Identification of by-products in support of process development of Muraglitazar

Full text of DOI:10.1002/mrc.4412

Special issue letter - application note
Received: 12 August 2015
Revised: 18 November 2015
Accepted: 7 January 2016
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/mrc.4412
Identification of by-products in support of
process development of Muraglitazar
Charles Pathirana,a* Andrew Rusowicz,a Russell Suda,a
Shankar Swaminathana,b and Venkatapuram Palaniswamya,c
Muraglitazar was being developed by Bristol-Myers Squibb for the treatment for type 2 diabetes and dislipidemia. Process
optimization included the minimization of the by-products. This endeavor was greatly facilitated by a clear understanding of
by-product identity. By-products were isolated by preparative chromatography and identified using NMR and MS. The identified
structures of the by-products provided useful information about the undesired side reactions, which were then minimized or
eliminated by altering the reaction conditions appropriately. Copyright © 2016 John Wiley & Sons, Ltd.
Keywords: NMR; 1H NMR; 13C NMR; MS; muaraglitazar; process development; by-products; structural elucidation
are carried over to the API. Identification of a few selected impuri-
ties encountered in the synthesis of muraglitazar is described here.
Introduction
Muraglitazar (1) was being developed by Bristol-Myers Squibb as a
peroxisome proliferator-activated receptor α/γ dual agonist for the
treatment of type 2 diabetes and dislipidemia (Fig. 1).[1–4] 1 was
being synthesized starting from an appropriately substituted oxazole
as shown in Scheme 1.[5,6] During this stage of process development,
characterization of intermediates and the final product is routinely
carried out. In addition, identification of the by-products observed is
also performed routinely for two reasons. First, the identification of
by-products in starting materials, isolated intermediates, and active
pharmaceutical ingredient (API) is required to fulfill regulatory require-
ments. Second, identification of by-products is also carried out to help
optimize the process and to achieve increased overall efficiency and
quality. Identification of any by-product provides increased knowl-
edge of subtle reactant–reagent interactions that may provide in-
sights into how to better control the reaction to achieve the desired
outcome. Therefore, it is crucial to identify the major by-products that
will enable the chemist to make adjustments to minimize their pro-
duction and increase the yield of the desired product. NMR plays a
key role in the identification of by-products and in combination with
mass spectrometry (MS) has been shown to be a comprehensive
approach for complete structural elucidation.[7]
By-product 4 was discovered from the API as the impurity with 1.4
relative retention time (RRT). By-products 9 and 10 were observed
at 1.1 and 0.9 RRT, respectively, in the pre-penultimate 2. By-
product 11 was discovered at 2.2 RRT in the penultimate 3.
Identification of by-product 4
The synthesis of muraglitazar (1) is shown in Scheme 1. An impurity
with a relatively high molecular weight of 980 Da at RRT 1.4 was
observed in the API. The high molecular weight and relatively
non-polar nature of the impurity suggested that this impurity is
resulting from an extended reaction during the final or a previous
synthetic step. The molecular formula of the isolated impurity was
determined to be C54H49N3O15 using electrospray-ionization
high-resolution mass spectrometry (ESI HRMS) [(M + H)+: observed
– 980.3231, calculated – 980.3242]. A ketone carbonyl carbon at δ
189.6 could be assigned using the 1H-13C HMBC correlations arising
from the protons ortho to the aromatic carbon it is attached to at δ
8.08. The entire fragment containing the ketone and the
monosubstituted phenyl ring was easily identified by comparing
the NMR data of 4 and 1 (Table 1). The acetal methine was assigned
using the 1H (δ 7.03, s) and 13C (δ 97.5, CH) chemical shifts. A corre-
lation from this proton to the ketone carbonyl carbon was missing.
However, three-bond correlations through oxygen to the two
The following describes the combined utilization of NMR and MS
to achieve complete identification of several by-products during
process optimization of muraglitazar (1).
