Thermally enhanced enzymatic proteolysis for rapid 18O labeling in proteomics

Article abstract of DOI:10.1016/j.ijms.2011.07.005

To streamline protocols for protein identification and better understand the contributions of purely thermal versus non-thermal effects, we utilize a microwave oven and a PCR thermocycler for rapid heating to accelerate enzymatic digestion of proteins. When performed in H₂¹⁸O, rapid heating results in efficient C-terminal ¹⁸O atom labeling of the proteolytic peptides. The approach is illustrated on the example of several pure proteins using trypsin and other proteases for rapid digestion. MALDI TOF/TOF MS provides unambiguous identification of the individual tryptic peptides. We have performed a time-course study on the degree of ¹⁸O incorporation by varying the irradiation/heating times for each method. In order to gain insights into the mechanism of thermally enhanced trypsin digestion and ¹⁸O labeling we carry out experiments in which the two events - lysis and labeling - are decoupled. We also study the rates of ¹⁸O incorporation as a function of tryptic peptide C-terminal amino acid type and peptide length. Both heating methods are very rapid - in most cases digestion and incorporation of two ¹⁸O atoms into R-terminated tryptic peptides is completed in less than 5 min, thus considerably reducing the time for bottom-up proteomics including quantitation by ¹⁸O labeling.

Full text of DOI:10.1016/j.ijms.2011.07.005

International Journal of Mass Spectrometry 312 (2012) 24–29
Contents lists available at ScienceDirect
International Journal of Mass Spectrometry
journal homepage: www.elsevier.com/locate/ijms
Thermally enhanced enzymatic proteolysis for rapid 18O labeling in proteomics
Miquel D. Antoine∗, Nathan A. Hagan, Plamen A. Demirev
Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
To streamline protocols for protein identification and better understand the contributions of purely ther-
mal versus non-thermal effects, we utilize a microwave oven and a PCR thermocycler for rapid heating to
accelerate enzymatic digestion of proteins. When performed in H218O, rapid heating results in efficient
C-terminal 18O atom labeling of the proteolytic peptides. The approach is illustrated on the example of
several pure proteins using trypsin and other proteases for rapid digestion. MALDI TOF/TOF MS provides
unambiguous identification of the individual tryptic peptides. We have performed a time-course study
on the degree of 18O incorporation by varying the irradiation/heating times for each method. In order
to gain insights into the mechanism of thermally enhanced trypsin digestion and 18O labeling we carry
out experiments in which the two events – lysis and labeling – are decoupled. We also study the rates
of 18O incorporation as a function of tryptic peptide C-terminal amino acid type and peptide length.
Both heating methods are very rapid – in most cases digestion and incorporation of two 18O atoms into
R-terminated tryptic peptides is completed in less than 5 min, thus considerably reducing the time for
bottom-up proteomics including quantitation by 18O labeling.
Received 26 April 2011
Received in revised form 6 July 2011
Accepted 7 July 2011
Available online 3 August 2011
This paper is dedicated to Dr. Catherine
Fenselau, in recognition of her pioneering
role and seminal contributions in the
development of biological mass
spectrometry and its applications.
Keywords:
Proteomics
MALDI TOF
Isotopic labeling
Enzymatic digestions
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
step, e.g., different enzymes, pH, solvents, temperatures, times, etc.
