- Email: [email protected]
Effect of Gemfibrozil on lipids, apoproteins, and postheparin lipolytic activities in normolipidemic subjects
Effect of Gemfibrozil on Lipids, Apoproteins, Activities in Normolipidemic
and Postheparin Subjects
Lipolytic
Agostino Gnasso, Burkhard Lehner, Werner Haberbosch, Ottmar Leiss, Klaus von Bergmann, and Jan Augustin The lipid lowering agent Gemfibrozil was tested in 8 normolipidemic subjects during a three-month intake. Plasma triglycerides decreased by 41% and Very Low Density Lipoprotein (VLDL) triglycerides decreased by 54%. The reduction of plasma cholesterol, less marked (by 10%). was due to a decrease of Low Density Lipoprotein by 20% while High Density Lipoprotein (HDL) increased up to 30%. The separation of HDL demonstrated that only HDL, were increased. The determination of the apoproteins in plasma and lipoprotein fractions showed similar results with a decrease of apo 6 (by 20%) and an increase of apo A-l and apo A-II. mainly in the HDL, fraction. Plasma postheparin lipolytic activities (PHLA) were not influenced by the therapy and no correlation was found between these activities and any of the plasma or lipoprotein lipids. The apo C-lll/apo C-II ratio in VLDL decreased by 30%; however, no correlation was found between this ratio in plasma as well as VLDL and triglycerides. In addition, the lntra Venous Fat Tolerance Test did not demonstrate any improvement of the clearance of exogenous fat. The lipid lowering efficacy of Gemfibrozil, its collateral effects, and the possible mechanisms of action are discussed. B 1986 by Grune & Stratton, Inc.
H
YPERCHOLESTEROLEMIA is one of the major risk factors for the development of coronary heart disease.‘” The role of triglycerides in the progression of atherosclerotic lesions is less clear.7-9 Several studies have demonstrated that the elevation of triglyceride-rich lipoprotein in plasma is accompanied by a further increase in risk for vascular diseases; others have shown the contrary. During the intravascular hydrolysis of Very Low Density Lipoproteins by the action of lipolytic enzymes, surface components are transferred to discoidal nascent HDL for the formation of HDL.” This mechanism in part explains the inverse correlation between VLDL triglycerides and HDL cholesterol in the plasma compartment. It has been suggested that lipoprotein lipase (LPL) is involved in the interconversion of HDLl to HDLI.” An opposite role was suggested for hepatic triglyceride lipase (HTGL).12 Treatment with clofibrate derivatives and/or analogues in patients with hypertriglyceridemia results in a decrease in VLDL triglycerides and a rise in HDL cholesterol.‘3-15 This has been attributed to an activation of the lipolytic system. However, the stimulation in postheparin lipolytic activities after treatment was small even when therapy was substantial. Thus it is unclear whether a modification in lipase activities, measured in plasma after heparin injection during clofibrate therapy, is the cause or the consequence of decreased VLDL concentrations in plasma. The present study was designed in order to test the effect of the lipid lowering agent Gemfibrozil on plasma lipoproteins, lipolytic activities, and triglyceride clearance capacity in normals. The data demonstrate a profound influence of the drug on the concentration of VLDL and HDL components, with no major change in postheparin lipolytic activities. MATERIALS
AND METHODS
Eight healthy male volunteers in the age range of 34.5 f 3.2 years (mean years *SD). normolipemic with no drug intake, were recruited from the personnel of the clinic. This reduces the likelihood of confounding variables that could affect the outcome of this study. The isocaloric diet that was taken by all subjects in the hospital consisted of 40% fat, 40% carbohydrates, and 20% proteins. The P/S ratio was 0.33. Alcohol intake during the study was less than 20 g/d. Mean ideal body weight was 101.5 + 8.3% (range, 90% to 115%) Metabolism, Vol35,
before the study started and did not change at all during the following months. Six subjects were nonsmokers, two smoked moderately (10 to 20 cigarettes/d) and did not change their habits during the study. No modifications in physical activity habits were reported. The study was open and conducted on an outpatient basis, in accordance with the principles of the Helsinki Declaration. The nature of the study was fully explained and written consent was obtained from each subject. Gemfibrozil, 1,200 mg (600 mg b.i.d.), were administered for three months. Blood samples were collected on Friday morning in order to omit weekend effects before and at monthly intervals during the study and four weeks after the drug was discontinued. Twenty mL of blood for lipid and apoprotein analysis, 10 for zonal ultracentrifugation, and 20 for liver and renal function tests, hematologic parameters, and electrolytes were collected in the morning, after a 12-hour fast, from an antecubital vein, before 100 U of heparin per kg body weight were injected. Five mL blood were drawn 15minutes later to measure postheparin hpolytic activities (PHLA). The Intravenous Fat Tolerance Test (IVFTT), determination of Iithogenic index, lipid composition of cholecystokinin-stimulated fasting duodenal bile, and hepatic secretion of biliary lipids were performed before and after three months of therapy. At these time points, therefore, determinations of plasma lipids were performed, three times in three following days. During the study, instead, blood for lipid determinations was taken two times. The results reported are the mean of the different determinations. EDTA blood was centrifuged for 30 minutes at 2,500 rpm and 4 “C in a Beckman J6B centrifuge. Ten mL of the obtained plasma were filled in Beckman Quick Seal tubes (16 x 76 mm) overlayered with 0.9% NaCl and centrifuged (50,000 rpm) in a Beckman L8-70 preparative ultracentrifuge, using the Beckman 70.1 Ti rotor, for 22 hours at 4 “C. The Very Low Density Lipoproteins (VLDL) were obtained by slicing the tubes. The infranatant underwent identical treatment after adjustment of the density to 1.019 g/mL by addition of KBr. After further 22 hours ultracentrifugation, the IDL were separated in the same way. By additional steps at densities of 1.063,
No 5 (May), 1986: pp 387-393
From the Department of Medicine. University of Heidelberg, and Department of Medicine, University of Bonn. Address reprint requests to Agostino Gnasso. MD. Klinisches Institut fGr Herzinfarktforschung. Bergheimer Str 5X. DtT900 Heidelberg, FRG. 8 1986 by Grune & Stratton, Inc. 00264l49S/86/350%0002$03.00/0
387
388
GNASSO ET AL
1.125, and 1.210 g/mL the Low (LDL) and High Density Lipoprotein subfractions (HDL,/HDL3) were obtained. The densities were controlled by a Heraeus-Paar DMA 45 Densitometer. The height at which the tubes were sliced was obtained empirically by measuring the density and performing lipid electrophoresis and zonal ultracentrifugation of the obtained fractions. The VLDL, IDL, and LDL were perfectly separated from each other. Zonal ultracentrifugation demonstrated a slight overlap between the two HDL fractions obtained by sequential ultracentrifugation. However it was not possible to obtain a better separation of these two fractions varying the height of the cut. The lipoproteins with density cl.063 g/mL contained about 2% of albumin, while the HDLl and HDL, were free of apo B. In the latter fractions, however, albumin represented 4% to 8% of the total protein bulk. In the 1.21 g/mL infranatant, or lipoprotein-free plasma, 4% to 6% of the total triglycerides, 3% to 4% of total cholesterol, and about 5% of the different apoproteins were found. An aliquot of the obtained fractions was used to determine the lipid concentrations while the rest was dialyzed against distilled water (Visking Dialysis Tubing, type 8/32, Serva, Heidelberg, FRG), lyophilized and delipidated by 2 x 5 mL ethanol/ether (3:1, v:v) for 20 and 4 hours and finally 5 mL ether for 1 hour at - 18 OC. Isoelectric focusing (IEF) on ultrathin (0.3 mm) flat gels was used for separation and quantification of the apoproteins. The gels were prepared in our laboratory in the following way: a 7% acrylamide solution with 2% ampholynes (either pH 4-6 or 6-8, Serva, Heidelberg, FRG) was polymerized on plastic sheets, which then functioned as carriers. Anolyte and catholyte were 0.5 M H,P04 and 0.4 M NaOH, respectively. After a prerun of about 1 hour to allow the formation of the pH gradient, 20 pL of the delipidated proteins dissolved in 7 mol/L urea, Tris-HCI pH 8.3, were applied to the plate using rubber wells. Focusing time was 3 hours at 1000 V, amperage free. After this time the protein bands were fixed for 10 minutes with 10% trichloroacetic acid and five minutes with 7% acetic acid. Staining followed for 15 minutes in 1% Light Green (Serva, Heidelberg, FRG) and destaining for 3 x 10 minutes with 7% acetic acid. Normally the background was clear after this procedure. A typical pattern for apo
APO VLDL
APO HDL
IPHI 0
VLDL and apo HDL is shown in Fig 1. The density of the stained bands, determined by an LKB 2202 Ultra Scan Laser Densitometer, was compared with that of standards of known concentrations. The purified apo-proteins, used as standards, were obtained from plasma of normal, for apo A-I and apo A-II, or hypertriglyceridemic subjects, for C-peptides, by gel filtration, affinity chromatography, and preparative isoelectric focusing. The purity of the apoproteins was controlled with excess amounts of proteins by SDS-gel electrophoresis and IEF. In addition no cross-reaction with antibodies against other apoproteins, respectively, was observed. Standard curves were performed with every isolated apoprotein. When the concentration of the sample to be tested fell out of the rectilinear part of the standard curve, appropriate dilutions were made. The C-III isoforms showed identical chromogeneity, however different from apo C-I and apo C-II. The coefficients of variations for apo A-I, A-II, C-I, -C-II, and -C-III were 3.73%, 2.64%, 2.19%, 2.37%, and 1.44%, respectively. Apo A-I and apo A-II, in plasma and in the HDL fractions, were determined also by immunodiffusion (Immuno AG, Vienna, Austria). The monospecificity of the apo A-I and apo A-II antibodies was detected with pure A-I and A-II apolipoproteins. No crossreaction was observed. The coefficients of variations were 5.35% for apo A-I and 6.10% for apo A-II. Quantification of apo B was performed by immunodiffusion on NOR-partigen plates (Behringwerke AG, Marburg, FRG) (cv 4.72%). Cholesterol, triglycerides, and phospholipids were measured using commercially available kits (Boehringer, Mannheim, FRG). HDL cholesterol was determined also after precipitation of the apo-B-containing lipoproteins by addition of phosphotungstic acid and magnesium ions (Boehringer, Mannheim, FRG). Zonal ultracentrifugation was carried out according to Patsch et al.16 HDL subclasses were separated in nonlinear sodium bromide gradients ranging in density from 1.00 to 1.40 g/mL, using a Beckman Ti-14 rotor, at 44,000 rpm and 4 ‘C for 22 hours. The activities of lipoprotein lipase (LPL) and hepatic triglyceride lipase (HTGL) were measured by immunoassay with goat antibodies against HTGL as described earlier.” The IVFTT was performed by injecting 1 mL Intralipid pro kg body weight within 1 minute (Intralipid 10% Vitrum, Deutsche Kabi Vitrum, Munich, FRG). Blood samples were collected before and every five minutes up to 40 minutes after injection, centrifuged, and turbidity was measured in plasma, diluted 1:11 with 0.9% NaCl at 691 nm.‘* The elimination rate of the fat was calculated by the following formula: I,F, = at + c
1 Al
0
where FL represents the difference between the extinctions of the samples obtained after and before Intralipid injection. Triglyceride concentration was determined from the same samples. The results of the lithogenic index, lipid composition of gall bladder, and hepatic bile before and at the end of drug intake are published in this journal.t9 The results were analyzed by Friedman Test and the significance of the differences between the time points was evaluated by Wilcoxon test. Relationships among plasma lipid, apoprotein, and lipase values were evaluated by Spearman’s rank correlation coefficients?’
