CLIN. CHEM. 25/10, 1787-1790 (1979)
Measurement of Acetate in Human Blood by Gas Chromatography:
Effects of Sample Preparation, Feeding, and Various Diseases
C. D. Tollinger, Hendrik J. Vreman, and Michael W. Weiner
acetate concentrations in whole blood,
serum, and plasma by a modification of a previously described method involving vacuum distillation and gas
chromatography. The mean acetate concentration of fresh
venous plasma from 27 normal subjects was 51 ± 5
tmol/L (95% confidence limits ranged from 0 to 103
imol/L). The acetate concentrations of serum and plasma
incubated for 2 h at either 4 #{176}C
or 27 #{176}C
were the same.
The acetate concentration of whole blood incubated at 27
was significantly greater than that of blood incubated
at 4 #{176}C.
This change may have resulted from the production of acetate by erythrocytes or from the hydrolysis of
acetate esters. Storage of plasma at -20 #{176}C
for 24 h
significantly increased acetate concentrations from 26 ±
6 imol/L to 63 ± 4 tmol/L. After the subjects consumed
a standard breakfast, venous plasma acetate concentrations increased from 58 to 97 zmol/Lat 30 mm. Acetate
concentrations in arterial plasma exceeded those in venous plasma. Plasma acetate concentrations were not
significantly altered in patients with malignancy or diabetes
mellitus, but severe liver disease and severe acidosis were
both associated with increased acetate concentrations.
These preliminary observations suggest that plasma acetate concentrations may be altered in several disease
states.
We measured
Additional Keyphrases:
normal values
serum) plasma
comparison
values in patients with cancer, liver disease,
diabetes mellitus, acidosis
variation, sources of
.
.
Recently we described
a new method for measuring
acetate
in biological materials
(1). The technique
involves the use of
low temperature,
high-vacuum
microdistillation,
and gas
chromatography
of the free acid. The lower limit of detection
for the method is approximately
10 iimol of acetate per liter,
and at concentrations
of 100 zmol/L the standard
deviation
is less than 5% of the mean. Here we report results of our
analyses
for acetate in plasma, serum, and whole blood, and
describe the effects of sample preparation
and storage. Effects
of feeding and of certain diseases on plasma acetate concentrations
are also described.
Materials and Methods
Subject Population
The
women
clinical
normal
subjects
used in this study were men and
volunteers,
ages 18-60 years, who showed no outward
symptoms
of illness or disease.
Stanford University School of Medicine, Department of Medicine,
Veterans Administration Medical Center, 3801 Miranda Ave. (1C601),
Palo Alto, CA 94304.
Received Jan. 29, 1979; accepted June 12, 1979.
Diabetes mellitus patients (n = 33) were identified
from
information
in their Veterans Administration
clinical records.
All of these patients
were adults with either adult or juvenile-onset
diabetes mellitus, being treated with insulin or oral
hypoglycemic
agents. None had overt ketoacidosis.
Patients
being treated for solid tumors and lymphomas
(n
= 81) were selected
from the Stanford
Hospital
Oncology
Clinic.
Patients
with severe liver disease were identified
by review
of hospital
records and had a serum bilirubin
concentration
exceeding
20 mg/L (n = 21).
Severe metabolic
acidosis in hospitalized
patients
(n = 26)
was defined as an arterial blood pH of <7.2.
Acetate Concentrations
Components
in Various Blood
Thirty milliliters
of whole blood was obtained from each of
seven normal fasting human subjects’
(five women, two men,
ages 21-40), and each sample was divided into six aliquots.
Aliquot 1 was heparinized
whole blood, which was incubated
for 2 h at 4 #{176}C.
Aliquot 2 was heparinized
whole blood, which
was incubated
for 2 h at room temperature.
Aliquots 3 and 4
were used to obtain serum: aliquot 3 was whole blood without
added heparin, which was allowed to clot for 2 h at 4 #{176}C,
and
aliquot 4 was whole blood without added heparin, which was
allowed to clot at 2 h at room temperature.
The serum was
obtained
by centrifuging
aliquots 3 and 4 at 1800 X g for 10
mm at 4 #{176}C.
To obtain plasma, we heparinized
aliquots 5 and
6 and immediately
centrifuged
them in the same way. The
plasma from aliquot 5 was kept at 4 #{176}C
for 2 h and aliquot 6
was kept at room temperature
for 2 h. After treatment,
all
samples
were deproteinized
and analyzed
as described
below.
