Improved LC-MS method for the determination of fatty acids in red blood cells by LC-orbitrap MS

Article abstract of DOI:10.1021/ac103093w

We report a new method for fast and sensitive analyses of biologically relevant fatty acids (FAs) in red blood cells (RBC) by liquid chromatography mass spectrometry (LC-MS). A new chemical derivatization approach was developed forming picolylamides from FAs in a quantitative reaction. Fourteen derivatized FA standards, including saturated and unsaturated FAs from C14 to C22, were efficiently separated within 15 min. In addition, the use of a recently introduced benchtop orbitrap mass spectrometer under positive electrospray ionization (ESI) full scan mode showed a 2-10-fold improvement in sensitivity compared with a conventional tandem MS method, with a limit of detection in the low femtomole range for saturated and unsaturated FAs. The developed method was applied to determine FA concentrations in RBC with intra- and interday coefficients of variation below 10%.

Full text of DOI:10.1021/ac103093w

TECHNICAL NOTE
pubs.acs.org/ac
Improved LCꢀMS Method for the Determination of Fatty Acids in Red
Blood Cells by LCꢀOrbitrap MS
Xingnan Li and Adrian A. Franke*
Cancer Research Center of Hawaii, 1236 Lauhala Street, Honolulu, Hawaii 96813, United States
S Supporting Information
b
ABSTRACT: We report a new method for fast and sensitive
analyses of biologically relevant fatty acids (FAs) in red blood
cells (RBC) by liquid chromatography mass spectrometry
(LCꢀMS). A new chemical derivatization approach was devel-
oped forming picolylamides from FAs in a quantitative reaction.
Fourteen derivatized FA standards, including saturated and
unsaturated FAs from C14 to C22, were efficiently separated
within 15 min. In addition, the use of a recently introduced
benchtop orbitrap mass spectrometer under positive electro-
spray ionization (ESI) full scan mode showed a 2ꢀ10-fold
improvement in sensitivity compared with a conventional tandem MS method, with a limit of detection in the low femtomole range
for saturated and unsaturated FAs. The developed method was applied to determine FA concentrations in RBC with intra- and
interday coefficients of variation below 10%.
atty acids (FAs), particularly long chain polyunsaturated fatty
acids (PUFAs), are present in living cells and body fluids in
a 10-fold increase in sensitivity has been achieved by the precharged
quaternary ammonium salt of the trimethylaminoethyl ester
(TMAE).9,10 However, this derivatization requires harmful re-
agents, and the products lack good chromatographic resolutions.
In this study, we established a simple, fast, accurate, and sensitive
ESI-LCꢀMS method for measuring biologically and clinically
important FAs, by comparing different derivatizing reagents,
including nitrogen-containing amines and alcohols, for positive
ESI-LCꢀMS detection. Moreover, we applied a new benchtop
orbitrap MS using exact full mass scan, as an alternative method
for tandem MS that allowed exact mass measurements for
unambiguous analyte identification without collision induced
fragmentation. Finally, we applied the developed method in red
blood cells (RBC), since FA composition in RBC provide a
useful marker for oxidative stress, cardiovascular disease, and
other chronic diseases.11
F
the form of free acids, esters, triglycerides, or phospholipids, play
a key role in metabolic pathways, regulating human cardiovas-
cular and immune systems,1 and are crucial in brain develop-
ment. Accurate measurement of FAs, therefore, has important
physiological and clinical implications. The most common quanti-
fication of FAs was previously performed by gas chromatogra-
phy/mass spectrometry (GC/MS) using electron impact (EI)
ionization after methyl ester derivatization.1,2 Recently, liquid
chromatographyꢀmass spectrometry (LCꢀMS) has been
widely used for FA analysis.3 Electrospray ionization (ESI) in
combination with tandem mass spectrometry or the FTICR
(Fourier transform-ion cyclotron resonance) technique4,5 have
offered an alternative way to ionize and detect nonvolatile and
heat-sensitive FAs. However, the drawbacks of low specificity in
the negative MS detection mode and requirement of postcolumn
alkalization after acidic chromatographic separation have hin-
dered a simple way of fatty acid analysis.
