Archives of Biochemistry and Biophysics 589 (2016) 27e37

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / ya b b i

Review article

The Caenorhabditis elegans lipidome

A primer for lipid analysis in Caenorhabditis elegans
Michael Witting a, *, Philippe Schmitt-Kopplin a, b
a

Research Unit Analytical BioGeoChemistry, Helmholtz Zentrum Mnchen, German Research Center for Environmental Health GmbH, Ingolstadter
Landstrae 1, 85764 Neuherberg, Germany
b
Lehrstuhl fr Analytische Lebensmittelchemie, Technische Universtitat Mnchen, Wissenschaftszentrum Weihenstephan, Alte Akademie 10, 85354
Freising-Weihenstephan, Germany

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 14 April 2015
Received in revised form
2 June 2015
Available online 10 June 2015

Lipids play important roles in biology, ranging from building blocks of membranes to signaling lipids. The
nematode and model organism Caenorhabditis elegans has been used to explore lipid metabolism and
several techniques for their analysis have been employed. These techniques include different possibilities
ranging from visualization of lipid droplets, analysis of total fatty acids to analysis of complex lipids
using lipidomics approaches. Lipidomics evolved from metabolomics, the latest off-spring of the
omics- technologies and aims to characterize the lipid content of a given organism or system.
Although being an extensively studied model organism, only a few applications of lipidomics to C.
elegans have been re- ported to far, but the number is steadily increasing with more applications
expected in the near future. This review gives an overview on the C. elegans lipidome, lipid classes it
contains and ways to analyze them. It serves as primer for scientists interested in studying lipids in
this model organism and list methods used so far and what information can be derived from them.
Lastly, challenges and future (methodological) research directions, together with new methods
potentially useful for C. elegans lipid
research are discussed.
2015 Elsevier Inc. All rights reserved.

Keywords: Lipidomics
Caenorhabditis elegans
Lipid analysis
Model organism

Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
Lipidomics, the comprehensive analysis of lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Caenorhabditis elegans, a versatile model organism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
The C. elegans lipidome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
Fatty acids and amides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
Glycerophospholipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Sphingolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
31
Glycerolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Steroids and related substances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Glycolipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Prenol lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Others classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Applied lipids analysis methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Tools for imaging of lipids and lipid deposits in C. elegans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Lipid extraction and fractionation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Mass spectrometry based methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Challenges and future research directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
Dispatch the lack of C. elegans lipid standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

* Corresponding author.
E-mail address: michael.witting@helmholtz-muenchen.de (M. Witting).
http://dx.doi.org/10.1016/j.abb.2015.06.0 03
0003-9861/ 2015 Elsevier Inc. All rights reserved.

28

M. Witting, P. Schmitt-Kopplin / Archives of Biochemistry and Biophysics 589 (2016)

27e37

MALDI imaging to understand spatial distribution of lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Knowledge transfer and standardization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36

Introduction
Lipidomics, the comprehensive analysis of lipids
Lipids play essential roles cross all kingdoms in biology and
their analysis in health and disease has a long tradition. In the
post- genomic era more emphasis is put on functional analysis of
genes and genomes using tools of functional genomics:
transcriptomics, proteomics and metabolomics, all of them
studying a different subpart of a cells or organisms interior.
Because lipids play such an important role as building blocks of
membranes, energy storage or second messengers and signal
transducer, the comprehensive analysis of lipids evolved from
metabolomics as individual disci- pline called lipidomics.
In contrast to general assumption that lipids represent a rather
homogenous class, several tens of thousands different lipid structures are possible [1]. Lipids are present in a large diversity, from
ubiquitous bulk phospholipids in the membranes to highly specic
signaling lipids and span several orders of magnitude in concentration. Additionally, different lipid classes have large differences in
polarity (e.g.
PC(18:0/16:0) has
a log P of 11.94 and
TG(16:0/16:0/
16:0) a log P of 22.08) or molecular mass (e.g. palmitic acid:
255.231857,
[M H]-compared
to
CL(10 -[18:2(9Z,12Z)/
18:2(9Z,12Z)],30 -[18:2(9Z,12Z)/18:2(9Z,12Z)]):
1447.964950,
[M H]-), which dramatically complicates their analysis.
Today, lipidomics is mainly based on Mass Spectrometry (MS),
but also Nuclear Magnetic Resonance (NMR) has been employed
[2]. Two different kinds of MS analysis can be differentiated: (i)
shotgun lipidomics directly infuses raw lipid extracts into the
MS and uses combinations of neutral loss and product ion
scans to differentiate and identify lipids and is mostly carried out
on ion trap
MS offering MSn capabilities, but also Quadropole-Time of Fligth
(Q-ToF) instruments are employed. Major advantage of this technology is its high-throughput possibility. However, missing
sepa- ration of isomeric lipids is the major disadvantage of this
approach. (ii) Lipid proling using Liquid ChromatographyeMass
Spectrom- etry (LCeMS) is the second most used technique so far.
The use of chromatographic separation enables resolving of
isomers including cis/trans isomers [3]. In most cases, reversed
phase (RP) separations using C8 or
C18 columns with an
Acetonitrile-iso-Propanol (ACN- iPrOH) gradient are employed
[3e5], but also
application of HILIC and
normal
phase
separation is common [6,7]. Lipidomics is applied in any eld of
biological research ranging from cell cultures [8], microorganisms
[9e11], plants [12,13] and mammals [14].
Caenorhabditis elegans, a versatile model organism
The soil-dwelling nematode C. elegans has become a major
model organism in biology. It offers several experimental advantages, including a fast reproductive cycle, a translucent body,
known cell lineage and a sequenced genome, which was the rst
of a
multicellular organism
[15].
Furthermore,
its
hermaphroditic nature allows raising a large number of isogenic
animals in short time.

The full developmental cycle of C. elegans from eggs to fertile

adults takes about 3 days at 20
four

C. The worm develops through

M. Witting, P. Schmitt-Kopplin / Archives of Biochemistry and Biophysics 589 (2016)

27e37
larval stages, named L1 to L4. Under unfavorable
conditions, like
overcrowding or food scarcity it can enter an alternative, nonfeeding and long-lived, developmental stage, the dauer stage
(dauer, german for enduring). After conditions ameliorate dauer
larvae normally develop through L4 larvae to adults without
compromises in adult life span.
C. elegans is used as model organism in different research elds
(e.g. toxicology, developmental biology, neurobiology, hostpathogen interactions, aging research and others). A large array of
tools was developed over the past years involving automated
worm-sorting, micro arrays for transcriptomics, imaging technologies, transgenics or tissue specic gene inactivation [16]. Further
details on the biology of the worm can be found in the review by
Hulme and Whitesides [17].
Although extensively studied, metabolomics and lipidomics
investigations on C. elegans are scarce, but number of publications
related to this topic is increasing steadily. Within in this review
we aim to give a general overview on lipids present C. elegans.
This information was inferred from genetic information on lipid
meta- bolism readily available since many years and results from
different publication. From this we deduced lipid classes and
species that might be present in the worm. Furthermore, we
introduce methods for their analysis and how they have been used
so for analysis of the C. elegans lipid content. Finally, challenges and
needs for successful, future lipidomic investigations are discussed.
This review serves as primer for scientists interested in
analyzing lipids in the model organism C. elegans.