Results and discussion
*
Correspondence to: Charles Pathirana, Pharmaceutical Development, Bristol-
Myers Squibb, New Brunswick, NJ 08903, United States. E-mail: charles.
pathirana@bms.com
By-products are generally encountered during process develop-
ment for the synthesis of API. International Conference on Harmo-
nization guidelines require knowledge of all the impurities present
in the API at or above 0.1% level. Impurities that fall into this cate-
gory are selected for identification. The level of the impurities is
determined as an area percent from the analysis performed using
reversed phase HPLC. The impurities that are selected for identifi-
cation can be from the API and also from the intermediates if they
a Pharmaceutical Development, Bristol-Myers Squibb, New Brunswick, NJ, 08903,
United States
b Current address: Valeant Pharmaceuticals, 400 Somerset Corporate Blvd,
Bridgewater, NJ, 08807, United States
c
Current address: Novatia, 54 Walker Lane, Newtown, PA, 18940, United States
Magn. Reson. Chem. (2016)
Copyright © 2016 John Wiley & Sons, Ltd.

C. Pathirana et al.
impurity. It may react with p-hydroxybenzaldehyde to afford 6,
which appears to be stabilized by the extra carbonyl group
attached to the acetal carbon. If 6 is present during the synthesis
of 2 as shown in Scheme 1, compound 7 would be generated
instead of 2. Compound 7 would afford compound 8 if carried
through the next step. Evidence for these transformations was
observed in the LC–MS analysis of 2 and 3. LC–MS analysis of 2
showed a chromatographic peak of a substance possessing a molec-
ular weight of 561 Da [(M + H)+ 562.3 Da] consistent with the presence
of 7 in 2. LC–MS analysis of 3 showed a chromatographic peak with a
molecular weight of 707 Da [(M + H)+ 708.6 Da] consistent with the
presence of the corresponding impurity 8 in 3. The particular lot of
p-hydroxybenzaldehyde could not be analyzed for the presence of
5 or 6 because it was consumed in the reaction.
Figure 1. Structure of muraglitazar (1).
aromatic carbons at δ 155.1 attached to acetal oxygens were ob-
served. The placement of the acetal carbon next to the ketone car-
bon was supported by the Liquid chromatography–tandem mass
spectrometry (LC–MS/MS) fragments at m/z 649 and 621 (Fig. 2).
Because of restricted rotation in the carbamate moiety, rotamers
were observed, indicated by the doubling in the resonances be-
longing to protons and carbons around the carbamate moiety.
Most prominent were the two methylene protons attached to the
carbamate nitrogen. The chemical shifts of these CH2 groups in
one rotamer were δ 4.01 and 4.43 (H-28, both singlets) and δ 3.92
and 4.56 (H-27, both singlets) in the other. Protons on both of these
CH2 groups showed three-bond HMBC correlations to C-30 at δ 155.7
and 155.1. HMBC spectrum also showed a correlation from H-28 to C-
29 at δ 170.7 and 171.0. Correlations from H-32/36 and H-37 to C-34
and from H-33/35 to C-31 established the p-methoxyphenoxy moiety
attached to the carbamate. Furthermore, the fragment containing
C-28 to C-37 in 4 was confirmed when its NMR data matched very
closely with the NMR data of the fragment containing C-20 to C-29
including C-1 in 1. The link between this fragment in 4 to the rest
of the molecule was revealed by the correlations from H-27 to
C-24 and C-23/25. All of the NMR data were assigned using results
from 1D (1H and 13C) and 2D (COSY, 1H-13C HMQC, 1H-13C HMBC)
experiments and are included in Table 1. Fragmentation in LC–
MS/MS provided further support for the assigned structure (Fig. 2).
Identification of by-products 9 and 10
Examination of Scheme 1 suggests that contaminant 5 in the
p-hydroxybenzaldehyde could be a plausible source of this
Two of the impurities identified from 3 indicated that excess gly-
cine methylester was responsible for their formation. Compound
Scheme 1. Synthesis of muraglitazar (1).
wileyonlinelibrary.com/journal/mrc
Copyright © 2016 John Wiley & Sons, Ltd.
Magn. Reson. Chem. (2016)

Identification of by-products by NMR in process development
Table 1. NMR assignments of 1 and 4
1
4
Assignment
δ H
4.43, bs, 2H* 4.56, bs*
—
7.25, d,8.5, 2H* 7.28,d,8.5*
δ C
δ H
—
—
δ C
189.6
126.2
1
50.8, 50.9*
129.1, 129.3*
129.4
2
3, 7
4, 6
5
8.08, d, 8.7, 2H 132.2
6.91, m, 2H
114.4, 114.5c* 7.10, d, 8.7, 2H 114.9
—
4.19, t,6.8, 2H* 4.20, t,6.8*
157.7
66.2
—
163.4
67.0
8
4.33, t, 6.3, 2H
4 (contd.)
