Application of “High-Performance”

CLIN. CHEM. 26/10,
1499-1503
(1980)
of “High-Performance”LiquidChromatographyto the Study of
Application
Sphingolipidoses
M. David Uliman,’ Reed E. Pyeritz,2 Hugh W. Moser,3 David A. Wenger,4 and Edwin H. Koiodny5
Quantitative
high-performance
liquid chromatographic
analysis of perbenzoylated
sphingolipids
has been used
to study the correlations
of body chemistry to clinical
phenomena. Plasma sphingolipids were isolated from 32
Gaucher (/3-glucosidase
deficiency) and six Fabry (agalactosidase
deficiency)
patients by solvent partition and
chromatographic
separation on silicic acid columns.
Plasma sphingolipids from a patient undergoing plasmaexchange were separated from interfering lipids with reversed-phase columns. Liquid-chromatographic
analysis
of sphingolipids provides useful supportive information for
diagnoses because affected individuals are shown to
possess increased
circulating
concentrations
of the
pathognomonic sphingolipid. We also used this technique
to monitor sphingolipid concentrations in plasma and urine
sediment during plasma exchange of a patient with Fabry’s
disease. Regular plasma exchanges produced and maintained decreased
concentrations
of sphingolipids
in
plasma, but near pre-exchange concentrations
were observed within days after the therapy was terminated.
Additional
Keyphrases:
Gaucher’s disease
inherited
anomalies
diseases of enzyme deficiencies
Fabry’s disease . sphingolipids
analysis of urine sediment
The contribution of “high-performance”
liquid chromatography (HPLC) to the understanding of the sphingolipidoses is becoming increasingly important. Our report will
demonstrate some potential areas in which the clinical biochemistry laboratory can utilize analytical HPLC of perbenzoylated sphingolipids to obtain clinical information on these
diseases.
The sphingolipidoses
are caused by lysosomal enzyme deficiencies, and clinical symptoms manifest themselves through
the resulting accumulation
of unmetabolized
sphingolipids.
The discovery of lysosomal enzyme deficiencies
has led to
enzyme assays that can be used for the pre- and postnatal
diagnosis of the specific diseases (1). Although the enzyme
assays provide reasonably reliable results, the clinician would
be greatly assisted by supportive biochemical information
on
the accumulation
and distribution
of the disease-specific
sphingolipids
and their precursors when “borderline”
activities are measured. Early quantitative
HPLC analysis of
sphingolipids
could also help the physician decide what enzyme assays should be performed. Further, because detection
of an enzyme deficiency
does not necessarily predict the
clinical course of the disorder, especially when various subtypes (e.g., infantile, juvenile, and adult) exist, biochemical
Center for Disease Control, Atlanta, GA 30333; current address:
V.A.Hospital, Bedford, MA 01730.
2 Johns Hopkins
Hospital,
Baltimore, MD 21205.
John F. Kennedy Institute, Baltimore, MD 21205.
University of Colorado, Health Sciences Center, Denver, CO
80220.
5E. K. Shriver Center, Waltham, MA 02154.
Received May 12, 1980; accepted June 26, 1980.
information
relative to the probable clinical
order would be valuable to the clinician
course of a disand the family
counselor in their respective health care roles. Finally,
clinical
evaluation
of the efficacy
of therapeutic
attempts
the
in
the sphingolipidoses is difficult because clinical symptoms,
which can take years to become manifest, may not show an
immediate clinical response to therapeutically altered body
chemistry. The HPLC analysis of tissue and body fluid
sphingolipid fluctuations during therapeutic attempts will
allow the correlation of those fluctuations to the short- and
long-term clinical response to therapy.
Before the development of HPLC, quantitative analysis of
the sphingolipids
required preparative thin-layer chromatography and subsequent measurement
of hexose by destructive
colorimetric
or gas-liquid-chromatographic
techniques (2, 3). Insufficient sensitivity and laborious
work-ups
for those procedures, combined with the low concentrations
of sphingolipids in most tissues and body fluids, made largescaleprocessing
of specimens impractical; this encumbered
research into sphingolipid biochemistry and correlationsto
clinicalsymptomatology in sphingolipidoses.The relatively
convenient and sensitive quantitative HPLC analysis for
perbenzoylated sphingolipids, with ultraviolet detection at
230 nm, will help the clinical biochemistry laboratory apply
fundamental sphingolipid biochemistry to diagnosis, prognosis, and therapeutic monitoring of sphingolipidoses.
