Download Aspects of organ protein, amino acid and glucose

Clinical Science (2003) 104, 127–141 (Printed in Great Britain)
Aspects of organ protein, amino acid and
glucose metabolism in a porcine model
of hypermetabolic sepsis
Maaike J. BRUINS, Nicolaas E. P. DEUTZ and Peter B. SOETERS
Department of Surgery, Maastricht University, Maastricht, The Netherlands
A
B
S
T
R
A
C
T
Although glucose and protein metabolism have been investigated extensively in experimental
models of hypodynamic sepsis, relatively little information is available regarding the compensated stage of sepsis. We investigated interorgan amino acid and glucose metabolism in a
porcine model of compensated hyperdynamic sepsis. Fasting catheterized pigs received
endotoxin (Escherichia coli lipopolysaccharide ; 3 µg : h−1 : kg−1 ; intravenous) or saline (controls)
and volume resuscitation over 24 h to reproduce hyperdynamic sepsis. Primed-constant infusions
of p-aminohippurate and 3H-labelled isotopes were used to measure glucose, amino acid and
protein metabolism across the portal-drained viscera, liver and hindquarters (to represent
muscle) at 0 and 24 h of endotoxaemia. Whole-body protein and glucose flux were increased
during hyperdynamic compensated sepsis. In endotoxaemic pigs, visceral protein was conserved,
and hindquarter protein breakdown exceeded the increase in liver protein synthesis, resulting
in net whole-body protein loss. Endotoxaemia increased hindquarter and visceral glycolysis and
branched-chain amino acid transamination. The rate of efflux of glutamine and alanine from the
hindquarters was higher than anticipated from protein breakdown, indicating de novo synthesis
of these amino acids during endotoxaemia. In addition to the hindquarters, the portal-drained
viscera provided substantial gluconeogenic amino acids and lactate to the liver. Although
increased liver glutamate release constitutes an important nitrogen-sparing mechanism and
carbon skeletons are effectively being cycled in glucose, net body protein is lost through
increased ureagenesis during the hyperdynamic stage of sepsis. Specific amino acid requirements
may develop in compensated hyperdynamic sepsis that is characterized by maintained organ
perfusion and increased substrate utilization at the expense of body protein.
INTRODUCTION
Septic shock is a major cause of death in the intensive care
unit. The initial cardiovascular response following infection is characterized by systemic vasodilation, usually
resulting in a hyperdynamic state with an elevated heart
rate and normal or slightly decreased blood pressure. In
the subsequent phase, a hyperdynamic circulation related
to the massive vasodilation is characterized by haemodynamics as decreased or borderline blood pressure,
Key words : branched-chain amino acids, endotoxaemia, enteral nutrition, fasting, glucose metabolism, intestine, liver, muscle,
protein metabolism, swine.
Abbreviations : AA, sum of measured amino acids ; BCAA, sum of branched-chain amino acids ; GNAA, sum of gluconeogenic
amino acids ; I, infusion rate ; NB, net balance ; PAH, p-aminohippurate.
Correspondence : Dr Maaike J. Bruins, Unilever Health Institute, Unilever Research, P.O. Box 114, NL-3130 AC Vlaardingen,
The Netherlands (e-mail maaike.bruins!unilever.com).
# 2003 The Biochemical Society and the Medical Research Society
127
128
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
fever and hyperventilation. When blood pressure is not
maintained by fluid resuscitation during this period, this
results in a hypodynamic circulation related to vasoconstriction, characterized by compromised cardiac output and coldness. Successful treatment, including fluid
resuscitation during this period, however, may prevent
progression to the hypodynamic stage and result in
compensated sepsis, with typically elevated cardiac output and low systemic vascular resistance.
One of the most critical metabolic features of hyperdynamic sepsis is extreme hypermetabolism and catabolism [1,2]. The metabolic response of the body
following sepsis comprises mobilization of substrates
from the periphery to be utilized by visceral tissues [3]
and immune cells [4,5], resulting in loss of lean body
mass. Patients with sepsis demonstrate accelerated protein degradation [6,7] and increased transamination and
oxidation of amino acids in skeletal muscle [8]. Consequently, gluconeogenic and other amino acids are released in increased amounts into the circulation. Also,
glycolysis of glucose to form lactate is increased in septic
patients [9,10]. Amino acids and lactate that are released
into the circulation become available for accelerated liver
glucose production. During gluconeogenesis, carbon
skeletons from amino acids and lactate (alanine cycle plus
Cori cycle) are converted into glucose, while the amino
groups are used in ureagenesis, resulting in net nitrogen
loss excreted as urea [2]. The glucose produced in this
response is the main energy source for cells involved in
the host immune response during sepsis [11]. In addition
to accelerated gluconeogenesis, peripheral mobilized
amino acids serve as substrates for central organs,
including liver (for acute-phase protein synthesis) [12]
and enterocytes [11]. Despite effective cycling of carbon
skeletons between the periphery and the liver in the
appearance of glucose, the response to sepsis is energetically expensive, with the end result being the net
degradation of whole-body protein, specifically muscle
protein.
Even when large amounts of carbohydrate, fat and
protein or amino acids are supplied, loss of lean body
mass of septic patients is difficult to prevent [13–16].
Over the past few years, new strategies in nutritional
support aiming at maintenance of lean body mass have
been developed. Adequate caloric and protein nutritional
support of critically ill patients may be the optimal
approach to prevent malnutrition by meeting the elevated
energy requirements and maintaining body nitrogen and
acute-phase protein synthesis in the hypermetabolic
response. Therefore it is of great importance from a
clinical point of view to gain more insight into the
quantitative changes in organ protein and glucose metabolism and interorgan substrate fluxes that occur in the
response to hyperdynamic sepsis.
Because of its property of provoking a generalized
pro-inflammatory response in the infected host [17],
# 2003 The Biochemical Society and the Medical Research Society
endotoxin is often used to reproduce a sepsis-like
condition in experimental models. The haemodynamic
features of sepsis models in rodents, however, differ
substantially from those in large animals. In rodents,
endotoxin is often bolus injected in large doses, whereas
in most pig models of sepsis, endotoxin is infused
continuously in moderate amounts. Whereas large
primed doses of endotoxin elicit a hypodynamic cardiovascular response (i.e. hypotension) more representative of decompensated sepsis or septic shock, small
endotoxin doses elicit a hyperdynamic cardiovascular
response [18]. In the present study, pigs were infused
continuously with low doses of endotoxin and were
fluid-resuscitated to compensate for intravascular fluid
losses, to achieve a hyperdynamic circulation representative of compensated sepsis in patients receiving modern
intensive care.
The alterations in protein and glucose metabolism that
occur under septic conditions, and their underlying
mechanisms, have been described extensively in experimental models of acute sepsis. Only a few animal studies
have described protein and glucose metabolism in a
hyperdynamic and hypermetabolic model of compensated sepsis.
More knowledge regarding the alterations in protein
and glucose metabolism and specific requirements present during the hypermetabolic response to sepsis may be
of help in composing new strategies in nutritional
support. We assessed changes in glucose utilization and
production rates and in protein synthesis and degradation
rates both at a whole-body level and across the hindquarters, portal-drained viscera and liver in a porcine
model of hyperdynamic sepsis.
METHODS
Animals
The Animal Ethics Committee of the Maastricht University approved all animal work, and experimental
protocols complied with local guidelines for the use
of experimental animals. Female pigs (offspring of
Yorkshire and Dutch Landrace species) were 3 months
old and weighed 20–22 kg. The animals were given 1 kg
of regular pig feed (Landbouwbelang, Roermond, The
Netherlands ; 16 % crude protein) daily, supporting a
growth rate of approx. 300 g\day. Pigs had free access to
water. Before surgery, pigs (n l 14) were randomly
assigned to one of the two treatment groups.
Surgical procedure
Food was withheld from the night before surgery.
