Effect of Acute Uremia on Triglyceride Kinetics in the Rat R. Gregg,
C. E. Mondon,
E. P. Reaven, and G. M. Reaven
Plasma triglyceride (TG) levels were elevated 24 hr after the production of acute uremia in rats. The effect of acute uremia on TG production rate was estimated by determining the rate of TG accumulation following Triton WR 1339 inhibition of lipoprotein removal, by measuring hepatic TG secretion rate during in situ liver perfusion, and by quantifying hepatocyte very low density lipoprotein content with the electron microscope. The results of all three of these approaches indicated that TG synthesis and secretion were decreased
in acute uremia, suggesting that the associated increase in plasma TG levels had to result from a removal defect. This hypothesis was tested directly by injecting prelabeled very low density lipoprotein TG into acutely uremic and control rats, and measuring its rate of disappearance from plasma. The ti of removal in acutely uremic rats was found to be approximately twice that of control, confirming the hypothesis that the rise in plasma TG levels in acute uremia is due to a defect in removal of TG from plasma.
A
TTENTION was drawn to the presence of hypertriglyceridemia in nonnephrotic patients with chronic renal failure in 1968,1m3and the increased incidence of hypertriglyceridemia in such patients has been repeatedly confirmed.4m’6 An elevated plasma triglyceride (TG) level can be due theoretically to an increase in rate of TG entry into plasma and/or a decrease in rate of TG removal from plasma: evidence has been presented in support of both of these mechanisms. Thus, the presence of elevated levels of plasma insulin in this situation has led to the suggestion that the hypertriglyceridemia is secondary to insulin-induced increases in hepatic TG production.3.‘7m’9 On the other hand, measurements of plasma postheparin lipolytic activity (PHLA) and very low density lipoprotein (VLDL) TG kinetics have suggested that the hypertriglyceridemia is secondary to a decrease in VLDL-TG removal rate from plasma. 3,4~8~‘3,‘7.‘xzo Either or both of these abnormalities may exist, but at present the majority of the evidence available is indirect. In an effort to obtain more direct evidence concerning the cause of hypertriglyceridemia in uremia, we have instituted a series of studies in which several experimental approaches have been used to define TG kinetics in uremic rats. In order to study the simplest model, we have begun by investigating the changes that occur in acutely uremic rats, a situation which has received relatively little attention. I9 In the current study, TG production rate was estimated by determining the rate of TG accumulation following Triton WR 1339 block-
From the Depariment of Medicine. Stanford University School of Medicine and Veterans Administration Hospital, Palo Alto, Calif Receivedfor publication March 17. 1976. Supported in part by National Institutes of Health Contract No-I-AM4-2213 from the Artihcial Kidnqv and Chronic Uremia Centers Branch, and by the Veterans Administration. Dr. Reaven is a Veterans Administration Medical Investigator, MRIS #7363. Reprint requests should be addressed to Gerald M. Reaven. M.D., V.A. Hospital (III-M). 3801 Miranda Avenue, Palo Alto, Calif: 94304. 0 I976 b.v Grune & Stratton. Inc. Metabolism, Vol. 25, No. 12 (December), 1976
1557
1558
GREGG
ET AL.
ade of TG removal, by measuring hepatic TG secretion rate during in situ liver perfusion, and by quantifying hepatocyte VLDL-TG content with electron microscopy. TG removal from plasma was directly measured by determining the disappearance rate of injected prelabeled VLDL-TG. The results of all four experimental approaches were consistent, and strongly suggest that VLDL-TG synthesis and secretion are decreased in acutely uremic rats, and hypertriglyceridemia in this situation is secondary to a decrease in TG removal from plasma. MATERIALS
AND METHODS
Protocol Female Sprague-Dawley rats, weighing 130~~150 g, were placed on a fat-free diet, consisting of 30”~ of calories as casein and 70°, as sucrose, 4- 7 days before surgical production of acute uremia. Rats were maintained on a 12-hr light (6 a.m. ~6 p.m.)/dark cycle. Food was removed at 8 a.m. the day of surgery, and rats were operated on 6- 8 hr later. Animals were lightly anesthetized with ether, and were either sham operated or made acutely uremic by bilateral ligation of the renal artery and vein via a midline incision. They were allowed access to water overnight, but received no food. All experiments were performed 24 hr after surgery.