[7,8,19,20]. For quantification, the relative ion abundances of the
The increased risk of biological terrorism in recent years has
highlighted the necessity for more efficient, rapid and robust meth-
ods to detect and identify bio-threat agents. Mass spectrometry,
including bottom-up and top-down proteomics, is becoming a
method of choice to successfully achieve this goal [1]. Currently,
most strategies for bottom-up proteomics rely on chemical or
enzymatic digestion, while quantitative proteomics encompasses
metabolic, enzymatic or chemical incorporation of stable-isotope
labels resulting in discernible mass shifts between corresponding
peptides and proteins in the two sample pools [2–8]. Isotope label-
ing with 18O atoms during proteolysis is one such qualitative and
quantitative bottom-up proteomics strategy [9–14]. It is based on
the incorporation of one or two 18O atoms during N- or C-terminal
specific endoprotease digestion of a protein in H218O [15–18]. Since
incorporation of two 18O atoms results in a more easily measured
4 Da shift between corresponding peptide pairs from the two sam-
ples (control versus labeled), C-terminal specific proteases (e.g.,
trypsin) are most often used. The proteolysis and 18O labeling can
be successfully decoupled to achieve optimal conditions for each
control versus isotopically shifted peptides are measured and cor-
related to the respective precursor proteins. The relative merits
and draw-backs of the 18O labeling technique have been discussed
in a number of reviews [4–7,9,20]. Proteolytic 18O incorporation
has been demonstrated to be compatible with shotgun proteomics
[21] and this labeling technique has been developed into a use-
ful tool in the relative quantification of complex protein mixtures,
determination of post-translational modifications, intra-molecular
protein cross-linking, and protein–protein interactions and com-
plexes [8,12,19]. Specific software algorithms for estimates of 18
O
incorporation efficiencies and peptide pair ratios [22,23] have been
combined with traditional proteomics tools to fully exploit the
advantages of this technique.
Typically, in a laboratory setting, when trypsin is used for pro-
teolysis and combined with 18O labeling, the reaction duration is
between 12 and 24 h in order for complete digestion and incor-
poration of two 18O atoms to occur. In order to devise a rapid
18O labeling method for quantitative applications in proteomics
for on-site screening, trypsin proteolysis [24,25] of model proteins
in either 16O or 18O water is performed using a microwave oven
or a PCR thermocycler [26]. For proteins investigated, ranging from
ubiquitin to bovine serum albumin (BSA), we observe very efficient
proteolysis and 18O incorporation on a time scale of less than 5 min
in either heating device. MALDI TOF/TOF MS is employed to identify
the individual tryptic peptides. We evaluate the 18O incorporation
∗
Corresponding author at: Johns Hopkins University Applied Physics Laboratory,
MS 21-N454, 11100 Johns Hopkins Rd., Laurel, MD 20723, USA.
Tel.: +1 443 778 5718.
E-mail address: miquel.antoine@jhuapl.edu (M.D. Antoine).
1387-3806/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijms.2011.07.005

M.D. Antoine et al. / International Journal of Mass Spectrometry 312 (2012) 24–29
25
rates as a function of tryptic peptide C-terminal amino acid type,
2.3. Mass spectrometry
peptide length, and microwave irradiation time. In order to gain
insights into the mechanism of thermally enhanced trypsin diges-
tion and 18O labeling experiments are carried out in which the two
events – lysis and labeling – are decoupled. Rates of 18O incorpo-
ration as a function of tryptic peptide C-terminal amino acid type
and peptide length are also investigated. Both heating methods are
very rapid – in most cases significant digestion and incorporation
of two 18O atoms into R-terminated tryptic peptides is achieved in
less than 5 min, thus considerably reducing the time for bottom-up
proteomics including quantitation by 18O labeling.
Positive ion MALDI mass spectra were acquired on an Autoflex
TOF/TOF instrument (Bruker Daltonics, Billerica, MA). Standard
accelerating voltage (∼20 kV) and delayed extraction (150 ns)
conditions were used in linear or reflectron TOF mode. For tan-
dem TOF/TOF MS, the LIFT method was employed, as previously
described [27]. Briefly, ions were desorbed by a N2 laser and delay-
extracted at 6 kV in the first leg of the instrument (where the
ions undergo unimolecular dissociation). The respective precur-
sor ion and the fragment ions were then selected (isolated) in an
ion gate, before re-acceleration in the second (reflectron) leg to
23 keV total energy for a singly charged precursor ion. Typically,
several thousand single laser shot traces were summed for each
MS/MS spectrum by manually scanning the laser beam across an
individual sample spot. The MS and MS/MS peptide spectra were
initially calibrated by using peptide and protein standards (average
mass error in reflectron mode <0.5 Da across the entire mass/charge
range below 3500). The linear and reflectron mode spectra were
subsequently recalibrated with several tryptic peptide fragments
used as internal mass standards (average mass error in reflec-
tron mode <0.02 Da). Standard bioinformatics software (Bruker
Biotools) was used to identify initially the proteolytic fragments
observed. In order to assign sequences to unidentified fragments
presumably resulting from disulfide bridges kept intact, the Web-
accessible Links software (a part of the collaboration MS3D Portal
– http://ms3d.org) was used [28].
