We are pursuing three lines of research. First, we are developing the Atomic Force Microscope (AFM) for studies of host- parasite interactions. Second, we are developing blind deconvolution techniques to deblur light microscope images as an alternative and/or supplement to confocal microscopy. Although the AFM has been used to study living cells, technology to utilize effectively the AFM for long-term studies of living cells has not been developed. In addition, most of the AFM techniques available such as phase imaging and force measurements were developed for the material sciences where imaging is performed in air not liquid. This necessitated that we develop interpretations for these techniques in a liquid environment. This work has taken approximately 2 years but is now coming to an end. We are utilizing what we have learned in two ways. First, we are performing """"""""proof of concept"""""""" experiments such as the quantitative kinetic analysis of mitotic cell division. This is a fundamental cell biological issue with a great deal of background information but substantial unanswered questions. Second, we are using phase imaging, a variant of tapping mode AFM imaging, where the phase lag of the cantilever oscillation relative to the signal sent to the cantilever's piezo driver is the basis for image generation. Phase images can be generated as a consequence of variations in material properties such as adhesion, friction, and viscoelasticity. However, the usefulness of phase imaging for biological materials has been explored only superficially. We systematically analyzed the phase images of living homoiothermic vertebrate cells immersed in culture medium under highly controlled environment conditions at various tapping forces together with studies of freshly cleaved mica, glass, and collagen control samples imaged under identical conditions as controls. We found that the peripheral regions of COS-1 cells consistently showed a more negative phase shift than the glass substrate in phase images at the tapping forces of set-point amplitude: free amplitude (Asp/A0) = 0.6 - 0.8. In addition, at all Asp/A0 values suitable for phase imaging, tapping frequency appears to be high enough to ensure that phase shifts are governed primarily by stiffness. Consequently, phase imaging is capable of high resolution studies of the cellular surface by detecting localized variations in stiffness. We produced phase images of a bifurcating fiber in COS-1 cell cytoplasm at a lateral resolution of ca. 30 nm. Third, wide-field fluorescence images obtained with the light microscope suffer from severe out-of-focus glare. This problem can be partially, but not completely, eliminated by confocal microscopy. However, confocal microscopy introduces other problems. For example, if the microscope is set up in a truly confocal mode, the signal to noise ratio of the resulting image will be very poor. In addition, examination of the specimen by confocal microscopy can result in considerable photobleaching. Removing residual glare from either wide field fluorescence or confocal images has, heretofore, been performed using computer- driven image processing algorithms which make certain basic assumptions such as a constant Point Spread Function (PSF). However, it can be demonstrated that the PSF varies throughout the specimen; it is not constant. We have been utilizing computer-based methods which make and require no basic assumptions. This technique is known as """"""""blind deconvolution."""""""" We have had limited, but spectacular, success with the technique. As a proof of concept, we deconvolved a dividing vertebrate cell infected with Toxoplasma gondii. We observed details which are either not apparent or less distinct than observed using confocal microscopy.
Showing the most recent 10 out of 21 publications