Materials and Methods
Specimens
Blood specimens from Gaucher’s disease [/3-glucosidase (EC
3.2.1.21)
3.2.1.22)
deficiency]
and Fabry’s disease [a-galactosidase
deficiency]
containing
tubes.
were collected in green-top
(EC
heparin-
Plasma was isolatedby centrifugingthe
specimens at 1000 X g for 10 mm.
Plasma and urine sediment obtained during plasma exchanges were from a 29-year-old man with Fabry’s disease.
Plasma-exchange
therapy
was initiated
in an attempt
to lower
circulating
amounts of galactosyllactosylceramide
(GalLac-Cer). After collection of baseline data (4) the patient was
subjected to four sessionsof plasma exchange over the course
of 12 months, each sessionconsistingof three 2-L exchanges
spaced two days apart. Data from the second plasma-exchange
session are presented here to illustrate
the ability of HPLC
to monitor fluctuations in circulating sphingolipid. The third
sessionwas a sham exchange to testthe clinical efficacy of the
procedure.
Isolation and Derivatization
of Neutral
Sphingolipids
from Plasma and Urine Sediment
Plasma sphingolipids from Gaucher and Fabry patients
were isolated as described by Ullman and McCluer (5).
Sphingolipids from plasma samples obtained during plasma
exchanges were isolated from total lipid extracts with the use
of reversed-phase,
rapid sample-preparation
columns
(C18-Sep-Pak; Waters Associates, Milford, MA 01757). The
lipid extract from 0.25 mL of plasma was dissolved in 1 mL
of methanol/acetone (95/5 by vol) and placed onto a rapid
CLINICAL CHEMISTRY, Vol. 26. No. 10, 1980
1499
the culture tube was capped and shaken vigorously with a
mechanical shaker for 15 mm. The contents of the tube were
filtered through a sintered-glass funnel fitted with a disc of
No. 1 filter paper and into a 100-mL filter flask. The culture
tube was rinsed with 5 mL of chloroform/methanol
(1/1), and
the rinse was filtered and collected with the previous fraction.
The contents of the filter flask were transferred to another
20 X 150 mm screw-cap culture tube, and the solvent was removed at room temperature under a stream of nitrogen. The
filter funnel residue was returned to the original 20 X 150mm
screw-cap culture tube and re-extracted as described above,
but with chloroform/methanol
(2/1). The extract was combined with the first residue and the solvent removed as before.
After the extract was redissolved in 24 mL of chloroform/
methanol (2/1), 6 mL of 8.8 g/L KC1 was added. The contents
were mixed and centrifuged to clarify the biphasic layers. The
upper phase was removed and discarded, and the lower phase
was washed with 12 mL of methanol/water (1/1), mixed, and
centrifuged as before. The upper phase was again withdrawn
and discarded. The lower phase was dried at room temperature under a stream of nitrogen and redissolved in 30 mL of
chloroform. We analyzed 0.5-mL aliquota of this solution after
each had been subjected to mild alkaline hydrolysis (2). The
urine-sediment sphingolipids were derivatized by perbenzoylation (5) and quantitated by HPLC.
w
U,
z
0
a.
U,
‘U
>
HPLC
TIME (0.01 MINI
Fig. 1. Chromatogram
of perbenzoylated plasma sphingolipids,
illustrating the usual increase In adult Gaucher Gic-Cer
Procedure
Quantitative HPLC analysis was performed by a slight
modification of the method described by Ulinian and McCluer
(6). Perbenzoylated sphingolipids were dissolved in carbon
tetrachloride, injected onto a pellicular silica gel (Zipax; E. I.
Dupont de Nemours, Inc., Wilmington, DE 19899) column (2.1
mmX 50cm), and eluted with a 10-mm linear gradient of 2/98
to 17/83 (by vol) dioxane/hexane solution with a flow rate of
3 mL/min. Absorbance at 240 nm was recorded. The variable-wavelength ultraviolet detector (Model 970A; Tracor
Instruments, Austin, TX 78767) used for the analyses was
equipped with a high-pressure, flow-through reference cell
to balance the residual absorbance of dioxane in the gradient
(6).