’ EXPERIMENTAL SECTION
Several chemical derivatization methods6,7 have since been
developed to improve the ESI-LCꢀMS detection responses in
the positive detection mode. Methyl ester derivatives for measur-
ing fatty acids have been hampered by the lack of specificity due
to the difficulty in fragmentation in tandem mass analysis.
Johnson et al.8 have developed a sensitive ESI-LCꢀMS method
for the measurement of free fatty acids using dimethylaminoethyl
ester (DMAE) derivatives which contain a nitrogen tag that was
easy to protonate. This FA-DMAE derivative showed fast and
complete ionization under positive ESI mode and rapid frag-
mentation by collision induced dissociation (CID) which provided
useful information when measured by tandem MS. Furthermore,
Fatty Acids Derivatization. Chemicals, reagents, standards
preparation, and fatty acids extraction from red blood cells (RBC)
information is provided in supplemental notes in the Supporting
Information. Both the FA standards and RBC samples under-
went the same extraction, hydrolysis, and derivatization steps
before LCꢀMS analysis. The derivatization procedure was
modified from Johnson’s method.9 In brief, 150 μL of standard
(0.1ꢀ100 μg/mL) and 20 μL of internal standard mixture were
Received: November 15, 2010
Accepted: March 8, 2011
Published: March 23, 2011
r
2011 American Chemical Society
3192
dx.doi.org/10.1021/ac103093w Anal. Chem. 2011, 83, 3192–3198
|

Analytical Chemistry
TECHNICAL NOTE
mixed and dried under nitrogen. To the dried residue was added
200 μL of oxalyl chloride (2 M in dichloromethane), and the
mixture was incubated at 65 °C on a heating block for 5 min and
then dried under nitrogen. To the residue was added 150 μL of
dimethylaminoethanol, 3-picolylamine, or 3-pyridylcarbinol, re-
spectively (1% in acetonitrile, v/v) to form the dimethylaminoethyl
ester (FA-DMAE), 3-picolylamide (FA-PA), and 3-picolinyl
ester (FA-PE) derivatives (Figure 1), respectively. The mixture
was incubated at room temperature for 5 min, followed by drying
under nitrogen to give the derivatized FAs. The FA-DMAE product
was further converted to trimethylaminoethyl ester (FA-TMAE)
by incubatingwith 150μL of methyl iodide (50% in methanol, v/v)
at room temperature for 5 min, followed by drying under nitrogen.
The dried FAs derivatives were dissolved in 1000 μL of ethanol and
further diluted up to 10-fold with ethanol prior to LCꢀMS analysis.
LCꢀMS Analysis. The analysis was performed using a model
HTC Pal autosampler (Leap Technologies, Carrboro, NC)
connected to a model Accela ultra-HPLC system in combination
Figure 1. Fatty acid derivatization reaction scheme and their adducts fragmentation by MS/MS analysis.
