|
|
||||||||
1Laboratory of Animal Physiology and Endocrinology, Faculty of Veterinary Medicine, University of Concepción, Chillán; 4Laboratory of Endocrinology and Metabolism, 3Faculty of Medicine, Institute of Maternal and Child Research, University of Chile, Santiago, Chile; and 2Reproductive Sciences Program, University of Michigan, Ann Arbor, Michigan
Submitted 10 March 2005 ; accepted in final form 8 June 2005
| ABSTRACT |
|---|
|
|
|---|
prenatal programming; insulin sensitivity index; glucose metabolism; small for gestational age; polycystic ovary syndrome
Although there is no human model that can be used to experimentally produce PCOS, clinical observations in women with congenital adrenal hyperplasia caused by classical 21-hydroxylase deficiency are suggestive of excess prenatal androgen action. Such women exhibit anovulation, ovarian hyperandrogenism, luteinizing hormone hypersecretion, polycystic-appearing ovaries, and insulin resistance despite normalization of adrenal androgen excess after birth (6, 21). Studies with animal models provide experimental evidence for prenatal testosterone (T) excess in the etiology of PCOS (1, 42). For example, experimentally produced prenatal T excess leads to the manifestation of metabolic and endocrine changes during adulthood in female monkeys that resemble those observed in human PCOS (1). Similarly, sheep treated with T during fetal life exhibit cycle anomalies, ovarian morphology, and gonadotropin secretion comparable to those of women with PCOS (9, 3436, 42). These observations suggest that T excess during early life, whether derived from fetal or extrafetal (maternal or external environment) sources, may contribute to the PCOS phenotype during adulthood. Therefore, an experimental animal model in which the fetus is exposed to excess T can contribute to our understanding of the etiology of PCOS.
Studies in prenatal T-treated female sheep (T-females) have focused predominantly on the hypothalamic-pituitary-gonadal axis during prepubertal development (36, 43) and adult reproductive life (9, 10, 34, 35, 36, 42). However, investigations on the effects of prenatal T excess on insulin sensitivity index (ISI), a common perturbation in PCOS women, are lacking. Recent studies in rhesus monkeys (14) provide unequivocal evidence in support of impaired pancreatic cell function and insulin sensitivity in T-females. Because these studies were conducted with an older cohort of monkeys, the developmental onset and progression of this disruption are not known. The aim of the present study was to build upon the monkey studies (14) and determine the impact of prenatal T treatment on developmental changes in ISI.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Assessment of insulin secretion and insulin resistance. To compare developmental changes in circulating insulin levels and insulin resistance of T-females with control (C) females, three ages were studied (5 wk, when newborns were still nursing; 20 wk, which is 2 mo postweaning; and 30 wk of age, during the peripubertal period). The average onset of puberty was 28.4 ± 1.5 and 24.9 ± 1.8 wk of age in C and T-females, respectively. For comparison, a group of normal males born of the same flock were also included; they have not been treated before birth. It should be noted males begin puberty at 10 wk of age (43).
Glucose and insulin tolerance tests. To assess insulin sensitivity, an intravenous glucose tolerance test (IVGTT) was done on 8 C-females, 8 T-females, and 7 males at 5, 20, and 30 wk of age. Only singleton female or male offspring (no twins) were used. Before the experiment (1 or 2 days), lambs were placed in individual crates in the animal experimentation room to accommodate them to new surroundings. Indwelling catheters (18 gauge; Arrow International, Reading, PA) were placed under local anesthesia in both jugular veins (32). Food was withdrawn at 2000. To acclimatize the animals to handling and blood sampling, serial blood samples were withdrawn from the jugular vein catheter the day before the IVGTT. All IVGTT were initiated at 0800 the next day. During the IVGTT, 300 mg/kg glucose were infused over 2 min via one of the indwelling jugular catheter (33). Blood samples (1 ml) were collected from the contralateral jugular vein at 15, 10, 0, 3, 5, 7, 10, 13, 15, 17, and 20 min. Blood samples intended for insulin measurements were collected in heparinized tubes on ice. A second set of tubes containing heparin and sodium fluoride was used for plasma glucose determinations.
