Higher-energy collision-induced dissociation for the quantification by liquid chromatography/tandem ion trap mass spectrometry of nitric oxide metabolites coming from S-nitroso-glutathione in an in vitro model of the intestinal barrier

Article abstract of DOI:10.1002/rcm.8287

Rationale: The potency of S-nitrosoglutathione (GSNO) as a nitric oxide (NO) donor to treat cardiovascular diseases (CVDs) has been highlighted in numerous studies. In order to study its bioavailability after oral administration, which represents the most convenient route for the chronic treatment of CVDs, it is essential to develop an analytical method permitting (i) the simultaneous measurement of GSNO metabolites, i.e. nitrite, S-nitrosothiols (RSNOs) and nitrate and (ii) to distinguish them from other sources (endogenous synthesis and diet). Methods: Exogenous GSNO was labeled with ¹⁵N, and the GS¹⁵NO metabolites after conversion into the nitrite ion were derivatized with 2,3-diaminonaphthalene. The resulting 2,3-naphthotriazole was quantified by liquid chromatography/tandem ion trap mass spectrometry (LC/ITMS/MS) in multiple reaction monitoring mode after Higher-energy Collision-induced Dissociation (HCD). Finally, the validated method was applied to an in vitro model of the intestinal barrier (monolayer of Caco-2 cells) to study GS¹⁵NO intestinal permeability. Results: A LC/ITMS/MS method based on an original transition (m/z 171 to 156) for sodium ¹⁵N-nitrite, GS¹⁵NO and sodium ¹⁵N-nitrate measurements was validated, with recoveries of 100.8 ± 3.8, 98.0 ± 2.7 and 104.1 ± 3.3%, respectively. Intra- and inter-day variabilities were below 13.4 and 12.6%, and the limit of quantification reached 5 nM (signal over blank = 4). The permeability of labeled GS¹⁵NO (10–100 μM) was evaluated by calculating its apparent permeability coefficient (P_app). Conclusions: A quantitative LC/ITMS/MS method using HCD was developed for the first time to selectively monitor GS¹⁵NO metabolites. The assay allowed evaluation of GS¹⁵NO intestinal permeability and situated this drug candidate within the middle permeability class according to FDA guidelines. In addition, the present method has opened the perspective of a more fundamental work aiming at studying the fragmentation mechanism leading to the ion at m/z 156 in HCD tandem mass spectrometry in the presence of acetonitrile.

Full text of DOI:10.1002/rcm.8287

Higher-energy Collision Dissociation for the quantification by liquid chromatography-
tandem ion trap mass spectrometry of nitric oxide metabolites coming from S-
nitrosoglutathione in an in vitro model of intestinal barrier
Yu Haiyan1, Bonetti Justine1, Gaucher Caroline1, Fries Isabelle1, Vernex-Loset Lionel2,
Leroy Pierre1, Chaimbault Patrick2,*
1 Université de Lorraine, CITHEFOR, F-54000 Nancy, France
2 Université de Lorraine, LCP-A2MC, F-57000 Metz, France
* Correspondence to: Chaimbault Patrick, SRSMC UMR 7565, ICPM, 1 boulevard Arago,
57070 Metz, France
E-mail: patrick.chaimbault@univ-lorraine.fr
Abstract
RATIONALE: The potency of S-nitrosoglutathione (GSNO) as a nitric oxide (NO) donor to
treat cardiovascular diseases (CVDs) has been highlighted in numerous studies. In order to
study its bioavailability after oral administration, which represents the most convenient route
for the chronic treatment of CVDs, it is essential to develop an analytical method permitting
(i) the simultaneous measurement of GSNO metabolites, i.e. nitrite, S-nitrosothiols (RSNOs)
and nitrate and (ii) them to be distinguished from other sources (endogenous synthesis and
diet).
15
METHODS: Exogenous GSNO was labeled with N, and the GS15NO metabolites after
conversion to nitrite ion were derivatized with 2,3-diaminonaphthalene. The resulting 2,3-
naphthotriazole was quantified by liquid chromatography-tandem ion trap mass spectrometry
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/rcm.8287
This article is protected by copyright. All rights reserved.

(LC/ITMS/MS) in multiple reaction monitoring mode after Higher-energy Collision
Dissociation (HCD). Finally, the validated method was applied to an in vitro model of
intestinal barrier (monolayer of Caco-2 cells) to study GS15NO intestinal permeability.
RESULTS: A LC/ITMS/MS method based on an original transition (m/z 171 to 156) for
sodium 15N-nitrite, GS15NO and sodium 15N-nitrate measurements was validated, with
recovery of 100.8 ± 3.8, 98.0 ± 2.7 and 104.1 ± 3.3 %, respectively. Intra- and inter-day
variabilities were below 13.4 and 12.6 %, and the limit of quantification reached 5 nM (signal
over blank = 4). The permeability of labeled GS15NO (10-100 µM) was evaluated by
calculating its apparent permeability coefficient (Papp).
CONCLUSIONS: A quantitative LC-ITMS/MS method using HCD was developed for the
first time to selectively monitor GS15NO metabolites. The assay allowed evaluating GS15NO
intestinal permeability and situated this drug candidate within the middle permeability class
according to FDA guidelines. In addition, the present method has opened the perspective of a
more fundamental work aiming at studying the fragmentation mechanism leading to the ion
at m/z 156 in HCD tandem mass spectrometry in the presence of acetonitrile.
Key words: 15N-nitrite, 15N-nitrate, 15N-S-nitrosoglutathione, 2,3-diaminonaphthalene, LC-
ITMS/MS
This article is protected by copyright. All rights reserved.

INTRODUCTION
Nitric oxide (NO) is a gaseous messenger playing an important role in the vascular
system homeostasis1. Its origin includes both endogenous (catabolism of arginine catalyzed
by endothelial NO synthase)2,3 and exogenous (intake of drinking water and food)4 sources.
The endogenous production of NO decreases with ageing5 and cardiovascular diseases
(CVDs) related to endothelial dysfunction. In this case, an exogenous supply of NO is
necessary. Currently, drugs used as NO donors have major drawbacks such as oxidative
stress induction and tolerance phenomenon6,7. Many studies highlight the potency of S-
nitrosothiols (RSNOs) as NO donors, especially S-nitrosoglutathione (GSNO)8, because they
do not exhibit the side effects previously cited9. However, very few reports have been
focused on their bioavailability after oral administration10, which is their most convenient
route for the chronic treatment of CVDs. Then small intestine is the major barrier for the
efficient absorption of orally administered therapeutics11,12. Caco-2 trans-epithelial
permeability assay (monolayer of differentiated Caco-2 cells) is the in vitro model commonly
used for intestinal absorption studies of drugs 13,14. Thus, using this model is the first step to
evaluate the intestinal permeability of GSNO and its bioavailability.
In a biological environment, GSNO is easily decomposed by metal ions, reductants,
enzymes or light, resulting in a release of NO9. This reactive species is then rapidly oxidized
into nitrite and nitrate ions4. Moreover, GSNO undergoes a transnitrosation process
corresponding to NO moieties exchange with a cysteine residue of peptides and proteins,
resulting in a formation of RSNOs8. The different GSNO metabolites, i.e. nitrite, nitrate ions
and RSNOs, represent a NO reservoir9. S-nitrosation of proteins is also a post-translational
pathway playing a key role in many physiopathological processes8,9,15. GSNO metabolites
should be measured together because they give complementary information4,8,9. Thus, there is
a great need to develop appropriate assays in this field. Actually, very few assays allow the
determination of all cited NO species in the same methods16,17,18
.
Mass spectrometry is a widespread technique used for the quantification of metabolites
of interest in biological samples because of its high specificity of detection (multiple reaction
This article is protected by copyright. All rights reserved.

