On Thu, 6 Apr 2000, Gayle Callis wrote:

> The periodic acid oxidation was recommended as a way to POSSIBLY improve
> nuclear staining ...  with hematoxylin, and maybe making the nuclei have 
> darker staining. In my experience, that never seemed to work.
> ... Trying to recall the exact chemistry of oxidizing the
> nucleic acids, vicinal OH groups on the sugar, going to CHO groups, but
> with overoxidation, creating more COOH groups available to the hematoxylin
> stain?  John, comments here.  
------------------
  If I'm the appropriate John, then here goes. This will probably
  be another of my long, rambling communications. Like every other
  internet message, it is not peer-reviewed and therefore carries 
  no scientific clout. It's just a bit of junk-food for brains that
  might like this sort of stuff. 
   
  The use of periodic acid to revive nuclear staining is also
  described in the AFIP Manual (Luna, 1968). The rationale is not
  obvious. Periodic acid doesn't react with the sugar in DNA (no
  adjacent OH groups in deoxyribose). Alum-haematein does combine
  with DNA in vitro, and staining of nuclei by very dilute solutions
  requires the presence of DNA. Ordinary haemalum solutions stain
  nuclei perfectly well after chemical or enzymatic extraction of
  all the DNA (which abolishes all nuclear staining by basic dyes).
  After DNA extraction chromatin should consist only of strongly
  basic proteins. This assertion is confirmed by the enhanced
  stainability with alkaline solutions of acid dyes. This acidophilia
  is prevented by chemical blockade of the strongly basic guanidino
  side-chain of the amino acid arginine. Arginine blockade also
  prevents staining by ordinary alum-haematoxylin mixtures. 

  None of this explains the resuscitation of nuclear staining
  by treatment with periodate. It's conceivable that periodate
  might oxidize adjacent -OH and -NH2 groups (it does do this)
  in compounds formed by reaction of guanidino groups with
  formaldehyde over long periods of time. This sort of speculation
  isn't very constructive! Gayle's observation that the method
  doesn't work is good enough for me.
------------------------- 
  The important thing people should know about haematoxylin (meaning
  haemalum: the Al complexes of haematein) is that it is _not_ a basic
  dye, and stained nuclei or other things should not be called
  "basophilic."  A combination of a blue basic dye with eosin (any
  blood stain, for example) provides much more informative general
  oversight staining than H & E. All this was pointed out by RD Lillie
  in the 1950s, and he went on to develop azure-eosin methods for use
  on machines. Why are pathologists and others so sold on H & E as
  the routine method, when better and more easily performed methods
  have been available for so long? 

  If you are interested and want to form your own opinion, I can 
  provide proper published references for the assertions in the
  above paragraphs. Just ask, and I'll send a reading list. 
------------
  Gayle's comments about the Gerrits/Horobin paper and staining
  sections of GMA-embedded material were very timely.  All who
  do such staining should read that paper (or at the very least
  it's abstract on PubMed). Thanks for bringing it to everyone's
  attention.

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