003 I-399818712 104-0409$02.00/0 PEDIATRIC RESEARCH Vol. 21, No. 4, 1987 Copyright O 1987 International Pediatric Research Foundation, Inc I'rfnied in U.S. A. Pulsatile Release of Bioactive Luteinizing Hormone in Prepubertal Girls: Discordance with Immunoreactive Luteinizing Hormone Pulses EDWARD 0. REITER, DARLENE E. BIGGS, JOHANNES D. VELDHUIS, AND INESE Z. BEITINS Department of Pediatrics, Baystate Medical Center, SpringJield, Massachusetts 01 199 [E.O.R., D.E.B.]; Department of Medicine, University of Virginia Medical School, Charlottesville, Virginia 22908 [J.D.V.]; and Department qf Pediatrics of the University of Michigan Medical Center, Ann Arbor, Michigan 48109 [I.Z.B.] ABSTRACT. An assessment of pulsatile secretion of lu- the onset of puberty is an increase in the frequency, as well as teinizing hormone (LH), measured by both immunoassay amplitude, of GnRH pulses from the hypothalamus which is (I-LH) and rat interstitial cell testosterone production translated into gonadotropin secretory pulses from the pituitary. bioassay (B-LH), as well as of folicle-stimulating hormone Increased LH pulse frequency and amplitude, as measured by I- and glycoprotein hormone a-subunit was carried out in LH have been demonstrated in early pubertal children (1-4). In seven normal prepubertal and six normal premenarcheal addition, the mean I-LH levels have been shown to rise with pubertal girls. Samples were obtained at 20-min intervals advancement of puberty in both cross-sectional and longitudinal for a 6-h period. The hormone secretion profiles were studies (5, 6). analyzed by several computerized methods yielding pulse Pituitary gonadotropins are known to be heterogeneous. frequency and amplitude, interpulse basal levels, and per- Therefore, in addition to changes in the quantity of gonadotro- centage increments, with bio/immuno ratios calculated for pins, as measured by immunoassays, the quality of gonadotro- peak and basal concentrations. In these prepubertal girls, pins as measured by in vitro bioassays and on isoelectric focusing mean B-LH levels were 12% of I-LH, with B/I ratio of have been shown to change (7-10). In general the B-LH levels 0.13; 30% of samples were below assay sensitivity (0.10 in serum have exceeded those of I-LH in both cross-sectional mIU/ml) for B-LH, but all I-LH (1.25 mIU/ml) were (I I) and longitudinal studies during puberty in both boys and detectable. In the pubertal group, B-LH levels were 30% girls (12). In mature men and women the B/I ratio of LH is of I-LH, with mean B/I ratio of 0.24 and undetectable B- increased during endogenous LH pulses and there exists an LH in 29% of samples. Pulsatile secretion in prepubertal overall concordance between the B-LH and I-LH pulses in excess girls was found in five of seven (11150 min) for B-LH and of 80% (7, 8). six of seven (11212 min) for I-LH; only two of six pubertal In this study we examined and compared the LH pulse char- girls had detectable pulses. Discordance of B- and I-LH acteristics in normal prepubertal and pubertal girls. In particular, pulses were frequent, with 56% of B-LH pulses lacking an we wished to establish whether 1) B-LH was secreted in episodic I-LH pulse and 47% of I-LH pulses not having a B-LH pulses that exceeded the amplitude of I-LH pulses thereby mak- pulse. These data demonstrate that 1) both B- and I-LH ing detection more accurate and 2) the B- and I-LH was secreted are secreted episodically in prepubertal girls; 2) I-LH-like synchronously. material is present in higher concentrations than B-LH in these girls; and 3) substantial discordance of B- and I-LH pulses exist. Such findings lend further support for molec- MATERIALS AND METHODS ular heterogeneity of pituitary gonadotropins during the Subjects. Seven healthy, normal prepubertal girls and six pre- pubertal process and suggest that changes of biopotency menarcheal pubertal girls were recruited. The prepubertal girls may be important functional regulators of the early puber- were Tanner breast stage 1 and had a mean age of 8.6 f 0.2 tal process. (Pediatr Res 21: 409-413, 1987) (SEM) yr. The premenarcheal girls had either Tanner breast stage I1 (n = 4) and were 10.7 + 0.8 yr old or Tanner breast stage III/ Abbreviations IV and were 12.0 f 3.8 yr old. All had normal developmental LH, luteinizing hormone and health histories, physical examinations, and were not taking FSH, follicle-stimulating hormone medications. RIA, radioimmunoassay(s) The study had been approved by the Human Subjects Com- B-LH, bioactive luteinizing hormone mittee of Baystate Medical Center and details were explained to I-LH, immunoreactive luteinizing hormone the children and their parents. The volunteers and their parents GnRH, gonadotropin-releasing hormone signed an informed consent form before participation