PH 6 RESULTS Typical IEF pattern of apo VLDL and apo HDL from four Fig 1. different subjects. The samples were applied at the cathod. Run conditions and preparation of the protein samples are given in the
text.
The plasma lipid concentrations are given in Table 1. The triglycerides fell progressively during the drug intake up to 44% after three months, while a less marked decrease of
EFFECT OF GEMFIBROZIL ON LIPIDS
Table 1. Lipid and Apoprotein Were Obtained
389
Parameters
in Plasma From Normals Before, During, and After Gemfibrozil
Before and After 4,8,
Week
and 12 Weeks
of Gemfibrozil
Intake and 4 Weeks
4
0
Intake (mg ? SD) (Values
After Omission of the Drug)*
6
12
16
TG
111 f 48
82 + 31
64 + 23t
62 ? 21t
94 + 33
CHOL
177 f 42
157 f 24t
157 + 23t
156 -f- 28T
174 f 27
195 + 30
203 & 28
PHOS
210
Apo B
86.3
+ 16.6
68.6
700.1
* 16.8
111.4
Apo A-l lmmunodiffusion
+ 43
188 f 25t
187 ? 30
* 12.lT
67.1
+ 15.4
Apo A-II lmmunodiffusion
33.6
t 8.4
36.3
t 7.1
Apo A-l IEF
85.3
+ 17.5
99.7
+ 14.8
Apo A-II IEF
18.6 + 6.9
21.7
* 6.3
115.4 38.6
f
13.4t
+ 18.3 ? 4.9
69.7
+ 14.lt
122.3
f
41.3
104.5
f 26.8t
118.5
25.6
+ 7.7T
27.7
15.07
84? 106.4
11.9 t 9.7
r 4.2t
35.4
* 3.1
r 17.6t
91.1
+ 10.1
+ 6.lt
20.1
? 3.6
‘Whereas the data determined for apoproteins A-l and A-II in plasma after immunodiffusion represent total, free, and lipoprotein bound apoproteins, the results obtained by IEF are the sum of the values in the lipoprotein fractions. TP < 0.05.
cholesterol and phospholipids was observed (12% and 7%, respectively). Four weeks after the suspension of the drug only triglycerides were still lower than before treatment (15%). Plasma apo B, apo A-I, and apo A-II concentrations are shown in Table 1. For A apoproteins, both the results obtained by immunodiffusion and IEF are reported and they show similar trends, although the absolute values slightly differ. This is true mainly for apo A-II, which is found in part in the 1.21 g/mL infranatant after sequential ultracentrifugation. Apo B decreased during the drug intake by about 20% while the A apoproteins increased significantly especially after three months of therapy. A strong reduction in all VLDL components was observed without change in the ratio of the sum of measured apoproteins to lipids (Table 2). Triglycerides, cholesterol, and phospholipids fell, respectively, by 59%, 51%, and 48%. Apoproteins decreased in the same manner. Apo B was reduced by 50% (from 6.9 to 3.4 mg%) and also the C peptides progressively diminished. However, the ratio of apo C-III and apo C-II changed, the VLDL particles becoming relatively enriched in apo C-II. The apo C-III/ape C-II ratio fell, in fact, by 3 1% after 12 weeks of drug intake compared to the starting values. The decrease of VLDL lipids and apoproteins was almost linear during the three months of therapy with no plateau. IDL were present, as expected, in very small amounts and showed changes similar to VLDL, but highly increased after Gemfibrozil was discontinued. At this time, the IDL were relatively enriched in triglycerides. However, these differences were not statistically significant related to large standard deviations. The protein component of this lipoprotein Table 2. -
-.
Lipid and Apoprotein
Parameters
0
Week
c 39.9
in VLDL Before, During, and After Gemtibrozil 4
lti
70.4
CHOL
16.1 + 7.5
10.1 f 7.1
PHOS
26.9
20.1
+ 16
class was too small to allow quantification of single apoproteins (Table 3). The LDL fraction (Table 4) decreased by about 20% during the drug intake. The triglycerides were markedly reduced (by 28%) after four and eight weeks of Gemfibrozil, slightly increasing (10%) after 12 weeks. The cholesterol and phospholipids were reduced by 22% and 16%, respectively, after four weeks, remaining then constant. The apo B was reduced constantly by about 20%, the protein/lipid ratio of this lipoprotein fraction did not change. In contrast to VLDL the highest reduction of this lipoprotein class was observed already after four weeks of treatment. At the end of the study all these components reached the values observed before drug intake. HDL were analyzed after separation in HDLz and HDL, by sequential flotation, like the other lipoprotein classes. Zonal ultracentrifugation was performed to obtain a qualitative pattern of the HDL subclasses. In addition, HDL cholesterol was determined after precipitation of apo-Bcontaining lipoproteins. The precipitation method gave consistently higher levels of cholesterol than those obtained after ultracentrifugation, but the changes during the study were similar (Fig 2). HDL, did not show major differences in any of the lipid and apoprotein components during treatment (Table 5). Only in the zonal ultracentrifugation was a shift of this subclass to a slightly higher density observed (Fig 3). In the HDL, range, triglycerides remained constant, whereas cholesterol and phospholipids increased by 30% and 15%, respectively, after volunteers had been treated for three months with Gemfibrozil. HDL,-apo A-I increased by 44% and apo A-II by 36% (Table 6). The rise in C peptides was
47.9
c 25.3 * 10.1
34.9
+ 19.21
6.5 + 3?6+
Intake Lmg% i SD) 12
6 29.1
-c 14.7s
16 49.7
’
i- 25.
7.9 * 4.7’
8.3 + 8 17.9 ” 9.7
13.6 f 6.6,
13.9 + 6.8’
Apo B
6.9 + 2.8
4.9 * 1.9
3.9 * 1.5*
3.4 It 1.2*
5.1 + 1.9
Apo C- 1
0.6 2 0.3
0.4 + 0.2’
0.3 f 0.2”
0.2 _t 0.21
0.3 ? 0.
Apo C-II
0.5 * 0.24
0.39
I 0.27
0.32
t 0.22’
0.28
& 0.19*
0.36
I
+ 0.26
Apo C-III
1.8 _+0.8
1.2 _t 1.0
0.9 t 0.7’
0.6 ‘- 0.6’
1.3 + 1.0
Apo C-III/C-II
3.7 + 0.4
2.9 * 0.5’
2.7 k 0.5*
2.6 + 0.7*
3.4 + 0.3
lP < 0.05.
390
GNASSO ET AL
Table 3.
