Previous investigators have disagreed about whether hemalum stains DNA or its associated nucleoproteins. I review here the literature and describe new experiments in an attempt to resolve the controversy. Hemalum solutions, which contain aluminum ions and hematein, are routinely used to stain nuclei. A solution containing 16 Al ions for each hematein molecule, at pH 2.0-2.5, provides selective progressive staining of chromatin without cytoplasmic or extracellular "background color." Such solutions contain a red cationic dye-metal complex and an excess of Al ions. The red complex is converted to an insoluble blue compound, assumed to be polymeric, but of undetermined composition, when stained sections are blued in water at pH 5.5-8.5. Staining experiments with DNA, histone and DNA + histone mixtures support the theory that DNA, not histone, is progressively colored by hemalum. Extraction of nucleic acids, by either a strong acid or nucleases at near neutral pH, prevented chromatin staining by a simple cationic dye, thionine, pH 4, and by hemalum, with pH adjustments in the range, 2.0-3.5. Staining by hemalum at pH 2.0-3.5 was not inhibited by methylation, which completely prevented staining by thionine at pH 4. Staining by hemalum and other dye-metal complexes at pH ≤ 2 may be due to the high acidity of DNA-phosphodiester (pK ~ 1). This argument does not explain the requirement for a much higher pH to stain DNA with those dyes and fluorochromes not used as dye-metal complexes. Sequential treatment of sections with Al(SO) followed by hematein provides nuclear staining that is weaker than that attainable with hemalum. Stronger staining is seen if the pH is raised to 3.0-3.5, but there is also coloration of cytoplasm and other materials. These observations do not support the theory that Al forms bridges between chromatin and hematein. When staining with hematein is followed by an Al(SO) solution, there is no significant staining. Taken together, the results of my study indicate that the red hemalum cation is electrostatically attracted to the phosphate anion of DNA. The bulky complex cation is too large to intercalate between base pairs of DNA and is unlikely to fit into the minor groove. The short range van der Waals forces that bind planar dye cations to DNA probably do not contribute to the stability of progressive hemalum staining. The red cation is precipitated in situ as a blue compound, insoluble in water, ethanol and water-ethanol mixtures, when a stained preparation is blued at pH > 5.5.