2. Experimental
2.1. Materials
Bovine ubiquitin, equine cytochrome C, carbonic anhydrase,
bovine serum albumin (BSA) and trifluoroacetic acid (TFA) were
obtained from Sigma Chemical Co. (St. Louis, MO). MALDI
matrices – ␣-cyano-4-hydroxycinnamic acid (CHCA) and 3,5-
dimethoxycinnamic acid (SA) – were purchased from Laser BioLabs
(Sophia-Antipolis Cedex, France) and were used without addi-
tional purification. Trypsin, TPCK treated (proteomics grade), was
obtained from Worthington Biochemical (Lakewood, NJ) and endo-
proteinase Glu-C was from New England BioLabs (Ipswich, MA).
Both proteases were reconstituted in either deionized water or 18
O
water. 18O water (97%) was obtained from Cambridge Isotope Labo-
ratories (Andover, MA). Acetonitrile and water, both in HPLC grade
(Burdick and Jackson), were purchased from VWR (West Chester,
PA).
Caution: Care should be exercised during microwave-assisted
heating, e.g., the sample vials should be uncapped in order to avoid
explosion if a regular microwave oven is used.
3. Results and discussion
2.2. Thermally enhanced proteolysis
Several physical methods for accelerating enzymatic or chem-
ical digestion in proteomics have been discussed in recent years
[29,30]. Ultra-fast (<1 min) trypsin digestion of proteins by high
intensity focused ultrasound has been demonstrated [31]. Sim-
ilarly, high pressure cycling in the range of 5–35 kpsi has been
introduced for in-solution digestions of single proteins and com-
plex protein mixtures in about 1 min [32]. A rapid (<10 min)
sample processing involves high intensity focused ultrasound to
rapidly reduce and alkylate cysteines, digest proteins and then label
peptides with 18O [32]. Among the physical methods, microwave-
assisted proteolysis – enzymatic [24,25] or chemical [33,34] – has
become a very popular approach for reducing the overall sample
preparation times in proteomics. Using water and/or other sol-
vents, it has been shown that microwave-assisted digestion times
for complex proteins and mixtures are markedly reduced from sev-
eral hours to several minutes without reducing the tryptic enzyme
efficiency or specificity as demonstrated with BSA (Fig. 1). By apply-
ing microwave irradiation in H218O, both rapid trypsinization and
efficient 18O incorporation for proteins on a time scale of 2 min
or less is achieved. We detect and identify by database searching
more than 40 different tryptic peptides in the mass/charge range
from 500 to 3000 (Fig. S1, Supplementary material). These all cor-
respond to BSA-derived peptides with typically zero or one missed
trypsin-specific cleavage at a K- or R-amino acid carboxyl terminal,
thus providing more than 40% unique sequence coverage for BSA
For all proteolytic digestion reactions, sample vials were filled
with 150 ␮l of protein solution (30 ␮M in water, either H216O con-
trol or H218O) followed by an aliquot of trypsin at an enzyme
to substrate ratio of 1:100 (w/w) or for endoproteinase Glu-C
1:50 (w/w). For microwave-assisted proteolysis, glass sample vials
containing the reactants were partially covered, and placed in
a commercial microwave oven (Panasonic (900W), VWR) along
with a 100 ml glass beaker partially filled with water. Microwave
irradiation at the maximum oven power setting was done in inter-
vals of 30 s to prevent solvent evaporation. In between irradiation
times, sample vials were taken out of the microwave oven, capped
and vortexed, then uncapped and returned to the oven for addi-
tional irradiation. Digestion was stopped after a pre-selected time,
typically between 2 and 5 min, by immersion of the sample vial
in an ice bath for rapid cooling. PCR thermocycler experiments
were conducted using a Bio-RAD MJ Mini Personal Thermocycler
(Hercules, CA). Reaction protocols were similar to those for the
microwave reactions except volumes were reduced to accommo-
date the smaller size of the thermocycler tubes. Various cycling
protocols were investigated and it was determined that a constant
temperature of 65 ◦C for 5 min provided optimum protein diges-
tion results. Conventional proteolytic digestions were conducted
by placing reaction vials containing protein and enzyme in a water
bath at 37 ◦C for 24 h. All reactions were stopped by placing sam-
ple vials in an ice bath after the designated reaction time for each
method and reaction products were analyzed directly afterwards.
