Download Effect of increased free fatty acid supply on glucose metabolism and

Clinical Science ( I 992) 82, 2 19-226 (Printed in Great Britain)
219
Effect of increased free fatty acid supply on glucose
metabolism and skeletal muscle glycogen synthase activity
in normal man
A. B. JOHNSON, M. ARGYRAKI, J. C. THOW, B. G. COOPER, G.FULCHER and R. TAYLOR
Department of Medicine, Royal Victoria Infirmary, Newcastle upon Tyne, U.K.
(Received 6 Junell6 October 1991; accepted 21 October 1991)
1. Experimental elevation of plasma non-esterified fatty
acid concentrations has been postulated to decrease
insulin-stimulated glucose oxidation and storage rates.
Possible mechanisms were examined by measuring
skeletal muscle glycogen synthase activity and muscle
glycogen content before and during hyperinsulinaemia
while fasting plasma non-esterified fatty acid levels were
maintained.
2. Fasting plasma non-esterified fatty acid levels were
maintained in seven healthy male subjects by infusion of
20% (w/v) Intralipid (1 ml/min) for 120 min before and
during a 240 min hyperinsulinaemic euglycaemic clamp
(100 m-units h - ' kg-') combined with indirect calorimetry. On the control day, 0.154 mol/l NaCl was infused.
Vastus lateralis muscle biopsy was performed before and
at the end of the insulin infusion.
3. On the Intralipid study day serum triacylglycerol
(2.24 k 0.20 versus 0.67 f 0.10 mmol/l), plasma nonesterified fatty acid (395 f 13 versus 51 f 1 pmol/l),
blood glycerol (152 k 2 versus 11k 1 pmol/l) and blood
3-hydroxybutyrate clamp levels [mean (95% confidence
interval)] I81 (64-104) versus 4 (3-5) pmol/l] were all
significantly higher (all P<O.OOl) than on the control
study day. Lipid oxidation rates were also elevated
(1.07 f 0 .0 7 versus 0.27 f0.08 mg min-' kg-',
P< 0.001). During the clamp with Intralipid infusion,
insulin-stimulated whole-body glucose disposal decreased
by 28% (from 8.53 k 0.77 to 6.17 f 0.71 mg min-' kg-',
P<0.005). This was the result of a 48% decrease in
glucose oxidation (3.77k0.32 to 1.95 k0.21 mg min-'
kg-', P<O.OOl), with no significant change in nonoxidative glucose disposal (4.76 k 0.49 to 4.22 f0.57 mg
min-' kg-', not significant).
4. Basal and insulin-stimulated glycogen synthase
activities (13.1 f 1.9 versus 11.4 f 2.3% and 30.8 k 2.3
versus 27.6 f4.5%, respectively) were unaffected by the
increased plasma non-esterified fatty acid levels.
Similarly, basal (36.1 f2.7 versus 37.2 f 1.4 pmol/g)
and stimulated (40.0 k 0.6 versus 37.6 k 4.4 pmol/g)
muscle glycogen levels were unaltered. Insulin-stimulated
hexokinase activity was also not affected (0.52 f0.08
versus 0.60 f 0.08 units/g wet weight).
5. Maintenance of plasma non-esterified fatty acid levels
at fasting values resulted in an increase in lipid oxidation
and was associated with a decrease in insulin-stimulated
whole-body glucose uptake and glucose oxidation rates,
but no change in non-oxidative glucose disposal.
Increased plasma non-esterified fatty acid levels did not
appear to have a direct inhibitory effect on glycogen
synthase activity or storage of glucose as glycogen at
these insulin levels.
INTRODUCTION
It is now well established that peripheral tissue insulin
insensitivity is a characteristic feature of type 2 diabetes,
and the skeletal muscle is the main site of resistance [l].
Decreased insulin-stimulated glucose uptake, glucose
oxidation and glucose storage rates have all been well
documented in such patients [2, 31. The anti-lipolytic
action of insulin is also impaired in type 2 diabetes, resulting in elevated plasma non-esterified fatty acid (NEFA)
levels and increased lipid oxidation rates [4-61. Randle et
al. [7] suggested that the high plasma NEFA levels were
causally related to the abnormalities of carbohydrate
metabolism in type 2 diabetes via the glucose-fatty acid
cycle. A n essential feature of their proposal was that
increased NEFA oxidation imposed restrictions upon the
metabolism of glucose by heart and skeletal muscle,
thereby leading to insulin insensitivity. It is possible therefore that elevated plasma NEFA levels may explain, at
least in part, the decreased insulin-stimulated glucose
metabolism observed in type 2 diabetic patients.
A number of studies have demonstrated that the
glucose-fatty acid cycle operates in normal man, using an
infusion of Intralipid and/or heparin to elevate plasma
NEFA levels. The high plasma NEFA levels were
associated with decreased rates of insulin-stimulated
whole-body glucose uptake and glucose oxidation [8-151.
Key words: glucose uptake, glycogen synthase, insulin action, lipid oxidation, skeletal muscle.