The C. elegans lipidome

C. elegans is able to produce a wealth of different lipids. In 2013
Zhang et al. collected all lipid metabolic genes present in the
worm by comparative genomics. In total they found 471 lipid
genes curated from KEGG, literature research and orthologues of
human lipid genes. 237 of 471 of these genes are conserved in
humans, mice, rats or Drosophila. More specically 71 of these
conserved genes are related to human metabolic diseases. Using
the OMIM database most genes were annotated by
the
keywords Obesity, Diabetes, Metabolic disease and others.
Furthermore, 327 genes in C. elegans are orthologues to human
disease genes, which implies that lipid genes are also involved in
diseases different from meta- bolic syndrome. This collection was
complemented by a functional study using RNA interference
(RNAi), which revealed several phe- notypes, including growth
and developmental defects also for genes not reported so far
[18]. One particular example how such genes relate to human
disease genes was published by Hashmi et al., who used the
worms intestine as model system to study the role of Krppel-like
factors in fat regulation, cell death and phagocytosis [19].
Additional the role
of these factors in human metabolic
regulation was reviewed and KLF-3 for example acts selectively
on insulin components [20].
Although of impressive nature, this work and other genetic
studies gives no information on the actual chemical diversity of the
lipid content, which is very strongly condition dependent (e.g.
availability of building blocks like fatty acids). In this part we present different lipid classes found in C. elegans, their biosynthetic

29

pathways and selected biological actions. Table 1 summarizes

classes present in the worm along with their commonly used abbreviations and LipidMaps identiers, while Fig. 1 shows their
basic structure examples of them.
Fatty acids and amides
Except for steroids, fatty acids are one of the major building
blocks of complex lipids. C. elegans is able to synthesize a broad
variety of fatty acids on its own, including saturated (FA), monoand polyunsaturated (MUFA and PUFA) and mono methyl
branched fatty acids (mmBCFA). Straight chain fatty acids are
synthesized from acetyl-CoA as primer by condensation with
malonyl-CoA in contrast to mmBCFAs, which are produced from
branched chain primers derived from catabolism of branched
chain amino acids. Fatty acids biosynthesis has been extensively
studied and several elongases and desaturases and their
regulation have been
described [21,22]. A 13C isotope labeling approach demonstrated
which fatty acids are directly taken up from the worms diet and
which are synthesized de novo. The two cyclo propane fatty acid
C17D and C19D are exclusively taken up from bacterial diet and
C16:1n7 and C18:1n7 are synthesized de novo in minor amounts
of about 5%, whereas C16:0,
C18:0,
C18:1n9, C18:2n6 are
produced in increasing amounts ranging from 7.2% to 19% and
mmBCFAs are exclusively synthesized de novo [23,24].
Fatty acids not only serve as building blocks for higher lipids,
but have rather important biological roles. For example g-linolenic
and stearidonic acid are required for basal immunity during
infection with Pseudomonas aeruginosa [25], while oleic acid and
NHR-80/

HNF4 are needed for longevity in germline ablated worms [26].

The mmBCFA C17:0iso is essential for postembryonic growth of L1
larvae [27]. Additionally, C17:0iso and ACS-1 are shown to be
needed for inositol triphosphate (IP3) signaling for early embryonal
development [28].
Furthermore, it has been demonstrated that PUFAs are important for correct neurotransmission in C. elegans. Watts et al.
encountered non-normal behavior of fat-3 mutants during their
studies of fatty acid metabolism [29]. Depletion of PUFAs in fat-3
mutants leads to movement deciencies, inability to respond to
head-touch and egg laying defects similar to egl-1 mutants. Defects
in motility were rescued by arachidonic acid (C20:4n-6) and
docosahexaeonic acid (C22:6n-3), similarly expression of fat-3
under control of the neuronal promotor unc-119 could also
rescued the egl phenotype. Further experiments lead to the
conclusion that mutation of fat-3 leads to functional but not
developmental defects in the nervous system and insufcient
neurotransmitter release [30].
Composition of fatty acids also correlates with longevity. Fatty
acid proles from different long lived mutants were measured and
by correlating relative longevity with different parameters (e.g.
chain length or degree of unsaturation) different trends became
apparent. First, the amount of MUFAs increases, while PUFAs
decrease with longevity. In parallel a trend towards shorter chain
length is observed. Results were also in accordance with susceptibility to hydrogen peroxide [31].
One lipid class directly derived from fatty acids are fatty amides,
represented in C. elegans by N-acylethanolamides (NEA). This lipid
class includes the mammalian endocannabinoid anandamide and is

Table 1
Lipid classes present in C. elegans inferred from genetic information.
Lipid class

Subclass

Abrrev.

LM ID

Sphingolipids

Sphingosine
Sphiganine (Dihydrosphingosine)
Sphingosine 1-phosphate
Sphinganine 1-phospate
N-Acylsphingosine (Ceramide)
N-Acylsphinganine (Dihydroceramide)
Ceramide 1-phosphate
Sphingomyelin
Glucosylceramide
Lactoyslceramide

SPH

SM
GlcCer
LacCer

SP0101
SP0102
SP0105
SP0105
SP0201
SP0202
SP0205
SP0301
SP0501
SP0501

1-Acyl-sn-glycerol
1,2-Diacyl-sn-glycerol
Triacylglycerol

MG
DG
DG

GL01
GL02
GL03

Phosphatidic acid
Phosphatidylcholine
Phosphatidylserine
Phosphatidylethanolamine
Phosphatidylmonomethylethanolamine
Phosphatidyldimethylethanolamine
Phosphatidylinositol
Cardiolipin

PA
PC
PS
PE
MMPE
NMPE
PI
CL

GP10
GP01
GP03
GP02
e e
GP06
GP12

Steroids and related substances

Steroids
Dafachronic acids
Cholesterylester

ST
DA
CE

ST0101
ST0403
ST0102

Glycolipids

Maradolipids

Mar

Fatty acids

Straight chain fatty acids

Branched chain fatty acids
Unsaturated fatty acids

FA
FA
FA

FA0101
FA0102
FA0103

Fatty amides
Prenol lipids

Fatty amides
Isoprenoids
Quinonnes

FA-EA
PR
PR

FA0804
PR01
PR02

Others

Ascarosides

Ascr

Glycerolipids

Glycerophospholipids

Depicted class also includes fatty acids bound as alkyl or alkenyl.

