| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Medical Department M, Aarhus Kommunehospital (S.N., N.M., J.S.C., J.O.L.J.), and Medical Department C, Aarhus Amtssygehus (S.B.P.), Aarhus University Hospital, DK-8000 Aarhus C, Denmark
Address all correspondence and requests for reprints to: Dr. Jens O. L. Jørgensen, Medical Department M, Aarhus Kommunehospital, DK-8000 Aarhus C, Denmark. E-mail: . jolj{at}dadlnet.dk
Abstract
The lipolytic properties of GH are essential for the acute effects on glucose metabolism and insulin sensitivity, whereas its more long-term impact on substrate metabolism is uncertain. The aim of the study was to evaluate the influence of pharmacological antilipolysis on substrate metabolism during constant and continued GH exposure. Seven adult GH-deficient (GHD) patients were studied twice in a double-blind randomized order: 1) after 4 wk of acipimox treatment (250 mg, orally, three times daily) and 2) after 4 wk of placebo treatment. Daily GH replacement was continued throughout both study periods. At the end of each period glucose and lipid oxidation rates were assessed by indirect calorimetry, and the protein oxidation rate was estimated by urinary excretion of urea. Endogenous glucose production and whole body protein metabolism were assessed by isotope dilution techniques using tritiated glucose and stable phenylalanine and tyrosine isotopes, respectively. GH and IGF-I levels were not different between periods, whereas FFA and glycerol levels were distinctly suppressed after 4 wk of pharmacological antilipolysis [FFA, 256 ± 63 (acipimox) vs. 596 ± 69 (placebo) µmol/liter; P = 0.001]. Likewise, plasma levels of total and low density lipoprotein cholesterol as well as triglycerides were significantly reduced after acipimox. Despite this, lipid oxidation rates were identical at the end of the two treatment periods [589 ± 106 (acipimox) vs. 626 ± 111 (placebo) kcal/24 h; P = 0.698]. The total and oxidative rates of glucose as well as protein oxidation and urea excretion were identical at the end of the two treatment periods (P > 0.05). Phenylalanine flux, a measure of protein turnover, was increased [34.62 ± 1.83 (acipimox) vs. 33.15 ± 1.61 (placebo) µmol/kg·h; P = 0.049] as was phenylalanine incorporation into protein, a measure of protein synthesis [30.79 ± 1.67 (acipimox) vs. 28.97 ± 1.51 (placebo) µmol/kg·h; P = 0.035].
The following conclusions were reached: 1) prolonged antilipolysis by means of acipimox stimulates protein turnover without affecting net protein balance; and 2) acipimox in combination with constant GH exposure results in sustained suppression of circulating levels of FFA, glycerol, and triglycerides without a reduction in the rate of lipid oxidation. The site and origin of lipid fuels for oxidation during suppression of lipolysis remain to be determined.
BASED ON in vitro studies of rodent muscle preparations Randle (1) originally put forward the hypothesis that an inverse relation exists between the oxidation of lipids and carbohydrates. In parallel, data have been published to support an inverse relation between lipid and protein metabolism in human subjects (2, 3, 4). Artificial elevation of FFA levels by infusion of lipids or induction of intravascular lipolysis induces both hepatic and peripheral insulin resistance (5, 6, 7, 8, 9, 10, 11, 12). In addition, the lipid oxidation rate is increased, and the glucose oxidation rate is decreased by lipid infusion (9, 10, 13, 14). Conversely, short-term administration of acipimox, a nicotinic acid derivative with antilipolytic properties, lowers circulating FFA levels, reduces the lipid oxidation rate, and increases the glucose oxidation rate in healthy adults (15, 16) as well as in diabetic subjects (17). A concomitant protein-sparing effect has been recorded during the infusion of lipids and ketone bodies (2, 3, 4), whereas increased protein degradation during pharmacological antilipolysis produced by acipimox administration has been suggested by increased urinary excretion of urea (15).
The effect of long-term pharmacological antilipolysis has been studied in patients with noninsulin-dependent diabetes mellitus. These studies suggest that the suppression of FFA levels and the subsequent improvement in glucose metabolism are only transient (18, 19). The reason for this escape phenomenon is not clear, but the observed rebound increase in GH release is likely to ultimately offset the antilipolytic effects (20).
