If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The pathophysiology of hypertriglyceridemia is complex hampering effective therapeutic strategies. Increased central parasympathetic nerve activity was shown to inhibit hepatic triglyceride (TG) excretion via modulation of liver stearyl-CoA desaturase (SCD)-1 activity in rodents. We evaluated the impact of 7-h lactate clamping on VLDL-TG homeostasis in humans.
Eight normolipidemic, male subjects were subjected to a continuous infusion of l-lactate (target concentration 3 mmol/L) or saline for 7 h in random order on two separate occasions. TG kinetics in very low density lipoproteins (VLDL1 and 2) were measured after a bolus injection of [1,1,2,3,3]-2H5-glycerol. Palmitic acid (16:0) and palmitoleic acid (16:1) in VLDL1 and VLDL2 were measured as a reflection of liver SCD1 activity.
Plasma TG levels changed by 0.16 ± 0.09 mmol/L during lactate vs −0.15 ± 0.08 mmol/L during saline (P < 0.05). VLDL1 16:1/16:0 ratio increased to 1.2 ± 0.7 during lactate versus a decrease during saline by −1.5 ± 0.6 (p = 0.01). During lactate VLDL1-TG excretion was higher compared to saline (1604 [827–2870] versus 1285 [505–2155] μmol glycerol; p < 0.05), trending toward higher VLDL1-TG pool sizes during lactate (28%; p = 0.07 versus saline).
In normolipidemic men, 7-h l-lactate clamp increases, rather than decreases SCD1 activity and hepatic TG secretion leading to elevated plasma TG levels. These conflicting data between human and rodents on central regulation of hepatic TG excretion illustrate that experimental findings on the role of the central nervous system in lipid metabolism should be interpreted with caution.
A 7-h lactate clamp leads to an increase in TG secretion mainly in VLDL1.
A small increase in SCD1 activity coincides with an increase in TG in VLDL1.
These conflicting data between humans and rodents on regulation of hepatic TG excretion illustrate that experimental findings on regulation of TG metabolism should be interpreted with caution.
Increased production of VLDL triglycerides (TGs) by the liver is a hallmark of the metabolic syndrome and type 2 diabetes, leading to increased plasma TG levels, accumulation of pro-atherogenic apoB-containing lipoprotein particles and subsequent increased risk of cardiovascular disease [
]. Two distinct VLDL populations can be discriminated. VLDL1 particles, the major fraction secreted by the liver, are larger, less dense and contain more triglycerides, whereas VLDL2 particles are smaller and more cholesterol-enriched. The hepatic VLDL1 secretion rate from the liver is more sensitive to metabolic changes such as the presence of insulin resistance [
], are required as a major substrate for hepatic TG synthesis. A number of key molecules are essential in hepatic VLDL processing. Microsomal triglyceride transfer protein (MTP) enables ApoB containing lipoprotein assembly by promoting fusion of precursor apoB particles with lipid droplets to form a small VLDL particle [
]. The presence of monounsaturated fatty acids, derived from the action of stearoyl-CoA desaturase (SCD1) enables further loading of TGs into the mature VLDL particles. SCD1 is a key enzyme catalyzing the formation of monounsaturated long-chain fatty acids from saturated fatty acyl-CoAs [
A role for the central nervous system in controlling hepatic TG production was proposed in multiple studies. TG production in hyperinsulinemia was attenuated when infusing neuropeptide Y (NPY) intra-cerebroventricularly (icv) in mice [
]. Lactate administration, either by systemic intravenous injection or by injection directly into the brain, increased hypothalamic pyruvate levels leading to metabolic alterations in glucose and lipid homeostasis. Interestingly, icv administration of glucose, which is subsequently metabolized to lactate in the hypothalamic region, led to a profound reduction in hepatic VLDL-TG production in rodents. This effect was abolished following disruption of the hepatic vagal nerve. Mechanistically, CNS-stimulation of hepatic vagal nerve activity was proposed to inhibit hepatic TG production via inhibition of SCD1-activity [
]. These effects could also be reproduced in rodents by peripheral lactate clamping, where doubling of plasma lactate levels resulted in a ∼40% decrease of plasma TG concentration. Based on these findings, a prominent role for CNS-regulation in hypertriglyceridemia was put forward.