Results
sample-preparation
column that had been pretreated with at
least 10 mL of the methanol/acetone solvent. The eluate was
collected in a 13 X 100 mm screw-cap culture tube. A 1-mL
rinse of the sample tube was applied to the column and collected with the first eluate. The collected fraction was then
re-applied to the reversed-phase column. A 2-mL rinse of the
sample tube was also applied to the column, followed with an
additional 4 mL of solvent. The total eluate was collected and
evaporated at room temperature under a stream of nitrogen,
and the residue was subjected to mild alkaline hydrolysis, as
described by Vance and Sweeley (3). The crude sphingolipids
were then perbenzoylated according to the method of Ullman
and McCluer (5). Recovery of sphingolipids from the reversed-phase column was estimated by measuring [1-14C]N-stearoyl-glucosylsphingosine
that had been added to the
plasma total-lipid extract before the reversed-phase column
purification step.
Urine sediments were isolated by filtration of 12-h urine
specimens through No. 1 fluted filter paper (Whatman, Inc.,
Clifton, NJ 07014) that had been previously washed with
chloroform/methanol
(2/1 by vol) and dried. The filtrate and
filter paper were placed into a 20 X 150 mm screw-cap culture
tube and 30 mL of chloroform/methanol
(1/1 by vol) was
added. The filter paper was pulverized with a glass rod, and
1500
CLINICAL CHEMISTRY, Vol. 26, No. 10, 1980
The average of two experiments indicated that about 96%
of the [1-14C]N-stearoyl-glucosylsphingosine
was recovered
from the reversed-phase rapid sample-preparation
column.
The reproducibility of the analysis was comparable with that
reported previously (5).
Patients with adult Gaucher’s disease (Figure 1) frequently
possess high concentrations of glucocerebroside (Glc-Cer) in
plasma (7). HPLC analysis of plasma Glc-Cer concentrations
from 32 adult Gaucher patients, however, ranged from high
normal (4.41 zmol/L) to a sixfold increase (32.71 zmol/L) over
the range of control values (2.53 to 6.15 jtmol/L). Although the
range of plasma Glc-Cer concentrations in Gaucher’s disease
overlapped with control values, the ratio of Glc-Cer to dihexosylceramide (Lac-Cer) was invariably higher (1.81 to 5.78)
than control values ( 1.20). Plasmas from three of the adult
Gaucher patients were also examined after splenectomy and
were shown to have no significant change in their sphingolipid
concentrations. The analysis of plasma sphingolipids from
Fabry patients demonstrated
the anticipated
(8) increase in
Gal-Lac-Cer in all cases (Table 1).
We also used HPLC to follow alterations
of sphingolipid
concentrations in plasma and urine sediment during the
course of plasma exchanges in a patient with Fabry’s disease.
The patient’s plasma Gal-Lac-Cer exceeded that in controls,
but other sphingolipids were in the normal range (Figure 2).
Table 1. Plasma Sphlngoliplds in SIx Patients with Fabry’s Disease
Patlsnt
1
Glc-C.r
4.80
moUL
Oal-1.ac-Cr
5.52
GIoboslds
1.45
2
6.01
5.85
7.02
1.64
3
4.11
3.93
3.89
1.01
4
5.50
4.00
5.78
1.46
5
3.08
2.47
4.35
6.4 ± 1.5
3.34
3.31
1.7±0.3
6
Control (n
a
SpNM
Lac-Cr
4.30
4.50
=
5.1 ± 1.5
1.36
0.88
1.7±0.6
Values are mean ± SD.
Regular plasma exchange produced and maintained lowered
plasma Gal-Lac-Cer concentrations, but near pre-exchange
concentrations were observed five days after the therapy was
terminated
(Figure 3a). HPLC analysis clearly demonstrated
that the sham exchange did not affect circulating concentrations of Gal-Lac-Cer (Figure 35). From 22 to 29 mg of GalLac-Cer was removed in each plasma-exchange session.
HPLC
assay of urine-sediment
sphingolipids
from the
Fabry patient revealed a marked increase in Gal-Lac-Cer
(100-700 nmol/12 h) over control values (7 nmol/12 h) (Figure
4). Quantitation of urine-sediment
sphingolipids from the first
session
of plasma exchange showed that urine-sediment
Gal-Lac-Cer excretion increased after each exchange, and that
the increase approximated the increased urine volume accompanying each plasma exchange.
