Full Papers
the sample was, if necessary, brought up to the initial volume
(70 mL) with water. Each assay was repeated three times to pro-
duce three independent replicates.
cubated with CAR, HY, and PM at a 1:1 ratio, a 1:5 ratio with HY,
and 10:1 and 5:1 ratios with AG. Control samples contained the
single RCS or quencher at 800 mm concentration in 1 mm phos-
phate buffer, pH 7.4.
Intact protein analysis by microflow automated loop injection ESI-
MS: Intact ubiquitin (either in its unmodified or RCS-modified
forms) was analyzed by HRMS as already described.[17] Briefly, ubiq-
uitin solution (40 mL) recovered from filter units was mixed with
denaturing solution (40 mL) (H2O/CH3CN/HCOOH; 40/60/0.2 v/v/v).
Aliquots of the samples (5 mL) were injected into an LTQ-Orbitrap
XL mass spectrometer by using an ESI source (Thermo Scientific,
Milan, Italy). Samples were automatically injected by an Ulti-
mate 3000 RSLCnano system at a constant flow rate of 10 mLminꢁ1
of H2O/CH3CN/HCOOH; 70/30/0.1 v/v/v. Sample injection and spec-
tra acquisition were fully automated and were controlled by Xcali-
bur software (version 2.0.7, Thermo Scientific) and Chromeleon
Xpress software (Dionex, version 6.80). Source parameters: spray
voltage 1.8 kV, capillary temperature 2208C, capillary voltage 35 V,
tube lens offset 120 V. A list of 20 background ions was used as
lock mass values for real-time mass calibration.[35] Mass spectra
were acquired by the Orbitrap analyzer in the positive-ion mode
by using the profile mode, scan range m/z=110–2000, AGC target
5ꢁ105, maximum inject time 500 ms, resolving power 100000 (full
width at half maximum (FWHM) at m/z=400). Each sample was ac-
quired three times to obtain three technical replicates.
MS-based analysis of RCS/quencher reaction products: After
24 h incubation, all samples were diluted approximately 10-fold in
water and were directly infused into the LTQ-Orbitrap XL by using
a microliter syringe (Hamilton, Bonaduz, Switzerland) at a flow rate
of 10 mLminꢁ1. The ESI source (Thermo Scientific, Milan, Italy) was
set as follows: spray voltage 3.5 kV, sheath gas 5, capillary tempera-
ture 2758C, capillary voltage 35 V, tube lens offset 120 V. The lock
mass option was enabled. Spectra were acquired by using the
Tune Plus software (version 2.4 SP1, Thermo). Full mass spectra
were acquired by the Orbitrap (FT) analyzer in the positive-ion
mode by using the following settings: profile mode, scan range m/
z=60–600, AGC target 5ꢁ105, maximum inject time 500 ms, re-
solving power 100000 (FWHM at m/z=400). Full mass spectra
were acquired for 30 s to average signals. If the Orbitrap analyzer
did not detect the signal, full mass spectra were acquired by using
the LTQ-XL analyzer in the range m/z=50–600, with AGC target
5ꢁ104, maximum inject time 100 ms. Most abundant peaks were
fragmented to obtain MS/MS spectra by using a selection window
of 2.5 m/z. We used both CID and HCD fragmentation (with AGC
target=1ꢁ104 or 2ꢁ105, respectively); the collision energy was ex-
perimentally adjusted for each precursor to increase the signal of
the fragment ions (generally CE=35–55 V). Most fragmentation
products were acquired by using the Orbitrap analyzer at a resolv-
ing power 100000 (FWHM at m/z=400). In a few cases, if the Orbi-
trap analyzer did not detect the signal, we acquired MS/MS spectra
with the LTQ-XL analyzer. Full mass spectra were acquired for 30 s
to average signals.
Quantification of the extent of carbonylation: A dedicated Xcalibur
processing method was set to quantify the area under the +11
multicharged peaks,[17] localized in the ranges m/z=779.00–783.50
for unmodified ubiquitin, m/z=793.00–797.50 for HNE-modified
ubiquitin,
m/z=(784.25–785.75)+(789.50–791.00)+(794.75–
796.25) for GO-modified ubiquitin, m/z=784.00–787.00 for MGO-
modified ubiquitin, and m/z=(784.00–785.50)+(788.10–789.60) for
MDA-modified ubiquitin. Peak areas were automatically detected
and quantified postacquisition by using an Xcalibur Quan Browser
(Thermo Scientific). The percentage of modified ubiquitin con-
tained in each sample was computed according to Equation (1):
In silico generation of reaction product fragments: We assigned
a chemical structure to each product obtained from the RCS/
quencher reaction on the basis of the data obtained from the liter-
ature and on the highly accurate mass values observed by HRMS.
Elemental composition was obtained by
AUC-modified ubiquitin
% Modified ubiquitin ¼ ½
the Xcalibur/Qual Browser software by
using a 8 ppm tolerance. The structure
was drawn by using ChemSketch soft-
ꢂ ꢃ 100
ð1Þ
ðAUC-unmodified ubiquitin þ AUC-modified ubiquitinÞ
The percentage of modified ubiquitin obtained upon co-incubation
with RCS and carbonyl quenchers was normalized against the per-
centage of modified ubiquitin obtained in the control samples
(ubiquitin incubated with RCS without any quencher). The noise
present in the blank sample (ubiquitin incubated without RCS) was
subtracted.
ware (version 12.01, Advanced Chemistry Development, Canada)
and imported into the Highchem—MassFrontier software (ver-
sion 5.1, Thermo Scientific) that generated theoretical fragments
according to both general chemistry rules and HighChem fragmen-
tation library. We used the default setting, with the following ex-
ceptions: radical species were excluded from the analyses, maxi-
mum four fragmentation steps were allowed, and the minimum
mass value was 50 Da. The mass values of the theoretical frag-
ments were compared with the experimental values, with a toler-
ance equal to 8 ppm for spectra acquired by the Orbitrap analyzer
and 0.3 Da for spectra acquired by the LTQ-XL analyzer.
The percentages of modified ubiquitin were plotted against the
log concentration of quencher (expressed as quencher/RCS molar
ratio). The concentration of carbonyl quenchers able to inhibit the
formation of adducts on ubiquitin by 50% was obtained by inter-
polating the concentration–response curve (GraphPad Software
Inc., version 6.04): such concentration was termed UC50 and is
expressed as the quencher/RCS molar ratio.
Molecular modeling of the investigated RCS: Owing to their
markedly different flexibility, the stereoelectronic properties of the
carbonyl quenchers were investigated by exploiting different com-
putational strategies. Indeed, the more rigid sequestering agents
AG, HY, and PYR were analyzed by considering only the lowest-
energy conformation as derived by quenched MonteCarlo simula-
tions (generating 1000 minimized geometries by randomly rotating
the rotors) followed by PM7 semiempirical optimization, whereas
for the more-flexible CAR, 15 nonredundant favored conformations
were generated by the MonteCarlo search and optimized by semi-
empirical calculations. All so-generated conformations were finally
minimized at their ground state and in the gas phase by density
HRMS and computational analyses for elucidation of reac-
tion products
Sample preparation: Each RCS was incubated at 378C with each
carbonyl quencher at different molar ratios, which were previously
optimized to enhance the formation of detectable amounts of
products. HNE was incubated with all quenchers in a 1:1 ratio.
MGO and GO were incubated at a 1:1 ratio with AG, HY, and PM,
a 1:5 ratio with HY, and 10:1 and 5:1 ratios with CAR. MDA was in-
ChemMedChem 2016, 11, 1 – 13
11
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