Lipid Parameters
1
IDL
in IDL Before. During, and After Gamfibroril 2
TG
4.38
f 2.20
3.00
CHOL
2.88
k 1.46
1.00 i
1.07
DISCUSSION
In the present study Gemfibrozil was given to eight normolipidemic subjects and several parameters were folIowed up to four months to verify its lipid lowering efficacy and possible collateral effects. The drug efficiently lowered plasma triglycerides and, to a lesser extent, plasma cholesterol and phospholipids. The fall of triglycerides was mainly due to the VLDL fraction, which was relatively enriched in apo C-II. Plasma cholesterol was less reduced than LDL cholesterol, because the decrease of the LDL fraction was in part balanced by the rise of HDL cholesterol. But only the HDL, fraction was influenced by the therapy. Similar effects of clofibrate derivatives and analogues on plasma lipoprotein metabolism have been described. Both
Table 4.
Lipid and Apoprotain
TG
5
3.25
+ 1.83
3.38
+ 1.51
8.00
f 4.69
2.25
+ 1.49
1.50 + 0.76
3.57
f 2.23
influences on VLDL synthesis and catabolism have been taken into consideration?‘-” An increase of postheparin lipolytic activities, as observed in other trials,24*25was not confirmed in the present study, in fact even a slight decrease was detected. However, in the above mentioned trials, patients with hypertriglyceridemia were studied. In these patients, the slight PHLA increase following the drug intake might be the cause as well as the consequence of the normalization of plasma triglycerides. The present data, obtained from normolipemic subjects, for the first time demonstrate that the drug has probably no influence at all on PHLA. These results therefore strongly support the latter hypothesis. In addition, no correlation between lipoprotein lipase or hepatic triglyceride lipase and the major plasma lipids was found. On the other hand, the modified VLDL apoprotein composition, with a percentage increase of the LPL activator apo C-II and a decrease of the PHLA inhibitor apo C-III, is unlikely to explain an improved hydrolysis of these particles. No correlation could be found between the apo C-III/ape C-II ratio and triglycerides either in plasma or VLDL. Apo C-II is present in excess in VLDL.26 In addition, in this study, the strong reduction of the VLDL fraction was not followed by an increase of the IDL or LDL fraction, as expected after higher hydrolysis rate of VLDL. On the contrary, even these fractions showed a reduction. The IDL, already present in very small amounts, further decreased under Gemfibrozil and LDL components were reduced up to 20% (Tables 2,3). Moreover, the fat tolerance test, with similar extinction rates both before and after the medication, and even a decreased elimination rate for triglycerides, demonstrates that the capacity for removal of exogenous fat from the plasma compartment is, if at all, impaired after Gemfibrozil treatment. The decrease of the LDL fraction, maximally evident after four weeks of Gemfibrozil, despite the continuing decline of VLDL, could be the consequence of an adapted reduction of the LDL clearance in order to adjust a new steady state. This is further supported by a strong increase in IDL and LDL after withdrawal of the drug. The modifications observed in the HDL range give further evidence for this mechanism. Recently, as a consequence of the hydrolysis of triglyceride-rich lipoproteins, with transfer of surface material to preexisting HDLJ, an interconversion of these particles to
in LDL Before, During, and After Gemfibrozil
4
Intake (mg% + SD)
12
8
12.8 + 2.4’
12.6 i 3.9’
14.5 * 3.1
+ 29.9
84.8
+ 20.1*
87.5
+ 21.3*
89.1
‘- 22.3
67.4
i
15.5
56.4
+ 13.5’
57.3
r 15.71
58.4
t
77.9
t 14.5
61.5
f 11.1’
61.0
i
64.6
+ 13.7
17.6 -+ 5.7
CHOL
108.5
PHOS Apo 8 lP < 0.05.
Parameters
0
4
3
* 2.00
less pronounced and significant only for apoprotein C-II., Apoprotein C-III remained constant. An increased absorbance and a shift to a higher density was observed for HDL3 after separation by zonal ultracentrifugation (Fig 3). All these modifications were reversible four weeks after the end of the drug period. The postheparin lipolytic activities in plasma, LPL and HTGL, were in the normal range (Fig 4). After four weeks of treatment a significant decrease in LPL activity was detected, whereas no significant change for both activities was observed in the other periods. The IVFTT performed immediately before and after medication revealed no significant difference in the elimination rate of exogenous fat (Fig 5), expressed as change in plasma turbidity. However, if plasma triglycerides were measured at the same time intervals it was found that after the drug intake the bolus injection of Intralipid resulted in higher values than before the medication (Fig 6). The other blood parameters checked did not show any alteration in electrolytes, glucose, protein, and renal tests. A slight reduction of the white blood cells was observed in six subjects after four weeks on Gemfibrozil. The cell number remained then constant and returned back to previous values four weeks after suspension of the drug. In addition, slight elevation of transaminases in two subjects at one of the time points was found. Some patients complained of “transitory abdominal tension” during the study but none had to discontinue the drug for this reason.
Week
Intake (mg% + SD)
12.5*
15.7’
16 17.6 * 2 104 -c 24 68.6
f 13.8
76.5
+ 11.2
391
EFFECT OF GEMFIBROZIL ON LIPIDS
T
60
T
T
il 1 8 ;*
I
16 weeks
HDL chol after precipitation of apo-B-containing lipoFig 2. proteins (-1 and after separation of HDL by sequential ultracentrifugation j----L ‘P < 0.05.
lighter HDLz has been described.” In this study no major alterations of the HDLz fraction took place whereas HDL, showed remarkable modifications under drug exposure. The HDL, composition remained constant throughout the study and the increase was due to both apoproteins and lipids (Table 6). Therefore this rise cannot be explained by a transfer of material from triglyceride-rich lipoproteins. Most
4
0 Postheparin Fig 4. LPL (----I. ‘P < 0.06.
'2 weeks
8
lipolytic activities in plasma: HTGL I-_) and
bE 0.3 ----
week
0
----week
12
aba
0.2
33
f 0.1
I
I
5
A typical zonal ultracentrffugation Fig 3. after Gemfibrozil intake. Table 5. Lipid and Apoprotein W.Sk
0
pattern
before
Parameters
and
I
IO
I
15
I
I
20
25
1
35
I
LO
8
Intake fmg% i SD) 12
16
4.4 ‘r_0.7
4.0 +_ 0.8
3.3 * 2.1
3.5 + 1.2
3.1 + 0.9
CHOL
12.8 + 7.0
14.8 f 4.0
13.8 + 6.8
14.6 & 7.0
13.7 f 6.8
PHOS
17.5 t
16.9 + 9.8
TG
17.0 + 9.3
18.1 i 6.8
17.9 i
Apo C- 1
1.5 + 0.9
1.5 + 0.4
1.9 f
1.7
1.2 + 1.0
0.6 + 0.3
Apo C-II
0.6 + 0.4
0.8 f 0.5
0.7 + 0.6
0.5 2 0.5
0.5 f 0.3
Apo C-III
2.0 t 2.1
2.0 f
l
min
Fig 5. lntra Venous Fat Tolerance Test: E represents the difference between extinctions of plasma taken after and before lntralipid injection. (-1 Before and I----) after three months treatment with Gemfibrozil.
in HDL, Before, During, and After Gemfibrozil
4
I
30
11.9
12.0
1.3
1.9 * 1.7
1.4 +- 1.1
1.4 + 0.7
Apo A-l
14.8 f 8.0
17.2 + 5.1
16.2 f 9.6
17.1 * 9.3
15.1 f 7.4
Apo A-II
2.8 * 2.2
3.0 f 2.1
3.4 f 2.6
2.9 * 2.6
2.6 f 2.0
GNASSO ET AL
392
Table 6. Lipid and Apoprotein Parameters in HDL, Before, During, and After Gemfibroril Intake (mg% + SD) Week
0
a
4
6.8 f 1.8
TG
6.3 k 1.0
12
5.5 * 1.9
16
7.1 + 1.1
7.3 f 0.8
CHOL
28.9 +. 6.8
32.8
k 5.2
32.6
+ 7.1
37.6
+ 3.8”
28.9
e 1.6
PHOS
67.8
66.6
+ 12.3
69.3
+ 13.1
78.3
f
55.7
f 26.1
Apo C-l
+ 12.7
2.7 f 0.7
3.2 f
3.3 f
1.0
1.0
11.7*
3.4 + 1.1
2.1 f 0.2
Apo C-II
1.3 f 0.3
1.7 k 0.3
1.4 f 0.3
1.5 + 0.4
1.3 + 0.1
Apo C-III
3.4 + 1.2
4.3 f
4.2 + 1.4*
4.1 f
3.4 It 0.7
Apo A-l
70.2
Apo A-II
15.8 + 5.1
f 13.8
82.6
1.2
s 12.4
18.7 + 4.6
87.8
+ 20.8*
22.2
f 6.3.