Analytical
Recovery
of Acetate from Plasma
One milliliter of 0.40 mmol/L sodium acetate was added to
3 mL of each of 10 different
plasma samples. Aliquots of the
original plasma, the acetate standard,
and the plasma-standard mixture were deproteinized
and analyzed
as described
below. Percentage
recovery is defined as (excess amount determined/excess
amount added) X 100.
Effects of Freezing on Plasma Acetate
Blood was obtained from the same seven normal fasting
subjects described
above, heparinized,
and the plasma separated. Aliquots of this plasma were immediately
deproteinized
and analyzed
for acetate
Other aliquots
of plasma
were
promptly
frozen at -20 #{176}C,
stored for one to 30 days, thawed
at room temperature,
and then deproteinized
and analyzed
for acetate. To determine
the effects of freezing when plasma
All studies described in this report were approved by the Stanford
Committee on the use of human subjects for investigation. All subjects
provided written informed consent.
CLINICALCHEMISTRY,Vol.
25, No. 10, 1979
1787
acetate
concentrations
were increased,
we obtained
plasma
from patients
during hemodialysis
with 35 mmol/L acetate
in the dialysate.
Aliquots of this plasma were then immediately analyzed or frozen at -20 #{176}C
for 24 h, thawed, and analyzed.
Arterio-Venous
Differences
Arterial
blood was obtained
from seven normal
fasting
subjects
by direct arterial
puncture
and venous blood was
obtained
by venipuncture.
Arterial blood was obtained
from
seven hemodialysis
subjects before a hemodialysis
treatment,
from the arterio-venous
fistula or bovine heterograph;
venous
blood was obtained
by venipuncture
of the arm that did not
have an A-V fistula. All blood samples were heparinized
and
the plasmas were separated
and analyzed.
Acetate
Concentrations
in Normal Human Subjects
Blood was obtained from 87 fasting normal
immediately
heparinized,
and the plasma
plasmas from 27 subjects were immediately
analyzed
subjects
analyzed.
for acetate.
were frozen
human subjects,
separated.
The
deproteinized
Fig. ‘1.
Diagram of the 6-station vacuum distillation manifold
The 10/30 slip joint on the left leads to a McCleod vacuum gauge. The opening on the right is connected to the vacuum pump
and
The plasmas from the 60 other normal
for various intervals,
then thawed and
90 s, the station valve was closed and the receiving tube was
placed in the solid CO2-acetone
bath, while the distilling tube
was heated
Effects of Feeding on Plasma Acetate
Concentrations
After they had fasted overnight,
blood was obtained
from
10 normal subjects (four men and six women, mean age 27 ±
4 years), heparinized,
and the plasma separated.
Each subject
then consumed
a standard
breakfast
consisting of two pieces
of toast, 5 g of butter, one egg, and 250 mL of milk. Blood was
then obtained
at 30-mm intervals
for 3.5 h by use of an indwelling venous needle. Each sample of blood was heparinized, centrifuged,
and the plasma was deproteinized
and analyzed without delay.
Serum and Plasma Acetate Concentrations in
Patients with Diabetes, Cancer, Liver Disease, or
Acidosis
Fresh samples of venous serum were obtained from patients
with diabetes,
cancer, and liver disease. Arterial plasma was
sampled from patients
with severe acidosis.
Determination
of Acetate
Acetate in plasma, serum, or whole blood was determined
by a modification
of the method described previously (1). The
microdistillation
apparatus
has been simplified and expanded
to distill six samples at one time. The use of this vacuum distillation manifold
(Figure 1) eliminated
the requirement
for
the distillation
and receiving flasks used previously.
Instead,
regular culture tubes (10 X 75 mm, cat. no. 339-267; Curtin
Matheson
Scientific,
Inc.) with two opposing
3-mm depressions approximately
1 cm from the bottom were used at the
distilling
side and the same, but unmodified,
tubes were attached to the receiving side. The depressions
were necessary
to keep the frozen liquid plug in the bottom of the distillation
tube during the initial (melting) stage of the distillation.
The
tubes were connected
to the manifold by means of sleeves, 40
mm X 6mm (i.d.) X 11 mm (o.d.), made of latex rubber tubing.
The sample (100 tL) was mixed vigorously
in a 75 X 10 mm
glass distilling
tube with 50 L of deproteinizing
mixture
containing,
per liter, 100 g of sulfosalicyclic
acid (deproteinizing agent), 6 mol of formic acid (acidifying
agent), and 3.00
mmol of propionic
acid (internal
standard).
Previously,
the
precipitated
protein was removed by centrifugation,
but the
present modifications
made this step unnecessary.