Table 1. Parameters of Fatty Acid Standards and 3-Picolylamide (PA), 3-Picolinyl Ester (PE), and Dimethylaminoethyl Ester
(DMAE) Derivatives for Orbitrap Mass Spectrometry Quantification
FA-PA derivative
FA-PE derivative
FA-DMAE derivative
name
lipid name
formula
[M þ H]þ
formula
[M þ H]þ
formula
[M þ H]þ
Monounsaturated
373.32134
oleic acid (OA)
C18:1 n-9
C16:1 n-7
ω-6
C24H40N2O
C22H36N2O
C24H39NO2
C22H35NO2
374.30536
346.27406
C22H43NO2
C20H39NO2
354.33666
326.30536
palmitoleic acid (PLA)
345.29004
Polyunsaturated
371.30569
linoleic acid (LA)
C18:2 n-6
C20:2 n-6
C20:4 n-6
C24H38N2O
C26H42N2O
C26H38N2O
C24H37NO2
C26H41NO2
C26H37NO2
372.28971
400.32101
396.28971
C22H41NO2
C24H45NO2
C24H41NO2
352.32101
380.35231
376.32101
eicosadienoic acid (EDA)
arachidonic acid (AA)
399.33699
395.30569
ω-3
R-linolenic acid (ALA)
C18:3 n-3
C20:5 n-3
C22:5 n-3
C22:6 n-3
C24H36N2O
C26H36N2O
C28H40N2O
C28H38N2O
369.29004
393.29004
421.32134
419.30569
C24H35NO2
C26H35NO2
C28H39NO2
C28H37NO2
370.27406
394.27406
422.30536
420.28971
C22H39NO2
C24H39NO2
C26H43NO2
C26H41NO2
350.30536
374.30536
402.33666
400.32101
eicosapentaenoic acid (EPA)
docosapentaenoic acid (DPA)
docosahexaenoic acid (DHA)
Saturated
375.33699
347.30569
319.27439
stearic acid (SA)
palmitic acid (PA)
myristic acid (MA)
C18:0
C16:0
C14:0
C24H42N2O
C22H38N2O
C20H34N2O
C24H41NO2
C22H37NO2
C20H33NO2
376.32101
348.28971
320.25841
C22H45NO2
C20H41NO2
C18H37NO2
356.35231
328.32101
300.28971
Conjugated
369.29004
367.27439
R-eleostearic acid (ESA)
R-parinaric acid (R-PA)
C18:3 n-5
C18:4 n-3
C24H36N2O
C24H34N2O
C24H35NO2
C24H33NO2
370.27406
368.25841
C22H39NO2
C22H37NO2
350.30536
348.28971
Deuterated Internal Standards
docosahexaenoic acid-d5 (DHA-d5)
eicosapentaenoic acid-d5 (EPA-d5)
linoleic acid-d4 (LA-d4)
C28H33D5N2O
C26H31N2OD5
C24H34N2OD4
C26H30N2OD8
C22H22N2OD14
C24H39N2OD3
424.33707
398.32142
375.33080
403.35590
359.37791
378.35582
C28H32NO2D5
425.32109
399.30544
376.31481
404.33992
360.36193
379.33984
C26H36NO2D5
C24H34NO2D5
C22H37NO2D4
C24H33NO2D8
C20H25NO2D14
C22H42NO2D3
405.35239
379.33674
356.34611
384.37122
340.39323
359.37114
C26H30NO2D5
C24H33NO2D4
C26H29NO2D8
C22H21NO2D14
C24H38NO2D3
arachidonic acid-d8 (AA-d8)
palmitoleic acid-d14 (PLA-d14
)
stearic acid-d3 (SA-d3)
3193
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198

Analytical Chemistry
TECHNICAL NOTE
with a model Exactive orbitrap mass spectrometer (both from
Thermo Electron, Waltham, MA). Ten μL of the diluted FA
derivative was injected onto a Agilent Zorbax SC-C18 column
(3.0 ꢁ 50 mm, 1.8 μm, Agilent, Santa Clara, CA) using a mobile
phase consisting of 0.1% formic acid in water (A) and 0.1%
formic acid in MeCN (B) at a flow rate of 500 μL/min, with the
following linear gradient A/B (v/v): 0ꢀ7 min 35/65, 7.1ꢀ11
min 10/90, and 11.1ꢀ15 min 35/65 for the separation of FA-PA
and FA-DMAE derivatives or 0ꢀ5 min 20/80, 5.1ꢀ16 min 10/
90, and 16.1ꢀ20 min 20/80 for FA-PE derivatives separation.
