CLIN. CHEM. 33/1, 140-141
Technique
Samples
Edwin
(1987)
for Gas-Chromatographic
J. Zarllng
and Maureen
Measurement
The volatile alkanes, especially ethane and pentane, have
for several years been used as markers of lipid peroxidation
(1) in studies both in vivo and in vitro. These gases are
produced
during the peroxidation of specific
unsaturated
fatty
acids, and are thought to provide a noninvasive
marker of cellular lipid destruction. However,
the alkanes
are excreted in such small concentrations
(nanomoles
per
liter of air) that hitherto
it has been necessary
to concentrate the gas samples before analysis,
either by use of a
closed
system (used with small animals
and tissue preparation) in which the alkanes are accumulated
(2, 3) or by
alkane trapping (used with small animals
and humans)
in
which large volumes of gas are concentrated by condensation on cold precolumns
(4).Both methods, though effective,
are cumbersome
and provide readings
for gas samples
collected
over a period
of several minutes. In an effort to
detect alkanes in single-breath samples from humans,
we
devised an on-line method for gas-chromatographic
analysis
for alkanes that can detect nanomoles of alkane per liter of
air. We also used our method to measure and compare the
concentrations
of alkanes
in alveolar
and total
breath
samples from healthy humans.
and Methods
Standard
gas samples
containing
a mixture
of the alkanes
ethane, propane, butane, and pentane (Alltech Associates,
Inc., Deerfield,
IL) were diluted
with room air to produce six
different
concentrations
for each gas ranging
between 0 and
13 nmol per liter of air (approximately
310 nL/L). Two
samples of each concentration of alkane gas were stored in
polyethylene/polypropylene
syringes (Fortuna Syringe; Aldrich Chemical, Milwaukee,
WI) and analyzed for alkane
content by gas chromatography
within 5 h. For each alkane,
we calculated
a linear correlation
between the concentration
added
and
the
concentration
detected.
In a separate
CLINICAL
CHEMISTRY,
from Single-Breath
Vol. 33, No. 1, 1987
ment, designed to assess the stability
in polyethylene syringes,
we ifiled
of the alkanes stored
12 syringes
from
a
common
source containing
approximately
9.5 nmol of each
alkane
per liter. Groups of three syringes
were then analyzed for alkane
content
0, 10, 25, and 50 h later. The mean
measured
concentration
of each gas was calculated
at each
time point and compared with the value found in the zero
hour sample.
For within-run
and between-run
precision
studies, we used five air samples containing
low, moderate,
and high concentrations of added alkanes.
To establish normal values for concentrations
of each o
the alkanes in human
breath, we collected duplicate samples of alveolar breath and total exhaled
breath from eight
ostensibly
healthy men, ages 24 to 33 years. Both types ol
breath samples were collected at 09:00 h on the same day
after an overnight fast. Duplicate alveolar breath samples,
collected by use of Haldane-Priestly
tube (5), were stored in
50-mL
polyethylene
syringes
and analyzed
within
5 h,
“Total
breath” was sampled by collecting an entire exhalation into a gas-tight
bag, and then withdrawing
duplicate
50-mi
samples
into polyethylene/polypropylene
syringes.
For analysis
of alkanes,
we used a gas chromatograph
(Model
6000; Vanan
Instruments,
Sunnyvale,
CA) equipped
with a gas-sampling
valve, which included a 10-mi
sampling loop and a flame ionization
detector
reading
in the
range
of 10 nA/V. We flushed and ifiled the sampling loop
with 50-mi volumes of each gas sample. We eluted the
components from a 2-rn stainless
steel column (Chromosorl
102; Varian
Instruments)
by using nitrogen carrier gas at a
flow rate of 30 mLfmin. The injector
temperature
was 150 #{176}C
and the detector temperature
was 225 #{176}C.