monitoring (MRM) mode) and its sensitivity for trace level measurement. Because of its low
molecular mass, NO (30 u) has to be derivatized to allow its measurement in cells or
biological fluids (urine, plasma, etc). Indeed, it is not possible for many commercially
available mass spectrometers to reach such a low m/z value and the background noise is
usually high at low masses. As mentioned above, NO is rapidly converted into nitrite and
nitrate ions, which are the real derivatized compounds. For pharmacology studies, there is a
need to quantify NO species coming from administered NO donor drugs regardless of
whether they come from endogenous synthesis or diet intake. Mass spectrometry allows
distinguishing NO species coming from administered NO donors and those from endogenous
15
synthesis and diet intake, thanks to the incorporation of the stable nitrogen isotope N into
the NO donor (presently GS15NO). Indeed, the 15N isotope is stable and its natural abundance
level represents 0.37% of all nitrogen isotopes19. Thus, it is possible to monitor specifically
14
adducts labeled either with N (corresponding to the endogenous production and the diet
uptake) or with 15N (corresponding to the drug uptake).
In addition to the increase in molecular mass, the derivatization procedure facilitates
nitrite and nitrate ions analysis either by gas chromatography (the nitrite ion is transformed
into a volatile derivative) or by Reversed Phase Liquid Chromatography (nitrite ion is
transformed into a more hydrophobic derivative increasing the retention on this kind of
stationary phases). The GC-CI-MS method developed by Tsikas et al uses pentafluorobenzyl
bromide (PFBBr) for 15N-nitrite and 15N-nitrate derivatization. It has the benefit of measuring
derivatized nitrite and nitrate ions in the same chromatographic run20. The method in liquid
chromatography-tandem mass spectrometry (LC-MS/MS) reported by Hanff and coworkers
15
uses reduced L-glutathione (GSH) for N-nitrite derivatization and simultaneously analyzes
nitrite ion and GSNO21. This assay relies upon the S-nitrosation reaction between GSH as a
probe and the nitrite ion to yield GSNO. 2,3-diaminonaphthalene (DAN) is a fluorogenic
probe which reacts with nitrite ion to yield 2,3-naphthotriazole (NAT) adduct22 (Figure 1A).
It has been widely used with spectrofluorimetric direct detection23 or coupled with LC22 and
15
even LC-MS/MS associated with N-nitrite derivatization and 15N-nitrate after reduction via
This article is protected by copyright. All rights reserved.

enzymatic catalysis24,25,26 (Figure 1C). Currently, there is no report of using LC-MS/MS to
measure RSNOs after DAN derivatization. For RSNOs, mercuric ions are commonly used to
cleave the S-NO bond with further analysis of released nitrite ion27 (Figure 1B).
This paper presents the development of a DAN-based LC-MS/MS assay for selectively
monitoring all NO species metabolized from 15N-labeled GSNO (nitrite, nitrate ions and
RSNOs). For the first time, all these NO species will be quantified using the same analytical
approach, providing a significant contribution to understanding the mechanism of GSNO
metabolism. Currently, the publications related to a DAN-based LC-MS/MS assay24,25,26
involve either a triple quadrupole mass spectrometer or a Q-trap instrument (operated as a
triple quadrupole). This is easily understandable as triple quadrupole apparatus are well
known to be the most sensitive tandem mass spectrometers for quantification when they are
operated in MRM mode. However, this kind of spectrometer is not always available in the
laboratory, and here we evaluated the sensitivity of Ion Trap Mass Spectrometry (ITMS) for
the quantification of NO traces associated to the potential benefit of the Higher-energy
Collision Dissociation (HCD) mode in this field. This original fragmentation method which
cannot be used with a triple quadrupole mass spectrometer was compared with the traditional
Collisionally Induced Dissociation (CID) available on both Ion trap and triple quadrupole
mass spectrometers. Thus, the presented method involves liquid chromatography coupled to
tandem ion trap mass spectrometry (LC-ITMS/MS) for the selective quantification of NO
species metabolized from GS15NO. DAN will be used for the derivatization of free 15N-
15
15
nitrites and N-nitrites coming from N-nitrates and RS15NO species. The resulting adduct
will be further analyzed by LC-ITMS/MS. Finally, the method will be validated and applied
to study the intestinal permeability of GS15NO with use of an in vitro model.
This article is protected by copyright. All rights reserved.

EXPERIMENTAL
Chemicals and reagents
All reagents and standards are analytical grade and used without further purification.
15
DAN, GSH, sodium N-nitrate (Na15NO3), 2-naphthylamine (2-NA) and mercuric chloride
15
(HgCl2) were obtained from Sigma-Aldrich (Saint-Quentin-Fallavier, France); sodium N-
nitrite (Na15NO2) was from Cambridge Isotope Laboratories (Tewksbury, MA, USA) and
14N-NAT from Chemodex (St. Gallen, Switzerland). Ultrapure deionized water (> 18.2
MΩ·cm) was used for the preparation of all solutions.
S-nitrosoglutathione (GS15NO) synthesis
GS15NO was synthesized as already described28 for GS14NO: 20 mM GSH was
incubated with 20 mM Na15NO2 in 500 mM HCl at 4°C in the dark for 1 h. The reaction was
stopped by neutralization with NaOH (40%) and the reaction mixture was further two-fold
diluted with 500 mM phosphate buffer, pH 7.4. The final concentration was determined by
UV spectrophotometry at a wavelength of 334 nm (ε = 922 L·mol-1·cm-1).
2,3-Diaminonaphthalene (DAN) derivatization
The strategy of analysis using the same LC-ITMS/MS approach after DAN derivation of
15
the different GS15NO metabolites (15N-nitrites, N-nitrates and RS15NO) generated during
the in vitro study is depicted in Figure 2.
15N-nitrite standard solutions were prepared daily in Hank's Balanced Salt Solution
(HBSS; NaCl, KCl, glucose, KH2PO4, Na2HPO4, CaCl2, MgSO4, pH 7.4). Three hundred µL
of standard solutions were mixed with 60 µL of 0.105 mM DAN in 0.6 M HCl, incubated at
37°C for 10 min (Figure 1A). The addition of 36 µL of 1 M NaOH and 4 µL of 0.1 mM
NaOH stopped the reaction. The real samples were similarly treated and their analysis
provided the concentration of free 15N-nitrites in the in vitro model ().
15N-nitrate was enzymatically converted into 15N-nitrite by using Nitrate/Nitrite
Fluorometric Assay Kit (Cayman Chemical, Ann Arbor, MI, USA) with some modifications
15
(Figure 1C). Briefly, 40 μL of N-nitrate standard solutions were diluted with 120 µL of
assay buffer (Item N0. 780022). The resulting mixture was incubated at room temperature for
This article is protected by copyright. All rights reserved.

