Light Microscopy - Beyond the limits of optical resolution

Optics textbooks tell us that the maximum magnification obtained through a light microscope is 400x: the closest two distinct points can be and still be resolved is 0.2 mm. This limitation is the result of light being diffracted by the object under observation and because diffracted light interferes with the image.

It can be seen from the formula that the resolving power improves with shorter wavelength. Ultra violet microscopes have higher resolving power, and the electron microscope higher still, but the latter suffers from the inconvenience that the electron beam requires a vacuum. In spite of these limitations, modern techniques have enabled the observation of objects smaller than the above theoretical considerations would allow. This progress has largely been the result of advances in fluorescence microscopy.

Here, fluorescent molecules (fluorophores) in the object under study are visualised. This is achieved by illuminating the object with monochromatic light at the wavelength which excites fluorescence emission in the fluorescent molecules. The object is viewed by means of filters which exclude the illuminating wavelength, and select the wavelength of the light emitted by the fluorophores. Each fluorophore can be thought of as a light source. This means that diffraction limitation is not reached when a single fluorophore is imaged. At a sufficiently low density of fluorophores, single molecules can be imaged. The disadvantage of the technique is that suitable fluorophores have to be introduced into the sample. This may be difficult, and the fluorophores may modify the behaviour of the object under study.

In practice, many molecules have been designed to be introduced into living cells and which bind to specific regions of the cells, highlighting the distribution of cell compartments and even the localisation of individual proteins. Using several different fluorophores with different optical properties, images are obtained in living cells where different organelles appear in different colours. The most convenient approach to obtain fluorescent images is to use an epi-fluorescence microscope, where the illuminating light illuminates the object through the objective. A filter block installed in the microscope tube directs the excitation light to the objective and filters the light emitted by fluorescence in the object and collected by the objective.

Nowadays, these microscopes are often linked to video cameras and the images are stored digitally for electronic display and analysis. Considerable progress has also been made in the synthesis of fluorescent molecules with improved qualities. Factors of concern are the suitable wavelength for excitation and emission (light should not be absorbed too much by the cell itself), the quantum yield (how many photons of light are emitted for a given number of photons absorbed), the bleaching rate (in an intense beam of light, the fluorophores gradually lose their ability to fluoresce), water solubility and the chemistry and specificity of attachment of the fluorophore to cellular sites. Also, the image which is observed comes from a rather thin section of the object determined by the depth of field (d_field) as follows:

d_field = l /NA² = 0.3 mm

but the fluorophores which contribute to the light collected by the objective are molecules excited by the light in the whole cone of light exiting from the objective. This means that the fluorescent image is superimposed on a bright diffuse background, and may have poor contrast or even be useless unless special precautions are taken.

Confocal microscopy

A new technique was devised to overcome the problem of high background fluorescence: confocal laser scanning microscopy. The technique is based on an epi-fluorescence microscope where the illumination light is provided by a laser whose beam is expanded to fill the objective. The fluorescent light emitted by the object is collected by the objective and focused by a secondary lens in front of a photomultiplier tube which is used to collect the light signal. A pin-hole is placed at the focal spot of the secondary lens. This ensures that only light in the focal plane enters the pin-hole, and light originating from above and below the focal plane of the object is rejected by the pinhole. The light through the pin-hole corresponds to one spot on the object plane. To build up an image, the pin-hole, or more usually a set of mirrors, are scanned in the x and y directions, thereby digitally building up an image of the whole image plane. The light originates from a narrow thickness (usually a few micrometers) of the specimen, as set by the size of the pinhole. The digital images can then be manipulated and stunning three-dimensional walk-through images can be produced. [see graphical illustration]

A more detailed explanation can be found on http://www.cs.ubc.ca/spider/ladic/intro.html. The technique however does have problems. For example the amount of exposure of the sample to high intensity light which is required to build up an image is considerable, resulting in bleaching of the fluorophore. The technique is now used widely, and is constantly being refined and improved.