Discussion
Past procedures forthe quantitativeanalysisof sphingoli-
pids frequently involved a chromatographic step, usually on
a silicic acid column (3, 5), to isolate a sphingolipid fraction
from a total plasma lipid extract. Substantial amounts of
sphingolipids were lost, however, because of their irreversible
binding to the silicic acid; moreover, silicic-acid fractionation
of sphingolipids from plasma samples that had been obtained
during plasma exchange yielded highly variable recoveries,
for as yet unknown causes. Substituting a C18 reversed-phase
rapid sample-preparation column for the silicic acid column
E
....
0
LU
U)
E
C
I
z
2
3
4
5
0
a.
U,
‘U
>
6 7
DAY
8
9 10
Ii
C.,
U
0
SHAM
PLASMAPHERESES
-J
a
b.
CD 8.0
4
7.0
4
6.0
5.0
4.0
3.0
2.0
1.0
AM
PulA P1* PA
P1*
PIA
P1*
P*
12345678
PJA
P}A
PJA
PA
PJA
Pj
91011121314
DAY
TIME (0.01 MIN)
Fig. 2. Chromatogram
of perbenzoylated plasma sphingoliplds,
illustratIng the increase In Fabry disease Gal-Lac-Cer
Fig. 3. Alterations in plasma Gal-Lac-Cer (a) during second
session of plasma-exchange
therapy and (b) during sham
plasma-exchange therapy
CLINICALCHEMISTRY,Vol. 26, No. 10, 1980
1501
atively simple and reproducible
umn
Gel-Lea
-
Car
I
TIME
(0.01
MIN)
Fig. 4. Urine-sedIment perbenzoylated sphingolipids from Fabry
patient and control
to isolate crude sphingolipids from plasma total lipid extracts
eliminated major lossesand provided reproducible recoveries.
The sphingolipids were essentiallynot retained
by the reversed-phase
column, which provided
minimal opportunity
for the compounds
to be irreversibly
bound to the packing
material.
We had selected radlolabeled Glc-Cer for the recovery study because the sphingolipid
theoretically had the
greatest retention
on the column and, therefore,
had the
greatest opportunity
to become irreversibly bound. More
extensive recovery studies of all the major plasma sphingolipids will be reported
later.
The procedure is probably not acceptable for tissues or body
fluids that contain significant
quantities
of sphingomyelin,
because sphingomyelmn
(a) is collected with the sphingolipid
fraction; (b) is stable to alkaline hydrolysis; (c) does not yield
only one peak when derivatizedby the method used forthese
studies; and (d) is eluted with other sphingolipids in the
chromatographic system. The very low concentration
of
plasma sphingomyelin permitted the isolation
ofsphingolipids
by reversed-phase columns.
The quantitation
of urine-sediment
1502
sphingolipids
CLINICAL CHEMISTRY, Vol. 26, No. 10, 1980
is a rel-
technique, requiring no col-
fractionation.
The assumption that increased circulating Glc-Cer was in
some part responsible for the transport and deposition of
accumulated lipid in the sphingolipidoses suggested that
valuable clinical information would be provided by quantitative HPLC analysis of plasma sphingolipids. Quantitative
analysis revealed that absolute adult Gaucher plasma Glc-Cer
concentrations overlapped control values. Because the analytical procedure permitted the quantitation
of other circulating sphingolipids, however, we noted that the ratio of
Glc-Cer to Lac-Cer produced no overlap between patients and
controls. The increased Glc-Cer to Lac-Cer ratio has been
found to occur in some forms of Niemann-Pick disease (9),
and so does not constitute a unique diagnostic tool; it is useful,
however, as a preliminary test for suggested enzyme analysis
or confirmatory information. The consistent relative increase
of Glc-Cer in the plasma of adult Gaucher patients and the
absolute increase of Gal-Lac-Cer in the plasma of Fabry patients suggest that analytical HPLC may become a very useful
supplement to established diagnostic procedures for the
sphingolipidoses. Whether analytical HPLC of sphingolipids
will help identify the heterozygous states is yet to be determined.
The unaltered sphingolipid concentrations after splenectomy in adult Gaucher’s disease is information that can help
predict the anticipated postoperative clinical course of the
disorder. Therefore, analytical HPLC will contribute to more
accurate prognosis of Gaucher’s disease and other sphingolipidoses when the pathogenic mechanisms of the accumulated
sphingolipids are understood.