Intramuscular premedication was with 10 mg\kg azoperone (Stresnil2 ; Janssen Pharmaceutica, Beersse,
Belgium) ; 1 h later, anaesthesia was induced by a mixture
Organ substrate metabolism in a pig model of endotoxaemia
of N O\O (1 : 2, v\v) and halothane (0.8 %). After
#
#
intubation, the pigs were intravenously administered
6.25 mg\kg Lincomycin2 (A.U.V., Cuyk, The
Netherlands) as bactericidal prophylaxis and 12.5 mg\kg
Spectinomycin2 (A.U.V.) as bacteriostatic prophylaxis.
Flunixine2 (50 mg\kg ; Finadyne ; Schering-Ploegh,
Brussels, Belgium) was given as postoperative analgesic.
During surgery, anaesthesia was maintained by the
mixture of N O\O and halothane and by intravenously
#
#
infused Lactetrol2 (per ml : 5.76 mg of NaCl, 0.37 mg of
KCl, 0.37 mg of CaCl : 2H O, 0.2 mg of MgCl : 6H O,
#
#
#
#
5 mg of sodium lactate ; Janssen Pharmaceutica).
The surgical procedure has been described in detail
elsewhere [19,20]. In brief, after opening of a midline
incision, seven catheters were inserted into various blood
vessels. Two catheters were placed in the abdominal
aorta : one just above the bifurcation (A1) and one just
above the right renal vein (A2). Two catheters were
implanted in the inferior caval vein at the corresponding
positions (V1 and V2 respectively). Furthermore,
catheters were placed into the portal, hepatic and splenic
veins. A gastrostomy catheter was inserted into the
stomach. All catheters were tunnelled through the abdominal wall and skin. Also, a jejunal Bishop–Koop
stoma was constructed which was brought through the
abdominal wall. In brief, 20 cm distal to the ligament of
Treitz, the jejunum was transected and the anatomical
continuity of the remaining bowel was re-established by
an end-to-side anastomosis using a single-layer running
suture (PDS 4-0 synthetic polydioxanon ; Ethicon,
Norderstedt, Germany). During recovery of the pigs, an
inflated balloon catheter (size Ch.16 ; 3–5 ml ; Ru$ schGold, Kernen, Schwalbach, Germany) served to close the
stoma and to prevent leakage from the jejunum. The pigs
wore a canvas harness to protect the catheters and stoma,
and to allow ease of handling. Postoperative care of the
pigs was standard [20]. During handling, the pigs were
placed in a movable cage to get them accustomed to this
condition.
Experimental protocol
Nutrition was withheld for 8 h before the start of the
experiments. A total of 14 pigs received either lipopolysaccharide endotoxin (dissolved in saline ; Escherichia coli
serotype 055 : B5 ; Sigma Chemical Co., St. Louis, MO,
U.S.A.) at a rate of 3 µg : h−" : kg−" via the V2 catheter for
24 h. Control animals (n l 7) received saline at an
equivalent infusion rate. For the subsequent 24 h, all pigs
were infused with saline (30 ml : h−" : kg−" during the
first 8 h and 20 ml : h−" : kg−" during the subsequent
16 h) to replenish intravascular volume losses. Blood
samples were taken before and 24 h after the start of endotoxin or saline infusion. The body temperature of
the pigs was measured throughout the experimental
period.
Sample processing
Immediately after the blood samples were collected, the
blood was distributed into heparin-containing tubes
(Sarstedt, Nu$ mbrecht, Germany) on ice. The haematocrit
was determined. For the analysis of arterial pH, HCO −,
$
arterial partial pressures of oxygen and carbon dioxide,
and oxygen saturation, 200 µl samples of blood were
sealed in 1 ml airtight heparin-containing syringes and
analysed immediately using an automatic blood gas
system [Acid Base Laboratory (ABL3) ; Radiometer,
Copenhagen, Denmark]. Supernatants were kept on ice.
For the measurement of urea, glucose and lactate concentrations, blood was centrifuged (4 mC, 5 min, 8500 g),
and 900 µl of the resultant plasma was added to 90 µl of
3 mmol\l trichloroacetic acid solution to ensure stability
of the substances. For amino acid analysis, 500 µl of
plasma was deproteinized by mixing it with 20 mg of dry
sulphosalicylic acid. For the measurement of the
p-aminohippurate (PAH) concentration, 300 µl of whole
blood was added to 600 µl of 0.7 mmol\l trichloroacetic
acid solution ; the solution was mixed thoroughly and
centrifuged (4 mC, 5 min, 8500 g) followed by the collection of supernatant fluid. All samples were frozen in
liquid nitrogen and stored at k80 mC until analysed.
Biochemical analysis
To determine the PAH concentration, the supernatant
fluid of deproteinized whole blood was deacetylated at
100 mC for 45 min. The PAH concentration was detected
by standard enzymic methods and spectrophotometric
analysis with an automated Cobas Mira-S system
(Hoffmann-La Roche, Basel, Switzerland). To calculate
plasma PAH concentrations, the whole-blood PAH
concentration was corrected for the haematocrit value.
Before urea measurements, plasma ammonia was removed by conversion into glutamate. The urea was then
converted enzymically into ammonia by the addition
of urease. The ammonia formed in this reaction was
quantified by measuring the extinction of NADPH
utilized in the conversion of ammonia into glutamate.
Plasma amino acid, glucose and lactate concentrations
were measured with a fully automated HPLC system
(Pharmacia, Woerden, The Netherlands), using methods
reported previously [21,22]. After HPLC separation,
phenylalanine, valine and glucose fractions were collected
and counted for radioactivity in a liquid scintillation
spectrophotometer (Beckman) to calculate specific radioactivities (d.p.m.\nmol).
Concentrations of the acute-phase proteins haptoglobin and fibrinogen and of total protein in plasma were
measured with a Nephelometer (model BN 100 ; Dade
Behring Vertriebs GmbH and Co., Schwalbach,
Germany). Proteins were determined by using rabbit
antibodies against human haptoglobin and fibrinogen
(Dade Behring), as performed for human plasma samples.
# 2003 The Biochemical Society and the Medical Research Society
129
130
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
Standard curves for haptoglobin and fibrinogen were
constructed using a purified human standard (Dade
Behring) and a secondary porcine standard (Sigma
Chemical Co.). High correlation coefficients ( 0.9)
were found for the results obtained using the two
methods.
Infusion protocol
At 1 h before the endotoxin infusion was started, a
primed infusion protocol was conducted. An infusion of
25 mmol\l PAH (A1422 ; Sigma Chemical Co.) was
started at a rate of 40 ml\h through the splenic vein and
the A1 (abdominal aorta) catheter after an initial bolus
of 5 ml [20]. Directly after the primed constant infusion of
PAH, a priming dose (1 µCi\kg) followed by a constant
infusion (1 µCi : h−" : kg−") of L-[2,6-$H]phenylalanine
([$H ]phenylalanine) and L-[3,4-$H]valine ([$H ]valine)
#
#
and a primed (0.5 µCi\kg) constant infusion
(0.5 µCi : h−" : kg−") of D-[6-$H]glucose were given via
the venous (V2) catheter. [$H ]Phenylalanine and
#
[$H ]valine were both purchased from Amersham,
#
and [$H]glucose was from NEN Dupont, NET-100A
(Mechelen, Belgium). Before the start of the infusions,
background blood samples were taken. During the final
60 min of the infusion protocol, an isotopic plateau
(calculated slope of isotopic enrichment against time not
different from zero) was reached [23]. Steady-state
conditions for PAH were obtained 1 h after the primed
infusion of PAH (results not shown). Blood samples
were collected in triplicate at 15-min intervals during the
last 60 min of infusion of PAH and the isotopes [24].
Calculations
Under steady-state conditions, the disappearance rate of
a substrate equals its rate of appearance, defined as ‘ flux
rate ’ (Q ; production plus exogenous infusion). The
turnover rates of phenylalanine, valine and glucose were
calculated using the formula :
Q l I\SAA
(1)
in which I is the rate of infusion (d.p.m. : min−" : kg−") of
[$H ]phenylalanine, [$H ]valine or [$H]glucose, and SAA
#
#
is the specific radioactivity (d.p.m.\nmol) in arterial
plasma of the corresponding isotope [25]. The turnover
rate of phenylalanine was used as an indication of wholebody protein breakdown, as this amino acid cannot be
newly synthesized. The turnover rate of glucose represents the sum of glucose appearing in the circulation from
glycogen breakdown and gluconeogenesis. However, 6$H label may be lost to some extent (" 10 %) in the
phosphoenolpyruvate cycle, contributing to a corresponding overestimation of the measured rate of turnover
of glucose [25,26].
