Clinical Science (1995) 88, 235-242 (Printed in Great Britain)
235
Acute stimulation of albumin synthesis rate with oral
meal feeding in healthy subjects measured with
[ring2H,Iphenylalani ne
K. A. HUNTER, P. E. BALLMER*, S. E. ANDERSON, J. BROOM?, P. J. GARLICK$
and M. A. McNURLANS
Rowett Research Institute. Aberdeen, U.K., *Department of Internal Medicine. University of
Berne, Switzerland. ?Department of Clinical Biochemistry, University of Aberdeen, U.K., and
$Department of Surgery, SUNY, Stony Brook. New York, U.S.A.
(Received 29 March/l I October 1994; accepted 18 October 1994)
1. The short-term effect of oral feeding on albumin
synthesis rate was investigated in 12 healthy volunteers using two meal regimens. Albumin synthesis was
measured over 90min after injection of a ‘flooding’
amount (43 mg/kg body weight) of phenylalanine
enriched to 7.5, 10 or 15 atoms % with the stable
isotope [ring-’H,] phenylalanine.
2. In one set of subjects, consumption of five small
hourly meals resulted in a consistent and significant
increase (P<0.05) in albumin fractional synthesis
rate from a mean (kSEM) fasting value of 5.8
( f 0.4) %/day to 7.1 ( k0.4) %/day in the fed state.
3. A second study in which albumin synthesis was
measured 30min after consumption of a single larger
meal was carried out in another set of volunteers. The
fractional rate of albumin synthesis was again significantly elevated after feeding (P<0.05), rising from
7.1 (kO.4) %/day in the fasted state to 9.1
( kO.6) %/day in the fed state. In both studies, similar
responses were observed in the absolute rate of
albumin synthesis (mg day-’ kg-I).
4. Albumin secretion time was significantly shorter
(P<0.05) after feeding in both studies, suggesting
that the acute stimulation in albumin synthesis
observed after feeding may in part be mediated via a
post-transcriptional mechanism.
5. The response of total liver protein synthesis to oral
feeding was investigated in an animal model employing adult rats studied with a flooding amount of
[2,6-’H]phenylalanine.
6. The results indicated a stimulation of 20% with no
difference in the proportion of albumin synthesis
relative to total liver synthesis, determined from the
immunoprecipitation of albumin from the liver.
INTRODUCTION
Within clinical practice, the observed relationship
between serum albumin concentration and nutrient
intake has led to the widespread use of serum
albumin as an indicator of nutritional status [l, 23.
Although this approach has been criticized [3, 41,
serum albumin concentration is still routinely determined and used in conjunction with other measurements such as plasma transferrin as a marker of
protein-energy malnutrition. The measurement of
one of the processes which regulates serum albumin
concentration, albumin synthesis, can provide a
more dynamic insight into the response of albumin
metabolism to factors such as trauma, pregnancy
and nutrition [S-71 than would be obtained by
assessing changes in albumin concentration alone.
Before investigating the effects of nutrition on
albumin synthesis in the longer term, it is important
to establish how an intake of nutrients affects the
normal acute regulation of albumin synthesis.
Although there is a wealth of evidence suggesting
that albumin synthesis is sensitive to nutrient intake
in terms of long-term adaptations to varied amount
and composition [S, 91, these studies have not
examined the immediate response of albumin synthesis to nutrient intake. The lack of information
regarding acute regulation has arisen mainly
because of methodological inadequacies, as in the
past the techniques available for determining albumin synthesis rates in human subjects were carried
out over time intervals which were too long to
detect rapid changes in synthesis rate [e.g. 10, 111.
The ‘flooding’ technique for measurement of albumin synthesis rate developed by Ballmer et al. [12]
is carried out over a much shorter period and has
enabled us to examine the acute effects of oral
feeding on hepatic albumin production in healthy
human subjects. In the present study, the short-term
response of hepatic albumin output to oral feeding
was assessed by determining albumin synthesis rates
in healthy subjects during one of two feeding regimens and comparing these with their synthesis rates
measured in the post-absorptive state. The feeding
Key words: Albumin synthesis, feeding, stable isotopes.
Abbreviations: ASR, absolute synthesis rate; BMI, body mass index; FSR. fractional synthesis rate.
Correspondence: Kirsty A. Hunter, Rowett Research Institute. Bucksburn, Aberdeen, U.K.
236
K. A. Hunter et al.
regimens were made up of a number of small meals
consumed over several hours to provide a constant
level of nutrient intake and a single larger, more
physiologically sized meal, designed so that the
normal response of albumin synthesis after consumption of meals which were physiological in
terms of composition, size and method of ingestion
could be investigated. Albumin synthesis rates were
determined by the technique described by Ballmer
et al. [12] but modified to use the stable isotope
L-[*H,]phenylalanine. This isotope has a number of
methodological advantages compared with the isotope
used
previously,
~-[‘~C]leucine, and
the validation procedures carried out before
~-[~H,]phenylalanine was employed in metabolic
studies have been described in full elsewhere [13].
