Higher values of hepatic lipase activity in postmenopause: relationship with atherogenic intermediate density and low density lipoproteins

Abstract

Objective 

To investigate the enzymatic activity of hepatic lipase (HL) in postmenopausal women (PMW) and reproductive age women (RAW); and to evaluate the relationship between this enzyme and the atherogenic intermediate density lipoproteins (IDL) and low density lipoproteins (LDL), and antiatherogenic high density lipoproteins (HDL) and its subfractions (HDL₂ and HDL₃).

Design 

We studied 55 PMW receiving no hormonal treatment in a cross-sectional study in comparison with a control group of 55 RAW, matched by body mass index. Follicle-stimulating hormone was > 40 mUI/ml in PMW and 3–12 mUI/ml in RAW. PMW presented at least 1 year of natural menopause and no more than 10 years of amenorrhea with E₂ serum concentration <15 pg/ml.

Results 

HL activity was significantly higher in PMW versus RAW (14.0 ± 1.4 vs. 10.9 ± 0.4 μmol of fatty acids/ml of postheparin plasma, respectively, mean ± SEM, p < 0.001). In PMW, IDL cholesterol showed a positive correlation with LDL cholesterol (r = 0.28, p < 0.05), and HDL₂ cholesterol was inversely correlated with HL activity (r = 0.31, p < 0.05). HL was positively correlated with plasma concentration of LDL cholesterol in both groups (r = 0.27, p < 0.05). The higher values of HL activity and IDL cholesterol were independent of age.

Conclusions 

Higher HL activity is associated with a more atherogenic profile in PMW.

Coronary atherosclerosis is the most important cardiovascular disorder, and its incidence increases with age in both sexes. Risk factors have been less studied in women than in men. However, it is known that after menopause, women present an increase in the incidence rates of cardiovascular disease that rapidly rise to those in men. ¹ Numerous studies have demonstrated that the increase in cardiovascular risk after menopause and in premature menopause may be partly attributable to the drop in sex hormones. ^2–4

The decrease in estrogens is associated with the decrease in the number of low density lipoprotein receptors, named B:E receptors, which may explain the high levels of atherogenic lipoproteins, intermediate density lipoprotein (IDL) and low density lipoprotein (LDL), described after menopause. ⁴ Plasma concentration of IDL and LDL cholesterol depends on the amount of their respective precursors, very low density lipoprotein (VLDL) and IDL, and on the activity of lipoprotein lipase (LPL). On the other hand, hepatic lipase (HL) is a key enzyme in the conversion of IDL to LDL and of HDL₂ to HDL₃. HL activity may alter the pathway of reverse cholesterol transport because it has been postulated that increased HDL₂ cholesterol evidences a more effective antiatherogenic mechanism. This enzyme is subject to hormonal control. Many studies in vitro and in vivo have demonstrated that the enzyme is inhibited by exogenous estrogens and stimulated by exogenous androgens. ^5–7

However, limited data are available on similar effects of endogenous sex steroid hormones. ^8,9 In men with cardiovascular disease, it has been described that an increase in HL activity leads to a proatherogenic proclivity with decreased HDL cholesterol and formation of small and dense LDL particles. ¹⁰

Epidemiological studies point out the increase in LDL cholesterol in postmenopause, but plasma IDL concentrations, their structure, and the relationship between IDL and LDL have not been studied in a similar way; neither is there agreement on the plasma concentrations of total high density lipoprotein (HDL) and its subfractions HDL₂ and HDL₃.

Therefore, the aim of this study was to investigate the enzymatic activity of HL in postmenopausal women (PMW) receiving no hormonal treatment to evaluate the relationship between this enzyme and the atherogenic (IDL and LDL) and antiatherogenic (HDL) lipoproteins in a cross-sectional study in comparison with a control group of reproductive age women (RAW).