Abbreviations: G6P. glucose 6-phosphate; GS, glycogen synthase; GS,, physiologically active glycogen synthase; GS,, total glycogen synthase activity; NEFA, non-esterified
fatty acids.
Correspondence: Dr Andrew Johnson, Department of Medicine, University of Newcastle upon Tyne, The Medical School, Framlington Place, Newcastle upon Tyne
NE2 4HH, U.K.
220
A. B. Johnsone t at.
Decreased glucose storage was also observed in a number
of these studies [8, 9, 151. An effect on glucose storage
was not originally proposed by Randle et al. [7], who
suggested that increased NEFA metabolism might favour
the conversion of glucose to glycogen rather than to
lactate. This discrepancy could be explained by one of
two possible mechanisms. First, a direct effect of fatty
acids on liver glycogen synthase (GS, E C 2.4.1.11) has
been demonstrated, causing it to dissociate into subunits
[16]. A similar effect of fatty acids on skeletal muscle GS
would be expected to result in decreased GS activity and
thus decrease glycogen synthesis. Alternatively, increased
lipid oxidation may lead to increased intracellular acetylCoA and citrate levels. These metabolites then inhibit
earlier steps of glucose metabolism, eventually leading to
decreased glucose transport [17, 181 into muscle and a
reduction in both glucose oxidation and storage.
The present study was undertaken to determine the
effect of maintaining plasma NEFA levels at fasting values
in healthy man upon insulin-stimulated glucose metabolism, skeletal muscle glycogen synthesis and skeletal
muscle GS activity, both basally and after insulin stimulation.
MATERIALS AND METHODS
Subjects
Seven healthy male subjects were studied. Their age
was 4 2 5 5 years and their mean body mass index was
25.25 1.5 kg/m2 ( m e a n s k s ~ ~No
) . subject had a family
history of diabetes mellitus, was taking any medications or
undertook regular exercise. All were consuming a weightmaintaining diet before the study that contained at least
200 g of carbohydrate/day. The protocol was approved
by the Ethical Committee of the Newcastle upon Tyne
Health Authority, and written informed consent was
obtained from each subject before the study.
Experimental protocol (Fig. I)
All subjects were admitted to the metabolic unit on the
evening before the study, fasted overnight (12-14 h) and
were maintained on strict bed rest. Each subject was
studied on two occasions in random order. On the
morning of each study, two teflon cannulae (Venflon,
Viggo, Helsingborg, Sweden) were inserted under local
anaesthesia. The first was inserted into an antecubital vein
for the infusion of all test substances, and the second into
a contralateral dorsal hand vein in a retrograde fashion
for intermittent blood sampling. This hand was then
placed in a temperature-regulated box. Basal samples
were taken at least 30 rnin after cannulation.
At time - 120 rnin an infusion of either 0.154 mol/l
NaCl in water (control study day) or 20% (w/v) Intralipid
(Intralipid study day) was begun at a rate of 1 ml/min.
Between time - 30 rnin and time 0 min of this equilibration period, a needle biopsy (UCH-type needle) of the
vastus lateralis muscle was performed under local
anaesthesia in five of the subjects. The muscle tissue
-
Time (min)
-60
-120
0 1 5 4 m o l / l N.~
aCl'
I
or 20% (wlv) lntralipid
(I rnllmin)
Insulin infusion
(m-units h - l kg-')
h
'
Variable 20% (wlv)
glucose infusion
Muscle biopsy
60
0
'
I
'
I
180
120
'
'
I
240
'
'
q
Indirect calorimetry
Collection of
metabolic data
Fig. I. Protocol of the study
obtained by this technique (50-150 mg) was immediately
frozen in liquid nitrogen and was stored at - 70°C until
assay. Once the basal biopsy had been taken, an intravenous insulin infusion (Human Actrapid; Novo,
Basingstoke, Hants, U.K.) was commenced at time 0 rnin
at a rate of 100 m-units h-' kg-l. The blood glucose
concentration was allowed to fall to 4.0 mmol/l and was
maintained at that level by measuring the blood glucose
concentration and adjusting the rate of infusion of 20%
(w/v) glucose in water at 5 min intervals. The duration of
the insulin infusion was 240 min. A repeat biopsy of the
contralateral vastus lateralis muscle was performed
during the last 30 rnin of the insulin infusion.
Hepatic glucose production was not measured, as this
has been shown to be completely suppressed in normal
subjects during comparable studies [ 151. Throughout the
study blood samples were taken for determination of
serum insulin, plasma NEFA and blood intermediary
metabolite concentrations. Samples were taken at 30 min
intervals, except during the basal measurement period,
and from 180 to 210 min, when samples were taken at 15
min intervals. A portion was centrifuged immediately and
the plasma was aspirated and frozen for subsequent
measurement of NEFA levels.
Indirect calorimetry
During the basal ( - 150 to - 120 rnin) and equilibration
( - 60 to - 30 min) periods and during the insulin infusion
(90-120 and 180-210 min), oxygen uptake and carbon
dioxide production were measured by using a Datex
Deltatrac Metabolic Monitor (Datex, Helsinki, Finland).