Measurements Baseline detertninarions. Sham-operated and acutely uremic rats were anesthetized with sodium thiamylal, 60 mg/kg, and bled by aortic puncture into a syringe wetted with a 5”,, EDTA solution. The plasma was separated by centrifugation. frozen. and later assayed for TG.” BUN (BUN Analyzer. Beckman Instruments), glucose (Glucose Analyzer. Beckman Instruments), insulin,** and FFAZ3 levels. TG secretion rate (TGSR/. The TGSR was determined by the use of Triton WR 1339, which blocks the removal of intravascular lipoproteins.24V25 Thus, by measuring TG accumulation with time, TGSR can be determined. Baseline plasma TG concentrations (TGa) were measured in nonanesthetized rats on blood collection from the tip of the tail into capillary tubes rinsed with a 5”,, EDTA solution. Triton was then injected into the tail vein at a concentration of 600 mg/kg, an amount determined previously to produce maximal TG accumulation. Two hours later rats were anesthetized with sodium thiamylal and bled from the aorta into EDTA-wetted syringes. and plasma was again collected for determination of TG concentrations (TG2). TGSR was calculated from the following formula: TGSR
(mg/lOO g body weight/hr)
=
(TG2 - TGo)
x plasma
volume
2 hr where TG is expressed as mg/ml and plasma volume as ml/100 g body weight. Plasma volume. Plasma volume was determined on seven acutely uremic and seven shamoperated rats with the Evans blue dye dilution technique. Triton was given as described above, and Evans blue dye was injected into a cannulated right lumbar vein 2 hr after the Triton injection. Five minutes later a blood sample was drawn from the aorta into an EDTA-wetted syringe. The concentration of Evans blue was determined, and the plasma volume was calculated from the standard formula. There was no significant difference between sham-operated and uremic rats. Hepatic TG production. Hepatic TG production by livers isolated from acutely uremic and sham-operated rats was determined by in situ cyclic perfusion techniques previously described.26 The perfusing medium consisted of 904< defibrinated rat blood, obtained from retired male bicarbonate (KRB) breeding stock maintained on laboratory chow, and IO”, Krebs-Ringer buffer containing 3 g bovine albumin/l00 ml (hematocrit of 40). The initial circulating volume averaged 42 ml. All perfusions received a continuous infusion of porcine insulin at a rate of 33 mu/O.435 ml of KRB buffer per hour and oleic acid complexed with rat serum solution. The rat serum solution for infusion of oleic acid was prepared by circulating a 904” rat
1559
ACUTE UREMIA
blood perfusing medium described above through an isolated normal rat liver for 75 min to remove vasoconstrictor factors and freezing 15-ml aliquots of the recovered serum. For a perwith NaOH fusion experiment 800 Fmoles of oleic acid was dissolved in 959; ethanol, neutralized and KOH to a previously determined phenolthalein end point, and evaporated to near dryness in a boiling water bath. The residue was dissolved in 5 ml KRB buffer and mixed with 15 ml of rat serum solution thawed to room temperature. The mixture was then filtered through glass wool to remove macroaggregates, and provided sufficient volume of oleic acid-serum complex for three liver perfusions. The concentration of oleic acid averaged 28 rmole/ml and was infused at a rate of 2.31 ml/hr. Net TG production was calculated from net changes in perfusate concentration after correcting for addition of the infusate and sampling losses of 1.8 ml every 30 min. Electron microscopic evaluation of hepatocyte VLDL production. Tissue taken from the left lobe of the liver of six sham and six rats with acute uremia was fixed, processed, and subjected to ultrastructural morphometric analysis as outlined in detail in a previous report from this laboratory.” In brief, an estimate of the total number of Golgi-associated VLDL particles was obtained from six hepatocytes from each animal and an estimate of VLDL size was obtained from measurements of the diameters of all the VLDL particles in photographic enlargements of two randomly selected Golgi complexes from six hepatocytes of each animal. In sham-operated animals this means that roughly 1400 particles were counted and roughly 500 particles were measured per animal. The use of Golgi&VLDL content as one of several indicators of hepatocyte VLDL production appeared justified in view of the fact that Go1 i $ complexes have been described as temporary storage and packaging depots for export VLDL.