Zerfaoui M, Ronin C
UPR 9024 CNRS, Marseille, France.
Eur J Clin Chem Clin Biochem. 1996 Sep;34(9):749-53.
Glycoprotein hormones have long been known to display extensive polymorphism and changes in bioactivity according to the endocrine status of the patient. Structural analysis has shown that pituitary gonadotropins (lutropin and follitropin) and thyrotropin are synthesized and secreted as a panel of isoforms which differ in glycosylation, bioactivity and circulatory half-life. Ultrasensitive immunoassays could reveal that glycosylation of plasma hormones is structurally different from the pituitary stock so that the ratio of circulating glycoforms may vary according to the physiopathology of the pituitary axis. However, contradictory results between immunoassays have been often reported, suggesting that some plasma forms can escape recognition by monoclonal antibodies which have been raised to the pituitary or urinary antigen. When hormone levels do not correlate with clinical features, one can also suspect that inactive or hyperactive forms are being measured. At the molecular level, very limited information has been gained toward the expression of hormone epitopes as a function of carbohydrate structure. To address this issue, we have compared the recognition of pituitary and recombinant human thyrotropin by various polyclonal and monoclonal antibodies before and after neuraminidase treatment. Both, pituitary and recombinant thyrotropin bound to anti-alpha and anti-beta antibodies, demonstrating thereby that recombinant thyrotropin can be used to calibrate immunoassays. While removal of sialic acid did not alter the recognition of the recombinant hormone in various immunoassays, this treatment specifically abolished the binding of pituitary thyrotropin to anti-beta monoclonal antibodies. These findings show that immunoreactivity of circulating hormone glycoforms, which are often more sialylated than their pituitary counterparts, may very well account for variation depending on the antibodies used in the immunoassays.
长期以来，人们一直知道糖蛋白激素具有广泛的多态性，并根据患者的内分泌状态改变生物活性。结构分析表明，垂体促性腺激素（促黄体生成素和促卵泡生成素）和促甲状腺激素是以一组同工型的形式合成和分泌的，这些同工型在糖基化、生物活性和循环半衰期方面存在差异。超灵敏免疫测定可以揭示血浆激素的糖基化在结构上与垂体储备不同，因此循环糖型的比例可能根据垂体轴的生理病理学而变化。然而，免疫测定之间经常报道相互矛盾的结果，这表明一些血浆形式可能无法被针对垂体或尿液抗原产生的单克隆抗体识别。当激素水平与临床特征不相关时，人们也可以怀疑所测量的是无活性或活性过高的形式。在分子水平上，关于激素表位作为碳水化合物结构的函数的表达所获得的信息非常有限。为了解决这个问题，我们比较了在神经氨酸酶处理前后，各种多克隆和单克隆抗体对垂体促甲状腺激素和重组人促甲状腺激素的识别。垂体促甲状腺激素和重组促甲状腺激素都与抗α和抗β抗体结合，从而证明重组促甲状腺激素可用于校准免疫测定。虽然去除唾液酸并没有改变重组激素在各种免疫测定中的识别，但这种处理特异性地消除了垂体促甲状腺激素与抗β单克隆抗体的结合。这些发现表明，循环激素糖型的免疫反应性，其唾液酸化程度通常比垂体对应物更高，很可能解释了取决于免疫测定中使用的抗体的差异。