For MALDI MS, 0.5 ␮l of the control and 18O water-containing
digests were each deposited in separate sample wells on the steel
sample slide (Bruker Daltonics, Billerica, MA). The CHCA matrix
solution (0.5 ␮l, 10 mg/ml in acetonitrile:water:TFA, 70:30:0.1) was
then added to the wells and the sample was dried in air.
(
16O control sample, Table 1). No trypsin autolysis products were
detected under our experimental conditions (MALDI spectra of the
trypsin control reaction displayed an intense peak for the intact
trypsin molecular ion, while no low mass ions were observed).
In order to minimize the overall experimental time, we have
deliberately excluded Cys reduction and alkylation steps for
disulfide-bond containing proteins in the current sample prepara-
tion protocols. Our premise is to develop protocols for rapid, on-site

26
M.D. Antoine et al. / International Journal of Mass Spectrometry 312 (2012) 24–29
100
analysis in the event of a bio-threat response. For the last several
(a)
years, our researchteam has investigatedmass spectrometry-based
methods for on-site field analysis of suspicious white powders,
whereby analysis time is critical [35,36]. Based on our data for
BSA, most of the 17 intra-molecular disulfide bridges are pre-
served intact during rapid thermally assisted proteolysis. Several
tryptic peptide ions with masses reflecting the presence of intact
disulfide bonds – either intra-fragment or disulfide-linked non-
adjacent tryptic peptides are observed. These assignments have
been confirmed also by 18O labeling (vide infra). Evidence for
cleaved disulfides (reduced Cys) was found in only two very low
intensity peaks.
0
100
(b)
Comparing the baseline isotope-resolved spectra of the control
and 18O-labeled BSA digests, we observe a reproducible shift from
2 to 8 Da between the respective tryptic peptide masses in the
two samples. The number of 18O atoms incorporated during this
short time proteolysis depends on the type of C-terminal amino
acid (Fig. 2). Almost all K-terminated peptides incorporate exclu-
sively one 18O atom (Fig. 2a). The incorporation of the first 18O atom
occurs during the initial reaction step – cleavage of the amide bond
(vide infra). After a reaction time of 90 s, only two out of fifteen
K-terminated peptides have exchanged more than one 18O atoms
(Table 1). In contrast, the R-terminated peptides incorporate almost
0
m/z
1000
1500
2000
2500
3000
Fig. 1. Comparison of MALDI-TOF mass spectra of peptides resulting from trypsin-
digested BSA: (a) Conventional overnight (24 h) reaction and (b) 5 min microwave-
assisted digestion.
Table 1
List of identified tryptic peptides from microwave-assisted trypsinization of BSA and degree of 18O incorporation.