60 min after adding 20 µL of enzyme cofactor (Item N0. 780012) and 20 µL of nitrate
reductase (Item N0. 780010). To continue the derivatization reaction, the mixture was
immediately processed as described for the nitrite ion by adding 40 µL of 0.105 mM DAN
and incubating at 37°C for 10 min, followed by the addition of 24 µL of 1 M NaOH and 3 µL
of 0.1 mM NaOH. The real samples were similarly treated and their analysis provided the
15
15
cumulative concentration of free N-nitrites and N-nitrates () in the in vitro model. The
15N-nitrate concentration in the in vivo model () was then obtained by subtracting the
concentration of free nitrites (=-).
The quantification of RS15NOs is carried out using a calibration curve built with
GS15NO in presence of mercuric ions: 300 µL of GS15NO standard solution were incubated
with 60 µL of 0.105mM DAN solution containing 1.05 mM HgCl2 in 0.6 M HCl at 37 °C for
10 min. The addition of 36 µL of 1 M NaOH and 4 µL of 0.1 mM NaOH stopped the
reaction. The real samples were similarly treated to provide the cumulative concentration of
15
free N-nitrites and RS15NO () in the in vitro model. The RS15NO concentration in the in
vivo model () was then obtained by subtracting the concentration of free nitrites (=-).
All resulting solutions were frozen at −80 °C for a maximum period of 3 months, until
analysis by LC-ITMS/MS.
2-Naphthylamine (2-NA) was used as internal standard (IS). For quantitative analysis,
each sample was spiked with 10 µL of a 0.02 mM solution of 2-NA prepared in acetonitrile-
water (50/50, v/v).
Analysis of 2,3-naphthotriazole (NAT) by LC-ITMS/MS
A Dionex Ultimate 3000 liquid chromatograph (Dionex, Sunnyvale, CA, USA) was
coupled with an ion trap mass spectrometer (LTQ Velos Pro, Thermo Scientific, San Jose,
CA, USA) fitted with the HESI-II probe (electrospray) operated in positive ion mode. Data
were collected and analyzed with Tune Plus software (Thermo Scientific). The injection
volume was 20 L. The separation was carried out on a Thermo Scientific Extend-C18 (2.1 ×
150 mm ID, 3 μm) column at a temperature of 25 °C. The mobile phase was composed of A:
0.01 M acetic acid and B: acetonitrile, delivered at 200 μL·min-1 in a gradient mode. The
This article is protected by copyright. All rights reserved.

column was first equilibrated at 40% B. When the sample was injected, the percentage of B
increased to 80% over 3.5 min and held for 2.5 min. The column was re-equilibrated for 10
min between two sample runs. In these chromatographic conditions, NAT is eluted from the
column in 5.5 min and 2-NA in 6.6 min (Figure 3). In addition, during the first 3 min, the
mobile phase was directed to waste using a switching valve in order to limit the
contamination of the mass spectrometer by the matrix.
The electrospray voltage was +5.0 kV. The ion source and capillary temperatures were
280 and 250 °C, respectively. The sheath, auxiliary and sweep gas flows were set at 10, 5 and
0 (arbitrary units) respectively. For quantification, the best MS/MS conditions were reached
by optimizing the collision energy for each analyte in order to obtain the highest signal. The
selected reaction monitoring (SRM) method recorded the following transitions for the
15
14
analytes: m/z 171 to 156 for N-NAT or m/z 170 to 156 for N-NAT via their respective
acetonitrile adduct (m/z 212 and 211) formed in HCD cell (the formation of the adduct with
this acetonitrile and the choice of these transitions will be discussed later on in the result and
discussion section) obtained in HDC at 50 eV and m/z 144 → 117 for the IS (2-NA) in CID at
35 eV. The activation Q value was 0.250 and the isolation window was 1u for each transition.
Samples were randomized to prevent batch effect, paired and run in sequence.
Method validation and quality control
A six-point calibration curve (blank, 5, 10, 20, 50, 100, 200 nM) for Na15NO2 and
GS15NO, and a five-point calibration curve (blank, 250, 500, 1250, 2500, 5000 nM) for
Na15NO3 were made in HBSS. They were used to calculate the limits of quantification (LOQ),
to check the linear range and to determine the percentage of recovery. Background levels of
15
15N-nitrite, GS15NO and N-nitrate in HBSS corresponding to blanks were measured and
subtracted from all quantification values. Three concentrations (15, 50 and 180 nM) of
Na15NO2 and GS15NO and (300, 2500 and 4000 nM) of Na15NO3 in HBSS were analyzed six
times to determine the intra-day precision and accuracy. The number of replicates for the
inter-day validation was also six. The inter-day statistics were determined from three
independent analytical runs, each containing the calibration standards and two QC samples at
This article is protected by copyright. All rights reserved.

low, medium and high concentration levels. A six-point calibration curve of Na15NO2 in
HBSS was daily run to assess instrument performances.
Permeability studies of GS15NO on Caco-2 cell monolayer model
The permeability of GS15NO across the Caco-2 monolayer was evaluated in the apical-
to-basolateral direction in HBSS buffer containing Ca2+ and Mg2+ as previously described10.
Briefly, 2×106 cells·cm-² were seeded on cell culture inserts (Transwell®, Corning, NY,
USA) with 0.4 μm pore size disposed in a 12-wells plate. The medium was replaced every
two days during the first week and daily during the final days, until the differentiated cell
monolayer was formed (14-18 days, transepithelial electrical resistance (TEER)  500 Ω·cm-
2). TEER was measured with a Millicell®-Electrical Resistance System (Millipore, Billerica,
MA, USA).
For permeation experiments, 500 μL of HBSS containing 10, 25, 50 or 100 µM of
GS15NO were introduced into the apical (donor) compartment. The basolateral (receptor)
compartment was filled with 1.5 mL of HBSS. After 1 h of permeation at 37°C, both
compartments were harvested and diluted with HBSS to fit the ranges of calibration curves
15
15
for N-nitrite, GS15NO and N-nitrate analysis. The resulting dilutions were immediately
derivatized with DAN and stored at -80 °C until analysis by LC/ITMS/MS. After each
permeation study, the integrity of the Caco-2 cell monolayer was evaluated by TEER
measurement and sodium fluorescein (5 μM) permeability assessment.
The cumulative amounts of GS15NO permeated through the Caco-2 cell monolayer were
calculated from the concentrations measured in the basolateral compartment. The apparent
permeability coefficient (Papp) values were calculated using the following equation:
dQ/dt (mol·s-1) refers to the quantity of permeated 15NO species (dQ) in the basolateral
compartment at the time of quantification (dt), A refers to membrane diffusion area (1.12
cm2), and C0 refers to the initial concentration of GS15NO in the apical compartment (from
1.10-5 to 1.10-4 M).
This article is protected by copyright. All rights reserved.