To minimize
peak
broadening
from the large sample volume,
we achieved an
on-line “cold-trap”
concentration
of the gas sample
as fol.
lows. After
injection,
we kept the column temperature
a
50#{176}C
for 1 mm, increased
it by 50 #{176}C/min
to 100#{176}C,
the
allowed a gradual
temperature
increase
of 15 #{176}C/minove
the next 6 mm to produce
a final
temperature
of 190 #{176}C
which was held for 9 mm. The total run time was 17 mu
The alkane elution times in minutes
were 2.67 (ethane)
4.56 (propane),
6.74 (butane),
and 8.85 (pentane).
For statistical
comparisons
we used the two-tailed Sti
dent’s t-testfor group data.
Results
Analytical
recovery
for each
linear over the gas concentrations
slopes, intercepts,
standard
error
Table
1. Detection
of the
four alkanes
we
Table 1 lists th
of estimates, and correb
tested:
of Alkanes
Added
Intercept
(iSE)
experi-
Department
of Medicine,
Room 820 CSB, 840 S. Wood St.,
University of illinois at Chicago, Chicago, IL 60612; and Veterans’
Administration
Hospital, West Side, Chicago, IL 60612.
Received August
25, 1986; accepted October 6, 1986.
140
Alkanes
Clapper
For detection of small quantities of alkanes that are present
in expired breath, these gases have hitherto been concena
trated, either by passing large volumes of breath through a
liquid-nitrogen-cooled precolumn
or by use of a closed
collection system. Here, we describe a technique for analyzing small volumes of gas from single-breath
samples from
humans,
in which no precolumn
is required.
Results are
linearlyrelated to sample concentrations of ethane, propane,
butane, and pentane in the range 0 to 13 nmol per literof air
(r = 0.999).
Within-run
coefficients
of variation were <15%.
Breath samples could be stored for as long as 10 h without
loss of the alkanes. We also report alkane concentrations
in
samples of alveolar gas and total breath collected
from
normal subjects. This technique appears to be well suited for
measuring alkane concentrations in single-breath samples.
Materials
of Volatile
Slope (±SE)
Ethane
Propane
Butane
Pentane
1.00
±
0.99
±
1.04
±
0.98
±
0.01
0.01
0.01
0.03
to Room
r
nmol/L
0.50
0.36
0.45
-0.26
±
±
±
±
0.05
0.04
0.11
0.19
Air
Stsnd. error
of estImate
0.106
0.102
0.253
0.445
1.OC
lOG
.99
.99
tion coefficients
(r) comparing
alkane added to room air and
alkane detected.
Results
of our within-run
and between-run
precision studies are summarized
in Table 2. The storage
capability
of the plastic syringe
system
(Table
3) was such
that the concentration
of each gas remained
stable for at
least 10 h.
The average concentrations of each alkane found in
alveolar and total breath samples are shown in Table 4. Also
listed are the average concentrations of the alkanes in room
air from our laboratory on five different days.
Table 2. Precision
of Alkane
Detection
in Room
WIthIn run
Mean
Mean
SD,
±
SD,
± 0.01
1.51
0.56
± 0.08
cv, %
14.3
0.01
0.13
1.12
1.40
0.97
8.83
±
0.19
19.3
±
±
0.51
1.07
9.27
0.51
1.42
±
0.01
2.55
±
13.6
±
3.77
1.53
8.98
±
0.10
8.63
0.80
±
±
0.04
0.14
0.05
0.47
0.98
0.36
0.06
±
0.38
0.41
±
0.20
0.95
±
0.32
10.3
3.63
±
0.75
0.36
9.69
2.37
±
±
±
0.58
0.30
5.60
±
0.68
4.72
±
0.82
±
0.82
9.34
±
1.20
Ethane
Propane
Butane
±
±
16.1
n
t
0.75
1.16
9.57
CV,
nmol/L
Pentane
Air
Between run
nmol/L
%
14.2
7.21
9.83
12.2
5.12
human
subjects without
the dilution
effect caused by air
from tracheal dead-space. Approximately
one-third of a total
human breath sample consists of tracheal dead-space air (6).
Our observation
of higher alkane concentrations
in alveolar
gas samples
probably reflects the peak alkane concentrations excreted through
the lungs more nearly accurately
4.12
than do measurements
conducted on total breath samples.