**’ 9 VLDL-TG removal rate. The effect of acute uremia on VLDL-TG removal rate from plasma was studied by injecting VLDL-TG that had been prelabeled in vivo with 3H-glycerol. and measuring its rate of removal. Female donor rats, weighing 200-250 g were maintained on a fat-free diet for 4-7 days. On the day of the experiment their food was removed at 8-9 a.m.. and 4 hr later they were injected via the tail vein under light ether anesthesia with 400 kCi of 3H-2-glycerol (I.C.N., 2 Ci/mmole). The donor rats were exsanguinated under sodium thiamylal anesthesia 20 min later. (Previous experiments had shown maximal incorporation of 3H-glycerol into TG 20 min after injection, at which time 95”” of the lipid extractable radioactivity was incorporated into VLDL-TG as determined by ultracentrifugation and thin-layer chromatography.) Serum was separated by centrifugation, and 0.8-m] aliquots were injected without anesthesia into the tail veins of sham-operated or acutely uremic rats. In control experiments the plasma obtained from the donor rats was used to isolate VLDL by methods previously described from 30 our laboratory, and aliquots of isolated VLDL. rather than unextracted plasma, were then injected into the experimental rats. After administration of prelabeled VLDL-TG. the tail was amputated proximal to the site of injection and 0.4 ml of blood was collected into capillary tubes rinsed with a 5”{ EDTA solution at approximately 5, 15, 25, and 35 min after the injection. (These samples took I-2 min to collect and in all subsequent calculations the time used for a collection period was the mean of the beginning and ending times of the blood collection.) The plasma was separated by centrifugation and stored frozen until analyzed. A lipid extract” was made of each sample, evaporated to dryness, and the radioactivity measured by liquid scintillation counting (Beckman LS-235) using a standard toluene scintillation mixture. The natural log of the cpm of each sample was determined and plotted against time. A best fit line was determined by least squares regression analysis and the slope was determined. The following calculations were made: slope = FTR (fractional
turnover
rate)
ti = In Z/FTR. The mean t$ of eight control rats as determined by following the disappearance of prelabeled plasma was essentially identical to that determined following the injection of isolated VLDL-TG into eight other control rats. Therefore, in order to avoid the possible alteration of the VLDL that might occur as the result of the isolation procedure, we carried out our studies of VLDL-TG removal rate with prelabeled unextracted plasma. Statistical anal_vsis. The Student’s t test was used to make all statistical comparisons. All
GREGG
1560
Table 1. Effect of Acute Uremia on Baseline Measurements BUN (mg/lOO Sham
15*
Uremic
238
1 f
Pt
TG
(33)*
11 (29)
*The
numbers
statistical
tests
versity
FFA ml)
@M/liter)
4(33)
21 *
3(11)
114
f
4 (11)
0.71
*
0.05
(11)
124zt
6(29)
56 f
6 (8)
162
i
9 (8)
0.45
*
0.04
(8)
indicate
< 0.001
the number
of animals
< 0.001
< 0.002
studied.
t test.
the least squares regression analysis were performed on the Stanford Uniusing the Statistical Package for the Social Sciences (SPSS) computer
and
370-76
GILJCOX (mg/lOO
(d/ml)
c 0.02
in parentheses
by Student’s
ml)
AL.
f SE)
107+
< 0.001
tobtained
Insulin
(mg/lOO
ml)
(Mean
ET
computer
program.
RESULTS
The general metabolic effects of acute uremia on several important variables are given in Table I. These measurements were made 24 hr after surgery (and 30 hr of fasting), and indicate that there was a modest, but statistically significant elevation in fasting TG level. The hypertriglyceridemic effect of acute uremia was accentuated (an approximate twofold elevation over control) when both sham-operated and uremic rats were allowed free access to food. However. the uncertainty of both food intake and absorption in the acutely uremic rat led us to conduct all subsequent studies in fasted animals. Table I also indicates that acute uremia led to a significant rise in both plasma glucose and insulin levels. This combination of events could be interpreted as the development of insulin resistance, but the fact that free fatty acid (FFA) levels were lower in uremic rats also suggests that the insulin resistance does not include all insulinsensitive tissues. The results in Table 2 indicate that less TG accumulated in the plasma in the 2 hr following Triton WR 1339 blockade in acutely uremic rats than in shamoperated animals. Since plasma volumes were similar in the two groups of animals (see Materials and Methods), TGSR was significantly decreased in the acutely uremic rats. These results demonstrate that the elevated TG concentrations in acutely uremic rats cannot be due to increased TG production. Further evidence that TG production is not increased in acute uremia is seen in Figs. 1 and 2. Figure I summarizes the results of the hepatic perfusion experiments, in which TG secretion from isolated livers of sham-operated and uremic rats was studied. The results demonstrate that the rate of TG secretion Table 2. TG Secretion Rate (TGSR) of Sham-operated and Acutely Uremic Rats (Mean ATG’ (w/ml/2 Sham Uremic
(n = 20) (n =
17)
f SE)
Plasma Volume t hr)
9.6
zk 0.4
4.28
+
5.5
+ 0.4
4.10zt0.11
TGSR
g)
(w/100
0.12
20.4
i
0.7
11.2
*
0.9
(ml/100
< 0.001
PT *Increment volume/2 tBased
in
plasma
triglyceride
concentration
in 2 hr after
Triton
WR
1339.