For each concentration level evaluated (incubation with 10, 25, 50 and 100 µM of GS15NO),
the permeation experiments were carried out 3 times.
Statistical analysis
All results are expressed as the means ± standard deviation. The Student’s t-test was
used to determine significant differences (p < 0.05).
RESULTS AND DISCUSSION
-
Nitric oxide is very unstable and turns rapidly into nitrite ion (NO2 ) which is the real
measured compound. .In the field of LC, nitrites are often derivatized by the 2,3-
diaminonaphthalene (DAN). The resulting 2,3-naphthotriazole (NAT) derivative allows its
detection by fluorescence (Figure 1). The DAN-based fluorescence LC method is particularly
appreciated for the analysis of nitrites (and nitrates after enzymatic reduction) in all
biological samples because of its simplicity of use, the rapidity of the sample preparation, and
the easy automation29. However, the fluorescence detection does not allow the distinction of
isotope labeling and the quantification of NO metabolites in the nM range (LOD is around
100 nM with DAN-based fluorescence HPLC).
Currently, there are only three publications related to DAN-based LC-MS/MS assays
found in the literature24,25,26. All these experiments were carried out either with a triple
quadrupole or with a Q-trap instrument using the CID fragmentation mode. In 2011, Shin et
al.26 reached a LOQ of 4nM and their method was only validated for the measurement of
-
NO2 . In 2016, Chao et al.24 published the quantification by DAN-based LC-MS/MS in urine
-
-
of NO2 and of NO3 after on-line solid phase extraction. For the first time, the interference on
15
13
the N-labeled NAT due to the natural presence of C is evoked and estimated to be 2.2%
(in good agreement with the theoretical value) of the contribution of the followed transition
for this derivative. The same year, Axton et al.25 reached a LOQ of 1nM in water and in the
Dulbecco’s Modified Eagle Medium (DMEM). However, the values of interferences of in the
blank samples were not discussed whereas they were 91 and 339 nM for 14N-nitrites in water
13
15
and DMEM respectively. Thus, the interference due to C on the signal N-nitrite signal in
This article is protected by copyright. All rights reserved.

the blank samples should be around 2 nM in water and 7.5 nM in DMEM. These interference
levels have to be considered to appreciate the LOQ in the real samples.
In this paper, we evaluated the sensitivity of an Ion Trap Mass Spectrometer associated
to the Higher-energy Collision Dissociation mode for the quantification of NO metabolite
traces. To achieve this goal, we compared the performance of the HCD mode with the CID
mode traditionally used and the blank values were also discussed to provide the LOQ in the
real samples. The method developed and validated in this paper was also tested to monitor
NO freed from RSNOs.
Optimization of the LC-ITMS/MS detection - Comparison between CID and HCD
The NAT adduct resulting from the derivatization of nitrite ion with DAN was
quantified by LC/ITMS/MS. Its quantification at trace level supposed the optimization of the
whole analytical pathway. Among the existing literature on NAT analysis by reversed phase
LC/MS/MS, acetonitrile (ACN) and methanol (MeOH) have been used as organic modifiers
in the mobile phase in combination with various acidic buffers. In our study, the most intense
MS signal was obtained with a 10 mM acetic acid solution combined with ACN in gradient
elution (see experimental section for details). Acetonitrile improved the sensitivity of NAT
detection by a factor of 10 compared with MeOH. The analysis took 7 min with a re-
equilibration time of 10 min between two injections. The gradient helped to reduce peak
widths and thus improved the sensitivity.
Acetonitrile also showed another interesting effect when we compared the two modes of
fragmentation with the LTQ Velos Pro ion trap instrument: the commonly used Collisionally
Induced Dissociation and Higher-energy Collision Dissociation. The CID mode led to the
same fragmentation pathway already published when using a triple quadrupole
instrument24,25,26 (Figure 4A). The protonated 14N-NAT (m/z 170) yielded two product ions at
m/z 142 (loss of N2) and m/z 115 (the indene ion [C9H7]+ derived from m/z 142 and
comprising the rest of the naphthyl ring after the loss of HCN)25.
This article is protected by copyright. All rights reserved.

The fragmentation in HCD was slightly different from that in CID (Figure 4B). When
m/z 170 is isolated and MS/MS experiments are carried out on this ion, the ultimate product
ion at m/z 115 remained the same but the intermediate product ion was now observed at m/z
156. There is thus a difference of 14 u between this product ion and its precursor ion. This
difference, which cannot obviously be assigned to a loss of a single nitrogen atom (N) or a
methylene group (CH2), has another complex origin which has been resolved through several
complementary experiments detailed below (Figures 4C and 4D, and Table 1):
 When MeOH replaced ACN in the mobile phase, the CID and HCD spectra showed the
same transition. In both cases, the precursor ion [14N-NAT+H]+ produced two transitions
m/z 170 → 142 and m/z 170 → 115. Thus, the presence of ACN is responsible of the
original fragment at m/z 156.
 In this work, 15N-NAT is quantified in order to monitor NO coming from the parent drug.
Its corresponding HCD MS/MS spectrum is provided in Figure 4D. The precursor ion
[15N-NAT+H]+ at m/z 171 produced two fragments at m/z 156 and 115. The observed
fragments are the same as for [14N-NAT+H]+. The conclusion is that 15N is lost during the
fragmentation from m/z 171 to 156.
 In order to understand what happened in the HCD cell with ACN during the
14
15
fragmentation, the N- and N-labeled NAT were infused using the LC mobile phase
containing either ACN or d3-ACN. The results are displayed in Table 1, and Figures 4C
and 4D. The hypothesis was the formation of a [NAT+H+ACN]+ adduct, confirmed by the
14
presence of a peak at m/z 211 for N-NAT in ACN and m/z 214 in d3-ACN. This adduct
underwent a rearrangement during fragmentation leading to m/z 183 by N2 loss (m/z 186
with d3-ACN). Then, an additional HCN loss led to m/z 156 from m/z 183 or m/z 159 from
m/z 186 with d3-ACN. All these ions were perfectly visible on the mass spectra (even if
some of them exhibited a very low intensity). The values of m/z observed for each ion
15
were consistent with this explanation whatever the derivative (14N or N-NAT) and the
solvent condition (ACN or d3-ACN), as illustrated in Table 1 and Figure 4. From these last
experiments, we concluded that ACN probably strongly interacted with the [NAT+H]+ ion
This article is protected by copyright. All rights reserved.

in the HCD cell. This is probably due to the specific position of the HCD cell in the LTQ
Velos Pro mass spectrometer (see supplementary data, Figure S1), which is located
upstream of the CID trap. In this particular design, the parent ion is first isolated in the
linear iontrap then ejected back to the Q00 region where its fragmentation occurs. Then, the
generated fragments are directed toward the linear iontrap where they are analyzed. In this
process, it is easy to understand that the solvent vapors can enter the HCD cell making
possible the reaction between [NAT+H]+ and ACN to produce this original fragmentation
pathway. The fragmentation pathway can be summarized as follows ([14N-NAT+H]+ is
taken as example):
 In CID, m/z 170 losses N2 to give m/z 142 which in turn loses HCN to give m/z 115.
This pathway was checked by MS3 which allows the observation of the transition m/z
142 → 115.
 In HCD, m/z 170 reacts or interacts with ACN to give m/z 211 ([14N-NAT+H+ACN]+).
Then m/z 211 undergoes two consecutive fragmentations, one loss of N2 (m/z 183) and
one loss of HCN, to give the original m/z 156. Unfortunately, the signal intensities of
m/z 211 and 183 are too weak to check this pathway by MS3 but interestingly, these
neutral losses are precisely the same (and in the same order) as observed from m/z 170
in CID. The last step consists in the loss of the solvent molecule from m/z 156 to restore
m/z 115 also observed in CID. This final pathway was checked by an MS3 experiment
in CID at 35 eV, which allows the observation of the transition m/z 156 → 115. The ion
m/z 156 can be somehow described as the ACN adduct of m/z 115 [C9H7+ACN]+, then
m/z 183 is corresponding to the ACN adduct of m/z 142 (in CID) but the precursor of
these fragment ion m/z 211 ([14N-NAT+H+ACN]+) is not observed in the source (see
supplementary data, Figure S2) meaning that the ACN adduct is only formed thank to
the HCD process.
The reaction mechanism remains unclear at this time and its elucidation would deserve
future investigation but the combination of ACN in the mobile phase and the HCD mode
allowed us to reach a limit of quantification of 1 nM for the NAT adduct, by following the
This article is protected by copyright. All rights reserved.

