Production of Rare Phyto-Ceramides from Abundant Food Plant Residues

Article abstract of DOI:10.1021/acs.jafc.6b04275

Ceramides (Cers) are major components of the outermost layer of the skin, the stratum corneum, and play a crucial role in permeability barrier functions. Alterations in Cer composition causing skin diseases are compensated with semisynthetic skin-identical Cers. Plants constitute new resources for Cer production as they contain glucosylceramides (GluCers) as major components. GluCers were purified from industrial waste plant materials, apple pomace (Malus domestica), wheat germs (Triticum sp.), and coffee grounds (Coffea sp.), with GluCer contents of 28.9 mg, 33.7 mg, and 4.4 mg per 100 g of plant material. Forty-five species of GluCers (1-45) were identified with different sphingoid bases, saturated or monounsaturated α-hydroxy fatty acids (C15-28), and β-glucose as polar headgroup. Three main GluCers were hydrolyzed by a recombinant human glucocerebrosidase to produce phyto-Cers (46-48). These studies showed that rare and expensive phyto-Cers can be obtained from industrial food plant residues.

Full text of DOI:10.1021/acs.jafc.6b04275

Article
pubs.acs.org/JAFC
Production of Rare Phyto-Ceramides from Abundant Food Plant
Residues
Mathias Reisberg,† Norbert Arnold,‡ Andrea Porzel,‡ Reinhard H. H. Neubert,§ and Birgit Drager*,†
̈
†Department of Pharmaceutical Biology and Pharmacology, Institute of Pharmacy, Faculty of Natural Sciences I, Martin Luther
University Halle-Wittenberg, Hoher Weg 8, D-06120 Halle (Saale), Germany
‡Department of Bioorganic Chemistry, Leibniz Institute of Plant Biochemistry, Weinberg 3, D-06120 Halle (Saale), Germany
§Department of Pharmaceutical Technology and Biopharmaceutics, Institute of Pharmacy, Faculty of Natural Sciences I, Martin
Luther University Halle-Wittenberg, Wolfgang-Langenbeck-Str. 4, D-06120 Halle (Saale), Germany
S
* Supporting Information
ABSTRACT: Ceramides (Cers) are major components of the outermost layer of the skin, the stratum corneum, and play a
crucial role in permeability barrier functions. Alterations in Cer composition causing skin diseases are compensated with
semisynthetic skin-identical Cers. Plants constitute new resources for Cer production as they contain glucosylceramides
(GluCers) as major components. GluCers were purified from industrial waste plant materials, apple pomace (Malus domestica),
wheat germs (Triticum sp.), and coffee grounds (Coffea sp.), with GluCer contents of 28.9 mg, 33.7 mg, and 4.4 mg per 100 g of
plant material. Forty-five species of GluCers (1−45) were identified with different sphingoid bases, saturated or
monounsaturated α-hydroxy fatty acids (C15−28), and β-glucose as polar headgroup. Three main GluCers were hydrolyzed
by a recombinant human glucocerebrosidase to produce phyto-Cers (46−48). These studies showed that rare and expensive
phyto-Cers can be obtained from industrial food plant residues.
KEYWORDS: ceramide, glucosylceramide, glucocerebrosidase, apple pomace, wheat germs, coffee grounds, Malus domestica,
Triticum sp., Coffea sp.
psoriasis.4 The substitution of skin-identical Cers in the form of
semisolid dosage forms is expected to improve this balance.
INTRODUCTION
Ceramides (Cers) and glucosylceramides (GluCers) belong to
the class of sphingolipids and consist of a sphingoid base (long-
chain base; LCB) and an amide-linked fatty acid (Figure 1).
■
Various pharmaceuticals and cosmetics are on the market that
contain Cers. It is suggested that phytosphingosine-based Cers
have good stabilizing and water-binding properties because of
up to four hydroxyl groups in the head core group.7−10
Cers are commercially available, but production is semi-
synthetic and very expensive. The extraction from animal
tissues, such as bovine brain, porcine skin, or hen eggs, is
possible but comprises certain risks and constraints. Nowadays,
the sphingoid bases are often isolated from fermentation broth
of yeasts,11 followed by an acylation step to provide Cer
structures similar to those found in human skin. Therefore,
there is a need for cheaper alternatives in production of Cers. In
plants, free Cers occur only as a minor fraction of
sphingolipids.12 They are partial constituents of GluCer and
(glycosyl-) phosphoryl inositol ceramides that represent the
major sphingolipid classes in plant plasma membranes and
tonoplasts.12 Cers as parts of sphingolipids occur ubiquitously
in the plant kingdom; they influence membrane integrity and
permeability.13
Figure 1. General structure of typical plant GluCers, e.g., Glc-d18:2
h16:0 (6) and Cers and major ESI-MS/MS fragmentation patterns.
Modifications within the sphingoid base and fatty acid (chain lengths,
hydroxylation, and/or desaturation) occur in plant Cers. Positive (Y0,
Z0, O) and negative (Y0, Z0, U) ion fragments are shown according to
nomenclature from Ann and Adams.31
Cers play a crucial role in the skin by providing a major lipid
class (approximately 50%) of the outermost layer of the skin,
the stratum corneum.1 Up to now 18 different Cer classes in
human skin were already reported.2,3 They possess unique
functions in membrane stabilization and in barrier formation
against water loss.4 Psoriasis5 and atopic dermatitis6 coincide
with decreased Cer levels and alterations in Cer composition.
Evidence for decreased Cers as cause for skin diseases arises
from experimental mice with impaired Cer formation showing
Therefore, plants may provide an economical alternative
source for natural GluCers. Contents of 1−100 mg/100 g dry
mass were reported from different plant materials.14−16
Received: September 25, 2016
Revised: December 30, 2016
Accepted: January 24, 2017
Published: January 24, 2017
© 2017 American Chemical Society
1507
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
linoleic acid (≥98%, AppliChem, Darmstadt, Germany), β-sitosterol
(≥80%, Roth, Karlsruhe, Germany), β-sitosterol glucoside (≥98%,
available from in-house library), squalene (≥97%, Fluka, Neu-Ulm,
Germany), and triolein (glycerol trioleate, ≥99%, Sigma, St. Louis,
MO).
TLC Analysis. TLC analysis war carried out using chloroform:me-
thanol 85:15 (v/v) as the solvent system.25 For visualization,
developed TLC plates were sprayed with copper(II) sulfate (10%),
phosphoric acid (8%), and methanol (5%) in water and heated at 150
°C for 10 min. TLC reference mix was composed of 100 μg/mL of
each squalene, triolein, cholesteryl oleate, linoleic acid, β-sitosterol, β-
sitosterol glucoside, and GluCer Glc-d18:2 h16:0, in chloroform−
methanol 2:1 (v/v). Samples (10 μL) were applied as 6 mm bands
using Linomat IV (Camag, Muttenz, Switzerland). The distance from
the lower edge was 8 mm, that from the left side was 12 mm, and the
space between bands was 4 mm.
Specifically, some common edible plants, e.g., soy and rice, were
identified as rich sources.14,17 Huge amounts of industrial plant
waste accumulate every day. The value chain often ends as
animal feed, e.g., for apple pomace and wheat germs, or in
household refuse, e.g., for coffee grounds. Approximately 2.7
million tons of apple pomace (as 25% leftover18 of apple juice
production), 14.7 million tons of wheat germs (as 2% of the
whole grain after milling), and 9.2 million tons of coffee were
produced for consumption in 2015 worldwide (FAS, 2016).
GluCers were already reported as constituents in apple14,19 and
wheat.20,21 We aim to investigate whether these industrial food
byproducts contain sufficient amounts of GluCers to serve as
sources for phyto-Cer production.
The nomenclature proposed by Karlsson22 for individual
sphingolipids is in accordance with IUPAC description of lipids.
A typical plant GluCer is Glc-d18:2Δ4E,8E/Z h16:0, where “d”
stands for the dihydroxy sphingoid base (“t” for a trihydroxy
base). The carbon chain length and the degree of desaturations
Extraction and Purification. Dried plant material (A, 750.0 g; W,
300.0 g; C, 300.0 g) was extracted by isopropanol−n-hexane−water
(55:20:25 (v/v/v);12 A, 3.0 L; W and C, 1.2 L) three times using an
ultrasonic bath for 15 min. The extracts were combined and
evaporated to dryness (A, 99.2 g; W, 110.4 g; C, 34.4 g). This total
lipid extract was separated into polar and nonpolar compounds by
liquid−liquid partitioning with chloroform−methanol−water (1:1:1,
v/v/v; A and W, 0.6 L; C, 0.3 L). Additional 0.5−1.0% (w/v water
phase) sodium chloride improved separation. The resulting two phases
were additionally partitioned using a mixture of chloroform:methanol
1:1 (v/v) for the aqueous phase and a mixture of methanol:water 1:1
(v/v) for the organic phase (A and W, 0.2 L; C, 0.1 L). The organic
phases were combined and evaporated to dryness (A, 30.42 g; W,
30.37 g; C, 30.44 g). For comparison purposes, 20.0 g of each extract
were applied to CC on silica gel 60 (250.0 g, 3.8 × 60 cm).