METHODS

Patients and controls

A total of 126 Caucasic healthy women were studied. Sixty-three were PMW clinically evaluated at the Climacteric Section, Gynecology Department of the Internal Medicine Hospital, University of Buenos Aires, Argentina, and their ages ranged from 49 to 65 years (mean age 55 years), with at least 1 year of natural menopause and no more than 10 years of amenorrhea, and they received no hormonal or hypolipidemic treatment. The control group comprised 63 healthy women of reproductive age (RAW), recruited at random among the laboratory's staff and employees of the Clinical Biochemistry Department, and their ages ranged from 18 to 35 years (mean age 30 years), and they received no hormonal contraceptive treatment. All the women had a body mass index (BMI) below 27 kg/m². Due to the influence of BMI on lipid and lipoprotein metabolism and on HL activity, 55 women from each group were selected and matched according to their BMI, PMW 23.2 ± 2.0 and RAW 22.8 ± 1.6 (media ± SE, NS). Follicle-stimulating hormone (FSH) was > 40 mUI/ml in PMW and 3–12 mUI/ml in RAW. PMW had estradiol serum concentration < 15 pg/ml, clearly below the RAW value of 50–80 pg/ml. None smoked or had hypertension or diabetes, or had a history of hypothyroidism, neoplasia, or renal disorder. None presented with any kind of chronic disease or suspected tumors. There was no history of cardiovascular disorder or any other disease liable to affect lipid metabolism. In no case did alcohol consumption surpass 15 g/day. Patients and controls were not under a regular training exercise. The participants followed a typical Argentine diet, consisting mainly of red meat, vegetables, and fruits; none of them followed any special diet. Written informed consent was obtained from each subject before admission to the study, which was approved by the Ethics Committee of the Hospital.

Samples

Throughout, blood samples were collected from patients and controls by venipuncture, between 8:00 and 10:00 a.m., after 12 hours of fasting. In the RAW group, blood samples were drawn at the follicular phase (day 3–8) of the menstrual cycle. Serum was separated by centrifugation at 3000 rpm for 15 minutes, within 1 hour of having been extracted; disodium EDTA, 1.5 mg/ml of serum, and 0.1 mg/ml of sodium azide were added to inhibit lipoprotein peroxidative degradation and bacterial growth. Plasma was kept at 4°C until its processing, within 48 h.

For the determination of HL activity, after blood sampling, heparin (60 IU/kg body weight) was administered intravenously. Ten minutes later, blood obtained by venipuncture of the contralateral arm was collected in tubes on ice. Postheparin plasma (PHP) was obtained by centrifugation at 2500 rpm at 4°C for 10 minutes and kept at −70°C until its processing within 30 days.

General analytic methods

Total cholesterol and plasma triglycerides (TG) were determined in a Hitachi 727 autoanalyzer by enzymatic methods, standardized by a Boehringer Mannheim kit (Mannheim, Germany), with standard and controls in each run. Intra-assay coefficient of variation (CV) for total cholesterol was 1.11% and interassay CV 1.52%. TG intra-assay CV was 1.32% and interassay CV 2.55%. LDL cholesterol was determined by precipitation with a solution of polyvinylsulfate 10 g/L dissolved in polyethylenglycol (MW 600) at 25%, pH 6.7. ¹¹ Total HDL cholesterol was determined by precipitation of apo B containing lipoproteins with dextran sulfate (MW 50,000) and MgCl₂ 1 M ¹² and enzymatic measurement of cholesterol in the supernatant. HDL₃ cholesterol was determined by precipitation with dextran sulfate (MW 50,000) and MgCl₂ 2 M according to the method of Warnick ¹³ and enzymatic measurement of cholesterol in the supernatant. HDL₂ cholesterol was calculated as the difference between total HDL and HDL₃ cholesterol. VLDL cholesterol was calculated by subtracting the cholesterol concentration of the supernatant obtained after precipitation with polyvinylsulfate from the HDL cholesterol level.

Quality control was carried out from starting precipitation with low, medium, and high serum pool concentrations kept at −20°C. Intra-assay CV was 3.2% for total HDL cholesterol, 5.4% for HDL₃ cholesterol, and 4.7% for LDL cholesterol. Determinations of total cholesterol, TG, LDL cholesterol, and HDL cholesterol were under the control of the Randox International Quality Assurance Program (RIQA) traceable to the Dr. Russell Warnick Laboratory (Seattle, WA).