15
new “indirect” SRM transition (i.e. m/z 171 to 156 for N-NAT via the acetonitrile adduct
m/z 212 formation in the HCD cell). For the transition m/z 171 to 115, the LOQ was 10 nM in
HCD mode and only 20 nM in CID (best achievable results in CID). This value was within
15
the same range as in previous studies for the measurement of N-nitrite by LC-MS/MS25,26
.
The chosen transitions for each compound (including the IS, i.e. the 2-naphthylamine) are
reported in the experimental section. The method was then validated to monitor with
accuracy the NO species freed from GSNO in an in vitro model of intestinal barrier at trace
levels.
Interferences in blank samples and working LOQ
The final objective of this study is to assess the permeability of GSNO across the
intestinal barrier by means of an in vitro model (Caco-2 monolayer). During the incubation,
Caco-2 cells secreted many metabolites into the buffer, including unlabeled NO species.
Thus, the incubation medium turned into a complex biological matrix, which required a high
specific or selective method of analysis. In this field, LC-MS/MS has become the standard
technique, being much more specific, selective and even much more sensitive than most other
conventional spectroscopic techniques (UV and even fluorescence). The selectivity has been
tested by analyzing corresponding blanks (untreated cells).
15
When developing a reliable quantitative analysis of N-nitrite, there is a clear risk of
interference between 15N-NAT and 14N-NAT containing one 13C atom (due to its natural
abundance of 1.1%). Chao et al24 showed a 2.2%-interference due to 13C, which was in good
agreement with what is theoretically expected. We did not observe any interference when the
quantification was performed using fresh reactants and culture medium. Thus, a LOQ of 1
nM was obtained LC-ITMS/MS which is comparable to the LOQ obtained by Axton et al. on
their Q-Trap25. However, ambient nitrogen oxides (NOx) are rapidly able to dissolve into
them, leading to an interference in the experimental blanks (Figure 5). This interference was
estimated to be around 50 to 60 nM in 14NOx (using the transition m/z 170 to 156), which is
15
equivalent to the signal of 1 nM of N-NAT for the corresponding transition (m/z 171 to
This article is protected by copyright. All rights reserved.

156). Consequently, the working LOQ was set at 5 nM for the validation step, taking in
account the signal-over-blank ratio rather than the signal-over-noise one.
Nevertheless, this concentration level is satisfactory enough for the accurate quantification of
nitrites in the real samples.
Method validation
The next step of this study was the investigation of the linearity range, reproducibility,
accuracy, precision, interferences, percent recovery, and carryover of the developed LC-
ITMS/MS method.
The linearity range and the LOQ were determined from 3 standard curves of nitrite ions.
The LOQ was limited to 5 nM, as described in the previous section. The calibration curve of
15N-NAT adduct was linear between 5 and 200 nM for Na15NO2 and GS15NO and between
250 and 5000 nM for Na15NO3, with the coefficient of determination (r²) higher than 0.9956.
The intra- and inter-day variabilities of the method were evaluated by using quality control
samples and the results are summarized in Table 2. Three concentrations of Na15NO2 and
GS15NO (15, 50, 180 nM), and of Na15NO3 (300, 2500, 4000 nM) were spiked into HBSS (n
= 6), and the concentration was calculated by using a six-point calibration curve for Na15NO2
and GS15NO (5, 10, 20, 50, 100, 200 nM), and for Na15NO3 (250, 500, 1250, 2500, 5000 nM).
The accuracy was determined as the deviation (%) of the calculated concentrations from the
nominal concentrations. The precision was determined as the relative standard deviation
(RSD) of the six measurements. Satisfactory accuracies were obtained with intra- and inter-
day bias lower than 5.2 % and 5.3 % for Na15NO2, 1.9 % and 6.7 % for GS15NO, and 7.7 %
and 9.4 % for Na15NO3, respectively, over the studied concentration ranges. The intra- and
inter-day precision ranged from 3.0 to 7.1 % for Na15NO2, from 7.7 to 13.4 % for GS15NO
and from 5.6 to 8.7 % for Na15NO3 in HBSS. These ranges were within the criteria stated in
the FDA guidance on bioanalytical method validation (Table 2).
This article is protected by copyright. All rights reserved.

Additionally, the product NAT was found stable after storage at -80 °C for 3 months and
freezing/thaw for 2 cycles (data not shown).
Intestinal permeability of GS15NO in vitro
To evaluate the intestinal permeability of nitrate, nitrite and RSNOs, we needed to
determine their concentration in the apical and basolateral compartments after incubation
(Figure 6). To the best of our knowledge, for the first time, one single derivatization
technique is used to determine the three considered analytes. In our approach, all GSNO
metabolites, nitrites, nitrates and RSNOs, were derivatized to the same compound (NAT) and
then quantified by LC-ITMS/MS. Free nitrites were measured first. The nitrate concentration
was then determined after enzymatic reduction of nitrate ions subtracting the concentration of
free nitrites. Similarly, RSNOs concentrations are calculated by subtracting the concentration
of free nitrites from the sum “RSNOs + nitrites” measured after the S-NO bond cleavage with
mercuric ions. In the apical compartment, the decomposition of GSNO into nitrites and
nitrates was also checked by direct HPLC-UV as proposed by Parent et al for concentrations
in the micromolar range30. GSNO was found to be stable under the present operating
conditions (data not shown). Nevertheless, in the basolateral compartment, GSNO
degradation could not be investigated by HPLC-UV because of the very low measured
concentrations (nanomolar range). In this compartment, the transfer of the NO moiety of
GSNO on cysteine residues of peptides and proteins is possible, resulting in a formation of
other RSNOs. That is the reason why, in the basolateral compartment, it is preferable to use
the term RSNOs rather than GSNO to take into account all possible compounds reacting with
mercuric ions. Nevertheless, whatever the GS15NO fate in the basolateral compartment, it is
clear that the 15N-labeled species analyzed by our LC-ITMS/MS method originates from this
drug.
The TEER values measured before and after 1 h of GS15NO permeation through the
Caco-2 cell monolayer was higher than 500 Ω·cm-2. Meanwhile, the permeability percentage
of the sodium fluorescein was less than 5 %. These results indicated a confluent and
unalterated monolayer.
This article is protected by copyright. All rights reserved.