Substrate metabolism across the hindquarters, portal# 2003 The Biochemical Society and the Medical Research Society
drained viscera and liver was calculated using a twocompartment model, as described previously [25].
The portal-drained viscera are defined as the total of all
portal-drained organs, besides the spleen, stomach and
pancreas, mainly representing the intestines. The splanchnic area includes the portal-drained viscera and liver ;
therefore calculations for the liver were made by subtracting values for the portal-drained viscera from those
for the splanchnic area. Calculations of kinetics in muscle
were made on the assumption that the hindquarters
represent 50 % of whole-body muscle [27]. The plasma
flow rates across the organs were calculated based on the
principle of dye dilution using the formula :
Plasma flow rate l I\([PAH]Vk[PAH]A)
(2)
in which I represents the rate at which PAH is infused
and [PAH]V and [PAH]A are the concentrations of PAH
in the venous and arterial plasma of the organ respectively.
Substrate net balance (NB) was calculated using the
formula :
NB l plasma flow ratei([S]Vk[S]A)
(3)
in which [S]V and [S]A represent venous and arterial
plasma concentrations respectively of the substrate.
Therefore a positive NB represents net release or efflux,
and a negative NB represents net uptake or influx, of an
amino acid across the organ.
Tracer NB was calculated in a similar way, based on
tracer dilution across the organ, using the specific
radioactivity (d.p.m.\nmol) of the tracer in arterial (SAA)
and venous (SAV) plasma :
Tracer NB l
plasma flow ratei(SAAi[S]AkSAVi[S]V)
(4)
The disposal of a substrate by an organ represents
disappearance due to degradation and\or incorporation
into macromolecules, and is calculated as follows :
Disposal l tracer NB\SAV
(5)
In this calculation, SAV is assumed to represent the
precursor pool (intracellular SA) [28]. Because the NB of
a substrate across an organ is the difference between
production and disposal, production can be calculated as
follows :
Production l NBjdisposal
(6)
Since phenylalanine degradation in muscle and gut is low
[29], phenylalanine disposal in the hindquarters and
portal-drained viscera is a reflection of protein synthesis,
and phenylalanine production reflects protein breakdown. In the liver, valine disposal and production
represent protein synthesis and breakdown respectively,
since valine degradation in liver is considered to be low
because of low transamination activity [30–32].
Organ substrate metabolism in a pig model of endotoxaemia
The total amino acid concentration (AA) was calculated as the sum of measurable α-amino acids (glutamic
acid, asparagine, serine, glutamine, histidine, glycine,
threonine, citrulline, arginine, alanine, taurine, tyrosine,
valine, methionine, isoleucine, tryptophan, phenylalanine, ornithine, leucine and lysine), the sum of
gluconeogenic amino acids (GNAA) as all amino acids
occurring in protein except for leucine and lysine, and the
sum of branched-chain amino acids (BCAA) as the sum
of valine, leucine and isoleucine.
unaffected by endotoxin infusion, and the urinary output
of both groups increased during 24-h fluid infusions. No
significant time-dependent effects of endotoxin were
observed on arterial blood gas values or arterial haematocrit (PT×G l 0.07) (Table 2). The arterial carbon dioxide
pressure was lower (PT×G l 0.058) and the arterial pH
was significantly higher (PT×G l 0.03) in the endotoxininfused animals, indicating developing respiratory alkalosis. Although a group effect on plasma flow to the
portal-drained viscera was observed (PG l 0.02) due to
different baseline values, no time-dependent effects of
endotoxin challenge on organ plasma flows were apparent (Table 3).
Endotoxin infusion significantly decreased the plasma
concentrations of glucose, glutamine, glycine, arginine
and tyrosine, and increased the concentrations of lactate,
urea, alanine, taurine and phenylalanine (Table 4).
BCAA, GNAA and AA were all increased by endotoxin
infusion. Neither the plasma concentration of fibrinogen
(control, 1.04p0.07 g\l ; endotoxin, 0.99p0.12 g\l) nor
the plasma concentration of haptoglobin (control,
0.58p0.03 g\l ; endotoxin, 0.59p0.03 g\l) differed significantly between the two groups. A significantly lower
total protein concentration was observed in pigs
challenged with endotoxin (control, 31.3p2.3 g\l ; en-
Statistics
Results are presented as meanspS.E.M. The data were
subjected to general factorial ANOVA to assess group
effects (PG), time effects (PT) and interactions between
time and group (PT×G). P 0.05 was considered significant.
RESULTS
Endotoxin infusion resulted in a significant rise in body
temperature (Table 1) as compared with saline infusion.
The body weight and urinary output of the pigs remained
Table 1 Temperature, body weight and urinary output in the postabsorptive state before and 24 h after the start of saline
(control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Statistics by ANOVA : group (PG), time (PT) and groupitime (PTiG) effects.
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Temperature (mC)
Body weight (kg)
Urinary output (litres/day)
37.7p0.3
22.5p0.7
1.2p0.2
38.1p0.2
22.5p0.5
1.0p0.1
37.6p0.3
23.6p1.9
2.3p0.1
39.9p0.2
23.9p1.6
2.0p0.3
PG
PT
PTiG
0.00
0.05
Table 2 Arterial blood gas values in the postabsorptive state before and 24 h after the start of saline (control) or endotoxin
infusion
Data are meanspS.E.M. (n l 7). PaCO2, arterial pressure of carbon dioxide ; PaO2, arterial pressure of oxygen, SaO2, oxygen saturation. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Haematocrit (%)
pH
PaCO2 (kPa)
PaCO2 (mmHg)
PaO2 (kPa)
PaO2 (mmHg)
HCO3− (mM)
SaO2 (%)
29.1p1
7.44p0.02
5.64p0.46
42.3p3.45
12.0p0.8
90p6
27.3p1.8
97.0p0.6
28.7p0.8
7.44p0.01
5.58p0.12
41.8p1.5
12.7p0.6
95.2p1.5
27.7p1.4
97.3p0.4
27.9p1.3
7.40p0.01
5.36p0.29
40.2p2.2
11.9p0.8
89.2p6
21.8p3.2
96.7p0.9
24.5p0.5
7.46p0.01
4.69p0.22
35.2p1.65
13.0p0.5
97.5p3.8
24.1p1.1
97.2p0.4
PG
PT
PTiG
0.07
0.03
0.06
# 2003 The Biochemical Society and the Medical Research Society
131
132
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
Table 3 Organ plasma flows in the postabsorptive state before and 24 h after the start of saline (control) or endotoxin
infusion
Data are meanspS.E.M. (n l 7). Statistics by ANOVA : group (PG), time (PT) and timeigroup (PTiG) effects.
Flow (ml : min−1 : kg−1)
0h
24 h
ANOVA
Organ
Control
Endotoxin
Control
Endotoxin
PG
Portal-drained viscera
Liver
Muscle
28p4
11p6
25p3
36p5
10p3
30p6
26p4
22p4
32p7
34p1
22p4
43p4
0.02
PT
PTiG
Table 4 Arterial substrate concentrations in the postabsorptive state before and 24 h after the start of saline (control) or
endotoxin infusion
Data are meanspS.E.M. (n l 7). Individual GNAA are indicated by *, and essential amino acids by †. Statistics by ANOVA : group (PG), time (PT) and timeigroup
(PTiG) effects.
