Pawley, James B.

Phone: +1-608-263-3147

Fax: +1-608-262-9083

Email: jbpawley@facstaff.wisc.edu

Office: 223 Zoology Research

250 N Mills Street

Madison, WI, USA 53706

My research interests have always been at the interface between biological specimens and physical instrumentation, generally microscopes, particularly 3D microscopes. While at Madison, I have worked to improve the performance of both high-voltage transmission electron microscopes, and low-voltage, high-resolution scanning electron microscopes.

More recently, I have been concentrating on improvements in the confocal scanning light microscope. This instrument permits one to make images of a single plane in a thick, fluorescent specimen. By collecting data from many adjacent planes, it is possible to produce "real" 3D images. More importantly, and unlike any sort of electron microscopy, it is possible to produce such images from specimens that are living and changing.

The biggest limitation to the use of the confocal microscope on living specimens is that the interaction of the exciting light and the fluorescent dye molecules can produce toxic substances that damage the cell. The only solution is to count the fluorescent light more efficiently so that the amount of excitation needed to make a usable image is reduced. To this end, I worked with Dr. James Janesick, (formerly head of CCD Design at JPL) to develop a silicon-based photosensor called the CCDiode. This sensor would have had about 10 times higher quantum efficiency than that of the photomultiplier tubes now used in confocal microscopes. This project collapsed in a Dot.Com disaster but is now being reborn in the form of a collaboration with Cliff Weatherup of Marconi Ltd. (UK, now called "E2V"). I am working with E2V to develop a new and improved version of this device at present. E2V holds the patent on the "gain register," a new charge amplifier that is far superior to any previous amplifier for use in a detector for the confocal microscope. In conjunction with Prof. Christian Seebacher, at the University of Munich, we have now obtained the fist images using a prototype detector.

This project gained new emphasis when it became evident that a detector of similar capabilities might be useful for reading out a new form of high density CD storage medium. There is really now some chance that it might be made!!

While on my sabbatical in Sydney, I investigated other practical problems that affect the use of the confocal microscope on living cells: Backscattered light (BSL) and spherical aberration.

Light is backscattered whenever it passes from a medium of one refractive index (RI) to that of another RI. Cells are full of refractile structures and these scatter light back towards the objective. If this light can be collected, it can be used as a confocal signal defining the location of small, scattering structures in the cell. Collecting this signal not only provides more 3D information about the specimen without subjecting it to additional light, it also provides a built-in measure of the quality of your image and how this might be improved. This second capability comes about because many of the scattering objects in the cell are smaller than the resolution limit of the microscope. The image of such a small object is called the "point-spread function" of the microscope and knowing it can allow one to use a computer to "deconvolve" the collected data set to produce an image that more closely resembles the original object. Though long known, the use of deconvolution on confocal data sets has been limited by the absence of suitable point-spread function images in most fluorescent data sets. This lack can be supplied by collecting BSL data in parallel with the fluorescent signal. Since leaving Sydney, I have been working with Drs. Felix Margadant, Peter Torok and Colin Monks to implement such a system.

Although spherical aberration can be corrected for by careful design of the objective lens, this correction is only effective if certain, precise conditions are met regarding the size and RI of every element between the focus plane and the camera. It is now becoming clear that this condition is seldom met, particularly on cells in aqueous media but covered with a coverslip. Even when using a water lens with correction collar for coverslip thickness, problems are common. In fact, next to low detector quantum efficiency, it is undoubtedly the greatest practical limitation to high performance in live-cell confocal microscopy.

I have been working with Jay Margolis and Colin Monks to produce theoretical and practical measurements of the problem and the extent to which it can be overcome by the use of aberration correctors or changing the RI of the immersion medium. This eventually lead to a product (the SAC), that won an IR-100 award. Click here for project details.