NIH Public Access

Author Manuscript
Anal Biochem. Author manuscript; available in PMC 2011 June 1.
Published in final edited form as:
NIH-PA Author Manuscript

Anal Biochem. 2010 June 1; 401(1): 154–161. doi:10.1016/[Link].2010.02.023.

QUANTIFICATION OF CERAMIDE SPECIES IN BIOLOGICAL

SAMPLES BY LIQUID CHROMATOGRAPHY-ELECTROSPRAY
TANDEM MASS SPECTROMETRY

Takhar Kasumov1,2,*, Hazel Huang2, Yoon-Mi Chung3, Renliang Zhang4, Arthur J.

McCullough1,2, and John P. Kirwan1,2
1 Department of Gastroenterology & Hepatology, Cleveland Clinic, Cleveland, OH

2 Department of Pathobiology, Cleveland Clinic, Cleveland, OH

3 Department of Cell Biology, Cleveland Clinic, Cleveland, OH
4 Department of Research Core Services, Cleveland Clinic, Cleveland, OH
NIH-PA Author Manuscript

Abstract
We present an optimized and validated liquid chromatography-electrospray ionization tandem mass
spectrometric (LC-ESI-MS/MS) method for the simultaneous measurement of concentrations of
different ceramide species in biological samples. The method of analysis of tissue samples is based
on Bligh and Dyer extraction, reverse-phase HPLC separation and multiple reaction monitoring of
ceramides. Preparation of plasma samples also requires isolation of sphingolipids by silica gel column
chromatography prior to LC-ESI-MS/MS analysis. The limits of detection and quantification are in
a range of 5–50 pg/ml for distinct ceramides. The method was reliable for inter-assay and intra-assay
precision, accuracy and linearity. Recovery of ceramide subspecies from human plasma, rat liver and
muscle tissue were 78–91%, 70–99%, and 71–95%, respectively. The separation and quantification
of several endogenous long-chain and very-long-chain ceramides using two non-physiological odd
chain ceramide (C17 and C25) internal standards was achieved within a single 21 min
chromatographic run. The technique was applied to quantify distinct ceramide species in different
rat tissues (muscle, liver, and heart) and in human plasma. Using this analytical technique we
demonstrated that a clinical exercise training intervention reduces the levels of ceramides in plasma
of obese adults. This technique could be extended for quantification of other ceramides and
NIH-PA Author Manuscript

sphyngolipids with no significant modification.

Keywords
sphingolipid; ceramide; lipidomics; insulin resistance; apoptosis; ESI-MS/MS

*Address for correspondence: Takhar Kasumov, PhD, Department of Gastroenterology & Hepatology and Pathobiology, Lerner Research
Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue/NE 4-206, Cleveland, OH 44195, Phone: 216 444 4189, kasumot@[Link].
C18, C24, C18:1 and C24:1 ceramides are designated for ceramides containing saturated and monounsaturated fatty acids, stearic, oleic
and nervonic fatty acids, respectively.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting
proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.
 Kasumov et al. Page 2

Introduction
Ceramides play an important role in cell signaling, cell differentiation, proliferation and
NIH-PA Author Manuscript

apoptosis, and are generated by de novo synthesis from palmitoyl-CoA and serine, and serve
as a precursor for many sphingolipids [1]. As a secondary messenger in the sphingomyelin
transmembrane signaling pathway, ceramides are also formed by the neutral Mg2+-dependent
sphingomyelynase catalyzed hydrolysis of spingomyelin in cell membranes [2]. Several other
minor pathways [3], including hydrolysis of ceramide metabolites (galactosylceramide,
glycosylceramide and ceramide 1-phosphate) and ceramidase-catalyzed acylation of
sphingosine with distinct fatty acyl-CoAs also contribute to ceramide homeostasis [1].

Although it is not clear how the structure of individual ceramide species defines their
physiological functions, it has been shown that ceramides containing specific fatty acids are
generated in response to certain stimuli (tumor necrosis factor alfa or B-cell receptor induced
apoptosis), underscoring the structure/function relationship for different ceramide species [4].
It is known that C16 and C24 ceramide species are involved in cell death [5,6], while C18
ceramide inhibits cell growth [7].

Recent interest in ceramides has increased due to their newly identified role in insulin resistance
[8]. From in vitro studies, it is known that ceramides inhibit glucose uptake through inhibition
of Akt, a serine protein kinase mediator of insulin action [9]. In addition, ceramide content is
NIH-PA Author Manuscript

increased in muscle, liver and adipose tissue in insulin resistant animal models [10]. Elevated
total ceramide levels in human muscle [11] and adipose tissue has been linked to peripheral
insulin resistance [12]. Recently we demonstrated that plasma ceramides are increased in obese
subjects with type 2 diabetes and are associated with reduced insulin sensitivity [13]. However,
the definite intracellular origin and physiological significance of different species of plasma
ceramides are unknown. Since, intracellular partitioning of fatty acids and their metabolites
play a significant role in insulin resistance and lipotoxicity [14], it is important to identify the
role of specific ceramides in these processes.