9
2.91, t,6.8, 2H
25.6
132.7
2.96, d, 6.3, 2H 25.6 Assignment
δ H
7.03, m, 1H
δ C
97.5
10
—
—
—
132.6 20
11
—
145.1
145.5 21
—
155.1
12
2.34, s, 3H
9.8
2.35, s, 3H
—
10.1 22, 26
158.7 23, 25
127.3 24
7.03, m, 4H
116.9, 117.0*
129.3
13
—
158.4
7.28, d, 8.1, 2.7H* 7.32, d, 8.1, 1.3H*
—
14
—
7.90, d,7.9, 2H
7.46, m, 2H
7.46, m, 1H
3.92, bs, 2H* 3.99, bs
—
127.1
—
131.9
51.0, 51.1*
15, 19
16, 18
17
125.4
7.90, d, 2.0, 2H 125.7 27
4.43, br s, 2.7H* 4.56, br s, 1.3H*
129.0
7.48, m, 2H
7.48, m, 1H
129.6
130.3
130.0
20
48.3, 48.4*
170.4, 170.7*
154.5, 154.8*
144.5, 144.6*
28
3.92, br s, 1.3H* 4.01, br s, 2.7H*
48.9, 49.0*
170.7, 171.0*
154.7, 155.1*
144.7, 144.8*
122.7, 122.8*
114.4
21
29
—
22
—
30
—
23
—
31
—
6.96, m, 4H
24, 28
25, 27
26
6.99, d,8.8, 2H* 7.00, d,8.8 122.6, 122.5*
32, 36
33, 35
34
6.91, m, 2H
—
3.72, bs, 3H*3.73, bs*
114.2c
156.6
55.4
6.90, m, 4H
—
3.72, br s, 2H* 3.73, br s, 4H*
156.8
29
37
55.6
δ ppm, relative to DMSO-d5 at δ 2.49 for 1H and DMSO-d6 at δ 39.5 for 13C. Some resonances are doubled because of the presence of rotamers in the
carbamate moiety. The two chemical shifts arising from rotamers where observed are denoted with an asterisk. The multiplicity, J coupling constant in
Hz, and the integral for 1H resonances are given after the chemical shift value. The rotamer pairs are integrated together for 1 and separately for 4. c
denotes interchangeable. The fragment from C20 to C27 is not common to both molecules and therefore, data for that fragment of 4 are described
separately. C20 to C29 fragment of 1 corresponds to C28 to C37 fragment of 4 and data for those fragments are aligned in the same rows.
The numbering shown earlier is for convenience only and may not be consistent with IUPAC nomenclature.
for C25H30N3O4 – 468.2135]. Its structure determined by NMR indi-
cated that it arose from the Schiff base resulting from the reaction
between 2 and glycine methyl ester, further reacting with one mole-
cule of glycine methyl ester. The two new methines in 9 appeared as
a broad singlet atδ 4.92. At 60 °C, this resonance began to separate into
two resonances. The structural fragment common to both 2 and 9 was
easily identified from 1H NMR data (Table 2). The CH2 group derived
from glycine appeared as two AB doublets. One of them appeared
at δ 3.52 as a doublet with a 16 Hz coupling. And the other doublet
was masked by the broad water signal at δ 3.74. The two OMe groups
appeared as two broad singlets at δ 3.63 and 3.75. The side chain of 9
was established using the following correlations. The OMe at δ 3.75
and the resonance at δ 4.92 showed correlations to a carbonyl carbon
at δ 166.7. These are the correlations from H-24 and H-1 and/or H-22
to C-23. The protons on C-20 at δ 3.52 and 3.76 and H-25 at δ 3.63
showed correlations to C-21 at δ 168.3. Apart from having broad res-
onances, the 1H NMR spectrum of 9 did not indicate that it was a
mixture of diastereomers, but 13C NMR spectrum showed a set of
Figure 2. Liquid chromatography–tandem mass spectrometry
fragmentation of 4.
