CLIN. CHEM.26/3, 403-405
(1980)
Amniotic Fluid Phospholipids Measured by Continuous-Development
Thin-Layer Chromatography
Mark D. Kolins, Emanuel Epstein, W. Harold Civin, and Sheldon Weiner’
We describe a one-dimensional thin-layer chromatographic
system for separation of amniotic fluid phosphatidyicholine, sphingomyelin,
phosphatidylglycerol,
phosphatidylinositol, phosphatidylserine, and phosphatidylethanolamine
in amniotic fluid. We utilize short-bed continuous development and “high performance”
thin-layer chromatography. Phospholipids are detected with an antimony molybdate staining reagent and quantitated by transmittance
densitometry. This system is more sensitive to changes
in lecithin/sphingomyelin
ratios than are planimetric
evaluations.
AddItIonalKeyphrases: lecithinlsphingomyelin
ratio
fetal
status
Amniotic
fluid phospholipids
are currently
assayed as an
aid in predicting
fetal lung maturity.
The measurement
most
commonly
used in the clinical laboratory
is the ratio of
phosphatidylcholine
(lecithin)
to sphingomyelin.
Although
use of the lecithin/sphingomyelin
(L/S) ratio is satisfactory
when applied to normal pregnancies,
its predictive value is less
certain
in pregnancies
complicated
by maternal
diabetes
mellitus or toxemia (1, 2). This fact has led to investigations
of the potential
use of other amniotic fluid phospholipids
in
predicting
fetal lung maturity, and phosphatidylglycerol
assay
is now considered
to be important
in predicting
fetal lung
status (2-4). In addition,
in cases of respiratory
distress syndrome, the proportions
of phosphatidylserine
and phosphatidylethanolamine
increase.
The assay of these two phospholipids also may be helpful in distinguishing
the mature and
immature
fetal lung (3).
The one-dimensional
chromatographic
systems (4, 5) for
obtaining
L/S ratios and relative percentage
of phosphatidylglycerol
fail to resolve phosphatidylserine
and phosphatidylethanolamine.
In current
clinical and investigtional
laboratory
methods
for individual
quantitation
of major
amniotic
fluid phospholipids,
two-dimensional
thin-layer
chromatography
is used.
We describe
here a short-bed
continuous-development
technique
of “high-performance”
thin-layer
chromatography
in one dimension,
with which we can distinguish
the major
phospholipids
of amniotic
fluid
(phosphatidylcholine,
sphingomyelin,
phosphatidylinositol,
phosphatidylserine,
phosphatidylethanolamine,
and phosphatidyiglycerol).
Materials and Methods
Reagents
and Apparatus
We obtained
egg lecithin/bovine
sphingomyelin
mixtures
in weight ratios of 1/1, 2/1, 3/1, 4/1, and 5/1 from Applied
Science Labs., Inc., State College, PA 16801. Bacterial
phosphatidylglycerol
and bovine phosphatidylethanolamine
were
obtained
from Supelco, Bellefonte,
PA 16823. Porcine brain
Departments
of Clinical Pathology
and 1 Obstetrics
and Gynecology, William Beaumont
Hospital, 3601 West Thirteen
Mile Rd., Royal
Oak, MI 48072.
Received July 27, 1979; accepted Dec. 4, 1979.
phosphatidylserine
and porcine liver phosphatidylinositol
were supplied
by Miles Laboratory,
Elkhart,
IN 46514. We
used Merck Silica Gel 60 high-performance
thin-layer
chromatographic
plates (10 X 10 cm) from Scientific
Products,
Romulus,
MI 48174; an ACD-18 automatic
computing
densitometer from Gelman Instrument
Company, Ann Arbor, MI
48106; and a short-bed
continuous-development
chamber
from Regis Chemical Co., Morton Grove, IL 60053. The mobile
phase consisted of petroleum ether (boiling point range 20-40
oC)/chloroform/methanol/acetic
acid (3/5/1.5/1
by vol).