S1P
Cer
CerP

Fig. 1. Structures of lipid classes present in C. elegans (A) sphingolipids usually contain a C17iso sphingoid base. Different derivatives are possible, including S1P, Cer, CerP, SM,
GlcCer and LacCer, (B) Glycerolipids are the major energy storage in C. elegans. MG and DG also serve as precursor for other lipids. (C) Glycerophospholipids are the major lipid
class found as building blocks of biological membranes in high amounts. C. elegans is capable of synthesizing a wide range with different headgroups, e.g. PA, PC, PS, PE, MMPE,
DMPE and PI. (D) The bile acid like dafachronic acids (D7-DA shown as example) are important signaling molecules in development and nematodal longevity. (E) Maradolipids are
special lipids found in dauer larvae, only recently discovered (F) Fatty acids are used as building blocks of complex lipids, but also serve as signaling molecules, while fatty
amides are important signaling molecules, e.g. linking food to lifespan.

synthesized by acylation of the ethanolamine nitrogen of PEs followed by cleavage of the phosphate ester bond yielding a NEA
and phosphatidic acid. Eicosapentanoyl ethanolamide is reduced
upon starvation or dietary restriction and mediates diet effects
to
life- span [32].
In
a
metabolomics approach
oleolyethanolamide was identied to be increased in long lived
worms over expressing lipl4, and binds to the proteins LBP-8 and NHR-80, activates target
genes of NHR-80 and NHR-49 and promotes longevity [33].
Glycerophospholipids
Glycerophospholipids, as building blocks of membranes,
represent the largest portion of the lipidome. Chemically they are
built from two fatty acids esteried to the sn1 and sn2 positions
of glycerol and different head groups attached to sn3 position.
Addi- tionally, side chains at the sn1 position can be bound by
either an acyl, alkyl or alkenyl bond. Generally fatty acids bound
on the sn1 position have a low degree of unsaturation, while
different fatty acids can be found on the sn2 position. From
genetic analysis C. elegans is able to synthesize and use a high
diversity of different

head groups, resulting in many different lipid

classes:
phosphatidic acids (PA), phosphatidyl cholines (PC), phosphatidyl
ethanolamines (PE), phosphatidyl serines (PS), phosphatidyl
glycerols and glycerol phosphates (PG and PGP), and phosphatidyl
inositols (PI). Furthermore, cardiolipins (CL) are
found in
mitochondrial mem- branes and N-mono- and di-methyl
phosphatidyl ethanolamines (MMPE and DMPE) are intermediates
in the synthesis of PCs from PEs present in certain mutants or
knock-downs [34]. For the three major glycerophospholipid
classes (PC, PE and PS) biosynthesis pathways are strongly
connected (Fig. 2B). PCs can be synthesized on different routes.
First, the Kennedy pathway (blue in Fig. 2B) uses dietary
choline or phosphoryl choline derived from the phosphobase
methylation pathway [34,35] (red in Fig. 2B) and its activated
form CDP choline for coupling to diacylglycerols (DG). There is
some evidence for the existence of the BremereGreenberg
pathway in C. elegans [36]. PS synthesized from PC by exchange of
the choline head group with serine catalyzed by PSSY-1 [37] and
PE is derived from PS by decarboxylation. As major building blocks
of membranes,
glycerophospholipid
strongly
inuence
mechanical properties of cells
and membrane uidity. In
the previous

Fig. 2. Examples of different lipid synthesis pathway present in C. elegans. Genes encoding lipid enzymes are depicted in red next to the reaction the products catalyze. (A) A vast
amount of different fatty acids with different chain length and degree of saturation can be produced by C. elegans on its own, although it utilizes fatty acids derived from bacterial
diet directly. Exceptions are mmBCFA, which are almost exclusively synthesized by the worm itself. (B) Glycerophospholipids are the major building blocks of biological
membranes and the synthesis of different classes is strongly interwoven. The shaded orange region shows the phosphobase methylation pathway producing phosphoryl choline
from serine. The Kennedy pathway (blue region) directly utilizes dietary choline for synthesis of PC. Lastly, PS is produced from PC by exchanging the headgroups. Some evidence
for existence of the BremereGreenberg pathway (green region) has been found in C. elegans.

paragraph PUFAs were linked to correct neurotransmission. Newer

work shows that PUFAs bound to specic glycerophospholipids
are needed for neuronal cell mechanics and touch sensation. fat-3,
fat4, fat-1 and elo-1 had reduced touch response compared to wild
type worms. Mutants were supplemented with either arachidonic
acid (C20:4n-6), eicosopentaenoic acid (C20:5n-3) or both. Interestingly, only when both were supplied mutants behaved like the
wild type and in the next step it was evaluated if PUFAs bound to
phospholipids or in another form are needed for phenotypic
rescue. The enzymes MBOA-7 and MBOA-6 are responsible for
integrating PUFAs into PIs or PCs, PEs and PSs. mboa-7 mutants
and mboa-

6(RNAi) (knockout of mboa-6 is lethal) were fed either with

PE(18:0/20:4), PC(18:0/20:4) or PS(18:0/20:4) or different combinations. Only PE(18:0/20:4) and PS(18:0/20:4) showed signicant
enhancement. Lastly, isolated touch receptor neurons (TRNs) of
wild type and fat-1; fat-4 animals showed different membrane
mechanics [38].
Sphingolipids
Sphingolipids play important roles as both structural units and
signaling molecules. Structurally, they are built from a sphingoid

base, derived from condensation of a fatty acid (usually C16:0)

with serine and N-linked fatty acid. C. elegans on the contrary uses
C15:0 and produces a C17:0iso branched chain sphinganine
base (d17:0iso-SPA). However under certain conditions, like RNAi
of the gene let-767, glucosylceramides also contain C16 and
hydroxy- C17:0 iso sphingoid bases [39]. Interestingly, N-linked
fatty acid side chains contain 2-hydroxy fatty acids, mostly
C22:0-OH [40]. The biosynthesis pathway with all related genes is
shown in Fig. 2C.
Previous work showed that elo-5 loss-of-function mutations
lead to arrest in the early L1 stage. Zhu et al. demonstrated that
this mutation leads to disruption of a novel sphingolipidTORC1 pathway needed for normal development. Several genes
for syn- thesis of glucosyl ceramides (GlcCer) from d17:iso-SPA
were tested and exogenous supplementation with specic
sphingolipids could rescued the phenotype of elo-5( ) and splt1(RNAi), but not of fath1(RNAi) and cgt-1( ), cgt-3( ) (fath-1 encodes for fatty acid 2hydroxylase and cgt-1 and cgt-3 for ceramide glucosyl transferases), which suggests that d17iso-GlcCer is the mediator in
this pathway. Using a genetic screen a loss-of-function mutation of
nprl3 was detected to abolish this effect [41]. These results directly
link mmBCFA to a new signaling pathway. Other publications
demon- strated that ceramides are
required for
anoxia
protection [42], autophagy dependent life span extension [43] or
radiation induced apoptosis [44].
Glycerolipids
Excessive energy in C. elegans is stored in form of lipid
droplets, containing tri acyl glycerols (TGs), which are primarily
located in epidermal and intestinal cells (up to 35% of body dry
mass) and to a lesser extends in form of glycogen (3.3% of body
dry mass). Lipid storage vesicles are not overlapping with
lysosome-related organ- elles [45]. For mobilization of energy
from these resources the worm harbors several lipases, some
also linked to longevity. lipl-4 for example is needed for
induction of autophagy and is also required for life span
extension in germ line ablated animals [46]. Similarly, also
diacylglycerol lipases regulate lifespan [47]. Fat storages are
mobilized during starvation, for example dauer larvae accumulate
fat storages, which are utilized during prolonged star- vation
[48,49]. The full complement of peroxisomal and b-oxida- tion
genes is present in C. elegans for generation of energy from fat
storages, including a functional methylmalonyl-CoA epimerase for
degradation of mmBCFAs [50,51].
Steroids and related substances
The worm is a sterol auxotroph and therefore sterols have to
be supplied exogenously by the diet, whereby different sterols
can serve the needs of C. elegans. Interestingly, the needed
amount of sterols is too low to have structural roles, but is rather
needed for synthesis of steroid signaling molecules [52]. Indeed,
development is regulated by bile acid like steroid hormones
called dafachronic acids [53].
Several different forms of
dafachronic acid have been identied [53,54], but the exact
biosynthesis pathway is still not completely known. Dafachronic
acids not only regulate develop- ment, but are also required for
signaling in longevity. For example the long life of germline
ablated animals involves the presence of the DAF-9/DA/DAF-12
pathway [55e57]. Furthermore, dietary re- striction mediated
longevity also
relies on
dafachronic acid signaling, but
instead of DAF-12 its close homolog NHR-8 is required and
let-363/mTOR is essential for this mediation [58]. Beside this
bile acid like steroid hormones C. elegans contains different