In the Biopharmaceutics Classification System31, drugs are assigned within four classes
depending on their aqueous solubility and intestinal permeability. Three classes of intestinal
permeability are defined according to drug Papp value: high permeability compounds such as
propranolol show Papp value higher than 10-5 cm·s-1, while low permeability compounds such
as furosemide show Papp value less than 1·10-6 cm·s-1 32. Intermediate values correspond to
15
“middle permeability” drugs. It was found that N-nitrite has a Papp value in the range of
15
(0.32 – 0.47)×10-6 cm·s-1, RS15NOs in the range of (0.16 – 0.35)×10-6 cm·s-1, while N-
nitrate is in the range of (1.65– 1.79)×10-6 cm·s-1, which demonstrated that GS15NO belongs
to the class of middle permeability compounds. In addition, GSNO showed concentration-
dependent distribution in both compartments in the tested concentration range (10-100 µM)
(Figure 6), and concentration-independent permeability. Noteworthy, the present
LC/ITMS/MS method allows us to reach the low concentrations of NO-related species in the
basolateral compartment, even at the lowest dose of GSNO applied to the cellular barrier
model (10 µM). Applying fluorescence with the same derivatization process (DAN) limits the
permeability studies to the highest concentration of GSNO (100 µM). High levels of NO
species (nitrite, nitrate and RSNOs) in the blank demonstrate the necessity of using isotope
labeling to improve the method LOQ, and sample dilution contributes to a lower risk of
matrix interference.
CONCLUSIONS
We have developed and validated a LC-ITMS/MS method capable of selectively
15
monitoring NO species metabolized from N labeled GSNO as NAT adducts using an Ion
Trap Mass Spectrometer. Compared with other published approaches, this method combines
several original features such as (i) the use of a single derivatization technique for the three
considered analytes (nitrites, nitrates and RSNOs), (ii) a new described fragmentation pattern
that is at least as sensitive as (sometime even more sensitive than) the previously described
methods (in the nM range)24,25,26 on triple quadrupole mass spectrometers, and (iii) the first
report of a 15N labeling method applied to an intestinal barrier model.
This article is protected by copyright. All rights reserved.

The IT mass spectrometer enables us to reach a very satisfying sensitivity thanks to an
original “indirect” MS/MS transition (m/z 171 to 156 via an acetonitrile adduct m/z 212
formation) obtained in Higher-energy Collision Dissociation with a limit of quantification of
1 nM on 15NO freed from GS15NO. The working LOQ was set at 5 nM to overcome
interferences produced by 14NO present in the real samples. The assay, applied to the study of
the intestinal permeability of GS15NO with an in vitro model of this physiological barrier,
classifies the S-nitrosoglutathione in the middle permeability class according to FDA
guidelines.
The work was the preamble of future assays of NO species metabolized from labeled
GS15NO or other NO donors in the cells or tissues. It also opens the perspective of a more
fundamental work such as the study of the reactivity of the ion [NAT+H]+ in HCD tandem
mass spectrometry in the presence of acetonitrile.
Furthermore, the use of in vitro models of intestinal permeability is of key importance to
evaluate the drug (GSNO) bioavailability after oral administration. To the best of our
knowledge, this work provides the first report of a 15N-labeled drug using an intestinal
permeability model. In the future, the transposition of our approach to other biological
matrices (plasma, tissues, organs like aorta) using sample preparation protocols already
described33,34 constitutes a natural extension of the fields of application.
Acknowledgements
Authors are grateful to the program of China Scholarship Council for the financial support of
Ms Haiyan Yu’s PhD thesis (grant NO. 201506270166). Authors also thank Dr Philippe
Giummelly (CITHEFOR, Université de Lorraine) for the synthesis of GS15NO.
REFERENCES
1 Moncada S, Palmer RM, Higgs EA. Nitric oxide: physiology, pathophysiology, and
pharmacology. Pharmacol Rev. 1991;43(2):109–142.
This article is protected by copyright. All rights reserved.

2 Cooke JP, Dzau VJ. Nitric oxide synthase: Role in the genesis of vascular disease. Annu
Rev Med. 1997;48:489–509.
3 Vallance P, Leone A, Calver A, Collier J, Moncada S. Accumulation of an Endogenous
Inhibitor
of
Nitric-Oxide
Synthesis
in
Chronic-Renal-Failure.
Lancet.
1992;339(8793):572–575.
4 Lundberg JO, Weitzberg E, Gladwin MT. The nitrate-nitrite-nitric oxide pathway in
physiology and therapeutics. Nat Rev Drug Discov. 2008;7(2):156–167. Doi:
10.1038/nrd2466.
5 Loo B, Labugger R, Skepper JN, et al. Enhanced Peroxynitrite Formation Is Associated
with
Vascular
Aging.
J
Exp
Med.
2000;192(12):1731–1744.
Doi:
10.1084/jem.192.12.1731.
6 Münzel T, Steven S, Daiber A. Organic nitrates: Update on mechanisms underlying
vasodilation, tolerance and endothelial dysfunction. Vascul Pharmacol. 2014;63(3):105–
113. Doi: 10.1016/j.vph.2014.09.002.
7 Csont T, Ferdinandy P. Cardioprotective effects of glyceryl trinitrate: beyond vascular
nitrate
tolerance.
Pharmacol
Ther.
2005;105(1):57–68.
Doi:
10.1016/j.pharmthera.2004.10.001.
8 Broniowska KA, Diers AR, Hogg N. S-Nitrosoglutathione. Biochim Biophys Acta. BBA -
Gen Subj. 2013;1830(5):3173–3181. Doi: 10.1016/j.bbagen.2013.02.004.
9 Gaucher C, Boudier A, Dahboul F, Parent M, Leroy P. S-nitrosation/Denitrosation in
Cardiovascular Pathologies: Facts and Concepts for the Rational Design of S-nitrosothiols.
Curr Pharm Des. 2013;19(3):458–472.
10Wu W, Perrin-Sarrado C, Ming H, et al. Polymer nanocomposites enhance S-
nitrosoglutathione intestinal absorption and promote the formation of releasable nitric
oxide stores in rat aorta. Nanomedicine-Nanotechnol Biol Med. 2016;12(7):1795–1803.
Doi: 10.1016/j.nano.2016.05.006.
11Chan OH, Stewart BH. Physicochemical and drug-delivery considerations for oral drug
bioavailability. Drug Discov Today. 1996;1(11):461–473.
This article is protected by copyright. All rights reserved.