The tube
was attached
to the distillation
side of one of the six manifold
stations and frozen in a solid C02-acetone
bath for 1 mm. The
station valve was then opened to evacuate the station. After
1788
CLINICAL CHEMISTRY,
Vol. 25, No. 10, 1979
to 45 #{176}C
in a water bath. The remainder
of the
manifold
was kept at ambient
temperature.
The distillation
was essentially complete in 3 mm and was stopped after 5 mm,
by removing the distilling tube. The receiving tube was then
removed, the sample was melted, mixed, and transferred
to
a glass microvial (300 tL, no. 5080-87 12; Hewlett-Packard
Co.,
Avondale, PA 19311) and sealed. This method of purification
is rapid, specific, and quantitative,
requiring
no more than 8
mm for transfer
of six samples.
Using this procedure,
one
person can process 200 samples during a typical working day.
The acetic acid concentration
of the distillate was determined
by injection of 2 L of distilled sample into the gas-chromatographic
column in a Hewlett-Packard
Model 5830A gas
chromatograph,
equipped
with an Automatic
Sampler
(HP
no. 7671A) used in single-column
mode. The 180 cm X 2 mm
glass column was packed with 60/80 mesh Carbopack
B coated
with 3% Carbowax 20M and 0.5% H3P04 (Supelco, Bellefonte,
PA 16823). The column had H3P04-treated
glass-wool plugs
at each end. Before being used, the column was conditioned
at 200 #{176}C
for 24 hat a flow rate of 60 mL/min. The column was
operated
at a temperature
of 160 #{176}C.
The injection port was
maintained
at 200 #{176}C
and the flame ionization
detector
was
kept at 250 #{176}C.
The N2 carrier gas flow rate was 60 mL/min.
The attenuation
was 2, with a slope sensitivity
of 0.1. The
chromatograph
was standardized
with a mixture of, per liter,
667 zmol of sodium acetate and 1.00 mmol of propionic
acid
in 2 mol/L formic acid.
Statistics
All data are expressed as the mean ±SEM. Student’s
t-test
and the paired t -test were used to determine
statistical
significance between means. The term “significant”
is used when
p is <0.05.
Results and Discussion
Comparison of Results Obtained by Original
Modified Procedures
and
Forty-three
frozen plasma samples were thawed, then simultaneously
purified
by our previously
published
method
and by the modified
procedure
presented
in Methods.
We
found no significant
difference
in acetate concentrations
as
measured
by the two methods when the results were analyzed
by the paired t -test. Gas-chromatographic
parameters
and
limits
of detection
remained
unchanged
from
original
method.
Table 1#{149}
Acetate Concentrations
Whole blood
27 #{176}C
(igIL)
in Whole Blood, Plasma, and Serum of Seven Persons
Plasma
4
-;
Serum
27
4
-#{231}
27
4 #{176}C
#{176}C
69
66
37
50
50
55
28
57
15
26
38
15
34
44
25
30
34
28
26
43
21
20
61
25
38
11
31
10
37
06
31
10
26
03
30
02
Mean
38
29
29
29
27
30
SEM
08
07
05
05
06
08
!
Whole blood, plasma, and serum were incubated for 2 h at temperatures shown. Values are means of duplicate determinations.
Acetate
Concentrations
in Blood Components
Table 1 shows that there was no significant
different
between acetate concentrations
in plasma or serum incubated
at either 4 #{176}C
or room temperature.
Evidently
either serum
or plasma may be used, and either may be stored for several
hours at room temperature
without
affecting
acetate
concentrations.
When whole blood was incubated
at 4 #{176}C,
the
acetate concentration
was the same as for serum or plasma.
However, the acetate concentration
of whole blood incubated
at room temperature
for 2 h was significantly
increased
(p
<0.05) because of the increase in values for subjects 3 and 7.
This change may have resulted from the production
of acetate
by erythrocytes
(2) or from the hydrolysis
of acetate esters.
Recovery
of Acetate from Plasma
The analytical
recovery of 0.1 mmol/L
concentrations
of
sodium acetate added to each of 10 samples of fresh plasma
was 98.1± 1.2%.
Arterio-Venous
The arterial plasma acetate concentration
was significantly
greater than the venous plasma acetate concentration,
both
in the normal subjects and the dialysis patients. These results
(Table 3) indicate
trations
Acetate
this is a 142% increase.
Similar experiments
were then performed with plasma obtained
from patients
during hemodialysis with acetate dialysate.
In this case, the mean acetate
concentration
of the fresh plasma was 2284 ± 392 zmol/L;
after 24 h at -20 #{176}C
the mean concentration
rose by 173 to
2457 ± 416 zmol/L
(p = 0.01).