As it was not possible to determine whether the
acute changes in albumin synthesis after feeding
were the result of an overall stimulation of liver
protein synthesis or a specific response of albumin,
an animal experiment was undertaken to establish,
firstly, if there was a similar response to feeding in
the adult rat and, secondly, if this response was
unique to albumin or a non-specific one reflecting a
general response of total liver protein synthesis.
Part of this work has been published previously
in abstract form [14].
MATERIALS A N D METHODS
Materials and instrumentation
~-[~H,]phenylalanineand L-phenylalanine were
obtained from Tracer Technologies, Somerville, MA,
U.S.A., and Ajinomoto, Tokyo, Japan, respectively.
Reagents for decarboxylation and derivatization of
phenylalanine were obtained from Sigma, Poole,
U.K., and Pierce (U.K.), Cambridge, U.K. Mass
spectrometry was performed using a VG 12-253
quadrupole mass spectrometer coupled to a Hewlett
Packard 5890 gas chromatograph (g.c.m.s.) under
selected ion monitoring conditions.
Subjects and clinical protocols
Study protocols were approved by the Joint Ethical Committee of Grampian Health Board and the
University of Aberdeen, U.K., and subjects gave
written informed consent prior to commencement of
study.
Subjects recruited were in good health with blood
biochemical parameters (e.g. fasting lipids and lipoproteins, sodium, phosphate, urea, urate, aspartate
aminotransferase, y-glutamyltransferase and lactate
dehydrogenase) and body mass indices (BMIs)
(wt./h2) within the normal ranges. Six volunteers
[four male, two female, aged 3 4 1 3 (SEM) years,
BMI 21 f 13 took part in study 1, while another six
(all male, aged 23&2 years, BMI 2 3 k 1) were
recruited for study 2. In both studies, the subjects’
albumin synthesis rates were measured on two
separate occasions, once in the post-absorptive state
and once in the fed state. The two measurements
were made several weeks apart and in random
order.
In study 1, subjects received ~-[*H,]phenylalanine (43mg/kg) enriched to 10 atoms % as a 2%
(w/w) solution in 0.45% (w/w) saline on both
occasions. In study 2, the same solution was administered to subjects except that the enrichments
were 7.5 atoms % for the first measurement and 15
atoms ”/, for the second measurement. All solutions
were sterilized by filtration through a 0.22-pm-pore
single-use filter (Millipore, Molsheim, France) before
injection.
In both studies, meals were made up of white
bread rolls, butter, jam, cheese, cottage cheese and
fresh orange juice. In study 1, each meal contained
one-twelfth of the subject’s daily energy requirement, and in study 2 each meal provided one-third
of the daily requirement.
Study 1. For the post-absorptive measurements,
subjects had fasted for 10h and remained so
throughout the period of measurement. At 10.00
hours the isotope solution was given at a constant
rate over 10min via a cannula inserted in a forearm
vein. A series of blood samples were taken from a
second cannula inserted in a vein in the opposite
forearm before and at 5, 10, 15, 20, 30, 50, 70 and
90 min after the start of injection for determination
of plasma free phenylalanine enrichment and enrichment of phenylalanine isolated from albumin.
Measurements in the fed state were carried out
during the later stage of a feeding protocol in which
subjects consumed five small meals beginning at
07.00 hours and continuing one every hour until
11.00 hours. Injection of the isotope solution began
at 10.00 hours and blood samples were taken at the
times described above until 11.30 hours. Meals were
calculated to each contain one-twelfth of the individual’s daily energy requirement. Daily energy expenditure was estimated as 1.4 x basal metabolic rate,
where basal metabolic rate was derived from 1985
F A 0 equations [l5] using sex, age, weight and
height. In order to simulate a ‘typical’ British diet,
12% of energy was derived from protein, 48% from
carbohydrate and 40% from fat.
Study 2. Measurement of albumin synthesis rates
in the post-absorptive state were carried out as in
study 1. The feeding protocol for the determination
of albumin synthesis rate in the fed state differed in
that subjects received one large meal rather than the
series of small meals. This meal was of the same
composition of protein, fat and carbohydrate as
mentioned previously, but was of a more physiological size, containing one-third of the individuals’
predicted daily energy requirement, than the smaller
meals consumed in study 1. Subjects consumed the
meal at 09.30 hours, and after 30min the isotope
solution was injected and blood samples taken at
the same times as specified above.
