| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Medicine (V.V., S.V., H.Y.-J., M.J.T.), University of Helsinki, 00140 Helsinki; and Jorvi Hospital (H.H.), 02740 Espoo, Finland
Address all correspondence and requests for reprints to: Matti J. Tikkanen, M.D., Department of Medicine, Helsinki University Central Hospital, PB 340, 00290 Helsinki, Finland. E-mail: matti.tikkanen{at}hus.fi.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Only a few studies have analyzed concentrations of circulating estradiol fatty acid esters in humans. When determined by gas chromatography-mass spectrometry, serum concentrations of estradiol fatty acid esters were low, but detectable, in premenopausal women (n = 4) but undetectable in postmenopausal women (n = 10) (3). We have recently developed a method for determination of estradiol fatty acid esters in human body fluids using time-resolved fluoroimmunoassay (TR-FIA) (8). Although originally developed for measurement of estradiol fatty acid esters in serum from pregnant women, this method was found to be sensitive enough for use in nonpregnant women and in men. We set out to investigate whether estradiol replacement therapy, given either orally or transdermally, might increase the amount of circulating estradiol fatty acid esters in postmenopausal women. To our knowledge, there are no previous data addressing effects of estradiol replacement therapy on serum estradiol fatty acid ester concentrations. To study possible physiological functions, we also related the changes in serum estradiol fatty acid ester concentrations to the changes in measures of endothelium-dependent and -independent vasodilatation during treatment (9).
| Subjects and Methods |
|---|
|
|
|---|
To compare the estradiol fatty acid ester/estradiol ratios in menstruating and postmenopausal women, serum estradiol fatty acid ester concentrations were also measured in seven healthy 16-yr-old women. Blood was obtained in one subject during the follicular phase (d 2 of menstrual cycle) and in six subjects during the luteal phase (d 1626 of menstrual cycle). To compare the results with those obtained in men, pooled sera (purchased from the Finnish Red Cross, Helsinki, Finland; complement inactivated at 56 C for 30 min, serum stored at -20 C) from 5060 male donors were analyzed.
Measurement of serum estradiol fatty acid ester and estradiol concentrations
Serum estradiol fatty acid ester and estradiol concentrations were determined as described previously (8). In short, after adding [3H]estradiol-3,17ß-dioleate (26003500 dpm) as an internal standard to each sample for determination of recovery, serum samples (1 ml) and low-, medium-, and high-control samples (1 ml; estradiol-17ß-stearate added to pooled male serum, corresponding to estradiol concentrations of 110 pM, 257 pM, and 598 pM, respectively) were extracted with diethylether:ethyl acetate. Sephadex LH-20 column chromatography was then performed to separate estradiol fatty acid esters from nonesterified estradiol. After saponification of the estradiol ester fraction in methanolic KOH (potassium hydroxide) and washing in Sep-Pak C18 column, cholesterol and other lipid impurities were eliminated by Lipidex 5000 reversed-phase and Sephadex LH-20 column chromatography. The dry residues of the hydrolyzed estradiol ester fraction and the nonesterified estradiol fraction (the latter obtained from the first Sephadex LH-20 separation step) were concentrated 2-fold by dissolving in 0.5 ml buffer (control samples were dissolved in 1.0 ml buffer). To determine recovery, radioactivity in the estradiol ester fraction of each sample was measured by liquid scintillation counting. The concentration of estradiol in both fractions was analyzed by TR-FIA as previously described (8). In postmenopausal women, samples obtained at 0 and 12 wk were analyzed in the same assay. Serum estradiol concentrations in postmenopausal women were also measured by RIA (9).
Estradiol-17ß-stearate-containing control samples were analyzed in every assay. According to the coefficients of variation of these low-, medium-, and high-control samples, the interassay imprecision of the quantitative estradiol ester method in seven assays was 5.5%, 8.5%, and 13%, respectively. A male serum sample (pooled human male sera) was also analyzed in every assay as a control. The measured concentrations of hydrolyzed estradiol esters in the 2-fold concentrated male serum sample were within the working range of estradiol TR-FIA (working range, 381840 pM). The interassay imprecision of the 2-fold concentrated male serum sample in seven assays was 12%, and the intraassay imprecision was 6% (n = 5). The concentration of serum estradiol fatty acid esters in postmenopausal women, young women, and men was calculated by correcting for recovery of the tritiated internal standard and then normalizing to volume. The mean recovery of the tritiated internal standard in seven assays was 72% (95% confidence interval, 70.973.2%). No correction for blank was made. The measured concentration of estradiol esters in a 2-fold concentrated H2O sample (reagent blank) was repeatedly less than 11.5 pM (<3.13 pg/ml), the lowest calibrator in TR-FIA.