After the glucose challenge (20 min), 0.1 U/kg human insulin (Humulin; Elli Lilly) were given as a bolus, and blood samples were withdrawn at 23, 25, 27, 30, 33, 35, 37, 40, 50, 60, 80, 100, 120, 140, 160, and 180 min. Blood samples were collected and processed as during IVGTT and spun at 1,000 g at 4°C for 15 min, and plasma was harvested and kept frozen at 20°C.
Insulin and glucose measurements. Plasma concentrations of insulin were determined by double-antibody RIA using commercial kits (DSL, Webster, TX). Increasing concentrations of ovine plasma were found to dilute in parallel with the human insulin standard. Recovery of added insulin averaged 96.4 ± 3.4% (n = 5). Results are expressed relative to the human insulin standard provided with the kit, which has been calibrated to the World Health Organization International Reference Reagent for Insulin (code 83/500). Minimal detectable dose, defined as 90% of buffer control, was 3 µU/ml with intra- and interassay coefficients of variation of 6 and 10%, respectively. Plasma concentrations of glucose were determined by the glucose oxidase method using commercial kits (Weiner Laboratory). Intra- and interassay coefficients of variation were <5%.
ISI. The insulin response was determined by calculating the mean levels of insulin achieved during the first 20 min after glucose infusion and by calculating the incremental area under the curve (AUC) of insulin, calculated as the difference between the basal AUC and the AUC of insulin secreted in response to the glucose challenge during the first 20 min of the assay. AUC was calculated with the trapezoidal formula using a computer program based on an Excel spreadsheet.
Insulin sensitivity was assessed by measuring the fasting insulin-to-glucose ratio and composite insulin response (ISI-C), the glucose utilization constant, and the glucose disappearance after exogenous insulin. The insulin-to-glucose ratio was calculated from the mean values of insulin and glucose in plasma samples obtained at 15, 10, and 0 min of glucose infusion in each subject. ISI was calculated using the formula adapted from Matsuda and DeFronzo (29) where ISI = 10,000/square root of [(fasting glucose x fasting insulin) x (mean glucose x mean insulin during the first 20 min of the IVGTT)] (29, 33).
The glucose utilization constant (%/min) was calculated using the slope of the log linear regression of plasma glucose concentrations between 10 and 20 min postglucose administration. The effect of exogenous insulin on plasma glucose disappearance was determined with the formula of Grulet et al. (20):
glucose/glucose 20, where glucose 20 is the glucose concentration before exogenous insulin administration and
glucose is the variation between glucose 20 and the plasma glucose obtained at 40 min (20 min after the insulin administration), which is derived from the regression plot (20).
Statistical analysis. All variables were analyzed by ANOVA for repeated measures with treatment (C-female, T-female, male) as the main factor and age (5, 20, and 30 wk) as the repeated-measures factor using the GB-Stat version 6.5 statistical program. Pairwise post hoc comparisons were made by the Newman-Keuls test. Differences were considered statistically significant at a level of P < 0.05. Results are shown as means ± SE.
| RESULTS |
|---|
|
|
|---|
|
|
AUC of basal insulin measurements over 20 min (imputed from 15-min measures) and incremental AUC of insulin response to glucose challenge during the first 20 min also increased with age in the C-females from 5 to 30 wk of age (Fig. 3). No age-related changes in basal and incremental AUC were observed in T-females and males. Basal and incremental AUC were of comparable magnitude at all three ages in T-females and males. Basal and incremental AUC of insulin was significantly higher in T-females compared with C-females at 5 wk of age but not at other ages. Basal and incremental AUC of insulin secretion were also higher in males compared with C-females at 5 wk but similar between males and T-females.
