Evolution-on-a-Chip

Lab-on-a-chip technology is becoming increasingly commonplace in biological research. The discipline of evolutionary ecology would not seem at first glance a strong candidate to take advantage of these microscale-fabricated devices, but this thinking may soon change. Traditional studies of evolutionary ecology using bacteria have been performed on bacteria grown in reactors called chemostats. Chemostats are not heterogeneous environments and lack spatial structure, limiting the experiments a biologist can perform. To remedy this situation, Keymer et al., set out to fabricate microscale devices to study how bacteria adapt to different regions on a heterogeneous landscape. The fabricated “landscape” consists of a linear array of 85 square wells, 100 m square by 30 m deep in size, called microhabitat patches (MHPs) etched into silicon wafers and all connected by channels 5 m wide and 50 m long, allowing for migration of bacteria between MHPs. Nutrients were supplied to MHPs by 200-nm channels—too narrow for bacteria to pass through. By changing the number of etched nutrient channels going to specific MHPs they were able to change the landscape conditions, making some MHPs more nutrient rich than others. The results they obtained demonstrate the ability to use these devices to understand evolutionary processes at both a local (a single MHP) and global (the entire array of MHPs) level. Keymer et al., found that in a “flat” landscape, where all 85 MHPs had equal numbers of nutrient channels, local MHPs showed fluctuations in population size while over the entire array a metapopulation emerged that demonstrated a more constant growth rate. In more complex experiments where different numbers of nutrient channels were connected to MHPs, bacteria were able to colonize the entire landscape. Interestingly, this colonization process could be divided into distinct epochs based on bacterial expansions over time into all areas of the landscape, even adaptation to the nutrient limited MHPs. This current work demonstrates the utility of microscale-fabricated devices for studying evolutionary ecology and suggests in the future people might be describing “environments-on-a-chip.” –NB

- Keymer et al. 2006. Bacterial metapopulations in nanofabricated landscapes. Proceedings of the National Academy of Sciences of the USA 103:17290-17295.

Laser Sharp Gene Expression

The ability to control gene expression in a specific tissue or cell type allows researchers to better understand gene function in vivo. Turning on the expression of a transgene can be done using genetic systems, such as yeast GAL4/UAS, or simply using a heat-shock promoter. The GAL4/UAS system requires the GAL4 transcription factor to be under the control of a regulatory region to permit tissue specific expression, while induction of heat-shock promoter driven transgenes often leads to a global expression pattern. An obvious drawback of the GAL4/UAS system is the need for a specific regulatory region that may not be available for all tissues or organisms of interest, while the draw-back of heat-shock driven expression is lack of specificity in the expression pattern. Studies have shown single cell expression from a heat-shock promoter driven transgene can be induced using either heated needles or lasers. Ramos et al., are now moving one step further, showing that laser-induced gene expression can occur over a much larger population of specific cells. In studies using the butterfly (Bicycles anynana) wing as a model system, transgenic butterflies harboring a GFP transgene fused to the Drosophila hsp70 promoter were subjected to a variety of laser pulses; microslits were placed over the laser beam, allowing different patterns of light to be focused on the areas of interest. Laser light focused in a single line pattern on the wing resulted in green fluorescence in a subset of cells in the treated area in more than 75% of individuals. Ramos et al., went on to show that even more complex GFP expression patterns could be obtained by creating microstencils to cover the laser beam, demonstrating the potential of this technique to reproduce the more complicated gene expression patterns often observed during development. This work improves on methods to induce gene expression in large numbers of specific cells, expanding the repertoire of techniques available for gene function analysis. –NB

- Ramos et al. 2006. Temporal and spatial control of transgene expression using laser induction of the hsp70 promoter. BMC Developmental Biology 6:55.

