Slide Show:

Figure legends

Figure 1:A) Drosophila lethal mutations are maintained over a balancer chromosome. Balancer chromosomes are multiply inverted chromosomes, to minimize recombination events, that often carry lethal mutations so that homozygous balancer embryos die. Embryos produced by parents with the genotype mutation/balancer produce offspring of the following genotypes: homozygous mutant embryos (mutation/mutation), heterozygous embryos (mutation/balancer) and homozygous balancer embryos (balancer/balancer). Only the heterozygous mutant/balancer embryos can develop into adults and so maintain a stable stock. However, for any embryo collection from these parents, only 25% of the embryos contain the homozygous mutation of interest. Using a GFP marker gene on the balancer chromosome it is possible to identify this minority population of mutant embryos from their balancer containing siblings, in living embryos.

B) A schematic diagram of the embryo sorter outlining how the machine operates. Embryos are maintained in suspension in the embryo chamber using a magnetic stirrer. Solution is pumped through the embryo chamber at a fixed rate (~6 ml/min). Some of the suspended embryos randomly enter the flow of liquid that is exiting the chamber. This flow carries the embryos through the optical cuvette. As embryos pass through the cuvette they are illuminated by an argon laser (488nm). The emitted light passes through two focusing lenses and then through two dichroic mirrors. Light is detected at three points along the light path: 1) A diode, which detects all interruptions of the light. 2) PhotoMultiplier Tube-1 (PMT-1) detects the light between the 510 and 520nm dichroic mirrors. This light passes through a very narrow GFP filter (510+5nm) before reaching the PMT-1. This is the detector for the GFP signal. 3) PMT-2 detects all the light that passes through the 520nm dichroic mirror. This PMT detects some of the autofluorescence. A computer draws a peak for each embryo for the signals received by the diode and the two PMTs. If the GFP peak is above or below a defined threshold the computer sends a signal to the mechanical switch to direct the embryo flow to either the save or waste tubes.
 
 

Figure 2: Optical system used in the embryo sorter.

A) The optical cuvette is composed of a 0.4mm diameter square tube, through which the embryos pass. This tube is enclosed in a square chamber that is filled with immersion oil and contains two exposed sides or viewing windows, one for excitation by the laser and the second for the emitted light. This viewing windows are perpendicular to each other.

B) Schematic diagram of the 0.4mm square tube. The diameter of the tube restricts the number of possible orientations of the embryos. This results in a reduction in the amount of reflected light that is detected from the embryos. C) The optical cuvette is aligned with the optics chamber. This ensures that the viewing window is aligned with the focusing lens such that the emitted light from the embryo passes through the focussing lens and is received by the PMTs. D) End on view of the optics chamber when the optics cuvette has been removed. Note that the light emission is detected at right angles to the excitation light from the laser.

E) Side view of the inside of the optics chamber. The optics chamber contains the dichroic mirrors and filters. The emitted light (red arrow) passes through this chamber and is detected by the diode and the PMTs. F) End on view of the optics chamber where the optic cuvette has been replaced. The optic cuvette is aligned with the laser and covered with protective black plastic.

(Yellow arrows = direction of the embryo flow. Red arrows = direction of the light path)
 

Figure 3: The Mechanical switch.

A) End on view of the optics cuvette (fig 3A). The embryos exit the central hole in a droplet of solution before reaching the switch. B) The end of the cuvette (in A) attaches to a metal ring. The metal ring has an inflow of solution at high speed from a second pump. The drop of solution containing the embryo enters a sheath of liquid and exits through the central hole in a stream of liquid. C) The cuvette-fluid ring described in A and B is attached to the end of the optics chamber. D) Looking down on the embryo switch. This switch is aligned underneath the cuvette-fluid ring (i.e., underneath stream the blue arrow in C). The embryos flow down, in a stream of solution, from the fluid ring to the embryo switch. E) Side view of the embryo switch. The default position is for the switch to be aligned to the waste tube, it then moves to the right to collect a Save embryo.

(Yellow arrows = direction of fluid flow. Red arrows = direction of the light path)
 

Figure 4: The embryo density is maintained using a two-chamber system.