|
|
| DISCUSSION |
|---|
|
|
|---|
Effect of prenatal T treatment on birth weight. Reduced birth weights of T-females treated with 60 mg of T corroborate our earlier findings (28), where similar outcomes were achieved after prenatal treatment with a higher dose of T (100 mg two times weekly; see Ref. 28). The results of these two studies differ from an earlier study where a single administration of 100 mg T at 30, 40, or 50 days postcoitum had an opposite effect, resulting in the birth of heavier lambs (19). Such differences may stem from differences in timing and amount or duration of T exposure. Paradoxically, the C-males were not growth retarded. Absence of growth retardation in C-males as opposed to T-females may relate to exposure of male fetuses to endogenous T throughout gestation, with late-gestation T exposure facilitating tissue accretion. Recently, we found that males exposed to higher levels of T during midgestation have lower birth weight (28). The increase in proportional weight gain of males after 16 wk suggests that continued postnatal exposure to T from the active testes may be a contributing factor because T increases muscle mass (24).
Evidence exists in humans linking prenatal T exposure with growth retardation. Baby girls born to a Spanish American cohort of PCOS mothers were found to have lower weight compared with those born to mothers without PCOS (38). This raises the possibility that the chronic hyperandrogenemia present during pregnancy in PCOS (37) may have contributed to the reduced body weight. Higher levels of T have also been found in amniotic fluid of diabetic gestating mothers (4) and in congenital adrenal virilizing disorders (6). A cordiocentesis study of 114 pregnancies in humans also found that fetal serum T levels around midgestation (1925 wk) were elevated in the male fetal range in
4 of 10 female fetuses sampled (7).
Interestingly, the effect of prenatal T treatment on birth weight parallels the outcome achieved with maternal undernutrition. Food restriction during mid- to late gestation results in lower-birth-weight lambs (40). Small-for-gestational-age human infants in general appear to result from a restriction of the supply of nutrients and oxygen to the growing fetus arising from altered placental physiology (30). Whether prenatal T treatment impacts placental physiology and restricts nutrient transfer to fetuses remain to be determined. Consistent with this premise, preliminary studies at days 60 and 90 of gestation found advancement of placental differentiation in sheep treated with T from days 30 to 90 of gestation (3). Postnatally, the T-females in the present study exhibited catch-up growth similar to that found in our recent study (28) using a higher level of prenatal T exposure. Growth retardation and catch-up growth are viewed as risk factors for predisposition of adult diseases (5, 31).
Prenatal T treatment and postnatal insulin sensitivity.
Our findings document that prenatal T alters ISI during early postnatal life. The extent to which this is an extension of perturbations in utero remains to be determined. The similarity of ISI between T-females and males at all ages coupled with differences in ISI between males and C-females suggest that T by androgenic or estrogenic action may mediate differences in the developmental trajectory of pancreatic differentiation or function. The increased insulin responses of the T-females may be the result of an increased percentage of islet tissue in the pancreas, a compensatory
-cell response to overcome compromised pancreatic differentiation, or alternatively the result of impaired glucose sensing by the pancreatic
-cell, as reported for women with PCOS (13). To what extent the reduced insulin sensitivity is related to reduced birth weight is also unclear. In humans, the percentage of islet tissue in the pancreas at birth is directly correlated to birth weight (15, 41), leading to reduced basal insulin levels (15, 23).
Developmentally, the decline in insulin sensitivity in C-females from 5 to 30 wk of age is consistent with findings of Gatford et al. (17) in sheep. Maturational decreases in insulin sensitivity in C-females also parallel what has been found in rats (18) and humans (2). Lack of sex differences at 38 days of life (
5 wk of age) in the study by Gatford et al. (17) as opposed to our study may relate to the high variability in basal insulin levels of males in their study and consequent masking of differences or, alternatively, a function of altered developmental trajectories in the two breeds (Suffolk in our study and Merino in the study by Gatford et al.).
