Chromosome counting in anther filaments following DAPI staining
-based on the protocol and advice of Greg Copenhaver and Daphne Preuss (University of Chicago). It works for Arabidopsis thaliana and other Arabidopsis species.

2. After fixation, neutralize the sample in 50mM potassium phosphate buffer pH 7.0. You can do this for the whole flower in the microfuge tube but we prefer to first remove just the fixed anther filaments with sharp forceps and place them on a slide (don't let them dry out!) and cover them with a generous amount of neutralization buffer (using a transfer pipette).

3. Remove the buffer with a fine tip transfer pipette then pipette ~20 ul of DAPI stain (10 ng/ microliter DAPI in 80% glycerol) onto the filaments and place a cover-slip on top (The DAPI stock is a 1 mg/ml stock solution in water and the stain one uses is a 1:100 dilution made into 80 % glycerol). The glycerol helps prevent drying by evaporation so that you don't really need to seal the edges of the coverslip (but if you do seal them for longer term storage, we use clear nail polish painted along the edge of the coverslip where it meets the slide)

4. Observe the filaments using an appropriate DAPI filter and a fluorescence microscope-we use a 60 X or 100X oil immersion objective and 10X eyepiece. The cells you want to look for are elongated and the nuclei inside them will be elongated as well. Chromosomes (actually centromeric heterochromatin) will be visible as brightly staining DAPI spots within the nuclei. You will need to focus up and down to see all the spots because they are not all in the same focal plane. If your microscope can do it, optical sectioning followed by computer reconstruction of the complete image (3D projection) is useful. At Wash U. we use the computational optical sectioning microscope, but confocal would also work great. The actual counting of chromosomes is a matter of statistics-you won't be able to count every chromosome in every cell. Diploids are clearly discernable from triploids or tetraploids. Aneuploids are a bit tougher. The problem is that sometimes two spots are coincident. If you are trying to detect the absence or presence of one or a few chromosomes you will need to count spots in many cells and calculate a mean. Some anther filaments seem to be better than others. There may be a developmental component to this if at some point the cells and the nuclei may become flatter and more elongated.