621
Clinical Science (1991)81,621-626
Peripheral triacylglycerol extraction in the fasting and
post-prandial states
JENNIFER L. POTTS, RACHEL M. FISHER, SANDY M. HUMPHREYS, SIMON W. COPPACK,
GEOFFREY F. GIBBONS* AND KEITH N. FRAYN
Sheikh Rashid Diabetes Unit and *Metabolic Research Laboratory, Radcliffe Infirmary, Oxford, U.K.
(Received 12 Februaryjl4 June 1991; accepted 27 June 1991)
SUMMARY
1. Triacylglycerol extraction by subcutaneous adipose
tissue and forearm muscle was studied in nine normal
subjects after an overnight fast and after the consumption
of a mixed meal.
2. There was an inverse correlation between the total
plasma fractional triacylglycerol extraction across the
adipose tissue and the fasting arterial plasma triacylglycerol concentration. In contrast, there was no correlation between the lower fractional triacylglycerol
extraction across the forearm muscle and the fasting
plasma triacylglycerol concentration.
3. Chylomicron-triacylglycerol concentrations
in
arterial(ized) plasma increased post-prandially and
peaked at 240-300 min. There was a comparable
increase in the very-low-density lipoprotein-triacylglycerol concentration, peaking at 300 min.
4. Clearance of chylomicron-triacylglycerol by
adipose tissue increased after the meal (P<0.05). In
contrast, the clearance of very-low-density lipoproteintriacylglycerol by adipose tissue decreased post-prandially ( P < 0.05).
5. Although there was significant uptake of chylomicron-triacylglycerol by the forearm muscle postprandially, this was less than by the adipose tissue.
Very-low-density lipoprotein-triacylglycerol was unaffected by passage through the forearm muscle at any
time.
6. We conclude that the extraction of lipoproteintriacylglycerol by human adipose tissue is important in
determining the fasting plasma triacylglycerol concentration. Chylomicron-triacylglycerol, appearing in the
plasma post-prandially, may compete with very-lowdensity lipoprotein-triacylglycerol for clearance by
adipose tissue lipoprotein lipase, and this mechanism may
explain, at least in part, the post-prandial rise in very-lowCorrespondence: Dr Keith N. Frayn, Sheikh Rashid Diabetes
Unit, Radcliffe Infirmary, Oxford OX2 6HE, U.K.
density lipoprotein-triacylglycerol. Forearm muscle, in
contrast, appears to play a much smaller role in the
extraction of plasma triacylglycerol, especially that in the
very-low-density lipoprotein fraction.
Key words: adipose tissue, chylomicrons, forearm muscle,
lipoprotein lipase, triacylglycerol, very-low-density lipoprotein.
Abbreviations: AUCo-360,area under the curve from the
start of eating to 360 min later; LPL, lipoprotein lipase;
TAG, triacylglycerol, VLDL, very-low-density lipoprotein.
INTRODUCTION
Although an elevated concentration of triacylglycerol
(TAG) in plasma is a risk factor for atherosclerosis [l], we
understand little of the mechanisms which regulate the
plasma TAG concentration. Amongst apparently normal
subjects, a wide range of fasting plasma TAG concentrations is found. These are partly explained by differing
rates of endogenous TAG secretion, but the role of differences in TAG clearance has not been fully explored [2].
TAG enters the plasma either from exogenous (dietary)
sources as chylomicron-TAG, or from endogenous
sources, as very-low-density lipoprotein (VLDL)-TAG
secreted from the liver. Most of the TAG is removed from
these lipoprotein fractions by the lipoprotein lipase (LPL)
of peripheral tissues, especially adipose tissue and skeletal
muscle. The two forms of TAG-rich lipoprotein are not
independent. After a fat-containing meal, the plasma
TAG concentration rises to a greater extent than can be
accounted for by the appearance of chylomicron-TAG
alone [3,4]. Some of this rise is accounted for by particles
containing apolipoprotein B 100, presumably of hepatic
origin [5]. Since insulin, at least in vitro, acutely inhibits
hepatic VLDL secretion [6,7], it seems probable that the
post-prandial rise in the plasma VLDL-TAG concentra-
622
J. L. Potts et al.
tion reflects decreased clearance, perhaps because of
competition, by chylomicrons appearing post-prandially,
for hydrolysis by LPL [4]. This has not been directly
investigated.