To evaluate whether these phenomena could be extrapolated to the human setting, thereby offering potential new therapeutic targets to treat hypertriglyceridemic states, we subjected normolipidemic volunteers to a lactate clamp. Plasma lactate was clamped at 3 mmol/L during a period of 7 h. We measured [1,1,2,3,3]-2H5-glycerol incorporation in TG in both VLDL1 and VLDL2 fractions to assess TG production.
Eight normolipidemic, male subjects were recruited through advertisement. Subjects were healthy, had a median age of 24.5 years [23.3–33.5], an average BMI of 24.6 ± 3.2 kg/m2 and did not use any medication. Participants were excluded if they consumed more than two units of alcohol per day, had a BMI >30.0 kg/m2, used medication which could influence the glucose- or lipid metabolism, had diabetes or hypertension. The institutional review board of the Academic Medical Center of the University of Amsterdam, the Netherlands, approved the study. All subjects provided written informed consent. Clinical Trials.gov NCT-number NTR2522 identifies this study.
2.2 Experimental protocol
All study subjects were invited to the AMC on 2 separate days. They received a lactate infusion on one day and a control isotonic 0.9% saline infusion on the other day. The saline infusion was performed to exclude that the ingestion of a large amount of fluid would hamper the results. All subjects were admitted to the research unit after an overnight fast. Three days before the study subjects were asked to maintain a diet with at least 250 g of carbohydrates a day. One intravenous catheter was inserted into an antecubital vein for infusion; a second intravenous catheter was inserted retrogradely and put into a heated hand box to obtain arterialized venous blood [
]. Saline was administered during a 30-min adaptation phase, thereafter saline or l-lactate 30% (IVAC pump model 598) was given in random order during 2 separate study days at least two weeks apart. l-Lactate 30% was infused to reach a stable plasma lactate level of 3 mmol/L during 7 h with an average infusion rate of 2.52 mg/kg/min [
]. When the plasma lactate level had stabilized at 3 mmol/L, the participants received a bolus of 100 mL (5 mg/mL) [1,1,2,3,3-2H5] glycerol (>99% enriched; Cambridge Isotopes, Andover, MA, USA). The subjects did not consume any food but were allowed to drink water during the study day. Blood sampling started at 09:00 a.m. after the adaption phase. Blood sampling was done before the injection of the tracer and at 15, 30, 60, 90, 120, 150, 180, 240, 300, 360 and 390 min post-tracer injection. The bed-side blood gas analyzer (Osmetech OPTI™ CCA Analyzer, OPTI Medical Systems, Inc. Roswell, Georgia, USA) was used to monitor pH levels instantaneously to keep the pH under the level of 7.55. During the first hour, every 10 min pH-measurement was performed, thereafter every half an hour. With the bedside lactate analyzer (YSI 2300 STAT Plus™ Glucose & Lactate Analyzer, Xylem, New York, New York, United States) we assessed the plasma lactate level every 10 min until 3 mmol/L was reached. Lactate levels were further monitored throughout the clamp every 30 min. All blood samples were collected on ice in EDTA tubes, centrifuged within 10 min (T = 4 °C, 3000 rpm, t = 20 min) and stored at −80 °C until analysis. During infusion the subjects remained in resting position in a silent environment. Blood glucose was measured using a bed-side calibrated glucose sensor (YSI 2300 STAT S; YSI Yellow Springs, OH, USA). Triglycerides (Randox, UK) and NEFA (DiaSys) were measured with a commercially available colorimetric assay on the Cobas Mira system (Roche, Switzerland). LDL cholesterol levels were analyzed with a commercial assay from Wako (Neuss, Germany). Plasma apoB was measured with a turbidimetric assay on the Cobas Mira system (Roche, Switzerland). Plasma insulin was assessed using an ELISA (Mercodia).