Concentration ( µmol/l)
0h
24 h
ANOVA
Substrate
Control
Endotoxin
Control
Endotoxin
Glucose
Lactate
Urea
Glutamate*†
Asparagine*†
Serine*†
Glutamine*†
Histidine*
Glycine*†
Threonine*†
Citrulline
Arginine*†
Alanine*†
Taurine†
Tyrosine*†
Valine†
Methionine†
Isoleucine†
Tryptophan†
Phenylalanine†
Ornithine*
Leucine*†
Lysine
BCAA
GNAA
AA
4385p296
459p20
3205p478
126p23
27p3
111p6
232p27
40p4
449p82
179p26
79p7
90p8
167p17
32p5
46p4
336p37
21p4
158p13
21p3
48p2
118p15
192p27
84p13
716p71
1756p93
2465p175
4847p221
542p49
3282p678
148p28
31p2
119p7
273p22
43p3
486p74
159p14
79p6
94p5
200p16
29p4
45p5
339p14
25p2
121p7
20p2
46p4
98p10
166p10
116p15
627p28
1847p48
2477p104
5218p237
524p54
1368p457
62p10
24p2
109p7
269p13
43p2
480p56
153p16
58p3
106p4
118p14
27p4
55p4
347p29
26p3
158p21
18p2
56p4
79p10
183p15
92p7
688p54
2171p80
2470p95
4092p150
796p62
2729p227
60p3
22p3
95p9
210p20
43p4
349p42
113p15
42p4
70p3
158p19
47p4
38p3
294p26
21p3
114p10
16p2
88p6
58p4
161p10
102p10
569p41
1804p113
2096p141
dotoxin, 22.3p2.1 g\l ; PT×G l 0.05). Figure 1 shows that
endotoxin infusion significantly increased the wholebody flux (appearance and disappearance) rates of glucose
# 2003 The Biochemical Society and the Medical Research Society
PG
PT
PTiG
0.00
0.05
0.01
0.05
0.02
0.05
0.00
0.05
0.00
0.04
0.00
0.01
0.05
0.01
0.05
0.05
0.05
(Figure 1A), valine (Figure 1B) and phenylalanine (Figure
1C), the latter indicating increased whole-body protein
breakdown.
Organ substrate metabolism in a pig model of endotoxaemia
demonstrated that endotoxin infusion significantly increased (2-fold) phenylalanine production (protein
breakdown), but not phenylalanine disposal (protein synthesis), in the hindquarters. There was no time effect
(fasting period plus infusions) on protein synthesis,
breakdown or NB across the hindquarters (Table 7).
As measured using [$H ]valine, both valine disposal
#
(control, 592p155 nmol : min−" : kg−" ; endotoxin,
1590p116 nmol : min−" : kg−" ; P l 0.01) and valine production (control, 524p221 nmol : min−" : kg−" ; endotoxin, 1181p153 nmol : min−" : kg−" ; P l 0.01) were increased significantly by endotoxin infusion.
Portal-drained viscera
Both glucose influx into and lactate efflux from the
portal-drained viscera were increased significantly during
endotoxaemia (Table 8). Increased glucose influx was due
to increased glucose disposal (Table 9). Endotoxin
infusion reduced the influx of glutamine, but increased
the influx of glutamate and BCAA, into the portaldrained viscera (Table 8). Moreover, endotoxin infusion
significantly increased glycine and tyrosine efflux, but
decreased alanine and arginine efflux, from the portaldrained viscera (Table 8). GNAA and lactate released
(net) from the portal-drained viscera accounted for 9 %
of the net uptake of these substrates by the liver. The
protein synthesis and breakdown rates in the portaldrained viscera, as measured using [$H ]phenylalanine,
#
did not change significantly with time and were not
affected by endotoxin infusion (Table 10).
Liver
Figure 1 Whole-body appearance of (A) glucose, (B) valine
and (C) phenylalanine (protein breakdown) after a 24 h saline
(control) or endotoxin infusion
Significance of differences : *P
0.05, **P
0.01 (two-way ANOVA).
Hindquarters
Glucose influx into and lactate efflux from the hindquarters were increased in the endotoxaemic compared
with the control pigs (Table 5). [$H]Glucose measurements showed that endotoxin infusion increased glucose
influx into the hindquarters by increasing glucose disposal (Table 6). In addition, efflux of glutamine and
alanine from the hindquarters was increased in the
endotoxaemic pigs (Table 5). The usual influx of the total
sum of amino acids turned into efflux during endotoxin
challenge. [$H ]Phenylalanine measurements (Table 7)
#
Glucose efflux from the liver was increased in endotoxininfused animals (Table 11), which was due to 2-fold
higher glucose production (Table 12) as compared with
controls. In the control animals, liver glucose efflux
decreased over time (P l 0.02 ; one-way ANOVA) during fasting. In the endotoxin-infused pigs, lactate influx
was 1.8-fold higher than in control pigs. Endotoxin
infusion also increased by 3.5-fold the influx of GNAA,
of which glycine was taken up to the greatest extent,
followed by alanine and glutamine (Table 11). AA influx
was increased by " 7.5-fold during endotoxin infusion.
The urea efflux of endotoxin-treated pigs was 2-fold
higher than that of control pigs (Table 11), but this
difference did not reach statistical significance (PT×G l
0.08). Endotoxin administration increased liver BCAA
influx (Table 11) and valine disposal (as measured from
[$H ]valine ; Table 13), implying that protein synthesis in
#
this organ was increased.
DISCUSSION
Investigation of glucose and protein metabolism at both
the whole-body and the organ level offers important
# 2003 The Biochemical Society and the Medical Research Society
133
134
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
Table 5 Substrate flux across the hindquarters in the postabsorptive state before and 24 h after the start of saline (control)
or endotoxin infusion
Data are meanspS.E.M. (n l 7). Negative flux represents net uptake. Individual GNAA are indicated by *, and essential amino acids by †. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
Flux (nmol : min−1 : kg−1)
0h
24 h
ANOVA
PG
Substrate
Control
Endotoxin
Control
Endotoxin
Glucose
Lactate
Glutamate*†
Glutamine*†
Glycine*†
Citrulline
Arginine*†
Alanine*†
Tyrosine
Valine†
Isoleucine†
Phenylalanine†
Leucine*†
BCAA
GNAA
AA
k3678p585
3269p563
k1512p231
1253p322
398p244
4p89
k110p72
529p149
4p52
k155p125
k146p163
k40p27
k302p93
k3678p585
k611p369
1537p595
k4610p579
3029p210
k1719p118
1254p146
332p1128
k80p64
k254p76
326p151
k37p35
k279p115
k166p39
k88p26
k344p75
k4610p579
k789p206
1339p560
k3670p932
2438p685
k881p186
552p109
k298p53
k85p25
k168p42
398p87
k93p56
k429p115
k278p59
k24p23
k381p90
k1088p181
k2092p679
k2464p809
k7042p965
4785p645
k1087p280
1872p269
1235p183
80p19
205p40
2515p639
259p58
k409p86
424p180
236p11
249p66
264p192
7113p1411
8323p1516
PT
PTiG
0.04
0.03
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.05
0.00
0.00
Table 6 Glucose disposal, production and flux across the hindquarters in the postabsorptive state before and 24 h after the
start of saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Glucose disposal and production were measured from [3H]glucose. Negative flux represents net uptake. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
Disposal/production/flux ( µmol : min−1 : kg−1)
0h
24 h
ANOVA
PG
Parameter
Control
Endotoxin
Control
Endotoxin
Glucose disposal
Glucose production
Glucose flux
6.7p0.5
3.0p0.9
k3.7p0.6
5.5p0.6
0.9p0.4
k4.6p0.6
5.2p0.9
1.5p1.0
k3.7p0.9
8.5p1.0
1.5p0.8
k7.0p1.0
PT
PTiG
0.05
0.04
Table 7 Protein synthesis, degradation and NB across the hindquarters in the postabsorptive state before and 24 h after the
start of saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Protein synthesis and degradation were measured from [3H2]phenylalanine. Negative flux represents net synthesis. Statistics by ANOVA :
group (PG), time (PT) and timeigroup (PTiG) effects.