Before both studies, subjects were asked to complete a record of their dietary intake over several
Albumin synthesis and meal feeding
237
remove any traces. Isolation and derivatization of
phenylalanine has previously been described [181.
Phenylalanine was enzymatically converted to phenylethylamine by incubating with tyrosine decarboxylase (EC 4.1.1.25). After extraction of phenylethylamine, it was subsequently derivatized to the
heptafluorobutyryl derivative. A sample (1 pl) was
then injected on to the g.c.m.s. system operating
under electron impact conditions in the splitless
mode and the ions at m/z106 and 109, corresponding to the M 2 and M 5 ions, were monitored.
Serum albumin concentrations were measured
using the Bromocresol Green method [191. Plasma
insulin concentrations were determined by radioimmunoassay using serum, tracer and standards [20].
+
+
Calculations and statistics
I
2
3
4
Fig. I. SDS-PAGE analysis with silver staining of albumin isolated
from the serum of a healthy donor by differential solubility in
ethanol in the presence of trichloroacetic acid. Lanes 1-3 contain I .2,
0.8 and 0.4pg of isolated albumin and lane 4 contains I pg of human serum
albumin standard.
days using precision scales (Soehle) and a written
diary. This was to ensure that the habitual diets of
these subjects were not extreme in either protein or
energy content. Diaries were analysed using the
Microdiet computer package (Salford University,
U.K.).
Sample preparation
Determination of plasma free phenylalanine
enrichment was carried out essentially as described
by Calder and Smith [16] for leucine. Briefly, free
amino acids were isolated from plasma samples
using a cation exchange column, eluted with 4 mol/l
ammonium hydroxide and then lyophilized. The
tertiary-butyldimethylsilyl derivative was formed
and enrichment measured by g.c.m.s. The ions at
mlz336 and 341 corresponding to the M-57 and
M-57 5 ions were monitored.
To determine the 'H, enrichment of phenylalanine in albumin, albumin was isolated from serum
by differential solubility in absolute ethanol from
trichloroacetic-(lo%, w/w) precipitated protein [171.
The purity of isolated albumin was confirmed by
SDS/PAGE visualized by silver staining. A picture
of the gel is shown in Fig. 1. After a series of
washing procedures with 2% perchloric acid, albumin was hydrolysed in 6mol/l HCl for 24h at
110°C. The acid was subsequently removed by
rotary evaporation with a number of washes to
+
Albumin fractional synthesis rates (FSRs), that is
the percentage of the intravascular albumin pool
synthesized per day, were calculated using the
formula FSR=(P, -PI) x 100/A. The derivation
and validity of assumptions for the use of the
formula has been discussed in detail by Ballmer et
al. [12]. P, and PI represent phenylalanine enrichment in albumin at two time points (90 and 50min
respectively) and A is the area under the curve of
plasma free phenylalanine enrichment with time
between the time points P,-T, and P,-T,. T, is
the secretion time for albumin, that is the time
taken for albumin to be synthesized and processed
within the hepatocyte. It is determined by plotting
the change in enrichment in albumin with time and
carrying out linear regression analysis on the part of
the curve which is judged by eye to be linear,
usually between 50 and 90min (see Fig. 3), and
determining the intercept of this line with the time
axis.
Absolute synthesis rate, which is the amount of
albumin synthesized per day expressed as a weight,
was calculated as the product of the fractional
synthesis rate and the intravascular albumin mass.
Intravascular albumin mass was estimated from
measured plasma albumin concentrations and estimated plasma volume. Since it was not feasible to
measure plasma volume directly in these healthy
subjects because of the need for radioisotopes, a
normogram was employed which uses sex, age and
weight to give an estimate [21]. We consider this to
be a reasonable estimate for subjects in the fasted
and fed states, having consulted the literature [22],
and from our own observation that meal feeding in
conjunction with a mock flooding procedure did not
induce any significant change in haematocrit or
serum albumin concentration.
Data are presented as means 1 S E M unless
otherwise stated. Differences in the fasted and fed
states were evaluated by paired t-test and taken to
be significant if P<O.O5. Since the number of subjects was small, differences were also analysed nonparametrically using the Wilcoxon signed-rank test.
238
K. A. Hunter et al.
For all analyses, both tests gave the same significant
differences.
Determination of the acute effect of feeding on albumin
synthesis as a proportion of total liver synthesis in the
adult rat
Animals and protocol. Animals used were adult
female Wistar rats weighing 308 f6 g (mean
+SEM) that had been maintained on a standard
chow diet and divided into two groups of equal
body weight containing eight animals per group.