Mass detection was carried out after electrospray ionization
(ESI) in positive-ion full scan mode. The settings of the mass
spectrometer were as follows: spray voltage, 4.5 kV; capillary
temperature, 250 °C; maxium injection time, 250 ms; and scan
rage, 100ꢀ650. Nitrogen was used as sheath gas (pressure 30
units) and auxiliary gas (pressure 10 units). The in-source
collision induced dissociation energy (CID) was set at 5 eV to
dissociate dimers or sodium adducts, and the automatic gain
control (AGC) was set at balanced. Data acquisition and analysis
were performed using Thermo’s Xcalibur software. Detection of
the analyte was set within 10 ppm of the calculated mass. Table 1
showed the detailed formula and exact molecular weight of each
analyte for LCꢀMS full mass scan analysis.
LCꢀMS data were also obtained on an Accela ultra-HPLC
system connected to a TSQ Ultra Quantum triple quadrupole
mass spectrometer (Thermo Electron, Waltham, MA). The ion
source was the same with the orbitrap, and the capillary temp-
erature was 250 °C. CID was performed with the collision gas at
1.0 Torr. Scanning was performed in selected reaction monitor-
ing (SRM) mode where the precursor ion transition to a dia-
gnostically valuable and abundant fragment was recorded. For
FA-DMAE derivatives, SRM was performed from the precursor
[M þ H]þ (FA plus 72 Da from the dimethylaminoehtyl group)
to the corresponding [M þ H ꢀ 45]þ ion (derived from loss of
Figure 2. LCꢀMS chromatogram of FA-PA derivatives from standards (75 ng/mL for all analytes). Profiles were collected using orbitrap LCꢀMS
described in the Experimental Section. Quantitation was performed by the exact masses of [M þ H]þ ions (protonated molecular weight of FA plus 91
Da from the picolylamine group). The abbreviations of each analyte are shown in Table 1. Mass spectra of each peak are included in the Supporting
Information.
3194
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198

Analytical Chemistry
TECHNICAL NOTE
the dimethylamino group). The collision energy (CE) was set at
16eV for polyunsaturated FAs and 20 eV for all other FAs. For FA-
PA derivatives, SRM was monitored from the precursor [M þ
H]þ ion (FAs plus 91 from the picolylamine group) to the common
fragment at m/z 109 generated by the picolylamine moiety. The
CE was set at 24 eV for polyunsaturated FAs, 30 eV for all other
FAs. The fragmentation of FA-PE derivatives requires higher
collision energy; we chose to monitor the common fragment at
m/z 92 (mass of picolyl moiety, Figure 1) generated from the
precursor [M þ H]þ (FAs plus 92 from the picolyl ester group)
that requires less collision energy. The CE was set at 34 eV for all
FAs. Figure 1 showed the detailed derivatization reaction schemes
and their fragmentation in MS/MS analysis.
modified FAs to previously reported precharged quaternary
ammonium FA-TMAE derivatives and FA-DMAE products.
All products synthesized were highly soluble in ethanol which
caused us to use ethanol for all dissolution and dilution steps.
The derivatization reactions were carried out using a mod-
ification from Johnson’s method9 via a straightforward one-pot
two-step reaction procedure (Figure 1). A total of 14 FAs in-
cluding saturated FAs (14ꢀ18 carbons), ω-3 and ω-6 polyunsa-
turated FAs (PUFAs) (18ꢀ22 carbons), some mono- and
conjugated-FAs (16ꢀ18 carbon), and 6 internal standards were
derivatized to the corresponding 3 types of products as shown in
Table 1 (FA-TMAE data not shown). Free fatty acids were first
converted to the acyl chloride intermediate by the treatment of
oxalyl chloride, followed by coupling with an amine or alcohol to
form the corresponding amide or ester. The FA-DMAE deriva-
tive was further converted to the precharged quaternary amine
FA-TMAE derivative by reacting with methyl iodide. This de-
rivatizing method is fast and quantitative, without a complicated
handling procedure. Higashi et al.13 have investigated the cou-
pling reaction of 2-picolylamine with carboxylic acid using triphe-
nylphosphine (TPP) and 2,20-dipyridyl disulfide (DPDS) via a
one-pot synthesis step. We found, however, that the TPP/DPDS
catalyzed reaction products interfere with late eluting FAs (e.g., SA
and PA) under our LC conditions, and the reaction time is longer
(30 min) compared to the 10 min total reaction time via acyl chloride.