Disposal/production/flux (nmol : min−1 : kg−1)
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Phenylalanine disposal (protein synthesis)
Phenylalanine production (protein degradation)
Phenylalanine flux (protein NB)
337p32
297p22
k40p27
456p50
368p50
k88p26
308p18
283p31
k24p23
291p98
527p96
236p11
# 2003 The Biochemical Society and the Medical Research Society
PG
PT
PTiG
0.00
0.00
Organ substrate metabolism in a pig model of endotoxaemia
Table 8 Substrate flux across the portal-drained viscera in the postabsorptive state before and 24 h after the start of saline
(control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Negative flux represents net uptake. Individual GNAA are indicated by *, and essential amino acids by †. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
Flux (nmol : min−1 : kg−1)
0h
24 h
ANOVA
Substrate
Control
Endotoxin
Control
Endotoxin
Glucose
Lactate
Urea
Glutamate*†
Glutamine*†
Glycine*†
Citrulline
Arginine*†
Alanine*†
Tyrosine
Valine†
Isoleucine†
Phenylalanine†
Leucine*†
BCAA
GNAA
AA
k3933p1198
2464p1260
k1189p522
k284p168
k1077p307
1089p546
506p88
398p150
882p277
216p55
323p105
228p116
158p48
328p124
882p304
3102p987
3772p1332
k3926p901
2763p469
k1258p625
k123p51
k1348p152
1319p311
466p48
250p65
1033p128
90p34
252p88
208p84
152p32
326p90
786p169
2935p855
3635p1206
k4319p766
1919p769
672p251
k121p64
k691p61
k37p216
286p44
239p132
436p95
31p35
k39p236
159p62
79p28
201p56
321p86
126p224
764p336
k6170p951
4979p944
425p575
k344p88
k374p149
1043p383
345p55
65p14
186p93
348p74
k276p128
40p15
80p34
61p4
k175p88
793p349
2173p450
PG
PT
PTiG
0.05
0.06
0.03
0.04
0.05
0.05
0.05
0.05
0.00
0.03
0.03
0.03
0.05
0.05
0.05
Table 9 Glucose disposal, production and flux across the portal-drained viscera in the postabsorptive state before and 24 h
after the start of saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Glucose disposal and production were measured from [3H]glucose. Negative flux represents net uptake. Statistics by ANOVA : group
(PG), time (PT) or timeigroup (PTiG) effect.
Disposal/production/flux ( µmol : min−1 : kg−1)
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Glucose disposal
Glucose production
Glucose flux
8.3p2.5
4.4p2.0
k3.9p1.2
7.2p3.1
3.3p1.8
k3.9p0.9
4.1p0.9
k0.2p0.8
k4.3p0.8
6.3p1.2
0.1p1.2
k6.2p0.9
prospects for improved understanding and nutritional
control of the hypermetabolic disorders seen in patients
with sepsis. In the present hyperdynamic porcine model
of sepsis, organ glucose and protein metabolism and
inter-organ amino acid, glucose and lactate fluxes were
defined more specifically. In view of the likelihood that,
in patients, metabolic adaptations to prolonged starvation
accompany the metabolic response to sepsis, a
(semi)starved control group allowed us to distinguish the
effects of endotoxaemia from those of starvation. The
present model of hyperdynamic sepsis was characterized
by accelerated whole-body protein and glucose flux. The
PG
PT
PTiG
0.05
0.05
response to endotoxin differed substantially from that
seen in fasted control pigs, in which muscle and visceral
protein was conserved and glucose synthesis by the liver
was diminished, probably due to a progressive decrease
in energy expenditure and to fat becoming a major
energy-producing substrate. In this model of sepsis, small
endotoxin doses and fluid resuscitation were used,
reproducing cardiovascular changes characteristic of the
hyperdynamic stage of compensated sepsis. We have
shown (M. Poeze, M. J. Bruins, I. J. H. Vriens, G. A. M.
Ten Have, G. Ramsay and N. E. P. Deutz, unpublished
work) that this sepsis model is characterized by main# 2003 The Biochemical Society and the Medical Research Society
135
136
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
Table 10 Protein synthesis, degradation and NB across the portal-drained viscera in the postabsorptive state before and 24 h
after the start of saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Protein synthesis and degradation were measured from [3H2]phenylalanine. Positive flux represents net degradation. Statistics by
ANOVA : group (PG), time (PT) and timeigroup (PTiG) effects.
Disposal/production/flux ( µmol : min−1 : kg−1)
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Phenylalanine disposal (protein synthesis)
Phenylalanine production (protein degradation)
Phenylalanine flux (protein NB)
396p24
554p39
158p48
358p49
510p71
152p32
382p38
462p35
79p28
352p94
433p65
80p34
PG
PT
PTiG
0.03
Table 11 Substrate NB across the liver in the postabsorptive state before and 24 h after the start of saline (control) or
endotoxin infusion
Data are meanspS.E.M. (n l 7). Negative flux represents net uptake. Individual GNAA are indicated by *, and essential amino acids by †. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
Flux (nmol : min−1 : kg−1)
0h
24 h
ANOVA
Substrate
Control
Endotoxin
Control
Endotoxin
Glucose
Lactate
Urea
Glutamate*†
Glutamine†
Glycine†
Citrulline
Arginine†
Alanine†
Tyrosine
Valine†
Isoleucine†
Phenylalanine†
Leucine†
BCAA
GNAA
AA
16 033p2378
k12 294p1915
2594p315
5407p1307
k1112p381
k760p775
k67p128
k232p176
k2929p572
k195p80
561p275
490p261
k66p90
553p193
1759p791
k1620p547
5038p1051
19 057p1469
k13 298p1276
3231p342
6125p403
k1191p362
41p340
97p53
k68p74
k2579p495
k109p49
364p120
389p84
k47p49
416p118
1169p272
k1654p1233
5030p1240
9058p1834
k14 658p1859
2420p455
3202p543
k42p340
k481p1094
k13p53
k356p123
k1743p297
k178p77
k41p104
k79p210
k217p143
k101p246
k70p389
k3504p1303
k1191p1117
18 833p2030
k25 677p1768
6015p1016
4461p626
k1538p312
k2622p1031
k204p205
k543p46
k4010p445
k692p96
k550p64
k291p30
k541p23
k491p73
k1476p223
k12202p1444
k8922p1326
tained mean arterial pressure, blood flow to the central
organs and urinary output, increases in heart rate, cardiac
output and temperature, and decreases in systemic
vascular resistance and oxygen extraction by the splanchnic organs. The pig was chosen as a chronically instrumented large-animal model because organ physiology
and metabolic changes in this animal are in many respects
comparable with those observed in humans [33,34].
However, it has to be taken into account that protein
turnover in the young pig is higher than that in humans
[35,36], which, together with the use of a moderate subacute model of sepsis, may lead to relatively early
# 2003 The Biochemical Society and the Medical Research Society
PG
PT
PTiG
0.05
0.00
0.08
0.05
0.03
0.06
0.03
0.00
0.00
0.04
0.01
0.05
0.00
0.00
manifestation of endotoxaemia-induced modifications in
protein metabolism under feeding conditions.
Disturbances in glucose metabolism during sepsis are
not clearly understood. Acute experimental endotoxaemia resembling septic shock is characterized by insulin
resistance and impaired glucose uptake by peripheral
tissues [37,38], whereas hypermetabolism-associated endotoxaemia is associated with an increased rate of
clearance of glucose by most lymphocyte-rich tissues and
muscle [39,40]. This is in accordance with the enhanced
glucose disposal by the hindquarters and portal-drained
viscera found in the present hyperdynamic model of
Organ substrate metabolism in a pig model of endotoxaemia
Table 12 Glucose disposal, production and flux across the liver in the postabsorptive state before and 24 h after the start of
saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Glucose disposal and production were measured from [3H]glucose. Positive flux represents net release. Statistics by ANOVA : group
(PG), time (PT) and timeigroup (PTiG) effects.
Disposal/production/flux ( µmol : min−1 : kg−1)
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Glucose disposal
Glucose production
Glucose flux
k0.3p1.7
15.7p2.1
16.0p2.4
k0.2p1.4
18.9p1.4
19.1p1.5
k0.4p1.2
8.7p1.5
9.1p1.8
k0.2p1.2
18.6p1.7
18.8p2.0
PG
PT
PTiG
0.04
0.05
Table 13 Protein synthesis, degradation and NB across the liver in the postabsorptive state before and 24 h after the start
of saline (control) or endotoxin infusion
Data are meanspS.E.M. (n l 7). Protein synthesis and degradation were measured from [3H2]valine. Positive flux represents net degradation. Statistics by ANOVA :
group (PG), time (PT) and timeigroup (PTiG) effects.
