1
Endocrinology Printed in U.S.A.
Copyright © 1997 by The Endocrine Society
Reproductive Sciences Program (V.P., K.M., F.J.K., A.R.M.) and the Departments of Pediatrics (V.P.),
Physiology (F.J.K.), and Biostatistics (D.T.M.), University of Michigan, Ann Arbor, Michigan
48109-0404
424
FSH SECRETORY DYNAMICS 425
Materials and Methods identified; pulses beginning within 10 min (two samples) of each other
Experimental design were considered to be concordant. The average lag time between con-
cordant pulses was also estimated for each ewe. In addition, to assess
For determination of secretory patterns of FSH, blood samples col- the overall temporal relationships between hormone patterns, the cross-
lected from the peripheral and hypophyseal portal circulations of ovari- correlation for the above variables was calculated at different time lags
ectomized ewes from a previous study (14) were used. Details of the (autocross-correlation). The time lag that yields the highest cross-cor-
surgical procedures, hypophyseal portal sample collection, and circu- relation is an estimate of the overall time lag between two series. The
lating patterns of GnRH in hypophyseal portal blood and LH in both the average pulse lag time simply estimates the temporal relationship be-
peripheral and hypophyseal portal circulations have been described tween concordant pulsatile episodes. Within-animal comparisons, such
previously (12, 14). Briefly, adult Suffolk ewes were ovariectomized and as amplitude of GnRH-associated vs. non-GnRH-associated pulses, were
surgically fitted with an apparatus for collection of hypophyseal portal performed using paired t tests or the nonparametric Wilcoxon signed
blood, and the collection procedure was initiated approximately 1 week rank test.
later, using methods described previously (13). Integrated samples of
FIG. 2. Patterns of hypophyseal portal and jugular FSH patterns from two ovariectomized ewes (no. 1 and 2), both sampled during the breeding
season. Hypophyseal portal LH and peripheral LH patterns (14) from a previous study are provided in the lower panels for comparison. To
understand temporal relationships between FSH and GnRH, GnRH secretory patterns are overlaid (shaded patterns; scale not shown). Asterisks
identify statistically identified pulses of FSH. Arrows indicate the GnRH-associated bursts of FSH that occur on top of a previously triggered
episode of FSH release.
episodic burst of FSH in hypophyseal portal plasma. In con- pophyseal portal and peripheral circulation are summarized
trast, such an association was not evident between GnRH in Table 1. Shown in Table 1A for each ewe are the sampling
and jugular FSH. FSH remained elevated between episodes duration, number of portal and jugular FSH pulses identi-
of release, whether measured at the hypophyseal portal or fied, and number of pulses per h. Previously reported pulse
jugular levels. The interval between the onset of the GnRH numbers of GnRH from the same ewes (14) are provided in
burst and secreted FSH in the hypophyseal portal blood was parenthesis for comparison. Within each series, although the
very short, with initial increases occurring within the same number of LH pulses identified in the hypophyseal portal
5-min sample for both hormones. The subsequent fall in plasma was the same as the number identified for GnRH (12),
secreted FSH was rapid, but slower than that for GnRH. the number of FSH pulses identified in the hypophyseal
Various characteristics of FSH pulses identified in the hy- portal plasma far exceeded that for GnRH. The mean pulse
FSH SECRETORY DYNAMICS 427
amplitude of hypophyseal portal FSH is shown in Table 1B. in the periphery (7.8 6 2.3 vs. 26.2 6 5.5 min for hypophyseal
Portal FSH pulses were 18.0 6 5.5-fold greater in amplitude portal and peripheral FSH, respectively; Table 1B).