TGSR
=
ATG x
plasma
hr. upon
(see Materials fobtained
g/hr)
the and
determination Methods).
by Student’s
The
t test.
of
plasma
mean
volume
of each
group
in seven sham-operated was
used to calculate
and TGSR.
seven acutely
uremic
rats
ACUTE
1561
UREMIA
Fig. 1. TG secretion rate from livers of acutely uremic (o--- o, n = 6) and shamoperated (o----e, n = 6) rats perfused for 2 hr in situ. FFA levels were measured at each time point, and were never lower than 0.5 mmole/ ml in the uremic rats. The TG secretion rate was significantly lower with p < 0.05 at 30 min and p < 0.01 at 60, 90, and 120 min in the acutely uremic rats.
$0
30 Minutes
from livers of uremic rats was lower at every time point during the 2-hr experimental period. The decrease in the rate of TG release from liver into the medium could be due to a decrease in synthesis and secretion of hepatic TG. Alternatively, TG synthesis might be normal, or even increased, and the effect of acute uremia could simply be to inhibit TG release. However, the data in Fig. 2 indicated that the latter alternative is unlikely, in that ultrastructural morphometric analysis indicated that there was a decrease in the number of the VLDL-particles in the Golgi complexes of hepatocytes from acutely uremic rats, without any change in the size of the particles. Thus, the concentration of intrahepatocyte VLDL-TG was decreased, not increased, as would be anticipated if the defect in acute uremia were limited to inhibition of TG release. As such, these combined results suggest that acute uremia results in a decrease in both hepatic TG synthesis and secretion. Finally, the rate of TG removal from the plasma of sham-operated and acutely uremic rats was quantitated by measuring the half-time of disappearance (6) of endogenously prelabeled VLDL-TG. Figure 3 shows the disappearance of injected radioactive VLDL-TG from a typical experiment conducted on a sham-operated and acutely uremic rat. The results of all these GOLGI-VLDL
250-
zoo-
150-
100
50
L
NUMBER
GOLGI-VLDL
SIZE
(PM)
Fig. 2. Effect of acute uremia on hepatocyte VLDL number and size. The results indicate that the total number of Golgi-associated VLDL (left panel) per hepatocyte from acutely uremic rats (U) are reduced by 50% as compared to sham-operated controls (5)) whereas the size (diameter) of the Golgi-associated VLDL (right panel) remain unchanged. Bars represent mean (f SE) of values from six hepatocytes of each of six rats.
GREGG
1562
VLDL-TG
REMOVAL
ET AL.
RATE
1ooc
z v
tot
Sham Operated
l[
cl
5
Fig. 3. Rate of disappearance of prelabeled VLDL-TG from plasma of an acutely uremic rat (o----o) as compared to Q sham-operated control rat
10 15 20 25 30 35
(w-4).
TIME (Minutes)
experiments are summarized in Table 3, and provide uremia leads to a prolongation of the disappearance plasma.
direct evidence that acute rate of VLDL-TG from
DISCUSSION
Nitzan” has previously shown that a rise in plasma TG levels is observed in rats with acute renal failure, and our results confirm this finding. Nitzan also found elevated plasma levels of insulin in these animals, as did we, and suggested that the hypertriglyceridemia was partially due to an insulin-induced increase in TG production. Elevation of plasma insulin levels have also been noted in patients with chronic renal failure by Bagdade et al.,3~‘7,‘8 and they have also suggested that the hyperinsulinemia may lead to an increase in TG production. The relationship between hyperinsulinemia and increased TG levels has been noted in several studies in man, ‘z-3’ but we have pointed out that increased plasma levels of insulin do not necessarily lead to increased TG production. This seems to be true in the case of acute uremia in the rat, and is Table 3. FTR and ttof VLDL-TG Removal of Sham-operoted and Acutely Uremic Rats (Mean f SE) FTT, (min ) Shorn Uremic
(n = 6) (n = 8)
ttf
= fractional
turnover
rate
= In 2/FTR.
$Obtoined
0.131
*
0.012
5.3
*
0.076
zk 0.005
9.1
+ 0.6
by Student’s
t test.