Measurement of IDL cholesterol

IDL cholesterol measurement was carried out according to the method of Wikinski et al. ¹⁴ Briefly, a first electrophoretic run of the serum was made in agarose gel (Sigma A-6013; Sigma, St. Louis, MO) at 1% in veronal buffer pH 8.6, with an intensity of 3 mA per slide. Two hours later, IDL was precipitated with heparin-MgCl₂ by placing the slides in a reagent consisting of MgCl₂, heparin, and NaCl. Once the IDL area was identified (band β), LDL was separated by subjecting the slide to a second electrophoretic run, with an intensity of 4.8 mA per slide. The band of the gel corresponding to precipitated IDL was cut, and the cholesterol was extracted with butanol and measured according to the ferric chloride method, using a cholesterol standard in isopropanol (5 mg/dl).

Measurement of hepatic lipase activity

HL activity was determined in PHP by measurement of the oleic acid produced from the enzyme-catalyzed hydrolysis of a triolein emulsion carrying a [³H]triolein radiolabel (Amersham TRA.191; Amersham, Buckinghamshire, UK). ¹⁵ The assay mixture containing labeled and unlabeled triolein (Sigma T-7140) (1.3 μmol/ml of glyceryl trioleate with a specific activity of 10 × 10⁶ cpm/μmol) was mixed with 0.11 μmol/ml of l-α-lysophosphatidylcholine (Sigma L-4129), 0.2% bovine serum albumin (Sigma A-6003), in 0.2 M buffer (Tris-HCl pH 8.8 with NaCl 0.15 M). This mixture was incubated with PHP in saline solution 1:10 and NaCl 1 M for 30 minutes at 30°C. After incubation, the reaction was stopped and released fatty acids were isolated by extraction with a carbonate-borate buffer, pH 10.5. The [³H]oleic acid was quantified by counting with a Liquid Scintillation Analyzer (Packard 2100 TR; Packard Instruments, Meridian, CT).

HL activity was calculated from the difference in counts per minute between the blank and the sample. HL activity measurement was validated by an inhibition assay using an antiserum to human HL, kindly provided by Dr. E. Cavallero (Hôpital Henri Mondor, Creteil, France).

In each assay, seven determinations were carried out, each one in triplicate, three samples from the PMW group randomly paired with three samples from the RAW group. In each series of determinations a known reference sample was included in triplicate. Using triplicate analysis, intra-assay CV was 4% and observed interassay CV 9%. Due to the complexity of this assay, the CV is considered to be quite satisfactory.

Statistical analysis of results

Analysis of the results was carried out by the Student's t test for independent groups. The analysis of the HL activity was evaluated by the Mann Whitney U test for non-Gaussian distributions as required. The analytical results are expressed as means ± SEM. Given the statistically significant difference between the mean ages of the PMW and RAW groups, differences between groups were examined by analysis of covariance (ANCOVA) with age (years) as covariate. Correlations between variables were performed using the Pearson correlation test. A probability value lower than 0.05 (two-tailed) was considered significant. Data handling and statistical analysis were performed with the STATISTICA for Windows software program.

RESULTS

As seen in Table 1, mean serum concentration of total cholesterol in PMW exceeded the desirable maximum value of 200 mg/dl and was also significantly greater than that of the RAW. PMW presented higher mean plasma TG with regard to RAW, even though they did not exceed the desirable maximum limit of 200 mg/dl. ¹⁶ Concomitantly, higher VLDL cholesterol was observed in PMW versus RAW. In both groups a direct and highly significant correlation was found between plasma TG and VLDL cholesterol (r = 0.89 in PMW and r = 0.44 in RAW, p < 0.001). The plasma TG/VLDL cholesterol ratio was 5.6 and 5.0 in PMW and RAW, respectively (p = NS).

TABLE 1: Plasma concentration of lipids and lipoproteins in PMW and RAW controls (mg/dl; mean ± SEM)

IDL cholesterol was significantly higher in PMW, with the mean value in the general population's 90th percentile. ¹⁴ LDL cholesterol was also higher in the PMW group. IDL cholesterol showed a modest but significant positive correlation with LDL cholesterol in PMW (r = 0.28, p < 0.05) but not in RAW (Fig. 1). The higher the precursor's concentration of IDL, the higher the concentration of LDL.

FIG. 1.: IDL showed a significant positive correlation with LDL cholesterol in PMW (n = 55). The higher the precursor's concentration of IDL, the higher the concentration of LDL. This correlation remained significant after adjusting for age.

