Capillaries can enable high-throughput live-cell investigations, including those for drug discovery and development and for studying how certain environmental conditions such as pressure might affect cellular processes in basic research. Even though capillaries are useful for these popular applications, because they are cylindrical, their walls introduce aberrations during imaging applications.“Investigators often shy away from using capillaries because they worry about image quality,” assistant professor Paul Urayama said. He and members of his laboratory at Miami University in Oxford, Ohio, demonstrated that they can achieve high spatial resolution with a commonly used technique: spatial deconvolution. Or, as Urayama put it, “The point of the paper was to say, ‘Hey, conventional approaches do work in improving image quality in capillaries, and you can get subcellular spatial resolution.’”Fluorescence from objects smaller than cells ideally should appear as points of light in images but, under a microscope, these points often look like blurry blobs of light. Spatial deconvolution clears up blurry images using an algorithm that, in the case of an image of a lot of fluorescent dots blurred together, guesses how those dots would appear in an ideal image. Then it blurs the supposed ideal image using an aptly named point-spread function, a mathematical formula that describes the blurred object and refines the image by repeating this process over and over until there is a clear image with discrete points.To demonstrate that they can apply the technique to curved capillaries, the researchers captured and processed images of subcellular-size Fluoresbrite YG microspheres from PolySciences Inc. of Warrington, Pa. In the future, they will use fluorescently labeled living cells instead of fluorescent beads. They excited the microspheres with the 488-nm line of an argon lamp and recorded images with an Olympus IX-81 microscope in epi-illumination mode and a Cooke Corp. camera with a CCD array of 1376 × 1040 pixels, each 6.45 μm in size. The pixel size was important to factor in the deconvolution for measuring accurate distances between the beads. For deconvolution, stacks of images are taken downward in the Z, or axial, direction. At least 50 stacks must be compared to generate a resolved image. The Olympus microscope has a computer-controlled stage that moves in 10-nm increments in Z.The software they used for deconvolution was SlideBook 4.0 with a 3-D deconvolution module from Intelligent Imaging Innovations Inc. of Denver, which Olympus bundled for the researchers along with the microscope and camera. “You don’t need specialized high-pressure or capillary-based software to do this,” Urayama said.Urayama’s group plans to use a capillary sandwiched between plates to study the effects of extreme environmental conditions on cells. An opening in the plates provides a window for viewing living cells and other objects in the capillary. Adapted with permission of Review of Scientific Instruments.After capturing stacks of images of the fluorescent microspheres, they performed deconvolution as if they were imaging a flat surface and not a curved capillary, and they did not observe a difference. This shows that, for imaging cells, you can treat these curved capillaries as though they are flat, as reported in the June 2008 issue of the Journal of Microscopy.The investigators demonstrated this effect for a range of microscope objectives, with a 0.75 NA and 0.40-mm working distance, a 1.30 NA and 0.10-mm working distance, and a 0.50 NA and 7.40-mm working distance. “One reason to use capillaries is that the walls are thin enough to use high-NA optics,” Urayama said.The original image of the microspheres (left) was improved by deconvolution for a flat surface (middle) and by taking into account the curvature of the capillaries (right). There was no appreciable difference. Adapted with permission of the Journal of Microscopy.These experiments were performed at normal atmospheric pressure, but the researchers want to use these capillaries as pressure-tolerant chambers to study the effects of extreme environmental conditions on metabolic rates, particularly using the classical metabolic indicator NADH. Pressure is known to affect ligand binding, which means that it could have a lot of effects on biochemical pathways. And because pressure lowers the pH of any metabolic system, Urayama also plans on doing pH imaging.Contact: Paul Urayama, Miami University; e-mail: urayampk@muohio.edu.