To enable a study of the impact of prolonged suppression of lipolysis on the metabolic effects of GH we therefore investigated GH-substituted GH-deficient (GHD) adults after 4 wk of either acipimox or placebo administration in a double-blind, cross-over design.
Subjects and Methods
Subjects and design
Seven adult GHD patients were included in the study. One patient suffered from idiopathic childhood-onset pituitary failure, and six patients had hypopituitarism secondary to juxtasellar pathology and/or its treatment (Table 1
). The diagnosis of GH deficiency was ultimately based on a stimulation test (insulin tolerance test, arginine infusion, or both) with a mean peak GH concentration of 0.33 µg/liter (Table 1). The stimulation test was performed before the start of GH replacement in adulthood. All patients had been receiving GH replacement therapy for at least 6 months before the study, and additional pituitary replacement therapy, where indicated, had remained unaltered for the same period. The patients were instructed not to change dietary habits during the study period. The study was approved by the regional ethical committee system, and informed consent was obtained from all participants.
|
Indirect calorimetry
At the end of the study period oxygen consumption and carbon dioxide production were measured over a period of 30 min by indirect calorimetry using a computerized open circuit that measured gas exchange across a canopy (Deltatrac, Datex Instruments, Inc., Helsinki, Finland). From the measurement of gas exchange, resting energy expenditure and the respiratory quotient (RQ) were calculated. Glucose and lipid oxidation rates were calculated from the above-mentioned measurements (21) after correction for protein oxidation as estimated by the urinary excretion rate of urea.
Protein metabolism
At 0800 h an iv bolus injection of L-[15N]phenylalanine (0.7 mg/kg), L-[2H4]tyrosine (0.5 mg/kg,) and L-[15N]tyrosine (0.3 mg/kg; Cambridge Isotope Laboratories, Inc., Andover, MA) were given. This was followed by a continuous iv infusion of L-[15N]phenylalanine (0.7 mg/kg·h) and L-[2H4]tyrosine (0.5 mg/kg·h). The chemical, isotopic, and optical purities of the isotopes were tested beforehand, and the solutions were prepared under sterile precautions and were shown to be free of bacteria and pyrogens. L-[15N[rsqb Phenylalanine, L-[2H4]tyrosine, and L-[15N]tyrosine were measured as their t-butyldimethylsilyl ether derivatives under electron ionization conditions (22). In addition, concentrations of phenylalanine and tyrosine were measured by mass spectrometry using L-[2H8]phenylalanine and L-[13C6]tyrosine, respectively, as internal standards. For measurements of whole body phenylalanine kinetics, the equations reported by Thompson et al. (23) were used.
Phenylalanine flux (Qp) and tyrosine flux (Qt) were calculated as follows: Qflux = I[(Ei/Ep) - 1], where I represents the rate of tracer infusion (micromoles per kg/h), and Ei and Ep are the enrichment of the tracer infused and the plasma enrichment of the tracer at steady state, respectively.
The rate phenylalanine of conversion to tyrosine (Ipt) was calculated as follows: Ipt = Qt x ([2H4]Tyrei/[15N]Pheei) x [Qp/Ip + Qp)], where [2H4]Tyrei and [15N]Pheei represent the isotopic enrichments of the respective tracers in plasma, and Ip is the infusion rate of [15N]phenylalanine (micromoles per kg/h).
Phenylalanine incorporation into protein is calculated by subtracting Ipt from Qp, because phenylalanine is irreversibly lost by either conversion into tyrosine or incorporation into protein.
Endogenous glucose production
At 0800 h a bolus injection of [3-3H]glucose (20 µCi) was given. This was followed by a continuous iv infusion of [3-3H]glucose (10.4 µCi/h). Three hours were allowed for tracer equilibration before the assessment of endogenous glucose production. The specific activity of tritiated glucose was determined as previously described by Møller et al. (24). The glucose rate of appearance during steady state was calculated as the specific activity of glucose divided by the rate of infusion of tritiated glucose.