Chloroform−methanol (100:0 (v/v), 1 L; 90:10 (v/v), 1 L; 80:20 (v/
v), 1 L, flow 1.5−2.5 mL/min) was used for gradient elution. Fractions
of 10 or 20 mL were collected with a Cygnet fraction collector
(Teledyne Isco, Lincoln, NE). After TLC analysis by comparison with
the reference mix, fractions of each plant source containing GluCers
and β-sitosterol glucoside (A, 0.666 g; W, 0.278 g; C, 0.279 g) were
further separated by size exclusion chromatography on sephadex LH-
20 (100.0 g, 1.5 × 60 cm; solvent, dichloromethane−methanol 1:1 (v/
v), flow 0.12 mL/min). Fractions of 2.5 mL were collected after TLC
and GluCer enriched fractions were obtained (A, 0.222 g; W, 0.084 g;
C, 0.050 g). These fractions were further separated by preparative
HPLC-MS (flow 25 mL/min, column temperature 25 °C) using a
Zorbax Extend-C18 column (5 μm, 21.2 × 150 mm, 80 Å, Agilent,
Santa Clara, CA). The split ratio of fraction collector to mass
spectrometer was set to 500:1. The injection volume was 300−900 μL,
and fractions were pooled according to their retention time and m/z
value. Finally, GluCer enriched fractions were obtained from apple
pomace (fraction A1 (10.2 mg): 1−3; A2 (10.1 mg): 4-5; A3 (55.6
mg): 6; A4 (1.1 mg): 9; A5 (1.1 mg): 12; A6 (2.0 mg): 16−17; A7
(3.1 mg): 21−22, 24; A8 (19.5 mg): 27; A9 (10.7 mg): 30−31; A10
(24.8 mg): 34−35; A11 (3.4 mg): 38, 40; A12 (0.9 mg): 42−43),
wheat germs (fraction W1 (6.6 mg): 6−7; W2 (4.0 mg): 8; W3 (3.7
mg): 10−11; W4 (3.8 mg): 14−16, 19−20; W5 (16.7 mg): 20−21,
23; W6 (2.5 mg): 25; W7 (13.8 mg): 26−28; W8 (6.3 mg): 29−32;
W9 (5.1 mg): 33−37; W10 (2.3 mg): 38−40; W11 (1.0 mg): 41−44;
W12 (0.8 mg): 45) and coffee grounds (fraction C1 (1.0 mg): 6; C2
(3.0 mg): 27; C3 (1.8 mg): 30−31; C4 (2.0 mg): 35; C5 (0.8 mg):
38, 40).
of the sphingoid and acyl chains are given as numbers (18:2).
Δ8E/Z
Double bond positions and configuration are given as
.
The α-hydroxylation within the fatty acid is declared by “h”,
and the sugar moiety is denoted as “Glc” for glucose.
Differences in the chemical structures of human and plant
Cers comprise prevalence of sphinganine (d18:0), sphing-4-
enine (sphingosine, d18:1Δ4), and phytosphingosine (t18:0) in
human skin, whereas plants appear to contain sphing-8-enine
(sphingosine, d18:1Δ8), sphinga-4,8-dienine (d18:2Δ4,8), and
phytosphing-8-enine (t18:1Δ8) as major components.23 Plant
sphingoid bases exhibit typical chain lengths of C18, and fatty
acid chains range from C14−26, of which 90−95% are α-
hydroxylated.23,24 However, detailed structure differences still
need a comprehensive comparison.
Purification of phyto-Cers is necessary for the investigation
of functional equivalence. The industrial purification of
compounds out of complex matrices requires exhaustive,
efficient, and economical methods. For that purpose, an
efficient extraction process, isolation (liquid−liquid extraction,
column chromatography (CC), pHPLC-MS), and structure
elucidation (TLC, HPLC-MS, MS/MS, HR-MS, NMR) for
phyto-Cers are described here.
MATERIALS AND METHODS
■
Materials. Apple pomace (A, Malus domestica, apple cultivars) was
kindly provided by Becker Eislebener Fruchtsaft (Eisleben, Germany).
Wheat germs (W, Triticum sp., wheat cultivars) were obtained from
Dr. Grandel (Augsburg, Germany) and ground coffee beans (C, Coffea
sp., coffee varieties) from Jacobs Douwe Egberts (Bremen, Germany).
Coffee grounds were obtained after percolation. Analytical TLC was
carried out on precoated silica gel F254 aluminum plates from Merck
(Darmstadt, Germany). For adsorption chromatography, silica gel 60
(0.063−0.200 mm, AppliChem, Darmstadt, Germany) was used. Size
exclusion column chromatography was performed using sephadex LH-
20 (GE Healthcare, Solingen, Germany).
Methanol used for preparative HPLC was HPLC-grade (99.9%,
Roth, Karlsruhe, Germany) and for mass spectrometry analyses ULC/
MS-grade (99.98%, Biosolve BV, Valkenswaard, Netherlands). Water
was demineralized (σ = 0.055 μS/cm, TKA GenPure, Niederelbert,
Germany). Pyridine-d5 (99.5%) was purchased from Deutero
(Kastellaun, Germany). Cerezyme containing the recombinant
DNA-produced analogue of human β-glucocerebrosidase imglucerase
(EC 3.2.1.45) originated from Genzyme (Cambridge, MA). Sodium
taurocholate hydrate (≥97%) was bought from Sigma (St. Louis,
MO). A GluCer standard (from soy beans, Glc-d18:2 h16:0, ≥98%)
was obtained from Avanti Polar Lipids (Alabaster, AL). TLC reference
compounds were cholesteryl oleate (≥98%, Sigma, St. Louis, MO),
Analytical Enzymatic Assay for Cers. A total of 10 μg from each
GluCer fraction (A1−12, W1−12, C1−5) were hydrolyzed by
imiglucerase. The enzyme was suspended in 50 mM sodium citrate
buffer (pH 4.5, final volume 1 mL), and 413.4 μL of a 5 mg/mL
sodium taurocholate solution, 2 mM dithiothreitol, and 1 mM ascorbic
acid were added as protection against enzyme oxidation. A total of 5
μg of imiglucerase was added, and the reaction mixture was incubated
at 37 °C for 14 h. The reaction was quenched with 0.4 mL of
chloroform−methanol 2:1 (v/v), and the water phases were extracted
two more times (0.4 mL of chloroform−methanol 2:1 (v/v)). The
organic phases were combined and evaporated to dryness. Resulting
1508
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
Table 1. Stepwise Concentration of GluCers
a
initial mass
[g]
total lipid extract
[g]
organic phase
[g]
silica gel fraction
[mg]
sephadex fraction
[mg]
prep. HPLC fraction
[mg]
GluCers/100g
[mg]
plant material
apple pomace
GluCers w/w
wheat germs
GluCers w/w
coffee grounds
GluCers w/w
750.0
99.2
30.42
1.012
0.337
216.6
28.9
0.029%
300.0
0.218%
110.4
0.092%
34.4
0.712%
30.37
-
-
-
-
0.423
0.128
101.2
33.7
0.034%
300.0
0.333%
30.44
-
-
-
-
0.424
-
0.076
13,1
4.4
0.004%
0.038%
0.043%
-
-
-
a
Applied 20.0 g; the following fractions are calculated for the total organic phase; prep. HPLC − preparative HPLC.
Cers were recognized by comparison with authentic reference
standards by TLC.
in positive and negative ion mode (m/z 200.0−1000.0) with the
Xcalibur (V 2.0) software (Thermo Fisher Scientific, Bremen,
Germany).
Enzymatic Hydrolysis and Isolation of Ceramides. Fractions
containing suitable amounts of GluCer (A3, 6, 25.0 mg; A8, 27, 10.0
mg; A10, 35, 10.0 mg) were enzymatically hydrolyzed by scaling up
the analytical assay to 50 mL (37 °C, 14 h, final volume 50 mL, 20.65
mL of a 5 mg/mL sodium taurocholate solution, 100 μg of
imiglucerase, 20 mL of chloroform−methanol 2:1 (v/v) for
separation). The organic phase was separated by CC (30.0 g silica
gel 60, 1.5 × 38 cm, chloroform−methanol 9:1 (v/v) isocratic, flow 1
mL/min) and afforded Cers 46 (11.7 mg), 47 (5.9 mg), and 48 (4.1
mg).
LC-MS Analysis. For analytical LC-MS analysis, an Agilent 1260
Infinity LC System coupled to an Agilent 6120 B Quadrupole MS
(Agilent Technologies, Santa Clara, CA) was used. Separation was
achieved on a Zorbax Extend-C18 column (5 μm, 4.6 × 150 mm, 80 Å,
Agilent Technologies, Santa Clara, CA) with a precolumn Zorbax
Extend-C18 Analytical Guard Column (5 μm, 4.6 × 12.5 mm, Agilent
Technologies, Santa Clara, CA). The injection volume was 10 μL, and
column temperature was set to 30 °C. The elution solvent was
composed of methanol (A) and water (B), each with 0.1% formic acid.