9, isolated as the 1.1 RRT impurity in 3, exhibited a molecular weight
of 468 Da by LC–MS. Its molecular formula was determined by
HRMS [+ESI HRMS (M+ H)+: observed – 468.2142 Da, calculated
Magn. Reson. Chem. (2016)
Copyright © 2016 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

C. Pathirana et al.
Table 2. NMR assignments of 9 and 10
9
10 (major diastereomer)*
10 (minor diastereomer)*
Assignment
δ H
δ C
δ H
δ C
δ H
δ C
1
4.76, bs, 1H
60.2
121.4
130.8
115.5
159.6
66.6
4.32, bd, 4.4, 1H
—
56.7
130.6
127.7
113.9
157.5
66.2
3.92, d, 7.9 Hz 1H
58.2
131.2
127.1
114.3
157.8
66.2
2
—
—
3, 7
7.39, 8.2, 2H
7.20, d, 8.2, 2H
6.87, d, 8.2, 2H
—
7.25, d, 8.2, 2H
4, 6
7.00, d, 8.2, 2H
6.89, d, 8.2, 2H
5
—
—
8
4.22, t, 6.3, 2H
4.18, t, 6.7, 2H
2.90, t, 6.7, 2H
—
4.18, t, 6.7 Hz, 2H
9
2.93, t, 6.3, 2H
25.7
25.6
2.90, t, 6.7 Hz, 2H
25.6
10
—
132.8
145.4
10.1
132.7
145.1
9.9
—
132.7
145.1
9.9
11
—
—
—
12
2.36, s, 3H
2.33, s, 3H
—
2.33, s, 3H
13
—
158.6
127.3
125.6
129.3
130.4
46.1
158.4
128.4
125.5
129.1
130.1
49.0
—
158.4
128.4
125.6
129.1
130.1
48.5
14
—
7.89, d, 7.7, 2H
7.48, m, 2H
7.48, m, 2H
3.52, d, 16, 1H 3.76, m, 1H
—
—
—
15, 19
7.89 br d, 6.7, 2H
7.47 m, 2H
7.47 m, 1H
3.40, s, 2H
—
7.89, br d, 6.7 Hz, 2H
16, 18
7.47, m, 2H
17
7.47, m, 1H
20
3.20, d, 16.0, 1H 3.32, d, 16.0, 1H
21
168.3
54.7
169.0
59.1
—
4.14, m, 1H
—
168.4
60.8
22
4.76, bs, 1H
—
4.16, m, 1H
—
23
166.7
53.8
170.9
51.5
170.9
51.9
24
3.76, s, 3H
3.63, s, 3H
—
3.26, s, 3H
—
3.49, s, 3H
—
25
52.7
—
—
NH (amide)
—
8.06, bs, 1H
—
8.06, bs, 1H
—
* Major and minor diastereomers are present at a ratio of 7 : 4 in 10.
1
δ ppm, relative to DMSO-d5 at δ 2.49 for 1H and DMSO-d6 at δ 39.5 for 13C. The multiplicity, J coupling constant in Hz, and the integral for H
resonances are given after the chemical shift value.
The numbering shown earlier is for convenience only and may not be consistent with IUPAC nomenclature.
minor resonances. When left at 60 °C, 9 degraded to produce
glycine and 2 as indicated by the NMR data of the mixture.
The molecular weight of 10, 0.9 RRT impurity observed in 2, was
determined by LC–MS to be 435Da [(M + H)+ 436Da]. Identification
of 9 was helpful in the identification of 10 and the rationalization of
its formation as shown in Scheme 2. The aromatic region of the 1H
NMR spectra of both 9 and 10 was similar to that of 2. Both 9 and
groups (δ 3.75, s, 3H and 3.63, s, 3H), whereas 10 had only one
methoxy resonance (δ 3.26, s, 3H). This difference and the corre-
sponding difference of 32 Da in the molecular weights from LC–
MS suggested that the side chain of 9 has undergone cyclization
with elimination of methanol to produce 10. The NMR data of 10
clearly showed that it was a mixture of two diastereomers. The res-
onance assigned to the benzylic methine showed significantly dif-
ferent chemical shifts in the two diastereomers (δ 4.32 and 3.92).
The protons of the CH2 group appeared as two doublets at δ 3.20
1
10 had two methines and a CH2 assignable to the side chain. H
NMR spectrum of 9 had two resonances indicating two methoxy
Scheme 2. The origin of impurities 9 and 10.
wileyonlinelibrary.com/journal/mrc
Copyright © 2016 John Wiley & Sons, Ltd.
