| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Nutrition (W.S., H.C., H.J., F.M.S.), Harvard School of Public Health, and the Departments of Medicine (F.M.S.), and Obstetrics and Gynecology (B.W.W.), Brigham and Women’s Hospital, and Harvard Medical School, Boston, Massachusetts 02115
Address all correspondence and requests for reprints to: Frank M. Sacks, M.D., Department of Nutrition, Harvard School of Public Health, 665 Huntington Avenue, Boston, Massachusetts 02115. E-mail: fsacks{at}hsph.harvard.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The plasma Lp(a) concentration is associated with coronary heart disease (CHD) in most case-control studies in Caucasian populations (4). Data are sparse for other racial groups. Certain prospective epidemiological studies reported that Lp(a) is an independent risk factor for the development of CHD (5, 6, 7), whereas other such studies have not found Lp(a) to be a risk factor (8, 9). In women, case-control studies (10, 11, 12, 13, 14, 15, 16, 17) and a prospective study (18) have found that Lp(a) is associated with CHD. One of these studies found that Lp(a) was significantly associated with CHD in women who are less than 55 yr old but not in those who are 55 or older (16); although, in another study, Lp(a) was significant in postmenopausal and premenopausal women (17). In the prospective Framingham study, the significant relationship between Lp(a) and CHD in women was also unaffected by age or menopausal status (18).
Premenopausal women have lower Lp(a) levels than postmenopausal women (19, 20, 21, 22). Lp(a) levels increase in women after a natural or surgical menopause (20, 21, 22) and could contribute to the postmenopausal increase in CHD. Estrogen replacement therapy lowers plasma Lp(a) levels (22, 23, 24, 25, 26, 27, 28, 29, 30), and this action could contribute to the reduction in CHD in women who receive this therapy (31). The mechanism by which estrogen lowers the level of Lp(a) is not known. Estrogen decreases LDL levels by increasing LDL clearance by LDL receptors (32, 33, 34, 35). Theoretically, increased LDL receptor activity could also increase the clearance of Lp(a) by increasing the binding to cells of the apoB component of Lp(a). Overexpression of the human LDL receptor in transgenic mice lowers Lp(a) levels (36). In the estrogen-treated rat, Lp(a) clearance from plasma was increased, as was uptake of Lp(a) by liver parenchymal cells (37). However, in humans, differences in Lp(a) concentrations among individuals, or between Lp(a) isoforms within an individual, are determined primarily by the production rate of apo(a), rather than by its clearance rate (38, 39, 40, 41, 42). We conducted the present study to distinguish whether estrogen lowers Lp(a) levels by increasing clearance or decreasing production (flux) of apo(a). The specific mechanism could have clinical relevance. For example, reduced flux into plasma of this atherogenic lipoprotein is undoubtedly beneficial, whereas an increased clearance rate would require an interpretation based on sparse existing information on the cellular mechanisms for uptake and the tissues in which Lp(a) particles are catabolized.
Human Lp(a) metabolism is itself incompletely understood. For example, it is not established whether Lp(a) is secreted by the hepatocyte into plasma, or whether Lp(a) is formed at the surface of hepatocytes by joining newly synthesized apo(a) to preexisting LDL or very-low-density lipoprotein (VLDL) particles that have been circulating in plasma. Studies of cultured hepatocytes and animal models provide support for both of these processes giving rise to plasma Lp(a) (43, 44, 45, 46, 47). We studied the metabolism in plasma of the components of Lp(a), apo(a) and apoB100, in a subset of the women, to gain insight about the metabolism of Lp(a) particles in normal humans.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The eligibility criteria for the participants were postmenopausal status (defined as having amenorrhea for at least 12 months and a serum FSH level above 50 IU/L), good overall health, age 41–70 yr, and body mass index less than 35kg/m2. Informed consent was obtained, and the study was approved by the human subjects committee of Brigham and Women’s Hospital. The enrollment goal was 12 participants, and it was attained. One of the women had participated in our previous kinetic studies of estrogen replacement therapy (32, 35, 48). Eleven women were receiving hormone replacement therapy before starting the study, and the hormones were withdrawn for at least 4 weeks before starting the study. They were then randomized to 1 of 2 sequences of administration of estrogen and placebo in a double-blind cross-over protocol. Micronized 17ß-estradiol (2 mg, 1 tablet daily; Mead Johnson, Evanston, IL) or matching placebo was taken for 7 weeks each in random order. A 3-week period of no treatment separated the estrogen and placebo periods. Fasting blood for lipid analysis was obtained on 2 days in the final week of the 2 treatment periods. On the final day of each treatment period, the participants were admitted to the Clinical Research Center at Brigham and Women’s Hospital for a metabolic study, after an overnight fast. Our unpublished findings, in a trial that used the same type and dose of estrogen as the present study, showed that Lp(a) levels reached a new steady state after 3 weeks of estrogen treatment, because there was no additional lowering after 6 and 9 weeks (unpublished observations, B. W. Walsh and F. M. Sacks). In agreement, it was found that conjugated equine estrogens (0.625 mg) lowered Lp(a) at 3 weeks, by the same amount as at 4 weeks (30).
Participants were instructed to follow their usual dietary patterns during the study. They were fasting for 10–12 h when the tracer study began in the morning. A fat-free, leucine-free lunch and supper, providing 60% of daily energy, were given after 4 and 10 h of infusion, to prevent any abnormal metabolic effects of prolonged fasting, as has been our procedure in our previous studies (32, 35, 48). The metabolism of Lp(a) was evaluated by endogenously labeling its protein components, apo(a) and apoB100, by a primed, constant iv infusion of trideuterated leucine (purity > 99%; Tracer Technologies, Cambridge, MA), priming dose (4.2 µmol/kg), and a constant rate of 4.8 µmol/kg·h for 12 h. Blood specimens were obtained from a second iv catheter in the contralateral arm every 20 min for the first 2 h and hourly thereafter. Plasma taken during the infusion was stored at -80 C until use. We analyzed samples at 4, 6, 8, 9, 10, 11, and 12 h. No tracer enrichment above natural abundance was detectable during the first 3 h of tracer infusion in the first two participants.
Lp(a) preparation
Lp(a) was isolated by ultracentrifugation (density = 1.05–1.15). The density of 0.5 mL of plasma was raised to 1.05 g/mL by the addition of potassium bromide and the samples were spun for 24 h in a type 25 Ti rotor at 25,000 rpm in a L8–70 M ultracentrifuge (Beckman Instruments, Palo Alto, CA). The supernatant fraction was removed, and the infranatant density was raised to 1.15 g/mL and then centrifuged for 30 h. The Lp(a) fraction was aspirated in a 0.2-mL vol, dialyzed against phosphate-buffered saline-Tween 20 buffer for 4 h and then kept at 4 C until electrophoresis was performed, usually on the same day. Because we could not detect any more than an occasional trace of apo(a) on silver-stained gels in the d<1.050 fraction of our participants’ fasting plasma, we did not study Lp(a) in this fraction, which contained the VLDL and LDL.
Apo(a) isolation and preparation
The Lp(a) fraction, prepared by ultracentrifugation from 0.5 mL of plasma, was divided into two 100-µL aliquots, added to Laemmli sample buffer containing ß-mercaptoethanol, and applied to 2 adjacent wells, 6 mm wide in SDS 3–6% polyacrylamide gels, 1.5 mm thick (Integrated Separation Systems, Natick, MA). Electrophoresis was carried out for 4 h at a constant current of 35 milliamperes at 4 C using a vertical apparatus (Hoefer, SE600). Apo(a) and apoB100 bands were identified by silver staining and were verified by immunoblotting, as described below. Two apo(a) bands were clearly visualized by silver staining and separable from each other and from apoB100 in four of the participants (Participants 1, 2, 4, and 5), and both isoforms were excised for individual analysis. Participants 3, 6, and 10 had only one isoform visible and may have had a null phenotype or two very similarly sized phenotypes. For participant 9, each isoform, S1 and B, was isolated as an apo(a)-apoB100 complex using PAGE with a nonreducing sample buffer, because the apo(a) isoforms overlapped with apoB100 in the reducing gel. The isoforms of participant 11, S1 and B, migrated too closely on reducing or nonreducing gels to be distinctly separated, so the isoforms were excised together on a nonreducing gel and were studied as an apo(a)-apoB100 complex. Finally, one of the two isoforms for participants 7, 8, and 12 could not be analyzed because of insufficient isotopic enrichment. A range of 5–100 µg apo(a) was recovered from the gel, depending on the plasma apo(a) concentration.