Mean serum concentration of total HDL cholesterol in PMW failed to differ from that in RAW. Only three PMW subjects with increased plasma TG presented values of total HDL cholesterol lower than the desirable minimum of 40 mg/dl in plasma. Nor was any difference observed in the plasma concentration of cholesterol subfractions HDL₂ and HDL₃, or in total HDL/HDL₂ cholesterol ratio, between the groups (Table 1). On the other hand, VLDL cholesterol correlated inversely and significantly with total HDL cholesterol in both groups (r = 0.63, p < 0.001 and r = 0.39, p < 0.05, in PMW and RAW, respectively).

PMW presented greater HL activity than controls (PMW 14.0 ± 1.4 μmol/ml PHP.h; RAW 10.9 ± 0.4 μmol/ml PHP.h;p < 0.001, Table 2). The distribution of HL activity was non-Gaussian in both groups.

TABLE 2: HL activity in PMW and RAW controls (μmol fatty acids/ml PHP · h)

HL activity showed a significant positive correlation with LDL cholesterol in both groups, (r = 0.27, p < 0.05, similar results in both), whereas a modest significant inverse correlation existed between HDL₂ cholesterol and HL activity only in the PMW group (r = 0.31, p < 0.05) (Fig. 2).

FIG. 2.: Correlation between hepatic lipase (HL) activity and HDL2 cholesterol in PMW (n = 55). HDL2 cholesterol in PMW correlated inversely and modest significantly with HL activity (p < 0.05). Adjustment for age did not substantially alter this correlation.

When lipid and lipoprotein values were adjusted for age, the PMW group showed a significantly higher mean total cholesterol (p < 0.02), LDL cholesterol (p < 0.05), and IDL cholesterol (p < 0.02). However, TG and VLDL cholesterol concentration lost significance (p = 0.09) (Table 1). In both groups taken separately, no correlation was found between age and HL activity, and after the adjustment for age, HL activity remained higher in PMW as compared with RAW (p < 0.001).

After adjusting for age, HDL₂ cholesterol remained significantly related to HL activity, as well as IDL to LDL in PMW (r = 0.31, p < 0.05 and r = 0.28, p < 0.05, respectively), whereas the relationship between LDL cholesterol and HL activity was no longer significant in any group (r = 0.20, p = 0.1).

We did not find a significant correlation between plasma estradiol and HL activity in either group.

DISCUSSION

In this study we demonstrate higher HL activity and IDL cholesterol concentration in PMW without hormonal replacement therapy in comparison with women in their reproductive age. The concentration of total serum cholesterol in the PMW group was significantly greater than that in the RAW group. This difference was in the levels of VLDL, IDL, and LDL cholesterol, without a difference in the level of total HDL cholesterol. HDL₂ cholesterol showed an inverse significant correlation with HL activity in the PMW group only. After adjusting for age, HL activity, IDL cholesterol, LDL cholesterol, and total cholesterol remained significantly higher in the PMW group.

Hepatic lipase, an enzyme that regulates IDL to LDL, as well as catabolism of subfractions HDL₂ to HDL₃, has been rarely studied in PMW without hormonal replacement therapy. In groups with a lower number of patients, other authors ^8,17 did not find any differences in HL activity between pre-and postmenopausal women.

Depress et al. ¹⁸ found a correlation between BMI and HL activity in obese women. In our case, patients and controls were matched for BMI to avoid the effect of this condition on HL activity. However, in subjects with the common hepatic lipase gene (LIPC), it has been shown that HL activity is up-regulated by increasing amounts of intra-abdominal fat independently of total body adiposity. ¹⁹ Because we have not considered the evaluation of intra-abdominal fat in these women, we cannot rule out that the increase in HL activity in PMW could be a consequence of the accumulation of intra-abdominal fat, which can be observed in this population, even in women belonging to a group with an average BMI of 23 kg/m².

We found higher VLDL associated with plasma TG and VLDL cholesterol levels. The average plasma TG/VLDL cholesterol ratio was similar to that of the RAW group (5.6 vs 5.0), so that in both groups the VLDL particle, precursor of IDL, did not show a qualitative alteration such as an increase in cholesterol content.

