The impact of spectrin-lipid interactions on membrane biophysics
Lead Research Organisation:
University of Exeter
Department Name: Physics
Abstract
The membrane of the cell has a remarkable structure that ensures the optimal cell shape and elastic properties during its lifetime. The shape and elasticity of the cell are of crucial importance to its biological functions, and many diseases have been related to abnormalities in membrane function (both as a cause or a result of the disease). Therefore it is of primary importance to understand the factors controlling the mechanical properties and functions of cell membranes. The cell membrane consists of a double layer of different types of lipids, which are underlined by a regular mesh-like protein network attached to the lipid membrane via special protein junctions. The underlying protein network is built primarily of the protein spectrin which is able to form long filaments. However, little is known about the interactions between spectrin and the lipids forming the membrane. In this project, we set out to identify the primary lipid binding to spectrin, and how this interaction affects the elastic properties of the lipid membrane. As the cell membrane consists of several hundred different lipid types, we will investigate how the interactions between spectrin the the lipids could change the lipid distribution along the membrane, and whether spectrin is able to assemble small patches of lipids enriched in a particular lipid species. Finally, we will investigate how the membrane electrical properties and permeability to solutes are affected by the attachment of spectrin molecules. Our research will help understanding the cell membrane structure and functions in detail and show possible ways in which the protein membrane skeleton is able to change the behaviour of the membrane. This work may also lead to the discovery of a new type of artificial membranes, whose mechanical properties may be varied by incorporating different proteins. In the longer term, it will clarify some membrane abnormalities associated with serious and life-threatening diseases.
Organisations
| Description | The membrane of the red blood cell consists of two main components: a double layer of different types of lipids, and a regular mesh-like protein network attached to the lipid bilayer via special protein junctions. This structure endows the red cell membrane with the necessary mechanical properties (elasticity) ensuring it is able to withstand forces exerted in the blood flow and fulfilling its biological functions. The underlying protein network (the membrane skeleton) is built primarily of the protein spectrin which is able to form long filaments. Although the structure of this network and the nature of the attachment complexes to the lipid bilayer are well understood, little is known about the direct interactions between the spectrin molecules and the various lipids forming the membrane, as well as any role such interactions may play for the mechanical stability of the membrane. In this project we set out to investigate these interactions. In the first part of the investigation, we studied the interaction between spectrin molecules and two of the constituent lipid species found in the inner leaflet of the red cell membrane, phophocholine (PC) and phosphoserine (PS). These lipids were used to form several different types of artificial membranes (vesicles). Using fluorescence microscopy as an imaging tool, we demonstrated that spectrin attaches more strongly to lipid membranes containing a mixture of PC and PS molecules than to pure PC membranes. To further explore this interaction, we formed membrane from four different lipid components, PC, PS, sphingomyelin and cholesterol. This particular mixture forms membranes which are heterogeneous in nature, with distinct microdomains containing lipids in different compositions. In these membranes, spectrin showed specific affinity to domains enriched in PS, which confirmed the conclusions from the previous study. The binding however appeared to be weak, which may be due to the specific phase state of the PS-enriched lipid microdomains. Finally, we investigated the direct interactions between spectrin and the native lipid membrane of red blood cells. To achieve this, we first induced the relocation of the PS molecules from the inner leaflet of the membrane to the outer one, and fluorescently labelled this compound in order to visualise it using a fluorescence microscope. Then we added fluorescently labelled spectrin to the cell suspension. We found a preferential binding of spectrin molecules to membrane domains rich in PS. This demonstrated that spectrin preferentially binds to PS molecules in the plasma membrane of red cells, in a similar way to the artificial lipid membranes. The set of experiments described above suggested that labelling PS in lipid membranes with fluorescence markers may prevent the attachment of spectrin to the membrane due to the competing character of the binding between PS and the fluorophore, on the one hand, and PS and spectrin, on the other. To clarify this we performed experiments using flow cytometry. The results supported this hypothesis: whilst spectrin preferentially attaches to PS, this interaction was hindered if the membranes were labelled with a fluorophore attached to the PS molecules prior to the addition of spectrin. This led to the conclusion that in cases like this, other, preferably label-free, imaging techniques should be used to visualise protein-membrane interactions. One