The disappearance of differences in ISI between C-females and T-females at 20 and 30 wk of age appears to be a reflection of a reduction in insulin sensitivity of C-females as they approached puberty, since the ISI of T-females did not change with age. The age-related increase in basal insulin concentrations in C-females, which confirms recent work of Gatford et al. (17), and the corresponding decline in ISI-C are supportive of developing insulin resistance in C-females as they approach puberty. Interestingly, ISI of 30-wk C-females were comparable to that of 5-wk-old males. The directionality of changes in ISI in males and C-females may relate to the marked differences in the timing of puberty. Male lambs reach puberty at
10 wk of age and females at
30 wk (43). The similarity of the ISI in males and T-females may relate to the reported early onset of neuroendocrine puberty in T-females (43). The extent to which changes in peripheral ISI contribute to the neuroendocrine changes that subserve puberty or, alternatively, the extent to which puberty-associated changes in neuroendocrine and gonadal function impact on ISI remain to be understood.
Although the observed hyperinsulinemia in T-females is suggestive of developing insulin resistance, this was not evident from the glucose disappearance rate after insulin bolus or fasting glucose levels. The absence of changes in fasting glucose levels in our study differs from other studies (11, 17), which found an age-related decline in basal glucose levels in both sexes. It is possible that an early elevation in basal glucose may have been missed in our study as a consequence of advancement in developmental trajectory of mechanisms governing glucose homeostasis in this breed of sheep. Sheep, like other ruminants, are perceived as more insulin resistant than nonruminants, and they obtain most of their glucose supply by gluconeogenesis in the liver, using as substrate short-chain fatty acids (propionate, butyrate and acetate) synthesized by microorganisms in the rumen (8). As such, the glucose disappearance rate obtained by the change in plasma glucose concentrations in response to an insulin challenge in this study may not be a good measure of insulin-mediated changes in sheep and may require hyperinsulinemic-euglycemic clamp studies.
Our findings raise several questions to be addressed. What are the late-life consequences of such early perturbation in ISI? Are these perturbations a continuum of disruptions that occurred in utero? What is the critical period for pancreatic programming, and is this facilitated by androgenic or estrogenic action of T? On the basis of epidemiologic data and other experimental models (22, 39), growth retardation and catch-up growth observed in our T-females suggest that these animals may be at risk of developing insulin resistance during their adult life, especially as they get older. This is corroborated by the documentation of impaired pancreatic
-cell function and insulin resistance in older Rhesus monkeys treated prenatally with T (14). As such, the monkey and sheep studies complement each other, with sheep studies addressing developmental changes and monkey studies documenting adult disease onset. Life span studies comparing C-females and T-females are needed to establish the chronology of development of diseases involving the insulin/glucose homeostasis. In summary, our findings document that prenatal T excess leads to fetal growth retardation and impaired insulin sensitivity early in life in the growing female sheep.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
This work was presented in part at the 86th Annual Meeting of the Endocrine Society, June, 2004, New Orleans, LA.
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Demissie, M. Lazic, E. M. Foecking, F. Aird, A. Dunaif, and J. E. Levine Transient prenatal androgen exposure produces metabolic syndrome in adult female rats Am J Physiol Endocrinol Metab, August 1, 2008; 295(2): E262 - E268. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Veiga-Lopez, W. Ye, D.J. Phillips, C. Herkimer, P.G. Knight, and V. Padmanabhan Developmental Programming: Deficits in Reproductive Hormone Dynamics and Ovulatory Outcomes in Prenatal, Testosterone-Treated Sheep Biol Reprod, April 1, 2008; 78(4): 636 - 647. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. King, N. Bari Olivier, P. S. Mohankumar, J. S. Lee, V. Padmanabhan, and G. D. Fink Hypertension caused by prenatal testosterone excess in female sheep Am J Physiol Endocrinol Metab, June 1, 2007; 292(6): E1837 - E1841. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. J. Crespi, T. L. Steckler, P. S. MohanKumar, and V. Padmanabhan Prenatal exposure to excess testosterone modifies the developmental trajectory of the insulin-like growth factor system in female sheep J. Physiol., April 1, 2006; 572(1): 119 - 130. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |