Acetone/Isopropanol Photoinitiating System Enables Tunable Disulfide Reduction and Disulfide Mapping via Tandem Mass Spectrometry

Article abstract of DOI:10.1021/acs.analchem.8b04019

Herein, we report the development of a new photochemical system which enables rapid and tunable disulfide bond reduction and its application in disulfide mapping via online coupling with mass spectrometry (MS). Acetone, a clean and electrospray ionization (ESI) compatible solvent, is used as the photoinitiator (1% volume) in the solvent system consisting of 1:1 alkyl alcohol and water. Under ultraviolet (UV) irradiation (～254 nm), the acetone/alcohol system produces hydroxyalkyl radicals, which are responsible for disulfide bond cleavage in peptides. Acetone/isopropanol is most suitable for optimizing the disulfide reduction products, leading to almost complete conversion in less than 5 s when the reaction is conducted in a flow microreactor. The flow microreactor device not only facilitates direct coupling with ESI-MS but also allows fine-tuning of the extent of disulfide reduction by varying the UV exposure time. Near full sequence coverage for peptides consisting of intra- or interchain disulfide bonds has been achieved from complete disulfide reduction and online tandem mass spectrometry (MS/MS) via low energy collision-induced dissociation. Coupling different degrees of partial disulfide reduction with ESI-MS/MS allows disulfide mapping as demonstrated for characterizing the three disulfide bonds in insulin.

Full text of DOI:10.1021/acs.analchem.8b04019

Subscriber access provided by UNIV OF LOUISIANA
Article
Acetone/Isopropanol Photoinitiating System Enables Tunable Disulfide
Reduction and Disulfide Mapping via Tandem Mass Spectrometry
Sarju Adhikari, Xiaoyue Yang, and Yu Xia
Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b04019 • Publication Date (Web): 15 Oct 2018
Downloaded from http://pubs.acs.org on October 20, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 9
Analytical Chemistry
1
2
3
4
5
6
7
8
9
Acetone/Isopropanol Photoinitiating System Enables Tunable Disul-
fide Reduction and Disulfide Mapping via Tandem Mass Spectrome-
try
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
, ‡
†
*, †, ‡
Sarju Adhikari, Xiaoyue Yang, and Yu Xia
†Department of Chemistry, MOE Key Laboratory of Bioorganic Phosphorus Chemistry and Chemical Biology, Tsinghua
University, Beijing 100084, China
‡Department of Chemistry, Purdue University, West Lafayette, Indiana 47906, United States
ABSTRACT: Herein, we report the development of a new photochemical system which enables rapid and tunable disulfide
bond reduction and its application in disulfide mapping via online coupling with mass spectrometry (MS). Acetone, a clean
and electrospray ionization (ESI) compatible solvent is used as the photoinitiator (1% volume) in the solvent system consist-
ing of 1:1 alkyl alcohol and water. Under ultra-violet (UV) irradiation (~254 nm), the acetone/alcohol system produces hy-
droxyalkyl radicals, which are responsible for disulfide bond cleavage in peptides. Acetone/isopropanol is most suitable for
optimizing the disulfide reduction products, leading to almost complete conversion in less than 5 seconds when the reaction
is conducted in a flow microreactor. The flow microreactor device not only facilitates direct coupling with ESI-MS but also
allows fine-tuning of the extent of disulfide reduction by varying the UV exposure time. Near full sequence coverage for pep-
tides consisting of intra- or interchain disulfide bonds has been achieved from complete disulfide reduction and online tan-
dem mass spectrometry (MS/MS) via low energy collision-induced dissociation. Coupling different degrees of partial disulfide
reduction with ESI-MS/MS allows disulfide mapping as demonstrated for characterizing the three disulfide bonds in insulin.
As mass spectrometry (MS) becomes the tool of
choice for characterizing structural modifications in pro-
teins, disulfide mapping still presents a challenge for cur-
Alternative ion activation methods have been ex-
plored aiming to induce disulfide cleavages in the gas phase
so that intact disulfide bonds can be preserved before MS
analysis. Disulfide bond can be preferentially cleaved by
1
rent proteomics analysis workflows. The issue is directly
1
3,14
linked to the nature of gas-phase fragmentation chemistry
of protonated peptide ions under low energy collision-in-
duced dissociation (CID), the most available tandem mass
spectrometry (MS/MS) technique equipped on commercial
mass spectrometers. As depicted by the mobile proton
electrons (in electron capture dissociation (ECD)
and
1
5
electron transfer dissociation (ETD) ), photons (in 157
1
6
17,18
nm, 266 nm,
and 193 nm ultraviolet photodissociation
1
9
(UVPD) ), radicals (radical initiated peptide sequencing
(FRIPS) and TEMPO conjugated peptide ions), and oxi-
2
0
2
21
model, disulfide bond does not dissociate readily as com-
dants in ion/ion reactions . These methods can each be
pared to the amide bonds under low energy CID, often lead-
ing to no detectable sequence fragment ions for the back-
bone region inside a disulfide loop. This aspect not only
used as a stand-alone MS/MS technique or to be combined
with CID to improve sequence coverage.
3
Disulfide mapping strategies rely heavily on multi-
limits protein sequencing but also causes difficulty in disul-
fide mapping. To tackle this problem, disulfide bond reduc-
tion followed by alkylation has been employed as a routine
enzymatic digestion and partial disulfide reduction, with a
goal to produce peptide digests only consisting of inter-
chain disulfides. This type of structures is preferred since
sequence fragments around cysteine amino acid residues
associated with a disulfide bond are more likely to be gen-
erated from MS/MS, which is necessary for confident disul-
fide assignment. Such disulfide mapping approach has been
recently demonstrated for human serum albumin (17 disul-
4
sample preparation procedure in bottom-up proteomics
5
workflows. Along this line, fast disulfide cleavage via chem-
ical or electrochemical reduction has been developed which
could be coupled with online liquid chromatography-mass
6
,7
spectrometry (LC-MS). Besides, disulfide cleavage inside
the ionization source region has been demonstrated using
8
19
fide bonds) and serotransferrin (19 disulfide bonds).