12Badhan R, Penny J, Galetin A, Houston IB. Methodology for Development of a
Physiological Model Incorporating CYP3A and P-Glycoprotein for the Prediction of
Intestinal Drug Absorption. J Pharm Sci. 2009;98(6):2180–2197. Doi: 10.1002/jps.21572.
13Artursson P. Cell-Cultures as Models for Drug Absorption Across the Intestinal-Mucosa.
Crit Rev Ther Drug Carrier Syst. 1991;8(4):305–330.
14Alqahtani S, Mohamed LA, Kaddoumi A. Experimental models for predicting drug
absorption and metabolism. Expert Opin Drug Metab Toxicol. 2013;9(10):1241–1254.
Doi: 10.1517/17425255.2013.802772.
15Miersch S, Mutus B. Protein S-nitrosation: Biochemistry and characterization of protein
thiol–NO interactions as cellular signals. Clin Biochem. 2005;38(9):777–791. Doi:
10.1016/j.clinbiochem.2005.05.014.
16Bryan NS, Grisham MB. Methods to detect nitric oxide and its metabolites in biological
samples.
Free
Radic
Biol
Med.
2007;43(5):645–657.
Doi:
10.1016/j.freeradbiomed.2007.04.026.
17Marzinzig M, Nussler AK, Stadler J, et al. Improved Methods to Measure End Products of
Nitric Oxide in Biological Fluids: Nitrite, Nitrate, andS-Nitrosothiols. Nitric Oxide.
1997;1(2):177–189. Doi: 10.1006/niox.1997.0116.
18Tsikas D. Methods of quantitative analysis of the nitric oxide metabolites nitrite and nitrate
in human biological fluids. Free Radic Res. 2005;39(8):797–815. Doi:
10.1080/10715760500053651.
19Schmidt TC, Zwank L, Elsner M, Berg M, Meckenstock RU, Haderlein SB. Compound-
specific stable isotope analysis of organic contaminants in natural environments: a critical
review of the state of the art, prospects, and future challenges. Anal Bioanal Chem.
2004;378(2):283–300. Doi: 10.1007/s00216-003-2350-y.
20Tsikas D. Pentafluorobenzyl bromide—A versatile derivatization agent in chromatography
and mass spectrometry: I. Analysis of inorganic anions and organophosphates. J
Chromatogr B. 2017;1043:187–201. Doi: 10.1016/j.jchromb.2016.08.015.
This article is protected by copyright. All rights reserved.

21Hanff E, Boehmer A, Jordan J, Tsikas D. Stable-isotope dilution LC-MS/MS measurement
of nitrite in human plasma after its conversion to S-nitrosoglutathione. J Chromatogr B-
Anal Technol Biomed Life Sci. 2014;970:44–52. Doi: 10.1016/j.jchromb.2014.08.041.
22Li H, Meininger CJ, Wu G. Rapid determination of nitrite by reversed-phase high-
performance liquid chromatography with fluorescence detection. J Chromatogr B Biomed
Sci App. 2000;746(2):199–207. Doi: 10.1016/S0378-4347(00)00328-5.
23Marzinzig M, Nussler AK, Stadler J, et al. Improved methods to measure end products of
nitric oxide in biological fluids: Nitrite, nitrate, and S-nitrosothiols. Nitric Oxide-Biol
Chem. 1997;1(2):177–189. Doi: 10.1006/niox.1997.0116.
24Chao MR, Shih YM, Hsu YW, et al. Urinary nitrite/nitrate ratio measured by isotope-
dilution LC–MS/MS as a tool to screen for urinary tract infections. Free Radic Biol Med.
2016;93:77–83. Doi: 10.1016/j.freeradbiomed.2016.01.025.
25Axton ER, Hardardt EA, Stevens JF. Stable isotope-assisted LC–MS/MS monitoring of
glyceryl trinitrate bioactivation in a cell culture model of nitrate tolerance. J Chromatogr
B. 2016;1019:156–163. Doi: 10.1016/j.jchromb.2015.12.010.
26Shin S, Fung HL. Evaluation of an LC–MS/MS assay for 15N-nitrite for cellular studies of
l-arginine action.
J
Pharm Biomed Anal. 2011;56(5):1127–1131. Doi:
10.1016/j.jpba.2011.08.017.
27Saville B. A scheme for the colorimetric determination of microgram amounts of thiols.
Analyst. 1958;83(993):670–2. Doi: 10.1039/AN9588300670.
28Parent M, Dahboul F, Schneider R, et al. A Complete Physicochemical Identity Card of S-
nitrosoglutathione. Curr Pharm Anal. 2013;9(1):31–42.
29Jobgen WS, Jobgen SC, Li H, Meininger CJ, Wu G. Analysis of nitrite and nitrate in
biological samples using high-performance liquid chromatography. J Chromatogr B.
2007;851(1-2):71–82. Doi: 10.1016/j.jchromb.2006.07.018.
30Parent M, Dahboul F, Schneider R, Clarot C, Maincent P, Leroy P, Boudier A. A
Complete Physicochemical Identity Card of S-Nitrosoglutathione. Current Pharmaceutical
Analysis. 2013;9:31–42.
This article is protected by copyright. All rights reserved.

31U.S. Department of Health and Human Services Food and Drug Administration Center for
Drug Evaluation and Research (CDER). Waiver of in vivo bioavailability and
bioequivalence studies for immediate-release solid oral dosage forms based on a
biopharmaceutics classification system. Biopharmaceutics. 2015.
32Peng Y, Yadava P, Heikkinen AT, Parrott N, Railkar A. Applications of a 7-day Caco-2
cell model in drug discovery and development. Eur J Pharm Sci. 2014;56:120–130. Doi:
10.1016/j.ejps.2014.02.008.
33 Warnecke A, Luessen P, Sandmann J, et al. Application of a stable-isotope dilution
technique to study the pharmacokinetics of human N-15-labelled S-nitrosoalbumin in the
rat: Possible mechanistic and biological implications. J Chromatogr B-Anal Technol
Biomed Life Sci. 2009;877(13):1375–1387. Doi: 10.1016/j.jchromb.2008.11.035.
34Damacena-Angelis C, Oliveira-Paula GH, Pinheiro LC, et al. Nitrate decreases xanthine
oxidoreductase-mediated nitrite reductase activity and attenuates vascular and blood
pressure
responses
to
nitrite.
Redox
Biol.
2017;12:291–299.
Doi:
10.1016/j.redox.2017.03.003.
This article is protected by copyright. All rights reserved.

14
15
Table 1: Ion composition of the MS/MS spectra of the N-NAT or N-NAT adduct in CID
(35 eV) and HCD (50 eV) depending on the organic modifier composing the
chromatographic mobile phase (solvent S). The NAT adduct was infused in a mixture of a 10
mM-acetic acid aqueous solution containing 40% of the solvent S. The relative abundance
(%) of each ion is indicated in parentheses. Nd: not detected
Observed m/z for [NAT+H]+ precursor ion
Label MS/ Solv.
ing
MS
Sa
[NAT+
H]+
[NAT-
[C9H [NAT+
[NAT-
[NAT-N2-HCN
N2+H]+
]
H+S]+ N2+H+S]+ +H+S]+
+b
7
MeO
H
115(1
00)
170(38)
142(19)
Nd
Nd
Nd
Nd
Nd
Nd
CID
(35eV)
115(1
00)
14N-
NAT
ACN 170(36)
ACN 170(46)
142(18)
115(1
00)
Nd
211(0.1)
183(1.1)
156(46)
HC
D
d3-
115(1
00)
(50eV)
170(45)
ACN
Nd
214(0.1)
Nd
186(1.1)
Nd
159(45)
Nd
MeO
H
115(1
00)
c
171(38) 142(19)
CID
(35eV)
115(1
00)
c
ACN 171(36) 142(15)
Nd
Nd
Nd
15N-
NAT
115(1
c
c
ACN 171(45)
Nd
Nd
212(0.1)
183(1.1)
186(1.0)
156(40)
159(42)
HC
00)
115(1
00)
D
d3-
c
c
(50eV)
171(46)
215(0.2)
ACN
Remar a: The NAT adduct is infused in a mixture of a 10 mM-acetic acid aqueous solution containing
ks:
40% of solvent S (MeOH, ACN or d3-ACN)
b: This ion is also corresponding to [NAT-N2-HCN+H]+ (see also Axton et al25 and also Shin et
al26)
c: For these ions, the loss is 29 u because N2 is containing the labeled nitrogen (15N).
This article is protected by copyright. All rights reserved.