Experimental
monoisotopic mass (Da)
Theoretical monoisotopic
mass (Da)
Peptide sequence
Residue numbera
Number of 18
O
incorporated after
90 s digestion
508.38
572.46
634.48
649.45
508.25
572.36
634.38
FGER
QRLR
229–232
219–222
20–24
205–209
2
2
2
2
GVFRR
649.33
IETMRb
649.33
CASIQK
223–228
689.44
712.43
689.37
712.37
AWSVAR
SEIAHR
236–241
29–34
2
2
733.49
733.42
VLTSSAR
212–218
2
847.50
906.57
918.55
847.50
906.47
918.52
LSQKFPK
IETMREK
LRCASIQK
242–248
205–211
221–228
2
(n/a)c
1
922.48
922.49
AEFVEVTK
249–256
1
927.50
927.49
YLYEIAR
161–167
2
974.43
974.46
DLGEEHFK
37–44
1
977.53
977.45
NECFLSHK
123–130
1
990.59
990.56
EKVLTSSAR
210–218
2
1001.59
1083.58
1163.60
1193.60
1249.60
1283.71
1305.65
1347.53
1439.79
1479.79
1511.85
1001.59
1083.60
1163.63
1193.60
1249.62
1283.71
1305.72
1347.55
1439.81
1479.80
1511.84
1511.95
1567.74
1639.94
1692.94
1750.97
1889.00
1888.93
1940.82
2045.03
2439.10
2492.16
2492.11
2587.15
3085.43
ALKAWSVAR
YLYEIARR
LVNELTEFAK
DTHKSEIAHR
233–241
161–168
66–75
25–34
35–44
361–371
402–412
76–88 + disulfide
360–371
421–433
438–451
545–557
347–359
437–451
249–263
242–256
89–105
169–183
264–280 + disulfide
168–183
264–285 + disulfide
66–88 + disulfide
221–228, 267–280 + disulfide
223–228, 264–280 + disulfide
223–228, 264–285 + disulfide
2
2
1, 2
2
1
2
0, 1
1
2
2
FKDLGEEHFK
HPEYAVSVLLR
HLVDEPQNLIK
TCVADESHAGCEK
RHPEYAVSVLLR
LGEYGFQNALIVR
VPQVSTPTLVEVSRb
QIKKQTALVELLK
DAFLGSFLYEYSR
KVPQVSTPTLVEVSR
AEFVEVTKLVTDLTK
LSQKFPKAEFVEVTK
SLHTLFGDELCKVASLRb
HPYFYAPELLYYANK
VHKECCHGDLLECADDR
RHPYFYAPELLYYANK
VHKECCHGDLLECADDRADLAK
LVNELTEFAKTCVADESHAGCEKd
LRCASIQK-ECCHG DLLECADDR
CASIQK-VHKECCHGDLLECADDR
CASIQK-VHKECCHGDLLECADDRADLAK
1, 2
1567.74
1639.94
1692.90
1750.94
1888.94
2
2
1
1
2
1940.75
2044.99
2438.98
2492.02
1
1
1
1
2587.06
3085.67
3
3
a
Residue numbering for P02726|ALBU BOVIN including propeptide sequence (a.a. 1–24).
b
c
Probable sequence assignment based on 18O incorporation (not confirmed by LIFT MS/MS due to weak precursor ion signal).
Signal too weak to accurately determine 18O incorporation.
d
Sequence assignment based on 18O incorporation and confirmation by LIFT MS/MS.

M.D. Antoine et al. / International Journal of Mass Spectrometry 312 (2012) 24–29
27
100
100
Control H216O, 90s
(a)
(a)
FKDLGEEHFK
0
100
H218O, 90s
H218O, 5min
1258
(b)
(c)
0
100
0
100
0
0
1245
1255
1265
m/z
1246
1250
1254
m/z
100
Fig. 3. Resolved molecular ion region spectra (MALDI reflectron-TOF) illustrating
digestion/labeling time-course for a selected Lys-terminated peptide; control diges-
tion for 90 s in H216O (top), 90 s digestion in H218O (middle), and 5 min digestion in
H218O (bottom).
16O
(b)
LGEYGFQNALIVR
tetrahedral enzyme–peptide substrate complex intermediate fol-
lowed by nucleophilic attack of H218O [19,20]. The complex
formation and ultimately the completion of the reaction depend
on a number of factors: the substrate–trypsin binding, enzyme
concentration, peptide solubility, etc. [19]. On the timescale of
this experiment for BSA (from 30 to 90 s for microwave diges-
tion), we observe efficient incorporation of two 18O atoms for
R-terminated peptides, while longer microwave irradiation times
(e.g., 5 min) are needed to more efficiently incorporate the second
18O atom in K-terminated peptides (Fig. 3). Microwave-assisted
proteolytic digestion reactions in H216O and H218O were also con-
ducted using endoproteinase Glu-C. Efficient fragmentation of BSA
is observed after an irradiation time of 5 min as well as incorpo-
ration of one and two 18O-atoms, depending on the C-terminal
amino acid (Figs. S2 and S3, Supplementary material). Similar to
the results using trypsin at this longer irradiation time, incorpora-
tion of two 18O-atoms for lysing terminating peptides is observed.
These results highlight the overall robustness of the rapid thermally
assisted digestion approach, including its applicability to different
proteases. When conducted in parallel, additional confirmation of
protein identification can thus be achieved without sacrificing the
rapid analysis time.