Finding the right conditions for partial reduction and prote-
8
,9,10
electrospray ionization (ESI),
matrix-assisted laser de-
1
1
sorption ionization (MALDI), and reactive electrospray-
olytic cleavages while avoiding disulfide scrambling is criti-
1
2
16,22
assisted laser desorption/ionization (ELDI) . All above ap-
proaches when coupled with subsequent MS/MS (typically
low energy CID) deliver rich sequence information due to
opening the disulfide bridges; however, disulfide linkage in-
formation is lost if complete disulfide reduction is adopted.
cal for successful disulfide mapping.
The optimization
process is typically achieved by trial-and-error and the con-
ditions cannot be generalized for different protein systems.
Obviously, new methods that could either improve diges-
tion or partial disulfide reduction will greatly advance the
overall process for disulfide mapping.
ACS Paragon Plus Environment

Analytical Chemistry
Disulfides are known to be highly reactive towards
Page 2 of 9
without further purification. Deionized water was obtained
from a water purification system at 0.03 µS∙cm (Thermo Sci-
entific; San Jose, CA, USA).
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
radicals and the dominant reaction channel is disulfide
bond cleavage. Earlier efforts from our group have shown
2
3
that radicals produced from gas discharge or ultra-violet
9
,10
(UV) irradiation
can cleave disulfide bonds within
Peptide/Protein digestion and liquid chromatography
Trypsin digestion was conducted by mixing 10 µL
peptides near or in an ESI emitter. However, disulfide
cleavage efficiency is only moderate and often accompanied
by side reactions due to the lack of control of the reaction
kinetics. In this work, we have developed a new
photochemical reaction system for achieving efficient and
tunable disulfide bond cleavage in second-time scale. In this
system, acetone was used as a clean and MS compatible pho-
toinitiator upon 254 nm UV irradiation, while secondary hy-
droxyalkyl radicals resulting from alcohol co-solvent were
responsible for cleaving the disulfide bond (Scheme 1). This
photochemical reaction system was hyphenated with infu-
sion ESI via a flow microreactor. Structural analysis capa-
bility of this photochemical system was demonstrated by
coupling complete or partial disulfide cleavage with online
ESI-MS/MS via CID for peptides containing one or multiple
disulfide bonds.
of protein/peptide solution (10 µg/mL) with 15 µL of 20
mM ammonium acetate buffer, 5 µL acetonitrile, and 2 µL of
o
trypsin (1 µg/µL), followed by incubation at 37 C for 3
hours. The trypsin digest was separated on a reversed
phase LC (Shimadzu Corporation, Columbia, MD, USA). Col-
umn dimension was 4.6 mm x 150 mm with 5 µm C-18 pack-
ing material. Separation conditions for the digest are 0.1%
formic acid in water (solvent A) and 0.1% formic acid in
ACN (solvent B) with a linear gradient starting from 0% to
40% solvent B for 30 minutes at 0.5 mL/min flow rate.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Photochemical reactions in nanoESI emitter and in flow
microreactor
A low-pressure mercury (LP-Hg) lamp with an
emission band around 254 nm (BHK, Inc., Ontario, CA) was
utilized to initiate the photochemical reactions. For the
nanoESI setup, the lamp was placed orthogonally to the tip
and off-axis by approximately 2-4 cm of the borosilicate
Scheme 1. Proposed reaction pathways for disulfide bond
cleavage from using acetone/alkyl alcohol photoinitiating
system.
9
glass emitter, as reported previously. The flow microreac-
tor was made from UV-transparently coated fused silica ca-
a) Radical Initiation
pillary (100 μm i.d., 375 μm o.d.; Polymicro Technologies;
24,25
Phoenix, AZ).
The lamp was placed in parallel to the ca-
hν
pillary (~10 cm exposure length) at 0.5 cm distance. The
peptide solution was pumped through the microreactor (1-
10 µL/min), the end of which was connected to the inlet of
an infusion ESI source. All photochemical reaction setups
were enclosed in a cardboard box to prevent direct human
exposure to UV light.
+ •CH3
(
254 nm)
b) Formation of Hydroxyalkyl Radical
Mass Spectrometry
•
CH3
MeOH: R = R = H EtOH: R = H, R = CH
3
QTRAP 4000 and 4500 hybrid triple quadru-
pole/linear ion trap MS (SCIEX, Concord, ON, Canada) were
used to perform preliminary studies with nanoESI emitter
reaction setup. For the online coupling of the flow microre-
actor with infusion ESI-MS/MS, X500R TOF (SCIEX) was
used. TIMS TOF (Bruker, Germany) was used to study par-
tially disulfide reduced insulin products.
1
2
1
2
IPA: R = R = CH
3
1
2
c) Cleavage of S-S
Pep-S-S-Pep
Pep-SH + Pep-S-CR R OH
1
2
RESULTS AND DISCUSSION
EXPERIMENTAL SECTION
Acetone/alkyl alcohol photoinitiating system for disul-
fide cleavage
Materials
Previously, our lab reported the use of hydroxyl
radicals (•OH) for initiating hydroxyalkyl radical formation
inside a nanoESI emitter, where alcohol was used as a co-
solvent for nanoESI of disulfide peptides.9 Disulfide
cleavage was the main reaction channel, however, the
reaction yield was limited due to relatively low
Oxidized glutathione, reduced glutathione, metha-
nol (MeOH), ethanol (EtOH), isopropyl alcohol (IPA), ace-
tone, acetonitrile (ACN), TPCK treated trypsin, oxytocin, in-
sulin from the porcine pancreas, and ribonuclease B were
purchased from Sigma-Aldrich (St. Louis, MO, USA). Soma-
tostatin-14 and selectin binding peptide were purchased
®
from Aladdin (Shanghai, China) and GL Biochem Ltd.