than jugular FSH pulses. Further, GnRH-associated pulses of Table 2 summarizes average concentrations of FSH in the
FSH were larger (P , 0.01) than non-GnRH-associated pulses hypophyseal portal and peripheral circulations during the
of FSH (57.7 6 10.8 vs. 43.4 6 11.4 ng/ml, respectively; Table 6-h collection period and the relative proportions of the basal
1C). Such a relationship was not evident when the analysis and episodic components. The total FSH level measured in
was conducted on jugular FSH pulses (3.6 6 0.8 vs. 3.2 6 0.3). hypophyseal portal blood during the 6-h collection period
The disappearance time of FSH in the hypophyseal portal was approximately 21-fold greater than that measured in
plasma, although not as rapid as that reported for GnRH peripheral blood (2919.8 6 1023.5 vs. 138.9 6 38.2 ng/ml,
(1.7 6 0.2 min), was much faster than that observed for FSH respectively). Total FSH measured in the basal compartment
428 FSH SECRETORY DYNAMICS Endo • 1997
Vol 138 • No 1
TABLE 1. Characteristics of FSH pulses identified in hypophyseal portal and peripheral circulation
A. Number of FSH pulses identified in hypophyseal portal and jugular circulation
of FSH secretion during the 6-h collection period averaged The time lag relationship between the onset of GnRH and
2245.1 6 897.0 and 101.5 6 36.5 ng/ml in hypophyseal portal hypophyseal portal FSH pulses closely paralleled that of
and jugular circulations, respectively. Although basal GnRH hypophyseal portal LH (1.2 6 0.3 min for LH and 1.1 6 0.4
concentrations were reported to be at or near the limit of min for FSH). No time lag existed between the onsets of LH
detection (14), the mean baseline of hypophyseal portal FSH and FSH pulses (0.4 6 0.2 min).
was 8.2 6 2.4-fold greater than that estimated for peripheral
FSH (68.4 6 16.6 vs. 9.1 6 1.0 ng/ml, respectively). A sub-
Discussion
stantially greater proportion of FSH comprised the basal
component compared with the episodic component (72.9 6 Progress in understanding the secretory nature of FSH has
3.8% vs. 27.1 6 3.8%, respectively; P , 0.01). A similar re- been limited by the inability to assess secretory patterns of
lationship was noted for jugular FSH (69.2 6 6.5% vs. 30.8 6 FSH at a site close to its production. In the majority of studies
6.5%, respectively; P , 0.01). in which secretory patterns of FSH were assessed from the
Figure 5 summarizes the pulse concordance relationships peripheral circulation, the very nature of an episodic pattern
of GnRH and FSH (solid bars), with values for GnRH and LH of FSH secretion appears suspect. As a consequence, many
provided for comparison in the background (striped bars). As studies that have characterized LH patterns in depth are
reported earlier, a virtual one to one relationship exists be- either limited by infrequent measurements of FSH or by the
tween GnRH and LH, whether LH was measured at the use of mathematical procedures to deconvolve the nature of
hypophyseal portal (12) or peripheral level (14). Similarly, secretion (9, 10, 19 –21). This difficulty can be overcome if
assessment of FSH in hypophyseal portal blood revealed a blood is acquired close to the site of secretion before the
close association between GnRH and hypophyseal portal secretory products are subjected to dispersion and clearance
FSH, with 93% of the GnRH pulses found to be associated in circulation. Recently, we determined that hypophyseal
with FSH pulses. This relationship was less pronounced portal blood, in addition to serving as a source material for
when GnRH and jugular FSH pulses were compared (79% defining secretory patterns of hypothalamic secretions (11,
concordance). Episodes of FSH secretion were identified in 13), provides a means to define secretory patterns of pituitary
the absence of corresponding GnRH pulses (31.5% in hy- secretions (12). Exploiting this powerful approach we find, as
pophyseal portal and 45% in peripheral blood). has been surmised (2– 6), that FSH is secreted in two modes:
FSH SECRETORY DYNAMICS 429
TABLE 2. Average concentrations of FSH in the hypophyseal portal and peripheral circulations and relative proportions of basal and
episodic components
A. Total and basal components of FSH
current with GnRH pulses. In contrast to studies in the mare of FSH-releasing factor was first provided by Igarashi and
(33), in which 35% of the identified GnRH pulses had no McCann (48). Although considerable anatomical and bio-
concurrent FSH or LH pulses, almost all (93%) of the GnRH chemical evidence supports this possibility (49), and partial
pulses in this study were associated with FSH pulses. Such separation of a separate FSH-releasing activity has been
a close relationship was, however, not evident when FSH achieved (50), no specific FSH-releasing factor has been iso-
pulses were identified in the peripheral circulation. The very lated or found to be released into hypophyseal portal blood.