The gradient elution program (flow rate 1.2 mL/min) was as follows:
linear gradient from 0 min A−B 97:3 (v/v) to 20 min A−B 100:0 (v/
v), followed by isocratic elution for 10 min A−B 100:0 (v/v) and an
equilibration step (5 min A−B 97:3 (v/v)) to the next run. The APCI/
MS parameters were set as follows: discharge voltage +4.5 kV,
discharge current 5 μA, gas temperature 200 °C, vaporizer temperature
450 °C, nebulizer gas (nitrogen) 35 psig, and drying gas (nitrogen) 12
L/min. Additionally, a collision-induced dissociation (CID) of 150 V
was used. Spectra were recorded in positive and full scan mode (TIC,
150.0−1000.0 m/z). Chromatograms and spectra were evaluated with
the Agilent ChemStation (B.04.03) software (Agilent, Santa Clara,
CA).
ESI-HRMS and ESI-MS/MS Analyses. Positive ion high resolution
ESI mass spectra were obtained from an Orbitrap Elite mass
spectrometer (Thermo Fisher Scientific, Bremen, Germany) which
was equipped with a heated electrospray ion source (spray voltage
+4.0 kV, capillary temperature 275 °C, source heater temperature 45
°C). Nitrogen was used as sheath gas. The sample solutions (1 μg/mL
in methanol) were injected continuously with a 500 μL of Hamilton
syringe pump (flow rate 5 μL/min). Fourier transform MS resolution
was 30 000. The instrument was externally calibrated by the Pierce
LTQ Velos ESI positive ion calibration solution from Thermo Fisher
Scientific (Bremen, Germany). The data were evaluated with the
Xcalibur (2.7 SP1) software (Thermo Fisher Scientific, Bremen,
Germany).
ESI tandem mass spectra were measured with a Thermo Finnigan
LCQ classic ion trap mass spectrometer (Thermo Fisher Scientific,
Bremen, Germany). It was equipped with an electrospray ion source
that was used in positive and negative mode. The MS conditions were
optimized with a GluCer standard substance to give the following
parameters (positive/negative mode): spray voltage 4.5 kV, capillary
voltage −20/+8.5 V, sheath gas 80/80 AU (nitrogen), capillary
temperature 220/220 °C, tube lens offset −20/+35 V. Samples (1 μg/
mL in methanol) were injected with a 250 μL Hamilton syringe pump
(flow rate 5 μL/min). MS/MS scans were recorded with CID between
35 and 40% (helium, isolation width 1.0−2.0). Spectra were recorded
NMR Analysis. Samples were dried before use in a desiccator with
KOH for at least 48 h. All samples were dissolved in 0.65 mL of
pyridine-d5 with TMS as internal standard. 1D- (1H, 13C, 1D-TOCSY)
and 2D- (COSY, TOCSY, HSQC, HMBC) NMR spectra were
measured at 25 °C with a Varian VNMRS 600 MHz NMR
spectrometer (Agilent Technologies, Santa Clara, CA). Proton and
carbon NMR frequencies were 599.828 and 150.840 MHz. Chemical
shifts δ [ppm] were referenced to the internal tetramethylsilane (δ 1H
= 0.000 ppm) and internal pyridine-d5 (δ 13C = 123.5 ppm),
respectively. Spectra were processed with MestReNova (8.0.0)
software (Mestrelab Research, Escondido, CA).
RESULTS AND DISCUSSION
■
Extraction of GluCers. Three industrial waste materials (A,
apple pomace; W, wheat germs; and C, coffee grounds) were
investigated. The procedure of GluCer enrichment could be
monitored by TLC (see Supporting Information). Different
lipid classes were assigned according to the reference mixture.
GluCer bands of low intensity were detected in the total lipid
extract and in the organic phase. Fractionation by the first
column chromatography on silica gel removed excessive
amounts of triglycerides, sterols, and free fatty acids. The
second column chromatography on Sephadex LH20 separated
sterol glycosides from GluCer. Total lipid extract of wheat
germs (36.8% w/w) was higher than that of apple pomace
(13.2% w/w) and coffee grounds (11.5% w/w) (Table 1).
However, all organic phases showed similar amounts
(approximately 30.4 g) whereas apple pomace had a higher
initial mass (750.0 g) in contrast to wheat germs and coffee
grounds (300.0 g). The liquid−liquid extraction could remove
nearly 2/3 of the total lipid extract of apple pomace and wheat
germs in the form of polar compounds. In contrast, this
extraction procedure could only eliminate 1/10 of the coffee
grounds total lipid extract, and polar compounds were already
extracted by percolation beforehand. The lipid contents after
extraction and liquid−liquid extraction were similar for apple
pomace (4.1% w/w), wheat germs (10.1% w/w), and coffee
grounds (10.2% w/w) (Table 1).
Quality of Separation Techniques. In the first separation
CC on silica gel 60, GluCers and sterol glucosides eluted
together. Only a coarse separation was possible but necessary
for an efficient and economical method. The ratio of sample
and stationary phase was only 8:100 (20.0:250.0 g, mobile
phase 3 L) and not suitable for a complete purification of
GluCers. Higher volumes of mobile phase and stationary phase
(2000 g of silica gel and 10 L of mobile phase for 20 g of
extract) should be used in a single CC step following standard
procedures. Size exclusion chromatography on sephadex LH-20
eventually separated sterol glucosides from GluCers (ratio:
1509
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
sample/stationary phase 0.3−0.7:100) and resulted in highly
GluCer enriched fractions. Application of reusable sephadex
LH-20 has economical advantages. The combination of two
chromatographic separation types is efficient to reduce
production costs for GluCer enriched fractions. Subsequent
preparative HPLC-MS afforded GluCers fractions A1-12, W1-
12, and C1-5. GluCers peaks were not baseline-separated. LC-
peaks with shoulders indicate more than one GluCer within
one fraction, but compounds could be distinguished individ-
ually by m/z as [M + H]+ ions (Figure 2). Each fraction
Thus, Glc-d18:2 h22:0 (30) and Glc-t18:1 h23:0 (31) eluted
together within one fraction (Figure 2, A9). Wheat germs and
coffee ground fractions showed similar separation results
(Supporting Information).
Structural information was obtained from ESI-HRMS
measurements. All compounds were measured as [M + Na]+
ions (Table 2). Errors were between 0.0000 and 4.6510 ppm
with most errors ≤2 ppm and greater errors for trace
compounds. ESI-MS/MS experiments showed [M + Na]+
ions in positive and [M − H]− ions in negative mode. Major
MS/MS fragments are presented in Figure 1 (all detected
fragments in Supporting Information) and were assigned
according to Ann and Adams.31 MS/MS fragmentation
revealed sugar cleavage as well as intact Cer structures (Y0
[M − C6H10O5 + Na]+, Z0 [M − C6H10O5 − H2O + Na]+) and
fragments of sphingoid bases (O) and fatty acids (U) for every
compound (Table 2). Acid chain lengths were assigned by
fragment U (e.g., h16:0 → m/z 271.5, h22:0 → m/z 355.4,
h24:0 → m/z 383.5). Desaturations within the fatty acids were
indicated by fragment U (e.g., h24:0 → m/z 383.5, h24:1 →
381.5, h24:2 → m/z 379.5) and only detected in wheat germs.
Sphingoid bases were identified according to the emerging
fragment O (e.g., d18:1 → m/z 484.2, d18:2 → m/z 482.2,
d18:3 → m/z 480.2, t18:1 → m/z 500.3).
Unknown derivatives of sphingadienine (d18:2) and
phytosphingenine (t18:1) type LCB were identified in trace
amounts in apple pomace (Figure 2). ESI-HRMS measure-
ments led to the assumption of additional oxygen groups
because of calculated total formulas containing 10 and 11
oxygens (Table 2). MS/MS data indicated an additional +14
amu in the fragment O (m/z 496.2) and suggested an oxo
group within the LCB. This LCB was attached to a h16:0 fatty
acid (U, m/z 271.4). An additional water elimination in
negative mode (Z0′, m/z 528.4) reinforced the structure
assignment to give Glc-oxo-d18:2 h16:0 (3). Another derivative
with an additional hydroxyl group within the LCB was
determined (O, m/z 498.3) as OH-d18:2. It was associated
with the same fatty acid (h16:0, U m/z 271.4). Two more
unknown GluCer classes were identified as 2OH-d18:2 and
OH-t18:1. 2OH-d18:2 showed +32 amu (O, m/z 514.3) in
contrast to d18:2. In addition to fragment Z0′ (e.g., m/z 658.6
for Glc-2OH-d18:2 h24:0 (24)) this class also showed an
additional ion Z0″ (e.g., m/z 640.5 for Glc-2OH-d18:2 h24:0
(24)) due to a water elimination. The second class was
determined as OH-t18:1 because of a + 16 amu difference of
the LCB (O, m/z 516.4) in contrast to t18:1. Fatty acids were
h16:0 (m/z 271.4), h22:0 (m/z 355.6) and h24:0 (m/z 383.5).