(
Shanghai, China), respectively. The single letter sequence
and disulfide bond connecting pattern for each peptide are
listed in Supporting Information, Table S-1. d -Methanol
CD OH, 99.5 atom % D) and d -acetone (CD COCD , 99.9
3
(
3
6
3
3
atom % D) were purchased from J&K Scientific Ltd. (Beijing,
China). All commercially purchased chemicals were used
ACS Paragon Plus Environment

Page 3 of 9
Analytical Chemistry
a) 1% Acetone, UV OFF
c) 1% Acetone, 50% MeOH, UV ON
e) 1% Acetone, 50%EtOH, UV ON
1
2
3
4
5
6
7
8
9
2
+
[M+2H]
07.2
ɣECG
3
338.1
308.1
1
00
50
0
ɣECG
ɣECG
100
50
0
100
50
0
2
SCH OH
ɣECG
ɣECG
SH
308.2
S-CH(CH )OH
3
3
52.1
2
80
340
400
280
340
m/z
400
280
340
m/z
400
m/z
b) 1% Acetone, UV ON
d) 1% Acetone, 50% MeOH, UV ON
f) 1% Acetone, 50% IPA, UV ON
ɣECG
m/z 338.1
m/z 308.2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
50.1
100
100
50
0
308.1
1
00
SC(O)CH3
ɣECG
m/z 307.2
ɣECG
SCH
3
ɣECG
SCH
ɣECG
5
0
50
0
SH
3 2
S-C(CH ) OH
2
C(O)CH
3
322.1
364.1
3
08.0
366.2
0
2
80
340
400
2
80
340
m/z
400
0
1
2
3
4
m/z
UV irradiance (min)
Figure 1. Positive ion mode MS spectra of oxidized glutathione (10 µM) prepared in solutions containing 1% acetone as a
photoinitiator. a) Before and b) after UV exposure of the aqueous solution in nanoESI emitter. c) UV irradiation of the peptide
in H
2
O:MeOH ((v/v) =1:1) and d) extracted ion chromatogram (XIC) for the reaction shown in c). UV irradiation of the peptide
O:EtOH ((v/v) =1:1) and f) H O:IPA ((v/v) =1:1).
in e) H
2
2
concentrations of primary radicals (•OH) formed from 185
nm UV photolysis of ambient air.
improve the reaction yield is to use more efficient radical
initiators, such as 2,2-dimethoxy-2-phenylacetophenone
DMPA) and benzophenone, which are commonly used in
organic synthesis and radical polymerization. These
compounds go through Norrish type I cleavage at the α-car-
bon of the carbonyl group and form primary radicals for
UV irradiation;18,32 the absence of the disulfide cleavage
without acetone addition in our study could be due to low
photon flux of the UV lamp. The above set of experiments
clearly suggests that Norrish Type I cleavage of acetone in-
deed happens for 1% of acetone upon 254 nm UV irradia-
tion. Similar reaction phenomenon was observed with as
low as 0.1% (v%) of acetone addition, albeit with longer UV
irradiation time. Although disulfide bond cleavage is 100%
by reacting with radicals derived from photolysis of acetone,
the presence of multiple reaction channels points out the
need to convert the first-generation radicals into less reac-
tive secondary radical species. In this study, we focused on
forming hydroxyalkyl radicals via using acetone as a
photoinitiator.
9
,26
An obvious way to
(
27
2
8,29
radical initiation upon UV irradiation.
They, however,
are not directly compatible with MS analysis due to non-vol-
atile nature and relatively high concentrations (in mM) typ-
ically employed in the reaction system, which would cause
severe interference to MS analysis. Acetone is the simplest
ketone and it breaks down to acetyl radical and methyl rad-
3
0,31
ical upon UV irradiation (Scheme 1).
Considering that
Formation of hydroxyalkyl radicals was explored
acetone has relatively low boiling point, low proton affinity,
and it is miscible with most ESI solvent systems for peptide
analysis, acetone should be well suited to be used as a
photoinitiator.
To test the idea, 1% acetone (volume%) was added
to an aqueous solution of oxidized glutathione (10 µM) and
loaded onto the nanoESI reaction setup (Figure S1, Support-
ing Information). Before UV irradiation, only doubly proto-
by adding alkyl alcohol precursor as co-solvent in the reac-
tion system. Figure 1c shows the nanoESI spectrum of oxi-
dized glutathione in MeOH/H O ((v/v) = 1:1, 1% acetone)
solvent system upon UV irradiation. As a big contrast to Fig-
ure 1b where no methanol was added, only two distinct re-
action products were observed, corresponding to the re-
duced (m/z 308.2) and S-hydroxymethyl modified glutathi-
one ions (m/z 338.2). S-hydroxymethyl product was veri-
fied by detecting signature formaldehyde loss (30 Da) from
2
2
+
nated peptide ions ([M+2H] ) at m/z 307.2 was observed
(Figure 1a). After the lamp was turned on for about 2
minutes, intact peptide ions completely disappeared and
new peaks resulting from disulfide bond cleavage domi-
nated (Figure 1b). The major products are reduced glutathi-
9
low energy CID (Figure S2, Supporting Information). The
extracted ion chromatogram (XIC) showed that the intact
glutathione had a sharp drop from 100% to 0% within 30
seconds (UV irradiation 1.1-1.5 min), accompanied by steep
increase of the reduced peptide (m/z 308), the ion signal of
which maximized at 1.2 min UV irradiation and then
dropped to reach a steady state within 20 s. During this
time, the ion signal of the hydroxymethyl substitution prod-
uct (m/z 338) increased rapidly and stayed constant
afterwards. These kinetic data suggest that reduced thiol
might be further converted to S-hydroxymethyl product af-
ter its initial formation. Indeed, almost identical spectrum
to Figure 1c was obtained when the oxidized glutathione
was replaced by reduced glutathione for the reaction
+
one ([γECG+H] , m/z 308.0) and radical substitution prod-
ucts at cysteinyl sulfur, including methyl radical (•CH
22.1), acetyl radical (•C(O)CH , m/z 350.1), and acetonyl
radical (•CH C(O)CH , m/z 361.1), respectively. The identity
3
, m/z
3
3
2
3
of these products were verified by accurate mass measure-
ments and MS/MS experiments. The same reaction was re-
peated without addition of acetone. No reaction products
associated with disulfide bond cleavage were observed,
even with prolonged UV exposure (~5 min). Homolytic
cleavage of disulfide bond can be induced by 193-266 nm
ACS Paragon Plus Environment

Analytical Chemistry
Figure S3, Supporting Information). No definitive evidence
Page 4 of 9
(
demonstrated that photochemical reactions can be greatly
accelerated relative to bulk reactions when conducted in a
flow microreactor due to significantly improved photon ef-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
could be obtained for tracing the singly protonated thiyl
radical ions since they overlap with the doubly protonated
glutathione ions at m/z 307.2.
