FASPRI: a new method for increased spatial resolution in surface plasmon imaging of unlabelled living cells

Surface plasmon resonance (SPR) can be used to monitor molecular events such as the binding of an antibody to its target. It is a highly sensitive optical method. In this method, a parallel beam of light is directed into a glass prism, coated on the hypotenuse face with a metal (often gold), so that it reflects off the metal film and emerges from the prism. At a certain angle of incidence, the light excites collective oscillations of free electrons knows as surface plasmons and the gold ceases to reflect light. This is seen as a sharp dip in the plot of reflected light intensity versus angle of incidence. The position and the depth of the dip change dramatically when an antibody binds to the antigen-coated gold, and the degree of binding can thus be measured accurately, even when the bound antibody layer is only one molecule thick. Attempts to form high-resolution microscope images with SPR have failed because the beam of light reflected is parallel, and such beams cannot form detailed images.

We propose here to transform SPR into a high-resolution microscope method for imaging events involving small numbers of molecules in living cells without the need to label them with specific dyes. The reflected beam in the standard SPR method is essentially reflected by a metal mirror. It is a basic physical principle that in such a situation the incidence and reflected light interfere to produce a so-called standing wave, which has zero intensity at the mirror surface. It has been established since the 19th century that no light can be detected at the mirror surface. It may be useful to think of the standing waves that can be produced by hand in a skipping rope tethered to a wall at one end. At the wall no motion can be detected in the rope.

We propose to place a thin fluorescent layer made of fluorescent organic dye or nanoscale crystals between the glass and the gold film. This will ordinarily not fluoresce because it is in the aforementioned zone of zero intensity. However, in any region where SPR occurs, the reflected beam will be substantially reduced in intensity (i.e. the mirror will cease to reflect) and instead of a standing wave, there will be an ordinary propagating wave passing into the gold layer, creating resonance. Fluorescence will then be excited. Since the fluorescence will radiate in all directions it will be ideal for imaging: we will use a microscope specifically optimised for high-resolution and sensitive imaging of light radiating from fluorescent particles. We call this approach FASPRI, which stands for Full-Aperture Surface Plasmon Resonance Imaging, since filling the full aperture of the lens is our key improvement over other attempts.

First, we will sandwich a fluorescent layer between a gold film and a microscope coverslip. Next, we will check that the fluorescence appears at the plasmon resonance angle, which we will measure by reflectance. We will then adapt our existing fluorescence microscope and then perform FASPRI imaging of living algal and mammalian cells.

Surface plasmon resonance (SPR) is the most sensitive binding assay method known. It is often used to measure antibodies in solution against a monolayer of antigen molecules, giving measurable indications of changes in concentration of the order of 10 pg/ml without labelling. The commonest apparatus is the Kretschmann-Raether prism configuration in which p-polarized light is directed through a prism of high-refractive index glass and undergoes internal reflection in a thin metallic film. The biological material in aqueous solution contacts the gold and alters its reflectivity through the effect of specimen mass concentration on surface plasmons in the metal. As well as fluid assay, SPR has also been used for functional imaging of intact cells. However, because of optical limitations, the spatial resolution of these techniques is poor, even when high numerical aperture objective lenses are used.

We propose here a new approach that uses a combination of a standing wave and fluorescence for high-resolution SPR imaging of intact living cells. We will use a basic physics principle that fluorescence cannot occur at the mirror surface. In regions of the metal film where SPR occurs it ceases to be a mirror and fluorescent emission is released and fills the aperture of a high aperture lens. We will adopt this approach and use fluorescence as a proxy for SPR, allowing high spatial resolution label-free SPR imaging in living cells, potentially at high speed. Because we can use the full aperture of the lens for detection, we call this technique Full-Aperture Surface Plasmon Resonance Imaging (FASPRI). We will create the metal-fluorescent layers on glass coverslips, test their performance on a non-imaging optical bench rig, then we will incorporate this into an existing microscope equipped with a high speed sCMOS camera. We will perform FASPRI of algal and mammalian cells to prove principle and evaluate the trade-off between sensitivity and imaging speed.

This application aims to discover whether a photophysical effect predicted by theory can be realized in practice. If it works, the scientific, biomedical and societal effects will be literally incalculable. Assuming success, the following may then be expected to occur.

1. The University of Strathclyde seeks to apply for a primary patent for the use of the effect and for implementations in the form of instruments, with the aid of the major companies that are now in the field of SPR measurement.
2. The applicant will conduct biomedical testing on the microscope apparatus and will seek to optimize the metal/fluorophore sandwich which is the hardware basic to the method. Fabricating the sandwich will use known and available methods but is not elementary: appropriate noble metal and fluorophores must be selected and deposited at the optimal thickness.
3. The implementation as a microscope described here would find immediate use as an incremental development of instruments already available for measuring multiple samples. What this could mean in diagnostics, for example, is that a few microlitres of blood serum from a patient would be enough to measure the presence and binding affinities of hundreds of antibodies specifically and simultaneously. Seropositivity could be established even in cases where collection of blood could consist only of tiny volumes (e.g. in premature births). Conversely, antigens could be screened in tiny samples. The small size of the necessary antigen array would have the further advantage that conventional microscope objectives of high N.A. (meaning high collection efficiency) could be used, perhaps making the test more sensitive than current methods. Forensic testing might also benefit from this.
4. Since images of the type we hope to create of living cells have never previously been possible, we have to deduce from previous evidence that it will be possible to detect individual large viruses near to cell membranes and the fine processes known from electron microscopy to extend from platelets during blood clotting. Such images would be invaluable for observing dynamic events in virus release from cells and the explosion of platelets during the formation of blood clots, with implications for diagnosis and study of thrombosis.
5. Although the above implementations concern microscopy, it is also possible that the basic new capability of fluorescence emission detecting surface plasmon resonance would lead to a new generation of sensors. These could be miniature versions of Kretschmann-Raether prisms fabricated on the ends of optical fibre probes and capable of measuring ligands in body fluids within the patient's body.

Although these are speculations at present, if this method works I am confident that the interest of the world will quickly focus on it and many new implementations will follow.