Due insufficient amounts of compounds, the position of the
functional groups could not be assigned. Similar compounds
have not been reported from plants. Recent studies announced
a Cer with unknown location of a hydroxyl group in human
skin (Cer NT, tetra18:0 18:0).2 Their complete structures need
to be fully elucidated in future studies.
Figure 2. Apple pomace GluCer fractions obtained by HPLC-APCI/
MS as extracted ions [M + H]+; compound numbers as given in Figure
3; a, assumed compounds.
contained up to 6 GluCers. Lipids are usually extracted26 and
isolated27 using chloroform and methanol as solvents. Chloro-
form is potentially carcinogenic and reprotoxic, and it is
therefore necessary to avoid this solvent during large scale
production.
GluCer Contents. Apple pomace provided a GluCer
content of 28.9 mg, wheat germs of 33.7 mg, and coffee
grounds of 4.4 mg per 100 g of plant material. GluCer contents
between 17 and 100 mg per 100 g of dry weight were reported
in different apple cultivars.15,16 Therefore, pomace of specific
apple varieties should be further studied to enhance the GluCer
yield. GluCers between 19.3 and 25.9 mg per 100 g of dry
weight (calc. as Glc-d18:2 h16:0, 714 g/mol) were recognized
from different wheat sources (whole grain,28 bran,21 and
flour14). The highest lipid content is located in wheat germs.
Thus, wheat germs exceeded these amounts with 33.7 mg of
GluCers per 100 g of plant material. Maximum GluCer
amounts of 28.4 to 35.1 mg per 100 g of dry weight (calc. as
Glc-d18:2 h16:0, 714 g/mol) are reported for soy beans used
for plant GluCer production.29,30 Thus, apple pomace and
wheat germs may serve as new GluCer resources, while coffee
grounds are no suitable source (4.4 mg per 100 g of dry
weight).
MS Analyses. In analytical HPLC-APCI/MS, GluCers were
detected as [M + H]+ ions or [M − 18 + H]+ in the case of one
GluCer class due to fragmentation conditions (CID 150 V)
(Table 2). The extracted ion chromatograms of apple pomace
GluCers are shown in Figure 2. Phytosphingosine (t18:1) and
sphingadienine (d18:2) represented the two major GluCer
sphingoid bases. The GluCers were separated according to their
fatty acid chain lengths as well as to their sphingoid bases.
1H NMR Fingerprinting. As most fractions contained more
than one GluCer, NMR measurements were accomplished as
1H NMR fingerprinting. The core structures sphingosine
(d18:1 as Glc-d18:1 h16:0, 8), sphingadienine (d18:2 as Glc-
d18:2 h16:0, 6), and phytosphingenine (t18:1 as Glc-t18:1
h22:0, 27) were investigated for relevant chemical shifts after
extensive structure elucidation by 1D (1H, 13C) and 2D
(HSQC, HMBC, COSY, TOCSY) NMR experiments. The
configuration of double bonds was determined according to the
1
size of vicinal H,1H coupling constants or 13C chemical shifts
1510
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
Figure 3. Individual GluCers (1−45) isolated from apple pomace (A), wheat germs (W), and coffee grounds (C). Intact GluCers were not reported
for apple pomace and coffee grounds yet; intact GluCer species identified before in wheat are denoted. Glc - β-glucose; a, position not proven; b,
suggested structure.
of adjacent carbons. E-double bonds have vicinal coupling
constants of 12−16 Hz, Z-configuration rather 8−10 Hz.32 In
long acyl chains adjacent carbons show chemical shifts of 27−
28 ppm for a Z-olefinic double bond and 32−33 ppm for an E-
olefinic double bond.33 The configurations were determined by
comparison with published 13C chemical shifts of the GluCers
as 2S, 3R, 2′R for sphingadienine (d18:2)34 and sphingosine
(d18:1)21 and 2S, 3S, 4R, 2′R for phytosphingenine (t18:1).35
LCB (NH, Δ4,5, Δ8,9), fatty acid (2′), and sugar (1″ and 5″)
proton signals could be assigned (Figure 4). All fractions
obtained from apple pomace, wheat germs, and coffee grounds
by HPLC (Figure 2; Supporting Information Figures 4, 5) were
monitored by NMR fingerprinting. Figure 4 contains a wheat
germ fraction with 6 GluCers (W4) as an example (see
due to NH and sugar signals (d18:1, d18:2, t18:1). An
additional NH signal was assumed to belong to d18:3, but this
assumption could not be proven due to the missing pure
reference substance. Double bonds at positions C-4,5 and C-8,9
were also detected. The attached sugar moiety was glucose due
to chemical shifts, and the β-glucosidic linkage was indicated by
the coupling constants (7.8 Hz).34 Fatty acids were always α-
hydroxylated (C-2′). The unsaturation of fatty acids could not
be clearly distinguished as these signals fell together with other
olefinic signals from C-8,9. In summary, 1H NMR finger-
printing as a simple and fast method supported identification by
mass spectrometry.
GluCer in Different Plant Materials. Together, 45
GluCer were elucidated (A, 22; W, 35; C, 7; Figure 3 and
Table 2). Glc-d18:2 h16:0 (6), Glc-t18:1 h22:0 (27), and Glc-
t18:1 h24:0 (35) were the main compounds of all three plant
1
Supporting Information Figures 19−21 for H NMR finger-
printing of all fractions). Three different LCBs were assigned
1511
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
a
Table 2. MS Identification of GluCers and Cers
ESI-HRMS
LC-APCI/MS
ESI-MS/MS
m/z
formula
RT
m/z
m/z
[M − acyl
[M − sug − acyl
source
compound
[M + Na]+
[M + Na]+
[min] [M + H]+ [M + Na]+
+ Na]+ O
[M − H+]−
− H]−
U
W
Glc-d18:1 h16:0 (8)
Glc-d18:1 h18:0 (14)
Glc-d18:1 h20:0 (25)
Glc-d18:1 h22:0 (33)
Glc-d18:1 h24:0 (41)
Glc-d18:2 h15:0 (4)
Glc-d18:2 h16:0 (6)
Glc-d18:2 h18:0 (10)
Glc-d18:2 h19:0 (15)
Glc-d18:2 h20:0 (21)
Glc-d18:2 h21:0 (26)
Glc-d18:2 h22:0 (30)
Glc-d18:2 h23:0 (34)
Glc-d18:2 h24:0 (38)
Glc-d18:2 h25:0 (42)
Glc-d18:2 h26:0 (45)
Glc-d18:2 h18:1 (7)
Glc-d18:2 h20:1 (11)
Glc-d18:2 h22:1 (23)
Glc-d18:2 h24:1 (29)
Glc-d18:2 h25:1 (36)
Glc-d18:2 h26:1 (39)
Glc-d18:3 h20:0 (13)
Glc-t18:1 h16:0 (5)
Glc-t18:1 h20:0 (16)
Glc-t18:1 h21:0 (22)
Glc-t18:1 h22:0 (27)
Glc-t18:1 h23:0 (31)
Glc-t18:1 h24:0 (35)
Glc-t18:1 h25:0 (40)
Glc-t18:1 h26:0 (43)
Glc-t18:1 h22:1 (18)
Glc-t18:1 h23:1 (20)
Glc-t18:1 h24:1 (28)
Glc-t18:1 h25:1 (32)
Glc-t18:1 h26:1 (37)
Glc-t18:1 h28:1 (44)
Glc-t18:1 h24:2 (19)
738.5496
766.5805
794.6127
822.6459
850.6772
722.5198
736.5346
764.5661
778.5836
792.5967
806.6135
820.6277
834.6432
848.6604
862.6759
876.6927
762.5526
790.5822
818.6113
846.6441
860.6609
874.6749
790.5811
754.5463
810.6072
824.6237
838.6401
852.6540
866.6706
880.6863
894.7024
836.6228
850.6364
864.6549
878.6714
892.6862
920.7188
862.6389
752.5254
750.5123
768.5236
852.6189
880.6498
854.6351
882.6660
574.4810
676.5853
704.6165
C40H77NO9Na
C42H81NO9Na