Investigation of the physiological function of distinct ceramides requires an accurate, precise

and sensitive method for their quantification in plasma and tissue biopsy samples. Currently
ceramides are analyzed by an enzymatic diacylglycerol (DAG) kinase assay [15], by thin-layer
chromatography (TLC) detection [16], high performance liquid chromatography (HPLC)
[17,18], and gas chromatography mass spectrometry (GC-MS) analysis after derivatization
[19]. In some cases the TLC-isolated ceramides are hydrolyzed, and the released fatty acids
and sphingosine are analyzed after derivitization by gas-liquid chromatography [20], or by
HPLC [21]. However, these methods are cumbersome and time consuming. While some of
these approaches may be used to analyze total ceramide, they do not provide data on individual
NIH-PA Author Manuscript

ceramide species [15,21]. In addition, existing methods provide discrepant results for ceramide
levels in different biological samples [22–24]. Recent developments in electrospray ionization
tandem mass spectrometry (ESI-MS/MS) have helped resolve some of these limitations.
Several lipidomic studies have used ESI-MS/MS technologies to analyze ceramides and related
sphingolipids in biological samples [25,26]. However, in many instances, the described
methods attempt wide range lipidomic analyses, that make the validation and optimization of
measuring multiple individual anlytes difficult. Recently, an optimized and validated tandem
mass spectrometry analysis of a single C18 ceramide in cell culture was reported [27]. As part
of our research on lipid related mechanisms of insulin resistance, we have optimized and
validated a reverse-phase liquid chromatography coupled ESI-MS/MS technique for the
simultaneous measurement of multiple ceramide species in different biological matrices. We
have verified this technique for ceramide profiling and quantification in healthy human plasma
and different rat tissues. We also demonstrate that this technique detects changes in human
plasma ceramide levels after a clinical exercise intervention. This technique may also be

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 3

extended for quantification of other ceramides and related sphingolipids without significant
modification to the methodology.
NIH-PA Author Manuscript

Material and Methods

Materials
Pure standards of endogenous ceramide subspecies (C14, C16, C18, C18:1, C20, C24 and
C24:1) and non-naturally occurring internal standards (C17 and C25 ceramides) were obtained
from AVANTI Polar Lipids Inc. (Alabaster, AL, purity > 99%). HPLC grade solvents for ESI-
MS/MS and lipid extraction were purchased from Fluka (Milwaukee, MO). All other
chemicals, including Merck grade silica gel (9385, 230–400 mesh, 60 Å) were from Sigma-
Aldrich (St. Louis, MO).

Standard solutions
The stock solutions of ceramides were prepared at 1 mg/ml in chloroform. After dilution with
ethanol, a working solution of C14, C16, C18, C18:1, C20, C24 and C24:1 mixture was
prepared at 714 ng/ml for each ceramide species in one solution. This solution was divided
into 0.5 ml fractions and saved at (−80°C) until analysis. A mixture of C17 and C25 solution
in ethanol at concentrations of 1000 ng/ml and 2000 ng/ml, respectively, was prepared as an
internal standard solution. Fifty μl of internal standard solution corresponding to 50 ng of C17
NIH-PA Author Manuscript

and 100 ng of C25 was used for each analysis.

Analytical methods
Validation studies
Linearity: Calibration curves for measurement of ceramide concentrations were constructed
in a 50 μl plasma matrix with different concentrations of each endogenous ceramide (0–178
ng/ml for C14, 0–357 ng/ml for C16, C18, C18:1 and C20 and 0–714 ng/ml for C24 and C24:1)
in a mixed solution containing seven ceramide standards. These ranges cover all ceramide
species found in biological fluids and tissues. All samples were spiked with constant amounts
of C17 and C25 ceramide internal standards. Linear regression equations were derived from
calibration curves, and were used to calculate ceramide concentrations in plasma and tissue
samples. Analytical sensitivity was estimated from the calibration curve slopes. The limit of
quantification (LOQ) was defined as the lowest point in the calibration curve with a signal/
noise ratio equal to 10. The limit of detection was defined as a signal/noise ratio equal to 3.

Accuracy and Recovery

Accuracy and Recovery: Plasma Recovery: Three sets of samples were prepared for an
NIH-PA Author Manuscript

accuracy and recovery study of ceramide concentrations. The first set of samples included 100
μl of a solution of different ceramides in ethanol at four different amounts: 5.6, 22.3, 44.5 and
178 ng for C14, C16, C18, C18:1 and C20, and 15.6, 31.3, 62.5 and 250 ng for C24 and C24:1.
The second set of samples consisted of duplicates of 50 μl plasma samples. For the third set,
samples with different amounts (same as in set 1) of ceramides were each spiked with 50 μl of
plasma. All samples were spiked with C17 (50 ng) and C25 (100 ng) ceramides as internal
standards. All plasma samples were aliquoted from the same pooled plasma. After
quantification of individual ceramides in each sample, the recovery of the assay was calculated
as the ratio of analyte (ceramide) concentration in spiked plasma to the sum of the non-spiked
plasma and pure standards.

Tissue recovery: Recovery for the ceramide isolation from tissue was assessed using a
ceramide standard mixture with and without powdered tissue homogenate (rat liver and muscle)
in parallel with tissue samples, which were not spiked with ceramide standard mixture. All

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 4

samples were spiked with C17 and C25 internal standards. The recovery of ceramides from
tissues was calculated using the same procedure as for recovery of ceramides from plasma.
The precision, i.e., intra-assay and inter-assay reproducibility of the method was determined
NIH-PA Author Manuscript

by multiple analyses of a plasma sample from pooled plasma. Intra-assay variability was
determined by analyzing one sample five times. Inter-assay was established by processing the
same sample in five different preparations on different days over 2 weeks.

Human Study—We investigated ceramide levels in five healthy adults who participated in
our ongoing studies on the effect of exercise on insulin sensitivity. Subjects were obese but
had normal glucose tolerance. All individuals gave their written informed consent to partake
in the study, and all protocols were approved by the Institutional Review Board of the Cleveland
Clinic. Plasma for ceramide analysis was obtained prior to and after a 12-week clinical exercise
intervention. For the three days immediately prior to and during the last three days of the
intervention, all subjects resided in our Clinical Research Unit. They were placed on isocaloric
diets based on estimated basal metabolic rates to ensure weight stability prior to metabolic
testing. Exercise was supervised and consisted of aerobic training for 1 hr/day, 5 d/wk at 80%
of HRmax (~70% VO2max). Fasting blood samples (1 ml, in EDTA coated tubes) were obtained
from a dorsal hand vein before and after the intervention. Blood samples were centrifuged
immediately at 4°C and the separated plasma was stored at (−80°C) until analysis.