Other assays
Plasma glucose was determined in duplicate using a glucose analyzer (Beckman Coulter, Inc., Palo Alto, CA). Whole blood glycerol, 3-hydroxybutyrate (3-OHB), lactate, and alanine were analyzed by fluorometric enzymatic methods (25).GH and cortisol were assayed by a time-resolved fluoroimmunoassay (Delfia, Wallac, Inc., Turku, Finland), whereas IGF-I and IGF-II were determined by an in-house, time-resolved fluoroimmunoassay (26). Insulin and C peptide were measured by commercially available immunoassays (DAKO Corp., Glostrup, Denmark). FFA were measured by a colorimetric method. Plasma glucagon was measured by an in-house assay (27). The urea concentration in urine and plasma was determined by a commercially available kit (COBASINTEGRA, Roche, Hvidovre, Denmark). Plasma cholesterol, high density lipoprotein (HDL) cholesterol and triglycerides (TG) were measured by standard enzymatic kits. Low density lipoprotein (LDL) cholesterol was calculated by the Friedewald’s equation (28). The urea nitrogen synthesis rate (UNSR) was calculated as previously described (29). Urine was collected between 0800–1100 h, before which the subjects were instructed to empty their bladder. During the collection period the subjects were given tap water orally. Body composition was measured by bioelectrical impedance.
Lipoprotein lipase (LPL) activity was determined essentially as previously described (30, 31). Briefly, 500 mg adipose tissue were homogenized in a buffer containing 0.25 M sucrose and 1.0 mM K2EDTA (pH 7.4) at 4 C, and the homogenate was centrifuged for 20 min at 12,000 x g at 4 C. The LPL activity in the postmitochondrial supernatant was determined by estimating the specific hydrolysis of [14C]triolein after 60 min of incubation. FFA were extracted from the incubation mixture as previously described (30) and were measured by liquid scintillation counting.
All results for circulating hormones and metabolites are the mean of triplicate analyses at the end of the study period (150–180 min). All results are expressed as the mean ± SE. Differences in the analytes between acipimox and placebo periods were analyzed by paired t test when data were normally distributed. Otherwise, Wilcoxon’s nonparametric test was used (applicable for the following analytes: cortisol, LPL activity, FFA, 3-OHB, glucose and lipid oxidation, and tyrosine flux). All calculations were made using the computer program SPSS, version 10.0 (SPSS, Inc., Chicago, IL). P < 0.05 was considered statistically significant.
Results
Body composition
Body weight (kilograms) was not different at the end of two periods [89.9 ± 5.4 (acipimox) vs. 90.4 ± 5.3 (placebo); P = 0.490]. Body composition, as measured by bioelectrical impedance (
), was also unchanged [429 ± 17 (acipimox) vs. 442 ± 17 (placebo); P > 0.05].
Circulating hormones and metabolites
The circulating levels of GH, IGF-I, and IGF-II did mot differ at the end of the two treatment periods (Table 2). Insulin and C peptide levels tended to be lower after acipimox treatment than after placebo, whereas glucagon and cortisol levels were identical (Table 2).
|
). Plasma levels of glucose as well as gluconeogenic substrate precursors were similar on the two occasions (Table 3).
|
|
The circulating levels of total cholesterol, LDL cholesterol, and TG were decreased by acipimox treatment [cholesterol, 5.17 ± 0.19 (acipimox) vs. 5.77 ± 0.28 (placebo) mmol/liter (P = 0.036); LDL cholesterol, 3.26 ± 0.09 (acipimox) vs. 3.77 ± 0.17 (placebo) mmol/liter (P = 0.032); TG, 1.60 ± 0.34 (acipimox) vs. 2.04 ± 0.41 (placebo) mmol/liter (P = 0.003)]. By contrast, HDL cholesterol was unchanged after acipimox [1.19 ± 0.12 (acipimox) vs. 1.09 ± 0.08 (placebo) mmol/liter; P = 0.274]. Adipose tissue LPL lipase activity was not affected by acipimox treatment [459 ± 73 (acipimox) vs. 375 ± 53 (placebo) cpm; P = 0.310].
Substrate metabolism (Table 3b
)
Resting energy expenditure as well as the oxidative rates of protein and glucose were identical during acipimox compared placebo to administration. Likewise, endogenous glucose production was similar after acipimox and placebo treatments (Table 3b), and no differences in the urinary urea excretion rate or the UNSR were recorded (Table 3b).
