Abstract. Two international reference preparations, the First International Reference Preparation of Human Pituitary Gonadotrophins (FSH and LH/ICSH) for Bioassay (identified hereafter by its code no.: 69/104) and the First International Standard of Urinary FSH and LH (ICSH) for Bioassay (hereafter: hMG 1st IS) and an aqueous extract of human pituitary glands (hPE) were fractionated in triplicate by isoelectric focusing on sucrose density gradient (Ampholine, pH range 2.5–10.0). The FSH activity was monitored in each fraction by an in vitro bioassay, a radioimmunoassay and a receptor binding assay. Unfractionated 69/104 was used as standard in each assay system. Compared with the hPE, the activity profiles of the 69/104 and hMG reference preparations were spread over a significantly wider pH range; the biological activity eluted in the pH range of 3.5–5.0 was of the order of 90% for hPE, 72% for 69/104 and less than 60% for hMG. Major variations in biological to immunological (B/I) and biological to receptor binding (B/R) activity ratios were observed in the individual electrofocusing fractions. The B/I ratios ranged from 0.8 to 2.2 (69/104), 1.0–5.7 (hMG) and 1.3–6.0 (hPE), respectively. The variation in the corresponding B/R ratios was: 0.5–1.7 (69/104), 0.58–4.0 (hMG) and 0.55–1.5 (hPE). When equal aliquots from each of the electrofocusing fractions were combined and this 're-constituted' pool was compared with the starting material, significant differences were observed in the B/I ratios: 1.63 instead of 1.0 (69/104), 2.58 vs 2.34 (hMG) and 2.72 vs 1.32 (hPE). There was no significant change in the mean B/R ratios before and after electrofocusing: 1.0 vs 1.02 (69/104), 1.09 vs 1.02 (hMG) and 1.05 vs 1.04 (hPE). The mean recovery of biological activity in each of the three reconstituted pools of the three preparations was 97% for 69/104, 88% for hMG and 98% for hPE, and the corresponding recoveries of receptor binding activities were 95%, 94% and 98%, respectively. In contrast, the mean recovery of immunological reactivity was only 55% (69/104), 80% (hMG) and 47% (hPE), respectively. The reduction in the immunological reactivity of the 69/104 preparation without any apparent loss of biological or receptor binding activities following electrofocusing indicates that its immunoreactivity is unstable under mild experimental conditions, which do not influence its biological and receptor binding activities. Hence, whereas this preparation might possibly be suitable as a standard for in vitro bioassays and receptor binding assays, it is unsuitable as a standard for the radioimmunoassay of hFSH.

Abstract. The LH potencies of 12 preparations of highly purified human pituitary LH, from 6 laboratories, were estimated by 2 in vivo bioassays and an in vitro bioassay in terms of the International Reference Preparation of Human Pituitary Gonadotrophins (FSH and LH) for Bioassay (coded 69/104); and by immunoassay in terms of the International Reference Preparation of Human Pituitary Luteinizing Hormone for Immunoassay (IRP; coded 68/40). The LH potencies varied between preparations, including the IRP (68/40), from 864 to 5740 IU/mg by seminal vesicle weight gain (SVW) assay; from 1510 to 11500 IU/mg by ovarian ascorbate depletion (OAAD) assay; from 4490 to 14500 IU/mg by in vitro (testicular interstitial-cell testosterone production) bioassay; and from 2030 to 9180 IU/mg by immunoassay. Estimates of protein content were based on the assumption that the absorbance of LH at 280 nm (A 1% 1 cm) was 6.0. The LH potency of most preparations was highest by in vitro bioassay and lowest by SVW assay. The correlation between activities determined by SVW and OAAD assays was more marked than that between estimates by OAAD assay and in vitro bioassay; there was no correlation between estimates by SVW assay and in vitro bioassay. The slopes of the log dose-response curves of preparations in the OAAD assay were positively correlated with their potencies by OAAD assay and negatively correlated with the slopes of their log dose-response curves in the SVW assay. The qualitative differences between preparations are considered to be a reflection of the heterogeneity of LH and of its modification by different purification procedures. The present data, together with the different patterns of heterogeneity found in some of these preparations by isoelectric focusing in a separate study, suggest that the more basic molecular forms of LH, which are preferentially purified during the isolation of LH free from FSH and TSH, have shorter plasma survival times than the more acidic forms. The LH immunoreactivities of all preparations were significantly correlated with their potencies estimated by each of the in vivo bioassays but not with those estimated by in vitro bioassay. The ratios of in vitro bioactivity (in terms of IRP (68/40)): immunoreactivity varied between preparations from 0.53–1.5. The FSH content of each preparation was less than 2% (w/w) by bioassay and immunoassay. Most preparations were more potent by in vitro bioassay than by in vivo bioassay, which contrasted with, and complemented, findings for purified FSH preparations. This indicated that, as in the case of LH, the more basic molecular species of FSH are associated with lower ratios of in vivo: in vitro bioactivity than are the more acidic species. This study provides the most comprehensive comparison available of the activities of purified preparations of LH isolated from frozen and acetone-dried human pituitary glands in different experienced laboratories. These data are needed for selecting material for an international reference preparation of LH for immunoassay on the basis of high LH potency by in vivo bioassay, recommended by the WHO as a criterion for the identity of the hormone and for its freedom from contaminants. The consequences of the heterogeneity of LH are considered for the purification of the reference material and for the suitability of the latter for the various types of specimens which require LH assays.