Disposal/production/flux (nmol : min−1 : kg−1)
0h
24 h
ANOVA
Parameter
Control
Endotoxin
Control
Endotoxin
Valine disposal (protein synthesis)
Valine production (protein degradation)
Valine flux (protein NB)
671p135
1103p114
432p95
649p60
1013p62
520p83
595p65
554p226
k41p104
1014p70
464p70
k550p64
sepsis. During sepsis, glucose is an important energy
substrate for lymphoid, nerve and muscle tissue [40,41].
An increased energy demand by gut enterocytes for the
de novo synthesis of purines and pyrimidines and ribose
sugars for nucleic acid synthesis, and of activated gutassociated monocytes\lymphocytes and spleen lymphocytes [41], is likely to account for the increased visceral
glucose utilization observed during hyperdynamic endotoxaemia. During endotoxaemia, glucose serves as substrate for accelerated glycolysis in muscle tissue. The
increased lactate efflux from the hindquarters and portaldrained viscera during endotoxaemia indicates only
partial glucose oxidation in these tissues during sepsis,
despite enhanced blood flow. The hypothetical changes
in glucose and lactate fluxes between the portal-drained
viscera, the liver and the muscle (estimated from the
hindquarters) induced by hyperdynamic endotoxaemia
are depicted in Figures 2(A) and 2(B) respectively. During
endotoxaemia, total glucose influx into the portal-drained
viscera and muscle (Figure 2B) exceeded glucose efflux
from the liver. Since renal gluconeogenesis was abolished
in endotoxaemic rats and dogs [42–45], it is not expected
that the kidney contributes substantially to glucose
supply. Increased glucose consumption, exceeding glucose production, in hyperdynamic sepsis may be at the
origin of the reduced glucose availability reflected in
PG
PT
PTiG
0.02
0.04
reduced glucose levels. Moreover, the liver consumed
more lactate than was released from the portal-drained
viscera and muscle (as estimated from the hindquarters)
together (compare Figures 2B and 2A), implying that
other organs and erythrocytes participate in the surplus
of lactate production.
Continuous endotoxin infusion accelerated protein
breakdown without affecting protein synthesis in the
hindquarters, accounting for net protein catabolism. If
the hindquarters approximate 50 % of whole-body
muscle, this increase in muscle protein degradation
largely (" 86 %) accounted for the increase in wholebody protein degradation, emphasizing that catabolism
of muscle protein primarily determines whole-body
protein catabolism. Whereas the regulatory mechanisms
that affect muscle protein degradation during injury and
sepsis have been studied extensively [46], endotoxininduced changes in protein synthesis remain relatively
unidentified. A number of rat studies using the caecal
ligation and puncture sepsis model have shown impaired
protein synthesis in muscle [47–49], whereas the present
hyperdynamic porcine model of sepsis showed unchanged protein synthesis in the hindquarters. It therefore appears that changes in muscle protein turnover rate
during sepsis depend on the severity of the septic insult
and circulatory response implicated in the sepsis model.
# 2003 The Biochemical Society and the Medical Research Society
137
138
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
Figure 2 Fluxes of glucose (GLC) and lactate (LAC) across muscle, portal-drained viscera and liver during a 24 h infusion of
(A) saline (control) and (B) endotoxin
Fluxes are given in µmol : min−1 : kg−1 body weight. Muscle flux is estimated as 2i that across the hindquarters. The thickness of the arrows is proportional to
the substrate flux.
Figure 3 Fluxes of alanine (ALA), glutamate (GLU) and glutamine (GLN) across muscle, portal-drained viscera and liver during
a 24 h infusion of (A) saline (control) and (B) endotoxin
Fluxes are given in µmol : min−1 : kg−1 body weight. Muscle flux is estimated as 2i that across the hindquarters. The thickness of the arrows is proportional to
the substrate flux.
This possibly depends on whether or not the septic
animal is adequately resuscitated. Moreover, differences
in protein turnover rates may occur depending on the
isotopic method used, i.e. a flooding dose compared with
the continuous infusion method [50].
The hypothetical glutamate, glutamine and alanine
inter-organ fluxes are illustrated in Figures 3(A) and 3(B).
Release of both glutamine and alanine from the hindquarters was increased during endotoxaemia (Figure 3B
compared with Figure 3A). Assuming that the hind# 2003 The Biochemical Society and the Medical Research Society
quarters represent 50 % of the muscle tissue, and that in
porcine muscle tissue the molecular phenylalanine\
glutamine and phenylalanine\alanine ratios are 3.8 : 1 and
3.4 : 1 respectively [51], efflux of glutamine and alanine
from the hindquarters was " 0.4 and " 1.3 µmol :
min−" : kg−" higher than would be anticipated from
protein breakdown during endotoxaemia. This indicates
that, during sepsis, intracellular glutamate serves increased de novo synthesis of alanine and glutamine.
During endotoxaemia, the net output of alanine and
Organ substrate metabolism in a pig model of endotoxaemia
glutamine from the muscle exceeded the net consumption by the splanchnic organs (Figure 3B compared
with Figure 3A). The ‘ excess ’ glutamine is probably used
by the kidney and lymphoid tissue [52], while the excess
alanine may be involved as a substrate for brain and nerve
tissue [53]. Net efflux of BCAA from the hindquarters
was observed during endotoxaemia, indicating that the
rate of release of BCAA derived from accelerated protein
degradation exceeded the increase in BCAA consumption by the hindquarters. Increased BCAA transamination and further oxidation is a major source of
glutamate and energy [54], and thereby constitutes a
major cause of amino acid loss during sepsis [55,56].
The net consumption of glutamate and BCAA by the
portal-drained viscera was increased, implying that under
endotoxaemic conditions the transamination and oxidation of BCAA, in which glutamate is used as an
intermediate, constitutes an important source of energy
in the visceral organs. Endotoxaemia reduced both
visceral glutamine net uptake and concomitant alanine
net release, despite maintained visceral blood flow.
Possibly, reduced glutaminase activity in the gut [57,58]
accounts for the decreased conversion of glutamine into
glutamate and subsequent glutamate transamination
into alanine. Whereas under normal conditions glutamine
is a major precursor of citrulline in the gut, under endotoxaemic conditions the visceral citrulline efflux may be
derived from arginine via the endotoxin-induced nitric
oxide synthase pathway. Accordingly, visceral arginine
consumption was elevated during endotoxaemia. The
portal-drained viscera released a substantial amount of
glycine in response to sepsis, which accounted for a large
part of the uptake of gluconeogenic precursors by the
liver. The high glycine efflux rates may originate from
biliary glycoprotein-rich mucins, which support the gut
microbial clearance system [59]. Both visceral protein
synthesis and protein degradation remained unaffected
during hyperdynamic endotoxaemia. This contrasts with
the increased intestinal protein synthesis demonstrated in
rats that were subjected to caecal ligation and puncture
[60]. Because of peritoneal contamination, caecal ligation
and puncture may elicit a more severe stress response
involving acute-phase protein synthesis in the portaldrained viscera, accounting for the discrepancy in the
observations.
During accelerated gluconeogenesis, amino acid and
lactate carbons serve as precursors for glucose, whereas
the nitrogen moieties of the deaminated amino acids that
are detoxified are responsible for increased urea formation, as is observed during endotoxaemia [61]. The
amino acids that are taken up by the liver are involved in
not only gluconeogenesis, but protein synthesis as well.
Liver protein synthesis in this endotoxaemia model was
increased, reflecting the enhanced synthesis of acutephase reactive proteins [3,62]. Although we could not
detect any changes in haptoglobin or fibrinogen con-
centrations, the total concentration of acute-phase
proteins increased during endotoxaemia. However, since
plasma protein concentrations depend on the intravascular pool size, and thus vary with hydration state,
fluid shifts and renal excretion rate, they do not appear to
be a good reflection of protein synthesis rates under
disease conditions such as sepsis. Glutamate release from
the liver was enhanced during sepsis, constituting an
important nitrogen-sparing mechanism, since glutamate
serves as a nitrogen intermediate for the synthesis of
alanine and glutamine in the muscle. During sepsis, only
60 % of the glutamate efflux from the liver was taken
up by muscle (estimated from the hindquarters) and
visceral organs (Figure 3B), suggesting a major role for
the nervous system and the kidney in utilization of the
excess.