Sample Preparation for Human Plasma Ceramide Analysis: Plasma samples were
NIH-PA Author Manuscript

defrosted and 50 μl aliquots were transferred to ice-cold screw-capped glass tubes placed on
ice. Samples, in parallel with standard solutions, were spiked with 50 ng of C17 and 100 ng of
C25 ceramides and were extracted with 2 ml of a chloroform/methanol (1:2) mixture according
to the protocol of Bligh and Dyer [28]. Phases were broken by adding 0.5 ml chloroform and
0.5 ml water. The lower organic fraction was removed and the remainder was extracted with
an additional 1 ml of chloroform. The pooled organic phase was dried under nitrogen gas and
the residue was reconstituted in 500 μl of methylene chloride and loaded onto a silica gel
column packed with 2 ml of silica gel suspension in methylene chloride. Columns were washed
with 1 ml of methylene chloride and ceramides were eluted with 2×2 ml of 30% isopropanol
in methylene chloride. Eluent was dried under nitrogen gas and the residue was reconstituted
in acetonitrile/2-propanol (60:40, v/v) containing 0.2% formic acid and analyzed by mass
spectrometry.

Animal Study—All animal procedures were approved by the Institutional Animal Care and
Use Committee (IACUC) at the Cleveland Clinic and were performed in accordance with NIH
guidelines. Male Sprague-Dawley rats (200–300 g) were purchased from Charles River
Laboratories (Wilmington, MA), and were housed in our animal care facility with a 12:12 h
light-dark cycle. The animals had free access to food (Harlan Teklad NIH rat diet) and water.
NIH-PA Author Manuscript

After three days of acclimation the animals were euthanized with sodium pentobarbital (120
mg/kg). The liver, heart and skeletal muscle were harvested rapidly and the tissues were
continuously sprayed with ice-cold saline. Tissues were immediately freeze-clamped on
aluminum blocks precooled in liquid nitrogen and saved at −80 °C until analysis.

Sample Preparation for Rat Tissue Ceramide Analysis: Frozen tissue samples were
powdered under liquid nitrogen. For each analysis an aliquot of tissue powder (7–15 mg wet
weight) was suspended in ice-cold saline solution (500 μl, 1M NaCl) and homogenized with
a glass mortar and pestle. Calibration curves for ceramide standards using C17 and C25
ceramides as internal standards were prepared and processed in parallel with tissue samples.
Ceramides from tissue homogenates were extracted using a similar protocol as for plasma.
Briefly, 2 ml of an ice-cold chloroform: methanol (1:2, v/v) mixture was added to homogenized
tissue and vortexed at 4°C. The pooled organic phase was filtered using a glass wool filter

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 5

packed in a pasteur pipette to remove any solid particles. Collected eluent was dried and the
residue was reconstituted in HPLC elution buffer and analyzed by mass spectrometry. All
ceramide measurement experiments were normalized for wet tissue weight.
NIH-PA Author Manuscript

High Performance Liquid Chromatography-Mass Spectrometry Analysis: Ceramide

species were quantified by LC-ESI-MS/MS. For optimization, the mixture of ceramide
standards was infused directly into the mass spectrometer and all source parameters and
ionization conditions were adjusted to improve the sensitivity of the assay. Extracted samples
(25 μl) were injected into a Waters HPLC (2690 Separation Module, Waters Corp., Milford,
MA) and separated through an Xperchrom 100 C8 column (2.1 × 150 mm, 5 μm, P.J. Cobert
Associates, St. Louis, MO). The ceramides were resolved using a gradient starting from 50%
mobile phase A (water containing 0.2% formic acid) at a flow rate of 0.3 ml/min for 1 min, to
100% mobile phase B (acetonitrile/2-propanol (60:40, v/v) containing 0.2% formic acid) over
3 min at a linear gradient, and then with 100% B for 12 min. The column was then equilibrated
for 5 min with 50% mobile phase B. The HPLC column effluent was introduced onto a
Micromass triple quadrupole mass spectrometer (Quattro Ultima, Waters Inc., Beverly, MA)
and analyzed using electrospray ionization in positive mode. The configurations for mass
spectrometry were: capillary voltage, 3.0 kV; cone voltage, 40 V; source temperature, 120°C;
and desolvation temperature, 250°C. The flow rate of the nitrogen gas in the cone and
desolvation gas were 86 L/h and 700 L/h, respectively. Argon gas was used for collision-
induced dissociation having the collision cell pressure at 8.43 × 10−4 mbar. Analyses were
NIH-PA Author Manuscript

performed using electrospray ionization in the positive-ion mode with multiple reaction
monitoring (MRM) to select both parent and characteristic daughter ions specific to each
analyte simultaneously from a single injection. The MS/MS transitions (m/z) were 510→264
for C14, 538→264 for C16, 552→264 for C17, 564→264 for C18:1, 566→264 for C18,
594→264 for C20, 648→264 for C24:1, 650→264 for C24, and 664→264 for C25. The data
were acquired using MassLynx software (version 4.1, Manchester, UK).

Quantification: Ceramide subspecies were quantified using calibration curves and the ratios
of the integrated peak areas of ceramide subspecies and internal standards. C17 ceramide was
used as an internal standard for quantification of C14, C16, C18, C18:1, and C20 subspecies.
Concentrations of C24 and C24:1 were quantified using C25 as an internal standard. Total
measured ceramide was calculated from the sum of C14, C16, C18:1, C18, C20, C24:1 and
C24 ceramide subspecies.

Data are presented as mean ±SD. Paired t-tests were used to examine the effects of exercise
on plasma ceramide levels. Significance was accepted when P<0.05.

Results
NIH-PA Author Manuscript

Tandem mass spectrometry optimization

The full scan MS/MS spectra were recorded in a positive-ion mode during direct infusion of
ceramide standards. All of the investigated ceramide species give simple positive-ion full scan
spectra in acetonitrile/2-propanol (60:40, v/v) containing 0.2% formic acid, confirming that at
acidic pH in the ion spray ionization process, ions are mainly produced by protonation of neutral
molecules (M+H+) [29–31]. Precursor ions for each ceramide were separated and scanned in
the first quadrupole (Q1), fragmented in the second quadrupole (Q2) and the product ions were
scanned in the third quadrupole (Q3). Fig. 1 shows typical product ions spectra for some
specific ceramide species. Consistent with the published literature [30,32], we found that
collision-induced ionization of all studied ceramide species generates stable product ions with
m/z 282 and 264 corresponding to the loss of an amide-bound acyl-group, and one or two
molecules of water, respectively (Fig. 2). In addition, ceramides generate stable ions at a m/z

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 6

difference of 18, accounting for the loss of a water molecule from the precursor ion. Since m/
z 264 is a high intensity ion, we optimized the electrospray parameters and collision energy to
increase the intensity of each precursor ion in combination with m/z 264 in MRM mode for
NIH-PA Author Manuscript

sensitive analysis of all species.

Liquid chromatography
Our orientation experiment with HPLC-MS/MS analysis of plasma and tissue samples after
Bligh and Dyer lipid extraction without further purification revealed the existence of interfering
peaks on HPLC chromatography and low sensitivity in mass spectrometry detection for plasma
but not for tissue samples. Therefore we developed an easy-to-use silica chromatography
approach for the isolation of sphingolipids from other abundant plasma lipids. This resulted in
a significant improvement in the HPLC chromatography and yielded higher sensitivity in the
plasma sample analysis. As a result, the chromatographic method proved suitable for separation
of all analyzed ceramides within a single run on a C8 reverse-phase column using MS/MS
detection. After testing different acidic organic eluent systems (pure and mixture of acetonitrile,
methanol and isopropanol gradients with acidic water) we found that the mixture of acetonitrile/
2-propanol (60:40, v/v) containing 0.2% formic acid gradient with acidic water allows
separation of 9 ceramide species (7 physiological ceramides and 2 non-naturally occurring
internal standards) on a C8 reverse-phase column in less than 5 min during a 21 min run; this
includes column re-equilibrium (Fig. 3). Although the MRM mode with different transitions
allows specific identification of coeluting ceramides, the chromatographic separation allows
NIH-PA Author Manuscript

discrimination between 13C isotope contributions from lower to higher molecular masses
[31]. These results indicate that the chromatographic separation that we have described
improves the specificity of the assay.

Assay validation—Calibration curves for a mixture of ceramide concentrations were made

by adding increasing amounts of ceramides to constant amounts of C17 and C25 ceramide.
The curves were linear over a range of 2.8–178 ng for C14, 2.8–357 ng for C16, C18, C18:1
and C20 ceramides, and 5.6–714 ng for C24 and C24:1 ceramides. The limit of detection, limit
of quantification (LOQ) and linear regression parameters of calibration curves are presented
in Table 1. These ranges cover all concentrations of ceramides found in human plasma and
tissues.

Results from the accuracy study for the determination of C14, C16, C18:1, C18, C20, C24:1
and C24 concentrations are presented in Fig. 4. We plotted different concentrations of ceramide
species that had been determined in plasma against the amount of these ceramides added to
plasma. The resulting lines of identity were linear with slopes and regression coefficients close
to 1. The y intercepts correspond to the plasma content of each ceramide, which is in good
NIH-PA Author Manuscript

agreement with literature values [33]. Based on these analyses, the calculated percentage
recoveries of different ceramide species from plasma were between 78±7% and 91±11% (Table
2). Intra-assay and inter-assay coefficient of variation for the assay were 0.2–3.3% and 0.2–
7.3%, respectively (Table 3).

Likewise, as shown in Table 4 the recovery of ceramide species from rat liver and muscle tissue
was relatively good (70–90%). The ceramide content determined in non-spiked tissue in this
recovery study is in agreement with literature values [17,23], and is comparable with the liver
and muscle ceramide levels determined from 5 different rats (Table 5).

After optimization and validation, we applied this technique to the determination of ceramide
levels in human plasma and different rat tissues. Concentrations of C14, C16, C18:1, C18, C20,
C24:1 and C24 ceramide species in plasma of healthy human adults before and after a twelve
week clinical exercise intervention are presented in Table 6. The major ceramide species were
C16, C24 and C24:1, corresponding to ~37%, ~28%, and ~9% of total measured ceramides.

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 7

The medium-chain C14 ceramide had the lowest concentration of the seven measured ceramide
species (~5% of total measured ceramides). These values are similar to those found for plasma
ceramides reported in previous publications [33,34].
NIH-PA Author Manuscript

Table 5 represents the results of our tissue analysis of the same ceramide species from 5 rats
in comparison with literature values determined by different methods [17,23,35]. In agreement
with the literature, our results show that ceramides containing palmitic and stearic acids (C16
and C18 ceramides) are the predominant ceramide species in rat myocardium. In contrast to a
previous report [35], our results demonstrate that C20, C24 and C24:1 also have high
abundance in rat heart.

Additionally, we found that the abundance of ceramides with C16, C24 and C24:1 fatty acids
was also high in the liver, which is also in good agreement with literature values [23]. In contrast
to rat myocardium, C18 ceramide only accounted for ~2% (~1.9 % in this study, compared to
~2.3 % in literature [23]) of total measured liver ceramides in the rat. Analysis in rat muscle
tissue revealed high levels of long-chain and very-long-chain fatty acid containing ceramides.
The content of C18 ceramide accounted for approximately one quarter of the total ceramides
in muscle. With the exception of C14 ceramide (2.6±0.4 ng/g w.w.), all other measured
ceramides accounted for substantial fractions of the total ceramide content. Thus, rat heart,
liver, and muscle analysis, demonstrates that C16, C18, C20 and C24 are major saturated
ceramides in all tissues. From two unsaturated ceramides measured in this study C24:1 is
NIH-PA Author Manuscript

abundant in all three tissues, while C18:1 has higher abundance only in rat muscle.

Discussion
In this report we present an optimized and validated technique for the simultaneous
quantification of a wide range of ceramide species in biological samples. Utilization of two
non-physiological ceramides as internal standards (C17 and C25) allowed quantification of
ceramides containing long-chain (C14, C16, C18, C18:1 and C20) and very-long-chain (C24:1
and C24) fatty acids in a single run. Analysis of the concentration profile of multiple ceramide
species requires 50 μl of plasma or ~10 mg (wet weight) of tissue. Thus, the high sensitivity
of this technique makes ceramide analysis possible, not only in biological fluids, but also in
small size human tissue biopsy samples. The simultaneous measurement of ceramide species
in plasma and different tissue samples will advance the investigation of the physiological
significance of ceramides, and open the possibility that plasma ceramides can be used as
biomarkers of insulin resistance, and/or apoptosis.

The wide concentration range of different ceramides in plasma and tissue (from nM to μM
concentrations) requires a sensitive and versatile method for their analysis. Several HPLC
NIH-PA Author Manuscript

[18], HPLC-MS [24,29,36] and gas GC-MS [19] approaches have been described for
determining concentrations in biological samples. Although HPLC and GC-MS instruments
are easily available and widely used for measuring ceramide concentrations, all of these
techniques require derivatization prior to analysis. In contrast, HPLC-MS assays are more
sensitive and do not require derivatization. This simplifies the assay and helps to avoid
experimental errors generated at different levels of the analysis. Recent HPLC-MS based
lipidomic and sphingolipidomic initiatives offer comprehensive profiling of multiple lipid
compounds [25,26]. However, optimization of the absolute quantification of individual
analytes has not yet been established. Since different lipids have distinct physicochemical
properties, one of the major challenges in lipid analysis is the selection of an appropriate
internal standard for quantification of individual analytes. Liebisch and colleagues used non-
naturally occurring C8 ceramide as an internal standard for quantification of long-chain and
very-long-chain ceramides using calibration lines constructed for a few ceramide species (C16,
C18:1 and C24:1) [30]. These data were then used to calculate concentrations of the closest

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 8

related species. When these observations are combined with the results from our study (Table
1, column 5) it is evident that different ceramide species have different mass spectrometric ion
responses. This necessitates the construction of individual calibration curves for each species.
NIH-PA Author Manuscript

In addition, our initial analysis of pure ceramide standards revealed ion suppression at higher
concentrations; this was substantial for C24 and C24:1 ceramides. When we used a C17
ceramide as a sole internal standard the result was distortion of the calibration lines toward an
exponential curve for these specific species (data not presented). Therefore, we elected the C25
ceramide as the structurally-closest non-naturally occurring ceramide internal standard for the
quantification of C24 and C24:1.

Another improvement in the ceramide analysis was achieved by purification of plasma

ceramides on silica chromatography after total lipid extraction, prior to LC-MS/MS analysis.
This step allows elution of sphingolipids in one fraction after washout of the most abundant
and non-polar lipids, using methylene chloride. The result was good liquid chromatography
separation and sensitive analysis of ceramide species in plasma. However, the lower lipid
content of tissue samples allowed us to eliminate the silica column chromatography step.

The validation of this technique demonstrates that it is linear, accurate, precise, and free of any
appreciable plasma and tissue matrix effect. The use of MRM mode based tandem mass
spectrometry provides sufficient selectivity and sensitivity to the method. This easy-to-use
analysis of ceramides bodes well for its routine application in basic and clinical translational
NIH-PA Author Manuscript

research. The separation of several ceramide species by HPLC in a short single run (21 min
for one sample analysis cycle) will allow this technique to be applied to high-throughput
analysis.

After validation and optimization we applied this technique to the quantification of ceramide
species in human plasma and rat tissues. The extracellular ceramide concentrations could be
affected by Zn2+ dependent acidic spingomyelinase in blood and tissues [37]. If not stabilized,
sphingomyelinase may generate ceramide from sphingomyelin in blood and tissue samples
after collection. Therefore, in order to circumvent the post-collection ceramide metabolism,
blood samples were collected in tubes containing ion chelating EDTA (in order to deactivate
Zn2+ dependent acidic spingomyelinase), centrifuged immediately, and the plasma was stored
at −80°C until analysis. It has been shown that fluorinated anesthetics (isoflurane and
desflurane), but not pentobarbital stimulate renal cortical ceramide expression through the
sphingomyelinase cascade [38]. Therefore, we used pentobarbital to anesthetize the rats. The
high sensitivity of the method allowed us to assess the concentration profiles of multiple
ceramide species in small sample volumes. The technique was verified by comparing our
results for the different ceramide levels with values reported in the literature using different
techniques. We also tested this assay by measuring changes in plasma ceramide levels in the
NIH-PA Author Manuscript

same subjects before and after a clinical exercise intervention. The method enabled us to detect
significant reductions in C14, C18:1, and C24 ceramide levels. Further, recent data on carbon
tetrachloride- and lipolysaccharide-treated rats indirectly suggest that plasma ceramides
originate from necrotized liver [23,24,39]. Establishing a direct link between plasma and
hepatic ceramides in humans with liver disease may allow the use of plasma ceramides as
biomarkers of hepatic dysfunction and cell death.