The quantitative roles of different tissues in TAG clearance also await clarification. The LPL activities of muscle
and adipose tissue are regulated in an approximately
reciprocal manner, the LPL of adipose tissue being stimulated by insulin, whereas muscle LPL is stimulated during
starvation [8]. Although this might suggest that muscle
LPL is more active in the fasting state, we have found
TAG extraction by adipose tissue to be greater than that
by forearm muscle even after an overnight fast [3,9].
We have therefore sought to elucidate some of these
aspects of TAG clearance in normolipidaemic subjects,
concentrating on the relationships between the fasting
plasma TAG concentration and peripheral TAG clearance, and the relative roles of adipose tissue and muscle
(as assessed in the forearm) in the clearance of chylomicron- and VLDL-TAG in the post-absorptive and postprandial states. Other, more 'nutritional', aspects of these
studies have been reported elsewhere [3].
METHODS
Subjects
Nine healthy subjects (four female) were studied. Their
ages ranged from 28 to 64 years and their body mass
indices from 19.3 to 26.7 kg/m2. Before the study the
subjects fasted and abstained from smoking for at least
12 h. All studies were carried out in a temperaturecontrolled room at 23°C. The studies were approved by
the Central Oxford Research Ethics Committee.
Experimental design
Cannulation of the venous drainage of the subcutaneous adipose tissue of the anterior abdominal wall was
carried out as described previously [3, lo]. A cannula was
then inserted retrogradely into an antecubital vein
draining deep forearm tissues and was kept patent by a
slow infusion of saline (150 mmol/l NaCI). A third
cannula was inserted either into a vein draining a hand
which was warmed in a box at 60-70°C to provide
arterialized samples (six subjects) or into a radial artery
(three subjects).
The subjects rested for at least 30 min before the
samples were taken. Three sets of samples were then
taken from each of the three sites at 20 min intervals. The
subjects then consumed a mixed meal, as described in
earlier studies [3].The meal had an energy content of 3.1
MJ (740 kcal) of which 47% came from carbohydrate and
41% from fat, and was eaten in 20 min. Further sets of
samples were taken at 30 min, 60 min, 90 min, 2 h, 3 h,
4 h, 5 h and 6 h after the start of the meal. Throughout
the experiment adipose tissue blood flow was measured
by the disappearance of 133Xe,and forearm blood flow
was assessed by strain-gauge plethysmography as
described previously [3].
Analytical method
Blood samples were taken into heparinized syringes
(Monovette; Sarstedt, Leicester, U.K.).
Chylomicrons were separated by layering 2 ml of
plasma under a solution of density 1.006 g/ml followed by
centrifugation for 30 min in a Beckman 50.1SW rotor
[Beckman Instruments (U.K.) Ltd, High Wycombe, Bucks,
U.K.; g, = 58 4501. The chylomicrons were otained by
removing the top layer of the tube in a Beckman tubeslicer. VLDLs were prepared either by sequential
flotation or by a density gradient technique. For the former, the chylomicron infranate was transferred to a Beckman 6 ml polyallomer bell-top tube, which was filled with
solution of density 1.006 g/ml and centrifuged for 16 h in
a Beckman 50.4 T i rotor in two concentric rings: inner
ring, g,, = 145 000; outer ring, g,,= 172 000. For density
gradient separation, 2 ml of the infranate was adjusted to
a density of 1.21 g/ml by the addition of 0.653 g of KBr.
Of this, 1.5 ml(l.815 g) was weighed into a bell-top tube
as before and was overlayed with the 1.006 g/ml density
solution. Centrifugation was then carried out in a 50.4 T i
rotor for 3 h: inner ring, g,,=227000; outer ring,
g,,= 269 000. In both cases the VLDLs were obtained by
removing the top layer of the tube by slicing.
TAG was measured in plasma, chylomicrons and
VLDLs by an enzymic colorimetric method with correction for free glycerol, adapted to an IL FLS Multistat 111
Micro-Centrifugal Analyser [ 111.