This uses a computerized open-circuit system to measure
gas exchange through a clear plastic 25 litre canopy
placed over the subject's head. Flow through the hood
was calibrated by the gas injection method and quantitative alcohol burning. Carbon dioxide concentration was
measured by an infra-red analyser, and oxygen concentration by a fast differential paramagnetic sensor. The
Deltatrac was calibrated with a gravimetric carbon
Fatty acid supply and glucose metabolism
dioxide and oxygen mixture before the start of each
measurement period. The monitor has a precision of
3.1% for oxygen consumption and 2.6% for carbon
dioxide production. Urine was collected from - 120 to
240 min for the determination of total nitrogen excretion.
Calculations
The whole-body glucose disposal rate was measured as
the glucose infusion rate (mg min-' kg-I) from 90 to 120
and 180 to 210 min of the insulin infusion. Substrate
oxidation rates were calculated over the same periods
from the respiratory exchange measurements and urinary
nitrogen excretion according to standard stoichiometric
equations [ 191. Non-oxidative glucose metabolism was
calculated by subtracting the glucose oxidation rate from
the total body glucose uptake rate during the clamp.
22 I
increase in absorbance at 340 nm due to NADPH formation.
Other analyses
All blood glucose concentrations were measured by the
glucose oxidase method (Yellow Springs Glucose
Analyser; Clandon Scientific, London, U.K.). Serum
insulin concentration was measured by a double-antibody
radioimmunoassay [22]. Serum C-peptide concentration
was measured by a radioimmunoassay kit [23] (Novo
Industri, Bagsvaerd, Denmark). Blood for intermediary
metabolites was immediately deproteinized in 0.5 mol/l
perchloric acid, and the extract was stored at - 20°C until
analysis by automated fluorimetric methods [24]. Plasma
NEFA level was measured enzymically by using a centrifugal analyser [25]. Serum triacylglycerol concentration
was measured using standard enzymic methods. Urinary
nitrogen was measured by the Kjeldahl's method [26].
Muscle GS and glycogen assay
Statistics
Muscle samples were weighed while still frozen and
were homogenized with a Polytron homogenizer (Polytron
Kinematica, Lucerne, Switzerland) in an ice-cold Tris/
HC1 buffer (1 ml of buffer/50 mg of frozen muscle), pH
7.3 containing 100 mmol/l KCl, 2 mmol/l EDTA, 2
mmol/l dithiothreitol, 25 mmol/l KF and 0.1% (v/v)
Triton X-100. Rabbit liver glycogen (2.0 g/l; Boehringer
Mannheim, Germany) was added to 800 pl of the
homogenate and mixed well. This was then centrifuged at
10000 g for 30 s in a microcentrifuge. The supernatant
was used for the assay of GS activity. The rest of the
homogenate was refrozen in liquid nitrogen and was
stored at - 70°C for the later assay of hexokinase activity
and glycogen content.
GS activity was assayed by measuring the incorporation of UDP-~-[U'~C]glucose
into glycogen at 30°C using
the modified method of Golden et al. [20]. Physiologically
active GS (GS,) was assayed in the absence of glucose
6-phosphate (G6P), whereas total GS activity (GS,) was
measured at a saturating concentration of G 6 P (6.7
mmol/l). Enzyme activity is expressed in units/g wet
weight, where 1 unit is the amount of enzyme activity
required to convert 1 pmol of substrate into glycogen/min
at 30°C. The percentage fractional activity is calculated as
the activity GS, divided by GS, multiplied by 100.
For the glycogen assay, 50 pl of muscle homogenate
was incubated with 100 p l of 0.2 mmol/l acetate buffer
containing amyloglucosidase (0.1 mg/ml; Sigma) at 42°C
for 90 min. Appropriate sample blanks were also run. The
tubes were immediately spun for 30 s at 10 000 g and the
supernatant glucose content was assayed by the hexokinase method using a centrifugal analyser [2 11. Glycogen
content is expressed as pmol of glucose/g muscle wet
weight.
Hexokinase activity was assayed using a modification
of the method of Easterby & Quadri [21]. The G 6 P
produced in the reaction was coupled to G 6 P dehydrogenase. The reaction was followed by measuring the
All results are expressed as means k SEM unless otherwise indicated. The 3-hydroxybutyrate data was logtransformed for statistical analysis. Statistical analyses
were performed by using Student's paired t-tests and
analysis of variance as appropriate. The sample size and
power calculations were performed by the methods of
Armitage &Berry [27].
RESULTS
Blood glucose and serum insulin concentrations
Fasting blood glucose concentrations were comparable
on both Intralipid and control study days, and were unaltered by the infusion of Intralipid (4.8k0.1 versus
4.8 k 0.1 mmol/l). Fasting serum insulin concentrations
were also comparable on both study days and were
similarly unaffected by the Intralipid infusion (4.5 k 0.9
versus 4.4 k 1.0 m-units/l). The blood glucose concentrations (meansfsD) for the periods 90-120 min and
180-210 min were 4.0k0.1 and 4.0k0.1 mmol/l on
both study days, respectively. The coefficient of variation
of the blood glucose concentration was 3.6 k 0.5% and
2.9k0.5% from 90 to 120 min on the Intralipid and
control study days, respectively, and 2.5 k 0.4% and
2.4 k 0.4%, respectively, from 180 to 210 min. After the
start of the insulin infusion, serum insulin concentrations
were comparable from 30 to 240 min (118k11 and
112 k 9 m-units/l, respectively).