0
100
18O
0
1476
1480
1484
S
1488
m/z
100
(c)
16O
S
CASIQK
VHKECCHGDLLECADDR
0
100
18O
Additional experiments aimed at decoupling microwave-
assisted trypsinization from microwave-assisted 18O labeling were
performed in order to better understand the mechanisms of both
processes. Initially, microwave-assisted trypsin digestion of BSA
was carried out in 16O-water for 90 s. The sample was then dried,
re-suspended in 18O-water and subsequently irradiated again for
additional 90 s. In this set of experiments, efficient incorporation
of two 18O atoms is observed for the enzymatically cleaved R-
terminated peptides upon microwave irradiation in the presence
of trypsin (Fig. 4). However, the time scale of this experiment has
been insufficient for even one 18O-atom exchange in most of the
K-terminated peptides (Fig. 4). All K-terminated peptides incorpo-
rating only one 18O atom during trypsinization in H218O (Table 1),
have not exchanged any 18O atoms in this decoupling experi-
ment either. This observation demonstrates that the same two-step
18O labeling reaction conducted using traditional methods also
occurs during microwave-assisted trypsinization. In the first step,
only one 18O-atom is incorporated in the C-terminal upon amide
bond cleavage. The second step, requiring formation of the serine
protease–tryptic peptide substrate complex, is more efficient for
the R-terminated peptides (compared to K-terminated peptides).
It can proceed independently of peptide bond cleavage and thus
0
2580
2590
2600
2610
m/z
Fig. 2. Comparison of MALDI reflectron-TOF spectra (the resolved molecular ion
region) of selected peptides resulting from a 90 s microwave-assisted tryptic diges-
tion in H216O (top panel of each) or H218O (bottom panel of each): (a) Lys-terminated
peptide, (b) Arg-terminated peptide, and (c) disulfide-linked fragment containing
both Lys- and Arg-terminated peptides.
exclusively two 18O atoms (Fig. 2b). No direct effects of trypsin
cleaved peptide length with 18O incorporation efficiency have been
observed for either peptide type (Table 1). As already demonstrated
[19,20], the mechanism explaining the proteolytic incorporation of
two 18O atoms in the C-terminal carboxyl group during proteolysis
is a two-step reaction. In the first step, one 18O atom is incorpo-
rated during peptide bond cleavage. The second step – the exchange
of the second 18O atom in the tryptic peptide – is the rate limit-
ing. This exchange reaction proceeds through the formation of a

28
M.D. Antoine et al. / International Journal of Mass Spectrometry 312 (2012) 24–29
100
100
1480.1
1249.9
16O
16O
(a)
(b)
FKDLGEEHFK
LGEYGFQNALIVR
0
0
1484.1
1249.9
100
100
18O
18O
0
0
1249
1251
1253
1255
m/z
1480
1484
1488 m/z
Fig. 4. Comparison of MALDI TOF spectra (the resolved molecular ion region) of selected peptides in the decoupling experiment: 90 s microwave-assisted tryptic digestion
in H216O (top panel of each); subsequent drying, resuspension in H218O and 90 s irradiation (bottom panel of each): (a) Lys-terminated peptide – no exchange; and (b)
Arg-terminated peptide – two 18O atoms exchanged. For comparison purposes, the spectra of the same peptides as in Fig. 1(a) and (b) are plotted here as well.
results in substitution of up to two 18O atoms for R-terminated
peptides. We note that the time required for complete (two atom)
18O exchange was more than an order of magnitude longer for K-
versus R-terminated peptides, as reported for the original decou-
pling experiments [19]. It has been further argued by Yao et al. that
the 18O substitution reaction in the second step is somewhat sim-
ilar to amide bond formation [19]. In this context, it should also
be noted that microwave heating has been demonstrated to sig-
nificantly reduce reaction times, improve yields of optically pure
products and the overall reproducibility in solid state peptide syn-
thesis [37].
In additional microwave trypsinization experiments, the over-
all microwave irradiation time is extended from 90 s to 5 min.
This time was still not sufficient for complete incorporation of
18O atoms in all K-peptides under these experimental conditions.