HPLC has been successfully utilized to monitor the decrease and gradual re-equilibration of plasma sphingolipids
during plasma exchanges. Should this or other potential
therapeutic modalities that regulate circulating amounts of
sphingolipids prove to be of clinical benefit, quantitative
HPLC analysis of sphingolipids could become a useful clinical
assay to monitor the maintenance of sphingolipid concentrations in the most therapeutically advantageous range.
The convenient analysis of urine-sediment sphingolipids
has also been demonstrated. Because urine sediment contains
a large constituency of renal epithelial cells in certain disorders, it can be envisaged as an “indirect” noninvasive biopsy
of the renal system. Measurement
of urine-sediment
sphingolipids during the course of chronic therapy could reveal alterations in intracellular content of the pathognomonic
lipids. Difficulties in urine-sediment analyses arise from the
variation of urine content and volume with daily events. More
useful information on intracellular sphingolipid content in
the sphingolipidoses will be gained when a quantifiable
marker for urine-sediment cellular composition (e.g., renal
epithelial cell-specific protein) is found.
The relevance of HPLC to the clinical biochemistry laboratory in the future holds great potential. Recent improvements in the derivitization of sphingolipids (10) may some day
make them amenable to automated analysis. A procedure for
the preparative HPLC isolation of sphingolipids (11) may lead
to development of uniform standards and substrates for
clinical biochemistry laboratories that conduct sphingolipidoses-related assays. Our understanding
of sphingolipid
biochemistry and its correlation to clinical phenomena will
determine the extent to which HPLC can be of value in the
clinical biochemistry laboratory.
We gratefully acknowledge the expert technical assistance of Mr.
Ricky Akins. D.A.W. is supported in part by NIH grants HD 08315,
HD 10494,and NS 10698, and by a Research Career Development
Award (NSOO1O8); H.W.M. is supported in part by grants MS-13513
and HD-10981 from the Public Health Service.
References
1. Brady, R. 0., Enzymological approach to the lipidoses. Anal. Clin.
Lab. Sci. 7, 105 (1976).
2. Warren, K. R., Schafer, I. A., Sullivan, J. C., et al., The effects of
N-hexyl-O-glucosyl
sphingosine on normal cultured human fibroblasts: A chemical model for Gaucher’s disease. J. Lipid Res. 17, 132
(1976).
3. Vance, D. E., and Sweeley,.C. C., Quantitative
determination
of
the neutral glycosyl ceraniides in human blood. J. Lipid Res. 8,621
(1967).
4. Ullman, M. D., Pyeritz, R. E., McCluer, R. H., and Moser, H. W.,
Biochemical response of Fabry’s disease to plasmapheresis.
Trans.
Am. Soc. Neurochem.
10,65 (1979).
5. UlIman, M.D., and McCluer, RH., Quantitative
analysis of plasma
neutral glycosphingolipids by high performance liquid chromatography of their perbenzoyl derivatives. J. Lipid Res. 13, 371 (1977).
6. Ullman, M. D., and McCluer, R. H., Quantitative
of perbenzoylated
microanalysis
neutral glycosphingolipid by high performance
liquid chromatography
with detection at 230 nm. J. Lipid Res. 19,
910 (1978).
7. Brady,R. 0.,Glucosylceramidelipidoses:
Gaucher’sdisease.
In
The Metabolic Basis of Inherited Disease, J. B. Stanbury, J. B.
Wyngaarden,
and D. S. Fredrickson, Eds. McGraw-Hill, New York,
NY, 1978,pp 731-746.
8. Desnick,R. J.,Klionsky,B.,and Sweeley,C. C.,Fabry’sdisease.
In The Metabolic Basis of Inherited Disease (see ref. 8), pp 810840.
9. Dacremont, G., Kint,S. A., Carton, D., and Cocquyt, G., Glucosylceramide in plasma of patients with Niemann-Pick disease. Clin.
Chim. Acta 52, 365 (1974).
10. Gross, S. K., and McCluer, R. H., High performance liquid
chromatographicanalysisof neutralglycosphingolipids
as their
per-O-benzoylderivatives.
Anal. Biochem. 102,429 (1980).
11. McCluer,R. H.,and Ullman, M. D.,Preparativeand analytical
high performance liquid chromatographyof glycolipids. ACS Symp.
Ser. 128,20 (1979).
CLINICALCHEMISTRY,Vol. 26, No. 10, 1980
1503