Currently, most, if not all reported LCꢀMS methods for FA
analysis use tandem MS in SRM mode. This technique provides
good sensitivity for FAs in biological samples but has several
restrictions. First, some FA derivatives, such as methyl ester
derivatives, are hard to fragment after ionization and, therefore,
cannot be detected by SRM MS; second, increasing the number
of analytes may lead to compromised sensitivity. This is based on
the scan events on a particular time period: the more scan events,
the less scan time there is for a particular SRM transition. Third,
our experiment showed that tandem MS lacks selectivity due to
the SRM selection of noncharacteristic product ions. Last but not
least, unlike full scan mode monitoring, tandem MS performs
’ RESULTS AND DISCUSSION
Several derivatization methods have been previously investi-
gated for the improvement of LCꢀMS analyses of FAs, including
precharged quaternary amines or an electron-capture pentafluor-
obenzyl moiety.7,9,12 Those modifications were found to be
valuable as to improve ionization and sensitivity, but they also
have some limitations, for example, the use of harmful reagents
and multiple derivatization steps,9 long analysis run,7 lack of complete
separation of the physiological significant fatty acids,10 and not
being applicable to saturated fatty acids.12 To overcome these
problems, we intended to find a chemical tag with the following
properties: (1) high proton-affinity (for example, an amine or a
pyridine group which are readily ionized to be sensitively detected
by positive ESI); (2) simple, mild, and quantitative derivatization
reaction; (3) commercial availability, low cost, low toxicity, envir-
onmentalfriendliness, and high stability of derivatization reagents
and products; (4) good chromatographic properties. On the basis of
these requirements, 3-picolylamine (PA) and 3-pyridylcarbinol
(PE) were selected as the derivatization reagents mainly because
they possess a readily ionizable pyridine group and can be easily
attached to the carboxylic acid moiety via an amide or an ester
linkage, respectively. Furthermore, to compare the chromato-
graphic resolution and limit of detection of the pyridine tag, we
Table 2. Comparison between the On-Column Limit of Detection (LOD) of 3-Picolylamide (PA), 3-Picolinyl Ester (PE), and
Dimethylaminoethyl Ester (DMAE) Derivatives of FAs Analyzed by Orbitrap (Exactive) and Tandem (TSQ) Mass Spectrometersa
FA-PA derivatives
FA-PE derivatives
FA-DMAE derivatives
tR
orbitrap LOD
(fmol)
TSQ LOD
(fmol)
tR
orbitrap LOD
(fmol)
TSQ LOD
(fmol)
tR
orbitrap LOD
(fmol)
TSQ LOD
(fmol)
name
(min)
(min)
(min)
OA
PLA
LA
8.6
3.6
4.7
9.4
4.6
2.9
2.9
5.5
4.3
11.3
7.0
2.9
3.9
2.6
11
280
30
9.0
6.0
6.0
9.6
5.4
4.4
3.6
6.1
4.4
14.0
8.9
5.7
5.2
3.3
44
43
84
63
5.8
2.8
3.5
6.8
3.4
2.3
2.2
3.8
3.2
9.7
4.8
2.4
2.9
2.1
30
22
180
30
7
6
96
20
81
22
70
EDA
AA
8
98
11
34
37
65
9
101
55
20
176
165
106
70
63
167
39
ALA
EPA
DPA
DHA
SA
25
30
13
20
38
21
24
190
62
15
89
64
135
13
4
40
42
140
135
59
23
5
130
138
28
112
35
26
148
97
PA
20
34
MA
13
12
45
28
23
R-ESA
cisPA
1.3 nmol
8.4 nmol
NDb
ND
ND
ND
ND
ND
ND
ND
ND
ND
a The abbreviation of each analyte is shown in Table 1. b ND: Not determined.