Previous reports have also indicated variable rates of incorpora-
tion of 18O atoms in labeled peptides, with the underlying cause
directly attributable to the reaction conditions [7,8]. Thus it is
anticipated that with additional optimization of reaction param-
eters reported herein, namely increasing reaction time, complete
labeling of lysine terminating peptides is achievable [8]. We deter-
mine that trypsin was not only intact after prolonged microwave
irradiation (no autolysis products observed) but it also retained
its activity and specificity (as evidenced by the decoupling exper-
iments). Since the reactions were performed at near neutral pH
(pure water), no back exchange by 16O is observed. The incorpora-
tion of one or two 18O atoms on a short time scale is reproducible.
Analogous results were obtained for other proteins investigated.
This approach allows relative quantification studies to be per-
formed, as well as K- versus R-terminated peptides to be discerned.
In several instances, initial identification of BSA tryptic peptides in
16O water very close in mass was not possible, since the observed
peptide mass was close to more than one predicted peptides masses
from in silico digestion (Table 1, peptide(s) with experimental m/z
649.45, 1511.85 and 2492.02). In these cases, when one of the
predicted peptides terminates in K and the other in R, the num-
ber of experimentally incorporated 18O atoms provides the correct
assignment for the observed peptide. Some of these assignments
are independently confirmed by MS/MS data of the respective unla-
beled peptide (Fig. 5). We note that 18O-specific software analysis
algorithms [19] can be adapted to exploit the advantages of this
rapid labeling technique. In some instances, mass shifts of 6 or 8 Da
have been observed. These occur as a result of 18O incorporation in
the C-termini of two non-adjacent tryptic peptides held together
100
100
b 23
(a)
S
S
LVNELTEFAKTCVADESHAGCEK
0
100
H
a 1
(b)
b 4
T K A
T
L
L
y 23
b 5
a 5
y 14
y 13
y 12
b 3
y 15
0
m/z
500
1000
1500
2000
2500
0
1000
1500
2000
2500
3000
m/z
Fig. 5. MALDI MS/MS (LIFT) spectrum of the unlabeled tryptic peptide at m/z
2492.02. The sequence of the peptide inferred from this spectrum demonstrates
Fig. 6. MALDI TOF mass spectra of peptide fragments generated from the trypsin
digestion of ubiquitin by (a) microwave-assisted digestion and (b) PCR thermocycler
digestion.
that it is Lys-terminated (in agreement with the observed number of exchanged 18
atoms – only one, Table 1).
O

M.D. Antoine et al. / International Journal of Mass Spectrometry 312 (2012) 24–29
29
by an intact disulfide bond. Therefore, the proposed rapid digestion
Appendix A. Supplementary data
and 18O incorporation method can be implemented for identifica-
tion of these and other types of cross-linked peptides in studies of
protein–protein interactions [38].
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ijms.2011.07.005.
Most of the microwave-assisted experiments, described above,
have been repeated in the PCR thermocycler with very similar
results. For the proteolytic digestion of ubiquitin using trypsin, a
similar degree of protein cleavage is obtained, however, slight vari-
ations in relative ion abundances of cleaved fragments is observed
(Fig. 6). This similarity in proteolysis products indicates that the
same thermal activation mechanisms of protein digestion and
labeling operate in both heating devices. The rate at which the
reaction mixture is initially heated, as well as the final temper-
ature of the reaction mixture affects digestion and 18O labeling
efficiency. One advantage of using a PCR thermocycler in bottom-
up proteomics is its ubiquity and accessibility in multiple labs,
and its affordability, compared to specialized microwave hydrolysis
reactors.
References
[1] P. Demirev, C. Fenselau, J. Mass Spectrom. 43 (2008) 1441.
[2] S.P. Gygi, B. Rist, S.A. Gerber, F. Turecek, M.H. Gelb, R. Aebersold, Nat. Biotechnol.
17 (1999) 994.
[3] Y. Oda, K. Huang, F.R. Cross, D. Cowburn, B.T. Chait, Proc. Natl. Acad. Sci. U. S. A.
96 (1999) 6591.
[4] M.B. Goshe, R.D. Smith, Curr. Opin. Biotechnol. 14 (2003) 101.