2.3 Preparation of l+-lactate infusion solution
30% l+-Lactic acid solution (w/w) and a 34.2% sodium-lactate solution were manufactured under GMP conditions (Pharmacy department, HAGA hospital, The Hague, The Netherlands). The final lactate infusion mixture was prepared by mixing 2.5 L of lactic acid solution 30%, 14.3 L of sodium-lactate 34.2% and 2.05 L water. The final pH was 4.8. Subsequently, water for injection was added to a final volume of 20 L. After filtration, the solution was filled in glass bottles of 100 mL and sterilized (15 min, 121 °C). Prior to administration, the lactate infusion mixture was diluted 1:9.5 with water for injection.
2.4 Isolation of VLDL1 and VLDL2 fractions
VLDL1 and VLDL2 were isolated from 4 mL frozen plasma by cumulative ultracentrifugation using a discontinuous salt gradient using an SW41 rotor in Beckman L3-50 ultracentrifuge (Beckman Inc., Palo Alto, CA 94034) [
]. The centrifuge tubes were coated using polyvinylalcohol and isopropanolol (Merck, Darmstadt, Germany). To increase the density to 1.21 g/mL Solid KBr (0.14 mg/mL) was added. A discontinuous gradient was formed by addition of 2.5 mL of d = 1.065 g/mL, d = 1.020 g/mL and d = 1.006 g/mL KBr solutions respectively. The chylomicron fraction was collected in the top 0.5 mL after centrifugation of 32 min at 36,000 rpm, 4 °C. VLDL1 and VLDL2 were isolated after respectively 3 h and 28 min and 17 h at 4 °C and 36,000 rpm. TG concentrations in all lipoprotein fractions were determined by using a commercially enzymatic assay (Randox, Belfast, Ireland) on a Cobas Mira autoanalyzer (ABX, Montpellier cedex, France). Fractions were frozen at −80 °C.
2.5 Measurement of 16:1/16:0 ratio as a reflection of SCD1 activity
Quantification of plasma fatty acid esters was performed using a HP7890A Gas Chromatograph equipped with an HP7683 Injector and a HP5975C Mass Selective Detector (Agilent Technologies, Santa Clara, United States). Chromatography was performed using an HP-5MS fused silica capillary column (30 m × 0.25 mm inner diameter, 0.25 μm film thickness, Agilent Technologies, Santa Clara, USA). The GC–MS conditions were as follows: carrier gas, helium at a flow-rate of 1.1 mL/min; injector temperature, 250 °C, split mode; oven temperature 140 °C, increased at 5 °C/min to 300 °C, and held for 10 min. The mass spectrometer was operated under negative chemical ionization mode with methane as reactant gas. The ion source temperature and the quadrupole temperature were 150 °C and 106 °C respectively. A SIM program was used for mass spectrometry quantification. Fatty acids were extracted from acidic plasma with ethylacetate before its analysis by gas chromatography. The dried and evaporated samples were derivatized with bis-(trimethylsilyl) trifluoroacetamide containing 1% trimethylchlorosilane. The SCD1 index was determined by calculating product:substrate ratios (16:1/16:0) using the quantitated values for palmitoleic acid and palmitate.
2.6 Measurement of 2H5 glycerol enrichment in TG in VLDL1 and VLDL2
Enrichment of [1,1,2,3,3,2H5] glycerol in TG present in VLDL1 and VLDL2 was measured as described in Ref. [
]. In short, 125 μL VLDL1 or VLDL2 was incubated with 125 μL lipase solution (1.25 mg lipase, Sigma–Aldrich, St Louis, USA) and 0.21 mg colipase (Sigma) 10 mL albuman (Sanquin, Amsterdam, The Netherlands) for 16 h at 37 °C. After incubation the glycerol was isolated using ion exchange column chromatography as described in Ref. [
] with minor modifications. A single plasma compartment for VLDL1 and VLDL2 was used and an additional glycerol-phosphate pool between glycerol and TG. This resulted in the compartmental model shown in Fig. 1, in which pool sizes Q8, Q7, Q6, etc. refer to compartments, 8, 7, 6, etc and fractional turnover rates k6,7, k4,6, k3,4, etc. refer to the fractional transfer rate (FTR) of glycerol from compartment 7 to compartment 6, from 6 to 4, from 4 to 3, etc.