3195
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198

Analytical Chemistry
TECHNICAL NOTE
only with high sensitivity and selectivity for known targets but
cannot be applied successfully to nontargeted analytes. In contrast,
the orbitrap mass spectrometers including the recently introduced
benchtop models have high resolving power, fast scan speed, high
in-scan dynamic range, and accurate mass analysis. They over-
come the problems associated with tandem MS and provide a
sensitive and selective way for analyzing all analytes including
those that were not targeted within a recorded HPLC run. Also,
reinterrogation of recorded analyses is a great advantage, espe-
cially for precious clinical, epidemiologic, and other biological
samples with limited specimen amounts. In our study, we intended
to take advantage of the orbitrap technology to analyze a wide
range of biological and clinical relevant fatty acids based on accurate
mass measurements and evaluated whether a recently introduced
benchtop model would fulfill the required needs.
including 14 FA standards and 6 deuterated internal standards,
were derivatized and successfully separated with a retention time
between 2 and 14 min (Figure 2). ESI orbitrap MS was performed
in positive full scan mode with a mass scan range from m/z 100 to
650 and 5 eV in-source CID to avoid adduct or dimer formation
without using higher energy collision induced dissociation (HCD)
in the collision cell. Quantitation was carried out using exact masses
of the protonated [M þ H]þ ions for FA-PA (FA plus 91 from
the picolylamine group), FA-PE (FA plus 92 from the pyridinyl
methanol group), and FA-DMAE (FA plus 72 from the dimethy-
laminoethyl group) or the Mþ ion (FA plus 86 from the trimethy-
laminoethyl group) for FA-TMAE derivatives. The calibration
curves were linear (R2 > 0.99) in the experimental concentration
range from 1.5 to 1500 ng/mL. Among the four different FA
derivatives, FA-PA derivatives provided intense protonated
molecules in the positive ESI and exhibited the best sensitivity
and chromatographic property. The limit of detection (LOD)
was in the low femtomole range for the saturated and unsaturated
FAs, 2ꢀ4-fold more sensitive than the DMAE method, largely
caused by the high proton-affinity pyridyl group (Table 2). The
Among many columns tested, good separation of the FA de-
rivatives were achieved on an Agilent Zorbax SC-C18 reverse
phase column using a mobile phase consisting of 0.1% formic acid in
water and acetonitrile that exhibits a better chromatographic
resolution than using methanol. A mixture of 20 fatty acids,
Figure 3. Typical orbitrap MS chromatogram of fatty acids extracted from 20 μL of red blood cells (diluted in 10 mL of AcN) after 3-picolylamide (PA)
derivatization. The abbreviation of each analyte is shown in Table 1.
3196
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198

Analytical Chemistry
TECHNICAL NOTE
Table 3. FA-PA Method Validation with NIST SRM3274-FAs in Botanical Oilsa
saturated FAs
unsaturated FAs
SRM oil
borage
MA
PA
SA
OA
LA
0.49 (0.62 ( 0.11)
100(110 ( 12)
33.6 (33.1 ( 4.0)
17.8 (18.3 ( 0.838)
27.9 (30.4 ( 2.4)
19.3 (20.9 ( 1.1)
152.4 (148.7 ( 8.7)
69.9 (68.9 ( 3.7)
396 (374 ( 35)
787 (742 ( 24)
160 (171 ( 11)
212 (160 ( 14)
evening primrose
0.344 (0.363 ( 0.03)
0.277 (0.271 ( 0.008)
0.246 (0.206 ( 0.025)
54.7 (58.2 ( 6.1)
42.7 (44.8 ( 5.0)
52.8 (56.4 ( 5.5)
flax
165.3 (165.7 ( 6.2)
157.2 (166.8 ( 7.8)
perilla
a Certified concentration value was expressed as a mass fraction (mg/g) at 95% level of confidence and noted in bold in the parentheses. Reference values
were also given in normal typeface in parentheses. The abbreviation of each analyte is shown in Table 1.