101.4 24.8
1.5,
+ 13.2’
75.8
f 5.0’
17.5 + 2.4
f 8.5
lP -c 0.05.
likely, as a result of Gemfibrozil treatment, the number of HDL, particles in plasma increases. The reason for this effect may be due to increased synthesis of apo A-I and apo A-II demonstrated for patients with hypertriglyceridemia.” An inverse relation between plasma triglycerides and HDL cholesterol in normals and patients with hyperlipoproteinemias has been previously reported.2*-30 In our study there was a slight inverse correlation (r = - Sl) between plasma triglycerides and HDLz cholesterol which however disappeared during and after treatment. In earlier studies mainly an increase in HDL2 cholesterol after Gemfibrozil intake was observed.““’ On the other hand, dietary treatment of patients with type IV hyperlipoproteinemia with normalization of plasma triglycerides did not result in any change of low HDL cholesterol concentrations.32 It was concluded that
I 1
5
I
10
0
-
week
----
week12
I
I
I
I
I
15
20
25
30
LO
+
min
lntra Venous Fat Tolerance Test: The values represent Fig 6. the differences between TO concentrations of plasma taken after and before lntralipid iniection. (-_) Before and (----1 after three months treatment with Gemfibrozil.
in this particular group of patients no interrelationship of the metabolism of VLDL and HDL occurs. These data also suggest a direct influence of the drug on HDL synthesis. Besides ample differences in the methodology it has to be pointed out that the aforementioned studies have been conducted with a heterogenous group of hyperlipidemic subjects. Our data indicate that at least in normals Gemfibrozil acts exclusively on the HDL3 subclass. The C apoproteins have been regarded as an entity that is more or less equally distributed between VLDL and HDL.33T34 Although ultracentrifugal effects cannot be excluded, in this study neither before nor during Gemfibrozil exposure was this pattern observed. In fact, 80% of the C peptides were bound to HDL without treatment and even 90% with treatment. The ratio of the different C apoproteins was constant throughout the study in HDL2 and HDL3. In VLDL, however, the percentage modification was observed exclusively between apo C-I and apo C-II; the contribution of apo C-III remained unchanged. These data suggest that in normals HDL are the major reservoir of C apoproteins in plasma. Moreover Gemfibrozil intake gives evidence for an independent metabolism of the different peptides. In conclusion, Gemfibrozil strongly affects plasma lipoproteins in normals resulting mainly in a decrease in VLDL and a rise in HDL3. Whereas the influence on VLDL is probably mediated by an inhibition of the synthesis of these particles, further investigations are needed to clarify the role of this drug on HDL metabolism. The effect of Gemfibrozil on the parameters reported in this study in patients with different types of hyperlipoproteinemias is currently under investigation. ACKNOWLEDGMENT
The authors thank 1. Geldmacher, E. Glatting, M. Miltner, D. Schraube for their excellent technical assistance.
and
REFERENCES
1. Cohn PF, Gabbay SI, Weglicky WB: Serum lipid levels in angiographically defined coronary artery disease. Ann Int Med 84:241-245, 1976. 2. Goldstein JL, Hazzard WR, Schrott HG, et al: Hyperlipidemia in coronary heart disease. I. Lipid levels in 500 survivors of myocardial infarction. J Clin Invest 52:1533-l 544, 1973 3. Goldstein JL, Schrott HG, Hazzard WR, et al: Hyperlipidemia in coronary heart disease. II. Genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder, combined hyperlipidemia. J Clin Invest 52: 1544-l 568, 1973
4. Hazzard WL, Goldstein JL, Schrott HG, et al: Hyperlipidemia in coronary heart disease. III. Evaluation of lipoprotein phenotypes of 156 genetically defined survivors of myocardial infarction. J Clin Invest 52:1569-1577, 1973 5. Gotto AM, Gorry GA, Thompson JR, et al: Relationship between plasma lipid concentrations and coronary artery disease in 496 patients. Circulation 56:875-883, 1977 6. Gordon T, Kannel WP, Castelli WP, et al: Lipoproteins, cardiovascular disease and death. Arch Int Med 141:1128-l 131, 1981
EFFECT OF GEMFIBROZIL
ON LIPIDS
393
7. Bailar JC III: Cause and effect in epidemiology. What do we know about hypertriglyceridemia? N Engl J Med 302:1417-l 718, 1980
8. Hulley SB, Rosenman RH, Bawol RD. et al: Epidemiology guide to clinical decisions: the association between triglycerides coronary heart disease. N Engl J Med 302:1383-1389, 1980
as a and
9. Boberg ML. Wallentin H: High density lipoprotein and other lipoproteins in normolipidemic and hypertriglyceridemic (type IV) men with coronary artery disease. Eur J Clin Invest 11:433440, 1981 10. Tall AJ, Small DM: Plasma high density lipoproteins.
N Engl
J Med 299:1232.1978
I 1. Patsch JR, Gotto AM Jr, Olivecrona T, et al: Formation of high density lipoprotein Z-like particles during lipolysis of very low density lipoproteins in vitro. Proc Natl Acad Sci USA 75:4519, 1978 12. Kuusi T. Saarinen P. Nikkila EA: Evidence for the role of hepatic endothelial lipase in the metabolism of plasma high density lipoprotein in man. Atherosclerosis 36:589, 1980 13. Noring L. Kjellin KG, Ledermann H: Effects of bezafibrate on low HDL-cholesterol in patients with ischemic cerebrovascular disease: A pilot study. J Intern Med Res 9:319. 1981 14. Wallace RB, Hunninghake DB, Reiland S, et al: The alterations of plasma high density lipoprotein cholesterol levels associated with consumption of selected medications. The Lipid Research Clinics Program Prevalence Study. Circulation 62:77, 1980 (suppl) 15. Patsch
JR,
Yeshurun
D, Jackson
RL,
et al: Effects
of
clofibrate, nicotinic acid and diet on the properties of the plasma lipoproteins in a subject with type III hyperlipoproteinemia. Am J Med 63:1001, 1977 16. Patsch JR. Sailer S, Kostner G. et al: Separation of the main lipoprotein density classes from human plasma by rate-zonal ultracentrifugation. J Lipid Res 15:356-366, 1974 17. Greten H, Laible V, Zipperle G, et al: Comparison of assay methods for selective measurement of plasma lipase. Atherosclerosis 265633572. 1977 18. Rijssner S: Studies on an Intravenous Fat Tolerance Test. Methodological, experimental and clinical experiences with Intralipid. Acta Med Stand Suppl 564:1-24, 1974 19. Leiss 0, von Bergmann K, Gnasso A, et al: Effect of Gemfibrozil on biliary lipid metabolism in normolipemic subjects. Metabolism 34:74-82, 1985 20. Sachs 1973, p 244 21. Capuzzi
L: Angewandte DM,
Lackmann
Statistik. RD.
Berlin, Alexander
Springer
Verlag,
J, et al: Rapid
modulation of lipogenesis by clofibrate in rat and monkey hepatocytes. Biochim Biophys Acta 409:144155, 1975 22. Berndt J, Gaumert R, Still J: Mode of action of the lipid lowering agents, clofibrate and BM 15075. on cholesterol biosynthesis in rat liver. Atherosclerosis 147-l 52, 1978 23. Heller F, Harvengt C: Effects of Clofibrate, Bezafibrate, Fenofibrate, and Probucol on plasma lipolytic enzymes in normolipidemic subjects. Em J Clin Pharmacol 25:57-63, 1983 24. Kashyap ML: The effect of Gemfibrozil on plasma lipids and apoproteins in man. Vascular Medicine Jan/Mar 16-2 1, 1984 25. Nikkila EA, Huttunen JK, Engholm C: Effect of clofibrate on postheparin plasma triglyceride lipase activities in patients with hypertriglyceridemia. Metabolism 26:179-186, 1977 26. Haberbosch W, Poli A, Baggio G, et al: Apolipoprotein C-II deficiency: The role of apolipoprotein C-11 in the hydrolysis of triacylglycerol-rich lipoprotein. Biochim Biophys Acta 739:49-60, 1984 27. Rissebah AH, Alfarsi S, Adams PW, et al: Transport kinetics of plasma free fatty acid, very low density lipoprotein, triglycerides, and apoprotein in patients with endogenous hypertriglyceridemia. Effects of 2,2-dimethyl, 5(2,5-xylyloxy) valeric acid therapy. Atherosclerosis 24:199-218, 1976 28. Miller GJ, Miller NE: Plasma-high-density-lipoprotein concentration and development of ischemic heart disease. Lancet 1: 1619,1975 29. Gordon TW, Castelli P, Hjortland MC, et al: High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Amer J Med 62:7G7-7 14,1977 30. Castelli WP, Doyler JP, Gordon T, et al: High density lipoprotein cholesterol and other lipids in coronary heart diseaseThe Cooperative Lipoprotein Phenotyping Study. Circulation 55:767, 1977 31. B#rresen AL, Berg K, Dahlin G, et al: The effect of Gemfibrozil on human serum apolipoproteins and on serum reserve cholesterol binding capacity (SRCBC). Artery 9/ 1:77-86, 198 1 32. Witzum JL, Dillingham MA, Giese W, et al: Normalization of triglycerides in type IV hyperlipoproteinemia fails to correct low levels of high-density lipoprotein cholesterol. N Engl J Med 303:907-9 14, 1980 33. Berman M, Hall M III, Levy RI, et al: Metabolism of apo-B and apo-C lipoproteins in man: kinetic studies in normal and hyperlipoproteinemic subjects. J Lipid Res 19:39-56, I978 34. Hugh MW, Fidge NH, Nestel PJ, et al: Metabolism of C-apoproteins: Kinetics of C-II, C-III, and C-Ill2 and VLDLapolipoprotein B in normal and hyperlipoproteinemic subjects. J Lipid Res 22:1235-1246, 1981
and Postheparin Subjects
Lipolytic
Agostino Gnasso, Burkhard Lehner, Werner Haberbosch, Ottmar Leiss, Klaus von Bergmann, and Jan Augustin The lipid lowering agent Gemfibrozil was tested in 8 normolipidemic subjects during a three-month intake. Plasma triglycerides decreased by 41% and Very Low Density Lipoprotein (VLDL) triglycerides decreased by 54%. The reduction of plasma cholesterol, less marked (by 10%). was due to a decrease of Low Density Lipoprotein by 20% while High Density Lipoprotein (HDL) increased up to 30%. The separation of HDL demonstrated that only HDL, were increased. The determination of the apoproteins in plasma and lipoprotein fractions showed similar results with a decrease of apo 6 (by 20%) and an increase of apo A-l and apo A-II. mainly in the HDL, fraction. Plasma postheparin lipolytic activities (PHLA) were not influenced by the therapy and no correlation was found between these activities and any of the plasma or lipoprotein lipids. The apo C-lll/apo C-II ratio in VLDL decreased by 30%; however, no correlation was found between this ratio in plasma as well as VLDL and triglycerides. In addition, the lntra Venous Fat Tolerance Test did not demonstrate any improvement of the clearance of exogenous fat. The lipid lowering efficacy of Gemfibrozil, its collateral effects, and the possible mechanisms of action are discussed. B 1986 by Grune & Stratton, Inc.