= slope
(see Materials
and
0.5
C. E. Mondon,
E. P. Reaven, and G. M. Reaven
Plasma triglyceride (TG) levels were elevated 24 hr after the production of acute uremia in rats. The effect of acute uremia on TG production rate was estimated by determining the rate of TG accumulation following Triton WR 1339 inhibition of lipoprotein removal, by measuring hepatic TG secretion rate during in situ liver perfusion, and by quantifying hepatocyte very low density lipoprotein content with the electron microscope. The results of all three of these approaches indicated that TG synthesis and secretion were decreased
in acute uremia, suggesting that the associated increase in plasma TG levels had to result from a removal defect. This hypothesis was tested directly by injecting prelabeled very low density lipoprotein TG into acutely uremic and control rats, and measuring its rate of disappearance from plasma. The ti of removal in acutely uremic rats was found to be approximately twice that of control, confirming the hypothesis that the rise in plasma TG levels in acute uremia is due to a defect in removal of TG from plasma.
A
TTENTION was drawn to the presence of hypertriglyceridemia in nonnephrotic patients with chronic renal failure in 1968,1m3and the increased incidence of hypertriglyceridemia in such patients has been repeatedly confirmed.4m’6 An elevated plasma triglyceride (TG) level can be due theoretically to an increase in rate of TG entry into plasma and/or a decrease in rate of TG removal from plasma: evidence has been presented in support of both of these mechanisms. Thus, the presence of elevated levels of plasma insulin in this situation has led to the suggestion that the hypertriglyceridemia is secondary to insulin-induced increases in hepatic TG production.3.‘7m’9 On the other hand, measurements of plasma postheparin lipolytic activity (PHLA) and very low density lipoprotein (VLDL) TG kinetics have suggested that the hypertriglyceridemia is secondary to a decrease in VLDL-TG removal rate from plasma. 3,4~8~‘3,‘7.‘xzo Either or both of these abnormalities may exist, but at present the majority of the evidence available is indirect. In an effort to obtain more direct evidence concerning the cause of hypertriglyceridemia in uremia, we have instituted a series of studies in which several experimental approaches have been used to define TG kinetics in uremic rats. In order to study the simplest model, we have begun by investigating the changes that occur in acutely uremic rats, a situation which has received relatively little attention. I9 In the current study, TG production rate was estimated by determining the rate of TG accumulation following Triton WR 1339 block-
From the Depariment of Medicine. Stanford University School of Medicine and Veterans Administration Hospital, Palo Alto, Calif Receivedfor publication March 17. 1976. Supported in part by National Institutes of Health Contract No-I-AM4-2213 from the Artihcial Kidnqv and Chronic Uremia Centers Branch, and by the Veterans Administration. Dr. Reaven is a Veterans Administration Medical Investigator, MRIS #7363. Reprint requests should be addressed to Gerald M. Reaven. M.D., V.A. Hospital (III-M). 3801 Miranda Avenue, Palo Alto, Calif: 94304. 0 I976 b.v Grune & Stratton. Inc. Metabolism, Vol. 25, No. 12 (December), 1976
1557
1558
GREGG
ET AL.
ade of TG removal, by measuring hepatic TG secretion rate during in situ liver perfusion, and by quantifying hepatocyte VLDL-TG content with electron microscopy. TG removal from plasma was directly measured by determining the disappearance rate of injected prelabeled VLDL-TG. The results of all four experimental approaches were consistent, and strongly suggest that VLDL-TG synthesis and secretion are decreased in acutely uremic rats, and hypertriglyceridemia in this situation is secondary to a decrease in TG removal from plasma. MATERIALS
AND METHODS
Protocol Female Sprague-Dawley rats, weighing 130~~150 g, were placed on a fat-free diet, consisting of 30”~ of calories as casein and 70°, as sucrose, 4- 7 days before surgical production of acute uremia. Rats were maintained on a 12-hr light (6 a.m. ~6 p.m.)/dark cycle. Food was removed at 8 a.m. the day of surgery, and rats were operated on 6- 8 hr later. Animals were lightly anesthetized with ether, and were either sham operated or made acutely uremic by bilateral ligation of the renal artery and vein via a midline incision. They were allowed access to water overnight, but received no food. All experiments were performed 24 hr after surgery.