In 10 randomly selected pre- and post-prandial
samples of VLDLs and chylomicrons, apolipoproteins
B48 and BlOO were separated by SDS/PAGE. In fasting
samples there was a trace of apolipoprotein BlOO in the
chylomicron fraction, but there was no apolipoprotein
B48 in the VLDL fraction. In post-prandial samples only
apolipoprotein B48 was found in the chylomicron
fraction and only apolipoprotein BlOO in the VLDL
fraction.
Calculations and statistical analysis
For statistical analyses, the results from the three
samples taken before eating were averaged to give a 'mean
basal' value. Fractional extraction of TAG was calculated
as the arteriovenous difference as a percentage of the
arterial concentration. TAG flux was calculated as the
arteriovenous difference multiplied by the plasma flow,
and TAG clearance was determined as the product of
fractional extraction (as a fraction, not percentage) and
plasma flow. Using the trapezoid rule, areas under the
curves for the various measures of TAG removal were
calculated, from the start of eating to 360 min later
(AUCo-360). 'Incremental' areas were calculated by
subtracting the mean baseline value extrapolated over
360 min to reflect changes occurring post-prandially.
Results are shown as means k SEM.
RESULTS
The subjects had widely differing fasting arterial plasma
TAG concentrations (range 380-1810 pmol/l) and an
Peripheral triacylglycerol extraction
inverse relationship was observed ( r = - 0.815, PCO.01;
Fig. 1) between fractional TAG extraction across adipose
tissue and the fasting arterial plasma TAG concentration.
In contrast, there was no significant correlation between
the lower fractional TAG extraction across the forearm
muscle and the fasting plasma TAG concentration.
Fractional TAG extraction from the VLDL fraction by the
adipose tissue in the basal state was also significantly
greater than that by the forearm muscle ( P < 0.001).
The rise in total plasma TAG concentration, which
peaked at 240-300 min after the meal, was described in
our earlier paper [3]. This rise was associated with significant increases in both the arterial chylomicron-TAG concentration ( P < O . O l , incremental AUC,_,,, = 43 k 11
2000 0.
1600 0
.
1200 800 400
d
-
\ 0
o00
623
mmol l-' min) and the arterial VLDL-TAG concentration
( P < 0.05, incremental AUCo-360= 61 k 21 mmol 1- min)
(Fig. 2). The magnitudes of the rises in the arterial
chylomicron-TAG and VLDL-TAG concentrations were
not significantly different.
The arteriovenous difference for the removal of
chylomicron-TAG by adipose tissue (Fig. 3) was small and
non-significant after an overnight fast, but increased
significantly after the meal (Table 1).The arteriovenous
difference for removal of chylomicron-TAG by forearm
muscle (Fig. 3) was also significant, although variable, and
increased after the meal (Table 1).
In the basal state the arteriovenous difference across
adipose tissue for VLDL-TAG (Fig. 3) was significant
( P <0.01, mean basal arteriovenous difference= 51 k 14
,umol/l), but this declined after eating (Table 1).Concentrations of VLDL-TAG in forearm venous plasma were
very similar to those in arterial(ized) plasma and are not
shown in Fig. 3 for clarity; the arteriovenous difference
was not significant at any time.
These relationships were more clearly seen in terms of
fractional extraction. While the fractional extraction of
0
H
0
-5
0
5
10
15
20
Fractional TAG extraction (70)
Fig. 1. Correlation between plasma TAG concentration
and percentage fractional TAG extraction across adipose
tissue (a)and forearm muscle ( 0 )both measured after an
overnight fast. Results are for nine subjects. The correlation between plasma TAG concentration and fractional
TAG extraction was significant in the adipose tissue
( r = -0.815, P < O . O l ) , but not in the forearm muscle.
T
800 -
T I
700 -
500 400 -
600
-40 0
60
120
180
240
300
360
Time (min)
I
I
I
I
I
I
60
120
180
240
300
360
Fig. 3. Plasma concentrations of chylomicron-TAG ( a )
and VLDL-TAG ( b )before and after a mixed meal (0-20
min). Results are m e a n s k ~for
~ ~nine subjects. A ,
Fig. 2. Plasma chylomicron-TAG (A ) and VLDL-TAG
( A ) concentrations in arterial(ized)blood before and after
a mixed meal (0-20 min). Results are m e a n s k s ~for
~
nine subjects.