Although impaired oxygen extraction by the splanchnic organs and the hindquarters was observed in an
identical hyperdynamic porcine model of sepsis (M.
Poeze, M. J. Bruins, I. J. H. Vriens, G. A. M. Ten Have,
G. Ramsay and N. E. P. Deutz, unpublished work),
blood flow and oxygen delivery to these organs was
sustained or even increased. This implies that substrate
delivery to the organs is maintained during hyperdynamic endotoxaemia, although this does not ensure
intracellular substrate availability per se. Net protein loss
in muscle is effected by a doubling of protein degradation,
whereas net protein gain in the liver is effected largely by
an increase (1.7-fold) in protein synthesis. The trade-off
is negative, leading to net protein loss at a whole-body
level. Not only muscle, but also the portal-drained
viscera, participate significantly in gluconeogenic substrate delivery to the liver by releasing lactate and amino
acids. Recycling of carbons from gluconeogenic precursors to glucose by the liver constitutes an important
glucose-sparing mechanism. Moreover, accelerated gluconeogenesis and nitrogen loss as urea in hyperdynamic
sepsis suggest that depletion of essential amino acids may
develop during hypermetabolic compensated sepsis. In
addition, amino acids such as arginine, glutamate and
glutamine are not only used in gluconeogenesis, but are
also required as precursors for activated immune cells
and nerve tissues, which may contribute significantly to
the depletion of these immunoactive amino acids. Increased utilization of BCAA in the portal-drained viscera
and the hindquarters suggests that these essential amino
acids may in the long term become unable to meet the
energy requirements of these tissues.
In conclusion, the present data may assist in defining
increased substrate requirements during hypermetabolic
sepsis, based on which substrate supplements can be
recommended for septic patients. Our results suggest
that the requirements for BCAA, and specifically the
immunoactive amino acids, may be elevated, and that
supplementation of these substrates may be advantageous
in patients with hypermetabolic sepsis.
# 2003 The Biochemical Society and the Medical Research Society
139
140
M. J. Bruins, N. E. P. Deutz and P. B. Soeters
ACKNOWLEDGMENTS
We express many thanks to K. Slot, H. M. H. van Eijk
and J. L. J. M. Scheijen for analytical help, and G. A. M.
Ten Have for assistance during experimental procedures.
This study was supported by grant 902-23-098 from The
Netherlands Organization for Scientific Research.
20
21
22
REFERENCES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Wilmore, D. W. (1983) Alterations in protein,
carbohydrate, and fat metabolism in injured and septic
patients. J. Am. Coll. Nutr. 2, 3–13
Matthews, D. E. and Battezzati, A. (1993) Regulation of
protein metabolism during stress. Curr. Opin. Gen. Surg.
72–77
Clowes, G. H. A. (1986) Amino acid transfer between
muscle and the visceral tissues in man during health and
disease. In Problems and Potential of Branched-Chain
Amino Acids in Physiology and Medicine (Odessey, R.,
ed.), pp. 299–334, Elsevier Science Publishers B.V.,
Amsterdam
Newsholme, E. A., Newsholme, P. and Curi, R. (1987)
The role of the citric acid cycle in cells of the immune
system and its importance in sepsis, trauma and burns.
Biochem. Soc. Symp. 54, 145–162
Januszkiewicz, A., Essen, P., McNurlan, M. A., Ringden,
O., Garlick, P. J. and Wernerman, J. (2001)
Determination of in vivo protein synthesis in human T
lymphocytes. Clin. Nutr. 20, 181–182
Border, J. R. (1970) Metabolic response to short-term
starvation, sepsis, and trauma. Surg. Annu. 2, 11–34
Kinney, J. M., Gump, F. E. and Long, C. L. (1972)
Energy and tissue fuel in human injury and sepsis. Adv.
Exp. Med. Biol. 33, 401–407
Woolf, L. I., Groves, A. C. and Duff, J. H. (1979) Amino
acid metabolism in dogs with E. coli bacteremic shock.
Surgery 85, 212–218
Spitzer, J. J., Bagby, G. J., Meszaros, K. and Lang, C. H.
(1989) Altered control of carbohydrate metabolism in
endotoxemia. Prog. Clin. Biol. Res. 286, 145–165
Gore, D. C., Jahoor, F., Hibbert, J. M. and DeMaria, E. J.
(1996) Lactic acidosis during sepsis is related to increased
pyruvate production, not deficits in tissue oxygen
availability. Ann. Surg. 224, 97–102
Meszaros, K., Bojta, J., Bautista, A. P., Lang, C. H. and
Spitzer, J. J. (1991) Glucose utilization by Kupffer cells,
endothelial cells, and granulocytes in endotoxemic rat
liver. Am. J. Physiol. 260, G7–G12
Fischer, J. E. and Hasselgren, P. O. (1991) Cytokines and
glucocorticoids in the regulation of the ‘‘ hepato-skeletal
muscle axis ’’ in sepsis. Am. J. Surg. 161, 266–271
Streat, S. J., Beddoe, A. H. and Hill, G. L. (1987)
Aggressive nutritional support does not prevent protein
loss despite fat gain in septic intensive care patients.
J. Trauma 27, 262–266
Streat, S. J. and Hill, G. L. (1987) Nutritional support in
the management of critically ill patients in surgical
intensive care. World J. Surg. 11, 194–201
Michie, H. R. (1996) Metabolism of sepsis and multiple
organ failure. World J. Surg. 20, 460–464
Chiolero, R., Revelly, J. P. and Tappy, L. (1997) Energy
metabolism in sepsis and injury. Nutrition 13, 45S–51S
Qureshi, S. T., Gros, P. and Malo, D. (1999) The Lps
locus : genetic regulation of host responses to bacterial
lipopolysaccharide. Inflamm. Res. 48, 613–620
Fink, M. P. and Heard, S. O. (1990) Laboratory models
of sepsis and septic shock. J. Surg. Res. 49, 186–196
Deutz, N. E. P., Reijven, P. L. M., Athanasas, G. and
Soeters, P. B. (1992) Post-operative changes in hepatic,
intestinal, splenic and muscle fluxes of amino acids and
ammonia in pigs. Clin. Sci. 83, 607–614
# 2003 The Biochemical Society and the Medical Research Society
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Ten Have, G. A. M., Bost, M. C. F., Suyk-Wierts,
J. C. A. W., van den Bogaart, A. E. J. M. and Deutz,
N. E. P. (1996) Simultaneous measurement of metabolic
flux in portally-drained viscera, liver, spleen, kidney and
hindquarter in the conscious pig. Lab. Anim. 30, 347–358
Van Eijk, H. M. H., Rooyakkers, D. R. and Deutz,
N. E. P. (1993) Rapid routine determination of amino
acids in plasma by high-performance liquid
chromatography with a 2–3 µm Spherisorb ODS II
column. J. Chromatogr. 620, 143–148
Van Eijk, H. M. H., Rooyakkers, D. R. and Deutz,
N. E. P. (1996) Automated determination of polyamines
by high-performance liquid chromatography with simple
sample preparation. J. Chromatogr. 730, 115–120
Cook, H. T., Bune, A. J., Jansen, A. S., Taylor, G. M.,
Loi, R. K. and Cattell, V. (1994) Cellular localization of
inducible nitric oxide synthase in experimental endotoxic
shock in the rat. Clin. Sci. 87, 179–186
Deutz, N. E. P., Bruins, M. J. and Soeters, P. B. (1998)
Infusion of soy and casein protein meals affect interorgan
amino acid metabolism and urea kinetics differently in
pigs. J. Nutr. 128, 2435–2445
Wolfe, R. R. (1992) Radioactive and Stable Isotope
Tracers in Biomedicine. Principles and Practice of Kinetic
Analysis, Wiley-Liss, New York
Wajngot, A., Chandramouli, V., Schumann, W. C.,
Kumaran, K., Efendic, S. and Landau, B. R. (1989)
Testing of the assumptions made in estimating the extent
of futile cycling. Am. J. Physiol. 256, E668–E675
Deutz, N. E. P., Wagenmakers, A. J. M. and Soeters, P. B.
