Use of UV excitation in confocal laser scanning fluorescence microscopy

Abstract By making only minor modifications, we adapted a conventional confocal beam-scanning laser microscope for the recording of UV-excited fluorescence. The major, and most expensive, change is that we coupled an external UV argon ion laser, providing the wavelengths 334, 351 and 364 nm, to the microscope scanner. We also replaced some optical components to obtain improved transmission and reflection properties in the UV. Only easily obtainable and inexpensive off-the-shelf components were used. The most serious problem encountered was the chromatic aberration of the microscope objective when using both UV and visible wavelengths. This is of no consequence in conventional microscopy where good imaging properties are important only in the visible region. In confocal microscopy on the other hand, good imaging properties are necessary for both the exciting and fluorescent light. Rather than having new optics designed, we tried with simple means to reduce the effects of the chromatic aberration to a tolerable level. This was done by mechanical adjustments in the ray-path. In addition we also tested two mirror objectives, which are inherently free from chromatic aberrations. However, such objectives have rather limited numerical apertures and are not of the immersion type. Their value in biomedical applications is therefore limited. The objective most frequently used in our experiments was a 63/1.25 oil-immersion fluorite. Without any compensation this objective had a depth resolution in UV-excited confocal fluorescence that was an order of magnitude worse than when using visible-light excitation. The useful field of view was also very small due to lateral chromatic aberration. By simple means we managed to improve the depth resolution by a factor of 4.4, and at the same time increase the useful field of view substantially. Still, the depth resolution was worse than what is obtained using visible light excitation. We think this is due to the fact that after compensation the objective is working with an incorrect tube length. Using the modified instrument, we recorded specimens labelled with AMCA and Fluoro-Gold, obtaining 1.5 μm thick optical sections.