other sterols, where by

the actual content of individual
compounds is dependent on the sterol supplementation [59,60].

Glycolipids
A special class of glycolipids was recently discovered in dauer
larvae of C. elegans: 6,60 -diacyltrehaloses, called maradolipids.
Interestingly most of the detected lipids from this class contained
at least one mmBCFA side chain. Further experiments showed
that these special lipids are part of the dauer body structural
unit, the microvilli in the lumen. Using elo-5 RNAi in a daf-2
background or the daf-2;DDtps mutant microvilli size is reduced
or completely missing [61].
Prenol lipids
Prenol lipids are derived from Acetyl-CoA utilized in the
mevalonate pathway. In contrast to other animals, C. elegans lacks
the cholesterol producing branch of the mevalonate pathway, but
all others are highly conserved. This makes the worm an ideal
model organism for the study of the non-cholesterol producing
branches. Knockdown of any enzyme of the main branch
producing Franesyl-PP from Acetyl-CoA leads to embryonic
lethality. Likewise, inhibition of HMG-CoA reductase by statins also
leads to embryonic lethality [62]. Important functions of prenol
lipids are derivatiza- tion of adenosine in tRNAs for stabilization
of codoneanticodon interactions [63], production of coenzyme Q
(although also
sup- plied by bacterial diet) [64] or protein
modication [65]. For further information on the mevalonate
pathway in C. elegans refer to the review by Rauthan and Pilon
[66].
Others classes
Ascarosides are a special class of lipid derived molecules, which
have different functions, from dauer inducing pheromone (daumone) [67], male attractants [68] to signaling of aggregation [69].
Ascaroside signaling integrates building blocks from fatty acid,
amino acid metabolic and other pathways to form a modular
library of pheromones [70,71]. Peroxisomal
b-oxidation is
important for synthesis of ascarosides and requires daf-22 and
dhs-28 for pro- duction of short chain fatty acids from very long
chain fatty acid precursors. More recently mutation in the acylCoA oxidases acox1, -2 and -3 has been shown to lead to defects in ascaroside
signaling [72]. More information on this special class can be found
in the recent review of Frank Schro der [71].
Applied lipids analysis methods
C. elegans produces a vast amount of lipids covering a large
combinatorial and chemical space and the question rises how this
diversity can be accessed analytically? In the same way as lipids
also methods for their analysis are very diverse, ranging from imaging methods, analysis of single classes (e.g. fatty acid analysis
with GC) or proling of the (complete) lipid content using MS
based technologies. Plenty of work has been carried out on the
level of fatty acids. However, as examples above have shown
not only composition of fatty acids but also their origin or
relation to com- plex lipids is important. Comprehensive
analysis of intact lipids only recently joined the C. elegans
toolbox.
Still, several analytical chemistry methods have been used, on
which the current technology is built on. Major analysis technologies are imaging of lipid deposits in C. elegans using uorescent
dyes or other technologies, analysis of lipid fractions with TLC and
analysis of fatty acid composition with GC and GCeMS. However,
the two new MS based technologies, shotgun lipidomics and
LCeMS based lipid proling, are catching up. Depending on the
nature of the biological question that needs to be answered
different approaches or technologies might be selected. The

following paragraphs gives an overview on methods employed

with C. elegans so far and compares their advantages and disadvantages. Fig. 3 shows a ow chart with different possibilities for
analysis of lipids and Table 2 compares their advantages and
disadvantages.
Tools for imaging of lipids and lipid deposits in C. elegans
Classically, lipid deposits are labeled with uorescent dyes that
concentrate in hydrophobic environments. The phenoxazone dye
Nile red or BODIPY labeled fatty acids are used for imaging of lipid
deposits in cell lines or C. elegans, by mixing with the worms diet
[73,74]. However, broader analysis indicated that these dyes do
not label major fat storages in C. elegans, instead xation and
staining with oil red O correlated well with biochemical analysis
of lipid content [45]. Imaging in live animals is a useful tool to
follow in- dividuals over time. Choline containing metabolites and
lipid were imaged in live C. elegans using Stimulated Raman
Scattering (SRS) together with isotope-based metabolic labeling,
that shifted the
NeH (NeD) bond vibration to 2100 cm 1 in the cell silent Raman
window [75]. Similarly, alkyne-tagged small biomolecules have
been used in conjunction with SRS for live-cell imaging, including
C. elegans. 5-Ethynyl-20 -deoxyuridine, 5-Ethynyl uridine, LHomo- propargylglycine, Propargylcholine and 17-Octadecynoic
acid were used to follow de novo synthesis of DNA, RNA, proteins,
phospho- lipids or triacylglycerols respectively [76]. Another
example made use of SRS together with phenyl-diyne cholesterol,
which is stable and permits specic detection, to investigate
cholesterol storage in
wild type worms and chup-1 mutants, which are decient in
cholesterol uptake [77]. Though for these methods labels are
needed, whereas coherent anti-Stokes Raman scattering (CARS) is
not dependent on prior labeling. CARS was compared with
different fat labeling methods under different conditions, which
revealed that CARS is the method of choice, if available [78].
Several other studies reported to use of CARS as label-free
approach to follow

lipid metabolism in C. elegans [79e81].