H
YPERCHOLESTEROLEMIA is one of the major risk factors for the development of coronary heart disease.‘” The role of triglycerides in the progression of atherosclerotic lesions is less clear.7-9 Several studies have demonstrated that the elevation of triglyceride-rich lipoprotein in plasma is accompanied by a further increase in risk for vascular diseases; others have shown the contrary. During the intravascular hydrolysis of Very Low Density Lipoproteins by the action of lipolytic enzymes, surface components are transferred to discoidal nascent HDL for the formation of HDL.” This mechanism in part explains the inverse correlation between VLDL triglycerides and HDL cholesterol in the plasma compartment. It has been suggested that lipoprotein lipase (LPL) is involved in the interconversion of HDLl to HDLI.” An opposite role was suggested for hepatic triglyceride lipase (HTGL).12 Treatment with clofibrate derivatives and/or analogues in patients with hypertriglyceridemia results in a decrease in VLDL triglycerides and a rise in HDL cholesterol.‘3-15 This has been attributed to an activation of the lipolytic system. However, the stimulation in postheparin lipolytic activities after treatment was small even when therapy was substantial. Thus it is unclear whether a modification in lipase activities, measured in plasma after heparin injection during clofibrate therapy, is the cause or the consequence of decreased VLDL concentrations in plasma. The present study was designed in order to test the effect of the lipid lowering agent Gemfibrozil on plasma lipoproteins, lipolytic activities, and triglyceride clearance capacity in normals. The data demonstrate a profound influence of the drug on the concentration of VLDL and HDL components, with no major change in postheparin lipolytic activities. MATERIALS
AND METHODS
Eight healthy male volunteers in the age range of 34.5 f 3.2 years (mean years *SD). normolipemic with no drug intake, were recruited from the personnel of the clinic. This reduces the likelihood of confounding variables that could affect the outcome of this study. The isocaloric diet that was taken by all subjects in the hospital consisted of 40% fat, 40% carbohydrates, and 20% proteins. The P/S ratio was 0.33. Alcohol intake during the study was less than 20 g/d. Mean ideal body weight was 101.5 + 8.3% (range, 90% to 115%) Metabolism, Vol35,
before the study started and did not change at all during the following months. Six subjects were nonsmokers, two smoked moderately (10 to 20 cigarettes/d) and did not change their habits during the study. No modifications in physical activity habits were reported. The study was open and conducted on an outpatient basis, in accordance with the principles of the Helsinki Declaration. The nature of the study was fully explained and written consent was obtained from each subject. Gemfibrozil, 1,200 mg (600 mg b.i.d.), were administered for three months. Blood samples were collected on Friday morning in order to omit weekend effects before and at monthly intervals during the study and four weeks after the drug was discontinued. Twenty mL of blood for lipid and apoprotein analysis, 10 for zonal ultracentrifugation, and 20 for liver and renal function tests, hematologic parameters, and electrolytes were collected in the morning, after a 12-hour fast, from an antecubital vein, before 100 U of heparin per kg body weight were injected. Five mL blood were drawn 15minutes later to measure postheparin hpolytic activities (PHLA). The Intravenous Fat Tolerance Test (IVFTT), determination of Iithogenic index, lipid composition of cholecystokinin-stimulated fasting duodenal bile, and hepatic secretion of biliary lipids were performed before and after three months of therapy. At these time points, therefore, determinations of plasma lipids were performed, three times in three following days. During the study, instead, blood for lipid determinations was taken two times. The results reported are the mean of the different determinations. EDTA blood was centrifuged for 30 minutes at 2,500 rpm and 4 “C in a Beckman J6B centrifuge. Ten mL of the obtained plasma were filled in Beckman Quick Seal tubes (16 x 76 mm) overlayered with 0.9% NaCl and centrifuged (50,000 rpm) in a Beckman L8-70 preparative ultracentrifuge, using the Beckman 70.1 Ti rotor, for 22 hours at 4 “C. The Very Low Density Lipoproteins (VLDL) were obtained by slicing the tubes. The infranatant underwent identical treatment after adjustment of the density to 1.019 g/mL by addition of KBr. After further 22 hours ultracentrifugation, the IDL were separated in the same way. By additional steps at densities of 1.063,
No 5 (May), 1986: pp 387-393
From the Department of Medicine. University of Heidelberg, and Department of Medicine, University of Bonn. Address reprint requests to Agostino Gnasso. MD. Klinisches Institut fGr Herzinfarktforschung. Bergheimer Str 5X. DtT900 Heidelberg, FRG. 8 1986 by Grune & Stratton, Inc. 00264l49S/86/350%0002$03.00/0
387
388
GNASSO ET AL
1.125, and 1.210 g/mL the Low (LDL) and High Density Lipoprotein subfractions (HDL,/HDL3) were obtained. The densities were controlled by a Heraeus-Paar DMA 45 Densitometer. The height at which the tubes were sliced was obtained empirically by measuring the density and performing lipid electrophoresis and zonal ultracentrifugation of the obtained fractions. The VLDL, IDL, and LDL were perfectly separated from each other. Zonal ultracentrifugation demonstrated a slight overlap between the two HDL fractions obtained by sequential ultracentrifugation. However it was not possible to obtain a better separation of these two fractions varying the height of the cut. The lipoproteins with density cl.063 g/mL contained about 2% of albumin, while the HDLl and HDL, were free of apo B. In the latter fractions, however, albumin represented 4% to 8% of the total protein bulk. In the 1.21 g/mL infranatant, or lipoprotein-free plasma, 4% to 6% of the total triglycerides, 3% to 4% of total cholesterol, and about 5% of the different apoproteins were found. An aliquot of the obtained fractions was used to determine the lipid concentrations while the rest was dialyzed against distilled water (Visking Dialysis Tubing, type 8/32, Serva, Heidelberg, FRG), lyophilized and delipidated by 2 x 5 mL ethanol/ether (3:1, v:v) for 20 and 4 hours and finally 5 mL ether for 1 hour at - 18 OC. Isoelectric focusing (IEF) on ultrathin (0.3 mm) flat gels was used for separation and quantification of the apoproteins. The gels were prepared in our laboratory in the following way: a 7% acrylamide solution with 2% ampholynes (either pH 4-6 or 6-8, Serva, Heidelberg, FRG) was polymerized on plastic sheets, which then functioned as carriers. Anolyte and catholyte were 0.5 M H,P04 and 0.4 M NaOH, respectively. After a prerun of about 1 hour to allow the formation of the pH gradient, 20 pL of the delipidated proteins dissolved in 7 mol/L urea, Tris-HCI pH 8.3, were applied to the plate using rubber wells. Focusing time was 3 hours at 1000 V, amperage free. After this time the protein bands were fixed for 10 minutes with 10% trichloroacetic acid and five minutes with 7% acetic acid. Staining followed for 15 minutes in 1% Light Green (Serva, Heidelberg, FRG) and destaining for 3 x 10 minutes with 7% acetic acid. Normally the background was clear after this procedure. A typical pattern for apo
APO VLDL
APO HDL
IPHI 0
VLDL and apo HDL is shown in Fig 1. The density of the stained bands, determined by an LKB 2202 Ultra Scan Laser Densitometer, was compared with that of standards of known concentrations. The purified apo-proteins, used as standards, were obtained from plasma of normal, for apo A-I and apo A-II, or hypertriglyceridemic subjects, for C-peptides, by gel filtration, affinity chromatography, and preparative isoelectric focusing. The purity of the apoproteins was controlled with excess amounts of proteins by SDS-gel electrophoresis and IEF. In addition no cross-reaction with antibodies against other apoproteins, respectively, was observed. Standard curves were performed with every isolated apoprotein. When the concentration of the sample to be tested fell out of the rectilinear part of the standard curve, appropriate dilutions were made. The C-III isoforms showed identical chromogeneity, however different from apo C-I and apo C-II. The coefficients of variations for apo A-I, A-II, C-I, -C-II, and -C-III were 3.73%, 2.64%, 2.19%, 2.37%, and 1.44%, respectively. Apo A-I and apo A-II, in plasma and in the HDL fractions, were determined also by immunodiffusion (Immuno AG, Vienna, Austria). The monospecificity of the apo A-I and apo A-II antibodies was detected with pure A-I and A-II apolipoproteins. No crossreaction was observed. The coefficients of variations were 5.35% for apo A-I and 6.10% for apo A-II. Quantification of apo B was performed by immunodiffusion on NOR-partigen plates (Behringwerke AG, Marburg, FRG) (cv 4.72%). Cholesterol, triglycerides, and phospholipids were measured using commercially available kits (Boehringer, Mannheim, FRG). HDL cholesterol was determined also after precipitation of the apo-B-containing lipoproteins by addition of phosphotungstic acid and magnesium ions (Boehringer, Mannheim, FRG). Zonal ultracentrifugation was carried out according to Patsch et al.16 HDL subclasses were separated in nonlinear sodium bromide gradients ranging in density from 1.00 to 1.40 g/mL, using a Beckman Ti-14 rotor, at 44,000 rpm and 4 ‘C for 22 hours. The activities of lipoprotein lipase (LPL) and hepatic triglyceride lipase (HTGL) were measured by immunoassay with goat antibodies against HTGL as described earlier.” The IVFTT was performed by injecting 1 mL Intralipid pro kg body weight within 1 minute (Intralipid 10% Vitrum, Deutsche Kabi Vitrum, Munich, FRG). Blood samples were collected before and every five minutes up to 40 minutes after injection, centrifuged, and turbidity was measured in plasma, diluted 1:11 with 0.9% NaCl at 691 nm.‘* The elimination rate of the fat was calculated by the following formula: I,F, = at + c
1 Al
0
where FL represents the difference between the extinctions of the samples obtained after and before Intralipid injection. Triglyceride concentration was determined from the same samples. The results of the lithogenic index, lipid composition of gall bladder, and hepatic bile before and at the end of drug intake are published in this journal.t9 The results were analyzed by Friedman Test and the significance of the differences between the time points was evaluated by Wilcoxon test. Relationships among plasma lipid, apoprotein, and lipase values were evaluated by Spearman’s rank correlation coefficients?’