Arterial(ized) blood; 0 , forearm venous blood; adipose
venous blood. Some error bars are omitted for clarity, as
are forearm venous VLDL-TAG concentrations, which
were very similar to the arterial(ized)concentrations.
-40 0
Time (min)
.,
624
J. L. Potts et al.
Table 1. AUC,_,,, for TAG extraction by adipose tissue and forearm muscle in nine normal subjects
Results are expressed as m e a n s k ~ For
~ ~ ‘Incremental’
.
areas the mean baseline value (extrapolated over 360 min) has
been subtracted; these areas therefore represent changes occurring post-prandially. Statistical significance (paired t-test):
*I-’< 0.05, **I)< 0.01 for the difference from zero; t P < 0.05, tip< 0.01 for the difference between chylomicron- and
VLDL-TAG; $P< 0.05 for the difference between adipose tissue and forearm muscle.
Incremental AUC,,.3,,,
Total AUCII-JOO
Chylomicron-TAG
Arteriovenous difference (mmol I - ’ min)
Adipose tissue
Forearm muscle
lo3X Fractional extraction (%/min)
Adipose tissue
Forearm muscle
13.8 f2.5**
7.6 f2.7*$
8.6 f 1.8**
3.2 f0.9**$
VLDL-TAG
8.8 f5.4
- 0.7 f 3.6
Chylomicron-TAG
13.0+2.5**
7.1 f2.6*$
1.4 k 0.6ff
- 0.2 0.7f$
+
7.2 k 4.5
1.9 f 2.1
VLDL-TAG
- 11.6 k 7.lf
-4.1 f 3 . 9
- 3.4 f0.8**t
- 1.5 zk 0.9$
TAG flux (pmo1/100 ml of tissue)
Adipose tissue
Forearm muscle
TAG clearance (1/100ml of tissue)
Adipose tissue
Forearm muscle
24.1 f4.6**
10.2 f 3.6*$
- 2.6
15.3 zk 8.3
f4.4$
22.3 f 3.6**
9.0 f4.0$
18.2 f5.4**
4.4 f 1.8*$
3.2 k 1.5f
-0.8kl.0:
12.7 f5.5*
1.2 f4.4t
chylomicron-TAG across adipose tissue increased postprandially, that of VLDL-TAG decreased (Fig. 4).
Fractional extraction of chylomicron-TAG across the
forearm muscle tended to rise post-prandially (not
significant), but there was little extraction of VLDL-TAG
by the forearm muscle at any time.
The absolute flux and metabolic clearance of TAG are
shown in Table 1. These confirmed the picture described
above. Chylomicron-TAG uptake by adipose tissue
increased post-prandially, whereas VLDL-TAG clearance
decreased. The forearm muscle was consistently less
active in TAG removal than the adipose tissue.
DISCUSSION
LPL, the enzyme responsible for clearance of circulating
TAG, was first recognized in adipose tissue and heart [ 121,
but was later shown to be present with lesser activity (per
unit weight of tissue) in skeletal muscle [ 131. Its activity in
different tissues has usually been studied in biopsies, with
the attendant difficulties of extrapolation to the physiological state. These difficulties are particularly marked in
the case of LPL because only a proportion of the enzyme
extractable from tissues is in the physiologically active
form [ 14, 151. In a few studies, arteriovenous differences
have been measured across the forearm [16-181 or the
thigh [19] for extraction of endogenous or exogenous
TAG. The measurement of exogenous TAG extraction
was made, however, by nephelometry, which might not be
a valid technique for this application: it seems more likely
that particles shrink rather than disappear completely
during one passage through an LPL-containing tissue [20,
211, so the fraction of particles shrinking below the limit
of light-scattering size might not accurately reflect the
proportion of TAG extracted.
- 15.6 k 14.7
- 8.9 f6.0
- 4.4 k 1.5*t
-2.8f 1.4
Measurement of arteriovenous differences by a specific
method (with a correction for free glycerol), as in the
present study, appears to offer a means for the direct
estimation of the physiological activity of LPL in tissues in
vivo [3]. This study has confirmed our earlier finding that
the fractional extraction of plasma TAG by adipose tissue
is considerably greater than that by forearm muscle [3,9].