The diet was primarily wheat and maize based and
provided 17.9% crude protein and 13.3 MJ/kg
digestible energy. Food was removed from both
groups at 20.00 hours on the evening before the
experiment. The next morning, animals in the 'fed'
group were given access to their normal diet for
50min. Immediately after this a flooding amount of
[2,6-3H]phenylalanine (150 pmo1/100 g body weight,
30 pCi/lOO g; Amersham International, Amersham,
Bucks, U.K.) was injected into the tail vein for
determination of protein synthesis. Animals were
sacrificed exactly 10min later and the liver quickly
excised and frozen in liquid nitrogen [23]. The same
protocol was carried out for the other group of
animals, but they remained fasted until death.
Sample preparation. Measurement of albumin synthesis as a proportion of total liver protein synthesis
was carried out essentially as described by Pain et
al. [24]. Briefly, weighed samples of frozen liver
were homogenized in 0.35 mol/l sucrose in 0.05 mol/l
Tris-HC1 pH 7.4. Protein was precipitated from
100 pl of homogenate by adding 10% trichloroacetic
acid, and after a number of washing steps samples
were dissolved in 0.3 mol/l NaOH and used (without
hydrolysis) to determine radioactivity incorporated
into total protein.
An immunoprecipitation technique was used to
estimate the radioactivity incorporated into albumin. After centrifuging 500 p1 of homogenate to
obtain the post-nuclear supernatant, 250 p1 of sheep
anti-rat albumin (The Binding Site, Birmingham,
U.K.) was added. Samples were incubated for 1 h at
room temperature and then for 3 h at 4°C. A 100-pl
aliquot of a proteinA-Sepharose slurry (Sigma) was
then added to improve quantitative immunoprecipitation. After 45 min the samples were centrifuged
and the immunoprecipitate was washed several
times to remove contaminating radioactivity.
Radioactivity incorporated into albumin was then
measured after dissolving the immunoprecipitate in
0.3 mol/l NaOH.
The fractional rate of total liver protein synthesis
was determined as described by Garlick et al. [23]
from the specific activity of free phenylalanine in
liver homogenate and phenylalanine in protein.
Calculations. Albumin as a percentage of total
liver protein synthesis was calculated as the specific
radioactivity in albumin divided by the specific
O+
,
I
0
15
30
I
45
60
75
90
Time (mins)
"y
% 20
O+
0
I
I
!
I
I
I
15
30
45
Time (mins)
60
75
90
Fig. 2. Changes (means fSD) in plasma free phenylalanine enrichment in (a) study I and (b) study 2 in subjects in the fasted (& and
fed
states.
(m
radioactivity in total liver protein. The fractional
rate of liver protein synthesis ( k J , that is the
percentage of the total liver protein pool synthesized
per day, was determined using the formula k,=
S, x loo/& x t . S, is the specific radioactivity of
protein-bound phenylalanine, SA is the specific
radioactivity of the liver homogenate free phenylalanine and t is the incorporation time. Since t is so
short, correction for the change in precursor labelling over the measurement period is unnecessary and
the value for liver free homogenate specific activity
at the end of the 1Omin can be taken to calculate
S,t.
Statistical analysis
Data are presented as means +SEM. Differences
in the post-absorptive and fed states were assessed
by Student's t-test and taken to be significant if
P < 0.05.
RESULTS
Plasma free phenylalanine enrichment
Changes in mean plasma free phenylalanine
enrichment over the 90-min measurement period for
studies 1 and 2 are shown in Fig. 2. In both cases,
Albumin synthesis and meal feeding
Table I. Albumin fractional synthesis rates (FSRs, %/day) and
secretion times (Ts, min) in subjects after an overnight fast and
meal feeding in studies I and 2. Values are means +SEM. *P<0.05
versus fasted state.
-
:I
8 0.25
aJ
8.-=.
0.20
239
-
Fasted
FSR
L
Fasted
Fed FSR
Diff.
T,
Fed T,
Diff.
~~
Study I
Mean
SEM
/
Study 2
Mean
SEM
15
0
0.30
8 0.25
I
aJ
E
30
45
Time (mins)
60
75
T
J
0.20
5
.L
a8
c
.-5 0.15
I
-
g
c 0.10
.H
.-
=
0.05
0
0
15
30
45
Time (mins)
60
75
7.l*
7.1
9.1*
0.6
0.4
0.4
+ 1.3
0.4
+2.1
0.7
36.5
1.1
33.2*
0.6
-3.3
0.8
35.1
1.4
31.9*
1.1
-3.2
I .o
90
(b)
1
5.8
0.5
90
Fig. 3. Changes (means TSD) in phenylalanine enrichment in
albumin in (a) study I and (b) study 2 in subjects in the fasted (A)
and fed
states.
(n
plasma free phenylalanine enrichment had reached
around 85% of the injected phenylalanine enrichment by 10min after injection of the flooding
amount of phenylalanine, and enrichment fell
almost linearly thereafter.