Embryos are initially added to the high-density embryo chamber. Solution is pumped through the low-density chamber (blue arrows) and exits a tube to the optical cuvette. If the rate of embryo sorting drops below a defined threshold the computer sends a signal to the fluid valve switch. This fluid valve then directs the fluid to be pumped through the high-density chamber (black arrow). Some of the solution and embryo suspension will then leave the high-density chamber into the low-density chamber, resulting in the addition of embryos to the low-density chamber. When the rate of embryo sorting increases to the defined threshold, the computer will send a signal back to the fluid switch to re-direct the fluid flow back to the low-density embryo chamber. This will stop the addition of more embryos.

A) The panel on the left contains a typical embryo collection from a GFP-balancer strain of Drosophila. This mixed population of GFP containing embryos can be sorted to obtain a population of non-GFP expressing embryos or a population of GFP expressing embryos.

B) Immunostains of wildtype and sorted twist mutant embryos. A molecularly characterized loss-of-function twist allele was placed in trans to a GFP-balancer chromosome. Embryo collections from this stock were sorted, selecting for non-GFP containing embryos, i.e. homozygous twist mutant embryos. The sorted embryos were formaldehyde-fixed and immunostained with dMef2 antibody (Drosophila Myocyte Enhancing Factor). The left panel shows a wildtype embryo immunostained with dMef2. The right panel shows an embryo that was sorted selecting for non-GFP expressing embryos, i.e., homozygous mutants, also immunostained with dMef2 antibody. The sorted embryo has no mesodermal staining, verifying that this is a homozygous twist mutant embryo.
 
 

Work Progress Summary 2000:

  We have begun to analyze gene expression throughout the life cycle of Drosophila. This will provide the first complete scan of gene expression for precise stages of development for an animals entire life span. It is an essential baseline study for analyzing mutants, since it provides the key comparison at all the relevant stages. The staging for embryogenesis was monitored We have divided the life cycle into 70 parts and have collected and prepared large quantities of RNA from each time. Each time point has been collected, independently, in triplicate. We will therefore be analyzing data from 210 experiments. The first hybridization of a full set of 70 probes has been completed and data are being collected now. The work is being done by a team from several labs. Participating people from my laboratory are Dr. Eileen Furlong and Brian Null.  Brian has produced microarrays of about 9000 spots of cDNA and EST clones that represent more than half of the Drosophila genes. Work is now underway to produce arrays containing all known and predicted Drosophila genes, and when these arrays are ready the full time course will be analyzed using them. We will eventually have data about more than 13000 genes at 70 time points, an unprecedented view of development. The data will be analyzed during the next few years using a wide range of informatics approaches.

An initial analysis, using Michael Eisenís clustering software, is shown in Fig. 2. The full clustered pattern is shown at the bottom, and a magnified view of a portion of the analysis is shown at the top. Gene expression changes were observed for about 1500 spots comparing the ten time points shown, and the genes were grouped according to which genes exhibited similar fluctuations in time. For example the genes that are highly expressed compared to the reference sample in the early embryonic timepoints show up as red in the top two lines and black or green in the lower lines. These genes include most of the known maternal genes that are transcribed during oogenesis and are contributed by the mother to the egg. Many unknown genes cluster with known maternal genes, so much will be learned about the nature of genes that are active maternally for the embryos benefit. Similarly we see known neural genes expressed as the nervous system is developing, and many unknown genes behave in ways that are similar to the known neural genes. These initial experiments show that our methods are working and informative.

We have also begun to analyze the effects of mutations on gene expression. In one initial study, we have compared embryos at three stages that have, or do not have, muscles developing.  We used twist mutants in comparison to wild-type embryos (Fig. 3), using the embryo sorter to isolate adequate quantities (~12000) of mutant embryos.  Each sample is compared to a reference sample which normalizes all the experiments. The reference sample is a mixture of RNA samples from many stages.  The results showed about 80 spots out of 9000 decreasing expression in the mutants and about 30 spots increasing expression, and we are now analyzing the genes identified in this analysis. We have also done the experiment in complementary fashion, hyperactivating twist function and therefore muscle development and seeing what genes change in those conditions. We have monitored a large set of control muscle and neural genes to see whether their expression amounts change as expected, and they do. Therefore the unknown genes that are changing will provide meaningful information. Our plan is to extend these analyses to many other mutants and compare the flow of genetic information in mutants and wild-type animals.