Plasma NEFA, serum triacylglycerol, blood glycerol,
3-hydroxybutyrate and intermediary metabolite concentrations
The plasma NEFA, blood glycerol and 3-hydroxybutyrate and serum triacylglycerol concentrations are
shown in Fig. 2. Fasting plasma levels of NEFA (408 f45
versus 3 6 9 f 4 8 pmolll), blood levels of glycerol
A. B. Johnson e t al.
222
0
4
control study day plasma NEFA concentrations were
suppressed completely during the insulin infusion to
50 -t- 1 pmolll. Plasma NEFA concentrations from 30 to
240 min were significantly different on the Intralipid and
control study days (P<O.001). On the Intralipid study
day blood glycerol concentrations increased to 149 f 14
pmol/l at time 0 min and then reached a steady-state concentration of 152 k 2 pmol/l, whereas on the control day
levels declined to 11+. 1 pmol/l. The difference between
study days was again significant ( P <0.001). Blood
3-hydroxybutyrate levels were also significantly different
from 30 to 240 rnin on the 2 study days ( P < O . O O l ) .
Serum triacylglycerol concentrations rose to 2.04 f0.20
mmol/l at time 0 min and then reached a steady-state concentration of 2.24 k 0.20 mmol/l from 30 to 240 rnin. On
the control study day, serum triacylglycerol concentrations decreased from 0 to GO min and then reached a
steady-state concentration of 0.62 f0.01 mmol/l during
the insulin infusion ( P <0.001 versus the Intralipid study
day).
Plasma lactate, pyruvate and alanine concentrations
were comparable in the fasting state and at 0 min in both
studies (Table 1). During the insulin infusion plasma
concentrations of lactate and alanine were lower on the
Intralipid study day (both P<0.001) (Table l ) , but the
difference in plasma pyruvate levels did not reach statistical significance (Table 1).
Whole-body glucose uptake and substrate metabolism
I-
o
-120
t
-60
0
60
120
180
240
Time (min)
Fig. 2. Time course of plasma NEFA (a), blood 3-hydroxybuty.
rate (b), blood glycerol (c) and serum triacylglycerol (d)
concentrations basally, during the equilibration period and during
the insulin infusion on the control study day ( 0 ) and during the
lntralipid infusion (0).Values are means +SEM.
Data on insulin-mediated whole-body glucose disposal,
oxidative glucose disposal, non-oxidative glucose disposal
and lipid oxidation during the Intralipid and saline infusions are shown in Fig. 3. During the equilibration
period, lipid oxidation was unchanged by the infusion of
Intralipid (0.94 5 0.08 versus 0.94 k 0.07 mg min-' kg-I).
Similarly, basal glucose oxidation was unaltered
(1.58 f 0.23 versus 1.58 f0.14 mg min-l kg-I). However, during the insulin infusion, lipid oxidation was significantly higher from 180 to 210 min on the Intralipid
study day (1.07 k 0.07 versus 0.27 f0.08 mg rnin-' kg- I,
P< 0.001). This was accompanied by a significant
decrease in the rate of insulin-stimulated whole-body
glucose uptake to 6.17 f 0.71 compared with 8.53 5 0.77
mg min-l kg-' (P<O.OOl). Over this period glucose
oxidation was also decreased to 1.95 f 0.21 compared
with 3.77 f 0.32 mg min-l kg- (P<O.OOl), but there was
no significant change in the rate of non-oxidative glucose
disposal (4.22 0.57 versus 4.76 f 0.49 mg min-I kg-I).
On the control study day there were significant
increases in the rates of whole-body glucose uptake
( P < 0.005) and oxidative ( P < 0.02) and non-oxidative
( P < 0.05) glucose metabolism and a significant decrease
in lipid oxidation ( P < 0.02) after increased time exposure
to insulin (180 to 210 min versus 90 to 120 min). On the
Intralipid study day there were also significant increases
in the rates of whole-body glucose uptake ( P <0.001) and
non-oxidative glucose metabolism ( P < 0.0 1), but no
changes in glucose or lipid oxidation rates with time.
'
*
(33.4 5 2.1 versus 37.4 f 3.2 pmol/l), blood levels of
3-hydroxybutyrate [mean (95% confidence intervals)] [47
(21-108) versus 40 (24-66) pmol/l] and serum levels of
triacylglycerol (0.86 k 0.06 versus 0.95 k 0.05 mmol/l)
were similar on the Intralipid and control study days.
After the start of the Intralipid infusion, plasma NEFA
concentrations rose to 821 5 94 pmol/l at time 0 min on
the Intralipid study day and then reached a steady-state
level of 395 f 13 pmol/l from 30 to 240 min. On the
Fatty acid supply and glucose metabolism
Table I , Blood lactate, pyruvate and alanine concentrations
during the basal period (fasting), before the start of the insulin
infusion (0 min) and during the insulin infusion (180-210 min).
