Quantifying intermolecular interactions as a basis of domain formation in membranes

Within the last two decades the field of membrane biology has witnessed an
increased interest in the function and organization of membrane lipids with a particular
focus on the possibility of these to demix into separate domains. The present
thesis aimed at providing quantitative information about intermolecular interactions
that may be responsible for the formation of such lipid domains in membranes.
Vesicular lipid model systems mimicking the composition of the plasma membrane
were biophysically characterized by means of modern microcalorimetric techniques
as a function of temperature and in the presence (or absence) of detergents.
For the formation and/or existence of one specific type of lipid domain, so called
lipid rafts, that are under intense scrutiny at present, cholesterol is reasoned to be of
paramount importance. To study differential interactions of cholesterol with different
lipids, three independent experimental assays for isothermal titration calorimetry
(ITC) in conjunction with a novel mathematical formalism to model these were
introduced. By means of reversible complexation with methylated–�–cyclodextrin
(cyd), sufficient amounts of the hydrophobic cholesterol molecule can be solubilized
in the aqueous phase. Thereby it became possible to study the thermodynamics of
either uptake of or release of cholesterol from lipid vesicles of various compositions.
As one important result a comprehensive set of quantitative data on cholesterol/lipid
interactions was obtained including for the first time also information on enthalpic
contributions to the differential interactions of cholesterol with different lipids.
Additionally, in these studies lipid/cyd interactions could be investigated and
suggestions on how to optimize cholesterol extraction from biological membranes
were made that could be derived from the different stoichiometries of the complexes
formed, i.e., lipid or cholesterol complexed to cyd, respectively.
The possibility to isolate detergent resistant patches is commonly used to argue
for the existence of (functional) domains in the original, detergent–free membrane.
This kind of reasoning does, however, neglect the possibility of detergent–induced
alteration or (in the worst case) induction of domains. In this context, a theoretical
model suitable to describe the selective solubilization of a membrane containing two
lipid domains (liquid ordered and liquid disordered) was developed. Based on equilibrium
thermodynamical relations it was shown that detergent–induced formation
of ordered membrane domains can occur if the detergent mixes nonideally with an
order preferring lipid and/or cholesterol. Furthermore, both the composition as well
as the mere existence of the liquid ordered domain was shown to be highly variable
upon addition of detergent to the membrane. A experimental study was carried out
in parallel to these theoretical simulations with the goal to better understand the mixing
of a commonly used nonionic detergent with different lipid/cholesterol systems.
In order to allow for a quantitative discussion of the experimental results obtained,
a theory for nonideal mixing in multicomponent lipid/detergent system was developed
that accounts for nonideality in terms of simple pair interaction statistics. The
parameters collected imply that a separation of ordered from disordered membrane
domains can under certain circumstances occur. A crucial parameter governing the
abundance and composition of detergent–resistant membrane patches appeared to
be the unfavourable interaction of cholesterol with detergent. Taken together, these
two studies provided additional evidence against the simple identification of lipid
rafts with detergent resistant membrane patches.
The third part of this thesis was devoted to a characterization of different phase
equilibria employing a rather new experimental technique, pressure perturbation
calorimetry (PPC). A micellar sphere–to–rod transition was characterized in terms
of a large set of structural, volumetric, and thermodynamic parameters including the
first published data on the change in partial molar volume of a detergent occurring
upon the transition. Subsequent to this study, the question whether binary mixtures
of an unsaturated lipid and cholesterol should be better described in terms of a phase
separation (liquid ordered and liquid disordered phases) or of gradual changes in
largely homogenous membranes was addressed with the help of PPC experiments.
The possibility of cholesterol to condense lipids not only laterally but also with respect
to volume was measured in this study for the first time. Information on the
number of condensed lipids per cholesterol were obtained by comparing the results
of simulations of expansivity curves according to three theoretical models appropriate
to be applied in this context. It was found that the behaviour of the binary mixtures
investigated is best described in terms of submicroscopic demixing rather than
true phase separation or random mixing.