The importance of TAG clearance by adipose tissue in
determining the plasma TAG concentration is shown by
the relationship between TAG fractional extraction across
adipose tissue and the arterial(ized) plasma TAG concentration in the fasting state (Fig. 4). No such relationship
was observed between the much lower fractional extraction across the forearm muscle, and the fasting plasma
TAG concentration. It should be noted that the human
forearm muscle contains a mixture of red and white
muscle fibres as do other human muscles [22].It is known
from animal experiments that LPL activity is greater in
red fibres than in white [ 131. Our results are therefore not
necessarily representative of the total muscle mass.
By the separation of lipoprotein fractions, we have
shown clearly that the rise in plasma TAG concentration
post-prandially reflects increases in both plasma chylomicron- and VLDL-TAG concentrations. Our studies were
not able directly to address the question of hepatic VLDL
secretion. We were able, however, to test the hypothesis
that the post-prandial rise in the plasma VLDL-TAG concentration represents at least in part the result of
decreased removal, perhaps through competition for LPL
by chylomicron-TAG. The metabolic clearance of VLDLTAG by the adipose tissue decreased somewhat postprandially, whereas that of chylomicron-TAG rose
considerably, presumably reflecting the post-prandial
stimulation of LPL activity by insulin [3, 231. Although
forearm muscle cleared chylomicron-TAG, showing that
chylomicron-TAG is a substrate for muscle LPL, the
Peripheral triacylglycerol extraction
625
Miss M. L. Clark for technical assistance, and Dr T. D. R.
Hockaday for advice and encouragement. The Sheikh
Rashid Diabetes Unit is supported by the Oxford
Diabetes Trust. G.F.G. is a member of the MRC External
Scientific Staff.
8
v
REFERENCES
30
t
0
60
120
180
240
300
360
Time (min)
.,
Fig. 4. Fractional extraction of TAG across adipose
tissue ( a )and forearm muscle ( b )before and after a mixed
meal (0-20 min).
0 , chylomicron-TAG; 0,0 , VLDLTAG. Results are mean sf^^^ for nine subjects. The
values for fractional extraction of TAG in the postprandial period were averaged over adjacent pairs of
samples in order to increase the precision of
measurement and are shown at the mean time of the two
samples.
extraction was less than that by adipose tissue. VLDLTAG, however, was almost completely unaffected by
passage through the forearm muscle, suggesting that
VLDL-TAG is not a good substrate for mixed muscle
LPL, as in the forearm. This is not in agreement with
observations on VLDL-TAG extraction by the trained
human thigh [ 193, although the femoral venous drainage
taken in that study to represent drainage from thigh
muscle may include venous drainage from subcutaneous
adipose tissue in addition. It does, however, tally with
various observations which suggest that VLDL-TAG is
not a substrate for muscle energy production during
exercise, whereas chylomicron-TAG may make some contribution [24].
ACKNOWLEDGMENTS
J.L.P. held a Glaxo Group Research studentship. We
thank the British Heart Foundation for financial support,
1. Castelli, W.P. The triglyceride issue: a view from
Framingham. Am. Heart J. 1986; 112,432-7.
2. Lewis, B. Hyperlipidaemia. In: Cohen, R.D., Lewis, B.,
Alberti, K.G.M.M. & Denham, A.M., eds. The metabolic
and molecular basis of acquired disease. London: Baillitre
Tindall, 1990: vol. 1,860-920.
3. Coppack, S.W., Fisher, R.M., Gibbons, G.F. et al. Postprandial substrate deposition in human forearm and
adipose tissues in viva Clin. Sci. 1990; 79,339-8.
4. Lewis, G.F.O., Meara, N.M., Soltys, P.A. et al. Postprandial
lipoprotein metabolism in normal and obese subjects:
comparison after the vitamin A fat-loading test. J. Clin.
Endocrinol. Metab. 1990; 71,1041-50.
5. Cohn, J.S., McNamara, J.R., Cohn, S.D., Ordovas, J.M. &
Schaefer, E.J. Postprandial plasma lipoprotein changes in
human subjects of different ages. J. Lipid Res. 1988; 29,
469-79.