In conclusion, we have developed and validated an HPLC coupled tandem mass spectrometry
analysis for the quantification of ceramide species in plasma and tissue samples. This technique
could be used to study the regulation of ceramide homeostasis and its relation to diseases
associated with altered lipid metabolism.

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 9

Acknowledgments
We thank Thomas Solomon for helpful discussions during preparation of the manuscript, and the CRU staff for their
NIH-PA Author Manuscript

help with the human studies. This work was supported in part by National Institutes of Health grants RO1 AG12834,
CTSA UL1-RR024989, and by the Cleveland Clinic Foundation.

Abbreviations used
ESI-MS/MS electrospray ionization tandem mass spectrometry
MRM multiple reaction monitoring
HPLC high performance liquid chromatography
GC-MS gas chromatography mass spectrometry
LOQ limit of quantification

References
1. Huwiler A, Kolter T, Pfeilschifter J, Sandhoff K. Physiology and pathophysiology of sphingolipid
metabolism and signaling. Biochim Biophys Acta 2000;1485:63–99. [PubMed: 10832090]
2. Clarke CJ, Snook CF, Tani M, Matmati N, Marchesini N, Hannun YA. The extended family of neutral
sphingomyelinases. Biochemistry 2006;45:11247–11256. [PubMed: 16981685]
NIH-PA Author Manuscript

3. Ohanian J, Ohanian V. Sphingolipids in mammalian cell signalling. Cell Mol Life Sci 2001;58:2053–
2068. [PubMed: 11814056]
4. Pewzner-Jung Y, Ben-Dor S, Futerman AH. When do Lasses (longevity assurance genes) become
CerS (ceramide synthases)?: Insights into the regulation of ceramide synthesis. J Biol Chem
2006;281:25001–25005. [PubMed: 16793762]
5. Osawa Y, Uchinami H, Bielawski J, Schwabe RF, Hannun YA, Brenner DA. Roles for C16-ceramide
and sphingosine 1-phosphate in regulating hepatocyte apoptosis in response to tumor necrosis factor-
alpha. J Biol Chem 2005;280:27879–27887. [PubMed: 15946935]
6. Seumois G, Fillet M, Gillet L, Faccinetto C, Desmet C, Franđcois C, Dewals B, Oury C,
Vanderplasschen A, Lekeux P, Bureau F. De novo C16- and C24-ceramide generation contributes to
spontaneous neutrophil apoptosis. J Leukocyte Biol 2007;81:1477–1486. [PubMed: 17329567]
7. Koybasi S, Senkal CE, Sundararaj K, Spassieva S, Bielawski J, Osta W, Day TA, Jiang JC, Jazwinski
SM, Hannun YA, Obeid LM, Ogretmen B. Defects in cell growth regulation by C18:0-ceramide and
longevity assurance gene 1 in human head and neck squamous cell carcinomas. J Biol Chem
2004;279:44311–44319. [PubMed: 15317812]
8. Holland WL, Summers SA. Sphingolipids, insulin resistance, and metabolic disease: new insights from
in vivo manipulation of sphingolipid metabolism. Endocrinol Rev 2008;29:381–402.
9. Pickersgill L, Litherland GJ, Greenberg AS, Walker M, Yeaman SJ. Key role for ceramides in
NIH-PA Author Manuscript

mediating insulin resistance in human muscle cells. J Biol Chem 2007;282:12583–12589. [PubMed:
17337731]
10. Samad F, Hester KD, Yang G, Hannun YA, Bielawski J. Altered adipose and plasma sphingolipid
metabolism in obesity: a potential mechanism for cardiovascular and metabolic risk. Diabetes
2006;55:2579–2587. [PubMed: 16936207]
11. Adams JM 2nd, Pratipanawatr T, Berria R, Wang E, DeFronzo RA, Sullards MC, Mandarino LJ.
Ceramide content is increased in skeletal muscle from obese insulin-resistant humans. Diabetes
2004;53:25–31. [PubMed: 14693694]
12. Kolak M, Westerbacka J, Velagapudi VR, Weagsèater D, Yetukuri L, Makkonen J, Rissanen A,
Heakkinen AM, Lindell M, Bergholm R, Hamsten A, Eriksson P, Fisher RM, Oresic M, Yki-
Jearvinen H. Adipose tissue inflammation and increased ceramide content characterize subjects with
high liver fat content independent of obesity. Diabetes 2007;56:1960–1968. [PubMed: 17620421]

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 10

13. Haus JM, Kashyap SR, Kasumov T, Zhang R, Kelly KR, Defronzo RA, Kirwan JP. Plasma ceramides
are elevated in obese subjects with type 2 diabetes and correlate with the severity of insulin resistance.
Diabetes 2009;58:337–343. [PubMed: 19008343]
NIH-PA Author Manuscript

14. Alkhouri N, Dixon LJ, Feldstein AE. Lipotoxicity in nonalcoholic fatty liver disease: not all lipids
are created equal. Exp Rev Gastroenterol Hepatology 2009;3:445–451.
15. Preiss JE, Loomis CR, Bell RM, Niedel JE. Quantitative measurement of sn-1,2-diacylglycerols.
Methods Enzymol 1987;141:294–300. [PubMed: 3037247]
16. Gorska M, Dobrzyn A, Zendzian-Piotrowska M, Namiot Z. Concentration and composition of free
ceramides in human plasma. Horm Metab Res 2002;34:466–468. [PubMed: 12198604]
17. Dobrzyn A, Gorski J. Ceramides and sphingomyelins in skeletal muscles of the rat: content and
composition. Effect of prolonged exercise. Am J Physiol Endocrinol Metab 2002;282:E277–285.
[PubMed: 11788358]
18. Yano M, Kishida E, Muneyuki Y, Masuzawa Y. Quantitative analysis of ceramide molecular species
by high performance liquid chromatography. J Lipid Res 1998;39:2091–2098. [PubMed: 9788256]
19. Tserng KY, Griffin R. Quantitation and molecular species determination of diacylglycerols,
phosphatidylcholines, ceramides, and sphingomyelins with gas chromatography. Anal Biochem
2003;323:84–93. [PubMed: 14622962]
20. Dobrzyn A, Knapp M, Gorski J. Effect of acute exercise and training on metabolism of ceramide in
the heart muscle of the rat. Acta Phys Scand 2004;181:313–319.
21. Blachnio-Zabielska A, Baranowski M, Zabielski P, Gorski J. Effect of exercise duration on the key
pathways of ceramide metabolism in rat skeletal muscles. J Cell Biochem 2008;105:776–784.
NIH-PA Author Manuscript