C44H85NO9Na
C46H89NO9Na
C48H93NO9Na
C39H73NO9Na
C40H75NO9Na
C42H79NO9Na
C43H81NO9Na
C44H83NO9Na
C45H85NO9Na
C46H87NO9Na
C47H89NO9Na
C48H91NO9Na
C49H93NO9Na
C50H95NO9Na
C42H77NO9Na
C44H81NO9Na
C46H85NO9Na
C48H89NO9Na
C49H91NO9Na
C50H93NO9Na
C44H81NO9Na
C40H77NO10Na
C44H85NO10Na
C45H87NO10Na
C46H89NO10Na
C47H91NO10Na
C48H93NO10Na
C49H95NO10Na
C50H97NO10Na
C46H87NO10Na
C47H89NO10Na
C48H91NO10Na
C49H93NO10Na
C50H95NO10Na
C52H99NO10Na
C48H89NO10Na
C40H75NO10Na
C40H73NO10Na
C40H75NO11Na
C46H87NO11Na
C48H91NO11Na
C46H89NO11Na
C48H93NO11Na
C34H65NO4Na
C40H79NO5Na
C42H83NO5Na
9.2
12.3
16.9
21.0
25.4
7.3
716.5
744.6
772.6
800.7
828.7
700.5
714.5
742.6
756.5
770.5
784.7
798.7
812.6
826.7
840.7
854.6
740.6
768.5
796.7
824.7
838.6
852.6
768.5
732.6
788.5
802.6
816.6
830.8
844.7
858.6
872.7
814.5
828.7
842.7
856.6
870.7
898.6
840.6
730.5
728.5
746.5
830.6
858.7
814.5
842.5
552.5
654.6
682.7
738.6
766.6
794.6
822.6
850.7
722.5
736.6
764.6
778.6
792.6
806.6
820.6
834.7
848.6
862.7
876.6
762.7
790.6
818.5
846.6
860.7
874.5
790.6
754.6
810.6
824.5
838.7
852.7
866.7
880.6
894.7
836.6
850.5
864.6
878.6
892.6
920.7
862.6
752.5
750.5
768.4
852.6
880.6
854.7
882.5
574.5
676.6
704.6
484.2
484.3
484.2
484.2
484.2
482.2
482.2
482.2
482.2
482.2
482.3
482.2
482.3
482.2
482.2
482.2
482.2
482.3
482.2
482.3
482.2
482.2
480.2
500.3
500.3
500.3
500.3
500.2
500.3
500.3
500.2
500.3
500.3
500.2
500.3
500.2
500.3
500.3
498.3
496.2
514.3
514.3
514.3
516.3
516.4
320.2
338.1
714.5
742.6
770.6
798.6
826.6
271.5
299.5
327.4
355.4
383.5
257.4
271.3
299.4
313.5
327.5
341.4
355.5
369.6
383.5
397.5
411.5
297.4
325.3
353.4
381.5
395.6
409.6
327.5
271.4
327.6
341.6
355.5
369.7
383.5
397.6
411.6
353.5
367.5
381.5
395.5
409.5
437.7
379.5
271.4
271.4
271.4
355.5
383.5
355.6
383.5
271.5
355.5
383.5
W
W
W
W
d
A
698.5
A, W, C
8.4
712.6
740.6
754.6
768.6
782.6
796.6
810.3
824.6
838.7
852.7
W
11.1
12.4
15.1
17.4
19.5
21.3
23.3
25.6
28.8
8.5
W
A, W
W
A, W, C
A, W
A, W, C
A, W
W
d
W
738.5
W
11.2
15.2
19.4
21.6
23.4
12.2
7.3
766.6
794.5
822.6
836.6
850.6
766.6
730.6
786.6
800.6
814.7
828.6
842.6
856.6
870.7
812.6
826.5
840.6
854.8
868.6
896.4
838.6
728.5
726.5
744.4
828.6
856.6
830.7
858.6
550.6
652.6
680.6
W
W
W
W
W
A
A, W
12.6
15.1
17.6
19.5
21.4
23.8
25.8
12.7
15.0
17.7
19.6
21.6
25.8
13.5
4.9
A
A, W, C
A, W, C
A, W, C
A, W, C
A, W
W
W
W
W
W
W
W
A
b
Glc-OH-d18:2 h16:0 (1)
b
A
Glc-oxo-d18:2 h16:0 (3)
5.3
b
A
Glc-2OH-d18:2 h16:0 (2)
5.1
b
A
Glc-2OH-d18:2 h22:0 (12)
11.8
16.5
9.7
b
A
Glc-2OH-d18:2 h24:0 (24)
b
c
c
A
Glc-OH-t18:1 h22:0 (9)
b
d
A
Glc-OH-t18:1 h24:0 (17)
11.8
9.9
A
d18:2 h16:0 (46)
t18:1 h22:0 (47)
t18:1 h24:0 (48)
A
21.1
24.6
A
338.1
a
b
Compound numbers (bold) as given in Figure 3. RT, retention time; acyl, acyl moiety; sug, sugar moiety. Proposed structure (location and
c
d
stereochemistry of double bonds and functional groups not known). Detected as [M − H2O + H]+ with CID 150 V. Ions not detected in MS1 but
indirectly by MS/MS measurements
materials. The GluCers consisted mainly of three LCB (d18:1,
d18:2, t18:1, all C18), of α-hydroxylated fatty acids (C15−28,
saturated, mono- and polyunsaturated) and of β-glucose.
Dihydroxy (d18:1, d18:2) LCB were rather bound to short
chain fatty acids (≤20) and trihydroxy (t18:1) to long-chain
fatty acids (≥22).36 In mammals sphingosine (d18:1Δ4) is also
mostly attached to short chain and phytosphingosine (t18:0) to
long chain fatty acids.37 All Δ4,5 double bonds were
determined as E, Δ8,9 as E and Z configurations. Fatty acid
desaturations were only Z-configured. Sphingosine (d18:1) and
1512
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
The position of desaturation could not be identified here.
Unsaturated fatty α-hydroxylated acids were defined as Z-type
and ω-9 series, which was formerly reported in chilling-resistant
plants like wheat.40 GluCers with fatty acids h28:1 (44) and
h24:2 (19) were identified in wheat for the the first time. Intact
GluCers were detected in coffee grounds after roasting and
brewing, which gives evidence for the stability of GluCer
species. However, their content was low; thus, coffee grounds
do not appear suitable for future phyto-Cer production.
Enzymatic Hydrolysis with Imiglucerase. A crucial step
of the phyto-Cer production is the sugar hydrolysis. Acidic
hydrolyses were not satisfying for GluCer sugar cleavage
because of an almost complete loss of trihydroxy ceramides
based on phytosphingosine and phytosphingenine. This was
stated before, when a two-step oxidation−reduction−hydrolysis
was investigated GluCers.41 Thus, enzymatic hydrolysis was
tested. A specific human recombinant glucocebrosidase
(imiglucerase, Cerezyme) was used for hydrolysis. All GluCer
fractions (A1−12, W1−12, C1−5) were hydrolyzed. TLC was
used for survey (Figure 5). All compounds could be cleaved
with an active enzyme; the GluCer bands disappeared and Cer
bands appeared with the same intensity. GluCers were not
hydrolyzed by the inactivated enzyme. There are no reports on
enzymatic hydrolysis of plant GluCers with human recombi-
nant glucocebrosidase so far. Glycosylceramidase activity was
analyzed in plants before.23 However, no glucocerebrosidase
was assigned in plants yet.
Figure 4. Major 1H NMR signals for structure identification (600
MHz, pyridine-d5). Hydrogens bound to sphingoid nitrogen assigned
as NH, hydrogens bound to carbon indicated by carbon number (′ for
fatty acid carbon, ′′ for sugar carbon) as in Figure 1; sample W4: Glc-
d18:3* h20:0 (13), Glc-d18:1 h18:0 (14), Glc-d18:2 h19:0 (15), Glc-
t18:1 h20:0 (16), Glc-t18:1 h22:1 (18), Glc-t18:1 h24:2 (19); *,
signal not proven (no reference available).
unsaturated fatty acids were only detected in wheat germs.
Some new LCB derivatives were identified in apple pomace by
mass spectrometry but need further proof due to their trace
levels.
In apple pomace, Glc-d18:2Δ4,8 h16:0 (6) and Glc-t18:1Δ8
h22/23/24:0 (27, 31, 35) were the main compounds, which is
consistent with previous studies.15,19 Wheat germs showed a
more complex range of GluCers. Sphingadienine (d18:2Δ4,8),
sphingosine (d18:1Δ8), and phytosphingenine (t18:1Δ8) were
found to be the major LCBs and h16:0, h18:0, and h20:0 the
main fatty acids. This was also reported before in wheat
grains.28 Glc-d18:2Δ4,8 h16:0 (6), Glc-d18:2Δ4,8 h20:0 (21), and
Glc-d18:1Δ8 h16:0 (8) were also reported to be the main
compounds in wheat bran21 and germs.20 Trace levels of
sphingatrienine (d18:3, 13) were identified in wheat germs.
This sphingoid base was detected before in tobacco leaves,38
rice, and maize39 as sphinga-4,8,10-trienine (d18:3Δ4E,8E,10E) .