2
4,25
ficiency using flow path in µm-dimension.
Flow micro-
The reaction phenomena for solvent systems in-
volving larger alkyl alcohols, viz. EtOH and IPA, were similar
to that observed in the MeOH solvent system. That is, 100%
disulfide cleavage was achieved and only two reaction prod-
ucts were detected (Figure 1e and 1f). MS CID of the sub-
stitution products, m/z 352 (Figure 1e) and m/z 366 (Figure
reactor is also advantageous regarding precise control of re-
action time so as to reduce detrimental side-reactions from
prolonged UV exposure as found by using a static nanoESI
setup. UV-transparently coated fused silica capillary was
employed to construct the flow microreactor, which was
connected directly to an infusion ESI source for online MS
experiments as shown in Figure 2a.
2
1
f) showed abundant loss of 44 Da and 58 Da, confirming
the formation of the respective S-hydroxyalkyl product
Figure S2, Supporting Information). A distinct difference
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
a)
from the Acetone/MeOH system is that the yield of reduced
thiol was increased significantly from 32% (MeOH) to 80%
(EtOH) and 93% (IPA).
Flow
Reaction
Solution
microreactor
The possible reaction pathways accounting for di-
sulfide cleavage using acetone/alcohol photoinitiating sys-
tem are proposed in Scheme 1. Direct disulfide/photon in-
teractions are considered negligible due to low photon flux
and orders of magnitude lower concentration of the peptide
relative to acetone. The reaction pathway is likely to be ini-
tiated by photochemical decomposition of acetone into me-
ESI
MS
UV
inlet
2
+
b) No UV
[M+2H]
4
67.2
1
00
AGCK
A: 376.2
3
0
thyl/acetyl radicals. In the presence of a large excess of al-
cohol, acetone should have little chance to interact with
peptide directly, while reaction with alcohol solvent is the
main reaction channel. In the case of methanol, the
5
0
TFTSC
B: 556.2
formation of hydroxymethyl radical (•CH
gen abstraction from the C-H bond of methanol is supported
by the D COH experiments (Figure S4, Supporting Infor-
2
OH) due to hydro-
0
3
70
430
490
m/z
550
600
3
c) 3 s UV
mation). For larger alcohols, hydrogen abstraction is gener-
+
[B+H]
+
[A+H]
ally preferred from the α-position due to its lower C-H bond
558.2
1
00
50
0
3
78.2
3
3
dissociation energy. For example, the BDE of C
CH C-HOH is 380.7 kJ/mol, while that of the C -H is 394.6
kJ/mol.
α
-H in
(
3
)
2
β
3
4
Mechanistic studies involving reactions of
3
3
hydroxyalkyl radicals with disulfides have been scarce.
3
70
430
490
550
600
Data generated from gas-phase reactions and theoretical
calculations suggest that carbon-centered radicals cleave
m/z
d)1
00
3
5,36
disulfide bond via a substitution mechanism.
Based on
Intact (S-S)
the observed products, we hypothesize that hydroxyalkyl
radical attacks the disulfide bond, forming thiyl radical and
S-hydroxyalkyl at the cleavage site. The thiyl radical can fur-
ther abstract a hydrogen atom from hydroxyalkyl radical
Reduced Products
50
0
3
7
(BDE, ~150 kJ/mol for O-H in •C(CH
3
) OH), forming re-
2
duced thiol and aldehyde/ketone as stable products. Hy-
droxyalkyl radicals can also react with a reduced thiol to
form S-hydroxyalkyl product and the reactivity follows the
0
1
2
3
4
5
UV Irradiation (sec)
order of •CH
2
OH > •CH(CH
3
)OH > •C(CH
3
2
) OH, as supported
Figure 2. a) A flow microreactor setup for coupling ace-
tone/IPA imitated disulfide bond reduction with ESI-MS.
ESI-MS spectra of 10 µM of trypsin-digested somatostatin-
from reaction data of reduced glutathione. It is worth not-
ing that the pathways described above are intended to de-
scribe the overall reaction phenomenon. Detailed mecha-
nistic investigations are beyond the scope of the current re-
port.
1
4 in H O:IPA (1:1) with 1% acetone b) before UV and c)
2
after 3 s UV irradiation in positive ion mode. d) Plot of %con-
version of the intact peptide to disulfide reduced species as
a function of UV irradiation time.
Disulfide reduction in a flow microreactor
Comparing the data for disulfide cleavage in Figure
1 using different alkyl alcohol as co-solvent, IPA solvent sys-
tem stood out for producing reduced disulfide product with
almost 100% conversion. Therefore, 1% acetone in 1:1
Such a reaction system was tested with a peptide
containing an interchain disulfide bond (sequence shown
the inset of Figure 2b). Figure 2b and 2c compare the ESI-
MS spectra before and after UV irradiation in positive ion
mode. Clearly, after 3s UV irradiation, the intact peptide
IPA/H
2
O was chosen as the optimal solvent system for pep-
2
+
tide and protein analysis. Previously, our group has
ions ([M+2H] , m/z 467.2) were 100% consumed, while
ACS Paragon Plus Environment

Page 5 of 9
Analytical Chemistry
ions corresponding to the two reduced peptide chains, ions (+15, m/z 955.1) produced rich backbone fragmenta-
1
2
3
4
5
6
7
8
9
[
A+H] (m/z 378.2) and [B+H] (m/z 558.2), dominated in
+
+
tion for sequence identification (Figure S7, Supporting In-
formation,).
the post-reaction spectrum. The progress of reduction was
monitored as a function of the UV irradiation time. As
shown in Figure 2d, the yield of disulfide reduction in-
creases monotonically as UV irradiation time increases and
it reaches 100% after 4 s. Above kinetic data demonstrate
that the degree of disulfide bond reduction could be tuned
by adjusting the UV exposure time of the peptide in the flow
microreactor, which is particularly useful for achieving par-
tial disulfide reduction for peptides containing multiple di-
sulfide bonds.