A single bolus of 100 \ g=m\ g of gonadoliberin (LRH) was administered intravenously to 8 post-menopausal and 9 normally menstruating women and blood was withdrawn before and 30, 60, 120 and 240 min after LRH stimulation. The plasma samples obtained at different time intervals from women showing a sufficiently high response to LRH (menopausal: 8, menstruating: 3) were combined and 2 ml samples of each pool were fractionated in triplicate by electrofocusing on sucrose density gradient. In addition, two plasma pools, obtained 30 min following LRH stimulation, one from 4 normally menstruating women (exhibiting a relatively low LH-response) and the other from 2 normally menstruating women aged 40, were analyzed in the same way in duplicate electrofocusing experiments. The hLH activity was determined in each electrofocusing fraction by an in vitro bioassay method following elution and purification by gel filtration. The LH activity was distributed as four major peaks at pI values of 7.10 \m=+-\0.05,7.58 \m=+-\0.06, 8.10 \m=+-\0.04and 8.54 \ m=+-\0.05 and a broad area of activity comprising a number of peaks in the pH range of 8.69\p=n-\9.50.The analysis of the data revealed marked differences in the relative distribution of the various molecular species present in the blood of menopausal women and of normally menstruating women. A molecular species exhibiting a pI value of 7.10 was invariably present (10\p=n-\15% of the total) in all samples of post-menopausal plasma (PMP) but was consistently absent from all samples of midcycle plasma (MCP). The amount of relatively 'less alkaline' material (eluted from pH range 7.37\p=n-\8.32)was significantly higher (P < 0.001) in the PMP samples compared to MCP samples. On the other hand, in the MCP samples the amount of relatively 'more alkaline' material eluted from the pH range 8.33\p=n-\9.50 was significantly (P < 0.001) higher (about 60% of the total recovered activity) compared to the PMP samples (about 30% of the total). Following LRH stimulation significant temporal changes were observed in the relative contribution of various molecular species to the hLH profile. A gradual increase, up to 60 min, in the material eluted in the pH range 6.87\p=n-\7.36in the post-menopausal plasma samples was accompanied by a gradual decrease in the material eluted in the pH range 7.84\p=n-\8.32.Two hours after LRH stimulation a significant drop was found in the material collected from pH range 8.33\p=n-\8.68,with a concomitant rise in the material eluted in the pH range 8.69\p=n-\9.50. This last mentioned shift was also observed in the plasma of normally menstruating women. It is concluded that major differences exist in the composition of biologically active hLH species present in the peripheral blood of post-menopausal and normally menstruating women. Moreover, significant temporal changes occur in the composition of circulating hLH species following stimulation by LRH both in post-menopausal and in normally menstruating women.

Eight highly purified human pituitary FSH preparations and purified preparations of the α-and β-subunits of FSH were fractionated by an electrofocusing technique in the pH range of 2·5–10·0 on a sucrose density gradient. The human (h) FSH activity in each of the eluted fractions was monitored by an in-vitro bioassay and a radioimmunoassay procedure. After electrofocusing, the overall recovery of the biological activity of the eight preparations was between 80 and 94% (mean 88%). On the other hand, the recovery of immunoreactivity ranged between 30 and 84% (mean 71%). A loss of over 85% of hFSH immunoreactivity was observed when the α- and β-subunits of hFSH were fractionated by the same procedure. The specific loss of varying amounts of immunoreactivity in all preparations during electrofocusing was also reflected by a proportional increase in the ratios of biological activity (B) to immunoreactivity (I); preparation A, which exhibited a loss of 70% of the immunoreactivity, had a threefold increase in its B/I ratio after electrofocusing. Significant differences were observed in the electrofocusing profiles of the eight preparations both in terms of their pI values and of their spread. The disparity in the relative distribution of hFSH activities in different pH regions suggested major differences in the carbohydrate moieties (sialic acid content) of the preparations studied, probably as a result of the chemical manipulations involved in the purification of the hormone. It is suggested that a combination of several (but certainly not all) of the preparations might serve as a provisional International Reference Preparation for hFSH radioimmunoassays.

No abstract