ApoB100 isolation from Lp(a)
ApoB100 was isolated from Lp(a) in a subset of 6 of the participants during both placebo and estrogen treatment (participants 2, 4, 5, 7, 10, and 12) by first separating Lp(a) on an agarose gel, transferring the band to a reducing PAGE system, and isolating the apo(a) and apoB100 separately. Lp(a), prepared by ultracentrifugation, was purified by agarose gel electrophoresis. Agarose gels were cast using Seaplaque (1.5%; FMC, Rockland, ME) dissolved in Tris (500 mmol/L), boric acid (160 mmol/L), urea (1 mol/L), and edetic acid (1 mmol/L), pH8.5. The stacking gel consisted of Seakem (1%; FMC) in Tris HCl (125 mmol/L), pH6.8. Gels were poured into glass plates, 15 x 18 cm (coated with GelBond). Lp(a), 100 µL, in Laemmli sample buffer without reductive reagent, was electrophoresed in running buffer, Tris (90 mmol/L), boric acid (90 mmol/L), SDS (0.1%), pH8.5, at 4 C with constant current of 30 milliamperes for 6 h in a vertical apparatus (Hoefer SE600). The Lp(a) bands were precisely located by quick blotting the agarose onto a nitrocellulose membrane (S&S, Keene, NH) and matching the stained membrane with the unstained gel. The nitrocellulose membrane was applied to the agarose gel for 10 min at room temperature, washed 3 times in Tris (20 mmol/L), NaCl (500 mmol/L), Tween-20 (0.3%), pH 7.5, and twice in deionized, distilled water, and then stained with colloidal gold (BIO-Rad, Hercules, CA). The Lp(a) bands on the agarose gel were excised, and the agarose was digested with ß-agarase (FMC) using the following procedure: The gel slices containing Lp(a) were equilibrated in 10 volumes of reaction buffer [Tris HCl (40 mmol/L), NaCl (40 mmol/L), EDTA (1 mmol/L), pH 6.0] for 2 h at room temperature. The reaction buffer was discarded; and the gel slices were melted at 65–70 C for 15–20 min, and then cooled to 45 C, maintaining the liquid state. ß agarase, 3 U for each 200 mg of 1.5% agarose gel, was added for 1 h. The melted, digested gel slices were applied to 3–6% SDS-PAGE in Laemmli sample buffer with ß-mercaptoethanol; electrophoresis was performed as described above; and the apo(a) and apoB100 bands were identified by silver staining. Further identification of apo(a) and apoB100 bands was carried out using immunoblotting with monoclonal anti-apo(a) or anti-apoB100, and staining with peroxidase-linked antimouse IgG. Recovery of the apo(a) or apoB100 from the gel averaged 76%.
Apo(a) isoform determination
Lp(a) was prepared by ultracentrifugation, and apo(a) was separated by SDS-PAGE, as described above. Apo(a) was electrotransferred to nitrocellulose membranes, which were incubated with a 1:2000 dilution of monoclonal anti-apo(a) (Biotechnology Research Inc., Rockville, MD). Three lanes containing seven distinct isoforms were run in each gel as standards (PerImmune, Inc., Rockville, MD), along with apoB100 prepared from human LDL, to define the phenotypes and molecular masses of apo(a). The bands were detected by incubating the blots with anti-IgG linked covalently to peroxidase. The molecular mass of each apo(a) band of the participants was estimated from a standard curve composed of the seven apo(a) standards of known molecular mass and apoB100 using a gel scanner (Model PDI325oe with ImageMaster software, Pharmacia Biotech, Huntington Station, NY). The molecular mass of apoB100, 550 kd, was the calibrator for the molecular masses of the apo(a) isoforms.
Plasma lipid and lipoprotein measurements
The plasma concentration of Lp(a) was determined by enzyme-linked immunosorbent assay that uses anti-apo(a) for binding and detection (Strategic Diagnostics, Newark, DE). Cholesterol and triglycerides were measured by enzymic methods using an autoanalyzer (Cobas Mira, Roche Diagnostics, Branchburg, NJ). High-density lipoprotein (HDL) was separated from the other lipoproteins by precipitation using sodium phosphotungstate. LDL-cholesterol was determined by ultracentrifugation.
Measurement of tracer enrichment and pool size
The apo(a) and apoB100 samples were hydrolyzed (110 C, 24 h) in 6 N HCl under nitrogen to prepare amino acids, as described previously (32). The isotopic enrichment was measured by a 5890 gas chromatograph and a 5988A mass spectrometer (Hewlett-Packard, Palo Alto, CA), as described previously (32). Enrichment was defined as the tracer/tracee ratio of the sample minus the measured natural abundance of d3-leucine.
The pool sizes of the apo(a) and apoB100 proteins were directly measured by quantitative gas chromatography/mass spectrometry. of apo(a) and apoB100 bands cut from the gel. For participants 9 and 11, whose apo(a) and apoB100 were not separable, as described previously, the Lp(a) band was cut after nonreducing PAGE, as described previously, and the mass of the apo(a)-apoB100 complex was determined by gas chromatography/mass spectrometry. The apo(a) mass was then calculated using the relative proportions of molecular masses of the apo(a) isoforms and apoB100. Norleucine was added to each gel slice before hydrolysis and derivitization. The quantity of leucine was then determined by comparison of the area under the leucine curve with the area under the norleucine curve, thereby correcting for losses. The mass of apo(a) for each isoform was calculated from the amino acid composition of the protease, kringle V, and kringle IV regions (49, 50). Pool size was the plasma concentration of apo(a) or apoB100 protein multiplied by plasma volume.
Calculation of metabolic rates
The apo(a) pool sizes and enrichment curves were used to design a kinetic model by interactive computer program, CONSAM31 (NIH, Bethesda, MD). A parallel model was used for apoB100. Kinetic models were developed using data from the initial participants and then were tested on composite enrichment curves that were computed from the means of the 12 participants for placebo and estrogen treatments. The final kinetic model had the following characteristics: 1) Parallel models were established for tracer and tracee for apo(a) and for apoB100 with equal rate constants. 2) Input of tracer and tracee mass was directed into the first of 2 compartments connected in series to simulate intrahepatic synthesis and secretion. The inputs of tracer and tracee mass were linked by a forcing function: input of tracer = precursor enrichment x input of tracee. The enrichment of the precursor for apo(a) and apoB100 of Lp(a) was approximated by the enrichment of apoB100 in fasting VLDL (VLDL Sf20–60) during 7–12 h of tracer infusion when plateau had been reached and maintained. The precursor enrichments were determined in each participant during placebo and estrogen treatment. 3) A single plasma apo(a) compartment received apo(a) from the second intrahepatic compartment. The rate constants between the first and second intrahepatic compartments, and between the second intrahepatic compartment and plasma, were made equal. 4) A single pathway cleared apo(a) from the plasma compartment. The metabolic rates of each apo(a)isoform and of apoB100 in Lp(a) were separately determined. Alternative models that used extravascular equilibration pools that exchanged with the plasma compartment were tested and found not to improve the fits obtained from this simple model.
Statistical analysis
Data are presented as mean ± SD, except as noted. Analysis was conducted with Excel 7 (Microsoft Corporation, Redmond, WA) for comparison of differences in plasma Lp(a) concentrations, and apo(a) and apoB100 metabolic rates between estrogen and placebo treatments were analyzed by paired t test using 2-sided significance levels. Pearson correlation coefficients were calculated to examine associations of reduction in Lp(a) levels with changes in kinetic parameters and other plasma lipid levels (Statistical Analysis System, SAS Institute Inc., Cary, NC). For participants 1, 2, 4, 5, and 9, who had two apo(a) isoforms analyzed separately, the data were summed to give the total apo(a) pool size and production rate.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
).
|
). Estrogen decreased Lp(a)
concentration in 11 of 12 participants (Table 2). The Lp(a) concentration on placebo
treatment was not related to the percentage change in Lp(a)
concentration during estrogen treatment (r = 0.09). There was a
trend for changes in Lp(a) and LDL-cholesterol concentrations during
estrogen treatment to be correlated (r = 0.48, P =
0.11). The mean percentage change in Lp(a) was -17% when the estrogen
treatment period followed the placebo period, compared with +22% when
placebo followed estrogen, indicating the absence of a carryover
effect.