+5
085.9
100
1
UV OFF
a)
+4
1357.1
50
+6
905.1
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
800
900
1000 1100 1200 1300 1400 1500
A+3H]3+
[
8
02.1
100
UV ON
Increasing sequence coverage for disulfide peptides
and proteins from complete reduction
b)
[C+2H]
2+
Selectin binding peptide is a naturally existing cir-
cular peptide due to an intrachain disulfide bond between
C1 and C9 residues (sequence shown in the inset of Figure
1084.5
5
0
B+H]+
[
858.4
892.4
1
202.5
2
3). As expected, no sequence ions were obtained from MS
0
8
00
900
1000
m/z
1100 1200 1300 1400 1500
CID of doubly protonated intact peptide ions (Figure S5,
Supporting Information). After 5 s UV irradiation, 100% di-
sulfide reduction was achieved for this peptide, as demon-
strated by the complete disappearance of intact peptide
c)
Intact Peptide:
Fragment with the other chain attached
A
C40 K P V N T F V H E S L A D V Q A V C58 S Q K
B
Y P N C95 A Y K
Chain A:
C
H I I V A C110 E G N P Y V P V H F D A S V
2
+
ions ([M+2H] , m/z 523.8, Figure 3a) and appearance of
C40 K P V N T F V H E S L A D V Q A V C58 S Q K
Y P N C95 A Y K
2
+
2
ions at m/z 524.7 ([red-M+2H] ) ions (Figure 3b). MS CID
of the reduced product at m/z 524.7 gave almost full series
of b- and y-type fragment ions, allowing sequence identifi-
cation of the peptide (Figure 3c). Similar reaction
phenomenon was observed with another peptide
containing intrachain disulfide bond (oxytocin, Figure S6,
Supporting Information).
Chain B:
Chain C:
H I I V A C110 E G N P Y V P V H F D A S V
Figure 4. MS spectra of trypsin-digested ribonuclease-B
subjected to the radical reaction: (a) before UV and (b) after
s UV irradiation in flow microreactor. (c) Sequence and
5
fragmentation maps of intact peptide and reduced A, B, and
C chains.
[M+2H]2+
23.8
UV OFF
[redM+2H]2+
524.7
b)
a)
5
We further applied complete disulfide reduction
for the analysis of protein digests, mimicking the bottom-up
proteomic approach. Ribonuclease-B, a small protein con-
taining four disulfide bonds (sequence shown in Figure S8,
Supporting Information) was subjected to trypsin digestion
and LC separation before disulfide reduction and ESI-
MS/MS analysis. One disulfide peptide was detected after
digestion, which contained three peptide chains connected
by two interchain disulfide bonds: C40-C95 and C58-C110 (se-
quence shown in Figure 4). The other predicted peptide
containing C26-C84 and C65-C72 disulfide bonds wasn’t
detected, possibly due to its resistance to proteolytic diges-
UV ON
520
530
520
530
m/z
m/z
[redM+2H]2+
c)
524.7
100
C I E L L Q A R C
SH
SH
b82+
y4
50
y5
b2
b3
y3
b4
y7
b5
y6
tion. Figure 4a and 4b compare the ESI-MS spectra of the
38
y2
b6
peptide before and after UV irradiation (5 s). Clearly, the
b7
6
+
5+
3+
intact peptide ions ([M+6H] , [M+5H] , [M+4H] ) com-
pletely disappeared after UV irradiation. The presence of
0
200
300
400
500
600
m/z
700
800
900 1000
3+
three separated peptide chains, viz. [A+3H] (m/z 802.1),
[B+H] (m/z 858.4), and [C+2H] (m/z 1084.5), confirmed
Figure 3. MS spectra of 10 µM of selectin binding peptide
+
2+
dissolved in H
2
O:IPA (1:1) with 1% acetone: a) before UV,
the presence of two interchain disulfide bonds in this pep-
2
b) after 5 s UV irradiation in flow microreactor, and c) MS
CID of reduced peptide ions at m/z 524.7 ([ M+2H] ). The
fragmentation map of the peptide is indicated in the inset.
6+
tide. The intact peptide ions ([M+6H] ) and the disulfide re-
red
2+
3+
+
2+
duced A, B, and C chains ([A+3H] , [B+H] , [C+2H] ) were
2
each subjected to MS CID for structural analysis (spectra
shown in Figure S9). The fragmentation map of the intact
and reduced peptide ions is summarized in Figure 4c. MS
Lysozyme (four disulfide bonds, 129 amino acid
residues) was tested as a model for larger proteins. Com-
plete disulfide reduction of all four disulfide bonds were
achieved within 5 s UV irradiation as compared to three-
2
2
hour reduction using DTT. MS CID of the reduced lysozyme
ACS Paragon Plus Environment

Analytical Chemistry
Page 6 of 9
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
+
4
+
1156.3
S-S
%
100%
0%
1
156.3
1145.3
a)
b)
c)
UV OFF
d)
1
00
Intact (•)
red-one (•)
red-two (•)
r^2
5
0
0%
6
+
0.9999
963.8
0
800
1000
1200
1400
11 15 65 0. 50
111 85 06 0. 5
1157.5
1158.5
4
+
UV ON (1.8 s)
3+
1156.5
S-S
%
53%
31%
16%
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
B
1145.9 e)
1
00
4+
1133.9
Intact (•)
red-one (•)
red-two (•)
r^2
B
5
+
8
50.9
1
156.5
6
+
5
0
1192.0 A2+
964.0
A3+
0.9999
0
800
1000
1200
1400
1600
155.5
1800
1
1156.5
1157.5
1158.5
UV ON (3.1 s)
S-S
%
12%
28%
60%
8
50.9
B
4+
f)
1157.1
Intact (•)
red-one (•)
red-two (•)
r^2
1
00
3+
B
1
133.9
5
+
3+
1157.1
1192.0
A
7
5
0
2
+
95.3
A
4+
146.2
0.9984
1
0
8
00
1000
1200
1400
11 165 05 0. 5
18 10 1056.5
1157.5
1158.5
m/z
m/z
Figure 5. Positive ion mode ESI MS spectra of porcine insulin in 1% acetone, IPA: H O ((v/v) = 1:1). a) Before UV, b) after
2
1.8 s, and c) 3.1 s UV irradiation in flow microreactor. d)-f) The zoomed-in region of 5+ insulin ions from a)-c). Decon-
voluted isotopic distributions of the intact (black dots), one disulfide reduced (green dots), and two disulfide reduced
(red dots) species are overlaid with the experimental data and their relative contributions are listed in the insets.