of the attractive possibilities would be to use Raman-based microscopy, such as Coherent Anti-Stokes Raman Scattering (CARS) imaging or Stimulated Raman Scattering (SRS) imaging, combined with Two-Photon Fluorescence (TPF) microscopy. Although the weak interactions between spectrin and lipid membranes did not allow us to use this set of techniques, we demonstrated the viability of such an approach using a different elastic protein, elastin, and investigated its attachment to artificial membranes (vesicles) and red cells (see below). Quantification of spectrin-membrane interactions was performed using Isothermal Titration Calorimetry (ITC), during which small vesicles (100 nm in size) were titrated in a protein solution. The results showed a weak attachment of spectrin to vesicles containing PC:PS in proportion 70:30. When the membranes lacked any PS, there was practically no binding of spectrin to the membrane. Previous studies from our and other groups suggested that direct spectrin attachment to the membrane may have an effect on the elasticity and lipid organisation of the lipid membrane. In the next stage of this project we investigated this possibility. Thin membranes are characterised by three distinct elastic constants which describe the resistance of the membrane to bending (bending modulus), shear deformation (shear modulus) and area dilation (area dilation modulus). The shear and area dilation modulus are convenient to measure using the so-called micropipette aspiration technique, whilst the bending rigidity of the membrane can be evaluated using thermal fluctuation spectroscopy, a variant of which we implemented in our laboratory. The vesicles we used were in liquid state hence their shear elastic modulus was zero. We therefore used the micropipette aspiration method to measure the membrane area dilation modulus, and thermal fluctuation analysis to measure the membrane bending modulus. Detailed measurements on vesicles aspirated in thin pipettes demonstrated that spectrin had no effect on the membrane area dilation modulus. Similar results were obtained also for the membrane bending modulus using the method of thermal fluctuation spectroscopy. A possible reason for this discrepancy between previous results reported in the literature and our measurements may be found in the purity of the extracted spectrin. For the purposes of our investigation, we developed an improved protocol for extracting spectrin from whole blood. A number of analytical techniques (Western blot, SDS-page gel and enzyme-linked immunosorbent assay (ELISA)) showed the presence of spectrin with high purity. We also extracted spectrin following the classical protocol found in the literature. Parallel experiments using these two extracts using Isothermal Titration Calorimetry lead to different results, showing that the pure spectrin fraction interacts more weakly with the lipid vesicles. The research under this project has yielded a number of important results that went above and beyond the expected outcome leading to new methods for investigation of interactions between proteins and lipid membranes. Some of the most important ones are summarised below: (1) We developed a novel two-channel detector for imaging phase-separated membranes with CARS microscopy. Membranes with distinct domains are often difficult to visualise, and the most widely used technique is fluorescence microscopy requiring a specific exogenous fluorescent marker to be incorporated in the membrane. CARS microscopy offers the possibility to image lipid membranes label free, using the CH vibration frequency. One way to achieve a contrast between different domains is to use selectively deuterated lipid species, which shifts the CH vibration frequency to that of CD. Using separately these two frequencies allows contrast to be achieved between different lipid microdomains in the same membrane. The problem is that the imaging at the two frequencies is preformed one after the other, which means that the shape and/or position of the vesicle as well as the domain structure may change due to diffusion and thermal fluctuation. Our new method performs the imaging simultaneously, allowing the capture of the whole image from the normal and deuterated parts of the membrane at the same time. (2) We developed a novel method, based on non-linear optical methods, for investigation of interactions between lipid membranes and proteins. The method is based on simultaneous label-free imaging of lipid membranes using Coherent Anti-Stokes Raman Scattering (CARS) microscopy, combined with Two-Photon fluorescence (TPF) microscopy to image the proteins. The method was tested on a model system consisting of giant lipid vesicles and the elastic protein alpha-elastin. In this particular case, every stage of the imaging was label-free since alpha-elastin is an auto-fluorescent protein that lends itself to two-photon fluorescence microscopy. This technique however can be adapted to proteins which are not auto-fluorescent, by appropriately labelling the protein molecules. The imaging data can be supplemented by measurements using other methods (such as ITC) to characterise better the protein-lipid interactions, as we demonstrated for alpha-elastin. (3) We also successfully used Stimulated Raman Scattering (SRS) microscopy to image bilayer lipid vesicles. This technique proved much more efficient than CARS in imaging lipid membranes, due to the much lower background signal, and discriminating between uni- and multi-lamellar membranes. We also demonstrated that