Other measurements
Serum and lipoprotein cholesterol, triglyceride, apolipoprotein B, and serum SHBG concentrations were measured as previously described (9, 10). Because oral (but not transdermal) estradiol treatment increases serum SHBG concentrations, the concentrations of free (nonprotein-bound) estradiol were calculated (11). In vivo endothelial function was assessed from blood flow responses to intrabrachial infusions of endothelium-dependent (7.5 and 15 µg/min acetylcholine) and endothelium-independent (3 and 10 µg/min sodium nitroprusside) vasodilators at 0 and 12 wk, as has been reported previously (9).
Statistical analyses
Data are expressed as mean (SEM), except for the estrogen concentrations, which are expressed as median (range). Wilcoxon signed ranks test (two tailed) was performed to compare data at baseline and after 12 wk of treatment within the groups. Mann-Whitney test was used to compare estradiol ester concentrations between the groups. Correlation analyses were performed using Spearmans nonparametric correlation coefficient.
| Results |
|---|
|
|
|---|
Serum total, LDL, and high-density lipoprotein (HDL) cholesterol; serum triglyceride; and apolipoprotein B concentrations at 0 and 12 wk are shown in Table 1
. The concentrations of serum estradiol fatty acid esters, serum nonesterified estradiol, and free (nonprotein-bound) estradiol are shown in Table 2
; and individual serum estradiol fatty acid ester concentrations at baseline and at 12 wk, in Fig. 1
. The concentration of serum estradiol fatty acid esters increased significantly after 12 wk of treatment in the oral estradiol group (P = 0.028) but remained unchanged in the transdermal estradiol and placebo groups (Table 2
). The serum estradiol fatty acid ester concentrations were significantly higher in the oral estradiol group, at 12 wk, than in the transdermal estradiol (P = 0.022) or placebo groups (P = 0.030). When serum estradiol ester concentration was expressed relative to total serum cholesterol, LDL cholesterol, or apolipoprotein B concentration, these ratios increased significantly (by 33%, 47%, and 40%, respectively) after oral but not transdermal estradiol or placebo treatment (Table 2
). In all three treatment groups, the median baseline concentration of estradiol fatty acid esters was higher than the median nonesterified estradiol concentration. At 12 wk, there was a positive correlation between serum total nonesterified estradiol and estradiol ester concentrations in the transdermal and oral estradiol treatment groups (n = 19, r = 0.53, P = 0.019). Serum SHBG concentrations increased significantly in the oral (P = 0.008), but not in the transdermal estradiol or placebo groups (Table 2
). The calculated median concentrations of free (nonprotein-bound) estradiol were similar in the oral and transdermal groups at 12 wk. Serum total nonesterified estradiol concentrations measured in the present study by TR-FIA correlated well (Spearmans correlation coefficient, 0.92, P < 0.001) with those previously measured by RIA (9).
|
|
|
The change in serum estradiol fatty acid ester concentration in all postmenopausal women (n = 26) at 12 wk vs. 0 wk was positively correlated to the changes in forearm blood flow responses: both to the changes in endothelium-dependent (7.5 and 15 µg/min acetylcholine; r = 0.41 and r = 0.43; P = 0.038 and P = 0.028, respectively) and endothelium-independent (10 µg/min sodium nitroprusside; r = 0.44; P = 0.024) vasodilatation. The correlation between the change in serum estradiol fatty acid ester concentration and the percent change (flow above basal compared with 0 wk) in sodium nitroprusside-induced (10 µg/min) vasodilatation during treatment (n = 26, r = 0.47, P = 0.015) is shown in Fig. 2
. In contrast, no significant correlations were observed between the change in serum nonesterified estradiol concentration and the changes in blood flow responses during treatment (9).