|
). The
delay in appearance and the subsequent slow incorporation of tracer
into apo(a) required an intrahepatic delay compartment consisting of 2
pools. Alternative approaches, such as linear regression, could not fit
the data as well as this compartmental model, even when an x-intercept
of 3–4 h was used. The precursor enrichments, approximated by the
plateau enrichment of apoB in VLDL Sf20–60, had a mean of 4.0%
(range, 3.2–4.8%) during placebo, and 4.1% (range, 3.2–4.9%)
during estrogen treatment (Fig. 1). These composite enrichment curves
were fit well by the compartmental model (Fig. 1). The slopes of the
tracer enrichment curves for placebo and estrogen were not obviously
different, and the fitted curves were nearly superimposible.
|
. Estrogen treatment reduced the mean apo(a) pool
size from 56 ± 24 nmol to 44 ± 21 nmol (P
< 0.01). The mean total residence time for the intrahepatic
compartments was similar during placebo and estrogen treatment,
5.3 ± 2.8 vs. 4.8 ± 2.4 h
(P = 0.24). The mean fractional catabolic rate (FCR)
for apo(a) during placebo was 0.362 ± 0.208 pools/day; and during
estrogen, the rate was 0.314 ± 0.186 pools/day (P
= 0.23). The mean difference in FCR between estrogen and placebo was
-0.048 ± 0.130/day. The FCRs during placebo and estrogen were
highly correlated among the participants (r = 0.79,
P < 0.005). This indicates relatively low biological
variation of Lp(a) FCR, and reliability in the laboratory and kinetic
modeling methods. In contrast to the lack of effect of estrogen on the FCR of apo(a), the mean production rate of apo(a) decreased significantly from 0.306 ± 0.233 nmol/kg·day during placebo to 0.206 ± 0.160 nmol/kg·day during estrogen treatment (P = 0.046). The mean difference was -0.100 ± 0.154 nmol/kg·day. The production rates of the individuals during placebo and estrogen were highly correlated (r = 0.75, P < 0.005).
Apo(a) isoform sizes ranged in apparent molecular mass from 456–933
kDa (Table 2). Two apo(a) bands were observed in nine participants, and
one band in three participants. The relationship among size, FCR, and
production rate for isoforms within persons, and the effect of estrogen
treatment, were studied in the five participants for whom two isoforms
could be separated and analyzed, as described in Subjects and
Methods. The mean pool sizes were slightly higher for the smaller,
compared with the larger, isoforms during each treatment (Table 3). There were trends toward higher
production rates for the smaller than for the larger size apo(a)
isoform within a person during placebo (P = 0.09) and
estrogen (P = 0.12), as well as for higher FCRs
(P < 0.01 for the placebo period, P =
0.06 for estrogen). The mean intrahepatic residence time was slightly
less for the smaller than for the larger isoform, 4.2 vs.
3.3 h (P = 0.11). The production rate of the
larger isoform was lower during estrogen treatment than during placebo
treatment for each participant, and the production rate of the smaller
isoform was lower during estrogen in four of the five participants,
with no significant differences between treatments. The
enrichment curves for participant 9 for the large and small isoform
during placebo and estrogen treatment illustrate the group trends for a
relationship between isoform size and metabolic rates (Fig. 2). During both placebo and estrogen
periods, the tracer enrichment in the large apo(a) isoform increased
more slowly and, at 12 h, had approximately a 50% lower value
than the small isoform. The enrichment curves during placebo and
estrogen for each isoform had nearly identical slopes (Fig. 2). The
enrichment curves of the average for the five participants is shown in
Fig. 3, also demonstrating the trend
toward higher FCR of the small isoform.