We used an antimony
molybdate
phosphate
reagent (6) to
stain phospholipids,
consisting
of a 10/1/3/6 (by vol) combination of 2.52 mol/L sulfuric acid, and the following aqueous
solutions:
8.2 mol/L
antimony
potassium
tartrate,
32.3
mmol/L ammonium
molybdate,
and 0.1 molfL ascorbic acid.
This reagent mixture is stable for 28 days at 4 #{176}C.
We combined 8 mL of this mixture with 1 mL of isopropanol
and 13
mL of distilled water to prepare the active staining reagent,
which must be used promptly.
Method
Amniotic fluid specimens were obtained by transabdominal
aspiration
between 30 and 42 weeks of gestation.
Fluids were
obtained
at cesarean
section by needle aspiration
before
rupture of the amniotic
membranes.
Meconiumand bloodstained fluids were discarded.
Acetone-precipitable
phospholipids
were isolated
from
amniotic fluids as described
by Gluck et al. (7). The isolated
phospholipids
were dissolved in 20 tL of chloroform,
and 10
was applied to the high-performance
thin-layer
chromatography
plates in streaks about 1 cm long. The plates were
placed at a reduced pressure
in a desiccator-oven
at room
temperature
for 5 mm. The mobile phase was placed in the
short-bed
continuous-development
chamber
and the plates
were positioned
to give the maximum
bed length. Plates were
developed for 50 mm, then blown dry with cool air from a hair
dryer. The antimony
molybdate
phosphate
reagent
was
poured over the plate and left at room temperature
for 1 mm.
The stained plates were dried with warm air from a hair dryer
for 5 mm to develop the phospholipid
bands. Phospholipids
were identified
by migration
identical with standards.
Each
phospholipid
was measured
by transmittance
densitometry
at 620 nm and expressed
as a percentage
of total phospholipids.
Results
AnalyticalVariables
Separation
of phospholipids.
We could clearly separate all
six of the major phospholipids
of amniotic
fluid with this
system (Figure 1).
Linearity.
Linearity was assessed by utilizing commercially
prepared
lecithin/sphingomyelin
mixtures.
Both our procedure with use of densitometry
and the method of Gluck et al.
(10) with planimetry
gave linear results for L/S weight ratios
of 1 through 5. The ratios by densitometry
showed a greater
response to changes in the lecithin and sphingomyelin
weight
mixtures
(Figure 2).
CLINICALCHEMISTRY,Vol.26,
No. 3, 1980
403
15
8
7
6
0
a
o5
-J
4
3
2
Fig. 1. Separation of the amniotic fluid phospholipids from amniotic fluid obtained
S, sphingomyelin;
1
at cesarean section
P1,phosphatidylinositol; PS,phosphatidylserine;
L, lecithin;
PE. phosphatidylethanolamIne;
P0, phosphatidylglycerol;
X, unidentifIed
Color stability.
The phospholipids
appear as blue bands
within 5 mm. The color is stable for at least 30 mm.
Accuracy
and precision.
Commercial
mixtures
of phosphatidylcholine
and sphingomyelmn at weight ratios of 1/1,2/1,
3/1,4/1, and 5/1 were assayed by the described
method. Corresponding
L/S ratios in duplicate
determinations
were 1.30
and 1.38, 1.97 and 1.93, 2.27 and 2.33, 3.08 and 3.06, and 3.98
and 3.82.
We assessed
within-day
precision
for the high L/S ratio
range by running 10 aliquots of pooled amniotic fluid obtained
at cesarean section. The mean L/S ratio was 9.13 (SD 0.88),
with a CV of 9.68%. The mean percentage
of phosphatidylglycerol (of total phospholipid)
was 3.23 (SD 0.59), with a CV
of 18.4%. The mean percentage
of phosphatidylmnositol
was
26.3 (SD 2.7), with a CV of 10.2%.
Within-day
precision
for the low L/S ratio range was assessed by isolating
10 aliquots of an equi-weight
mixture of
phosphatidylcholine
and sphingomyelin.