|
Serum estradiol fatty acid ester concentrations in 16-yr-old menstruating women (n = 7), pooled human male sera (obtained from 5060 donors), and all postmenopausal women at baseline (n = 26) are shown in Table 3
. The median serum estradiol fatty acid ester concentrations were in the same range in young menstruating and postmenopausal women, between 7580 pM (2022 pg/ml) (P = not significant). In men, despite a higher serum concentration of nonesterified estradiol than in postmenopausal women, the serum estradiol fatty acid ester concentration was lower than in postmenopausal women (P = 0.001) (Table 3
).
|
| Discussion |
|---|
|
|
|---|
The serum estradiol fatty acid ester concentrations were determined as equivalents of estradiol by immunoassay after complete chromatographic separation and hydrolysis of the nonpolar ester fraction. In contrast to our previous study that analyzed pregnancy serum (8), the serum samples of the present study obtained from men and nonpregnant women were concentrated 2-fold before measurement. As judged from reproducibility of determinations of low estradiol ester concentrations in men, and specificity of the antiserum, the estradiol ester concentrations presented for men and postmenopausal women are accurate. A limitation of the study was that it was not possible to get and analyze serum samples containing no estradiol esters; and thus, the possible confounding effects of serum matrix components on the measured estradiol ester levels could not be completely ruled out. Because the number of subjects was relatively small in this study, the results should be reproduced in a larger cohort.
When the ratios of serum estradiol fatty acid esters to serum nonesterified estradiol were compared, the ratio was 0.29 in young women, 0.75 in men, but 2.2 in postmenopausal women at baseline before any replacement therapy (Table 3
). These proportions are much higher than those we previously found during pregnancy (8). Thus, the estradiol fatty acid ester/estradiol ratio in blood seems to increase, rather than fall, with physiological decreases in serum estradiol concentrations. Because estradiol esters are synthesized in blood by HDL-associated lecithin:cholesterol acyltransferase (LCAT) (12, 13, 14), their substrate in men and postmenopausal women is presumably estradiol produced by peripheral aromatization of gonadal and adrenal androgens. The higher proportion of esterified estradiol in men, compared with young women (Table 3
), could be partly explained by the higher proportion of nonprotein-bound estradiol in men, compared with young women (15). The unexpectedly high proportion of estradiol ester in postmenopausal women could, in theory, represent a storage mechanism for estrogen during postmenopausal hormone deficiency (3). Also, some in vivo and in vitro studies have provided data of enhanced steroid fatty acid esterification associated with aging (16, 17). In a previous study using different methodology, serum estradiol ester concentrations in men and postmenopausal women were considered to be below the detection limit, as determined by mass spectrometry after saponification of the esters (3). In that same study, however, omental and sc adipose tissue of women menopausal for less than 12 yr contained substantial amounts of estradiol fatty acid esters (670 fmol/g), about two thirds of the amount determined in adipose tissue of premenopausal women (3). Even 1520 yr after menopause, there seemed to be relatively large amounts of estradiol esters in fat (3). The accumulation of esterified estradiol in adipose tissue after menopause suggests an important physiological function for estrogen esters (3).
Supraphysiological (micromolar) concentrations of 17ß-estradiol have been shown to inhibit both cellular and copper-mediated oxidation of LDL in vitro (18, 19). Physiological estradiol concentrations (nano- to picomolar) may also inhibit LDL oxidation in vitro in the presence of ascorbic acid (20). Estradiol concentrations of 1100 nM have been shown to inhibit LDL oxidation in vitro but only after estradiol had been transformed to estradiol fatty acid esters in plasma (6). Sack et al. (21) reported that administration of 17ß-estradiol intraarterially or transdermally to postmenopausal women was associated with an increased resistance of LDL to oxidation ex vivo. Other studies have either supported (22, 23, 24) or not supported (25, 26, 27) protection of LDL against oxidation by estrogen replacement therapy. These data thus raise the possibility that estradiol, especially in its esterified form, may, at physiological concentrations in plasma, protect lipoproteins from oxidation. In our study, the significant increase, by 47%, in the estradiol fatty acid ester/LDL cholesterol ratio during oral estradiol treatment suggests that the number of estradiol fatty acid ester molecules per LDL particle may have been increased (Table 2
). The increased estradiol fatty acid ester/LDL cholesterol ratio could, in theory, provide an explanation for the reported antioxidant protection of LDL during estrogen administration (21, 28).