Pharmacological antilipolysis induced a state of accelerated protein turnover. Whole body protein breakdown, as reflected by phenylalanine flux (QP) was elevated after acipimox administration [QP, 34.6 ± 1.8 (acipimox) vs. 33.2 ± 1.6 (placebo) µmol/kg·h; P < 0.05; Fig. 1]. By contrast, neither tyrosine flux (QT), nor phenylalanine conversion to tyrosine (Ipt) was significantly affected by acipimox (Fig. 1), and consequently, protein synthesis, as assessed by QP - Ipt, was elevated after acipimox treatment [QP - Ipt, 30.8 ± 1.7 (acipimox) vs. 29.0 ± 1.5 (placebo) µmol/kg·h; P = 0.04; Fig. 1].
|
In previous studies of long-term acipimox administration the antilipolytic action levels off after treatment for more than a few days. A possible explanation for this rebound lipolysis is a stimulatory effect of lowered FFA concentrations on GH secretion. The present design enabled an evaluation of the effect of prolonged pharmacological antilipolysis on glucose and protein metabolism in subjects with constant and stable GH levels.
Our study demonstrates sustained antilipolysis after 4 wk of acipimox administration, as evidenced by markedly suppressed levels of lipid intermediates. Despite this, the oxidation rates of both lipid and glucose as well as endogenous glucose production were not significantly different after acipimox. By contrast, a state of accelerated protein turnover was observed after acipimox. In this context it should be noted that with the present design intrapersonal variability of the parameters measured over 3–4 months may have introduced a certain amount of biological noise, and minor metabolic differences may have escaped detection (type 2 errors).
As mentioned, the maintenance of antilipolysis is in contrast to previous findings obtained in diabetic subjects (18, 19, 32). Importantly, GH, glucagon, and cortisol levels were reported to be elevated after both short-term and prolonged (4-wk) antilipolysis (15, 20, 33, 34). The lipolytic action of GH is well recognized (35), whereas glucagon does not seem to influence lipolysis in man (36), and the influence of glucocorticoids on lipolysis remains uncertain (37). The present study therefore suggests that stimulation of GH secretion may be important for rebound lipolysis during long-term pharmacological antilipolysis.
Our data indicate that antilipolysis stimulates proteolysis as well as protein synthesis, which has not previously been reported. Short-term infusion of lipids has been shown to induce hypoaminoacidemia (38) and to reduce proteolysis (3), and infusion of ketone bodies inhibits leucine oxidation and stimulates protein synthesis (4). In line with this, pharmacological antilipolysis increases protein breakdown, estimated by urinary urea excretion (15). It may be speculated that decreased FFA availability for oxidation during pharmacological antilipolysis evokes increased mobilization of protein. Another putative mechanism is decreased insulin secretion secondary to increased insulin sensitivity during FFA lowering. We observed borderline decrements in both insulin and C peptide concentrations in the circulation, and insulin has repeatedly been shown to inhibit protein breakdown (39, 40, 41). In addition, we saw unaltered rates of ureagenesis and phenylalanine to tyrosine conversion, indicating that increased whole body rates of protein synthesis counterbalanced the increase in protein breakdown. What precipitates this compensatory rise in protein synthesis remains unknown, in particular in view of the unchanged concentrations of GH and IGF-I and the marginally decreased insulin levels.