PH 6 RESULTS Typical IEF pattern of apo VLDL and apo HDL from four Fig 1. different subjects. The samples were applied at the cathod. Run conditions and preparation of the protein samples are given in the
text.
The plasma lipid concentrations are given in Table 1. The triglycerides fell progressively during the drug intake up to 44% after three months, while a less marked decrease of
EFFECT OF GEMFIBROZIL ON LIPIDS
Table 1. Lipid and Apoprotein Were Obtained
389
Parameters
in Plasma From Normals Before, During, and After Gemfibrozil
Before and After 4,8,
Week
and 12 Weeks
of Gemfibrozil
Intake and 4 Weeks
4
0
Intake (mg ? SD) (Values
After Omission of the Drug)*
6
12
16
TG
111 f 48
82 + 31
64 + 23t
62 ? 21t
94 + 33
CHOL
177 f 42
157 f 24t
157 + 23t
156 -f- 28T
174 f 27
195 + 30
203 & 28
PHOS
210
Apo B
86.3
+ 16.6
68.6
700.1
* 16.8
111.4
Apo A-l lmmunodiffusion
+ 43
188 f 25t
187 ? 30
* 12.lT
67.1
+ 15.4
Apo A-II lmmunodiffusion
33.6
t 8.4
36.3
t 7.1
Apo A-l IEF
85.3
+ 17.5
99.7
+ 14.8
Apo A-II IEF
18.6 + 6.9
21.7
* 6.3
115.4 38.6
f
13.4t
+ 18.3 ? 4.9
69.7
+ 14.lt
122.3
f
41.3
104.5
f 26.8t
118.5
25.6
+ 7.7T
27.7
15.07
84? 106.4
11.9 t 9.7
r 4.2t
35.4
* 3.1
r 17.6t
91.1
+ 10.1
+ 6.lt
20.1
? 3.6
‘Whereas the data determined for apoproteins A-l and A-II in plasma after immunodiffusion represent total, free, and lipoprotein bound apoproteins, the results obtained by IEF are the sum of the values in the lipoprotein fractions. TP < 0.05.
cholesterol and phospholipids was observed (12% and 7%, respectively). Four weeks after the suspension of the drug only triglycerides were still lower than before treatment (15%). Plasma apo B, apo A-I, and apo A-II concentrations are shown in Table 1. For A apoproteins, both the results obtained by immunodiffusion and IEF are reported and they show similar trends, although the absolute values slightly differ. This is true mainly for apo A-II, which is found in part in the 1.21 g/mL infranatant after sequential ultracentrifugation. Apo B decreased during the drug intake by about 20% while the A apoproteins increased significantly especially after three months of therapy. A strong reduction in all VLDL components was observed without change in the ratio of the sum of measured apoproteins to lipids (Table 2). Triglycerides, cholesterol, and phospholipids fell, respectively, by 59%, 51%, and 48%. Apoproteins decreased in the same manner. Apo B was reduced by 50% (from 6.9 to 3.4 mg%) and also the C peptides progressively diminished. However, the ratio of apo C-III and apo C-II changed, the VLDL particles becoming relatively enriched in apo C-II. The apo C-III/ape C-II ratio fell, in fact, by 3 1% after 12 weeks of drug intake compared to the starting values. The decrease of VLDL lipids and apoproteins was almost linear during the three months of therapy with no plateau. IDL were present, as expected, in very small amounts and showed changes similar to VLDL, but highly increased after Gemfibrozil was discontinued. At this time, the IDL were relatively enriched in triglycerides. However, these differences were not statistically significant related to large standard deviations. The protein component of this lipoprotein Table 2. -
-.
Lipid and Apoprotein
Parameters
0
Week
c 39.9
in VLDL Before, During, and After Gemtibrozil 4
lti
70.4
CHOL
16.1 + 7.5
10.1 f 7.1
PHOS
26.9
20.1
+ 16
class was too small to allow quantification of single apoproteins (Table 3). The LDL fraction (Table 4) decreased by about 20% during the drug intake. The triglycerides were markedly reduced (by 28%) after four and eight weeks of Gemfibrozil, slightly increasing (10%) after 12 weeks. The cholesterol and phospholipids were reduced by 22% and 16%, respectively, after four weeks, remaining then constant. The apo B was reduced constantly by about 20%, the protein/lipid ratio of this lipoprotein fraction did not change. In contrast to VLDL the highest reduction of this lipoprotein class was observed already after four weeks of treatment. At the end of the study all these components reached the values observed before drug intake. HDL were analyzed after separation in HDLz and HDL, by sequential flotation, like the other lipoprotein classes. Zonal ultracentrifugation was performed to obtain a qualitative pattern of the HDL subclasses. In addition, HDL cholesterol was determined after precipitation of apo-Bcontaining lipoproteins. The precipitation method gave consistently higher levels of cholesterol than those obtained after ultracentrifugation, but the changes during the study were similar (Fig 2). HDL, did not show major differences in any of the lipid and apoprotein components during treatment (Table 5). Only in the zonal ultracentrifugation was a shift of this subclass to a slightly higher density observed (Fig 3). In the HDL, range, triglycerides remained constant, whereas cholesterol and phospholipids increased by 30% and 15%, respectively, after volunteers had been treated for three months with Gemfibrozil. HDL,-apo A-I increased by 44% and apo A-II by 36% (Table 6). The rise in C peptides was
47.9
c 25.3 * 10.1
34.9
+ 19.21
6.5 + 3?6+
Intake Lmg% i SD) 12
6 29.1
-c 14.7s
16 49.7
’
i- 25.
7.9 * 4.7’
8.3 + 8 17.9 ” 9.7
13.6 f 6.6,
13.9 + 6.8’
Apo B
6.9 + 2.8
4.9 * 1.9
3.9 * 1.5*
3.4 It 1.2*
5.1 + 1.9
Apo C- 1
0.6 2 0.3
0.4 + 0.2’
0.3 f 0.2”
0.2 _t 0.21
0.3 ? 0.
Apo C-II
0.5 * 0.24
0.39
I 0.27
0.32
t 0.22’
0.28
& 0.19*
0.36
I
+ 0.26
Apo C-III
1.8 _+0.8
1.2 _t 1.0
0.9 t 0.7’
0.6 ‘- 0.6’
1.3 + 1.0
Apo C-III/C-II
3.7 + 0.4
2.9 * 0.5’
2.7 k 0.5*
2.6 + 0.7*
3.4 + 0.3
lP < 0.05.
390
GNASSO ET AL
Table 3.