The mean L/S ratio
L
Borer’s
Method
HP-TLC
Fig.3.Distribution
ofLIS ratios inpairedspecimens runbythe
presentmethod (HP-TLC)and thatofGlucketal.(10) (Borer’s
method)
was 0.85 (SD 0.08), with a CV of 9.56%. Day-to-day
precision
was assessed by isolating phospholipids
on four consecutive
days. Nine aliquots of the 1/1 mixture of lecithin and sphingomyelin were run each day. The mean L/S ratio was 0.96 (SD
0.20, CV 21.3%).
Clinical
Studies
We quantitated
phospholipids
in 22 specimens.
Phosphatidylglycerol
was identified in all these, with L/S ratios ranging
from 2.9 to 15.3. None of these infants subsequently
developed
the respiratory
distress syndrome.
L/S ratios were assessed
by our procedure
and by the
method of Gluck et al. (10) in 30 paired specimens.
There was
clustering
of ratios by the latter method,
in contrast
to the
broader distribution
by ours, which suggests that the present
method discriminates
better (Figure 3). Two specimens
obtained at cesarean section gave low L/S ratios (1.8 and 1.9) by
the method of Gluck et al. (10); our method gave higher values
(3.0 and 2.7, respectively).
Fluid from the former case also
contained
phosphatidyiglycerol.
Neither case developed
the
respiratory
distress syndrome.
Discussion
P1
-‘-
ss
Fig.2.Densitometric
traceofamnioticfluid
phospholipids
Abbreviations as in Fig. 1
404
CLINICAL
CHEMISTRY,
Vol. 26, No. 3, 1980
In this high-performance
thin-layer
chromatographic
system, silica gel is used as in conventional
thin-layer
chromatography,
but the particle size of the adsorbent
material
is
more uniform, about 7 sum. This allows for a densely packed
thin layer, which results in better resolution
between similar
phospholipids
(Figure 1). In conventional
thin-layer
chromatography,
in general, the particle distribution
is in the range
of 5 to 30 zm; this less-dense adsorbent
layer results in larger
and more diffuse spots.
The short-bed continuous-development
chamber can better
4
3
0
a
CO
-J
2
Slope
0.20
A
0.5
1
Commercial
2
Lecithin
3
(mg)mixed
4
5
with 1mg Sphingomyelin
Fig. 4. Sensitivity of the present method (HP-TLC) and that of
Gluck et al. (10) (Borer’s method) compared by use of commercially prepared L/S mixtures
The slope for the present method is nearly three times greater, indicating its
greater sensitivity to changes in iecithin/sphingomyelin concentration
resolve substances
with similar Rf’s. The Rf’s of substances
generally vary with solvent (mobile phase) strengths
(8), but
if solvent concentrations
are changed too much, the solvent
may traverse the chromatography
plate before the substances
have moved from the origin. The short-bed
continuous-development
chamber
overcomes
this difficulty
because the
thin-layer
chromatographic
plate protrudes
out of the otherwise sealed chamber, permitting
the solvent to move up the
plate and evaporate
outside the chamber. Therefore,
solvent
is continuously
moving up the plate at a constant velocity. In
addition,
as the distance to be traversed
by the solvent is decreased,
solvent
velocity
increases
(9). This continuous
movement
of solvent results in resolution
equivalent
to that
with a much longer chromatographic
plate or column, and the
high solvent velocity shortens
the time required.
We found phosphatidylinositol
and phosphatidylcholine
to have similar R1’s in solvent systems utilized by other investigators
(4). This was also true of phosphatidylsermne
and
phosphatidylethanolamine.
Thus, to achieve separation,
we
used the short-bed
continuous-development
system
and
added petroleum
ether to reduce the polarity
of the solvent.
The greater sensitivity
of the present method as compared
with commonly used planimetric
evaluation
of L/S ratios (10)
is emphasized
by comparison
of the slopes obtained
(Figure
4). The slope for the present method is nearly three times that
for the method of Gluck et al. (10), indicating
that our system
gives a greater change in the calculated
L/S ratio for a change
in the L/S weight mixtures.