|
|
|
). The mean FCR for apo(a) was
0.277 ± 0.194/day; and for apoB100, it was 0.270 ±
0.155/day during placebo treament, and 0.221 ± 0.112/day and
0.203 ± 0.120/day during estrogen treatment, with no significant
differences (Table 4). The FCRs of apo(a)
and apoB100 were highly correlated among the participants during
placebo (r = 0.94) or estrogen (r = 0.83) treatment. The mean
production rates of apo(a) and apoB100 of Lp(a) were similar during
either placebo or estrogen treatment. There were trends toward
decreases in pool size, production rates, and FCRs of apoB100 of Lp(a)
during estrogen, compared with placebo, treatment (P =
0.05–0.06).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our finding that estrogen does not change the apo(a) FCR in postmenopausal women contrasts with findings, in rats, that massive doses of ethinyl estradiol increased clearance of human Lp(a) and uptake by hepatic parenchymal cells (37). These doses of estrogen are known to increase hepatic LDL receptors in rats (34). Experiments in mice made transgenic for the human LDL receptor have shown that LDL receptors can clear Lp(a) from plasma (36). However, the extent of LDL receptor up-regulation by high-dose estrogen or in the transgenic mouse is much greater than is likely to occur under physiological hormonal replacement in women. Estrogen replacement therapy should raise LDL receptor levels from the relatively low levels resulting from aging and estrogen-deficiency to levels closer to premenopausal women. Other evidence suggests that variation in LDL receptor levels from low to average does not affect Lp(a) clearance. Lp(a) clearance in homozygous or heterozygous familial hypercholesterolemic patients is similar to normolipidemic controls (39, 40, 53). In hypercholesterolemic patients, treatment with HMG CoA reductase inhibitors, which up-regulates LDL receptors, does not lower Lp(a) concentrations (54, 55, 56, 57). Corticotropin, which increases LDL receptor activity in human hepatoma cells, did not affect the uptake of Lp(a) (58). The binding capacity and affinity of the human hepatic LDL receptor for Lp(a) is lower than for LDL (59). In view of the much higher plasma concentration of LDL than Lp(a) particles, competition between LDL and Lp(a) for LDL receptor binding would eliminate most receptor sites to which Lp(a) could bind (59). Therefore, the lack of an effect of postmenopausal estrogen replacement on apo(a) clearance is consistent with findings that the LDL receptor seems to play a minimal role in human Lp(a) metabolism. The VLDL receptor expressed on extrahepatic cells (60), and direct proteolysis of apo(a) in plasma (61), are mechanisms for Lp(a) clearance recently identified.
Our finding that estrogen alters Lp(a) levels by affecting the apo(a) production rate adds to the growing evidence that human Lp(a) levels are, in general, controlled by synthesis, rather than by clearance rates. Nicotinic acid treatment lowers the plasma levels of Lp(a) by reducing Lp(a) production (62). Lp(a) levels among normolipidemic or hyperlipidemic persons are significantly correlated with Lp(a) production, rather than with clearance (38, 39, 40, 41, 42). The higher plasma levels of Lp(a) isoforms of smaller size are related mainly to higher production rates (41, 42). We also found a trend toward higher production rates for the smaller, than for the larger, size isoform within a person, although the FCR of the small isoform also tended to be higher than the large isoform.
The present study, in its use of endogenous labeling of apo(a), has theoretical advantages, compared with radioiodinated labeling of Lp(a) but also potential limitations. In favor of endogenous labeling is that tracer enrichment is measured individually in apo(a) and apoB100 of Lp(a), with no possible contamination by small amounts of labeled lipid or small, dense LDL. Exogenous labeling of Lp(a) could also alter the metabolic properties of the particles. A potential limitation of endogenous labeling is the need to know the tracer enrichment in the liver of the amino acid precursors of apo(a). The precursor enrichment for apo(a) is assumed to be the same as that for apoB100, which was determined from its enrichment in plasma VLDL at plateau. It has been established that the leucine-apoB enrichment in VLDL is an accurate measure of the amino acid enrichment of other hepatic secretory proteins, such as albumin and fibrinogen (63). Simultaneous determination of the kinetics of plasma apolipoprotein A-I (predominantly synthesized in the liver), using radioiodinated or endogenous labeling, provided the same results, suggesting no serious bias with either method (64). The metabolic rates of hepatic secretory proteins, calculated using the leucine enrichment in apoB, are moreover independent of dietary protein intake (63). The relatively short period of tracer infusion in our study, 12 h, has the advantage that the results were not influenced by recycling of other plasma proteins into the apo(a) precursor pool. A theoretical disadvantage of a short time of study is that the enrichment curve of apo(a) may exhibit complexity in later hours, necessitating division of the plasma apo(a) pool into multiple compartments. However, studies using radioiodinated Lp(a) found relatively simple kinetics, because the urine-to-plasma ratio of radioactivity was constant for 10 days after injection of 125-I-Lp(a) (38), and the plasma decay curves for 125I-Lp(a) were fit with a classical 2-pool model (38, 62).