CID of the disulfide reduced chains led to almost 100% se-
quencing of the peptide, allowing determination of cysteine
residues at C40 and C58 in A chain, C95 in B chain, and C110 in
C chain. MS CID of the intact peptide produced a series of
sequence ions in A chain with intact B chain attached, i.e.
53% intact, 31% and 16% one disulfide and two disulfide
reduced species, respectively (Figure 5e, deconvolution
procedure in Supporting Information). When the UV irradi-
ation time was extended to 3.1 s, the remaining 5+ ion pop-
ulation only contained 12% intact insulin ions, while the
one and two disulfides reduced species was 28% and 60%,
respectively (Figure 5f). It is worth noting that partial disul-
fide reduction typically produced a mixture of different de-
grees of disulfide reduction; however, due to the capability
of online reaction monitoring the optimized condition for a
certain reduced population could be achieved rapidly. The
intact 5+ insulin ions and those resulting from different de-
grees of reduction (1.8 s and 3.1 s UV irradiation) were each
2
BAb
2
, BAb , BAb10, BAb11, BAb14, BAb15, BAb16, and BAb17,
9
suggesting a disulfide linkage at C40-C95. Several fragment
ions specific to the C chain were obtained, confirming its
presence in the intact peptide, which could only be linked to
A Chain through a disulfide linkage at C58-C110
.
Mapping disulfide bonds in insulin from tunable partial
reduction
2
A distinct advantage of the photochemical reaction
system is the capability of tuning the degree of disulfide re-
duction simply by adjusting the UV irradiation time (in sec-
onds). This tunability is critical in analyzing disulfide con-
necting patterns for intact proteins or peptides which con-
tain multiple disulfide bonds. Such aspect is demonstrated
by disulfide mapping of porcine insulin, which consists of
two interchain disulfide linkages between chain A and chain
subjected to MS CID for disulfide mapping. The lower m/z
region (500-900) of the CID data are summarized in Figure
6a-c and the corresponding fragmentation maps derived
from detected b- and y-type ions are summarized in Figure
6d-f. CID data in higher m/z region are supplied in the
2
Figure S10, Supporting Information. As expected, MS CID of
the intact insulin ions (5+, m/z 1156.1, Figure 6a) only pro-
duced sequence fragment ions in the exocyclic backbone re-
gions and no specific information regarding disulfide link-
B (AC
7
-BC
7
and AC20-BC19) and one intrachain linkage within
-AC11). Data in Figure 5 compare the ESI-MS
2
chain A (AC
6
age can be obtained. However, MS CID of the fully reduced
2+
5+
spectra of insulin before UV and after different periods of
UV irradiation. Before UV irradiation, the protonated insu-
lin ions, 5+ and 4+ (m/z 1156.3 and 1145.3), dominated in
the spectrum (Figure 5a). After 1.8 s UV irradiation, ions of
A chain (3+ at m/z 795.3 and 2+ at m/z 1192.0) and B chain
(4+ at m/z 850.9 and 3+ at m/z 1133.9) appeared as abun-
dant species (Figure 5b). Deconvolution of the isotope clus-
ter of the remaining insulin 5+ ions showed that it contained
A
and B ions produced full sequence coverage of each
chain, allowing pinpointing the positions of the six cysteine
residues (Figure S11, Supporting Information).
2
Figure 6b shows MS CID of the 5+ ions (m/z
1156.5) containing 31% one disulfide reduced and 16% two
disulfide reduced species. The presence of Bb7-18 ions sug-
gests an inter-chain disulfide had been cleaved before CID
and this bond should involve C residue in chain B. On the
7
ACS Paragon Plus Environment

Page 7 of 9
Analytical Chemistry
By112+
Fragment with the other chain attached
Fragment without the other chain
1
2
3
4
5
6
7
8
9
1
156.3
a)
d)
b ion
y ion
S
S
Bb5+
Ab5+
By9
By6+
2+By5+
Bb6+
G I V E Q C6 C7 T S I C11 S L Y Q L E N Y C20
N
By8+
A chain
S
S
S
S
B chain
5
00
600
700
800
900
732.5
733.0
733.5
F V N Q H L C7 G S H L V E A L Y L V C19 G E R G F F Y T P K A
Ab5+
1156.5
b)
Bb132+
732.9
e)
2+
Bb13
S
S
Bb112+By112+
Bb122+
By102+Ab6+
+
Ab7
733.4
Bb102+
By92+By5+
Bb142+
Bb6+
G I V E Q C6 C7 T S I C11 S L Y Q L E N Y C20 N
+
Bb162+
By6
Bb152+
S
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
By8+
Ab7
33.3
7
S
F V N Q H L C7 G S H L V E A L Y L V C19 G E R G F F Y T P K A
5
00
600
700
800
900
732.5 733.0
733.5
Ab5+
1157.1
c)
Bb132+
f)
732.9
Bb112+
+
Bb122+
Bb13
Bb152+
Bb142+
733.4
Bb102+
2+
+
G I V E Q C6 C7 T S I C11 S L Y Q L E N Y C20
N
Ab
6
B
5+
Ab
7
Ab7
By6+
S
Bb5+
By112+
Bb162+Ab9+
733.3
By92+
By5+
Bb6+
Ab8+
By8+
S
F V N Q H L C7 G S H L V E A L Y L V C19 G E R G F F Y T P K A
5
00
600
700
800
900
732.5 733.0
733.5
m/z
m/z
2
Figure 6. MS CID of 5+ insulin ion populations derived from a) intact, b) 31% one disulfide reduced and 16% two disul-
fide reduced species, and c) 28% one disulfide reduced and 60% two disulfide reduced species. The corresponding frag-
mentation maps are shown in d)-f).