SRS microscopy can also be combined with TPF microscopy for simultaneous imaging of lipid membranes and attached proteins. (4) For the first time, we demonstrated the feasibility of using non-linear optical microscopy (CARS, SRS) as an imaging tool for membrane micromechanical measurements. We combined the micropipette aspiration technique with CARS or SRS which allowed for an accurate resolution of the shape of the aspirated vesicle. This approach will be particularly useful when investigating the mechanical compliance of phase-separated membranes consisting of microdomains, by allowing a detailed imaging of the membrane lateral domain organisation and its evolution due to the applied stress. Currently, this could be achieved by a combination between fluorescence microscopy and another optical method (e.g. differential interference contrast), which may be cumbersome. (5) We developed a novel label-free method to investigate the mechanical properties of live cells and characterise in detail the deformation of each individual constituent of the cell (i.e. plasma membrane, cytoskeleton, cell nucleus) likely to contribute to the integral mechanical response of the cell. The method consists of two steps. Firstly, using non-linear optical techniques, hyperspectral imaging of each three individual cell constituent (plasma membrane, cytoskeleton, cell nucleus) are recorded and analysed. This allows a distinct spectral fingerprint for each constituent to be identified, which can be used to locate exactly the position and shape of each one of them in cells subjected to deformation in a micropipette. Secondly, the cell is aspirated in a micropipette and SRS is used to visualise the different cell constituents. Each of these constituent can be monitored in the course of the micromechanical measurement, and its shape and position in the micropipette to be visualised as the negative pressure in the pipette is increased and more of the cell in aspirated inside. This allows more information to be extracted to characterise the contribution of each of these cell constituents to the overall mechanical compliance of the cell. We demonstrated the applicability of this method using fibroblasts aspirated in a micropipette, and imaged using SRS microscopy. Parallel experiments using the classical fluorescence confocal microscopy and the same cells with fluorescently-labelled cytoskeleton were used as a means to verify the new technique and showed excellent agreement between the two techniques. Four manuscripts reporting the main results of this work are currently in preparation. |
| Exploitation Route | The main results from this project clarify the extent to which the most abundant protein of the membrane skeleton, spectrin, directly interacts with the lipid bilayer membrane. These findings will be of interest to researchers working in the area of fundamental biophysics, cell and molecular biology and biomedicine. In the course of the project we have also developed several innovative methodologies mainly based on non-linear optical microscopy, which can be used to investigate interactions between lipid membranes (native and artificial) and exogenous molecules such as proteins, toxins and other biologically active agents. The most important ones are: (1) We developed a novel two-channel detector for imaging phase-separated membranes with CARS microscopy. This is a label-free imaging technique suitable for heterogeneous membranes containing micro-domains. The advantage of the method is in the simultaneous acquisition of the image, avoiding artefacts due to domain diffusion and membrane thermal fluctuations. (2) We developed a novel method, based on non-linear optical methods, for investigation of interactions between lipid membranes and proteins. The method is based on simultaneous label-free imaging of lipid membranes using Coherent Anti-Stokes Raman Scattering (CARS) microscopy, combined with Two-Photon fluorescence (TPF) microscopy to image the proteins. This method could be the technique of choice when fluorescent labelling is undesired and should be avoided. This opens new possibilities for chemically specific imaging of biological samples, and in particular, membranes. (3) We also successfully used Stimulated Raman Scattering (SRS) microscopy to image bilayer lipid vesicles. We also demonstrated that SRS microscopy can also be combined with TPF microscopy for simultaneous imaging of lipid membranes and attached proteins. (4) For the first time, we demonstrated the feasibility of using non-linear optical microscopy (CARS, SRS) as an imaging tool for membrane micromechanical measurements. We combined the micropipette aspiration technique with CARS or SRS which allowed for an accurate resolution of the shape of the aspirated vesicle. This approach will be particularly useful when investigating the mechanical compliance of phase-separated membranes consisting of microdomains, by allowing a detailed imaging of the membrane lateral domain organisation and its evolution due to the applied stress. (5) We developed a novel label-free method to investigate the mechanical properties of live cells and characterise in detail the deformation of each individual constituent of the cell (i.e. plasma membrane, cytoskeleton, cell nucleus) likely to contribute to the integral mechanical response of the cell. This novel approach will be very useful for researchers aiming at detailed characterisation of the mechanical compliance of different cells. |
| Sectors | Healthcare,Pharmaceuticals and Medical Biotechnology |