With regard to the use of a trichrome technique to stain
  apoptotic nuclei and their fragments (apoptotic bodies),
  it would be sensible to do the method on sections of at
  least two different thicknesses and determine whether the
  abundance of apparently apoptotic cells (as a percentage
  of the total number of cells counted) changes with the
  thickness of the section. If this percentage does not vary
  with section thickness, the specificity of the technique
  would be supported.

  The reason for doing this control (or test) is that in
  entirely normal tissues it is possible to see nuclei of
  different colours in cells stained with certain mixtures
  of anionic dyes. In the distant past it was thought that
  this might indicate some functional differences or cyclic
  activities of the cells.  Then in 1955 L. Lison published
  a paper called "Staining differences in cell nuclei" in
  the Quarterly Journal of Microscopical Science 96: 227-237.
  He stained sections of rat's liver at several thicknesses
  from about 4 to 15 um with a mixture of aniline blue,
  orange G and PTA (as used in Mallory's method, or similar).
  In the thinnest sections all the nuclei were blue, and in
  the thickest about 40% were blue and 60% orange. The
  biggest jump in percentage was between thicknesses of 6 um
  and 10 um.

  Lison's deduction was that aniline blue anions could enter
  and stain only those nuclei that had been cut through by
  the microtome knife, and could not enter intact nuclei that
  were contained within the thickness of the section. Orange G
  (much more compact anions, and only about 1/4 the size of
  aniline blue) was able to penetrate and stain whole, uncut
  nuclei. Lison was one of the fathers of histochemistry. His
  book "Histochimie Animale" (1st edn was 1936) is one of the
  classics of the subject. Does anyone know if he's still
  around? The last I heard was about 1980 when he was an
  emeritus professor in Brazil - Sao Paulo, I think.

  This suggestion is not a criticism of the method published
  in December's J Histotechnol, which may well be a perfectly
  valid stain for apoptotic bodies and nuclei. The printed
  pictures certainly look convincing. It's just that the
  penetration and retention of dyes applied from mixtures is
  a complicated business, and is affected by section
  thickness as well as by fixation and other pre-treatments
  of the tissue. Apoptotic nuclei are small (pyknotic), and
  apoptotic bodies are so tiny that quite a high proportion
  will be entirely contained in a thin paraffin section.

  It's worth remembering that the only certain way to see
  apoptosis is by electron microscopy. Close LM examination
  (any nuclear stain; most often a haemalum) is 2nd best.
  The expensive methods - TUNEL and in situ nick translation -
  are less certain (because they also stain the later
  remains of cells that have died by necrosis), but they have
  the advantage of not colouring normal nuclei or other
  "background" objects. Such methods appeal to younger research
  workers who have not studied histology, feel uncomfortable
  with a microscope, and expect a computer to identify and count
  everything under the objective.  The remarks in this paragraph
  apply to sections, not to cell cultures.

  Two recent, good reviews of methods for detecting apoptosis are:
    Willingham, MC 1999. J Histochem Cytochem 47 (9), 1101-1109.
    Huppertz, B et 2 al 1999. Anat Embryol 200, 1-18.

 John A. Kiernan,
 Department of Anatomy & Cell Biology,
 The University of Western Ontario,
 LONDON,  Canada  N6A 5C1