Although
we did not obtain
identical values for commercial
L/S mixtures
and L/S ratios
determined
by our procedure,
there is a constant
proportion
between the two (Figure 4), as there also is for the method of
Glucketal.
(10).
The popularity
of the method of Gluck et al. lies in avoiding
the need for sulfuric acid charring;
our method also has this
advantage.
In addition,
densitometric
quantification
of
phospholipids
enables a more rapid assessment
of all phospholipid bands than planimetry.
With our system, one-dimensional
separation
of the major
phospholipids
currently
being evaluated
in predicting
fetal
lung status is possible.
Studies
by Hallman
et al. (3) have
suggested
that phosphatidylethanolamine
and phosphatidylserine
make up a significantly
higher percentage
of total
phospholipids
in infants who develop the respiratory
distress
syndrome. Isolation and quantification
of these phospholipids
from amniotic fluids may prove to be an indicator of immature
fetal lung status, in contrast to phosphatidylglycerol,
which,
when present, indicates mature fetal lung status (11, 12). The
converse, however, is not true: the absence of phosphatidylglycerol does not necessarily
indicate
an immature
lung.
Therefore,
phosphatidylethanolamine
and phosphatidylserine
may yield useful information
in complicated
pregnancies,
when the L/S ratio is frequently
high despite immature
lung
status (3).
References
I. Gluck, L., and Kulovich, M. V., Lecithin/sphingomyelin
ratios in
amniotic fluid in normal and abnormal pregnancy. Am J. Obstet.
Gynecol.
115, 539-546 (1973).
2. Cunningham, M. D., Desai, N. S., Thompson, S. A., and Greene,
J. M., Amniotic fluid phosphatidylglycerol
in diabetic pregnancies.
Am. J. Obstet. Gynecol. 131, 719-724 (1978).
3. Hallman, M., Feldman, B. H., Kirkpatrick,
E., and Gluck, L.,
Absence of phosphatidylglycerol
(PG) in respiratory distress syndrome in the newborn. Pediat. Res. 11, 714-720 (1977).
4. Tsai, M. Y., and Marshall, J. G., Phosphatidylglycerol
in 261
samples of amniotic fluid from normal and diabetic pregnancies, as
measured by one-dimensional
thin-layer chromatography.
Clin.
Chem. 25, 682-685 (1979).
5. Gotelli, G. R., Stanfill, R. E., Kabra, P. M., et al., Simultaneous
determination of phosphatidylglycerol and the lecithin/sphingomyelin
ratio in amniotic fluids. Clin. Chem. 24, 1144-1146 (1978).
6. Manual
of Methods
for Chemical
Analysis
of Water
and Wastes
(EPA-625/6-74-003),
U. S. Environmental
Protection
Agency,
Washington, D. C., 1974, pp 249-255.
7. Gluck, L., Kulovich, M. V., Borer, R. C., et al., Diagnosis of the
respiratory distress syndrome by amniocentesis. Am. J. Obstet. Gynecol. 109, 440-445 (1971).
8. Perry, J. A., A new look at solvent strength selectivity and continuous development. J. Chromatogr.
165, 117-140 (1979).
9. Shortbed/Continuous
Development
Technical
Manual,
Regis
Chemical Co., Morton Grove, IL.
10. Gluck, L., Kulovich, M. V., and Borer, R. C., Estimates of fetal
lung maturity. Clin. Perinatol.
1, 125 (1974).
11. Kulovich, M. V., Hallman, M. B., and Gluck, L., The lung profile.
I. Normal pregnancy. Am. J. Obstet. Gynecol.
135, 57-63 (1979).
12. Kulovich, M. V., and Gluck, L., The lung profile. II. Complicated
pregnancy. Am. J. Obstet. Gynecol.
135, 64-70 (1979).
CLINICAL
CHEMISTRY,
Vol. 26, No. 3, 1980
405