We compared the kinetic characteristics of Lp(a) in published studies
(Table 5). The mean FCRs for apo(a) in
the present study, 0.36 pools/day during placebo and 0.31 during
estrogen, are similar to the means for normolipidemic persons in
previous studies that used radioiodinated Lp(a) as a tracer, 0.26 to
0.33 (38, 39, 41, 42, 53) (Table 5). In contrast, Morrisett et
al. (65) found a mean apo(a) FCR of 0.16 in five normal persons,
using endogenous labeling with deuterated lysine rather than leucine.
Studies of apoB metabolism show no difference in results using these 2
amino acids for tracers (66). The individual FCRs in Morrisett et
al. (65) (0.085 to 0.267/day) were in the range for the women in
our study, and perhaps the difference in means is caused by the small
sample size. The mean production rate of Lp(a) in the present study,
4.0 mg/kg·day during placebo, is similar to the previous studies that
measured production rates in persons with normal plasma Lp(a) levels
(38, 39, 41). Overall, the similarity of Lp(a) kinetic parameters among
studies suggests that the impact of methodological differences is
small.
|
In conclusion, postmenopausal estrogen replacement lowers plasma apo(a) production, thereby reducing plasma Lp(a) concentration. This metabolic effect can be added to other lipoprotein effects of estrogen, including increased production of triglyceride, VLDL and LDL apoB, apo A-I, and LDL receptors (69). Reduction by estrogen in the rate of secretion into plasma of a potentially atherogenic lipoprotein particle is an evident benefit. Reduced production of Lp(a) could contribute to the lower cardiovascular disease rates in women receiving estrogen replacement therapy.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received January 30, 1998.
Revised May 14, 1998.
Accepted June 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
55 years of age. Circulation. 94:1263–1268.This article has been cited by other articles:
 |
|
 |
|
  E. Merki, M. J. Graham, A. E. Mullick, E. R. Miller, R. M. Crooke, R. E. Pitas, J. L. Witztum, and S. Tsimikas Antisense Oligonucleotide Directed to Human Apolipoprotein B-100 Reduces Lipoprotein(a) Levels and Oxidized Phospholipids on Human Apolipoprotein B-100 Particles in Lipoprotein(a) Transgenic Mice Circulation, August 12, 2008; 118(7): 743 - 753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Zheng, C. Khoo, J. Furtado, K. Ikewaki, and F. M Sacks Dietary monounsaturated fat activates metabolic pathways for triglyceride-rich lipoproteins that involve apolipoproteins E and C-III Am. J. Clinical Nutrition, August 1, 2008; 88(2): 272 - 281. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Zheng, C. Khoo, K. Ikewaki, and F. M. Sacks Rapid turnover of apolipoprotein C-III-containing triglyceride-rich lipoproteins contributing to the formation of LDL subfractions J. Lipid Res., May 1, 2007; 48(5): 1190 - 1203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Ramakrishnan Studying apolipoprotein turnover with stable isotope tracers: correct analysis is by modeling enrichments J. Lipid Res., December 1, 2006; 47(12): 2738 - 2753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lew, P. Komesaroff, M. Williams, T. Dawood, and K. Sudhir Endogenous Estrogens Influence Endothelial Function in Young Men Circ. Res., November 28, 2003; 93(11): 1127 - 1133. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Marcovina, M. L. Koschinsky, J. J. Albers, and S. Skarlatos Report of the National Heart, Lung, and Blood Institute Workshop on Lipoprotein(a) and Cardiovascular Disease: Recent Advances and Future Directions Clin. Chem., November 1, 2003; 49(11): 1785 - 1796. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. W. Wu and A. von Eckardstein Androgens and Coronary Artery Disease Endocr. Rev., April 1, 2003; 24(2): 183 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
References Circulation, December 17, 2002; 106(25): 3373 - 3421. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
—-, estrogen;
—, placebo.
—, small isoform, 551 kDa; 