Phenylalanine enrichment in albumin
Fig. 3 illustrates the increase in mean L['H,]phenylalanine enrichment in serum albumin
after injection of isoptope. After a delay of 3CL
40 min representing the secretion time, 'H, incorporation in circulating albumin became apparent.
Enrichment increased almost linearly thereafter
throughout the remainder of the measurement
period.
Albumin synthesis rates
Albumin fractional synthesis rates (FSRs, %/day)
in the post-absorptive and fed states are shown in
Table 1. The two feeding regimens employed in the
two studies (five small meals and one larger meal)
were both associated with a significant rise
( P <0.05) in albumin FSR, an increase upon feeding
being evident in all 12 subjects. Mean fasted albumin FSR of 5.8 ( k0.5) in study 1 and 7.1 (kO.4) in
study 2 increased to 7.1 ( L 0.4) in study 1 and 9.1
(kO.6) in study 2. The mean increase in albumin
FSR upon feeding was 24% ( k 9 ) of the fasting FSR
in study 1 and 31% (k11) in study 2. Individual
responses to feeding were considerable, increases
ranging from 9% to 74% of the fasting FSR.
In Table 2 albumin synthesis rates are expressed
in absolute terms, i.e. mgday-'kg-'. Also included
are details of estimated plasma volumes and serum
albumin concentrations. Feeding produced significant increases (P<0.05) in absolute synthesis rates
(ASRs) in both studies, mean fasting values
(mgday-'kg-') rising from 125 (k12) in study 1
and 152 (+9) in study 2 to 149 (*lo) in study 1
and 196 (kl2) in study 2 in the fed state.
The two different feeding regimens were both
associated with a significant reduction in albumin
secretion times when compared with fasting values
(see Table 1). The mean fasting secretion times of
36.5 ( + l . l ) m i n in study 1 and 35.1 (k1.4)min in
study 2 fell upon feeding to 33.2 (k0.6, P<0.05)
min in study 1 and 31.9 (kl.1, P<0.05)min in
study 2.
In both studies, injection of a flooding amount of
phenylalanine was followed by a transient rise in
plasma insulin concentration. The mean peak
increase in study 1 at 5min was 10.4p-units/ml in
the fasted state and 17.6p-units/ml in the fed state.
In study 2 the mean peak increase at 10min was
15.8 p-units/ml in the fasted state and 15.1 punits/
ml in the fed state.
Animal experiment
Fed rats ate 3.2kO.4g of diet during the 50-min
period directly before the protein synthesis measurements were carried out. This was approximately
one-fifth of their normal daily intake. One animal in
the fasted group was discounted owing to a poor
injection of isotope. Table 3 summarizes the effect of
short-term access to food on total liver protein and
albumin synthesis rates. Feeding resulted in a 20%
increase in the fractional rate of total liver protein
synthesis ( P < 0.01) compared with the fasted values.
K. A. Hunter et al.
240
-'
Table 2. Albumin absolute synthesis rates (AM, mgday kg- I), serum albumin concentrations (g/l) and estimated
plasma volumes (ml) in subjects after an overnight fast and meal feeding in studies I and 2 Values are means * S E N
*P <0.05 versus fasted state.
Fasted
Plasma volume
Fed
Serum albumin
concentration
ASR
Serum albumin
concentration
ASR
Study I
Mean
SEM
2913
188
46
I25
45
I
12
I
I49*
10
Study 2
Mean
SEM
3456
I65
46
I
I52
46
I
I96*
I2
9
Table 3. The effect of short-term feeding on the fractional rate of
total liver protein synthesis (%/day) and albumin as a percentage of
total liver protein synthesis in the adult rat. Values are means +SEM.
NS, not sienificant.
Albumin as a percentage of
total liver protein
synthesis
Fractional rate of
total liver
protein synthesis
Fasted
Fed
Significance
13.5&0.5
13.1 i0.6
NS
47.4k0.5
56.8f 1.4
Pt0.01
Albumin synthesis increased in line with the rest of
liver protein synthesis, with albumin synthesis as a
proportion of total liver protein synthesis being
similar in the fasted and fed animals.
DISCUSSION
In the present studies we have been able to
demonstrate a consistent and significant increase in
the rate of albumin synthesis in response to meal
feeding in healthy subjects. Stimulation of both the
fractional and estimated absolute synthesis rates was
apparent in both studies in which the two feeding
regimens were devised to investigate, firstly, the
cumulative effect of small meals designed to achieve
a steady state with a constant intake of nutrients
(study 1) and, secondly, the effect of a single large
meal representing a more physiological state in
which nutrients are consumed as distinct meals
(study 2). In both studies, albumin fractional synthesis rate rose significantly with feeding, increasing by
24% in study 1 and 31% in study 2. Similar
responses to that of the fractional synthesis rate
were seen in absolute rates in both studies.