It would be anticipated from previous studies with acipimox that lowering of FFAs is accompanied by a simultaneous suppression of the lipid oxidation rate. By contrast, the rate of lipid oxidation was unchanged despite distinctly lowered FFA levels. The reason for this discrepancy is not clear, but several possible explanations exist. The respiratory quotient could be falsely lowered by net ongoing gluconeogenesis during the administration of acipimox. Gluconeogenesis from alanine consumes CO2 without any exchange of O2, thus lowering the RQ. Whether increased gluconeogenesis would be of sufficient magnitude to explain our observations is doubtful, and it is noteworthy that urea excretion was unaltered, which speaks against quantitatively significantly elevated rates of gluconeogenesis. Alternatively, the RQ value indeed reflects an unchanged lipid oxidation rate despite lowered FFA levels. This suggests that the net lipid oxidation rate and the plasma FFA oxidation rate are regulated independently as previously reported (42, 43). It is also noteworthy that GH in vitro has been shown to directly stimulate lipid oxidation independently of its lipolytic actions (44). The source of lipid fuels for oxidation in our acipimox study could be triacylglycerol or ketone bodies, rather than circulating FFA. Ketone bodies, however, are an unlikely candidate in view of the low circulating levels. It is possible that the hydrolysis of circulating TG in the vasculature of tissues such as skeletal and cardiac muscle increased in the acipimox experiment. The low insulin levels could have mediated reciprocal regulation of LPL activity in adipose and muscle tissue, respectively. Reduced LPL activity in adipose tissue and increased activity in skeletal muscle tissue, which is seen during fasting (45), could explain the unaltered rate of lipid oxidation observed in our study. In our study LPL activity in fat biopsies from sc abdominal tissue was not significantly changed by acipimox. The circulating fasting levels of total and LDL cholesterol as well as total TG were significantly reduced by acipimox, whereas HDL cholesterol increased insignificantly. This is compatible with the previously reported effects of acipimox (32). On the whole, our data therefore do not provide an obvious explanation for the underlying mechanisms. Future studies including measurements of skeletal muscle LPL activity and turnover rates of lipoproteins, TG, and FFA are needed. It would also be informative to include control periods without GH and acipimox and to repeatedly measure lipid oxidation during each study period.
Endogenous glucose production and circulating gluconeogenic substrate concentrations were unchanged by acipimox treatment for 4 wk. In patients with noninsulin-dependent diabetes acipimox appears to stimulate gluconeogenesis from lactate despite unaltered hepatic glucose production (46), but in healthy subjects acipimox has been reported to stimulate endogenous glucose production (15). In the latter study FFA levels were suppressed, whereas GH levels were elevated. In our study FFA levels remained suppressed, whereas endogenous glucose production and GH levels were unchanged, which indirectly is compatible with a direct stimulatory effect of GH on hepatic glucose output, as previously reported (47).
In conclusion, our study demonstrates that long-term pharmacological antilipolysis persistently lowers FFA levels in GHD patients receiving stable GH substitution, which suggests that (increased) GH secretion may have accounted for the rebound increase in FFA levels observed in previous studies. Despite decreased FFA levels, the lipid oxidation rate was unchanged, which suggests alternative sources of lipid for oxidation. In addition, consistently lowering FFA levels promotes accelerated protein turnover. The physiological significance of these observations as well as the underlying mechanisms remain to be investigated.
Footnotes
Abbreviations: GHD, GH-deficient; HDL, high density lipoprotein; Ipt, phenylalanine conversion to tyrosine; LDL, low density lipoprotein; LPL, lipoprotein lipase; 3-OHB, 3-hydroxybutyrate; QP, phenylalanine flux; QT, tyrosine flux; RQ, respiratory quotient; TG, triglycerides; UNSR, urea nitrogen synthesis rate.
Received October 17, 2001.
Accepted March 7, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  M. B. Krag, L. C. Gormsen, Z. Guo, J. S. Christiansen, M. D. Jensen, S. Nielsen, and J. O. L. Jorgensen Growth hormone-induced insulin resistance is associated with increased intramyocellular triglyceride content but unaltered VLDL-triglyceride kinetics Am J Physiol Endocrinol Metab, March 1, 2007; 292(3): E920 - E927. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Vickers, P. L. Hofman, P. D. Gluckman, P. E. Lobie, and W. S. Cutfield Combination therapy with acipimox enhances the effect of growth hormone treatment on linear body growth in the normal and small-for-gestational-age rat Am J Physiol Endocrinol Metab, December 1, 2006; 291(6): E1212 - E1219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Svensson and B.-A. Bengtsson Growth Hormone Replacement Therapy and Insulin Sensitivity J. Clin. Endocrinol. Metab., April 1, 2003; 88(4): 1453 - 1454. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) and at the end of the placebo period (
). *, P < 0.05 for acipimox vs. placebo (by paired t test).