Total internal reflection microscopy

This technique reduces the excessive background fluorescence which originates from fluorescent molecules which are not in the area of interest, but which deteriorate the contrast of the image. 'Total Internal Reflection' (TIR) is what happens to a beam of light travelling through a dense medium such as glass, and encountering an interface with a less dense medium such as air or water. If the beam of light encounters the surface at a sufficiently glancing angle, the light beam will not be refracted through the interface and continue travelling through the less dense medium, but will be reflected at the interface and continue travelling into the glass. The angle where the transition between refraction and total internal reflection occurs is called the critical angle, and is given by f, the angle between the light beam and the normal to the interface, so that:

f = sin^-1(n_water/n_glass)

which for glass and water gives a value of 60.9 ¢X. At angles slightly greater than the critical angle, namely when TIR takes place, an electric field component of the light penetrates through the interface into the water. This electric field is light with unusual properties. This light, the 'evanescent wave' or 'evanescent field' has the same wavelength as the incident beam but has the unusual property of penetrating only a short distance into the water, no more than 1 micrometer. This means that fluorophores may be excited by the light in the evanescent field if they are close to the glass/water interface, but fluorophores further away in the bulk of the solution will not be excited. The result is that images with very low background fluorescence are obtained.

The image shows a myofibril obtained from a skeletal muscle, lying on the surface of a glass prism, in the evanescent field generated by a totally internally reflected laser beam of 532 nm. The myofibril lies in solution containing a low concentration of Cy3-ATP, an analogue of ATP, the substrate for the active site of the myosin molecule. Myosin molecules are arranged into filament arrays which interdigitate with arrays of actin molecules. This gives rise to the striated appearance of the myofibril. The image was taken by Ronnie Burns, NIMR, 1997.

The image above has a high contrast, and the advantage of this technique over confocal scanning is that the exposure to excitation light is slight, and images can be collected very rapidly. The technique is highly suited to measuring rapidly changing or rapidly moving objects. The technique can be used to observe the behaviour of single fluorophores as shown below:

Pseudocolor display of fluorescent spots representing 50 pM Cy3-EDA-nucleotide observed on a synthetic filament of rabbit skeletal muscle myosin. Myosin filaments were illuminated with 7 mW Ar-ion laser (wavelength = 514.5nm). The images were taken at 3-s interval (a-k). All images were rolling-averaged over 8 video frames. The position of the myosin filament were confirmed as a fluorescent image in the presence of 5 nM Cy3-EDA-ATP (panel l). Courtesy of Dr. Kazu Oiwa, Kansai Advanced Research Center, Kobe 651-24, Japan

The main disadvantage of this technique is that the specimen are constrained in a thin micrometer-thick layer on the glass surface. This may be an advantage for observing the cell membrane in contact with the glass surface, but is not suitable for observing a cell in situ.

Two-photon excitation fluorescence microscopy

A further technique has been developed to achieve observation of cells in their tissue. This technique makes use of the fact that fluorescence can be excited in fluorophores not only by photons at a wavelength corresponding to the absorption ban for fluorescence, but also by photons at a wavelength twice that of the absorption band. The latter absorption only produces fluorescence if two photons are aborbed in quick succession, namely within a few femtoseconds of each other. The consequence of this two-photon excitation technique is that fluorescence emission only takes place where the photon density I sufficient to meet the photon-flux requirements for two-photon excitation, namely at the focal point of a microscope objective through which light from a high repetition rate laser travels. The consequence is that at any one time, the fluorescence originates in a small volume of the object. Scanning in the x-y plane and through the depth of the objects allows 3-D reconstruction of single cells as shown below, in a living tissue.

Purkinje neurone in a living brain slice filled with fluorescein dextran imaged with two-photon excitation laser scanning microscopy. Size of the image is 200x160 microns. By courtesy of Dr. Karel Svoboda, Cold spring Harbor Laboratories, Cold Spring Harbor, NY 11724, USA.

We have briefly seen that developments in techniques have led to images which would have been undreamt of, only a few years ago. Progress on many fronts was necessary: there has been new chemistry for developing biologically useful fluorophores, there has been considerable progress in camera technology, to record as many of the available photons as possible. Computing development has enabled scanning techniques to be affordable in many laboratories, and computer-aided lens design, and new lens materials, have produced lenses of higher quality and numerical aperture than ever. There is little reason to doubt that this progress will continue, and single molecule imaging will become routine. New techniques are also developing. Perhaps the most significant one is Atomic Force Microscopy where again the driving force is the interface between physics and biology - and that is Biophysics.

[Electron Microscopy]