A number of assumptions were copied from Adiels et al. [
]; the average molar weight of TG is 885 and 1 mol TG equals 1 mol glycerol. Total plasma volume is 4.5% of bodyweight, the average plasma volume (V) was 3.7 L, which was used as a Bayesian value (±20%) for iterative modeling. Identical distribution volumes of VLDL1, VLDL2 and glycerol were assumed. Plasma glycerol enrichments were measured and thus FTRs k8,7 and k7,8 were estimated. A five-compartment unit with a delay of 0.3 h VLDL was used to model TG production (Fig. 1).
The concentration of [1,1,2,3,3]-2H5-glycerol at time point t (ct) was calculated as the product of the atom percent enrichment at this time point (APEt) and the mean glycerol concentration in a particular compartment. So for VLDL1 at time point t:
The pool sizes of [1,1,2,3,3]-2H5-glycerol in VLDL1, VLDL2, and plasma i.e. q1, q2, and q7 were at each point in time calculated as the products of their respective concentrations (Ct) at each time point and plasma volume (V). So for VLDL1 at time point t:
The flux of glycerol was calculated as the product of FTR and glycerol pool size. So for VLDL1 disposal at time point t:
Average time dependent curves were constructed for [1,1,2,3,3]-2H5-glycerol in VLDL1, VLDL2, and plasma of individuals from both infusions. The model was fitted to these curves using SAAM II software (version 1.2.1, SAAM Institute University of Washington). For this, plasma volume and k0,6 were introduced as Bayesian values that were corrected for bodyweight of the subjects and the estimated FTR as given by Adiels i.e. 3.7 ± 0.75 L and 19 ± 4.0 h−1, respectively. The FTRs were calculated as described by Adiels et al. for modeling of plasma glycerol.
The obtained estimated values of the FTRs, and V, except FTRs of interest i.e. k0,1, k0,2, and k2,1 were used as Bayesian values (including an error of 20%) in fitting the model to the curves of the individual subjects. As a result, estimations of all FTRs were obtained from all subjects in both infusions including the error in the estimation.
2.8 Statistical analysis
Results are presented as mean ± SD or median (interquartile range) in Table 1. For the lactate infusion and saline day results are presented as mean ± SEM. We used SPSS software, version 16.0 (Chicago, IL, USA) and GraphPad Prism Software version 5.0 (CA, USA). Differences were tested using a Student's t-test (two-sided, equal variances, paired) when the results were normally distributed, and Mann–Whitney test otherwise. A probability value P < 0.05 was considered to confer statistical significance.
Table 1Baseline characteristics of healthy subjects.
Healthy subjects (n = 8)
82.8 ± 16.6
Body mass index (kg/m2)
24.6 ± 3.2
Alanine transaminase (U/L)
Plasma glucose (mmol/L)
3.1 ± 0.5
0.9 ± 0.2
Data are presented as mean ± SD or median [IQR], TG: triglyceride; HDLc: high density lipoprotein cholesterol; LDLc: low density lipoprotein cholesterol; VLDL: very low density lipoprotein; apo: apolipoprotein.