Although there was no significant difference in the
degree of stimulation of albumin synthesis by the
two feeding regimes, the single large meal consumed
in study 2 appeared to stimulate albumin synthesis
to a slightly greater extent. There is a suggestion
with these data therefore that the increase in albumin synthesis observed after feeding may be sensi-
Difference in
ASR
+24
0
i43
13
tive to the amount of nutrients and the time over
which they are consumed.
Both the fractional and absolute rates in these
individuals are within the range of values reported
previously [11, 12, 25, 261. Individuals also varied in
the degree of stimulation in response to feeding,
with increases in the fractional rate as a percentage
of the fasted fractional synthesis rate ranging from
9% to 61% in study 1 and from 10% to 74% in
study 2. A similar range was also seen in albumin
absolute synthesis rate. These responses did not
appear to be related to age, fasted fractional synthesis rate or estimated habitual protein and energy
intakes.
An investigation such as the one carried out in
study 2 was only made possible by the use of the
flooding technique for the measurement of albumin
synthesis because of the metabolic non-steady state
which would have been present 30min after consumption of a large meal. By infusing subjects with
a large amount of labelled amino acid, this method
reduces fluctuations in the labelling of the precursor
for protein synthesis which would otherwise invalidate techniques requiring steady-state conditions
such as the [14C]carbonate or constant-infusion
methods. In addition, we have demonstrated previously [12] that the short period of measurement
employed makes the inclusion of transcapillary loss
of albumin as a confounding factor unnecessary.
To determine whether or not the observed effect
on albumin was specifically confined to albumin or
the result of a more generalized liver response, the
effect of short-term feeding on albumin synthesis in
an animal model was used, which allowed us to
determine the effect of nutrient intake on the ratio
of albumin to total liver synthesis. Intake of a
moderate amount of chow increased the fractional
rate of total liver protein synthesis from the fasted
value of 47 to 57%/day (Table 3). Albumin as a
percentage of total liver synthesis remained constant, however, at approximately 13%. These results
are in agreement with previous observations [27]
demonstrating that there is no selective stimulation
of albumin synthesis in rats with feeding. However,
because total liver protein synthesis was increased
Albumin synthesis and meal feeding
by feeding, the amount of albumin synthesized in
the liver and secreted into the plasma would be
greater in the fed rats than in the fasted rats. This
provides a plausible explanation for the increase in
albumin synthesis that was observed in the human
subjects.
There is evidence to suggest that regulatory
factors involved in the observed response of albumin synthesis to feeding include amino acid supply
to the liver [28-301, glucose and insulin [31, 321. In
addition, the route of ingestion of nutrients may be
an important factor in the response of albumin
synthesis to feeding. Recent work by Ballmer et al.
[33] has demonstrated that feeding short-term
(12 h), high-dose total parenteral nutrition to
healthy subjects does not promote albumin synthesis to any significant extent. The authors speculated
that the contrasting responses observed with enteral
and parenteral feeding may be a result of augmentation of the nutrient supply to the liver during
enteral feeding.
Although this is the first study to determine the
immediate response of albumin synthesis to feeding
in human subjects, there has been another study
which investigated the effects of enteral feeding over
a 6-h period. Subjects fed enterally for 6 h with a
mixed glucose and amino acid solution exhibited a
massive stimulation of albumin synthesis, the fractional rate of albumin synthesis increasing from a
mean fasted value of 1 2 k 2 to 23f3%/day in the
fed state when measured with a constant infusion of
['4C]leucine [lo]. The fasting albumin fractional
synthesis rate and the response to feeding observed
in that study are far greater than those we observed
in either study. Although the reason for this large
disparity is not obvious, a number of differences in
the technique of measurement and details of the
feeding protocol are potentially important. It is
possible, for example, that differences in the rate of
absorption or the composition of our meals, which
were made up of normal foodstuffs compared with
the semisynthetic diet used by De Feo et al. [lo],
may be partially responsible. Certainly it seems that
albumin synthesis may be sensitive to nutrient composition, as a number of studies carried out in postsurgery patients [34, 351 have shown. In addition,
their use of the plasma ketoacid of leucine to
estimate the specific activity of the precursor for
albumin synthesis may be of significance.