Lipid Parameters
1
IDL
in IDL Before. During, and After Gamfibroril 2
TG
4.38
f 2.20
3.00
CHOL
2.88
k 1.46
1.00 i
1.07
DISCUSSION
In the present study Gemfibrozil was given to eight normolipidemic subjects and several parameters were folIowed up to four months to verify its lipid lowering efficacy and possible collateral effects. The drug efficiently lowered plasma triglycerides and, to a lesser extent, plasma cholesterol and phospholipids. The fall of triglycerides was mainly due to the VLDL fraction, which was relatively enriched in apo C-II. Plasma cholesterol was less reduced than LDL cholesterol, because the decrease of the LDL fraction was in part balanced by the rise of HDL cholesterol. But only the HDL, fraction was influenced by the therapy. Similar effects of clofibrate derivatives and analogues on plasma lipoprotein metabolism have been described. Both
Table 4.
Lipid and Apoprotain
TG
5
3.25
+ 1.83
3.38
+ 1.51
8.00
f 4.69
2.25
+ 1.49
1.50 + 0.76
3.57
f 2.23
influences on VLDL synthesis and catabolism have been taken into consideration?‘-” An increase of postheparin lipolytic activities, as observed in other trials,24*25was not confirmed in the present study, in fact even a slight decrease was detected. However, in the above mentioned trials, patients with hypertriglyceridemia were studied. In these patients, the slight PHLA increase following the drug intake might be the cause as well as the consequence of the normalization of plasma triglycerides. The present data, obtained from normolipemic subjects, for the first time demonstrate that the drug has probably no influence at all on PHLA. These results therefore strongly support the latter hypothesis. In addition, no correlation between lipoprotein lipase or hepatic triglyceride lipase and the major plasma lipids was found. On the other hand, the modified VLDL apoprotein composition, with a percentage increase of the LPL activator apo C-II and a decrease of the PHLA inhibitor apo C-III, is unlikely to explain an improved hydrolysis of these particles. No correlation could be found between the apo C-III/ape C-II ratio and triglycerides either in plasma or VLDL. Apo C-II is present in excess in VLDL.26 In addition, in this study, the strong reduction of the VLDL fraction was not followed by an increase of the IDL or LDL fraction, as expected after higher hydrolysis rate of VLDL. On the contrary, even these fractions showed a reduction. The IDL, already present in very small amounts, further decreased under Gemfibrozil and LDL components were reduced up to 20% (Tables 2,3). Moreover, the fat tolerance test, with similar extinction rates both before and after the medication, and even a decreased elimination rate for triglycerides, demonstrates that the capacity for removal of exogenous fat from the plasma compartment is, if at all, impaired after Gemfibrozil treatment. The decrease of the LDL fraction, maximally evident after four weeks of Gemfibrozil, despite the continuing decline of VLDL, could be the consequence of an adapted reduction of the LDL clearance in order to adjust a new steady state. This is further supported by a strong increase in IDL and LDL after withdrawal of the drug. The modifications observed in the HDL range give further evidence for this mechanism. Recently, as a consequence of the hydrolysis of triglyceride-rich lipoproteins, with transfer of surface material to preexisting HDLJ, an interconversion of these particles to
in LDL Before, During, and After Gemfibrozil
4
Intake (mg% + SD)
12
8
12.8 + 2.4’
12.6 i 3.9’
14.5 * 3.1
+ 29.9
84.8
+ 20.1*
87.5
+ 21.3*
89.1
‘- 22.3
67.4
i
15.5
56.4
+ 13.5’
57.3
r 15.71
58.4
t
77.9
t 14.5
61.5
f 11.1’
61.0
i
64.6
+ 13.7
17.6 -+ 5.7
CHOL
108.5
PHOS Apo 8 lP < 0.05.
Parameters
0
4
3
* 2.00
less pronounced and significant only for apoprotein C-II., Apoprotein C-III remained constant. An increased absorbance and a shift to a higher density was observed for HDL3 after separation by zonal ultracentrifugation (Fig 3). All these modifications were reversible four weeks after the end of the drug period. The postheparin lipolytic activities in plasma, LPL and HTGL, were in the normal range (Fig 4). After four weeks of treatment a significant decrease in LPL activity was detected, whereas no significant change for both activities was observed in the other periods. The IVFTT performed immediately before and after medication revealed no significant difference in the elimination rate of exogenous fat (Fig 5), expressed as change in plasma turbidity. However, if plasma triglycerides were measured at the same time intervals it was found that after the drug intake the bolus injection of Intralipid resulted in higher values than before the medication (Fig 6). The other blood parameters checked did not show any alteration in electrolytes, glucose, protein, and renal tests. A slight reduction of the white blood cells was observed in six subjects after four weeks on Gemfibrozil. The cell number remained then constant and returned back to previous values four weeks after suspension of the drug. In addition, slight elevation of transaminases in two subjects at one of the time points was found. Some patients complained of “transitory abdominal tension” during the study but none had to discontinue the drug for this reason.
Week
Intake (mg% + SD)
12.5*
15.7’
16 17.6 * 2 104 -c 24 68.6
f 13.8
76.5
+ 11.2
391
EFFECT OF GEMFIBROZIL ON LIPIDS
T
60
T
T
il 1 8 ;*
I
16 weeks
HDL chol after precipitation of apo-B-containing lipoFig 2. proteins (-1 and after separation of HDL by sequential ultracentrifugation j----L ‘P < 0.05.
lighter HDLz has been described.” In this study no major alterations of the HDLz fraction took place whereas HDL, showed remarkable modifications under drug exposure. The HDL, composition remained constant throughout the study and the increase was due to both apoproteins and lipids (Table 6). Therefore this rise cannot be explained by a transfer of material from triglyceride-rich lipoproteins. Most
4
0 Postheparin Fig 4. LPL (----I. ‘P < 0.06.
'2 weeks
8
lipolytic activities in plasma: HTGL I-_) and
bE 0.3 ----
week
0
----week
12
aba
0.2
33
f 0.1
I
I
5
A typical zonal ultracentrffugation Fig 3. after Gemfibrozil intake. Table 5. Lipid and Apoprotein W.Sk
0
pattern
before
Parameters
and
I
IO
I
15
I
I
20
25
1
35
I
LO
8
Intake fmg% i SD) 12
16
4.4 ‘r_0.7
4.0 +_ 0.8
3.3 * 2.1
3.5 + 1.2
3.1 + 0.9
CHOL
12.8 + 7.0
14.8 f 4.0
13.8 + 6.8
14.6 & 7.0
13.7 f 6.8
PHOS
17.5 t
16.9 + 9.8
TG
17.0 + 9.3
18.1 i 6.8
17.9 i
Apo C- 1
1.5 + 0.9
1.5 + 0.4
1.9 f
1.7
1.2 + 1.0
0.6 + 0.3
Apo C-II
0.6 + 0.4
0.8 f 0.5
0.7 + 0.6
0.5 2 0.5
0.5 f 0.3
Apo C-III
2.0 t 2.1
2.0 f
l
min
Fig 5. lntra Venous Fat Tolerance Test: E represents the difference between extinctions of plasma taken after and before lntralipid injection. (-1 Before and I----) after three months treatment with Gemfibrozil.
in HDL, Before, During, and After Gemfibrozil
4
I
30
11.9
12.0
1.3
1.9 * 1.7
1.4 +- 1.1
1.4 + 0.7
Apo A-l
14.8 f 8.0
17.2 + 5.1
16.2 f 9.6
17.1 * 9.3
15.1 f 7.4
Apo A-II
2.8 * 2.2
3.0 f 2.1
3.4 f 2.6
2.9 * 2.6
2.6 f 2.0
GNASSO ET AL
392
Table 6. Lipid and Apoprotein Parameters in HDL, Before, During, and After Gemfibroril Intake (mg% + SD) Week
0
a
4
6.8 f 1.8
TG
6.3 k 1.0
12
5.5 * 1.9
16
7.1 + 1.1
7.3 f 0.8
CHOL
28.9 +. 6.8
32.8
k 5.2
32.6
+ 7.1
37.6
+ 3.8”
28.9
e 1.6
PHOS
67.8
66.6
+ 12.3
69.3
+ 13.1
78.3
f
55.7
f 26.1
Apo C-l
+ 12.7
2.7 f 0.7
3.2 f
3.3 f
1.0
1.0
11.7*
3.4 + 1.1
2.1 f 0.2
Apo C-II
1.3 f 0.3
1.7 k 0.3
1.4 f 0.3
1.5 + 0.4
1.3 + 0.1
Apo C-III
3.4 + 1.2
4.3 f
4.2 + 1.4*
4.1 f
3.4 It 0.7
Apo A-l
70.2
Apo A-II
15.8 + 5.1
f 13.8
82.6
1.2
s 12.4
18.7 + 4.6
87.8
+ 20.8*
22.2
f 6.3.