Values are meansf sw.Statistical significance: *P (0.00 I compared with
Table 2. Basal and insulin-stimulated GS and hexokinase activity
and skeletal muscle glycogen content on the lntralipid and
control study days. Values are m e a n s k w .
lntralipid
study day
Control
study day
GS activity (%)
Basal
Stimulated
13.1 f 1.9
30.8k2.3
I1.4f2.3
27.6k4.5
Glycogen content (pmollg)
Basal
Stimulated
36.I f2.7
40.0 50.6
37.2 f I .4
37.6t4.4
Hexokinase activity (unitslg)
Basal
Stimulated
0.60t0.07
0.55k0.06
0.60k0.06
0.511.0.07
control study day.
Concn. (pmoVl)
Lactate
Fasting
0 min
180-210 min
Pyruvate
Fasting
0 min
180-2I0 min
Alanine
Fasting
0 min
180-210 min
lntralipid
study day
Control
study day
562k 86
532 f64
645 k 38*
561 1.76
532f90
833f71
54.1 t8.2
41.6 f4.8
53.8k 3.5
55.3k7.3
44.6k 7.8
62.8? 4.0
252k22
224f 19
171 f24*
277 1.20
261 +30
234k41
GS activity, hexokinase activity and glycogen content
The activities of GS and hexokinase and the glycogen
content are shown in Table 2. Intralipid had no apparent
effect on the basal (13.1-1-1.9 versus 11.4-1-2.3%) or
stimulated (30.8 f 2.3 versus 27.6 -1- 4.5%) activities of
GS. Thus the degree of stimulation of fractional GS
activity (stimulated minus basal) was identical on both
study days (235% and 242% for the Intralipid and control
study days, respectively). GS, in the basal state
(1.46 f0.21 versus 1.21 k 0.17 units/g) was similar on the
Intralipid and control study days and did not change
during the insulin infusion (1.29 f 0.22 versus 1.09 f 0.34
units/g). There was also no significant difference in basal
or stimulated skeletal muscle glycogen content on both
study days (Table 3). Skeletal muscle hexokinase activity
did not change in the presence or absence of raised
plasma NEFA concentrations or during the insulin infusion.
DISCUSSION
In the present study we have demonstrated that
maintenance of plasma NEFA levels at approximately
fasting values resulted in a significant increase in lipid
oxidation during euglycaemic hyperinsulinaemia. Plasma
NEFA were mainly derived from the action of endothelial
lipoprotein lipase on Intralipid triacylglycerols. However,
NEFA released at the endothelial cell surface need not
necessarily enter the plasma pool and thus the plasma
NEFA concentrations might not reflect the entry rate of
NEFA into peripheral tissues. A 28% decrease in insulinstimulated whole-body glucose uptake ensued, and this
was the consequence of a 48% decrease in glucose oxida-
223
tion and a small but non-significant decrease in nonoxidative glucose disposal.
These findings are in agreement with previous reports.
However, the majority of these studies [ l l , 14, 151 were
performed at considerably higher plasma NEFA concentrations than the current study, higher even than the
fasting levels observed in newly presenting, untreated type
2 patients [5], thus making extrapolation to the diabetic
state difficult. Even though the present study was performed at mildly elevated plasma NEFA levels, the
steady-state plasma NEFA concentrations were twice
those reported in untreated type 2 patients after a
comparable time exposure to hyperinsulinaemia.
Although care must be taken when extrapolating from
this acute study which was performed at 4.0 mmol/l to the
diabetic state, our data strongly suggest that the
glucose-fatty acid cycle could operate in type 2 diabetes.
However, on the basis of the present results it is most
unlikely that this would account for more than 30% of the
insulin insensitivity seen in type 2 diabetes.
In this study, the majority of the decrease in insulinstimulated whole-body glucose disposal is accounted for
by inhibition of glucose oxidation. In a number of
previous studies, however, elevated plasma NEFA levels
have also been reported to inhibit non-oxidative glucose
disposal [8, 9, 151. As non-oxidative glucose disposal is
generally thought to reflect storage of glucose, these
studies suggest that glycogen synthesis may be decreased.
Most recently, the addition of palmitate to human skeletal
muscle preparations has been demonstrated to inhibit
glycogen synthesis, both basally and during insulin stimulation [28]. The mechanism underlying these observations
is uncertain, but a direct effect of fatty acids on GS has
been proposed as one possible explanation.
GS, the key regulatory enzyme in glucose storage, is
normally regulated covalently by phosphorylation/
dephosphorylation of specific serine residues, and
allosterically by G6P [29]. In studies in vitro fatty acids
may affect liver GS activity by causing dissociation into
subunits [16]. As this would be expected to bring about
decreased activity of GS by insulin in vivo, this
A. 6.Johnson et al.
224
--1.2
75.0
I
Y
M
Y
M
-
-
I
I
.-c
c
’5 0.8
E
El
El
v
C
._
.-e0,
0
0
-.2 0.4
2.5
-.