[PubMed: 18680146]
22. Turinsky J, O’Sullivan DM, Bayly BP. 1,2-Diacylglycerol and ceramide levels in insulin-resistant
tissues of the rat in vivo. J Biol Chem 1990;265:16880–16885. [PubMed: 2211599]
23. Yamaguchi M, Miyashita Y, Kumagai Y, Kojo S. Change in liver and plasma ceramides during D-
galactosamine-induced acute hepatic injury by LC-MS/MS. Bioorg Med Chem Lett 2004;14:4061–
4064. [PubMed: 15225726]
24. Ichi I, Nakahara K, Fujii K, Iida C, Miyashita Y, Kojo S. Increase of ceramide in the liver and plasma
after carbon tetrachloride intoxication in the rat. J Nutr Sci Vitamin 2007;53:53–56.
25. Schmelzer K, Fahy E, Subramaniam S, Dennis EA. The lipid maps initiative in lipidomics. Methods
Enzymol 2007;432:171–183. [PubMed: 17954217]
26. Sullards MC, Allegood JC, Kelly S, Wang E, Haynes CA, Park H, Chen Y, Merrill AH Jr. Structure-
specific, quantitative methods for analysis of sphingolipids by liquid chromatography-tandem mass
spectrometry: “inside-out” sphingolipidomics. Methods Enzymol 2007;432:83–115. [PubMed:
17954214]
27. Haynes TA, Duerksen-Hughes PJ, Filippova M, Filippov V, Zhang K. C18 ceramide analysis in
mammalian cells employing reversed-phase high-performance liquid chromatography tandem mass
spectrometry. Anal Biochem 2008;378:80–86. [PubMed: 18423390]
28. Bligh EA, Dyer WJ. A rapid and simple method for the determination of esterified fatty acids and for
NIH-PA Author Manuscript

total fatty acids in blood. Can J Biochem Physiol 1959;37:911–917. [PubMed: 13671378]
29. Gu M, Kerwin JL, Watts JD, Aebersold R. Ceramide profiling of complex lipid mixtures by
electrospray ionization mass spectrometry. Anal Biochem 1997;244:347–356. [PubMed: 9025952]
30. Liebisch G, Drobnik W, Reil M, Trèumbach B, Arnecke R, Olgemèoller B, Roscher A, Schmitz G.
Quantitative measurement of different ceramide species from crude cellular extracts by electrospray
ionization tandem mass spectrometry (ESI-MS/MS). J Lipid Res 1999;40:1539–1546. [PubMed:
10428992]
31. Bielawski J, Szulc ZM, Hannun YA, Bielawska A. Simultaneous quantitative analysis of bioactive
sphingolipids by high-performance liquid chromatography-tandem mass spectrometry. Methods
2006;39:82–91. [PubMed: 16828308]
32. Merrill AH Jr, Sullards MC, Allegood JC, Kelly S, Wang E. Sphingolipidomics: high-throughput,
structure-specific, and quantitative analysis of sphingolipids by liquid chromatography tandem mass
spectrometry. Methods 2005;36:207–224. [PubMed: 15894491]

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 11

33. Ichi I, Nakahara K, Miyashita Y, Hidaka A, Kutsukake S, Inoue K, Maruyama T, Miwa Y, Harada-
Shiba M, Tsushima M, Kojo S. Association of ceramides in human plasma with risk factors of
atherosclerosis. Lipids 2006;41:859–863. [PubMed: 17152923]
NIH-PA Author Manuscript

34. Groener JE, Poorthuis BJ, Kuiper S, Hollak CE, Aerts JM. Plasma glucosylceramide and ceramide
in type 1 Gaucher disease patients: correlations with disease severity and response to therapeutic
intervention. Biochim Biophys Acta 2008;1781:72–78. [PubMed: 18155675]
35. Baranowski M, Blachnio A, Zabielski P, Gorski J. Pioglitazone induces de novo ceramide synthesis
in the rat heart. Prostagl Lipid Mediat 2007;83:99–111.
36. Mano N, Oda Y, Yamada K, Asakawa N, Katayama K. Simultaneous quantitative determination
method for sphingolipid metabolites by liquid chromatography/ionspray ionization tandem mass
spectrometry. Anal Biochem 1997;244:291–300. [PubMed: 9025946]
37. Tani M, Ito M, Igarashi Y. Ceramide/sphingosine/sphingosine 1-phosphate metabolism on the cell
surface and in the extracellular space. Cell Signal 2007;19:229–237. [PubMed: 16963225]
38. Lochhead KM, Zager RA. Fluorinated anesthetic exposure “activates” the renal cortical
sphingomyelinase cascade. Kidney Int 1998;54:373–381. [PubMed: 9690203]
39. Lightle S, Tosheva R, Lee A, Queen-Baker J, Boyanovsky B, Shedlofsky S, Nikolova-Karakashian
M. Elevation of ceramide in serum lipoproteins during acute phase response in humans and mice:
role of serine-palmitoyl transferase. Arch Biochem Biophys 2003;419:120–128. [PubMed:
14592455]
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 12
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Fig. 1.
Product ions spectra of selected ceramide subspecies in positive ion detection mode.
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 13
NIH-PA Author Manuscript