(1999) Discrepancy between muscle and whole body
protein turnover. Curr. Opin. Clin. Nutr. Metab. Care 2,
29–32
Biolo, G., Fleming, R. Y. D., Maggi, S. and Wolfe, R. R.
(1995) Transmembrane transport and intracellular kinetics
of amino acids in human skeletal muscle. Am. J. Physiol.
268, E75–E84
Tourian, A., Goddard, J. and Puck, T. T. (1969)
Phenylalanine hydroxylase activity in mammalian cells.
J. Cell. Physiol. 73, 159–170
Shinnick, F. L. and Harper, A. E. (1976) Branched-chain
amino acid oxidation by isolated rat tissue preparations.
Biochim. Biophys. Acta 437, 477–486
Seglen, P. O. and Solheim, A. E. (1978) Effects of
aminooxyacetate, alanine and other amino acids on
protein synthesis in isolated rat hepatocytes. Biochim.
Biophys. Acta 520, 630–641
Ooiwa T., Goto, H., Tsukamoto, Y. et al. (1995)
Regulation of valine catabolism in canine tissues : tissue
distributions of branched-chain aminotransferase and 2oxo acid dehydrogenase complex, methacrylyl-CoA
hydratase and 3-hydroxyisobutyryl-CoA hydrolase.
Biochim. Biophys. Acta 1243, 216–220
Dodds, W. J. (1982) The pig model for biomedical
research. Fed. Proc. Fed. Am. Soc. Exp. Biol. 41, 247–256
Miller, E. R. and Ullrey, D. E. (1987) The pig as a model
for human nutrition. Annu. Rev. Nutr. 7, 361–382
Garlick, P. J., Burk, T. L. and Swick, R. W. (1976)
Protein synthesis and RNA in tissues of the pig. Am. J.
Physiol. 230, 1108–1112
Waterlow, J. C. (1980) Protein turnover in the whole
animal. Invest. Cell. Pathol. 3, 107–119
Lang, C. H., Dobrescu, C. and Meszaros, K. (1990)
Insulin-mediated glucose uptake by individual tissues
during sepsis. Metab. Clin. Exp. 39, 1096–1107
Lang, C. H. (1992) Sepsis-induced insulin resistance in
rats is mediated by a beta-adrenergic mechanism. Am. J.
Physiol. 263, E703–E711
Lang, C. H., Bagby, G. J. and Spitzer, J. J. (1984)
Carbohydrate dynamics in the hypermetabolic septic rat.
Metab. Clin. Exp. 33, 959–963
Meszaros, K., Lang, C. H., Bagby, G. J. and Spitzer, J. J.
(1988) In vivo glucose utilization by individual tissues
during nonlethal hypermetabolic sepsis. FASEB J. 2,
3083–3086
Newsholme, E. A. and Carrie, A. L. (1994) Quantitative
aspects of glucose and glutamine metabolism by intestinal
cells. Gut 35, S13–S17
Organ substrate metabolism in a pig model of endotoxaemia
42
43
44
45
46
47
48
49
50
51
52
Archer, L. T., Benjamin, B., Lane, M. M. and Hinshaw,
L. B. (1976) Renal gluconeogenesis and increased glucose
utilization in shock. Am. J. Physiol. 231, 872–879
Ardawi, M. S. M., Khoja, S. M. and Newsholme, E. A.
(1990) Metabolic regulation of renal gluconeogenesis in
response to sepsis in the rat. Clin. Sci. 79, 483–490
Maitra, S. R., Homan, C. S., Pan, W., Geller, E. R.,
Henry, M. C. and Thode, H. C. (1996) Renal
gluconeogenesis and blood flow during endotoxic shock.
Acad. Emerg. Med. 3, 1006–1010
Hallemeesch, M. M., Cobben, D. C. P., Deutz, N. E. P.
and Soeters, P. B. (1998) Liver increases glucose
production, whereas kidney decreases glucose production
after endotoxin challenge. Hepatology 28, 384A (abstract)
Price, S. R. and Mitch, W. E. (1998) Mechanisms
stimulating protein degradation to cause muscle atrophy.
Curr. Opin. Clin. Nutr. Metab. Care 1, 79–83
Hasselgren, P. O., Jagenburg, R., Karlstrom, L., Pedersen,
P. and Seeman, T. (1984) Changes of protein metabolism
in liver and skeletal muscle following trauma complicated
by sepsis. J. Trauma 24, 224–228
Breuille, D., Arnal, M., Rambourdin, F., Bayle, G.,
Levieux, D. and Obled, C. (1998) Sustained modifications
of protein metabolism in various tissues in a rat model of
long-lasting sepsis. Clin. Sci. 94, 413–423
O’Leary, M. J., Ferguson, C. N., Rennie, M. J., Hinds,
C. J., Coakley, J. H. and Preedy, V. R. (2001) Sequential
changes in in vivo muscle and liver protein synthesis and
plasma and tissue glutamine levels in sepsis in the rat.
Clin. Sci. 101, 295–304
Rennie, M. J., Smith, K. and Watt, P. W. (1994)
Measurement of human tissue protein synthesis : an
optimal approach. Am. J. Physiol. 266, E298–E307
Wu, G., Ott, T. L., Knabe, D. A. and Bazer, F. W. (1999)
Amino acid composition of the fetal pig. J. Nutr. 129,
1031–1038
Ardawi, M. S., Majzoub, M. F., Kateilah, S. M. and
Newsholme, E. A. (1991) Maximal activity of phosphatedependent glutaminase and glutamine metabolism in
septic rats. J. Lab. Clin. Med. 118, 26–32
53
54
55
56
57
58
59
60
61
62
Erecinska, M., Nelson, D., Nissim, I., Daikhin, Y. and
Yudkoff, M. (1994) Cerebral alanine transport and alanine
aminotransferase reaction : alanine as a source of neuronal
glutamate. J. Neurochem. 62, 1953–1964
Spydevold, O. and Hokland, B. (1981) Oxidation of
branched-chain amino acids in skeletal muscle and liver
of rat. Effects of octanoate and energy state. Biochim.
Biophys. Acta 676, 279–288
Groves, A. C., Woolf, L. I., Duff, J. H. and Finley, R. J.
(1983) Metabolism of branched-chain amino acids in dogs
with Escherichia coli endotoxin shock. Surgery 93,
273–278
Nawabi, M. D., Block, K. P., Chakrabarti, M. C. and
Buse, M. G. (1990) Administration of endotoxin, tumor
necrosis factor, or interleukin 1 to rats activates skeletal
muscle branched-chain alpha-keto acid dehydrogenase.
J. Clin. Invest. 85, 256–263
Souba, W. W., Herskowitz, K., Klimberg, V. S. et al.
(1990) The effects of sepsis and endotoxemia on gut
glutamine metabolism. Ann. Surg. 211, 543–549
(discussion 549–551)
Austgen, T. R., Chen, M. K., Dudrick, P. S., Copeland,
E. M. and Souba, W. W. (1992) Cytokine regulation of
intestinal glutamine utilization. Am. J. Surg. 163, 174–179
(discussion 179–180)
Choi, J., Klinkspoor, J. H., Yoshida, T. and Lee, S. P.
(1999) Lipopolysaccharide from Escherichia coli
stimulates mucin secretion by cultured dog gallbladder
epithelial cells. Hepatology 29, 1352–1357
Zamir, O., Hasselgren, P. O., Higashiguchi, T., Frederick,
J. A. and Fischer, J. E. (1992) Effect of sepsis or cytokine
administration on release of gut peptides. Am. J. Surg.
163, 181–185
Brosnan, J. T. (2000) Glutamate, at the interface between
amino acid and carbohydrate metabolism. J. Nutr. 130,
988S–990S
Von Allmen, D., Hasselgren, P. O., Higashiguchi, T. and
Fischer, J. E. (1992) Individual regulation of different
hepatocellular functions during sepsis. Metab. Clin. Exp.
41, 961–965
Received 3 October 2002; accepted 1 November 2002
# 2003 The Biochemical Society and the Medical Research Society
141