conjugated FAs, including cisPA and RESA, however, showed a
LOD in the nanomole range, possibly due to the labile nature of
the conjugated system during sample preparation. The FA-PE
derivatives also gave intense protonated [M þ H]þ signals as
base peaks in the positive ESI-MS mode, but the FA-PA derivatives
exhibited better results in terms of chromatography and LOD,
with the lipophilic FA-PE derivatives more retained on the
column that required a high portion of organic solvent for elution;
in addition, the LOD was 2ꢀ3-fold higher than the FA-PA de-
rivatives. Interestingly, contrary to the 10-fold increase in sensi-
tivity using precharged quaternary amine FA-TMAE derivatives
via methyl iodide treatment under a tandem mass method as
reported previously,9 we found FA-DMAEs and FA-TMAEs to
possess similar LODs in the femtomol range using orbitrap MS in
the full scan method. DMAE products are preferable to TMAE,
because the harmful reagent methyl iodide is not needed and an
additional derivatization step can be avoided. In summary, we
found that FA-PA derivatives are superior to PEs, DMAEs, and
TMAEs when measured by orbitrap MS, due to 2ꢀ4-fold lower
LODs, requiring short chromatographic runs (2ꢀ12 min) with
good resolutions.
consistency of the assay was evaluated by repeated analysis of quality
control (QC) RBC samples for 5 days. The intraday CVs for the
lyophilized and liquid RBC ranged from 4 to 19% (mean of 7%) and
5 to 18% (mean of 7%), respectively, on the basis of 10 analytes.
Dried RBC showed better interday CV ranges (4ꢀ26% with a mean
of 15%) than liquid RBC (11ꢀ22% with a mean of 20%) on the
basis of 10 analytes, possibly due to lack of homogeneity in the liquid
RBC. It is important to note that high CV values were only obtained
by the low concentrated analytes in RBC. The mean intra- and
interday CV of lyophilized RBC was improved to 5 and 8%, re-
spectively, on the basis of the 6 most concentrated FA analytes (i.e.,
conc >0.1 mg/g such as PA, SA, LA, AA, OA, and DHA). More
RBC material would be needed for the analysis of low level FAs.
The accuracy of the FA-PA method was validated using NIST
standard reference material SRM3274 (FAs in botanical oils).
Four different botanical oils were hydrolyzed, extracted, and
derivatized to FA-PAs (see Supporting Information), and the final
products were analyzed by orbitrap MS. The concentrations of
MA, SA, PA, OA, and LA were determined to be within the 95%
confident level, except that the LA concentrations in evening
primrose and perilla oils are slightly higher than that (Table 3).
For comparison purposes, the FA derivatives were also analyzed
on a triple quadrupole MS. The LC and ion source conditions
were the same as those used with the orbitrap. The quantification
was performed under SRM mode, and collision energy (CE) was
operated at different levels for the saturated and unsaturated FAs
as well as for FA-DMAE, PA, and PE derivatives to obtain the
optimal sensitivity for each analyte (see Experimental Section
part for detailed description). Compared to FA-PAs, the FA-PEs
required higher collision energies to form the 3-pyridylcarbinol
ion (m/z 110) at above 45 eV; we, therefore, chose to monitor
the transition to the next abundant 3-pyridyl methyl ion (m/z
92) which required less energy (CE below 34 eV). For FA-PA
derivatives, the needed CE was relatively mild, but saturated FAs
needed higher energy than unsaturated FAs. Protonated 3-pico-
lylamine (m/z 109) was selected for SRM transitions. The FA-
DMAE derivatives required less CE (16 eV) for fragmentation;
SRM was selected in this case following the transition from the
adduct to the loss of the dimethylamine moiety (Figure 1). The
calibration curves were linear (R2 > 0.99) in the experimental
concentration range from 0.75 to 150 nM. Overall, TSQ triple
quadruple MS/MS also provided a sensitive detection for FAs with
a LOD (signal-to-noise = 3) in the midfemtomole range, but the
orbitrap assay was found to be more sensitive than the tandem
mass assay with detection limits being 2ꢀ10-fold lower. Table 2
shows a detailed comparison between different LCꢀMS methods.