Fig. 2.
Structure and fragmentation pattern of ceramide; m/z 264 is the most abundant product ion
during collision induced ionization of ceramides.
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 14
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Fig. 3.
Typical chromatograms of standard mixture, plasma, and muscle ceramides, using multiple
reaction monitoring mode.
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 Kasumov et al. Page 15
NIH-PA Author Manuscript

Fig. 4.
Accuracy study for C14, C16, C18, C18:1, C20, C24 and C24:1 ceramides. Known amounts
of ceramide species were added to pooled plasma samples (50 μl of each) and the amounts of
each ceramide were subsequently quantified. The plot shows ng of ceramide (endogenous plus
added) versus ng of ceramide added. The quantification procedure is described in the text.
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 1
Linearity, analytical sensitivity and detection limit

Linear Regression Parameters

Ceramide Limit of Detection (ng/ml) Limit of Quantification (ng/ml) Internal Standard
Slope r2
Kasumov et al.

C14 0.005 0.01 C17 0.0239 0.999

C16 0.02 0.05 C17 0.0033 0.998
C18 0.01 0.02 C17 0.0102 0.998
C18:1 0.01 0.02 C17 0.0101 0.996
C20 0.005 0.01 C17 0.0070 0.998
C24 0.05 0.1 C25 0.0113 0.996
C24:1 0.2 0.5 C25 0.0162 0.991

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

Page 16
 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 2
Ceramide recovery from plasma. Data from accuracy study presented in Fig. 4 were used for calculation of recovery of ceramide species. The result represents
the average of recoveries from plasma at 5 different spiked ceramide concentrations. The calculation procedure is described in the Method’s section.

Ceramide C14 C16 C18 C18:1 C20 C24:1 C24

Kasumov et al.

Recovery (%) 78±7 91±11 83±7 91±10 85±8 79±7 84±5

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

Page 17
 Kasumov et al. Page 18

Table 3
Intra- and interassay variability for ceramide assay in rat liver samples
NIH-PA Author Manuscript

CV (%)
Ceramide
Inter-assay (n = 5) Intra-assay (n = 5)

C14 0.2 0.2

C16 3.7 2.9
C18 0.9 2.0
C18:1 0.2 0.1
C20 1.4 2.3
C24 3.7 3.3
C24:1 7.3 1.2
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 4
Ceramide recovery from rat tissues

Tissue Type Ceramide Tissue (ng/15 mg [Link].) Spike (ng) Detected Total ng Recovery (%)

Skeletal Muscle C14 4.5 260.5 217.4 82.1

Kasumov et al.

C16 41.2 287.0 233.5 71.1

C18 51.3 627.1 598.7 88.3
C18:1 4.4 245.7 204.0 81.6
C20 9.3 429.6 426.0 97.1
C24 26.1 133.6 150.9 94.5
C24:1 34.8 155.4 166.2 87.4

Liver Tissue C14 1.9 260.5 213.4 81.3

C16 287.0 173.5 394.8 85.7
C18 42.2 627.1 574.6 85.9
C18:1 2.6 245.7 174.1 70.1
C20 5.9 429.6 294.8 67.7
C24 769.3 133.6 786.9 87.2
C24:1 353.1 155.4 504.8 99.3

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

Page 19
 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 5
Ceramide content (nmol/g wet weight) in different rat tissues (n=5).

Ceramide Heart Liver Muscle (soleus)

Kasumov et al.

Present study Baranowski M. et al [35] Present study Yamaguchi M. et al. [23] Present study Dobrzyn A. et al. [17]

C14 4.4±0.1 7.9±1.3 8.9±3.5 - 2.6±0.4 5.7±0.7

C16 33.0±0.9 24.5±5.4 28.7±4.3 34.1±9.2 33.3±8.2 44.9±13.8
C18 19.8±3.8 25.1±5.5 5.5±0.4 7.0±1.9 42.5±3.0 54.6±15.6
C18:1 4.7±0.3 5.5±1.3 2.3±0.7 3.8±0.6 13.1±0.6 24.3±7.2
C20 21.0±3.4 5.2±0.8 4.1±0.9 - 17.2±1.4 -
C24 40.0±2.9 12.3±0.8 119.7±4.7 137.4±30.5 37.1±4.2 -
C24:1 28.0±3.5 - 105.2±13.3 69.8±22.2 29.5±5.0 5.3±1.1

Total 150.7±24.2 101±14.9 314.0±61.7 302.6±51.0 176.1±13.3 161.2±24.8

Anal Biochem. Author manuscript; available in PMC 2011 June 1.

Page 20
 Kasumov et al. Page 21

Table 6
Plasma ceramide levels in healthy adults before and after a 12-week clinical exercise intervention.
NIH-PA Author Manuscript

Ceramide Pre-exercise Post-exercise P-value

C14 0.26±0.06 0.20±0.09 0.03

C16 1.97±0.73 1.68±0.57 NS
C18 0.73±0.29 0.75±0.12 NS
C18:1 0.06±0.02 0.05±0.02 0.05
C20 0.35±0.07 0.33±0.06 NS
C24 1.48±0.49 0.99±0.25 0.01
C24:1 0.51±0.17 0.39±0.06 NS

Total 5.37±0.55 4.40±0.68 0.01

Data are presented as mean ± SE. All five subjects completed supervised aerobic exercise training: one hour per day, 5 d/wk, at 80% of HRmax (~70%
VO2max). Ceramide concentration are expressed in μmol/L.
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Anal Biochem. Author manuscript; available in PMC 2011 June 1.