101.4 24.8
1.5,
+ 13.2’
75.8
f 5.0’
17.5 + 2.4
f 8.5
lP -c 0.05.
likely, as a result of Gemfibrozil treatment, the number of HDL, particles in plasma increases. The reason for this effect may be due to increased synthesis of apo A-I and apo A-II demonstrated for patients with hypertriglyceridemia.” An inverse relation between plasma triglycerides and HDL cholesterol in normals and patients with hyperlipoproteinemias has been previously reported.2*-30 In our study there was a slight inverse correlation (r = - Sl) between plasma triglycerides and HDLz cholesterol which however disappeared during and after treatment. In earlier studies mainly an increase in HDL2 cholesterol after Gemfibrozil intake was observed.““’ On the other hand, dietary treatment of patients with type IV hyperlipoproteinemia with normalization of plasma triglycerides did not result in any change of low HDL cholesterol concentrations.32 It was concluded that
I 1
5
I
10
0
-
week
----
week12
I
I
I
I
I
15
20
25
30
LO
+
min
lntra Venous Fat Tolerance Test: The values represent Fig 6. the differences between TO concentrations of plasma taken after and before lntralipid iniection. (-_) Before and (----1 after three months treatment with Gemfibrozil.
in this particular group of patients no interrelationship of the metabolism of VLDL and HDL occurs. These data also suggest a direct influence of the drug on HDL synthesis. Besides ample differences in the methodology it has to be pointed out that the aforementioned studies have been conducted with a heterogenous group of hyperlipidemic subjects. Our data indicate that at least in normals Gemfibrozil acts exclusively on the HDL3 subclass. The C apoproteins have been regarded as an entity that is more or less equally distributed between VLDL and HDL.33T34 Although ultracentrifugal effects cannot be excluded, in this study neither before nor during Gemfibrozil exposure was this pattern observed. In fact, 80% of the C peptides were bound to HDL without treatment and even 90% with treatment. The ratio of the different C apoproteins was constant throughout the study in HDL2 and HDL3. In VLDL, however, the percentage modification was observed exclusively between apo C-I and apo C-II; the contribution of apo C-III remained unchanged. These data suggest that in normals HDL are the major reservoir of C apoproteins in plasma. Moreover Gemfibrozil intake gives evidence for an independent metabolism of the different peptides. In conclusion, Gemfibrozil strongly affects plasma lipoproteins in normals resulting mainly in a decrease in VLDL and a rise in HDL3. Whereas the influence on VLDL is probably mediated by an inhibition of the synthesis of these particles, further investigations are needed to clarify the role of this drug on HDL metabolism. The effect of Gemfibrozil on the parameters reported in this study in patients with different types of hyperlipoproteinemias is currently under investigation. ACKNOWLEDGMENT
The authors thank 1. Geldmacher, E. Glatting, M. Miltner, D. Schraube for their excellent technical assistance.
and
REFERENCES
1. Cohn PF, Gabbay SI, Weglicky WB: Serum lipid levels in angiographically defined coronary artery disease. Ann Int Med 84:241-245, 1976. 2. Goldstein JL, Hazzard WR, Schrott HG, et al: Hyperlipidemia in coronary heart disease. I. Lipid levels in 500 survivors of myocardial infarction. J Clin Invest 52:1533-l 544, 1973 3. Goldstein JL, Schrott HG, Hazzard WR, et al: Hyperlipidemia in coronary heart disease. II. Genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder, combined hyperlipidemia. J Clin Invest 52: 1544-l 568, 1973
4. Hazzard WL, Goldstein JL, Schrott HG, et al: Hyperlipidemia in coronary heart disease. III. Evaluation of lipoprotein phenotypes of 156 genetically defined survivors of myocardial infarction. J Clin Invest 52:1569-1577, 1973 5. Gotto AM, Gorry GA, Thompson JR, et al: Relationship between plasma lipid concentrations and coronary artery disease in 496 patients. Circulation 56:875-883, 1977 6. Gordon T, Kannel WP, Castelli WP, et al: Lipoproteins, cardiovascular disease and death. Arch Int Med 141:1128-l 131, 1981
EFFECT OF GEMFIBROZIL
ON LIPIDS
393
7. Bailar JC III: Cause and effect in epidemiology. What do we know about hypertriglyceridemia? N Engl J Med 302:1417-l 718, 1980
8. Hulley SB, Rosenman RH, Bawol RD. et al: Epidemiology guide to clinical decisions: the association between triglycerides coronary heart disease. N Engl J Med 302:1383-1389, 1980
as a and
9. Boberg ML. Wallentin H: High density lipoprotein and other lipoproteins in normolipidemic and hypertriglyceridemic (type IV) men with coronary artery disease. Eur J Clin Invest 11:433440, 1981 10. Tall AJ, Small DM: Plasma high density lipoproteins.
N Engl
J Med 299:1232.1978
I 1. Patsch JR, Gotto AM Jr, Olivecrona T, et al: Formation of high density lipoprotein Z-like particles during lipolysis of very low density lipoproteins in vitro. Proc Natl Acad Sci USA 75:4519, 1978 12. Kuusi T. Saarinen P. Nikkila EA: Evidence for the role of hepatic endothelial lipase in the metabolism of plasma high density lipoprotein in man. Atherosclerosis 36:589, 1980 13. Noring L. Kjellin KG, Ledermann H: Effects of bezafibrate on low HDL-cholesterol in patients with ischemic cerebrovascular disease: A pilot study. J Intern Med Res 9:319. 1981 14. Wallace RB, Hunninghake DB, Reiland S, et al: The alterations of plasma high density lipoprotein cholesterol levels associated with consumption of selected medications. The Lipid Research Clinics Program Prevalence Study. Circulation 62:77, 1980 (suppl) 15. Patsch
JR,
Yeshurun
D, Jackson
RL,
et al: Effects
of
clofibrate, nicotinic acid and diet on the properties of the plasma lipoproteins in a subject with type III hyperlipoproteinemia. Am J Med 63:1001, 1977 16. Patsch JR. Sailer S, Kostner G. et al: Separation of the main lipoprotein density classes from human plasma by rate-zonal ultracentrifugation. J Lipid Res 15:356-366, 1974 17. Greten H, Laible V, Zipperle G, et al: Comparison of assay methods for selective measurement of plasma lipase. Atherosclerosis 265633572. 1977 18. Rijssner S: Studies on an Intravenous Fat Tolerance Test. Methodological, experimental and clinical experiences with Intralipid. Acta Med Stand Suppl 564:1-24, 1974 19. Leiss 0, von Bergmann K, Gnasso A, et al: Effect of Gemfibrozil on biliary lipid metabolism in normolipemic subjects. Metabolism 34:74-82, 1985 20. Sachs 1973, p 244 21. Capuzzi
L: Angewandte DM,
Lackmann
Statistik. RD.
Berlin, Alexander
Springer
Verlag,
J, et al: Rapid
modulation of lipogenesis by clofibrate in rat and monkey hepatocytes. Biochim Biophys Acta 409:144155, 1975 22. Berndt J, Gaumert R, Still J: Mode of action of the lipid lowering agents, clofibrate and BM 15075. on cholesterol biosynthesis in rat liver. Atherosclerosis 147-l 52, 1978 23. Heller F, Harvengt C: Effects of Clofibrate, Bezafibrate, Fenofibrate, and Probucol on plasma lipolytic enzymes in normolipidemic subjects. Em J Clin Pharmacol 25:57-63, 1983 24. Kashyap ML: The effect of Gemfibrozil on plasma lipids and apoproteins in man. Vascular Medicine Jan/Mar 16-2 1, 1984 25. Nikkila EA, Huttunen JK, Engholm C: Effect of clofibrate on postheparin plasma triglyceride lipase activities in patients with hypertriglyceridemia. Metabolism 26:179-186, 1977 26. Haberbosch W, Poli A, Baggio G, et al: Apolipoprotein C-II deficiency: The role of apolipoprotein C-11 in the hydrolysis of triacylglycerol-rich lipoprotein. Biochim Biophys Acta 739:49-60, 1984 27. Rissebah AH, Alfarsi S, Adams PW, et al: Transport kinetics of plasma free fatty acid, very low density lipoprotein, triglycerides, and apoprotein in patients with endogenous hypertriglyceridemia. Effects of 2,2-dimethyl, 5(2,5-xylyloxy) valeric acid therapy. Atherosclerosis 24:199-218, 1976 28. Miller GJ, Miller NE: Plasma-high-density-lipoprotein concentration and development of ischemic heart disease. Lancet 1: 1619,1975 29. Gordon TW, Castelli P, Hjortland MC, et al: High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Amer J Med 62:7G7-7 14,1977 30. Castelli WP, Doyler JP, Gordon T, et al: High density lipoprotein cholesterol and other lipids in coronary heart diseaseThe Cooperative Lipoprotein Phenotyping Study. Circulation 55:767, 1977 31. B#rresen AL, Berg K, Dahlin G, et al: The effect of Gemfibrozil on human serum apolipoproteins and on serum reserve cholesterol binding capacity (SRCBC). Artery 9/ 1:77-86, 198 1 32. Witzum JL, Dillingham MA, Giese W, et al: Normalization of triglycerides in type IV hyperlipoproteinemia fails to correct low levels of high-density lipoprotein cholesterol. N Engl J Med 303:907-9 14, 1980 33. Berman M, Hall M III, Levy RI, et al: Metabolism of apo-B and apo-C lipoproteins in man: kinetic studies in normal and hyperlipoproteinemic subjects. J Lipid Res 19:39-56, I978 34. Hugh MW, Fidge NH, Nestel PJ, et al: Metabolism of C-apoproteins: Kinetics of C-II, C-III, and C-Ill2 and VLDLapolipoprotein B in normal and hyperlipoproteinemic subjects. J Lipid Res 22:1235-1246, 1981