discrete nature of the GnRH-associated bursts of FSH in the Whether the non-GnRH-associated excursions in FSH secre-
hypophyseal portal blood and the close time lag relationship tion represent responses to a yet to be identified hypotha-
between GnRH and FSH suggest that the primary factor lamic FSH-releasing factor remains to be determined.
responsible for induction of the GnRH-associated pulses of
tagogues) all suggest that the FSH we measure in pituitary portal plasma should be used merely as guides, providing an
portal blood reflects primarily secretion and not leakage insight into the dynamics of pituitary FSH secretion, and not
from damaged pituitary cells. As for the site of origin of the to calculate total secretion.
FSH in hypophyseal portal blood, there are three possibili- In summary, characterization of FSH secretory profiles in
ties: 1) active secretion of gonadotropes located in pars tu- the hypophyseal portal blood from ovariectomized ewes
beralis, 2) retrograde blood flow from pituitary to the hy- reveals a dual mode of FSH secretion, basal and episodic.
pothalamus, and 3) drainage from pituitary sinusoids. Furthermore, a close one to one relationship exists between
GnRH and FSH pulses, suggesting that GnRH is a regulator
of episodic FSH secretion. Finally, identification of non-
Secretion vs. clearance GnRH-associated excursions of FSH secretion supports the
18. Kushler RH, Brown MB 1991 A model for the identification of hormone 38. DePaolo LV, Bicsak TA, Erickson GF, Shimasaki S, Ling N 1991 Follistatin
pulses. Stat Med 10:329 –340 and activin: a potential intrinsic regulatory system within diverse tissues. Proc
19. Christman GM, Randolph JF, Kelch RP, Marshall JC 1991 Reduction of Soc Exp Biol Med 198:500 –512
gonadotropin-releasing hormone pulse frequency is associated with subse- 39. Wang Q-F, Khoury RH, Smith PC, McConnell DS, Padmanabhan V, Midgley
quent selective follicle-stimulating hormone secretion in women with poly- AR, Schneyer AL, Crowley WF, Sluss PM 1996 A two-site monoclonal an-
cystic ovarian disease. J Clin Endocrinol Metab 72:1278 –1285 tibody immunoradiometric assay for human follistatin: secretion by a human
20. Reame NE, Kelch RP, Beitins IZ, Yu M-Y, Zawacki C, Padmanabhan V 1996 ovarian teratocarcinoma-derived cell line (PA-1). J Clin Endocrinol Metab
Age effects and pulsatile LH secretion across the menstrual cycle of premeno- 81:1434 –1441
pausal women. J Clin Endocrinol Metab 81:1512–1518 40. Mather JP, Woodruff TK, Krummen LA 1992 Paracrine regulation of repro-
21. Loucks AB, Heath EM 1994 Dietary restriction reduces luteinizing hormone ductive function by inhibin and activin. Proc Soc Exp Biol Med 201:1–15
(LH) pulse frequency during waking hours and increases LH pulse amplitude
41. Ying S 1988 Inhibins, activins and follistatins: gonadal proteins modulating the
during sleep in young menstruating women. J Clin Endocrinol Metab
secretion of follicle-stimulating hormone. Endocr Rev 9:267–293
78:910 –915
42. Padmanabhan V, Van Cleeff J, Favreau PA, Midgley AR, Opposing effects
22. Urban RJ, Padmanabhan V, Beitins I, Veldhuis JD 1991 Metabolic clearance
of activin and follistatin on LH and FSH secretion from perifused ovine pi-