Three GluCers from apple pomace (Glc-d18:2 h16:0 (7);
Glc-t18:1 h22:0 (27); and Glc-t18:1 h24:0 (29)) were
individually hydrolyzed by the enzyme. The hydrolysis did
not proceed quantitatively (yield 51−74%) and was substrate-
dependent. Higher substrate concentrations in the higher μM
range (200−700 μM) were not hydrolyzed completely. On the
other hand, GluCers concentrations in the lower μM range
(12−14 μM) were totally converted. The same was observed
using an endo glucosylceramidase (EC 3.2.1.123) from leech.42
The enzyme could only transform approximately 20% of the
Figure 5. Enzymatic hydrolysis with glucocerebrosidase. Enzymatic hydrolysis of fractions from apple pomace (A; A), wheat germs (W; B), and
coffee grounds (C; C) with active (+) and inactive (−) enzyme.
1513
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
substrate (150 μM). Lower substrate concentrations (37.5 μM)
and a different detergent led to hydrolysis up to 34%.42
is a novel procedure for phyto-Cer production. Availability of
the imiglucerase is ensured by heterologous production;
however, costs for the human-identical enzyme used for
infusion therapy are high. Recombinant glucocerebrosidase is
produced in plants, plant cells, and mammalian cells. The
heterologous proteins require, in addition to rigorous
purification, enzymatic tailoring of glycosylation to ensure
nonimmunogenicity and macrophage targeting.46 For bio-
technological GluCer hydrolysis, glucocerebrosidase should
have optimal activity, but the protein structure does not need
tailoring after expression; thus, heterologous enzyme produc-
tion can be performed in a convenient microorganism,
providing cheap enzyme protein.
The cleavage of the sugar moiety was demonstrated by NMR
experiments (see Supporting Information). Sugar-bound hydro-
gens (H-1″ to H-6″) were not detectable after hydrolysis. The
structures of these Cers were also elucidated by HPLC-MS,
ESI-MS/MS, ESI-HRMS, and NMR experiments and were
consistent with their GluCer precursors. Cers could be analyzed
under the same conditions as GluCers. All relevant m/z are
shown in Table 2. 1H- and 13C-chemical shifts are presented in
Table 3 and Table 4 and were assigned by COSY, TOCSY,
1
Table 3. H-, 13C-Chemical Shifts (pyridine-d5) of
d18:2Δ4E,Δ8E/Z h16:0 (46)
Phyto-Cers and Their Possible Role in Dermatother-
apy. Phyto-Cers are similar but only partially identical to
human skin Cers (Figure 6). Phyto-Cers may differ in chain
lengths and degree of desaturation. Skin Cers typically show
sphingoid base chain lengths of C18 to 22 and fatty acids of
C16 to 30, of which C22 to 26 are most abundant.37,47 C16,
C20, C22, and C24 are the major fatty acid chain lengths in
plants. Sphingosine (d18:1) type Cers (e.g., Cer [AS, d18:1
h18:0] and [NS, d18:1 18:0]) represent approximately 17%
and phytosphingosine (t18:0) type Cers (e.g., Cers [AP, t18:0
h18:0] and [NP, t18:0 18:0]) approximately 31% of the
stratum corneum Cers. Other major Cer groups derive from 6-
hydroxysphingosine and from sphinganine (for human skin Cer
nomenclature see Figure 6 and Supporting Information).2
Cers contribute to the skin barrier function against water
loss. The hydroxylated head core groups of the Cers influence
the packing of stratum corneum lipids, and phyto-Cers [NP]
and [AP] have a high number of hydroxyl groups building
hydrogen bonds to increase the stability of the lipid phases.7,8
Plant Cers with three or four hydroxyl groups may serve as
stabilizing and water binding agents in therapy. Investigation of
the Cer composition of diseased skin, e.g., of psoriasis skin
patches or of skin from atopic dermatitis, revealed a decrease
specifically of phyto-Cers and other highly hydroxylated Cer
species.5,6 It is therefore obvious to attempt treatment with
phyto-Cers in particular. For this purpose, phyto-Cers need to
be available, and the structures and the composition of Cers
from plant sources need to be elucidated.
On the other hand, an α-hydroxy group (e.g., Cer [AP]) can
lead to decreasing hydrogen bonding48 and to a less stabilized
lamellar order, at least in artificial skin membranes.49 GluCers
with five hydroxyl groups described here (e.g., Glc-2OH-d18:2
h24:0 (24) and Glc-OH-t18:1 h24:0 (17)) could provide still
different features. Their function in plant cells and their
possible application in skin therapy must be further elucidated.
Further, benefits or disadvantages of desaturations in plant
sphingolipids, e.g., at position Δ8 or within the fatty acid (ω-9),
need investigation. It was shown that lipids with a Z-
configuration can stretch the lipid layers and embed water.50
However, monounsaturated free fatty acids led to a disturbed
lipid packing51 which impairs the integrity of the stratum
corneum. Therefore, comparative studies with phyto-Cers and
Cers established skin treatment are necessary. For such studies,
phyto-Cers need to be isolated. If they show beneficial effects
and are similar to each other in activity, complete separation of
individual phyto-Cer species will not be necessary. Enriched
fractions containing a defined spectrum of phyto-Cers may be
applied and can be economically produced.
δH [ppm] (M, J [Hz])
Δ8Z Δ8E
δC [ppm]
position
Δ8Z
Δ8E
Sphingoid Base
1a
1b
2
4.521 (m)
4.264 (m)
4.729 (m)
4.888 (m)
61.91
61.93
-
56.13
56.12
3
73.09
73.07
4
6.109 (m, 15.4)
6.040 (m, 15.4)
132.55
131.84
32.96
132.43
131.84
32.97
d
d
5
a
6
2.219 (m)
a
a
7
2.221 (m)
2.175 (m)
27.41
32.80
c
c
8
5.497 (m)
5.513 (m)
129.43
130.65
27.61
129.94
131.15
32.95
c
c
9
5.497 (m)
5.513 (m)
c
a
10
2.086 (m)
2.022 (m)
b
11
1.384 (m)
1.260 (m)
-
29.53−30.08
29.53−30.08
32.14−32.15
22.96−22.97
14.31
12−16
16
17
1.260 (m)
0.873 (t, 6.9)
18
NH
Fatty Acid
1′
8.392 (d, 8.9)
8.386 (d, 8.8)
-
-
175.38
175.37
72.57
35.85
2′
4.635 (dt, 8.0; 3.9)
72.55
35.84
a
3′a
3′b
4′a
4′b
5′
6′−14′
14′
15′
16′
2.262 (m)
a
2.077 (m)
-
-
1.842 (m)
1.764(m)
25.92
25.90
b
1.384 (m)
29.53−30.08
29.53−30.08
32.14−32.15
22.96−22.97
14.31
1.260 (m)
-
1.260 (m)
0.873 (t, 6.9)
a
b
1H chemical shifts from HSQC correlations. 1H chemical shifts from
c
HMBC correlations. 1H chemical shifts from TOCSY correlations.
d
Signals lay upon each other.