6. Durrington, P.N., Newton, R.S., Weinstein, D.B. &
Steinberg, D. Effects of insulin and glucose on very low
density lipoprotein triglyceride secretion by cultured rat
hepatocytes. J. Clin. Invest. 1982; 70,63-73.
7. Bartlett, S.M. & Gibbons, G.F. Short- and longer-term
regulation of very-low-density lipoprotein secretion by
insulin, dexamethasone and lipogenic substrates in cultured
hepatocytes. A biphasic effect of insulin. Biochem. J. 1988;
249,37-43.
8. Cryer, A. Tissue lipoprotein lipase activity and its action in
lipoprotein metabolism. Int. J. Biochem. 1981; 13,525-41.
9. Coppack, S.W., Frayn, K.N. & Humphreys, S.M. Plasma
triacylglycerol extraction in human adipose tissue in viva
effects of glucose ingestion and insulin infusion. Eur. J. Clin.
Nutr. 1989; 43,493-6.
10. Frayn, K.N., Coppack, S.W., Humphreys, S.M. & Whyte,
P.L. Metabolic characteristics of human adipose tissue in
viva Clin. Sci. 1989; 76,509-16.
11. Humphreys, S.M., Fisher, R.M. & Frayn, K.N. Micromethod for measurement of sub-nanomole amounts of
triacylglycerol. Ann. Clin. Biochem. 1990; 27,597-8.
12. Robinson, D.S. Lipoprotein lipase - past, present and
future. In: Borensztajn, J., ed. Lipoprotein lipase. Chicago:
Evener, 1987: 1-13.
13. Borensztajn, J., ed. Lipoprotein lipase. Chicago: Evener,
1987: 133-48.
14. Pradines-Figukres, A., Vannier, C. & Ailhaud, G. Lipoprotein lipase stored in adipocytes and muscle cells is a
cryptic enzyme. J. Lipid Res. 1990; 31, 1467-76.
15. Nilsson-Ehle, P. Measurements of lipoprotein lipase activity.
In: Borensztajn, J., ed. Lipoprotein lipase. Chicago:
Evener, 1987: 59-77.
16 Kaijser, L. & Rossner, S. Removal of exogenous triglycerides in human forearm muscle and subcutaneous
tissue. Acta Med. Scand. 1975; 197,289-94.
17 Rossner, S., Eklund, B., Kaijser, L., Olsson, A.G. &
Walldius, G. Removal of exogenous plasma triglycerides in
forearm muscle and subcutaneous tissue of hyper- and
normotriglyceridaemic men. Eur. J. Clin. Invest. 1976; 6,
299-305.
18. Hailer, S., Jauch, K.-W., Gunther, B., Wolfram, G., Zollner,
N. & Heberer, G. Arterial deep venous difference of
lipoproteins in skeletal muscle of patients in postoperative
state: effects of medium chain triglyceride emulsion. J.
Parenter. Enteral Nutr. 1988; 12,377-81.
626
J. L. Potts et al.
19. Kiens, B. & Lithell, H. Lipoprotein metabolism influenced
by training-induced changes in human skeletal muscle. J.
Clin. Invest. 1989; 83,558-64.
20. Eisenberg, S. Regulation of the ApoBlOO cascade. In:
Suckling, K.E. & Groot, P.H.E. Hyperlipidaemia and
atherosclerosis. London: Academic Press, 1988: 65-72.
21. Gibbons, G.F. Assembly and secretion of hepatic very-lowdensity lipoprotein. Biochem. J. 1990; 268, 1-1 3.
22. Johnson, M.A., Polgar, J., Weightman, D. & Appleton, D.
Data on the distribution of fibres types in thirty-six human
muscles. An autopsy study. J. Neurol. Sci. 1973; 18,
111-29.
23. Sadur, C.N. & Eckel, R.H. Insulin stimulation of adipose
tissue lipoprotein lipase. Use of the euglycemic clamp
technique. J. Clin. Invest. 1982; 69, 1 1 19-25.
24. Nikkila, E.A. Role of lipoprotein lipase in metabolic adaptation to exercise and training. In: Borensztajn, J., ed. Lipoprotein lipase. Chicago: Evener, 1987: 187-99.