Measurements Baseline detertninarions. Sham-operated and acutely uremic rats were anesthetized with sodium thiamylal, 60 mg/kg, and bled by aortic puncture into a syringe wetted with a 5”,, EDTA solution. The plasma was separated by centrifugation. frozen. and later assayed for TG.” BUN (BUN Analyzer. Beckman Instruments), glucose (Glucose Analyzer. Beckman Instruments), insulin,** and FFAZ3 levels. TG secretion rate (TGSR/. The TGSR was determined by the use of Triton WR 1339, which blocks the removal of intravascular lipoproteins.24V25 Thus, by measuring TG accumulation with time, TGSR can be determined. Baseline plasma TG concentrations (TGa) were measured in nonanesthetized rats on blood collection from the tip of the tail into capillary tubes rinsed with a 5”,, EDTA solution. Triton was then injected into the tail vein at a concentration of 600 mg/kg, an amount determined previously to produce maximal TG accumulation. Two hours later rats were anesthetized with sodium thiamylal and bled from the aorta into EDTA-wetted syringes. and plasma was again collected for determination of TG concentrations (TG2). TGSR was calculated from the following formula: TGSR
(mg/lOO g body weight/hr)
=
(TG2 - TGo)
x plasma
volume
2 hr where TG is expressed as mg/ml and plasma volume as ml/100 g body weight. Plasma volume. Plasma volume was determined on seven acutely uremic and seven shamoperated rats with the Evans blue dye dilution technique. Triton was given as described above, and Evans blue dye was injected into a cannulated right lumbar vein 2 hr after the Triton injection. Five minutes later a blood sample was drawn from the aorta into an EDTA-wetted syringe. The concentration of Evans blue was determined, and the plasma volume was calculated from the standard formula. There was no significant difference between sham-operated and uremic rats. Hepatic TG production. Hepatic TG production by livers isolated from acutely uremic and sham-operated rats was determined by in situ cyclic perfusion techniques previously described.26 The perfusing medium consisted of 904< defibrinated rat blood, obtained from retired male bicarbonate (KRB) breeding stock maintained on laboratory chow, and IO”, Krebs-Ringer buffer containing 3 g bovine albumin/l00 ml (hematocrit of 40). The initial circulating volume averaged 42 ml. All perfusions received a continuous infusion of porcine insulin at a rate of 33 mu/O.435 ml of KRB buffer per hour and oleic acid complexed with rat serum solution. The rat serum solution for infusion of oleic acid was prepared by circulating a 904” rat
1559
ACUTE UREMIA
blood perfusing medium described above through an isolated normal rat liver for 75 min to remove vasoconstrictor factors and freezing 15-ml aliquots of the recovered serum. For a perwith NaOH fusion experiment 800 Fmoles of oleic acid was dissolved in 959; ethanol, neutralized and KOH to a previously determined phenolthalein end point, and evaporated to near dryness in a boiling water bath. The residue was dissolved in 5 ml KRB buffer and mixed with 15 ml of rat serum solution thawed to room temperature. The mixture was then filtered through glass wool to remove macroaggregates, and provided sufficient volume of oleic acid-serum complex for three liver perfusions. The concentration of oleic acid averaged 28 rmole/ml and was infused at a rate of 2.31 ml/hr. Net TG production was calculated from net changes in perfusate concentration after correcting for addition of the infusate and sampling losses of 1.8 ml every 30 min. Electron microscopic evaluation of hepatocyte VLDL production. Tissue taken from the left lobe of the liver of six sham and six rats with acute uremia was fixed, processed, and subjected to ultrastructural morphometric analysis as outlined in detail in a previous report from this laboratory.” In brief, an estimate of the total number of Golgi-associated VLDL particles was obtained from six hepatocytes from each animal and an estimate of VLDL size was obtained from measurements of the diameters of all the VLDL particles in photographic enlargements of two randomly selected Golgi complexes from six hepatocytes of each animal. In sham-operated animals this means that roughly 1400 particles were counted and roughly 500 particles were measured per animal. The use of Golgi&VLDL content as one of several indicators of hepatocyte VLDL production appeared justified in view of the fact that Go1 i $ complexes have been described as temporary storage and packaging depots for export VLDL.