However, all these methods give only very unspecic information on total lipid classes, but not on individual lipid species. MS
based imaging is interesting alternative for metabolite and lipid
imaging directly in tissues without the necessity of prior labeling.
Matrix Assisted Laser Desorption IonizationeMass Spectrometry
(MALDIeMS) and Time of Flight-Secondary Ion Mass Spectrometry
(ToF-SIMS) based techniques have been used with C. elegans in two
proof-of-principle publications [82,83]. For routine application of
this technology several developments have to be made, especially
spatial resolution in MALDIeMS has to be in an acceptable range to
visualize different tissues (see also Challenges and Future research
directions).
Lipid extraction and fractionation
If individual lipids or lipid classes have to be analyzed they need
to be extracted from a C. elegans sample. The two most employed
lipid extraction methods are the Folch [84] and Bligh and Dyer
protocols [85]. Both extraction protocols are based on chloroform/
methanol, yet often replaced with protocols using MTBE [86,87],
due to the carcinogenic nature of chloroform and environmental
concerns. Comparison of extraction protocols, in which also
C. elegans eggs were included as one particular sample, exhibited
that the new method has similar yields. In contrast to chloroform,
MTBE based extractions build a two phase system with the
organic solvent on the top and cell debris on the bottom
particularly useful for robotic automation [86].
Lipid composition of C. elegans can be assessed on class level in
the simplest way by the use of thin layer chromatography (TLC).
This method is easy to use and no specialized equipment is needed
and can be performed analytically or preparative for isolation of
specic lipid classes. Quantication is achieved by general or specic staining (e.g. glycolipids [61] or phospholipids [88]) and
densitometry. Use of 2D-TLC, adding a second separation

Fig. 3. Different possibilities for analysis of lipids and lipid deposits in C. elegans have been described. This ow chart shows all major techniques and approaches with example
literature. Different combinations of steps can be employed to reach an optimal result, e.g. extraction, fraction and measurement of a specic class or fraction.

Table 2
Summary of advantages, disadvantages and information content of different technologies for analysis of C. elegans lipids.
Method
Imaging (e.g.
Fluorescence labeling,
SRS CARS)

Advantages

Disadvantages

Direct analysis
Non-invasive,

live

imaging

Information content

No information on lipid composition

Localization and amount of labeled and/or detectable

For analysis of specic lipids or tracers with

lipids

possible (CARS, SRS)

tags are needed

TLC

No special equipment needed

Fast, easy and cheap

Only limited information

GCeFID/GCeMS

High resolution separation

isomers
Absolute
quantication
different fatty acids

of

Amount of lipid classes

Only amenable to volatile lipids or lipids that

Fatty acid composition

can be made volatile by derivatization

Amount of isomeric fatty acids

(absolute) quantities of different lipids

of

Shotgun lipidomics

Fast, high-throughput
capable
(analysis time of several minutes)
Absolute quantication possible

No separation of isomers

LCeMS

Separation of isomeric molecules

Absolute quantication possible

Long analysis time, typically between 15 to

Absolute or relative quantities of different lipids, also

25 min

for (chromatographically) separable isomers

Ultrahigh
resolution
allows
separation of isobaric molecules
Fast and high throughput capable

No

DIeHReMS

differentiation

molecules

dimension, increases resolution of this technique [61]. Solid phase

extraction (SPE) represents an alternative for lipid fractionation
and is mostly based on NH2 phases for separation of different
lipid classes [88]. Also more advance materials for isolation of
specic lipid classes (e.g. S1P) are available [89].
Information on lipid composition from these methods is
limited, but they are useful for enrichment of specic lipid classes
and can be combined with more sensitive and specialized
methods, like GCeMS or shotgun lipidomics.
Mass spectrometry based methods
Mass spectrometry is a powerful tool to decipher between
different lipid species and allows precise relative and/or absolute
quantication. It can be either used without (shotgun) or with
chromatographic separation. Based on the nature of chromatography different lipids can be analyzed.
GC and GCeMS is the most employed techniques for fatty acid
analysis in C. elegans. Fatty acids are transferred to volatile methyl
esters by esterication of free or trans-esterication of bound fatty
acids. MS can be used in conjunction with isotopic labeling to
determine rate of de novo synthesis [23]. GCeMS is often
employed in conjunction with a prior fractionation using either
TLC or SPE to assess the fatty acid composition of specic
fractions. Despite, in- formation on
specic fatty acid
combinations in complex lipids is lost. Results are usually
expressed as percentages of detected fatty acids. GC allows high
resolution separations, including baseline separation of iso and
ante-iso fatty acids or different double bond isomers. A major
problem when analyzing fatty acids with GCeMS is in source
fragmentation of PUFAs under EI conditions. Using chemical
ionization can overcome this and also allows locating double
bonds in PUFAs [90,91]. A novel LC-MS based method might
replace this approach in future, offering higher sensitivity using a
charge reversal derivatization of free fatty acids with N-(4aminomethylphenyl)pyridinium, although not applied to
C. elegans yet [92].
The analysis of the fatty acid composition of intact lipids is
getting more important, because there is only weak correlation
between bound and free fatty acids. Liquid based techniques offer
the possibility to introduce intact lipids into the MS. Two different
approaches can be used. First, shotgun lipidomics directly infuses
the raw lipid extract without prior chromatographic separation
and analysis is performed on
tandem mass spectrometers
utilizing

between

isomeric

Relative quantication of isobaric lipids

Exact mass and molecular formula for unknown
lipids

multiple neutral losses and product ion scans to determine lipid

composition of different signals. Up to know this is the mostly
employed analysis method for C. elegans lipidomes.
Shotgun lipidomics in conjunction with multiple precursor and
neutral loss scans and data dependent acquisition (DDA) was used
for the analysis of the C. elegans lipidome. It was demonstrated
that using this technique accurate identication and quantication
can be performed in one single direct infusion experiment. The
experiment was carried out on a Q-ToF experiment equipped with
an automated nanospray chip ESI source (NanoMate) allowing long
spraying times with a minimal consumption of sample extract.
Phospholipids were extracted according to Bligh and Dyer and
TAGs were isolated by 2D-TLC as described by Matyash et al. [93].
Limits of quantication (LOQ) in the lower nM range were
achieved. In total 90 glycerophospholipids and 35 TGs with
unique molecular formulae were detected in a single experiment.
MS/MS data indi- cated that behind each individual molecular
composition of one TG mass, 5 to 15 isobaric species can be found.
Using systems of linear equations relative abundances of each
specic isobar could be calculated [94]. The same group used
multivariate analysis of high- resolution survey MS scans for
lipidomics screening on an Orbitrap instrument similarly
equipped with a NanoMate for long infusion times. For proof of
concept RNAi of two genes, encoding putative methyl
transferases, was conducted. Analysis showed presence of MMPE
and DMPE in the corresponding extracts, proong methyl
transferase activity of the two genes called pmt-1 and pmt-2 [34].
Dedicated software tools in combination with shotgun lipidomics enables screening of novel lipids. The software LipidXplorer with its Molecular Fragmentation Query Language
(MFQL) was used to specically search for maradolipids in extracts
from C. elegans dauer larvae in comparison to L3 larvae. Analysis
revealed a novel class of lipids called lyso-maradolipids having only
one fatty acid moiety specically enriched in dauer larvae.
Furthermore relative abundances of fatty acid side chains could be
calculated for marado and lyso-maradolipids [95]. Due to missing
chromatography direct infusion experiments
benet from
employing high-resolution instrumentation. Ishida et al. performed
direct infusion experiments using the NanoMate chip ESI source
coupled to a Bruker APEX III ICR-FT/MS for analysis of C. elegans
PEs. High-resolution of the employed MS differentiated between
diacyl- and alkyl-acyl PEs solely on mass [96]. DI-ICR-FT/MS
hold great promises for high throughput deep metabotyping [97]
and will be potentially applied in future C. elegans lipidomics
investigations.