The small sample
volume
requirement
also makes our
method adaptable
for assaying rapid sequential
samples for
small laboratory
animals,
although
some sample dilution
from tracheal dead-space gas is likely. Our method can also
be used for in vitro experiments
conducted in small-volume
51 7
33:6
6.00
12.7
17.3
12.9
containers.
= 5 each.
Table
3. Stability
of Alkane
Stored
Storage
0
Ethane
±
8.90
10.1
±
duratIon,
0.26
0.35
0.51
9.37
9.58 ± 0.43 11.0
a Mean ± SD, nmol/L of air, n
b Significantly
different (p <0.05)
±
±
±
±
0.12
0.05
8.91
8.12
0.28 7.67
1.25 9.42
We thank Thomas J. Layden, M.D., for his helpful comments
and
review
of this manuscript,
and Carol Lane for her secretarial
assistance.
Supported
in part by grants from the Nil
(CA40002701) and the Veterans’
Administration
(FCP 041-55-103).
Syringes5
h
50
25
±
Butane
Pentane
9.50
in Plastic
10
9.42
8.73
Propane
ducible
and varied
linearly
with concentrations
over the
range
tested.
Our
method
is sensitive
enough
to detect
alkanes in ambient
air samples. We are not aware of any
published
method
for alkane
detection
in single-breath
samples or large-volume
gas collections
reported
in sufficient detail to permit a comparison
with our method.
The polyethylene
syringes used for sample storage in this
experiment
provided stable results for at least 10 h, after
which the concentrations
of some alkanes began to decline.
Preliminary
experience
with rubber-containing
disposable
syringes indicated
concentrations
were not stable in those
devices for more than 1 h.
Because our analysis method requires only 50-mi sample
volumes, it is well suited
for alveolar breath samples from
±
0.32
±
0.21
0.21
±
b
±
±
±
8.75
3.45
8.07
±
9.99
±
2.01
3.
from 0 storage
value.
References
1. Litov
RE,
Irving
DH, Downey JE, Tappel AL. Lipid peroxidation: a mechanism involved in acute ethanol toxicity as demonstrated by in vivo pentane production in the rat. Lipids 1978;13:305-7.
2. Muller
Table 4. Alkane
Human
Subjects
Concentration
in Breath from Normal
after Fasting
(n = 8) and in Room Air
(n = 5)
Alveolar
breath
Total
Mean
Ethane
Propane
±
SD, nmol/L
0.88
±
0.04
0.85
±
0.81
0.64
±
Butane
Pentane
3.7
±
0.20
0.09
1.2
0.87
0.54
2.4
±
Room air
breath
A, Sies H. Assay of ethane and pentane
organs and cells. Methods Enzymol 1984;105:311-19.
0.80
±
±
0.44
±
0.39
0.57
±
0.12
1.45
±
0.88
±
0.71
0.03
±
0.06
isolated
3. Koster U, Albrecht D, Kappus H. Evidence of carbon tetrachloride and ethanol-induced
lipid peroxidation
in vivo demonstrated
by
ethane production
in mice and rats. Toxicol Appl Pharmacol
1977;41:639-48.
4. Hempel
of aIr
0.04
0.08
from
V. May R, Frank
formation during halothane
1980;52:989-92.
5. Mets G, Gassull
H, Rammer
anaesthesia
H, Koster U. Isobutene
in man. Br J Anaesth
MA, Leeds AR, Blendis
TM, Jenkins
NA.
A
simple method of measuring
breath hydrogen in carbohydrate
malabsorption
by end-expiratory
sampling.
Clin Sci Mo! Med
iscussion
1976;50:237-40.
Our system for alkane
analysis
in breath or gas samples
rovided rapid results without the cumbersome
use of either
i precolumn
or rebreathing
apparatus.
Results were repro-
6. West JB. Disturbances
of respiratory
functions:
In: Harrison’s
principles of internal medicine, 8th ed. New York: McGraw Hill
Book Co., 1977:1333.
CLINICAL
CHEMISTRY,
Vol. 33, No. 1, 1987
141