**’ 9 VLDL-TG removal rate. The effect of acute uremia on VLDL-TG removal rate from plasma was studied by injecting VLDL-TG that had been prelabeled in vivo with 3H-glycerol. and measuring its rate of removal. Female donor rats, weighing 200-250 g were maintained on a fat-free diet for 4-7 days. On the day of the experiment their food was removed at 8-9 a.m.. and 4 hr later they were injected via the tail vein under light ether anesthesia with 400 kCi of 3H-2-glycerol (I.C.N., 2 Ci/mmole). The donor rats were exsanguinated under sodium thiamylal anesthesia 20 min later. (Previous experiments had shown maximal incorporation of 3H-glycerol into TG 20 min after injection, at which time 95”” of the lipid extractable radioactivity was incorporated into VLDL-TG as determined by ultracentrifugation and thin-layer chromatography.) Serum was separated by centrifugation, and 0.8-m] aliquots were injected without anesthesia into the tail veins of sham-operated or acutely uremic rats. In control experiments the plasma obtained from the donor rats was used to isolate VLDL by methods previously described from 30 our laboratory, and aliquots of isolated VLDL. rather than unextracted plasma, were then injected into the experimental rats. After administration of prelabeled VLDL-TG. the tail was amputated proximal to the site of injection and 0.4 ml of blood was collected into capillary tubes rinsed with a 5”{ EDTA solution at approximately 5, 15, 25, and 35 min after the injection. (These samples took I-2 min to collect and in all subsequent calculations the time used for a collection period was the mean of the beginning and ending times of the blood collection.) The plasma was separated by centrifugation and stored frozen until analyzed. A lipid extract” was made of each sample, evaporated to dryness, and the radioactivity measured by liquid scintillation counting (Beckman LS-235) using a standard toluene scintillation mixture. The natural log of the cpm of each sample was determined and plotted against time. A best fit line was determined by least squares regression analysis and the slope was determined. The following calculations were made: slope = FTR (fractional
turnover
rate)
ti = In Z/FTR. The mean t$ of eight control rats as determined by following the disappearance of prelabeled plasma was essentially identical to that determined following the injection of isolated VLDL-TG into eight other control rats. Therefore, in order to avoid the possible alteration of the VLDL that might occur as the result of the isolation procedure, we carried out our studies of VLDL-TG removal rate with prelabeled unextracted plasma. Statistical anal_vsis. The Student’s t test was used to make all statistical comparisons. All
GREGG
1560
Table 1. Effect of Acute Uremia on Baseline Measurements BUN (mg/lOO Sham
15*
Uremic
238
1 f
Pt
TG
(33)*
11 (29)
*The
numbers
statistical
tests
versity
FFA ml)
@M/liter)
4(33)
21 *
3(11)
114
f
4 (11)
0.71
*
0.05
(11)
124zt
6(29)
56 f
6 (8)
162
i
9 (8)
0.45
*
0.04
(8)
indicate
< 0.001
the number
of animals
< 0.001
< 0.002
studied.
t test.
the least squares regression analysis were performed on the Stanford Uniusing the Statistical Package for the Social Sciences (SPSS) computer
and
370-76
GILJCOX (mg/lOO
(d/ml)
c 0.02
in parentheses
by Student’s
ml)
AL.
f SE)
107+
< 0.001
tobtained
Insulin
(mg/lOO
ml)
(Mean
ET
computer
program.
RESULTS
The general metabolic effects of acute uremia on several important variables are given in Table I. These measurements were made 24 hr after surgery (and 30 hr of fasting), and indicate that there was a modest, but statistically significant elevation in fasting TG level. The hypertriglyceridemic effect of acute uremia was accentuated (an approximate twofold elevation over control) when both sham-operated and uremic rats were allowed free access to food. However. the uncertainty of both food intake and absorption in the acutely uremic rat led us to conduct all subsequent studies in fasted animals. Table I also indicates that acute uremia led to a significant rise in both plasma glucose and insulin levels. This combination of events could be interpreted as the development of insulin resistance, but the fact that free fatty acid (FFA) levels were lower in uremic rats also suggests that the insulin resistance does not include all insulinsensitive tissues. The results in Table 2 indicate that less TG accumulated in the plasma in the 2 hr following Triton WR 1339 blockade in acutely uremic rats than in shamoperated animals. Since plasma volumes were similar in the two groups of animals (see Materials and Methods), TGSR was significantly decreased in the acutely uremic rats. These results demonstrate that the elevated TG concentrations in acutely uremic rats cannot be due to increased TG production. Further evidence that TG production is not increased in acute uremia is seen in Figs. 1 and 2. Figure I summarizes the results of the hepatic perfusion experiments, in which TG secretion from isolated livers of sham-operated and uremic rats was studied. The results demonstrate that the rate of TG secretion Table 2. TG Secretion Rate (TGSR) of Sham-operated and Acutely Uremic Rats (Mean ATG’ (w/ml/2 Sham Uremic
(n = 20) (n =
17)
f SE)
Plasma Volume t hr)
9.6
zk 0.4
4.28
+
5.5
+ 0.4
4.10zt0.11
TGSR
g)
(w/100
0.12
20.4
i
0.7
11.2
*
0.9
(ml/100
< 0.001
PT *Increment volume/2 tBased
in
plasma
triglyceride
concentration
in 2 hr after
Triton
WR
1339.