P
a
;0.0
3 0.0
u
min
rnin
rnin
min
Fig. 3. Basal and insulin-stimulated rates of lipid oxidation, glucose oxidation, glucose disposal and glucose storage
during a 240 min euglycaemic clamp performed on the control ( Q ) and on the lntralipid (B) study days. Values are
means +SEM.
mechanism has been put forward to explain the decreased
non-oxidative glucose disposal rates observed in these
studies [30]. In the present study, however, we have
demonstrated that acutely raising plasma NEFA levels to
approximately 820 ,umol/l had no effect on fasting GS
fractional activity. Maintenance of plasma NEFA levels at
fasting values also had no demonstrable effect on insulinstimulated GS activity. Furthermore, the consequent
increased lipid oxidation rates had no significant effect
upon insulin-stimulated non-oxidative glucose disposal. It
could be argued that the steady-state plasma NEFA levels
achieved in this study were not sufficiently high to affect
GS activity. However, similar findings to ours have
recently been reported in the presence of markedly
elevated plasma NEFA levels [31], and it thus seems
unlikely that fatty acids inhibit non-oxidative glucose
disposal via this mechanism. Nevertheless, an effect of
fatty acids could possibly be masked if GS, and GS, were
to decrease in parallel. This can be discounted though, as
GS, activity was unaltered by Intralipid both basally and
during insulin infusion.
Alternatively, it has been suggested that NEFA may
exert their effect on non-oxidative glucose disposal via the
glucose-fatty acid cycle. Elevation of plasma NEFA levels
during the infusion of insulin leads to an increase in lipid
oxidation. This in turn results in an increase in intracellular acetyl-CoA and citrate concentrations producing
inhibition of pyruvate dehydrogenase and phosphofructokinase activities, respectively. Glucose oxidation and
glycolysis are inhibited and G 6 P accumulates. Earlier
steps of glucose metabolism are inhibited, eventually
leading to decreased glucose transport into skeletal
muscle, and thus decreased, glycogen synthesis may
result. Randle et al. [ 7 ] , however, originally hypothesized
that the converse might occur, and proposed that
decreased rates of glucose flux through the oxidative
pathway would favour redirection of glucose metabolism
towards glycogen synthesis. A number of studies have
demonstrated that the glycogen content of rat heart,
diaphragm and skeletal muscle preparations increased
when plasma NEFA levels were elevated, thus supporting
their hypothesis [32, 331. Our data demonstrate for the
first time that the glycogen content of human skeletal
muscle was unaffected by increased plasma NEFA concentrations, both basally and during hyperinsulinaemia. In
fact, there was a small non-significant increase in muscle
glycogen content during the Intralipid infusion, whereas
non-oxidative glucose disposal decreased slightly.
It might be considered that a type 2 statistical error is
present when interpreting the skeletal muscle enzyme
data given the small number of subjects who underwent
paired biopsy. However, we estimated that the power of
the study to show activation of GS during either clamp
was approximately 90% for a two-sided 5% significance
test, indicating that the likelihood of a type 2 error is
small. Further calculations based upon the muscle biopsy
data of five of the subjects indicated that we would need
as many as 110 subjects to observe a true difference in GS
Fatty acid supply and glucose metabolism
activation between the two studies at the 5% level in a
two-sided test. In view of the invasive nature of the study
it was felt unjustified to submit further subjects to biopsy.
Additional support for the operation of the
glucose-fatty acid cycle comes from the intermediary
metabolite data. On the Intralipid study day, decreased
glucose flux through the glycolytic pathway is suggested
by the lower clamp blood pyruvate and lactate concentrations. The significant decrease in blood alanine concentrations is also consistent with decreased flux through the
glycolytic pathway, as pyruvate is a major precursor of
alanine. Our data therefore confirm the findings, at much
higher plasma NEFA concentrations ( = 3 mmol/l), of
Ferrannini et al. [14]. Their suggestion that decreased
blood alanine levels are secondary to decreased alanine
release from muscle rather than increased liver alanine
uptake has been recently confirmed [34]. Overall, reduced
glucose flux through the glycolytic pathway would be
expected to favour redirection of any glucose taken up
towards the storage pathways, and decreased glycogen
synthesis would only arise when glucose uptake became
rate-limiting.
In the present study we did not measure the activity of
pyruvate dehydrogenase, the key regulatory enzyme of
glucose oxidation. Previous work on human pyruvate
dehydrogenase in vivo has produced discrepant findings,
with some groups reporting activation by insulin [35],
whereas others have found no change [36]. These discrepancies may well reflect the complexity of the enzyme
system and assay methodology. We have measured
skeletal muscle hexokinase activity, an important regulatory enzyme in the glycolytic pathway, and demonstrated that elevated plasma NEFA levels had no direct
effect on its activity. However, preparation of the muscle
extract and dilution in the assay will have decreased the
concentration of its allosteric inhibitor, G6P. Unfortunately, it is extremely difficult to examine the effect of
fatty acids upon the glycolytic and oxidative pathways in
human skeletal muscle in vivo, as the use of freeze-clamp
techniques to determine metabolic crossover points is not
practicable. Furthermore, it is unlikely that needle-biopsy
specimens will be snap-frozen in under 2 s [ 3 7 ] .Muscle
metabolism will continue during this period, again making
interpretation of data extremely difficult.