Baseline characteristics of the 8 participants are presented in Table 1. The participants had median plasma TG level of 0.8 [0.6–1.1] mmol/L and a mean LDL cholesterol level of 3.1 ± 0.5 mmol/L. Plasma apoB values were 0.9 ± 0.2 g/L. TG concentration in the isolated VLDL1 fraction was 0.98 ± 0.12 mmol/L and in the isolated VLDL2 was 0.67 ± 0.1 mmol/L. During the lactate clamp a 5-fold increase was achieved in plasma lactate levels during steady state (3.06 ± 0.05 mmol/L versus 0.65 ± 0.02 mmol/L during saline infusion) (Fig. 2), whereas plasma glucose concentrations during lactate infusion remained unchanged at an average of 4.88 ± 0.03 mmol/L versus 4.81 ± 0.06 mmol/L (P = 0.28) during the saline infusion. No changes were observed during both infusions periods for plasma insulin (Fig. 2C).
3.2 Triglyceride, VLDL-TG and 16:1/16:0 ratio
During the lactate clamp plasma TG levels increased from 0.95 ± 0.13 mmol/L to 1.11 ± 0.17 mmol/L (difference = 0.16 ± 0.09 mmol/L, P < 0.05) (Fig. 3A). TG content in isolated VLDL1 increased from 0.98 ± 0.06 mmol/L to 1.12 ± 0.05 mmol/L (P < 0.05), whereas TG content in isolated VLDL2 was unaffected (0.67 ± 0.08 to 0.79 ± 0.10 mmol/L) (P = 0.11). During saline infusion plasma TG levels decreased from 0.90 ± 0.18 to 0.75 ± 0.12 mmol/L (−0.15 ± 0.06 mmol/L; P < 0.05; Fig. 3A) without significant changes in TG content in VLDL1 (0.72 ± 0.1 to 0.81 ± 0.1 mmol/L; P = 0.12) and VLDL2 (0.65 ± 0.1 to 0.64 ± 0.1 mmol/L; P = 0.9). No changes in plasma FFA were observed (Fig. 3B). The 16:1/16:0 ratio in isolated VLDL1 fraction during lactate infusion increased from baseline to 1.2 ± 0.74 (P < 0.01 vs. saline infusion; Fig. 3B). Saline infusion resulted in a decrease of the 16:1/16:0 ratio in VLDL1 from baseline to end of saline infusion (change −1.5 ± 0.55; P < 0.05 vs baseline). 16:1/16:0 ratio in VLDL2 did not change during either infusion.
3.3 Hepatic VLDL-TG fluxes
The glycerol enrichment curves in plasma, VLDL1 and VLDL2 during saline and lactate infusion are shown in Figs. 4 and 5. The total turnover of labeled glycerol in the administrated pool was comparable (Fig. 4C), as was the fraction of labeled glycerol that was directly converted, probably to glycerol-phosphate i.e. 98.5% (data not shown). On average less than 1% of the glycerol-phosphate was used for plasma TG synthesis.
During lactate infusion we observed a 34% increase in glycerol-phosphate flux (k4,6) into the hepatic TG fraction as compared to saline infusion leading to an increased VLDL-TG assembly (k3,4) and an increased flux into the hepatic TG pool (k5,4) (Fig. 1). In line, there was a trend toward a higher glycerol secretion rate into VLDL1-TG (FLUX1,3) during lactate infusion compared to saline infusion (1604 μmol (827–2870) vs. 1285 μmol (505–2155), respectively; P = 0.07 Figs. 4D and 5A). The incorporation of TG directly in VLDL2 (FLUX2,3) was 18% higher during lactate infusion (median saline: 311 μmol (117–475) versus lactate: 465 μmol (42–1019)), but this difference was not statistically significant (Fig. 5A). Correspondingly, both the VLDL1-TG as well as the VLDL2-TG pool sizes were increased by 28% and 18% during lactate (Fig. 5A).
During lactate infusion we observed a significant decreased transfer of labeled glycerol from VLDL1 to VLDL2 (FLUX2,1), leading to a higher VLDL1-TG pool (saline: 804 μmol (412–1634) vs lactate 504 μmol (245–533); P < 0.01; Fig. 5A). Similarly, the clearance of VLDL2-TG (FLUX0,2) during lactate infusion decreased by 22% (saline 1158 μmol glycerol (670–2006) versus lactate: 972 μmol glycerol (546–1430)), although the difference in VLDL2-TG levels did not reach statistical significance.