Magn. Reson. Chem. (2016)

Identification of by-products by NMR in process development
and 3.33 in the minor diastereomer, while these protons appeared
as a singlet at δ 3.40 in the major diastereomer. HMBC spectrum
showed correlations to C-21 from H-20 and to C-23 from H-1,
H-22, and H-24 in both isomers. These correlations established
the piperazine ring. The molecular formula of C24H25N3O5 for 10
was confirmed by HRMS [+ESI HRMS (M+ H)+: observed –
formation from one molecule of glycine reacting with two mole-
cules of 2. The presence of carbomethoxy group was confirmed
by the correlations from the methyl protons (δ 3.47) and the adja-
cent CH2 protons (δ 3.75) to the carbonyl carbon (δ 166.6) in
1H-13C HMBC. The arrangement of the three CH2 groups as assigned
was confirmed by the correlations of the protons and carbons from
three CH2 groups to each other in 1H-13C HMBC.
1
436.1851 Da, calculated – 436.1872]. As indicated by H NMR, the
two diastereomers in 10 were present at a ratio of 7 : 4.
Identification of by-product 11
A third impurity observed in 3 at 2.2 RRT appeared to arise from a
reaction of 3 with 2. The molecular weight of impurity 11 was
determined from LC–MS to be 671 Da [(M + H)+ 672 Da]. HRMS pro-
vided its molecular formula of C41H42N3O6 [+ESI HRMS (M+ H)+:
observed – 672.3079 Da, calculated – 672.3074]. The fragment from
C-1 to C-17 was identified by comparing NMR data with the NMR
data of 1 (Table 3). However, the integral of the singlet assigned
to the methyl group in this fragment compared with the three-
proton singlet of the carbomethoxy group indicated a ratio of 2 : 1.
This indicated a symmetry in 11 and was consistent with its
Conclusion
Several impurities were encountered during the process develop-
ment of muraglitazar (1). Identification of the structures of a few
of them is described herein. Based on the structures of the impuri-
ties, it was deduced that excess reagents and active functionalities
in the products are responsible for the generation of the impurities.
A higher quality API was produced after appropriate adjustments to
the process were made to control the production of impurities
based on knowledge gained by the characterization of the impuri-
ties. Thus, impurities 9 and 10 could be eliminated by carefully
controlling the amount of glycine methylester hydrochloride. Impu-
rities arising from contaminants in the reagents (reactants), for exam-
ple, 4, are avoided by stringent quality control of the reagents.
Table 3. NMR assignments of 11
Assignment*
δ H
δ C
1
4.23, s, 4H
—
57.8
121.6
133.2
114.9
159.5
66.5
2
3, 7
4, 6
5
7.36, d, 7.8, 4H
6.94, d, 7.8, 4H
—
Experimental
Chromatography
8
4.20, t, 6.2, 4H
2.88, t, 6.2, 4H
—
Analytical HPLC was performed using a reversed phase YMC Pro
C18 column (4.6 × 150 mm, 3 μm) and a gradient of H2O and MeCN
(both containing 0.05% TFA) at ambient temperature. The flow rate
was 1.0 ml/min. Peaks were detected using UV at 280 nm.
Preparative chromatography was performed on a Waters 500 in-
strument using a reversed phase YMC-Pack Pro C18 (20× 250 mm,
5 μm) column using a gradient of H2O and MeCN (both containing
0.1% TFA). Flow rate was 15 ml/min, and peak detection was per-
formed using 280 nm UV.
9
25.6
10
11
12
13
14
15, 19
16, 18
17
20
21
22
132.7
145.6
9.9
—
2.30, s, 6H
—
158.7
125.7
125.6
129.3
130.5
50.8
—
7.85, br d, 7.1, 4H
7.46, m, 4H
7.46, m, 2H
3.78, s, 2H
—
NMR
166.6
52.8
NMR spectra with 1H detection were acquired at 25 °C on a Bruker
DRX 400 NMR spectrometer using a 3-mm 1H/13C Nalorac inverse
probe with a z-gradient. 13C spectra were obtained at 25 °C on a
3.47, s, 3H
Bruker DRX 400 NMR spectrometer using a 3-mm H/13C Nalorac
1
direct detection probe. NMR data were acquired using samples in
3-mm tubes containing approximately 5–10 mg of each compound
dissolved in ~0.2 ml of DMSO-d6. The spectra are referenced to the
solvent resonances (DMSO-d5 at 2.49 ppm for 1H, DMSO-d6 at
39.5 ppm for 13C).
* Only the unprimed atom labels are included in the table, but atoms
labeled with primed and unprimed labels have the same chemical
shifts. The 1H NMR integrals indicate the overlapped resonances by
showing a value corresponding to both resonances.