HSQC, and HMBC correlations. Structure elucidation
confirmed (2S,3R,4E,8E/Z)-2N-[(2′R)-2′-hydroxyhexadeca-
noylamino]-4(E),-8(E/Z)-octadecadiene-1,3-diol (d18:2
h16:0, 46), (2S,3S,4R,8E/Z)-2N-[(2′R)-2′-hydroxydocosanoy-
lamino]-8(E/Z)-octadecadiene-1,3,4-triol (t18:1 h22:0, 47),
and (2S,3S,4R,8E/Z)-2N-[(2′R)-2′-hydroxytetracosanoylami-
no]-8(E/Z)-octadecadiene-1,3,4-triol (t18:1 h24:0, 48) as the
hydrolyzed phyto-Cers (Figure 6). These compounds were
isolated before from plants as trace substances.43−45 The
enzymatic hydrolysis of these phyto-Cers from GluCers was
not reported so far. As Cers contents are essentially lower than
GluCers amounts in plants,12 hydrolysis by glucocerebrosidase
An upscaling process for phyto-ceramides as rare and
expensive natural products appears possible now and should
1514
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
1
Table 4. H-, 13C-Chemical Shifts (pyridine-d5) of t18:1Δ8E/Z h22:0 (47) and t18:1Δ8E/Z h24:0 (48)
t18:1Δ8E/Z h22:0
δH [ppm] (M, J [Hz])
Δ8Z Δ8E
t18:1Δ8E/Z h24:0
δH [ppm] (M, J [Hz])
Δ8Z Δ8E
δC [ppm]
δC [ppm]
position
Δ8Z
Δ8E
Δ8Z
Δ8E
Sphingoid Base
1a
4.531 (dd, 4.5; 10.8)
62.01
62.03
4.533 (dd, 4.6; 10.8)
62.00
62.03
1b
4.444 (dd, 5.0; 10.8)
5.139 (m)
-
4.446 (dd, 5.0; 10.8)
5.143 (m)
-
2
52.97
76.89
72.97
33.98
52.99
76.87
72.92
33.86
52.97
76.88
72.97
33.98
52.98
76.86
72.92
33.86
3
4.364 (dd, 4.5; 6.8)
4.307 (ddd, 2.6; 6.8; 9.2)
4.367 (dd, 4.4; 6.8)
4.310 (ddd, 2.6; 6,8; 9.3)
4
a
a
5a
2.335 (m)
2.335 (m)
a
a
5b
1.977 (m)
-
-
1.977 (m)
-
-
a
a
6a
2.052 (m)
26.92
26.76
2.054 (m)
26.91
26.76
a
a
6b
1.812 (m)
1.809 (m)
a
a
a
a
7
2.261 (m)
2.196 (m)
27.94
33.33
2.261 (m)
2.196 (m)
27.94
33.33
c
c
a
a
8
5.558 (m)
5.559 (m)
130.38
130.25
27.60
130.83
130.72
33.01
5.558 (m)
5.556 (m)
130.38
130.25
27.60
130.83
130.72
33.01
c
c
a
a
9
5.475 (m)
5.505 (m)
5.480 (m)
5.518 (m)
a
a
a
a
10
2.091 (m)
2.015 (m)
2.089 (m)
2.013 (m)
b
b
11
1.371 (m)
1.283 (m)
-
29.54−30.12
29.54−30.12
32.13−32.15
22.96
1.356 (m)
1.283 (m)
-
29.54−30.12
29.54−30.12
32.13−32.14
22.96
12−16
16
17
1.283 (m)
0.874 (t, 7.0)
1.283 (m)
0.875 (t, 6.8)
18
14.31
14.31
d
d
NH
8.603 (d, 9.0)
-
8.607 (d, 8.9)
-
Fatty Acid
1′
2′
3′a
3′b
4′a
4′b
5′
6′−20′/22′
20′/22′
21′/23′
22′/24′
-
175.23
72.47
35.74
-
-
175.23
72.47
35.74
-
4.639 (m)
4.641 (dd, 3.7; 8.0)
a
a
2.246 (m)
2.246 (m)
a
a
2.061 (m)
2.053 (m)
a
a
1.802 (m)
25.86
25.84
1.804 (m)
25.86
25.84
a
a
1.740 (m)
-
1.737 (m)
-
b
b
1.371 (m)
29.54−30.12
29.54−30.12
32.13−32.15
22.96
1.381 (m)
29.54−30.12
29.54−30.12
32.13−32.14
22.96
1.283 (m)
-
1.283 (m)
-
1.283 (m)
0.874 (t, 7.0)
1.283 (m)
0.875 (t, 6.8)
14.31
14.31
a
b
c
d
1H chemical shifts from HSQC correlations. 1H chemical shifts from HMBC correlations. 1H chemical shifts from TOCSY correlations. Signals
overlapping.
Figure 6. Isolated phyto-Cers and human skin Cers. Cer, ceramide; S, sphingosine; P, phytosphingosine; A, α-hydroxylated fatty acid; N,
nonhydroxylated fatty acid.
be pursued. With the methods described here, phyto-Cers can
be obtained from plant material.
1515
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
(4) Breiden, B.; Sandhoff, K. The role of sphingolipid metabolism in
cutaneous permeability barrier formation. Biochim. Biophys. Acta, Mol.
Cell Biol. Lipids 2014, 1841, 441−52.
(5) Motta, S.; Monti, M.; Sesana, S.; Caputo, R.; Carelli, S.; Ghidoni,
R. Ceramide composition of the psoriatic scale. Biochim. Biophys. Acta,
Mol. Basis Dis. 1993, 1182, 147−151.
(6) Ishikawa, J.; Narita, H.; Kondo, N.; Hotta, M.; Takagi, Y.;
Masukawa, Y.; Kitahara, T.; Takema, Y.; Koyano, S.; Yamazaki, S.;
Hatamochi, A. Changes in the ceramide profile of atopic dermatitis
patients. J. Invest. Dermatol. 2010, 130, 2511−4.
(7) Imokawa, G.; Akasaki, S.; Hattori, M.; Yoshizuka, N. Selective
recovery of deranged water-holding properties by stratum corneum
lipids. J. Invest. Dermatol. 1986, 87, 758−61.
(8) Pascher, I. Molecular arrangements in sphingolipids. Con-
formation and hydrogen bonding of ceramide and their implication on
membrane stability and permeability. Biochim. Biophys. Acta, Biomembr.
1976, 455, 433−51.
(9) Imokawa, G.; Kuno, H.; Kawai, M. Stratum-corneum lipids serve
as a bound-water modulator. J. Invest. Dermatol. 1990, 96, 845−851.
(10) Di Nardo, A.; Wertz, P.; Giannetti, A.; Seidenari, S. Ceramide
and cholesterol composition of the skin of patients with atopic
dermatitis. Acta Derm.-Venereol 1998, 78, 27−30.
(11) Maister, H. G.; Rogovin, S. P.; Stodola, F. H.; Wickerham, L. J.
Formation of extracellular sphingolipids by microorganisms: IV. Pilot-
plant production of tetraacetylphytosphingosine by Hansenula ciferrii.
Appl. Microbiol. 1962, 10, 401−6.
(12) Markham, J. E.; Jaworski, J. G. Rapid measurement of
sphingolipids from Arabidopsis thaliana by reversed-phase high-
performance liquid chromatography coupled to electrospray ionization
tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2007, 21,
1304−14.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jafc.6b04275.
■
S
GluCer purification of apple pomace, wheat germs, and
coffee grounds (TLC); detected GluCer species of wheat
germs and coffee grounds (HPLC-MS); detected MS/
MS fragments of GluCers and Cers; 1D and 2D NMR
spectra and chemical shifts of GluCers and Cers; main
and trace compound differentiation (HPLC-MS);
isolated Cer compounds; nomenclature system of skin
Cers (PDF)
AUTHOR INFORMATION
Corresponding Author
*(B.D.) Telephone: +49 341 97 30100. Fax: +49 341 97 30109.
E-mail: birgit.draeger@zv.uni-leipzig.de.
■
ORCID
Birgit Drager: 0000-0002-0615-5674
̈
Funding
M.R. kindly acknowledges the support of the German
Academic Exchange Service (DAAD) and of the German
Federal Ministry for Education and Research (BMBF) via
Project “Welcome to Africa” as well as the support of the
Biosolutions GmbH, Halle.
Notes
The authors declare no competing financial interest.
(13) Markham, J. E.; Lynch, D. V.; Napier, J. A.; Dunn, T. M.;
Cahoon, E. B. Plant sphingolipids: function follows form. Curr. Opin.
Plant Biol. 2013, 16, 350−7.
(14) Sugawara, T.; Miyazawa, T. Separation and determination of
glycolipids from edible plant sources by high-performance liquid
chromatography and evaporative light-scattering detection. Lipids
1999, 34, 1231−1237.
(15) Takakuwa, N.; Saito, K.; Ohnishi, M.; Oda, Y. Determination of
glucosylceramide contents in crop tissues and by-products from their
processing. Bioresour. Technol. 2005, 96, 1089−92.
(16) Ogawa, T.; Migita, H.; Shimada, S.; Ichida, J.; Osada, K. The
structure and level of glucosylceramide in apple pomace. Nippon
Shokuhin Kagaku Kogaku Kaishi 2014, 61, 251−257.
ACKNOWLEDGMENTS
■
M. Reisberg is grateful for helpful advice and the enzyme assay
conditions from Konrad Sandhoff (Life and Medical Sciences
Institute (LIMES), Membrane Biology & Lipid Biochemistry
́
Unit, c/o Kekule-Institute of Chemistry and Biochemistry,
Bonn, Germany) and provision of Cerezyme® by Christoph
Baerwald (University Hospital Leipzig, Department for Gastro-
enterology and Rheumatology, Rheumatology Unit, Leipzig,
Germany). The technical assistance from Anja Ehrlich, Manuela
Woigk and Annegret Laub is kindly appreciated.
(17) Vesper, H.; Schmelz, E. M.; Nikolova-Karakashian, M. N.;
Dillehay, D. L.; Lynch, D. V.; Merrill, A. H., Jr. Sphingolipids in food
and the emerging importance of sphingolipids to nutrition. J. Nutr.
1999, 129, 1239−1250.
(18) Shalini, R.; Gupta, D. K. Utilization of pomace from apple
processing industries: a review. J. Food Sci. Technol. (New Delhi, India)
2010, 47, 365−71.
(19) Whitaker, B. D. Analysis of plant cerebrosides by C18 and C6
HPLC. In Physiology, biochemistry and molecular biology of plant lipids;
Williams, J., Khan, M., Lem, N., Eds.; Springer Netherlands:
Dordrecht, 1997; pp 143−145.
ABBREVIATIONS USED
■
A, apple pomace; APCI, atmospheric pressure chemical
ionization; C, coffee grounds; CC, column chromatography;
Cer, ceramide; CID, collision-induced dissociation; ESI,
electrospray ionization; FAS, Foreign Agricultural Service
(United States Department of Agriculture); GluCer, glucosyl-
ceramide; LCB, long-chain base; MS, mass spectrometry;
NMR, nuclear magnetic resonance; RP, reversed phase; TIC,
total ion count; TLC, thin layer chromatography; TMS,
tetramethylsilane; W, wheat germs
(20) Sullards, M. C.; Lynch, D. V.; Merrill, A. H., Jr.; Adams, J.