In summary, the present study has demonstrated that
maintenance of plasma NEFA levels at fasting values decreased insulin-stimulated glucose metabolism by approximately 30%. This was due to a 48% decrease in glucose
oxidation and a non-significant decrease in glucose
storage. Elevated plasma NEFA concentrations had no
demonstrable direct effect on the measured activity of GS
either basally or during insulin stimulation. Similarly,
skeletal muscle glycogen content was also unaffected by
the elevated plasma NEFA levels.
ACKNOWLEDGMENTS
We thank our subjects for their co-operation, Linda
Ashworth, Lindsay Brigham, Peter Robinson, Lesley
225
Morrison and Louise Charlton for performing the assays
of insulin, C-peptide and intermediary metabolites; Sister
Margaret Miller and Sister Mavis Brown for the care of
our patients, and Peter Kelly for his help in performing the
statistical analyses for this paper. This work was
supported by Glaxo Group Research. A.B.J. was the
recipient of a Wellcome Clinical Fellowship.
REFERENCES
I. DeFronzo, R.A., Gunnarson, R., Bjorkman, O., Olsson, M. & Wahren, 1.
Effects o f insulin on peripheral and splanchnic glucose metabolism in noninsulin dependent (type 2) diabetes mellitus. J. Clin. Invest. 1985; 76,
148-55.
2. Golay, A., DeFronzo, R.A., Ferannini, E. et al. Oxidative and non-oxidative
glucose metabolism in non-obese Type 2 (non-insulin dependent) diabetic
patients. Diabetologia 1988; 3 I , 585-9 I.
3. Boden, G., Ray, T.K., Smith, R.H. & Owen, O.E. Carbohydrate oxidation and
storage rates in obese non-insulin dependent diabetic patients: effect of
improving glycaemic control. Diabetes 1983; 32,982-7.
4. Chen, Y.-D.I., Golay, A., Swislocki, A.L.M. & Reaven, G.M. Resistance t o
insulin suppression o f free fatty acid concentrations and insulin stimulation of
glucose uptake in non-insulin dependent diabetes mellitus. J. Clin. Endocrinol.
Metab. 1987; 64, 17-2 I.
5. Johnson, A.B., Argyraki, M., Thow, J.C. et al. Impaired activation o f skeletal
muscle glycogen synthase in NIDDM is unrelated t o the degree of obesity.
Metab. Clin. Exp. I99 I; 40, 252-60.
6. Felber, J.P., Ferrannini, E., Golay, A. et al. Role o f lipid oxidation in pathogenesis of insulin resistance o f obesity and Type 2 diabetes. Diabetes 1987;
36, I341-50.
7. Randle, P.J., Garland, P.B., Hales, C.N. & Newsholme, E.A. The glucose fatty
acid cycle: its role in insulin insensitivity and the metabolic disturbances o f
diabetes mellitus. Lancet 1963; i,785-9.
8. Thiebaud, D., DeFronzo, R.A., Jacot, E. et al. Effect of long chain triglyceride
infusion on glucose metabolism in man. Metab. Clin. Exp. 1982; I I, 1128-36.
9. Felley, C.P., Felley, E.M., van Melle, G.D., Frascarolo, P., Jequier, E. & Felber,
J.P. Impairment of glucose disposal by infusion of triglycerides in humans:
role o f glycaemia. Am. J. Physiol. 1989; 256, E747-52.
10. Rouselle, J.,Buckert, A., Pahud, P., Jequier, E. & Felber, J.P. Relationship
between glucose oxidation and glucose tolerance in man. Metab. Clin. Exp.
1982; 31,866-70.
I I, Lee, K.U., Lee, H.K., Koh, C.S. & Min, H.K. Artificial induction of intravascular lipolysis by lipid-heparin infusion leads t o insulin resistance in man.
Diabetologia 1988; 31, 285-90.
12. Vouillamoz, D., Temier, E., Jequier, E. & Felber, J.P. Importance of substrate
comDetition in the mechanism o f insulin resistance in man. Metab. Clin. Exp.
I 98i; 36,7 I 5-20.
leauier. E.. Chabot. V., Buber, V. & Felber, I.P. Carbohydrate and
13. Gomez, F., ~.
lipid oxidation in normal human subjects: its influence on glucose tolerance
and insulin response t o glucose. Metab. Clin. Exp. 1972; 21, 381-91.
14. Ferrannini, E., Barrett, E.J., Bevilacqua, S. & DeFronzo, R.A. Effect o f fatty
acids on glucose production and utilisation in man. J. Clin. Invest. 1983; 72,
1737-47.
15. Bonadonna, R.C., Zych, K., Boni, C., Ferrannini, E. & DeFronzo, R.A. Time
dependence of the interaction between lipid and glucose in humans. Am. J.