Table 2. Effects on Plasma Acetate
Concentrations of Storage at -20 #{176}C
Acetate, Mmol/L
30-day storage
Plasma obtained during hemodialysis
Fresh (n = 14)
24-h storage
a,,
<0.01.
Data expressed as mean ±SEM.
Concentrations
The mean
acetate
with other
26 ± 6
63 ± 4
66 ± 18
2284 ± 392
2457 ± 416
substrates,
of acetate.
in Normal
concentration
Plasma
in fresh
plasma
from 27
vals, were also analyzed
separately.
The mean acetate concentration
was 56 * 5 zmol/L with a range of 0 to 149 mol/L
Table 3. Arterial and Venous Acetate
Concentrations in Fresh Plasma from Fasting
Subjects
Plasma
acetate, hmol/i.
Arterial
Venous
Normal subjects
72±7
p<0.Ol
56±5
Hemodialysis patients
p<0.OOl
37±3
46±5
26±7
19±7
Data expressed as mean ±SEM.
Table 4. Acetate Concentrations in Venous
Plasma of Normal Subjects and Patients with
Various Diseases
Subjects
Mean ± SEM
Normal
Normal plasma
Fresh (n = 7)
24-h storage
low, compared
extraction
normal subjects was 51 ± 5 jmol/L
(range, 2 to 99 smol/L)
(Table 4). The 95% confidence limits (defined as the mean ±
2 SD) for acetate concentrations
in normal fresh venous
plasma are 0-103
zmol/L. Plasma samples from 60 other
normal subjects, which had been at -20 #{176}C
for various inter-
Table 2 shows that storage of plasma for 24 h at -20 #{176}C
plasma acetateassociated
withfreezing
was only37 zmol/L,
that even though plasma acetate concen-
are relatively
there is peripheral
Effects of Freezing on Plasma Acetate
Concentrations
significantly
increased
acetate
concentrations
from 26 ± 6
mol/L
to 63 ± 4 tmol/L.
Additional
storage for 30 days at
-20 #{176}C
did not significantly
change plasma acetate concentrations.
Possibly some acetate esters are hydrolyzed
during
the freezing
or thawing
process.
Although
the change in
Difference
Range
imol/L
(n = 27) Fresh samples
(n = 60) Frozen samples
2-99
51 ± 5
56 ± 42
0-149
43 ± 14
2-1080
Patients with malignancy
(n = 81) Frozen samples
Patients with liver disease
(n = 21) Frozen samples
92±11
62-302
Patients with diabetes mellitus
(n = 33) Frozen samples
27 ± 3
4-575
CLINICALCHEMISTRY,Vol. 25, No. 10, 1979
1789
EFFECTS OF FEEDING ON PLASMA ACETATE
CONCENTRATIONS (n 10)
0.1
plasma acetate concentrations
appeared
Although these analyses were performed
been frozen, the acetate concentrations
exceeded
-J
the concentrations
Diabetes
of frozen normal plasma.
reportedly
in experimental
animals
(6-8).
patients
with diabetes
mellitus
0.1(
0
E
E
a)
a
a)
C)
a
a
E
a
acetate
0
30
60
90
120
150
11
Time (mm.) after feeding
Effects of consuming a standard breakfast on plasma
acetate concentrations
Fig. 2.
Data (Inmmol/L) expressed as mean ± SEM. At 30 mm. the plasma acetate
concentrations significantly exceeded fasting
values(p <0.001)
and the95% confidencelimitsare0 to 133imol/L.The wider
range of the frozen samples
may be due to the previously
noted
effectoffreezing.
Effects of Feeding on Plasma Acetate
Concentrations
Figure 2 shows that 30 mm after the subjects
ingested
a
standard
breakfast,
their plasma acetate concentrations
rose
significantly,
from 58 ± 7 to 97 ± 13 mmol/L, declining after
1 h to slightly below fasting values. These results suggest that
subjects should fast for at least an hour before blood is sampled for acetate determination.
Plasma Acetate Conceptrations
Diseases
in Various
Cancer. Hepp et al. (3) reported
in vitro produce large quantities
that certain tumors studied
of acetate, suggesting
that
serum
acetate
concentrations
may be increased
in patients
with cancer. For this reason, we measured
serum acetate
concentrations
in 81 patients with various malignancies.
The
mean serum acetate value of 43 ± 14 imol/L
was not significantly different
from normal. Two patients
had plasma acetate concentrations
>0.1 mmolfL: the value for a patient with
carcinoma
of the rectum was 1.080 mmol/L, that for a patient
with carcinoma
of the lung was 468 zmol/L. For the other 79
patients,
the mean venous plasma acetate concentration
was
25 ± 2 tmol/L,
which is significantly
lower than the mean for
fresh or frozen normal
venous
plasma.
