Inexpensive Method for Viewing Fluorescent DiI-Labeled Cells with a Dissecting Microscope

Fluorescent tracers and bioluminescent proteins allow researchers to label specific cells and to study their biological properties in vivo or in vitro. However, the epifluorescence compound microscopes typically used to observe labeled cells have a short working distance, and they are not suited for observing and manipulating labeled cells in vivo. Commercial epifluorescence attachments are available for dissecting microscopes, but they are quite expensive (several thousand dollars). We describe modifications to a standard dissecting microscope equipped with a fiber-optic illuminator that provide sufficient sensitivity to detect small clusters of fluorescent cells in intact living tissue and that cost less than $15. Using this approach, we have been able to greatly increase the fraction of fluorescently labeled neurons placed into primary cultures. This technique is also likely to be useful to researchers who want to transplant labeled cells in vivo. We have previously used the relatively nontoxic fluorescent lipophilic tracer [1,1′-dioctadecyl-3,3,3′,3′-tetramethylindo-carbocyanine perchlorate; DiI-C18-(3)] to retrogradely label sympathetic preganglionic neurons in the spinal cord of chick embryos by making injections into the sympathetic chains (1–5,7). The preganglionic neurons were then placed into cell culture for studies of synapse formation (5), neurotransmitter receptor regulation (1,2) or axon outgrowth (8). A limitation of this approach was that it was common to also label somatic motor neurons, because their axons pass very close to the sympathetic chain. With careful dissection with tungsten needles, it was possible to separate the spinal cord into two pieces, in each of which the only labeled cells were of the desired type. However, carrying out experiments was time-consuming because it was necessary to section and observe the material left behind to verify that only the desired population had been placed into cell culture. Furthermore, the labeled neurons comprised only a small portion of the cultured cells. Our yield of labeled cells could be increased by removing unlabeled cells using fluorescence-activated cell sorting (1), but the time, effort and cost that went into cell sorting were also high. We therefore needed a method that would allow us to observe clusters of DiI-labeled neurons in the living spinal cord, so that we could remove the appropriate cell populations under visual inspection. Rather than purchasing a commercial epifluorescent filter attachment, we have developed simple and cheap modifications to a standard dissecting microscope. We use Roscolux plastic filters that are designed for theatrical lighting as both excitation and barrier filters (Rosco Filters, Port Chester, NY, USA [available at most local theatrical distributors and from Edmund Scientific, Barrington, NJ, USA]). In keeping with the wide variety of lighting effects used in the theater, a huge number of filters are available, each with well-defined spectral characteristics. These filters come in sheets that are 40–130 μm thick and can easily be cut to the desired size and shape. Filters are made from either co-extruded polycarbonates or deep-dyed polyester; they are heatresistant up to temperatures of 160° or 125°C, respectively (although the manufacturer’s instructions warn that they should not directly contact the light source). For determining the appropriate combination of filters to use with a particular dye, a sample book with 11⁄2× 31⁄4-in pieces of 100 different filters (available for under $10) is very convenient. Once appropriate filters are identified, a typical 20× 24-in sheet costs only $5.95. Mercury or xenon arc lamps are typically used in fluorescence microscopy because they have a high luminous density, but a fiber-optic 150-W Model 180 Fiber-lite Quartz Halogen Illuminator (EKE lamp; Dolan-Jenner, Lawrence, ME, USA) at its maximum setting provided sufficient illumination (40 000 ftc) with minimal heat production. To attach the excitation filter to the light source, a circular piece of filter was placed into the bottom of a 35-mm Petri dish, which was then attached with Velcro to the fiber-optic lens. The Velcro allowed for quick insertion and removal of the filters. On one of our dissecting microscopes, a removable ring in front of the objective lens was used to mount the barrier filter; on the other, the barrier filter was placed inside a 60-mm Petri dish and then attached to the microscope body with Velcro. Figure 1A shows a cross section of a chick spinal cord 16 h after injection of DiI-C18-(3) into the sympathetic chains on both sides of the embryo, viewed with white light produced by the fiberoptic illuminator with no filters in place. This section was cut from living tissue with a Beaver Microsharp Blade (Becton Dickinson Labware, Bedford, MA, USA). In Figure 1B, the same section is viewed with the appropriate plastic filters in place. In this preparation, both preganglionic neurons (medial-dorsal cluster, arrows) and somatic motor neurons (ventral-lateral cell cluster, arrowheads) were labeled. Neither population of labeled neurons was visible with unfiltered white light (Figure 1A). Thus, these filters allowed the visualization of labeled cells, so that the isolation of preganglionic or motor neurons could be done more easily. Our previous experiments (3) demonstrated that DiI-labeled neurons survive well when placed into culture (Figure 1D); spinal cords that had been viewed through the plastic filters prior to dissection gave rise to the expected number of viable, fluorescently labeled neurons, so there is likely to be little or no excess cell death associated with the procedure. One important issue regarding this simple approach for visualizing fluorescently labeled material is its sensitivity. Figure 1 (C and D) demonstrates that single cells can be successfully imaged. In these experiments, we labeled E7 chick dorsal root ganglia neurons with DiI during dissociation as described by Honig and Hume (3) and then placed them into low-density cultures. Figure 1C shows the cultured neurons as they appear through the dissecting scope with the appropriate filters in place. Figure 1D shows the same field viewed with a Model IM35 Inverted Epifluorescent Microscope (Carl Zeiss, Thorn-