Physiol. 1989; 257, E49-56.
16. Wititsuwannakul, D. & Kim, K.H. Mechanism of palmityl coenzyme A
inhibition o f liver glycogen synthase. j. Biol. Chem. 1977; 252,7812-17.
17. Randle, P.J. Fuel selection in animals. Biochem. SOC.Trans. 1986; 14,
799-806.
18. Randle, P.J., Kerbey, A.L. & Espinal, J. Mechanisms decreasing glucose oxidation in diabetes and starvation: role o f lipid fuels and hormones. Diabetes
Metab. Rev. 1988:. 4.623-38.
.
19. Frayn, K.N. Calculation o f substrate oxidation rates in vivo from gaseous
exchange. J.Appl. Physiol. 1983; 55,628-34.
20. Golden, S., Wals, P.A. & Katz, J, An improved procedure for the assay o f
glycogen synthase and phosphorylase in rat liver homogenates. Anal.
Biochem. 1977; 77,436-45.
226
A. B. Johnson e t al.
21. Easterby, J.S. & Quadri, S.S. Hexokinase type 2 from rat skeletal muscle.
Methods Enzymol. 1977; 90, 11-15,
22. Soeldner, S.]. Critical variables in the radioimmunoassay of serum insulin
using the double antibody technique. Diabetes 1965; 14, 77 1-9.
23. Heding, L.G. Radioimmunological determination o f human C-peptide in
serum. Diabetologia 1975; I I,541-8.
24. Harrison, J.,Hodson, A.W., Skillen, A.W., Stappenbeck, R., Agius, L. &
Alberti, K.G.M.M. Blood glucose, lactate, pyruvate, glycerol, 3-hydroxybutyrate and acetoacetate merasurements in man using a centrifugal analyser
with a fluorimetric attachment. J.Clin. Chem. Clin. Biochem. 1988; 26, 141-6.
25. Knox, D.P. &Jones, D.G. Automated enzymic determination of free fatty
acids by centrifugal analysis. J. Autom. Chem. 1984; 6, 152-4.
26. Hawk, P.B. Practical physiological chemistry. In: Kjeldahl method. 12th ed.
Toronto: Blakiston, 1947: 8 14-22.
27. Armitage, P. & Berry, G. The size o f a statistical investigation. Statistical
methods in medical research. 2nd ed. Oxford: Blackwell Scientific Publications, 1987: 179-85.
28. Argyraki, M., Wright, P.D., Venables, C.W., Proud, G. & Taylor, R. In vitro
study o f human skeletal muscle strips: effect o f non-esterified fatty acid
supply on glucose storage. Metab. Clin. Exp. 1989; 38, I 183-7.
29. Cohen. P. Muscle glycogen synthase. In: Boyer, P.D. & Krebs, E.G., eds. The
enzymes. 3rd ed. London: Academic Press Inc., 1986: vol. 17, 46 1-97,
30. DeFronzo, R.A. The triumvirate: 6-cell, muscle, liver. A collusion responsible
for NIDDM. Diabetes 1988; 37,667-87.
31. Koivisto, V.A., Saloranta, C., Falholt, K. & Harkonen, M. Enhanced lipid
oxidation does not reduce non-oxidative glucose disposal o r muscle
glycogen synthase activity in healthy man [Abstract]. Diabetes 1989;
38 (Suppl. 2), 89A.
32. Randle, P.J., Newsholme, E.A. & Garland, P.B. Regulation of glucose uptake
by muscle. 8. Effects of fatty acids, ketone bodies and pyruvate and o f
alloxan-diabetes and starvation, on the uptake and metabolic fate o f glucose
in rat heart and diaphragm muscles. Biochem. J. 1964; 93,652-65.
33. Jenkins, A.B., Storlien, L.H., Chisholm, D.J.& Kraegen, E.W. Effects o f
nonesterified fatty acid availability on tissue specific glucose utilisation in rats
in vivo. J.Clin. Invest. 1988 82, 293-299.
34. Walker, M., Fulcher, G.R., Catalano, C., Petranyi, G., Orskov, H. & Alberti,
K.G.M.M. Physiological levels of plasma non-esterified fatty acids impair
forearm glucose uptake in normal man. Clin. Sci. 1990; 79, 167-74.
35. Mandarino, L.J., Wright, K.S., Verity, L.S. et al. Effects o f insulin infusion on
human skeletal muscle pyruvate dehydrogenase, phosphofructokinase, and
glycogen synthase. Evidence for their role in oxidative and non-oxidative
glucose metabolism. J. Clin. Invest. 1987 ; 80, 655-63.
36. Kruszynska, Y., Williams, N., Perry, W.M. & Home, P. The relationship
between insulin sensitivity and skeletal muscle enzyme activities in hepatic
cirrhosis. Hepatology 1988; 8, I 6 15-19,
37. Yki-Jarvinen, H., Shah, K., Ren, J.M. & Koivisto, V.A. Localisation of ratelimiting defect for glucose disposal in skeletal muscle of insulin-resistance
Type I diabetic patients. Diabetes 1990; 39, 157-67.