Interestingly, the change in serum estradiol fatty acid ester concentration in all postmenopausal women during treatment was positively correlated with the changes in blood flow responses to endothelium-dependent and -independent vasoactive drugs in forearm resistance vessels. The significance of these correlations is unclear, because such a relationship between serum estradiol ester concentration and vascular function does not prove causality. As reported previously in the same group of women (9), oral estradiol treatment markedly improved endothelium-dependent and -independent blood flow responses, but there was no correlation between serum estradiol concentrations and measures of vascular responses, apparently because both transdermal and oral estradiol increased free (nonprotein-bound) estradiol concentrations similarly. As shown in the present study, only oral estradiol increased serum estradiol fatty acid ester concentrations. These data add another potential candidate to the list of putative mechanisms whereby oral (but not transdermal) estradiol increase blood flow. However, apart from the increase in serum estradiol esters, increases in HDL cholesterol and serum estrone concentrations and decreases in LDL cholesterol, lipoprotein (a), and serum free testosterone concentrations are other possible causes for beneficial effects of oral estradiol treatment on endothelial function (9).
One possible explanation for the increase in median serum estradiol fatty acid ester concentration in the oral estradiol group only could be that orally, but not transdermally, administered estradiol undergoes hepatic first-pass metabolism. The liver is capable of synthesizing estradiol fatty acid esters upon incubation with estradiol (12, 29), but the role of the liver in metabolism of estradiol fatty acid esters in humans remains unknown. Another possible explanation could be that the higher estradiol ester concentration during oral, compared with transdermal, treatment might have resulted from the greater estrogen dose administered orally. Although serum nonprotein-bound estradiol levels were the same in both treatment groups, serum total estradiol and estrone concentrations (and presumably several other metabolites) were much higher during oral treatment. Consequently, the exposure of the liver to estrogen was greater, as demonstrated by the elevated HDL3 level during oral therapy, which could have facilitated estradiol fatty acid ester formation. Several experimental studies indicate that serum estradiol fatty acid esters, as well as the fatty acid esters of some other steroids, are synthesized in HDL particles (30, 31, 32), particularly in the HDL3 subfraction (14, 32), in an enzymatic reaction mediated by LCAT (13, 14). Estradiol fatty acid esters may be further transferred from HDL particles to LDL (30, 31). LDL particles are potent modulators of vascular function. LDL, especially oxidized LDL, blunts endothelium-dependent vasodilatation in vitro (33), but lowering of LDL cholesterol with statins has been shown, in multiple studies, to enhance endothelium-dependent vasodilatation (34, 35). The possibility that the ability of LDL particles to modulate vascular function depends on their content of estradiol fatty acid esters warrants further study based on the present data.
In conclusion, the results of this study suggest that serum estradiol fatty acid ester concentration may be increased by oral estradiol therapy. The data also raise the possibility that estradiol fatty acid esters, perhaps when incorporated in LDL particles, contribute to enhanced vascular function during oral estradiol treatment.
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: HDL, High-density lipoprotein; LCAT, lecithin:cholesterol acyltransferase; LDL, low-density lipoprotein; TR-FIA, time-resolved fluoroimmunoassay.
Received June 7, 2002.
Accepted October 18, 2002.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
F. Gungor, I. Kalelioglu, and A. Turfanda Vascular Effects of Estrogen and Progestins and Risk of Coronary Artery Disease: Importance of Timing of Estrogen Treatment Angiology, June 1, 2009; 60(3): 308 - 317. [Abstract] [PDF] |
||||
![]() |
M. Badeau, V. Vihma, T. S. Mikkola, A. Tiitinen, and M. J. Tikkanen Estradiol Fatty Acid Esters in Adipose Tissue and Serum of Pregnant and Pre- and Postmenopausal Women J. Clin. Endocrinol. Metab., November 1, 2007; 92(11): 4327 - 4331. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Vihma, A. Tiitinen, O. Ylikorkala, and M. J. Tikkanen Quantitative Determination of Estradiol Fatty Acid Esters in Lipoprotein Fractions in Human Blood J. Clin. Endocrinol. Metab., June 1, 2003; 88(6): 2552 - 2555. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |