Small genes encoding fluorescent proteins (e.g. EGFP, mCherry and dsRed), are used extensively as visible markers to study the location and timing of gene expression. Fluorescent fusion proteins are especially well-suited to studying in vivo gene expression patterns in the transparent C. elegans (Tursun et al., 2009). Perhaps one of the most significant limitations to their use is the high cost of the equipment needed to observe fluorescent proteins microscopically - the fluorescence dissecting stereomicroscope for screening and picking mutant worms and epi-fluorescent compound microscopes for more detailed studies. Commercially available fluorescence dissecting stereomicroscopes typically cost between US$12,000 and US$50,000. They are offered by only the "high-end" microscope companies, based upon their most expensive dissecting scopes, and incorporate premium-priced mercury arc-lamp illuminators, power supplies and epi-fluorescence modules. In a previous study, we demonstrated that significant cost savings could be achieved in EGFP detection by substituting high-flux blue spectrum LEDs for the mercury arc-lamps, and incident illumination for epi-illumination, if appropriate high-quality optical filters were employed (Papp et al., 2009). The resultant US$7,000 prototype system was compared with a US$50,000 top-of-the-line Leica fluorescence dissecting stereomicroscope system using the same plate of EGFP positive worms and the systems gave similar results. In the current study, we investigate the possibility of extending our findings to fluorophores with different spectral properties, such as dsRed and mCherry. We examined various LEDs combined with different excitation and emission filters. While the commercially available high-flux green spectrum LEDs that we studied do not have emission peaks that line up well with the excitation peaks of the red fluorophores, we were still able to demonstrate a highly usable level of detection by using appropriate optical filters. Since the human eye and fluorescence cameras can detect very low light levels, the absolute fluorescence level is not as important as the fluorescence signal-to-noise ratio. We will investigate sources of noise and present data (possibly with a demonstration) regarding techniques to improve the signal:noise ratio by reducing background fluorescence.

Affiliation:
- Tritech Research, Los Angeles, CA. www.TritechResearch.com