1H and 13C spectra were acquired with a 30° pulse and spectral
widths of 12 or 14 and 250 ppm, respectively. For all the 1H
1
detected 2D spectra, H spectral width was set at 12 or 14 ppm.
δ ppm, relative to DMSO-d5 at δ 2.49 for 1H and DMSO-d6 at δ 39.5 for
13C. The multiplicity, J coupling constant in Hz, and the integral for
1H resonances are given after the chemical shift value.
COSY spectra were acquired with 128 F1 increments and eight
scans per increment. Sinebell weighting in each dimension and
linear prediction to 1 k points in F1 were applied to yield 1 k × 1 k
matrix before Fourier transformation. HMQC spectra were acquired
with 256 increments in F1 and 32 scans per increment and an
The numbering shown earlier is for convenience only and may not be
consistent with IUPAC nomenclature.
Magn. Reson. Chem. (2016)
Copyright © 2016 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/mrc

C. Pathirana et al.
average-single-bond coupling constant of 140 Hz. Processing pa-
rameters were the same as for COSY. HMBC spectra were acquired
with 256 increments in F1 and 32 scans per increment. The experi-
ment was optimized for 8 Hz long-range coupling. Squared sinebell
weighting in each dimension and linear prediction to 1 k points in
F1 were applied to yield a final 1 k ×1 k matrix prior to Fourier
transformation.
References
[1] D. Barlocco. Curr. Opin. Investig. Drugs 2005, 6(4), 427–34.
[2] B. R. Henke. J. Med. Chem. 2004, 47(17), 4118–27.
[3] J. Li, L. J. Kennedy, Y. Shi, S. Tao, X.-Y. Ye, S. Y. Chen, Y. Wang,
A. S. Hernández, W. Wang, P. V. Devasthale, S. Chen, Z. Lai, H. Zhang,
S. Wu, R. A. Smirk, S. A. Bolton, D. E. Ryono, H. Zhang, N.-K. Lim,
B.-C. Chen, K. T. Locke, K. M. O’Malley, L. Zhang, R. A. Srivastava,
B. Miao, D. S. Meyers, H. Monshizadegan, D. Search, D. Grimm,
R. Zhang, T. Harrity, L. K. Kunselman, M. Cap, P. Kadiyala,
V. Hosagrahara, L. Zhang, C. Xu, Y.-X. Li, J. K. Muckelbauer, C. Chang,
Y. An, S. R. Krystek, M. A. Blanar, R. Zahler, R. Mukherjee,
P. T. W. Cheng, J. A. Tino, J. Med. Chem. 2010, 53(7), 2854–64.
[4] V. P. Hosagrahara, G. Chandrasena, S. Y. Chang, B. Koplowitz,
N. Hariharan, P. T. W. Cheng, W. G. Humphreys, Xenobiotica 2006,
36(12), 1227–38.
[5] A. Rusowicz, G. C. Lane, M. Saindane, H-J. Chung, M. F. Malley, PCT Int.
Appl. WO 2005113521.
[6] P. T. W. Cheng, P. Devasthale, Y. T. Jeon, S. Chen, H. Zhang, PCT Int. Appl.
WO 2001021602.
[7] G. E. Martin, in Analysis of Drug Impurities (Eds: R. J. Smith, M. L. Webb),
Wiley, Oxford, UK, 2007.
Liquid chromatography–mass spectrometry
Liquid chromatography–mass spectrometry spectra were acquired
using a Thermo TSQ-700 instrument. Samples were introduced
using analytical LC conditions described earlier. Mass spectra were
acquired with + ESI (ionization voltage 4000 V) and a mass range of
120 to 1200 Da. LC–MS/MS spectra were acquired with 29 or 40 eV
collision energy.
Accurate mass measurements were obtained on a Waters QTOF-2
instrument using leucine enkephalin at 556.2771 as a lock mass
(internal standard).
Acknowledgements
Supporting Information
Authors wish to thank Dr. Scott A. Miller for accurate mass spec-
trometry support and reviewing the manuscript and Drs. Thomas
V. Raglione and Dave Lloyd for reviewing the manuscript.
Additional supporting information may be found in the online ver-
sion of this article at the publisher’s web site.
wileyonlinelibrary.com/journal/mrc
Copyright © 2016 John Wiley & Sons, Ltd.
Magn. Reson. Chem. (2016)