in the study. study protocol. All of the girls had an intravenous needle placed in a forearm vein 30 min before the sampling was started. Subjects were resting comfortably and did not appear anxious or The mechanisms responsible for the initiation and tempo of distressed. Blood samples were collected every 20 min for a total puberty remain elusive. It is currently believed that central to of 360 min. The sera were separated and kept at -20" C until assayed. Received September 8, 1986; acccptcd Dcccmbcr 2, 1986. Assays. In each sample LH, FSH, and glycoprotein hormone Reprints Edward 0. Reiter, M.D., Dcpartrnent of Pediatrics, Baystate Medical a-subunit were measured by specific RIA the characteristics of Center, 759 Chestnut Street, Springfield, MA 01 199. Supported by a research grant from Baystate Medical Center (E.O.R., D.E.B.) which have been previously published (l3). B-LH was measured RCDA-IK04I-ID00634 (J.D.V.), and HDI 85 12 (I.Z.B.). by means of the rat interstitial cell assay (1 1, 13) in each sample. 409 410 REITER ET AL. All of the samples from each subject were included in one assay. sampling period, and the mean level for the entire sampling. Integratcd serum pools from the whole study were used for the Smoothed overall baseline concentrations to correct for presence measurement of serum E2 and estrone sulfate by means of of pulses were estimated by a weighted moving average across sensitive and specific RIA (14). all the data series using a program of Merriam and Wachter (16). The assay for serum I-LH was a double antibody RIA utilizing The method of Clifton and Steiner (17) which performs iterative iodinated LER960 and IRP-2 hMG (Second International Ref- data scans to identify significant pulses exceeding a threshold erence Preparation of Human Menopausal Gonadotropin) as the value which undergoes continuing adjustment to avoid missing reference standard. The same standard was used for the estima- true pulses but minimizing the false influence of assay noise was tion of serum I-FSH and B-LH. The respective sensitivities of also utilized. This latter analytic technique was used to check the the I-LH and B-LH assays were 1.25 and 0.1 mIU/ml, respec- modified methodology of Santen and Bardin (15) and demon- tively. The a-subunit was measured by means of a specific RIA strated high concordance (approximately 85%) of pulse desig- which had a sensitivity of 37 pg/ml using LER 1793 for assay nation. standard and LER 1946 for iodination material (1 3). Pulse data are given as mean + SE. Comparisons between I- Stati.stica1 ana1yse.s. The serum LH secretion profiles were and B-LH values were analyzed by within-subject comparisons analyzed by several methods. We modified the program of using a paired two-tailed Student's t test. I- and B-LH pulses Santen and Bardin (1 5) setting a threshold for minimum increase were considered concordant whenever the pulses overlapped in from nadir to peak of more than four times the pertinent time. individual intraassay coefficient of variation as the requirement for definition of a pulse. Such a criterion reduces the false- RESULTS positive rate to approximately 1 %. The intraassay coefficients of variation used for this analysis were: 8.1% for B-LH over the Subjects. The group of seven prepubertal girls have to be range of 2 to 8 mIU/ml, 16.3% over the range of 1 to 2.0 mIU/ considered "late" prepubertal. In a previous study we character- ml, and 18.3% over the range of 0.1 to 1.0 mIU/ml; 9.7, 8.1, ized B-LH in 15 prepubertal girls who were 6.2 + 0.8 yr old with and 10.1% for the LH, FSH, and a-subunit immunoassays in serum E2 levels of 3.9 f 0.4 pg/ml (I I). Although the present the relevant range for the potency estimates made in sera from group of girls also had Tanner breast stage I on physical exami- these girls. nation, they were older (mean of 8.6 yr) and had significantly The output from the Santen and Bardin (15) program also higher serum E2 levels [7.7 f 0.1 (SE) pg/ml (p< 0.001)]. Serum provides percentage increments of pulses (nadir-to-peak in- estrone sulfate levels were also higher (1 67 + 1 l versus 97 + pg/ creases), integrated area under the gonadotropin curve over the ml) in the older group. Table 1. Characteristics ofpulsatile secretion of bioactive LH relative to irnrnunoreactive LH in prepubertal girls Amplitude characteristics Mean Area Pulse frequency* Interpulse Basal Basal E2 Subiect (mIU/ml) (mIU/ml. min) (no.16 h) Peak % Increment (mIU/ml) B/I B/I (~g/ml) I B-LH 0.73 248 3 1.3 118 0.67 0.2 1 0.2 1 8 I-LH 3.5 1126 1 4.6 7 1 3.2 L B-LH 0.36 I-LH 2.89 3 B-LH 0.50 I-LH 5.84 4 B-LH 0.63 I-LH 2.90 5 B-LH 0.63 I-LH 3.96 6 B-LH 0.10 I-LH 2.7 1 7 B-LH 0.10 IND I-LH 3.8 1 IND Mean rt SEM B-LH 0.43 + 0.10 I-LH 3.6 + 0.4 * Frequency in whom significant pulses were detected.