Several other studies likewise employed shotgun lipidomics

[28,41e43].
The major drawback of shotgun lipidomics is the missing differentiation between isomeric lipids, for which reason chromatographic separation is needed. Traditionally normal-phase
separations using silica columns and non-polar solvents for elution
(e.g. iPrOH, hexane, chloroform, ethyl acetate or mixtures) have
been used for analysis of lipid classes and was also employed to
C. elegans [28]. However it requires dedicated equipment free of
water due to immiscibility of different organic solvents with
water, which would lead bad
chromatographic performance,
unstable baselines and other problems. Hydrophilic Interaction
Liquid Chromatography (HILIC) is an
interesting alternative
approach us- ing water miscible organic solvents for lipid class
separation [6,7].
The predominant method in lipidomics nowadays is based on
RP separation and an ACN-iPrOH gradient, which allows detection
of a broad range of different lipid classes and species within a
single run [4]. Castro et al. for example studied the impact of
daf-2 mu- tation on the metabolome and lipidome. For lipid
analysis GCeMS for total fatty acids and LC-MS for intact lipids
was utilized. LC-MS analysis was performed on a C8 column
with an ACN-iPrOH gradient. Correlation analysis between
total fatty acids and LCeMS data revealed a central role for
C20:2
in
lipid
remodeling between triacylglycerols and
phospholipids [98]. We have recently described a UPLCeMS based
method for comprehensive analysis of the C. elegans lipidome
using a sub-2 mm core shell particle C18 column. This method
shows very high reproducibility as evaluated with lipid standard
materials and C. elegans lipid extract, even be- tween different
column batches and days. Fig. 4 shows a typical chromatogram
derived from this method. Use of DDA allows collection of MS/MS
spectra of different lipids on a large scale already during
proling runs [5]. Furthermore, an automated approach for
analysis of this large scale MS/MS data collection was developed.
This novel approach, called LipidFrag, is based on the in silico
fragmentation engine MetFrag. A classier model based on true
positive and false positive annotations of neutral precursor

masses from MS/MS spectra of lipid standards was used to evaluate

performance and calculation of reliability of in silico fragmentation
results. Using this novel workow ambiguous result can be ltered
out [Witting et al., unpublished].
Challenges and future research directions
Although only a limited number of investigations on comprehensive lipid analysis in C. elegans are published so far, much has
been achieved. The worm as model system allows easy setup of
experiments related to fat and lipid metabolism and storage and a
lot of knowledge has been collected over the last years. Still,
several lipid classes lack comprehensive measurements at the
moment. Sphingosine kinases and lyases play important roles as
shown by different studies [99], but comprehensive and reliable
detection of their educts and products is still not achieved.
Menuz et al., were able to detect C17iso sphingosine-1-phosphate
and also found a 10- fold
increase in sphk-1 mutants [42].
Potentially novel sample preparation methods can overcome
this limitation, which also enabled detection new forms of S1P in
Drosophila [89]. Lastly, direct detection and identication of intact
lipids using either shotgun or LCeMS based lipidomics will
replace long isolation, derivatization and measurement of fatty
acids and will lead to new insights. However, several points
have to be raised and should be addressed for successful future
investigations.
Dispatch the lack of C. elegans lipid standards
The Metabolomics Standard Initiative (MSI) dened several
levels of condence for identication of metabolites and lipids,
whereby the highest level (level 1) can be only achieved by comparison with an authentic standard under the same analytical
conditions [100]. Such standards are not commercially available
for certain lipids present in C. elegans. Especially for the
maradolipids and several dafachronic acids no
commercial
standards are avail- able. Though, synthesis protocols for both
have been described

Fig. 4. Example chromatogram derived from a C. elegans lipid extract analyzed with the method from Witting et al. in positive ionization mode [5]. Three different clusters of lipid
features can be detected. Detected classed referred in gure are independent of ionization mode.

[54,101e103]. Interestingly, ceramides with 2-OH-fatty acid

are available, but not with C17:0iso base. However, such
standards would be needed for denite identication and
comparison with ceramides present in C. elegans. Synthesis of
chemical reference standards for
each individual lipid
is
unrealistic and not feasible, but representative candidates from
each lipid class are needed at minimum. Normally, lipids from
the same class exhibit similar fragmentation spectra. Therefore
one or several standards can serve as proxy for identication.
Far more important is the need for suitable internal standards
for normalization and quantication. Because C. elegans is capable
of producing and using odd chain fatty acids, lipids normally
applied as internals standards like PC(17:0/17:0) are not suitable.
Isotopes labeled analogs are a possible solution (e.g. lipids with
deuterated side chains or 13C labeled side chains) to this issue, but
are
either not available or very
expensive.
MALDI imaging to understand spatial distribution of lipids
The nematode represents a multicellular organism with differentiated tissues. These tissues vary in their lipid content and
composition. The intestine is the major storage of lipid droplets
containing TGs. Similarly composition of phospholipids may vary
between tissues. Several imaging tools have been described to
locate and analyze lipid deposits, e.g. based on uorescent dyes or
CARS. These tools give no information on lipid composition, for
example different fatty acid side chains or their combination. So far
only phosphatidylcholine has been imaged using isotope labeling
in combination with SRS [75].
MALDIeMS based imaging is gaining popularity as alternative
tool to access lipid composition and spatial information in a single
experiment. It is widely used in tissue analysis of proteins as
alternative to classical histology. Application of metabolite imaging
using MS in increasing with several areas of application [104],
including lipid analysis [105]. MALDIeMS imaging to localize and
visualize molecular information in the whole worm has been also
described [82]. However, spatial resolution of available
instruments is still insufcient for use with C. elegans. First
results in our lab showed that metabolite imaging is theoretically
possible, but needs improvements in sample preparation and
spatial resolution (un- published). ToF-SIMS is an interesting
alternative approach offering very high spatial resolution. An
initial study used ToF-SIMS for analysis of C. elegans extracts,
but also
showed potential applica- bility in
imaging of
biomolecules [83]. Future developments to- wards higher spatial
resolutions and increased sensitivity in MALDIeMS will open new
elds in the analysis of C. elegans lipids.
Knowledge transfer and standardization
Currently no centralized database for C. elegans lipid exists. The
major database for lipidomics investigations is LipidMaps, containing near
40,000 unique
lipid
structures (status of
18.10.14) [106]. Still, major lipids present in C. elegans, e.g.
ceramides with C17:0iso sphingoid bases or maradolipids are
missing in
this database. WormBase is the current gold
standard for C. elegans biology and integration of metabolites,
lipids and metabolic and lipid pathways would greatly facilitate
dispersion of knowledge. One step in this direction was made
with the creation of the SMID- DB, storing information on C.
elegans signaling molecules [107].
Current efforts in our group are directed to generate a
comprehensive in silico library of lipids potentially present in
C. elegans, which can be validated and supplemented experimentally. A rst step toward integration of C. elegans lipid data was
done by the LipidMaps consortium by adding lipid related
genes and proteins of C. elegans to the database.