Divide and Conquer

Splitting the yeast transcriptional activator Gal4 into its DNA binding and transcription activation domains has been a favorite trick in molecular biology for over 15 years, most famously in the yeast two-hybrid system. In a recent paper in Neuron, Luan et al., show how a related strategy can be used to analyze neuronal networks in Drosophila. The authors’ approach relies upon transgenic flies expressing (i) the Gal4 DNA binding domain (DBD) fused to a leucine zipper motif; (ii) the Gal4 activation domain (AD) in tandem with a leucine zipper sequence capable of heterodimerizing with the Gal4DBD-Zip fusion; and (iii) a Gal4-regulatable reporter. Only cells that express the first two constructs would be capable of expressing the third component. If each of the Gal4 fusion constructs is under the control of a different promoter, then only a limited subset of cells will express both halves of the transcription factor and therefore the expression of the reporter will be highly defined. To demonstrate the validity of the system, the authors tested Gal4AD under the control of the Crustacean Cardioactive Peptide (CCAP) promoter and Gal4DBD driven by the choline acetyltransferase (Cha) promoter. This experiment pinpointed a population of some 18 neurons in whole mounts of the Drosophila CNS in which both the CCAP and Cha promoters were active. Of course, the Gal4-regulated gene need not be a reporter but could be a transgene capable of manipulating the system in some experimentally meaningful way. In that vein, Luan et al.,. used their defined expression strategy to selectively ablate neurons using reaper, a cell death gene. In a related approach, the authors also showed that they could perform enhancer trap studies. For this application, they used a VP16AD construct to make the enhancer trap lines, and in concert with a CCAP-driven Gal4DBD construct, were able to characterize two distinct neuronal subsets (comprising the subesophageal and thoracic ganglia and the brain and abdominal ganglia, respectively). This strategy for observing and manipulating neuronal subsets should allow more precise understanding of neuronal circuits, and may prove transferable to vertebrate models, including zebrafish and mice. –ND

- Luan et al. 2006. Refined spatial manipulation of neuronal function by combinatorial restriction of transgene expression. Neuron 52:425-436.

The Fruits of Fiber

Imaging deep brain regions in living animals may be performed noninvasively via MRI or PET. However, these techniques can at present offer only limited spatial and temporal resolution. Recently, Vincent et al., provided a convincing demonstration of the capabilities of an alternative: fibered fluorescence microscopy (FFM). This approach uses bundles of thousands of optical fibers, which are incorporated into microprobes 300 or 650 m wide. To test the resolution of FFM, a series of in vivo imaging experiments were performed in Thy1-YFP mice, which express a fluorescent marker in all sensory and motor neurons. After successfully imaging nerve cell bodies, individual nerve fibers, single axons, and neuromuscular junctions, Vincent et al., used FFM for repeated monitoring of the saphenous nerve in a crush injury model. Fifteen daily postinjury imaging sessions were conducted, and together recorded the hallmark signs of degeneration-regeneration that are characteristic of this model system. The specific advantages of FFM in this experimental model include more efficient data acquisition as compared to microscopic analysis of fixed nerves, and the ability to perform repeated measurements on the same animal (thus decreasing the number of subjects needed). Although the analysis of nerve regeneration required a small incision, noninvasive imaging can also be performed with FFM, as shown by data obtained by inserting the probe into the nasal cavity and visualizing the olfactory neuroepithelium. The most exciting results that are reported by Vincent et al., have to do with deep-brain imaging. These experiments used transgenic mice that express EGFP in particular brain regions or, for calcium imaging experiments, rats injected with a calcium indicator dye and given pharmacological or electrical stimulation. In these experiments, real-time imaging at cellular resolution was possible at a depth of up to 6 mm. Further refinements in technology (the use of laser pulses to reduce bleaching) and reagents (viral vectors expressing other reporter genes) should allow a wide variety of functional studies to be performed. Ultimately, it may also be possible to adapt this system to in vivo imaging in nonanesthetized, motile animals. –ND

- Vincent et al. 2006. Live imaging of neural structure and function by fibred fluorescence microscopy. EMBO Reports 7:1154-1161.