Biology Reference
In-Depth Information
10
11
cm
2
s
1
. For
a similar sized, but water-soluble protein (hemoglobin), the diffusion rate was much faster,
at 7
1 second and a eukaryotic liver cell in 1 minute. The diffusion rate was 5
10
7
cm
2
s
1
. The effective membrane viscosity was therefore 7
10
7
cm
2
s
1
/
10
11
cm
2
s
1
or ~10
3
to 10
4
times more viscous than water. Viscosity of the membrane
bilayer interior therefore resembles that of light machine oil.
The diffusion rate for membrane proteins was shown to span a large range, from that
approaching free phospholipids (~10
8
cm
2
s
1
) through to essentially being immobile
(bound to the cytoskeleton, ~10
12
cm
2
s
1
). Interactions of many trans-membrane proteins
with the cytoskeleton account for protein immobility and are now well established. A first
indication of this occurred in the erythrocyte where removal of the cytoskeleton increased
band 3 lateral diffusion 40 times. Phospholipid diffusion was shown to be independent of
the head group (i.e. all phospholipids have about the same fast diffusion rate of ~10
8
cm
2
s
1
), but diffusion is impacted by membrane phase (discussed below and in Chapter
10). For example, lipid diffusion in gel state bilayers is ~100 to 1,000 times slower than in
fluid, liquid crystalline state bilayers. Lipid diffusion is also known to be impacted by the
presence of cholesterol, generating the liquid ordered (l
o
) state. This must be considered
with regard to the structure, stability, and function of l
o
-state lipid rafts. In general,
membrane components diffused 10 to 100 times slower in biological membranes than in
protein-free lipid bilayer membranes, implying hindered diffusion due to crowding (see
Chapter 11).
It is generally accepted that biological membrane components, lipids, and proteins, are
heterogeneously distributed into countless numbers of bewildering and short-lived domains.
An early experiment by Michael Edidin employing lateral diffusion measurements sup-
ported the concept of membrane heterogeneity
[26]
. Edidin's experiment used FRAP to
follow the diffusion of two carbocyanine dyes, one with two short lipid chains (C
10
,C
10
-
DiI) and one with two long lipid chains (C
22
,C
22
-DiI). The esterified fatty acyl chains have
very different T
m
s. C-10 decanoic acid has its T
m
at 31.6
C while the T
m
of C-22 behenic
acid is 79.9
C (see Table 4.3). At 37
C C-10 would be in the melted, liquid crystalline state
while C-22 would be in the solid, gel state. Edidin predicted that the short chain dye (C
10
,
C
10
DiI) would partition into a more fluid (disordered) phase while the longer chain dye
(C
22
,C
22
DiI) would prefer a more ordered phase. If the membrane lipid bilayer was homo-
geneous, he reasoned, both dyes should have similar diffusion rates, but if the membrane
bilayer was heterogeneous, the two fluorescent probes should exhibit different diffusion
rates. When he tested this hypothesis on sea urchin and mouse eggs, he found large differ-
ences between the diffusion rates for each dye. His conclusion was that membrane bilayers
have considerable patchiness that must exist for at least a few minutes. The nature of this
patchiness is at the heart of membrane domain studies.
5
E.
LIPID TRANS-MEMBRANE DIFFUSION (FLIP-FL
OP)
Earlier in this chapter it was discussed that trans-membrane asymmetry is always absolute
for proteins and carbohydrates. Therefore their rate of trans-membrane diffusion or flip-flop
is zero. It never happens. However, lipids do exhibit partial lipid asymmetry and so one
would assume that lipid flip-flop might be possible. However, if flip-flop was very fast, lipid
