
tracking allow one to monitor such movements along curved
edge regions. There are really two methods to be used. One
is the relative concentration of particles at the edge versus
the flat regions of the cell. That is often a somewhat
misleading number in the case of the integrins. They rapidly
attach to the substratum and move away from the leading
edge causing a depletion from that leading region. However
one can analyse the diffusion time along the edge or the
probability of leaving the edge and going onto the flat region
of the membrane to obtain an estimate of the energy
difference at the edge. The two major functions of edge
concentration are the placement of control molecules such
as EMA at the point where they will first encounter new
substrate signals and the retention of substrate attachment
molecules, such as integrins, at the leading edge for maximal
benefit in migration.
Conclusions
We have here described three different domains of cellular
plasma membrane that are largely defined by the biochemical
organization of the cytoskeleton and associated proteins. In
AG = -2-4 kT
Figure 2. The energy differences that are expected for a freely
diffusing glycoprotein that is concentrated at edge regions.
 Cellular plasma membrane domains
two cases, i.e. the front versus the rear, these have important
physical parameters associated with them because the
attachments at the front are used to generate forces in the
nanonewton range on substrata. The tethers in the rear
likewise are generating significant forces on the substrata
anywhere from tens to hundreds of piconewtons. The edge
concentration of proteins consitutes a physical separation of
membrane glycoproteins forming a domain within the plain
of the membrane. All of these domains depend heavily upon
the cytoskeleton which in turn is controlledthrough processes
initiated at the plasma membrane. The quantitative
differences between front versus rear of motile cells or flat
versus curved membranes are only 4-20-fold at any time. For
the cells to make decisions with high fidelity to follow specific
paths during development or to respond to signalling
molecules, additional factors are needed such as a temporal
Mol Membr Biol Downloaded from informahealthcare.com by Mcgill University on 11/03/14
integration of signals. With the laser tweezers and single
particle tracking techniques we have the ability to now probe
different cellular domains in an attempt to gain a further
understanding of these processes.
Acknowledgement
This review was in part supported by NIH grants and a grant from
the Human Frontiers in Science Program.
For personal use only.
References
Abercrombie, M.. Heaysman, J. and Pergrum, S. (1970) The
locomotion of fibroblasts in culture. 111. Movements of particles on
dorsal surface of the leading lamella. Experimental Cell Research,
60, 437-444.
Baumrind, N. L., Parkinson, D., Wayne, D. B., Heuser, J. E. and
Pearlman, A. L. (1992) EMA: a developmentally regulated cell-
surface glycoprotein of CNS neurons that is concentrated at the
leading edge of growth cones. Developmental Dynamics, 194,
311-325.
Bretscher, M. S. (1984) Endocytosis: relation to capping and cell
locomotion. Science, 224, 681-686.
Bretscher. M. S. (1989) Endocytosisand recycling of the fibronectin
receptor in CHO cells. EMBO Journal, 8, 1341-1348.
91
Cooper, J. A. (1991) The role of actin polymerization in cell motility.
Annual Review of Physiology, 53, 585-605.
DiMilla, P., Stone, J., Abelda, S., Quinn, J. and Lauffenburger, D.
A. (1993) Maximal migration of human smooth muscle cells on
fibronectin and collagen type I V occurs at an intermediate
adhesiveness. Journal of Cell Biology. 122, 729-737.
Fechheimer, M. and Zigmond, S . H. (1993) Focusing on
unpolymerized actin. Journal of Cell Biology, 123. 1-5.
Gelles, J., Schnapp, B. J. and Sheetz, M. P. (1988) Tracking kinesin-
driven movements with nanometer-scale precision. Nature, 331,
450-453.
Heath, J. P. and Holifield, B. F. (1991) Cell locomotion: new research
tests old ideas on membrane and cytoskeletal flow. Cell Motility
and Cytoskeleton, 18, 245-257.
Kucik, D. F., Kuo, S. C.. Elson, E. L. and Sheetz, M. P. (1991)
Preferentialattachment of glycoproteins to the cytoskeleton at the
leading edge of lamella. Journal of Cell Biology, 115, 1029-1036.
Kuo. S. C. and Sheetz, M. P. (1993) Force of single kinesin molecules
measured with optical tweezers. Science, 260, 232-234.
Lauffenburger, D. A. (1991) Models for receptor mediated cell
phenomena: adhesion and migration. Annual Review of
Biophysical Chemistry, 20, 387-414.
Pollard, T. D., Doberstein, S. K. and Zot, H. (1991) Myosin I in cell
motility. Annual Review of Physiology, 53, 653-680.
Qian, H., Sheetz, M. P. and Elson, E. (1991) Analysis of diffusion
and flow in two-dimensional systems. Biophysical Journal, 60,
910-921.
Regen, C. M. and Horwitz, A. F. (1992) Dynamics of B1 integrin-
mediated adhesive contacts in motile fibroblasts. Journal of Cell
Biology, 119, 1347-1359.
Schmidt, C. E., Dai, J., Lauffenburger, D. A., Sheetz, M. P. and
Horwitz, A. F. (1995) Integrin-cytoskeletalinteractions in neuronal
growth cones. Journal of Neuroscience (In press).
Schmidt, C. E., Horwitz, A. F., Lauffenburger, D. A. and Sheetz, M.
P. (1993) Integrin-cytoskeletalinteractions in migrating fibroblasts
are dynamic, asymmetric, and regulated. Journal of Cell Biology.
123, 977-992.
Sheetz, M. P. (1994) Cell migration by graded attachment and
contraction. Seminars in Cell Biology, 5 , 149-155.
Sheetz, M. P., Turney, S., Qian. H. and Elson, E. L. (1989)
Nanometer-level analysis demonstrates that lipid flow does not
drive membrane glycoprotein movements. Nature, 340, 284-288.
Sheetz, M. P., Baumrind, N. L., Wayne, D. S. and Pearlman, A. L.
(1990) Concentration of membrane antigens by forward transport
and trapping in neuronal growth cones. Cell, 61, 231-241.
Symons, M. H. and Mitchison, T. J. (1991) Control of actin
polymerization in live and permeabilized fibroblasts. Journal of Cell
Biology, 114, 503-51 3.