other hand, the detection of BAy4, 6-8 ions allow for locating
an interchain disulfide bond between AC20 and BC19. Two se-
the partially reduced A chain ions even from using the nar-
rowest isolation window available on the instrument. Nev-
ertheless, comparisons of the CID data between the partially
reduced and the fully reduced A chain ions showed that the
quence ions were detected between AC
6
and AC11, viz. Ab
6
and Ab , albeit at relatively low intensities (inset of Figure
6b). This phenomenon is suggestive for an intrachain disul-
fide in this sequence region. MS CID of the 5+ insulin ions
containing 60% two disulfides reduced species is shown in
Figure 6c with unique ions highlighted in red. The presence
of abundant Ab6-9 and BAy11-15 ions clearly suggests that this
ion population contains structures with the intrachain di-
sulfide bond in chain A and an interchain disulfide bond
7
relative ion intensities of b -10 were significantly reduced
6
2
from the partially reduced A chain ion population while
those of exocyclic sequence ions were quite comparable, i.e.
4,5 and b11-17 (data shown in Figure S12, Supporting Infor-
mation). Consequently, the disulfide linkage between C
and C10 in chain A was determined.
b
6
(AC
7
-BC
7
) had been reduced. It is worth mentioning that
and Ab are significantly
the relative ion intensities of Ab
6
7
CONCLUSIONS
increased as the two disulfide reduced species increased
from 16% to 60% (insets of Figure 6b and 6c), supporting
that they are likely derived from the two-disulfide bond re-
duced population. Interestingly, no fragment ion relating to
structures containing reduced AC20-BC19 disulfide was ob-
served. If this disulfide bond had been cleaved, sequence
ions such as By12-23, ABb7-17, Ay1-10, or BAb11-20, should be pre-
sent. The above data clearly demonstrate that disulfide re-
A photochemical reaction system that allows for
rapid and tunable disulfide reduction has been developed
for characterizing peptides containing one or multiple di-
sulfide bonds by ESI-MS/MS. Acetone/IPA is determined as
the best photoinitiating system for disulfide bond reduction
with almost 100% yield achieved within less than 5 s UV ir-
radiation. The use of flow microreactor enables precise
control of the degree of disulfide reduction and online cou-
pling with infusion ESI-MS/MS. By pairing complete disul-
fide reduction with subsequent low-energy CID, near com-
plete sequence coverage is achieved for various peptides
containing one or multiple disulfide bonds. Furthermore,
different degrees of partial reduction can be achieved
simply by adjusting the UV irradiation time, greatly facilitat-
ing the disulfide linkage assignment, as demonstrated by
mapping the three disulfide bonds in insulin. For the pep-
tide systems studied herein, no evidence for disulfide bond
scrambling has been obtained, likely benefited from the fast
reaction and online detection system. Compared to conven-
tional disulfide mapping approach, the online photochemi-
cal system enjoys several distinct advantages, including full
compatibility with ESI-MS without introducing any chemi-
cal matrix, significantly shortened reaction time (from
hours to seconds), and flexibility in tuning the extent of
duction was slowest for the disulfide bond between AC20
-
BC19. The difference in the disulfide reduction kinetics could
be attributed to the different solvent accessibility of the
three disulfide bonds, in which the AC20-BC19 has been re-
ported to be least accessible in insulin under electrochemi-
3
9
cal reduction.
Although the two interchain disulfide bonds have
been pinpointed from MS CID of the partially reduced insu-
2
lin ions, the intrachain disulfide bond in chain A still needs
to be determined. To acquire such information, ions of par-
tially reduced chain A (2+, m/z 1191.0) which consisted of
an intact intrachain disulfide bond but with two other inter-
2
chain disulfide bonds being cleaved, were subjected to MS
CID. It is worth pointing out that a small portion of the fully
reduced A chain ions was inevitably mixed in the precursor
ion population and underwent beam-type CID together with
ACS Paragon Plus Environment

Analytical Chemistry
Page 8 of 9
disulfide reduction. These aspects are highly desirable for
(7) Mysling, S.; Salbo, R.; Ploug, M.; Jørgensen, T. J. D. Anal.
Chem. 2014, 86, 340-345.
(8) Cramer, C. N.; Kelstrup, C. D.; Olsen, J. V.; Haselmann, K. F.;
Nielsen, P. K. Anal. Chem. 2017, 89, 5949-5957.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
developing high throughput structural analysis capability
for proteins consisting of multiple disulfide bonds. As a pre-
liminary test, disulfide reduction in the microflow reactor
was online-coupled with reversed phase LC-MS. Disulfide
linkage analysis was successful for peptide digest of a model
protein. Detailed characterization of this system for bottom-
up protein analysis is currently undergoing and will be re-
ported separately.
(
1
(
9) Stinson, C. A.; Xia, Y. J. Am. Soc. Mass Spectrom. 2014, 25,
192-1201.
10) Durand, K. L.; Tan, L.; Stinson, C. A.; Love-Nkansah, C. B.;
Ma, X.; Xia, Y. J. Am. Soc. Mass. Spectrom. 2017, 28, 1099-1108.
