Abstract
The purpose of this communication is to clarify and correct some of the statements published in the article, “Immunocytochemical demonstration of αs1- and β-casein in mouse mammary glands at early stages of pregnancy” by T. Kanazawa and K. Kohmoto in the Journal of Histochemistry and Cytochemistry 52: 257–264, 2002. These authors state in the Summary that they were able to localize the caseins to the rough endoplasmic reticulum and the Golgi apparatus in immunocytochemically stained paraffin sections from lactating mammary glands. These two cell organelles are not resolved in Bouin's-fixed paraffin sections viewed under the light microscope. This requires preparation of the tissue for electron microscopy, as shown in the review article they cite in support of their interpretation (Burgoyne and Duncan 1998). Furthermore, in the Introduction the authors state, “mammary epithelial cells in vivo need to experience pregnancy and parturition to differentiate functionally.” This statement is completely without support in the scientific literature. In fact, functional differentiation, as defined by the synthesis of milk proteins in mammary epithelial cells, was demonstrated in vivo over 40 years ago in ovariectomized, hypophysectomized, and adrenalectomized female mice by the injection of insulin, hydrocortisone, and prolactin (Nandi 1958). This observation was later confirmed in explant cultures from unprimed nulliparous female mice placed in serum-free medium, but only in the presence of insulin, hydrocortisone, and prolactin (Juergens et al. 1965; Turkington et al. 1965; Stockdale and Topper 1966). Subsequent experimentation too vast to enumerate here has supported these observations, and it is a widely accepted tenet that all three of these hormones are required and sufficient for a lactogenic response in mammary epithelial cells, whether in vivo or in vitro. Topper's group demonstrated that insulin was essential over 20 years ago (Bolander et al. 1981). The authors dismiss these wide-ranging and multidisciplinary reports by stating “it was not clear whether these three hormones are truly sufficient for inducing differentiation in mammary epithelial cells.” Subsequently, they refer to their own unpublished work, which demonstrates induction of casein in cultures (serum-free) from cycling nulliparous mice with these three hormones to provide a straw-man hypothesis to justify repeating the work of others with monoclonal antibodies to mouse α- and β-casein.
In the Discussion, the authors once again chose to ignore the published literature that convincingly demonstrates that mammary epithelial cells from unprimed cycling nulliparous mice can be induced to synthesize caseins as well as other milk proteins (Vonderhaar et al. 1973) both in vivo and in vitro. They misquote the in vitro results of Tonelli and Sorof (1982) and of Durban et al. (1985) to support the contention that their unpublished demonstration of casein synthesis in primary mammary epithelial cells in serum-free medium is a seminal discovery and justifies the technical improvements for casein detection which they allege that their monoclonal antibodies provide. Furthermore, their claim that three phases of casein synthesis occur in the mammary epithelium during pregnancy is based on the presence of immunologically detectable intracellular casein and discounts the movement of the casein into the luminal space. In early pregnancy at 0–4 days, the acinar lumina are small and contain no secretory product. Subsequently, at days 6–10, the acinar lumina become distended with secretion and the epithelial cells themselves less secretory in appearance. At 14 days of pregnancy both lumina and epithelial cells are distended with secretory product. It has been shown that progesterone levels rise during pregnancy and prevent the translation of casein mRNAs (Guyette et al. 1979; Rosen et al. 1980). Therefore, this observation is also not novel.
Finally, in our 1981 paper (Smith and Vonderhaar 1981), we demonstrated by immunoelectron microscopy that casein was present within the cisternae of the rough endoplasmic reticulum in the epithelial cells within explants incubated with insulin, hydrocortisone, and prolactin but not in control cultures, in excellent agreement with our immunohistochemical results carried out on the same samples. Our immunohistochemical analysis of casein synthesis induction was carried out on deplasticized 1-μm sections by immunostaining methodology (Hogan and Smith 1982) essentially identical to that reported in the Kanazawa paper under discussion. The anti-casein antibody used was characterized in Smith et al. (1984) and was shown to immunoprecipitate only casein(s) from mouse skim milk.