An interesting observation arising from this work
is the consistent reduction in albumin secretion time
(T,) which occurs after feeding. The development of
the flooding technique has allowed us to measure
this variable in human subjects, and it has become
apparent from this and previous work carried out
by our group [33] that the increase in albumin
synthesis rate observed upon feeding may be
mediated through a reduction in the period between
the onset of albumin synthesis and its secretion into
the blood. In addition to the observed change in
secretion time, it is possible that the mechanism by
24 I
which albumin synthesis is rapidly augmented with
feeding is in part post-transcriptional and may
involve the protection of albumin messenger RNA
(mRNA) with the cytosol during fasting. Yap and
colleagues [36, 371 have shown that, in rats, a
short-term fast prompts a subcellular redistribution
such that albumin mRNA is stored as higher molecular weight nuclear precursors to cytoplasmic albumin mRNA. After refeeding fasted rats for 1h,
albumin mRNA sequences are rapidly transferred
back to polysomes and, as shown by Rothschild et
al. [38], the ability of the liver cell to synthesize
albumin can be restored after 15-30min of
refeeding.
In summary, sequential measurements of the rate
of albumin synthesis were carried out in healthy
subjects after an overnight fast and following consumption of either one large meal or a number of
smaller meals. The rate of albumin synthesis
increased in response to feeding in all subjects
studied, showing considerable inter-individual variation. It would appear from these data that, in
health, the rate of albumin synthesis in humans is
subject to significant fluctuations throughout a 24-h
period, with synthesis being attenuated during periods of fasting and responding rapidly to nutrient
intake.
The present studies have successfully illustrated
the use of the flooding technique for the measurement of albumin synthesis rate in a metabolic
situation in which other techniques would have
been inappropriate. It will be a useful tool for
investigating the acute effects of regulatory factors
such as nutrition and trauma on the synthesis of
body protein.
ACKNOWLEDGMENTS
We acknowledge the support of the Scottish
Office Agriculture and Fisheries Department and
Nestec Ltd, Switzerland.
REFERENCES
I. Whitehead RG, Coward WA. Lunn PG. Serum albumin concentration and the
onset of kwashiorkor. Lancet 1973; i: 62-6.
2. Whicher J, Spence C. When is serum albumin worth measuring?Ann Clin
Biochem 1987: 24: 572-80.
3. Haider M, Haider SQ.Assessment of protein-calorie malnutrition. Clin Chem
1984; 3& 1286-99.
4. Gray GE, Meguid MM. The myth of serum albumin as a measure of nutritional
status. Gastroenterology 1990; 99: 1845-51.
5. Dahn MS, Jacobs LA, Smith S, Lange P, Mitchell RA, Kirkpatrick JR. The
significance of hypoalbuminaemia following injury and infection. Am Surg 1985;51:
340-3.
6. Oluferni 0s.Whittaker PG, Halliday D, Lind T. Albumin metabolism in fasted
s u b j m during late pregnancy. Clin Sci 1991; 81: 161-8.
7. James WPT, Hay AM. Albumin metabolism: Effect of the nutritional state and the
dietary protein intake. J Clin Invest 1968; 4E 195&72.
8. Kelman L, Saucders SJ, Fruth L. Wicht S, Corrigal A. E f f m of dietary protein
restriction on albumin synthesis, albumin catabolism, and the plasma aminogram.
Am J Clin Nutr 1972; 25: 1174-8.
9. Castro CE, Saunden Sevall J. Hepatic level of rat albumin messenger RNA is
influenced by factors other than dietary protein. J Nutr 1985; 115: 491-5.
10. De Feo P, Horber FF, Haymond MW. Meal stimulation of albumin synthesis: a
significant contributor to whole body protein synthesis in humans. Am J Physiol
1992; 263: E794-9.
242
K. A. Hunter et al.
I I. Gersovitz M, Munro HN, Udall J, Young VR. Albumin synthesis in young and
elderly subjects using a new stable isotope methodology: Response to level of
protein intake. Metabolism 1980 2 9 1075-86.
12. Ballmer PE, McNurlan MA, Milne E, et al. Measurement of albumin synthesis
in humans: a novel approach employing stable isotopes. Am J Physiol 1990;
259 E797-803.
13. McNurlan MA, Essen P, Thorell A, et al. Response of protein synthesis in
human skeletal muscle to insulin: an investigation with L-PHl]phenylalanine.
Am J Physiol 199%276: EIO2-8.
14. Hunter KA, Ballmer PE, Anderson SE, Calder AG. Garlick PJ. McNurlan MA.
Increased albumin synthesis in healthy volunteers with feeding. Proc Nutr SOC
1993; 52: 9 IA.
15. FAO/WHO/UNU. Energy and protein requirements. Technical report 724.
Geneva: WHO, 1985.
16. Calder AG, Smith A. Stable isotope ratio analysis of leucine and ketoisocaproic
acid in blood plasma by gas chromatography/mass spectrometry. Use of
tertiary butyldimethylsilyl derivatives. Rapid Comm Mass Spec 1993; 2: 14-16.
17. Korner A, Debro JR. Solubility of albumin in alcohol after precipitation by
trichloroacetic acid: a simplified procedure for separation of albumin. Nature
(London) 1956 178: 1067.