Liver disease. Acetate is produced
by the gastrointestinal
flora, and acetate concentrations
in the portal venous blood
have been reported to be considerably
greater than values for
thesystemiccirculation
(4).Furthermore,the livercontains
the greatest amount of acetyl-CoA
Therefore,
we measured
plasma
severe liver disease (mean plasma
The mean value was 100 ± 12
mol/L).
Seven of these subjects
centrations
exceeding
100 tmol/L
synthetase
(EC 6.7.1.1) (5).
acetate in 21 patients
with
bilirubin, 62.7 ± 9.3 mg/L).
lLmol/L (range, 62 to 302
had plasma acetate
conand plasma bilirubin
and
1790 CLINICALCHEMISTRY,Vol. 25, No. 10, 1979
mellitus
concentration
was
to correlate roughly.
on samples that had
in these subjects also
impairs acetate metabolism
For this reason, we studied 35
of various severity. The mean
27 ± 15 imol/L
(range,
9 to 575
smol/L).
Only two subjects had concentrations
>100 tmol/L
(117 and 575 imol/L).
These results suggest that most patients
with diabetes
mellitus do not have increased
plasma
acetate concentrations.
Addosis.
Some patients develop metabolic
acidosis, which
is not completely accounted for by renal failure, ketoacidosis,
or lactic acidosis. Therefore,
26 patients with severe metabolic
acidoss
were investigated.
Their mean plasma acetate concentration
was 103 ± 14 lmol/L
(range 30-402). Ten of these
subjects had values greater than 100 /zmol/L, indicating
that
some severely
acidotic
subjects
can be expected
to have
above-normal
plasma acetate concentrations.
These preliminary
observations
suggest that plasma acetate
may be altered in some disease states. The clinical relevance
of these findings is uncertain.
Fasting
plasma acetate concentrations
represent a balance of endogenous
production
and
peripheral
metabolism.
Therefore,
additional
studies
that
directly
measure
the endogenous
production,
peripheral
metabolism,
and turnover
rate of acetate (9) are necessary th
determine
the cause of abnormal
plasma acetate concentrations.
We greatly appreciate the assistance of Drs. Richard Belier and
Vinod Assomull, who obtained the venous and arterial blood samples.
Susan Kirk (Palo Alto VA Hospital) provided some arterial blood-gas
samples. Mrs. Ann Olthof performed the feeding study, and Maureen
McTigue collected the patients’ blood samples. The microdistillation
apparatus was constructed
by Jan Van Gastel, Sr., 399 Vine St., Menlo
Park, CA 94025. This project was supported
by NIH Contract
N01AM52210
and the VA Medical Research Service.
References
I. Vreman,
Determination
H. .1., Dowling, J. A., Raubach, R. A., and Weiner, M. W.,
of acetate in biological material by vacuum microdistillation and gas chromatograply.
Anal. Chem. 50, 1138 (1978).
2. Hochheuser, W., Weiss, H., and Wieland, 0., Uber den Stoffwechsel der freien Essigsaure im Tierkorper and seine Beziehung zum
Atherogeneseproblem.
J. Clin. Chem. Clin. Biochem
2, 175 (1964).
3. Hepp, D., Prusse, E., Weiss, H., and Wieland, 0., Essigsaure als
Endprodukt
des aeroben Krebstoffwechsels.
Biochem. Z. 344, 87
(1966).
4. Buckley, B., Mammaliam
University,
Great Britain,
acetate metabolism.
Ph.D. thesis, Oxford
1972.
5. Knowles, S. E., Jarrett, I. G., Filsell, 0. H., and Ballard, J., Production and utilization of acetate in mammals. Biochem. J. 142,401
(1974).
6. Harper, P. V., Neal, W. B., Jr., and Hlavacek,
zation
in the dog. Metabolism
7. Ciaranfi,
2,62
G. R., Acetate
utili-
(1953).
F., and Fonnesu, A., Time course of injected acetate in
normal and depancreatized
dogs. I3iochem. J. 57, 171 (1954).
8. Jarrett, I. G., Filsell,O. H., and Ballard, F. J., Metabolic and endocrine interrelationships
in normal and diabetic sheep. Horm.
Metab. Res., Suppl. 4, 111 (1974).
9. Annison, E. F., and White, R. B., Further studies in the entry rates
of acetate and glucose in sheep; with specific reference to endogenous
production of acetate. Biochem. J. 84, 546 (1962).