TGSR
=
ATG x
plasma
hr. upon
(see Materials fobtained
g/hr)
the and
determination Methods).
by Student’s
The
t test.
of
plasma
mean
volume
of each
group
in seven sham-operated was
used to calculate
and TGSR.
seven acutely
uremic
rats
ACUTE
1561
UREMIA
Fig. 1. TG secretion rate from livers of acutely uremic (o--- o, n = 6) and shamoperated (o----e, n = 6) rats perfused for 2 hr in situ. FFA levels were measured at each time point, and were never lower than 0.5 mmole/ ml in the uremic rats. The TG secretion rate was significantly lower with p < 0.05 at 30 min and p < 0.01 at 60, 90, and 120 min in the acutely uremic rats.
$0
30 Minutes
from livers of uremic rats was lower at every time point during the 2-hr experimental period. The decrease in the rate of TG release from liver into the medium could be due to a decrease in synthesis and secretion of hepatic TG. Alternatively, TG synthesis might be normal, or even increased, and the effect of acute uremia could simply be to inhibit TG release. However, the data in Fig. 2 indicated that the latter alternative is unlikely, in that ultrastructural morphometric analysis indicated that there was a decrease in the number of the VLDL-particles in the Golgi complexes of hepatocytes from acutely uremic rats, without any change in the size of the particles. Thus, the concentration of intrahepatocyte VLDL-TG was decreased, not increased, as would be anticipated if the defect in acute uremia were limited to inhibition of TG release. As such, these combined results suggest that acute uremia results in a decrease in both hepatic TG synthesis and secretion. Finally, the rate of TG removal from the plasma of sham-operated and acutely uremic rats was quantitated by measuring the half-time of disappearance (6) of endogenously prelabeled VLDL-TG. Figure 3 shows the disappearance of injected radioactive VLDL-TG from a typical experiment conducted on a sham-operated and acutely uremic rat. The results of all these GOLGI-VLDL
250-
zoo-
150-
100
50
L
NUMBER
GOLGI-VLDL
SIZE
(PM)
Fig. 2. Effect of acute uremia on hepatocyte VLDL number and size. The results indicate that the total number of Golgi-associated VLDL (left panel) per hepatocyte from acutely uremic rats (U) are reduced by 50% as compared to sham-operated controls (5)) whereas the size (diameter) of the Golgi-associated VLDL (right panel) remain unchanged. Bars represent mean (f SE) of values from six hepatocytes of each of six rats.
GREGG
1562
VLDL-TG
REMOVAL
ET AL.
RATE
1ooc
z v
tot
Sham Operated
l[
cl
5
Fig. 3. Rate of disappearance of prelabeled VLDL-TG from plasma of an acutely uremic rat (o----o) as compared to Q sham-operated control rat
10 15 20 25 30 35
(w-4).
TIME (Minutes)
experiments are summarized in Table 3, and provide uremia leads to a prolongation of the disappearance plasma.
direct evidence that acute rate of VLDL-TG from
DISCUSSION
Nitzan” has previously shown that a rise in plasma TG levels is observed in rats with acute renal failure, and our results confirm this finding. Nitzan also found elevated plasma levels of insulin in these animals, as did we, and suggested that the hypertriglyceridemia was partially due to an insulin-induced increase in TG production. Elevation of plasma insulin levels have also been noted in patients with chronic renal failure by Bagdade et al.,3~‘7,‘8 and they have also suggested that the hyperinsulinemia may lead to an increase in TG production. The relationship between hyperinsulinemia and increased TG levels has been noted in several studies in man, ‘z-3’ but we have pointed out that increased plasma levels of insulin do not necessarily lead to increased TG production. This seems to be true in the case of acute uremia in the rat, and is Table 3. FTR and ttof VLDL-TG Removal of Sham-operoted and Acutely Uremic Rats (Mean f SE) FTT, (min ) Shorn Uremic
(n = 6) (n = 8)
ttf
= fractional
turnover
rate
= In 2/FTR.
$Obtoined
0.131
*
0.012
5.3
*
0.076
zk 0.005
9.1
+ 0.6
by Student’s
t test.
= slope
(see Materials
and
0.5