To make results from future lipidomics experiments in the

worm comparable a standard lipid extract would be helpful.
Similar efforts have been carried out to create a NIST standard
Metabolites in Human Plasma [108]. Such a standard lipid extract
could be run as additional quality control to validate results and
allows com- parison of different experiments. Potentially such a
standard ma- terial can be produced for the different life stages
and amount of specic lipids can be quantied and stored in an
above mentioned database following the example of the Human
Metabolome Data- base [109,110].
Conclusion
The nematode C. elegans is capable of synthesizing a large bunch
of different lipids from fatty acids to glycerol- and glycerophospholipids, steroid hormones, glycolipids and many more.
Different analytical methods have been employed to access this
large chemical diversity. C. elegans is a proven model organism for
lipid metabolism research. Lipids play a central role in the biology
of the nematode, but also in other model organisms and are often
linked intimately linked to reproduction and life span [111]. As
shown lipids, e.g. ceramides in a novel TORC pathway, can serve as
important signaling molecules. One might ask now based on this
observation: How specic is the activity of this pathway? What
lengths of the N-linked fatty acids are possible and do all of them
act in the pathway or are just a few of the active and all other are
produced transiently? The same questions can be asked for
different lipid classes. With the advent of the descriptive omics
technologies, including lipidomics these questions will be potentially answered. Furthermore, more and more lipid standards will
become available allowing direct activity testing of them and facilitates lipid identication. A major advantage of C. elegans as
model organism is that hypothesis can be tested in a straight forward manner with readily available mutants or feeding assays.
Using RNAi, specic genes can be knocked down at specic time
points without genetic mutation (knock-out), which makes it also
possible to study function of genes of which a knock-out would be
lethal during development.
These possibilities have to be combined with novel tools and
approaches already available or to be developed to fully exploit the
potential of C. elegans. However, rst steps into this direction are
made.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]

G. van Meer, EMBO J. 24 (18) (2005) 3159e3165.

H. Fernando, et al., Alcohol. Clin. Exp. Res. 34 (11) (2010) 1937e1947.
S.S. Bird, et al., Anal. Chem. 84 (13) (2012) 5509e5517.
C. Hu, et al., J. Proteome Res. 7 (11) (2008) 4982e4991.
M. Witting, et al., J. Chromatogr. A 1359 (2014) 91e99.
G. Boncompain, et al., PLoS One 9 (1) (2014) e86196.
M. Hol capek, et al., J. Chromatogr. A 1377 (2015) 85e91.
J. Li, et al., Anal. Chem. 85 (9) (2013) 4651e4657.
M. Jain, et al., Proc. Natl. Acad. Sci. 104 (12) (2007) 5133e5138.
D. Portevin, et al., MicrobiologyOpen 3 (6) (2014) 823e835.
E. Layre, et al., Chem. Biol. 18 (12) (2011) 1537e1549.
M. Zoeller, et al., Plant Physiol. 160 (1) (2012) 365e378.
Y. Okazaki, et al., Metabolomics 9 (1) (2013) 121e131.
M. Jove, et al., Sci. Rep. 3 (2013).
T.C.e.S. Consortium, Science 282 (5396) (1998) 2012e2018.
H. Qadota, et al., Gene 400 (1e2) (2007) 166e173.
S.E. Hulme, G.M. Whitesides, Angew. Chem. Int. Ed. 50 (21) (2011)
4774e4807.
Y. Zhang, et al., BMC Genomics 14 (1) (2013) 164.
S. Hashmi, et al., DNA Cell Biol. 27 (10) (2008) 545e551.
S. Hashmi, et al., Nutr. Metab. 10 (1) (2013) 31.
J.L. Watts, J. Browse, Proc. Natl. Acad. Sci. 99 (9) (2002) 5854e5859.
T.J. Brock, J. Browse, J.L. Watts, PLoS Genet. 2 (7) (2006) e108.
C.L. Perez, M.R. Van Gilst, Cell Metab. 8 (3) (2008) 266e274.
M. Kniazeva, et al., PLoS Biol. 2 (9) (2004) e257.
M. Nandakumar, M.-W. Tan, PLoS Genet. 4 (11) (2008) e1000273.

[26] J. Goudeau, et al., PLoS Biol. 9 (3) (2011) e1000599.