Finally, we applied the FA-PA orbitrap MS method to the
detection of FAs in RBCs (Figure 3). Due to the highly sensitive
nature of FA-PAs, the sample size could be significantly reduced
to around 20 μL of liquid or 5 mg of lyophilized RBCs. The
’ CONCLUSIONS
We developed a simple and fast fatty acid-picolylamine (FA-
PA) derivatization method for the sensitive analysis of fatty acids
using exact masses as measured by orbitrap mass spectrometry in
positive ESI full scan mode. In comparison with previously
reported methods using DMAE and TMAE derivatizations, the
FA-PA method had superior sensitivity with a limit of detection
in the low femtomole range, 2ꢀ4-fold lower than previous methods.
Efficient separation for 14 calibrated saturated and unsaturated
FAs was achieved within 15 min. In addition, the sensitivity and
selectivity were further improved 2ꢀ10-fold by the use of orbi-
trap MS in full scan mode, which offers great advantages for trace
amount analysis in addition to the potential of data reinterroga-
tion of nontargeted FAs using full scan data. Finally, oribitrap also
provides additional selectivity for nonreadily fragment molecules
such as protonated FA esters.
’ ASSOCIATED CONTENT
S
Supporting Information. Additional information as noted
b
in text. This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Telephone: (808) 586-3008. Fax: (808) 586-2970. E-mail:
adrian@crch.hawaii.edu.
3197
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198

Analytical Chemistry
TECHNICAL NOTE
’ REFERENCES
(1) Moldovan, Z.; Jover, E.; Bayona, J. M. Anal. Chim. Acta 2002,
465, 359–378.
(2) Seppanen-Laakso, T.; Laakso, I.; Hiltunen, R. Anal. Chim. Acta
2002, 465, 39–62.
(3) Rezanka, T. J. High Resolut. Chromatogr. 2000, 23, 338–342.
(4) Hughey, C. A.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2002,
74, 4145–4149.
(5) Leavell, M. D.; Leary, J. A. Anal. Chem. 2006, 78, 5497–5503.
(6) Johnson, D. W.; Trinh, M.-U. Rapid Commun. Mass Spectrom.
2003, 17, 171–175.
(7) Yang, W.-C.; Adamec, J.; Regnier, F. E. Anal. Chem. 2007,
79, 5150–5157.
(8) Johnson, D. W. Rapid Commun. Mass Spectrom. 2000, 14, 2019–
2024.
(9) Johnson, D. W.; Trinh, M.-U.; Oe, T. J. Chromatogr., B: Anal.
Technol. Biomed. Life Sci. 2003, 798, 159–162.
(10) Pettinella, C.; Lee, S. H.; Cipollone, F.; Blair, I. A. J. Chromatogr., B:
Anal. Technol. Biomed. Life Sci. 2007, 850, 168–176.
(11) O’Brien, D. M.; Kristal, A. R.; Jeannet, M. A.; Wilkinson, M. J.;
Bersamin, A.; Luick, B. Am. J. Clin. Nutr. 2009, 89, 913–919.
(12) Lee, S. H.; Williams, M. V.; DuBois, R. N.; Blair, I. A. Rapid
Commun. Mass Spectrom. 2003, 17, 2168–2176.
(13) Higashi, T.; Ichikawa, T.; Inagaki, S.; Min, J. Z.; Fukushima, T.;
Toyo’oka, T. J. Pharm. Biomed. Anal. 2010, 52, 809–818.
3198
dx.doi.org/10.1021/ac103093w |Anal. Chem. 2011, 83, 3192–3198