In the present study we show that a 7-h l-lactate clamp in human increases, rather than decreases hepatic VLDL-TG secretion, leading to higher TG concentrations in the VLDL1 fraction. SCD1 activity in VLDL1 was also higher during lactate infusion. These opposite findings in humans compared to those obtained in rodent models emphasize that data on central regulation of hepatic TG excretion in experimental models should be interpreted with great caution in the human setting.
Lactate has been suggested to interfere with lipolysis and glucose utilization, since lactate is metabolized into pyruvate by the enzyme lactate dehydrogenase [
]. In contrast, Lam et al. elegantly demonstrated that intra-cerebroventricular infusion of either glucose or lactate resulted in an increased vagal signaling to the liver, in turn leading to decreased hepatic TG excretion in rodents [
]. Inhibition of lactate generation by central administration of oxamate, an inhibitor of the enzyme lactate dehydrogenase, attenuated the observed effects, suggesting that the hypothalamus is instrumental in the observed effects [
] and the recently proposed astrocyte–neuron lactate shuttle hypothesis suggests that lactate is even actively transferred to neurons in the brain, particularly in the hypothalamic region where fibers innervating visceral organs originate [
]. Mechanistically, SCD1 activity in the liver, an important mediator involved in regulation of hepatic VLDL-TG secretion, was found to be responsible for the observed effects of icv lactate on hepatic VLDL-TG secretion.
In view of the therapeutic potential of such CNS-pathway regulating hepatic TG-excretion, we tested the validity of this concept in humans. To minimize the chance of a false negative finding, we generated a 5-fold increase in plasma lactate levels during a period of 7 h. In spite of the marked increase and prolonged duration of the lactate clamp, we observed an increase, rather than a decrease in hepatic VLDL-TG excretion. The increase in hepatic VLDL1-TG secretion was corroborated by an increased SCD1 activity, as shown by a higher plasma VLDL1 16:1/16:0 ratio. This increase of TG flux through the VLDL pathway, however, did not result in differences in VLDL subfraction clearance. Thus our data suggest that excess lactate is directly converted into pyruvate and used as substrate into the TCA cycle leading to increased citrate synthesis which after transport to the cytoplasm induces increased FFA and hence TG production.
Several aspects need closer consideration when interpreting the results of this study. A major limitation of the present study is that our study population was relatively small due to the costly nature and laborious lab procedures. Moreover, we used healthy controls and cannot establish whether results apply to individuals with abnormal plasma lipid levels. Despite this limitation, we observed significant and coherent data showing an increase rather than a decrease in hepatic VLDL1-TG excretion following a lactate clamp. Second, we studied healthy volunteers rather than patients characterized by perturbed lipid metabolism. Thus, we cannot exclude that in patients with increased hepatic VLDL excretion, such as e.g. familial combined hyperlipidemia, an effect could have been observed. Finally, our study protocol in humans does not allow us to demonstrate that peripheral lactate infusion reached steady state increase of lactate at the level of the hypothalamus, nor could we validate that lactate clamping resulted in direct changes in hepatic vagal nerve activity. However, in our study we did generate a 5-fold increase in plasma lactate levels during a period of 7 h, which can be expected to increase lactate concentration at the level of the hypothalamus.
In summary, plasma lactate clamping in healthy volunteers during a period of 7 h results in increased VLDL-TG excretion and increased VLDL1 SCD1 activity. Our data indicate that experimental findings on central regulation of hepatic TG-excretion in rodents cannot be extrapolated to the human setting.
We do not have any conflicts of interest to declare.
We would like to acknowledge the volunteers for participating in the current study protocol. We thank D. van Harskamp, A.W.M. Schimmel and L.N. Klaaysen for excellent technical assistance.
Are triglycerides important as a risk factor for coronary disease?.