[27] M. Kniazeva, T. Euler, M. Han, Genes Dev. 22 (15) (2008) 2102e2110.
[28] M. Kniazeva, et al., Genes Dev. 26 (6) (2012) 554e566.
[29] J.L. Watts, et al., Genetics 163 (2) (2003) 581e589.
[30] G.M. Lesa, et al., J. Cell Sci. 116 (24) (2003) 4965e4975.
[31] R.J. Shmookler Reis, et al., Aging (Albany NY) 3 (2) (2011) 125e147.
[32] M. Lucanic, et al., Nature 473 (7346) (2011) 226e229.
[33] A. Folick, et al., Science 347 (6217) (2015) 83e86.
[34] D. Schwudke, et al., Anal. Chem. 79 (11) (2007) 4083e4093.
[35] L.H. Palavalli, et al., Biochemistry 45 (19) (2006) 6056e6065.
[36] G. Lochnit, R. Geyer, Acta Biochim. Pol. 50 (4) (2003) 1239e1243.
[37] L. Nilsson, E. Jonsson, S. Tuck, Trafc 12 (12) (2011) 1839e1849.
[38] V. V asquez, et al., Cell Rep. 6 (1) (2014) 70e80.
[39] E.V. Entchev, et al., J. Biol. Chem. 283 (25) (2008) 17550e17560.
[40] D. Chitwood, et al., Lipids 30 (6) (1995) 567e573.
[41] H. Zhu, H. Shen, A.K. Sewell, M. Kniazeva, M. Han, Elife. 2 (2013) e00429.
[42] V. Menuz, et al., Science 324 (5925) (2009) 381e384.
[43] M.-B. Mosbech, et al., PLoS One 8 (7) (2013) e70087.
[44] X. Deng, et al., Science (New York, N.Y.) 322 (5898) (2008) 110e115.
[45] E.J. ORourke, et al., Cell Metab. 10 (5) (2009) 430e435.
[46] Louis R. Lapierre, et al., Curr. Biol. 21 (18) (2011) 1507e1514.
[47] Y.-H. Lin, et al., Aging Cell 13 (4) (2014) 755e764.
[48] V.B. ORiordan, A.M. Burnell, Biochemistry 92 (2) (1989) 233e238.
[49] W.G. Wadsworth, D.L. Riddle, Dev. Biol. 132 (1) (1989) 167e173.
[50] J. Khnl, et al., FEBS J. 272 (6) (2005) 1465e1477.
[51] R.J. Chandler, et al., Mol. Genet. Metab. 89 (1e2) (2006) 64e73.
[52] T.V. Kurzchalia, S. Ward, Nat. Cell Biol. 5 (8) (2003) 684e688.
[53] D.L. Motola, et al., Cell 124 (6) (2006) 1209e1223.
[54] P. Mahanti, et al., Cell Metab. 19 (1) (2014) 73e83.
[55] H. Hsin, C. Kenyon, Nature 399 (6734) (1999) 362e366.
[56] T.M. Yamawaki, et al., PLoS Biol. 8 (8) (2010) e1000468.
[57] J.R. Berman, C. Kenyon, Cell 124 (5) (2006) 1055e1068.
[58] M. Thondamal, et al., Nat. Commun. 5 (2014).
[59] D.J. Chitwood, R. Lozano, W.R. Lusby, J. Nematol. 18 (1) (1986) 9e17.
[60] F. Broue , et al., Aging Cell 6 (1) (2007) 87e94.
[61] S. Penkov, et al., Angew. Chem. Int. Ed. 49 (49) (2010) 9430e9435.
[62] C. Morck, et al., Proc. Natl. Acad. Sci. 106 (43) (2009)
18285e18290. [63] B.C. Persson, et al., Biochimie 76 (12) (1994)
1152e1160.
[64] T. Jonassen, P.L. Larsen, C.F. Clarke, Proc. Natl. Acad. Sci. 98 (2) (2001)
421e426.
[65] P. Ranji, et al., PLoS One 9 (6) (2014) e100033.
[66] M. Rauthan, M. Pilon, Lipids Health Dis. 10 (1) (2011) 243.
[67] P.-Y. Jeong, et al., Nature 433 (7025) (2005) 541e545.
[68] J. Srinivasan, et al., Nature 454 (7208) (2008) 1115e1118.
[69] E.Z. Macosko, et al., Nature 458 (7242) (2009) 1171e1175.
[70] J. Srinivasan, et al., PLoS Biol. 10 (1) (2012) e1001237.

[71] Frank C. Schroeder, Chem. Biol. 22 (1) (2015) 7e16.

[72] X. Zhang et al., Acyl-CoA oxidase complexes control the chemical message
produced by Caenorhabditis elegans, in: Proceedings of the National Academy of Sciences, 2015.
[73] K. Ashra, et al., Nature 421 (6920) (2003) 268e272.
[74] H.Y. Mak, et al., Nat. Genet. 38 (3) (2006) 363e368.
[75] F. Hu, et al., Analyst 139 (10) (2014) 2312e2317.
[76] L. Wei, et al., Nat. Meth. 11 (4) (2014) 410e412.
[77] H.J. Lee, et al., Sci. Rep. 5 (2015).
[78] K. Yen, et al., PLoS One 5 (9) (2010) e12810.
[79] G.J. Tserevelakis, et al., PLoS One 9 (1) (2014) e84431.
[80] T. Hellerer, et al., Proc. Natl. Acad. Sci. 104 (37) (2007) 14658e14663.
[81] T.T. Le, et al., J. Lipid Res. 51 (3) (2010) 672e677.
[82] G. Patti, et al., Metabolomics 10 (4) (2014) 737e743.
[83] F.M. Geier, et al., Surf. Interface Anal. 45 (1) (2013) 234e236.
[84] J. Folch, M. Lees, G.H.S. Stanley, J. Biol. Chem. 226 (1) (1957) 497e509.
[85] E.G. Bligh, W.J. Dyer, Can. J. Biochem. Physiol. 37 (8) (1959) 911e917.
[86] V. Matyash, et al., J. Lipid Res. 49 (5) (2008) 1137e1146.
[87] S. Chen, M. Hoene, J. Li, Y. Li, X. Zhao, H. Haring, E.D. Schleicher, C. Weigert,
G. Xu, R. Lehmann, J. Chromatogr. A 1298 (2013) 9e16.
[88] P. Meister, et al., Genome Biol. 12 (12) (2011) R123.
[89] P. Narayanaswamy, et al., Anal. Chem. 86 (6) (2014) 3043e3047.
[90] C.K. Van Pelt, J.T. Brenna, Anal. Chem. 71 (10) (1999) 1981e1989.
[91] A.L. Michaud, et al., Anal. Biochem. 307 (2) (2002) 348e360.
[92] J.G. Bollinger, et al., J. Lipid Res. 54 (12) (2013) 3523e3530.
[93] V. Matyash, et al., PLoS Biol. 2 (10) (2004) e280.
[94] D. Schwudke, et al., Anal. Chem. 78 (2) (2006) 585e595.
[95] C. Papan, et al., Anal. Chem. 86 (5) (2014) 2703e2710.
[96] M. Ishida, et al., Rapid Commun. Mass Spectrom. 18 (20) (2004) 2486e2494.
[97] M. Witting, et al., Anal. Bioanal. Chem. 407 (4) (2015) 1059e1073.
[98] C. Castro, et al., Mol. BioSyst. 9 (7) (2013) 1632e1642.
[99] J.P. Chan, Z. Hu, D. Sieburth, Genes Dev. 26 (10) (2012) 1070e1085.
[100] L. Sumner, et al., Metabolomics 3 (3) (2007) 211e221.
[101] V.A. Sarpe, S.S. Kulkarni, J. Org. Chem. 76 (16) (2011) 6866e6870.
[102] R. Csuk, et al., Tetrahedron Lett. 54 (18) (2013) 2274e2276.
[103] K.K. Sharma, et al., Mol. Endocrinol. 23 (5) (2009) 640e648.
[104] N. Sun, et al., Proteomics (2014) p. n/a-n/a.
[105] A. Ly, et al., Histochem. Cell Biol. (2014) 1e10.
[106] M. Sud, et al., Nucleic Acids Res. 35 (suppl 1) (2007) D527eD532.
[107] Boyce Thompson Institute, C.U. C. elegans Small Molecule Identier Database
(SMID DB), 2011, Available from: <http://smid-db.org>.
[108] Y. Simo n-Manso, et al., Anal. Chem. 85 (24) (2013) 11725e11731.
[109] D.S. Wishart, et al., Nucleic Acids Res. 41 (Database issue) (2013)
D801eD807.
[110] D.S. Wishart, et al., Nucleic Acids Res. (37 Database) (2009) D603eD610.
[111] M. Hansen, T. Flatt, H. Aguilaniu, Cell Metab. 17 (1) (2013) 10e19.