18. Calder AG, Anderson SE, Grant I, McNurlan MA, Garlick PJ. The
determination of low drphenylalanine enrichment (0.0024.09 Atom Percent
Excess), after conversion to phenylethylamine, in relation to protein turnover
studies by gas chromatography (electron ionisation mass spectrometry). Rapid
Comm Mass Spec 1992; 6: 4214.
19. Doumas BT. Watson W, Biggs HG. Albumin standards and the measurement
of serum albumin with bromocresol green. Clin Chim Acta 1971; 31: 87-96.
20. MacRae JC, Bruce LA, Hovall FDDeB, et al. Influence of protein nutrition on
the response of growing lambs to exogenous bovine growth hormone. J
Endocrinol 1991; I)o: 53-61,
21. Dagher FJ, Lyons JM, Finlayson DC, Shamsai J. Moore FD. Blood volume
measurement: a critical study. Prediction of normal values: controlled
measurement of sequential changes: choice of a bedside method. Adv Surg
1965; I: 69-109.
22. Marigold JH, Gilmore IT, Thompson RPH. Effects of a meal on plasma
clearance of [“C]glycocholic and lndocyanine Green in man. Clin Sci 1981; 61:
325-30.
23. Garlick PJ, McNurlan MA, Preedy VR. A rapid and convenient technique for
measuring the rate of protein synthesis in tissues by injection of
[IHIphenylalanine. Biochem J 1980; 192: 719-23.
24. Pain VM, Clemens MJ, Garlick PJ. The effect of dietary protein deficiency on
albumin synthesis and on the concentration of active albumin messenger
ribonucleic acid in rat liver. Biochem J 1978; IR: 129-35.
25. Olufemi OS, Humes P. Whittaker PG, Read MA, Lind T, Halliday D. Albumin
synthetic rate: a comparison of arginine and alpha-ketoisocaproate precursor
methods using stable isotope techniques. Eur J Clin Nutr 1990; 44: 351-61.
26. Rothschild MA, Oratr M, Schreiber SS. Albumin synthesis. N Eng J Med 1972;
286: 748-57.
27. Maurice M, Lardeux B, De Saint-Steban C, Bourdel G, Feldman G. Influence of
diets with different levels of protein and energy on liver albumin content in
the rat. J Nutr 1986; 1 1 6 213241.
28. Kirsch RE, Saunders SJ. Frith L, Wicht S, Kelman L, Brock IF. Plasma amino
acid concentration and the regulation of albumin synthesis. Clin Nutr 1969 12:
1559-62.
29. Kelman L, Saunders SJ, Wicht S, et al. The effects of amino acids on albumin
synthesis by the isolated perfused rat liver. Biochem J 1972; 129 805-9,
30. Yap SH, Hafkenscheid JCM. Effect of starvation on the synthesis rate of
albumin in vivo and its relation t o the concentration of amino acids in the
peripheral blood, the portal concentration and in the liver cystolic fraction.
Ann Nutr Metab 1981; 25: ISM.
31. Princen JMG, Mol-Backx GPBM, Yap SH. Restorative effects of glucose
refeeding on reduced synthesis of albumin and total protein and on
disaggregated polyribasomes in liver of starved rats: Evidence of a
post-transcriptional control mechanism. Ann Nutr Metab 1983; 27: 182-93.
32. Wittman IS, Lee K, Miller ON. Dietary and hormonal influences on rat
polysome profiles: fat, glucose and insulin. Biochim Biophys Acta 1969 174
536-43.
33. Ballmer PE, McNurlan MA, Anderson SE, Garlick PJ. The effect of short term
total parenteral nutrition on albumin synthesis rates in healthy volunteers.
Clin Nutr 1992; II (Suppl.): I.
34. Skillman JJ, Rosenoer VM, Smith PC, Fang MS. Improved albumin synthesis in
postabsorptive patients after amino acid infusion. N Engl J Med 1976 295
10374.
35. Skillman JJ, Rosenoer VM, Young JB, Young VR, Long PC, Munro HN. Energy
intake can determine albumin synthesis in man after surgery. Surgery 1985; 97:
271-7.
36. Yap SH. Stair RK, Shafritz DA. Effect of a short term fast on the distribution
of cytoplasmic albumin messenger ribonucleic acid in rat liver. J Biol Chem
1978; 253 494450.
37. Shafritz DA. Yap SH, Stair RK. Regulation of albumin synthesis in rat liver.
Mol Biol Rep 1979 5 71-8.
38. Rothschild MA, Oratz M. Mongelli J. Effects of a short term fast on albumin
synthesis studied in vivo, in the perfused liver and on amino acid incorporation
by hepatic microsomes. J Clin Invest 1968; 47: 2591-9.