In plasma, IDL concentration depends on the catabolism of VLDL by LPL, on HL activity, and on its catabolism via B:E receptors. The IDL cholesterol was higher in the PMW group, an alteration related to a greater supply of the VLDL precursor associated with the previously described increase of LPL activity. ⁸ Due to the lower number of B:E receptors associated with a deficit in estrogen secretion, IDL catabolism at the receptor level would be impaired and would not be completely compensated by the greater HL activity. IDL degradation to LDL is significantly related with IDL supply, so plasma concentration of LDL cholesterol again depends on HL activity in the presence of B:E receptor deficit. In contrast, in the RAW group, VLDL supply is smaller and IDL is partly catabolized in the receptors. As well LDL is generated due to normal HL activity. The positive correlation observed between LDL cholesterol and HL in the whole population is attributable to successive hydrolysis of the TG contained both in IDL and in LDL, leading to relative enrichment of the LDL particle in cholesterol. ¹⁰

Circulating IDL and LDL can reach the subendothelium but, due to their larger size as compared with LDL, reflux of IDL from the intima toward the circulation is smaller and, in contrast with LDL, particles may be taken up by macrophages to produce foam cells, without need of prior modification. ²⁰ Hodis et al. ²¹ demonstrated in a group of men that the progression of the atherosclerotic lesion is associated with increased IDL mass, even when changes in LDL, VLDL, HDL₂, or HDL₃ mass are not observed. The striking affinity of IDL for the proteoglycans of the arterial wall has also been reported, ²² especially the smallest and densest fraction, which is found to increase after menopause. ²³

We did not find differences in total HDL cholesterol between the PMW and the RAW groups. Results from the Lipid Research Clinics indicate that if HDL level is high before menopause, it is not decreased by this condition. ²⁴ Our results are also in agreement with that of the Italian Climacteric Research Group Study (ICARUS). ²⁵ However, HDL subfractions play a different role in cholesterol transport. In PMW, HDL₂ cholesterol showed a slightly significant inverse correlation with HL activity. In obese women, St. Amand et al. ²⁶ found a significant inverse correlation between HL and HDL cholesterol, whereas in nonobese women only a weak correlation was discerned between HL activity and HDL apo AI. Given the leading role of HDL₂ in reverse cholesterol transport and the fact that the enzyme is involved in the catabolism of this lipoprotein, the greater HL activity would play an atherogenic role, even though greater formation of preβ-HDL particles ²⁷ and qualitative alterations in lipoproteins with apo A cannot be ruled out. ^28,29 Besides, Hegele et al. ³⁰ have recently observed that the common promoter variation in LIPC, which has been reported to be related to changes in HL activity, is not always associated with variation in plasma HDL cholesterol concentration. Mahaney et al. ³¹ also reported that hepatic lipase activity was not linked to plasma total HDL cholesterol concentration in Mexican Americans. These results suggest that there must be environmental or genetic factors still unspecified that could influence plasma HDL cholesterol concentration.

In women in a state of pseudomenopause produced by GnRH agonists, the reduction in estradiol secretion is accompanied by an increase in the concentration of LDL. In contrast, the concentration of serum HDL remains unchanged, resulting in a higher LDL/HDL ratio. ³² In an important follow-up study on women from baseline to 3 years after reaching menopausal status, Longcope et al. ³³ found significant correlations between the lower estradiol levels and the increment in apo AI and apo B. In our group of PMW, the higher IDL cholesterol and LDL cholesterol levels provide further data as markers of the increase in the apo B containing lipoproteins.

In both groups serum estradiol was not associated with HL activity, which could perhaps be due to the wide fluctuations in RAW and/or to the low values in PMW. Moreover, we have not considered environmental, dietary, and/or other factors such as the promotor polymorphism, which are also important in determining HL activity.

In summary, the most relevant findings in this study on postmenopausal versus reproductive age women are the higher HL activity and plasma concentration of IDL cholesterol in PMW, which are independent of age.

When correlated with the higher LDL cholesterol, they display the atherogenic characteristic of this profile. However, because the intra-abdominal fat ratio was not measured, it is not possible to exclude that fat redistribution in body central regions in PMW could explain, at least in part, some of the reported differences in lipid profile.

Acknowledgments:

The skillful assistance of Laura Aisemberg in greatly appreciated.

This study was supported by a grant from the University of Buenos Aires (FA 085).