11) Fukuyama, Y.; Iwamoto, S.; Tanaka, K. J. Mass Spectrom.
(
2006, 41, 191-201.
(12) Peng, I. X.; Ogorzalek Loo, R. R.; Shiea, J.; Loo, J. A. Anal.
Chem. 2008, 80, 6995-7003.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
13) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am.
Chem. Soc. 1998, 120, 3265-3266.
14) Oh, H.; Breuker, K.; Sze, S. K.; Ge, Y.; Carpenter, B. K.;
McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 15863-
5868.
ASSOCIATED CONTENT
(
Supporting Information
The Supporting Information is available free of charge on
the ACS Publications website.
1
(15) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.;
Hunt, D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528-9533.
(16) Foley, S. F.; Sun, Y.; Zheng, T. S.; Wen, D. Anal. Biochem.
2008, 377, 95-104.
(17) Fung, Y. M. E.; Kjeldsen, F.; Silivra, O. A.; Chan, T. W. D.;
Zubarev, R. A. Angew. Chem. 2005, 117, 6557-6561.
Procedure for isotopic deconvolution, mass spectra data on
disulfide
bond
cleavage
using
acetone/alcohol
photoinitiating system (PDF).
AUTHOR INFORMATION
Corresponding Author
(
8
(
18) Agarwal, A.; Diedrich, J. K.; Julian, R. R. Anal. Chem. 2011,
3, 6455-6458.
19) Quick, M. M.; Crittenden, C. M.; Rosenberg, J. A.; Brodbelt,
J. S. Anal. Chem. 2018, 90, 8523-8530.
(20) Lee, M.; Lee, Y.; Kang, M.; Park, H.; Seong, Y.; Sung, B. J.;
Moon, B.; Oh, H. B. J. Mass Spectrom. 2011, 46, 830-839.
*
E-mail: xiayu@mail.tsinghua.edu.cn
(
8
21) Pilo, A. L.; McLuckey, S. A. Anal. Chem. 2016, 88, 8972-
979.
ORCID
(
(
(
22) Gray, W. R. Protein Sci. 1993, 2, 1732-1748.
23) Xia, Y.; Cooks, R. G. Anal. Chem. 2010, 82, 2856-2864.
24) Adhikari, S.; Xia, Y. Anal. Chem. 2017, 89, 12631-12635.
Yu Xia: 0000-0001-8694-9900
Author Contributions
(25) Adhikari, S.; Zhang, W.; Xie, X.; Chen, Q.; Xia, Y. Anal.
Chem. 2018, 90, 5239-5246.
All authors have given approval to the final version of the
manuscript
(
2
(
4
(
1
26) Creasey, D. J.; Heard, D. E.; Lee, J. D. Geophys. Res. Lett.,
000, pp 1651-1654.
27) Hoyle, C. E.; Bowman, C. N. Angew. Chem. Int. Ed. 2010,
9, 1540-1573.
28) Norrish, R. G. W.; Kirkbride, F. W. J. Chem. Soc. 1932,
518-1530.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
(29) Norrish, R. G. W. Trans. Faraday Soc. 1934, 30, 103-113.
(30) Norrish, R. G. W.; Crone, H. G.; Saltmarsh, O. D. J. Chem.
Soc. 1934, 1456-1464.
Financial support from National Natural Science Founda-
tion of China (No. 21722506, No. 21621003) and NIH
R01GM118484 is greatly appreciated. We acknowledge as-
sistance from D. Zhang for isotope deconvolution.
(
(
31) Haas, Y. Photochem. Photobiol. Sci. 2004, 3, 6-16.
32) Bookwalter, C. W.; Zoller, D. L.; Ross, P. L.; Johnston, M. V.
J. Am. Soc. Mass. Spectrom. 1995, 6, 872-876.
(
33) Akhlaq, M. S.; Murthy, C. P.; Steenken, S.; Von Sonntag, C. J.
REFERENCES
Phys. Chem. 1989, 93, 4331-4334.
(34) Luo, Y.-R. Handbook of bond dissociation energies in
organic compounds; CRC Press: Boca Raton, FL, 2003.
(
2
(
1) Lakbub, J. C.; Shipman, J. T.; Desaire, H. Anal. Bioanal. Chem.
018, 410, 2467-2484.
2) Wysocki, V. H.; Tsaprailis, G.; Smith, L. L.; Breci, L. A. J. Mass
Spectrom. 2000, 35, 1399-1406.
3) Lioe, H.; O'Hair, R. A. J. Am. Soc. Mass. Spectrom. 2007, 18,
(
2
(
35) Krenske, E. H.; Pryor, W. A.; Houk, K. N. J. Org. Chem.
009, 74, 5356-5360.
36) Anglada, J. M.; Crehuet, R.; Adhikari, S.; Francisco, J. S.;
Xia, Y. PCCP 2018, 20, 4793-4804.
37) Internet Bond-energy Databank (pKa and BDE)-iBonD:
http://ibond.chem.tsinghua.edu.cn, accessed July 2018.
38) Bernard, B. A.; Newton, S. A.; Olden, K. J. Biol. Chem. 1983,
(
1109-1123.
(
(4) Wall, J. S. J. Agric. Food. Chem. 1971, 19, 619-625.
(
5) Zhang, Y.; Fonslow, B. R.; Shan, B.; Baek, M.-C.; Yates, J. R.
Chem. Rev. 2013, 113, 2343-2394.
6) Zhang, Y.; Cui, W.; Zhang, H.; Dewald, H. D.; Chen, H. Anal.
Chem. 2012, 84, 3838-3842.
(
258, 12198-12202.
(39) Cramer, C. N.; Haselmann, K. F.; Olsen, J. V.; Nielsen, P. K.
Anal. Chem. 2016, 88, 1585-1592.
(
ACS Paragon Plus Environment

Page 9 of 9
Analytical Chemistry
1
2
3
4
5
6
7
8
9
Insert Table of Contents artwork here
ESI
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Acetone/IPA
tR<5 sec
9
ACS Paragon Plus Environment