Clustering of ErbB proteins
ErbB proteins belong to the epidermal growth factor (EGF) receptor family of receptor tyrosine kinases. The family has four members (ErbB1-4, also known as HER1-4), from which ErbB1, the founding member, is also called EGF receptor. ErbB1 is a receptor for EGF and other EGF-like ligands, whereas ErbB3 and ErbB4 are stimulated by different kinds of neuregulins (NRG, also known as heregulins, HRG). ErbB2 is a co-receptor for the other three members enhancing their signaling potency. According to the accepted theory, developed mainly for ErbB1, unstimulated, monomeric receptors dimerize upon growth factor binding, whose primary driving force is the conformational change induced by the ligand in the extracellular domain. The kinase domain of dimeric receptors is activated leading to transmembrane signaling.
Our work has led to the following discoveries:
- ErbB and other membrane proteins generate hierarchical clusters, among which dimers are only the smallest types
- EGF receptor (ErbB1) and ErbB2 behave substantially differently regarding their large-scale clustering properties and their response to growth factor stimulation (published paper)
The effect of the lipid environment of the cell membrane on the clustering and cell biological function of membrane proteins
The cell membrane cannot be considered a homogenous system from the standpoint of lipids or proteins since both types of molecules constitute assemblies of varying temporal and spatial stability. The hierarchical association of membrane proteins represent such a structure, but lipid microdomains also belong to them. Lipid rafts are such supramolecular organization of lipids, which are thermodynamically unstable, small (10-100 nm) structures. Proteins, lipids, the cytoskeleton and membrane turnover all contribute to their generation. They are similar, but not identical, in many respects to the liquid-ordered (Lo) domains in model membranes. The lipid environment of the cell membrane obviously influences the biophysical and cell biological properties of transmembrane proteins through their transmembrane domain. An important property of the cell membrane is the dipole potential, which is a positive potential of magnitude 200-500 mV in the interior of the cell membrane generated by the dipoles of lipids and membrane-associated water molecules. This electric field interacts with transmembrane protein (due to the dipole moment of their transmembrane domain), with ligands and molecules binding to cell surface receptors or being transported across the membrane.
We have established that
- the dipole potential significantly enhances the ligand-induced clustering and signaling of ErbB1 and ErbB2 (published paper)
- the dipole potential is biologically significantly larger in lipid rafts than in other parts of the membrane (published paper)
- increasing the sphingolipid content of the cell membrane enhances the size of liquid-ordered (Lo) domains and “traps” lipid and protein constituents residing in the liquid-disordered (Ld) domain inhibiting their mobility and function. Such an increase in the sphingolipid content of the cell membrane is present in Gaucher’s disease (published paper)
- the positive intramembrane dipole potential inhibits the cellular uptake of penetratin.
Penetratin is a cell-penetrating peptide having a potential role in efficient and targeted delivery of drugs. It enters the cytoplasm of cells either by direct membrane crossing or by endocytosis followed by traversing the membrane of the endo-lysosomal compartment (figure on the top). Due to the charged nature of penetratin, the positive intramembrane dipole potential inhibits in membrane crossing. Decreasing the dipole potential with atorvastatin treatment resulted in a significantly enhanced concentration of penetratin in the cytosol (figure at the bottom).
Medical implications of the cell membrane and membrane lipids
The cell membrane has immense medical importance since practically all drug molecules have their target in the membrane or must cross it if they have an intracellular binding site. In addition, the properties of membrane proteins are substantially influenced by the lipid composition of the cell membrane. These have medical implications for the following reasons:
- several diseases (e.g. lysosomal storage disorders, inherited or acquired dyslipidemias, malignant diseases) are accompanied by alterations in the lipid composition of the serum and the cell membrane, which impacts the functioning of membrane proteins (e.g. by mechanisms mentioned in point 2)
- the lipid composition of the cell membrane can be altered for therapeutical purposes, which can change the symptoms or course of several diseases.
We have established that
- sphingolipid accumulation, typical of Gaucher’s disease, also involving the cell membrane, induces several changes in the properties of the cell membrane, e.g. reduced fluidity, inhibited STAT signaling and endocytosis of non-raft proteins. Measurement of endocytosis was carried out using confocal microscopy as demonstrated by the figure below.
A – Labeling of the cell membrane with anti-CD14 monoclonal antibody (green) and DAPI (blue), which stains the nucleus. Red dots corresponds to nuclei. B – The cell membrane (blue) and the intracellular space (red) identified using the membrane and nuclear stains. C – Overlay of the fluorescence image on the membrane and intracellular masks.
- binding of trastuzumab (Herceptin), used in the therapy of ErbB2-overexpressing breast cancers, to the cell membrane is inhibited by overexpression of the MUC-4 mucin and overproduction of hyaluronan (published paper, published paper).
Methodological developments related to fluorescence labeling of cells and Förster resonance energy transfer
Förster resonance energy transfer (FRET) is often used in the investigation of protein clustering. In FRET, a donor fluorophore passes energy an acceptor within its vicinity of 2-10 nm. Since the efficiency of the process declines with the sixth power of the donor-acceptor distance, it can be used for measuring the clustering of fluorescently labeled proteins.
Fluorescence labeling of proteins is usually achieved by fluorescent monoclonal antibodies (see figure above) or with fluorescent protein constructs. We have achieved the following results in our method development efforts:
- We have shown that intensity-based FRET experiments carried out at high excitation intensities, commonly achieved in confocal microscopy, lead to severely underestimated FRET values if the analysis is carried out with the conventional equations disregarding fluorophore saturation. Fluorophore saturation is a situation when most fluorophores are in the excited state. The effect of this phenomenon on donor quenching is summarized in the figure below:
A complete treatment of this condition is available
- in this PDF presentation
- in a paper published in Anal. Chem.
- in the rFRET Matlab application
- we have developed a method for the investigation of the medium- and large-scale clustering of proteins based on flow cytometric measurement of homo-FRET (published paper)
- we have developed a method for the accurate estimation of the FRET efficiency in the presence of poor signal-to-noise ratio based on maximum likelihood estimation (published paper)
- we have proven using simulations that the mean of pixelwise FRET efficiencies calculated from microscopic measurements in the presence of poor signal-to-noise ratio is an unbiased estimator of the real FRET efficiency (published paper)
- we have proven that fluorescence labeling deteriorates antibody affinity and consequently the average degree of labeling (DOL) of the cell-bound antibody fraction is lower than that of the stock (published paper)