Table 2. Inter-day and intra-day variability of quality control samples for 15N-nitrite S-
15
15
15
nitrosoglutathione (GS15NO) and N-nitrate analysis with IS. N-nitrite, GS15NO and N-
nitrate were spiked into HBSS buffer (n = 6), and the concentration was calculated by using
related six-point calibration curve (5, 10, 20, 50, 100, 200 nM) of 15N-nitrite and S-
nitrosoglutathione (GS15NO), and five-point calibration curve (250, 500, 1250, 2500, 5000
nM) of 15N-nitrite. Accuracy was determined as the deviation (%) of calculated
concentrations from the nominal concentrations. Precision was determined as the relative
standard deviation (RSD) of the 6 measurements.
Intra-day
Inter-day
NO
species
Concentration
(nM)
Accuracy
Precision
(%)
Accuracy
Precision
(%)
(%)
(%)
15
50
105.2±6.4
99.0±6.8
6.1
6.9
104.7±7.0
94.7±2.8
6.7
3.0
15N-nitrite
GS15NO
180
15
98.6±5.3
5.4
102.7±7.3
93.3±11.8
96.0±7.4
7.1
98.1±13.1
99.1±10.9
100.0±12.7
103.0±6.9
100.8±5.6
107.7±8.5
13.4
11.0
12.7
6.7
12.6
7.7
50
180
300
2500
4000
101.5±11.4
103.0±7.0
100.6±8.7
109.4±8.8
11.3
6.8
15N-nitrate
5.6
8.7
7.9
8.1
This article is protected by copyright. All rights reserved.

Figure 1. Principal reactions involved in this study. A) Derivatization of nitrite ion with 2,3-
diaminonaphthalene (DAN) leading to a 2,3-naphthotriazole (NAT) derivative. The nitrogen
transferred from the nitrite ion is indicated with an asterisk (*); B) Release of nitrites from the
S-nitrosoglutathione (GSNO) catalyzed by mercuric ions and C) Reaction of nitrate
conversion into to nitrite ion catalyzed by a nitrate reductase in presence of cofactor (here
NADPH = nicotinamide adenine dinucleotide phosphate).
This article is protected by copyright. All rights reserved.

1) DAN derivatization
2) LC-ITMS/MS
-
 15NO2
subtraction of 
-

15NO2
+
-
15NO2
1) Nitrate reductase
2) DAN derivatization
3) LC-ITMS/MS
-
15NO3
-
15NO3
15NO3-
RS15NO

GS15NO metabolites
-

15NO2
+
1) Treatment with Hg2+
2) DAN derivatization
3) LC-ITMS/MS
RS15NO
RS15NO

Figure 2. Strategy of analysis of the different GS15NO metabolites obtained during the in
15
vitro study using the same LC-ITMS/MS approach. The concentration of free N-nitrites 
after incubation is directly obtained after DAN derivatization followed by LC-ITMS/MS.
Free 15N-nitrates are obtained after enzymatic (nitrate reductase) pretreatment of the
metabolite pool. The quantification by LC-MS/MS after DAN derivatization of this sample
provides the cumulative concentration  of free 15N-nitrites + 15N-nitrates (reduced into
nitrites). The concentration in free 15N-nitrates  is deduced from result  by subtracting the
concentration of free 15N-nitrites . The concentration of RS15NOs is obtained after the
pretreatment of the metabolite pool with Hg2+. The quantification by LC-MS/MS after DAN
15
derivatization of this sample provides the cumulative concentration  of free N-nitrites +
GS15NO (converted into nitrites, see reaction 2 in figure 1). The concentration of RS15NO 
is deduced from result  by subtracting the concentration of free 15N-nitrites .
This article is protected by copyright. All rights reserved.

100
80
60
40
20
0
NAT
2-NA
(IS)
0
1
2
3
4
5
6
7
Time (min)
Figure 3. Representative LC-ESI-ITMS/MS chromatogram (conditions given in the
experimental section) of the 2,3-naphthotriazole (NAT) derivative and the internal standard
(IS, 2-naphthylamine (2-NA)).
This article is protected by copyright. All rights reserved.

115
100
90
[C9H7]+
80 A) CID of 14N-NAT with ACN
70
60
50
40
30
20
10
0
[
14N-NAT+H]+
170
142
14N2
-HCN
-
50
60
70
80
90
100
110
120
130
140
m/z
150
160
170
180
190
200
210
220
115
100
90
80
70
60
50
40
30
20
10
0
B) HCD of 14N-NAT with ACN
-HCN
170
156
[14N-NAT+H+ACN]+
211
-
14N2
183
180
141
140
m/z
50
60
70
80
90
100
110
120
130
150
150
150
160
170
190
200
210
220
115
100
90
80
70
60
50
40
30
20
10
0
-HCN
C) HCD of 14N-NAT with d3-ACN
170
159
[
14N-NAT+H+d3-ACN]+
214
-
14N2
141
140
186
50
60
70
80
90
100
110
120
130
160
170
180
190
200
210
220
m/z
115
100
90
80
70
60
50
40
30
20
10
0
D) HCD of 15N-NAT with ACN
-HCN
171
156
[15N-NAT+H+ACN]+
212
- (14N15N)
141
140
183
180
50
60
70
80
90
100
110
120
130
160
170
190
200
210
220
m/z
Figure 4. Fragmentation of the protonated molecular ion of NAT diluted at 2 µM in the
chromatographic mobile phase. A) CID at 35 eV of [14N-NAT+H]+ (m/z 170) using
acetonitrile (ACN), B) HCD at 50 eV of [14N-NAT+H]+ using ACN, C) HCD at 50 eV of
[14N-NAT+H]+ using deuterated acetonitrile (d3-ACN) and D) HCD at 50 eV of [15N-
NAT+H]+ (m/z 171) using ACN. See also Table 1 for complementary information.
This article is protected by copyright. All rights reserved.

1200
1100
1000
900
800
700
600
500
400
300
200
100
0
5 nM of 15N-nitrite
Blank
Fresh buffer
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
Time (min)
Figure 5. Representative ion chromatograms under selected reaction monitoring (SRM)
15
transition of m/z 171 → 156, corresponding to derivatization of 5 nM N-nitrite in HBSS
buffer, blank of the in vitro study and freshly prepared HBSS buffer.
This article is protected by copyright. All rights reserved.

Figure
6. Permeation of 10, 25, 50 or 100 µM of 15N labeled S-nitrosoglutathione (GS15NO) through
an in vitro model of intestinal barrier (differentiated Caco-2 cell monolayer) during 1 h at 37
C. GS15NO metabolites (nitric oxide (NO) species concentrations) were quantified in both
compartments, n = 3. Results are expressed as mean ± SD.
This article is protected by copyright. All rights reserved.