Structure determination of soybean and wheat glucosylceramides by
tandem mass spectrometry. J. Mass Spectrom. 2000, 35, 347−53.
(21) Zhu, Y. D.; Soroka, D. N.; Sang, S. M. Structure elucidation and
chemical profile of sphingolipids in wheat bran and their cytotoxic
effects against human colon cancer cells. J. Agric. Food Chem. 2013, 61,
866−874.
REFERENCES
■
(1) Feingold, K. R. Thematic review series: skin lipids. The role of
epidermal lipids in cutaneous permeability barrier homeostasis. J. Lipid
Res. 2007, 48, 2531−46.
(2) t’Kindt, R.; Jorge, L.; Dumont, E.; Couturon, P.; David, F.;
Sandra, P.; Sandra, K. Profiling and characterizing skin ceramides using
reversed-phase liquid chromatography-quadrupole time-of-flight mass
spectrometry. Anal. Chem. 2012, 84, 403−11.
(3) Rabionet, M.; Gorgas, K.; Sandhoff, R. Ceramide synthesis in the
epidermis. Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2014, 1841,
422−34.
(22) Karlsson, K. A. Sphingolipid long chain bases. Lipids 1970, 5,
878−91.
(23) Lynch, D. V.; Dunn, T. M. An introduction to plant
sphingolipids and a review of recent advances in understanding their
metabolism and function. New Phytol. 2004, 161, 677−702.
1516
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517

Journal of Agricultural and Food Chemistry
Article
(24) Sperling, P.; Heinz, E. Plant sphingolipids: structural diversity,
biosynthesis, first genes and functions. Biochim. Biophys. Acta, Mol. Cell
Biol. Lipids 2003, 1632, 1−15.
(25) Ternes, P.; Wobbe, T.; Schwarz, M.; Albrecht, S.; Feussner, K.;
Riezman, I.; Cregg, J. M.; Heinz, E.; Riezman, H.; Feussner, I.;
Warnecke, D. Two pathways of sphingolipid biosynthesis are separated
in the yeast Pichia pastoris. J. Biol. Chem. 2011, 286, 11401−14.
(26) Folch, J.; Lees, M.; Sloane Stanley, G. H. A simple method for
the isolation and purification of total lipides from animal tissues. J. Biol.
Chem. 1957, 226, 497−509.
(27) Cahoon, E. B.; Lynch, D. V. Analysis of glucocerebrosides of rye
(Secale cereale L. cv Puma) leaf and plasma membrane. Plant Physiol.
1991, 95, 58−68.
(28) Fujino, Y.; Ohnishi, M. Sphingolipids in wheat-grain. J. Cereal
Sci. 1983, 1, 159−168.
(29) Gutierrez, E.; Wang, T.; Fehr, W. R. Quantification of
sphingolipids in soybeans. J. Am. Oil Chem. Soc. 2004, 81, 737−742.
(30) Wang, L.; Wang, T.; Fehr, W. R. HPLC quantification of
sphingolipids in soybeans with modified palmitate content. J. Agric.
Food Chem. 2006, 54, 7422−7428.
(31) Ann, Q.; Adams, J. Structure-specific collision-induced
fragmentations of ceramides cationized with alkali-metal ions. Anal.
Chem. 1993, 65, 7−13.
(32) Mitchell, T. N.; Costisella, B. NMR - From spectra to structures;
Springer: Berlin, Heidelberg, 2007.
(45) Hu, P. Y.; Zhang, D.; Pu, S. B.; Li, Z. L.; Zhou, H. H. Chemical
Constituents from the Root of Tripterygium wilfordii. Asian J. Chem.
2014, 26, 4344−4346.
(46) Limkul, J.; Iizuka, S.; Sato, Y.; Misaki, R.; Ohashi, T.; Ohashi, T.;
Fujiyama, K. The production of human glucocerebrosidase in glyco-
engineered Nicotiana benthamiana plants. Plant Biotechnol. J. 2016, 14,
1682.
(47) Masukawa, Y.; Narita, H.; Shimizu, E.; Kondo, N.; Sugai, Y.;
Oba, T.; Homma, R.; Ishikawa, J.; Takagi, Y.; Kitahara, T.; Takema, Y.;
Kita, K. Characterization of overall ceramide species in human stratum
corneum. J. Lipid Res. 2008, 49, 1466−1476.
(48) Corbe, E.; Laugel, C.; Yagoubi, N.; Baillet, A. Role of ceramide
structure and its microenvironment on the conformational order of
model stratum corneum lipids mixtures: an approach by FTIR
spectroscopy. Chem. Phys. Lipids 2007, 146, 67−75.
(49) Engelbrecht, T. N.; Schroeter, A.; Hauss, T.; Deme, B.; Scheidt,
H. A.; Huster, D.; Neubert, R. H. H. The impact of ceramides NP and
AP on the nanostructure of stratum corneum lipid bilayer. Part I:
neutron diffraction and H-2 NMR studies on multilamellar models
based on ceramides with symmetric alkyl chain length distribution. Soft
Matter 2012, 8, 6599−6607.
(50) Potts, R. O.; Francoeur, M. L. Lipid biophysics of water-loss
through the skin. Proc. Natl. Acad. Sci. U. S. A. 1990, 87, 3871−3873.
(51) Mojumdar, E. H.; Helder, R. W.; Gooris, G. S.; Bouwstra, J. A.
Monounsaturated fatty acids reduce the barrier of stratum corneum
lipid membranes by enhancing the formation of a hexagonal lateral
packing. Langmuir 2014, 30, 6534−43.
(33) De Haan, J. W.; Van De Ven, L. J. M. Configurations and
conformations in acyclic, unsaturated-hydrocarbons - 13C NMR study.
Org. Magn. Reson. 1973, 5, 147−153.
(34) Inagaki, M.; Harada, Y.; Yamada, K.; Isobe, R.; Higuchi, R.;
Matsuura, H.; Itakura, Y. Isolation and structure determination of
cerebrosides from garlic, the bulbs of Allium sativum L. Chem. Pharm.
Bull. 1998, 46, 1153−1156.
(35) Cateni, F.; Zilic, J.; Falsone, G.; Hollan, F.; Frausin, F.; Scarcia,
V. Preliminary biological assay on cerebroside mixture from Euphorbia
nicaeensis All. Isolation and structure determination of five
glucocerebrosides. Farmaco 2003, 58, 809−17.
(36) Fujino, Y.; Ohnishi, M.; Ito, S. Further studies on sphingolipids
in wheat grain. Lipids 1985, 20, 337−342.
(37) Wertz, P. W.; Schwartzendruber, D. C.; Madison, K. C.;
Downing, D. T. Composition and morphology of epidermal cyst lipids.
J. Invest. Dermatol. 1987, 89, 419−425.
(38) Sperling, P.; Franke, S.; Luthje, S.; Heinz, E. Are glucocerebro-
sides the predominant sphingolipids in plant plasma membranes? Plant
Physiol. Biochem. 2005, 43, 1031−8.
(39) Sugawara, T.; Duan, J.; Aida, K.; Tsuduki, T.; Hirata, T.
Identification of glucosylceramides containing sphingatrienine in maize
and rice using ion trap mass spectrometry. Lipids 2010, 45, 451−5.
(40) Imai, H.; Ohnishi, M.; Kinoshita, M.; Kojima, M.; Ito, S.
Structure and distribution of cerebroside containing unsaturated
hydroxy fatty acids in plant leaves. Biosci., Biotechnol., Biochem. 1995,
59, 1309−1313.
(41) Iga, S.; Iga, D. P.; Nigolescu, A.; Florentina, D.; Cîmpeanu, G.
Preparation of ceramide and sphingosine by chemical and biochemical
methods − An instrument for the evaluation of compatibility between
sphinganine and fatty acids. Rom. Biotechnol. Lett. 2011, 16, 6841−
6846.
(42) Wing, D. R.; Garner, B.; Hunnam, V.; Reinkensmeier, G.;
Andersson, U.; Harvey, D. J.; Dwek, R. A.; Platt, F. M.; Butters, T. D.
High-performance liquid chromatography analysis of ganglioside
carbohydrates at the picomole level after ceramide glycanase digestion
and fluorescent labeling with 2-aminobenzamide. Anal. Biochem. 2001,
298, 207−17.
(43) Zhang, W. K.; Xu, J. K.; Zhang, X. Q.; Yao, X. S.; Ye, W. C.
Sphingolipids with neuritogenic activity from Euphorbia sororia. Chem.
Phys. Lipids 2007, 148, 77−83.
(44) Wang, R. F.; Liu, R. N.; Zhang, T.; Wu, T. A new natural
ceramide from Trollius chinensis Bunge. Molecules 2010, 15, 7467−71.
1517
DOI: 10.1021/acs.jafc.6b04275
J. Agric. Food Chem. 2017, 65, 1507−1517