Cytochemical methods of hormone assay LUCILLE BITENSKY

Initially polypeptide hormones were assayed by their biological activity. Such conventional bioassays, however, are remarkably insensitive. For example, the Lipscomb and Nelson (1962) bioassay of corticotrophin (ACTH) cannot measure a concentration of less than about 100 pg/ml, while the normal circulating levels of the bioactive form of this hormone range from about 5 to 60 pg/ml. The introduction of radioimmunoassay increased the sensitivity with which hormones could be assayed quite remarkably, that of ACTH being able to measure as little as 10 pg/ml. Moreover, the number of assays which could be done in unit time by radioimmunoassay was far greater than could be achieved by conventional bioassay. The fact that radioimmunoassay fundamentally measures the number of specific molecules (or antigenic determinants) present rather than the functional activity of the hormone might be regarded as an improvement on bioassay, representing a precise chemical rather than a biological analytical tool. However, so long ago as 1967 it was recognised that immunoassays measure a composite of antigenic activity which is not necessarily related to the biological activity of the hormone. A special committee, convened by the World Health Organization, noted a need for 'micro-bioassays' which would have at least the same sensitivity as radioimmunoassay and which could be done in parallel with it (World Health Organization, 1975). Since then reports concerning the dissociation of immunoactivity and bioactivity (for example, Besser et al., 1971) have increased alarmingly. Thus gastrin has been shown to occur in a number of forms which show various biological activities (Yalow, 1974). Apart from this problem of assaying molecules which may have little or no biological activity, the sensitivity of radioimmunoassay may be insufficient to measure low normal circulating levels of some hormones or pathologically low levels that might occur in some diseases. Thus the population survey of Tunbridge et al. (1976) and the other studies of Petersen et al. (1975) clearly show that the circulating level of TSH in a considerable proportion of normal people may be below 1.0 ,uU/ml or even below 0.5 ,uU/ml. Yet this is the sensitivity level of even the best radioimmunoassays currently available for this hormone. Therefore there is clearly a need for a functional (that is, biological) assay system at least 10 times more sensitive than the current radioimmunoassays. Such bioassays could be performed in parallel with the more conventional radioimmunoassays to test whether the molecules detected by the latter were functional. Their greater sensitivity would also be of value when only small samples were obtainable, as in investigations ofhormones in neonates (Holdaway et al., 1973) or when many serial samples were required (Daly et al., 1974a), and when the hormonal levels were below the limits of precise measurement by radioimmunoassay.