[5] S. Julka, F. Regnier, J. Proteome Res. 3 (2004) 350.
[6] S.E. Ong, M. Mann, Nat. Chem. Biol. 1 (2005) 252.
[7] J.L. Capelo, R.J. Carreira, L. Fernandes, C. Lodeiro, H.M. Santos, J. Simal-Gandara,
Talanta 80 (2010) 1476.
[8] H. Loftheim, A. Asberg, L. Reubsaet, J. Chromatogr. A 1217 (2010) 8241.
[9] X.D. Yao, A. Freas, J. Ramirez, P.A. Demirev, C. Fenselau, Anal. Chem. 73 (2001)
2836.
[10] I.I. Stewart, T. Thomson, D. Figeys, Rapid Commun. Mass Spectrom. 15 (2001)
2456.
[11] K.J. Reynolds, X.D. Yao, C. Fenselau, J. Proteome Res. 1 (2002) 27.
[12] M. Miyagi, K.C.S. Rao, Mass Spectrom. Rev. 26 (2007) 121.
[13] C. Fenselau, X. Yao, J. Proteome Res. 8 (2009) 2140.
[14] J. Capelo, R. Carreira, L. Fernandes, C. Lodeiro, H. Santos, J. Simal-Gandara,
Talanta 80 (2010) 1476.
[15] K. Rose, L.A. Savoy, M.G. Simona, R.E. Offord, P. Wingfield, Biochem. J. 250 (1988)
253.
[16] T. Takao, H. Hori, K. Okamoto, A. Harada, M. Kamachi, Y. Shimonishi, Rapid
Commun. Mass Spectrom. 5 (1991) 312.
[17] M. Schnolzer, P. Jedrzejewski, W.D. Lehmann, Electrophoresis 17 (1996) 945.
[18] O.A. Mirgorodskaya, Y.P. Kozmin, M.I. Titov, R. Korner, C.P. Sonksen, P. Roep-
storff, Rapid Commun. Mass Spectrom. 14 (2000) 1226.
[19] X.D. Yao, C. Afonso, C. Fenselau, J. Proteome Res. 2 (2003) 147.
[20] C. Fenselau, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 855 (2007) 14.
[21] M.P. Washburn, D. Wolters, J.R. Yates, Nat. Biotechnol. 19 (2001) 242.
[22] C.J. Mason, T.M. Therneau, J.E. Eckel-Passsow, K.L. Johnson, A.L. Oberg, J.E. Olson,
K.S. Nair, D.C. Muddiman, H.R. Bergen, Mol. Cell. Proteomics 6 (2007) 305.
[23] A. Nefedov, M. Gilski, R. Sadygov, J. Proteome Res. 10 (2011) 2095.
[24] B.N. Pramanik, U.A. Mirza, Y.H. Ing, Y.H. Liu, P.L. Bartner, P.C. Weber, M.K. Bose,
Protein Sci. 11 (2002) 2676.
[25] S.S. Lin, C.H. Wu, M.C. Sun, C.M. Sun, Y.P. Ho, J. Am. Soc. Mass Spectrom. 16
(2005) 581.
[26] O.A. Turapov, G.V. Mukamolova, A.R. Bottrill, M.K. Pangburn, Anal. Chem. 80
(2008) 6093.
[27] D. Suckau, A. Resemann, M. Schuerenberg, P. Hufnagel, J. Franzen, A. Holle, Anal.
Bioanal. Chem. 376 (2003) 952.
[28] B. Schilling, R.H. Row, B.W. Gibson, X. Guo, M.M. Young, J. Am. Soc. Mass Spec-
trom. 14 (2003) 834.
[29] R.J. Carreira, M.S. Diniz, J.L. Capelo, Rapid Commun. Mass Spectrom. 25 (2011)
75.
[30] J.L. Capelo, R. Carreira, M. Diniz, L. Fernandes, M. Galesio, C. Lodeiro, H.M. Santos,
G. Vale, Anal. Chim. Acta 650 (2009) 151.
[31] D. Lopez-Ferrer, J. Capelo, J. Vazquez, J. Proteome Res. 4 (2005) 1569.
[32] D. Lopez-Ferrer, K. Petritis, K. Hixson, T. Heibeck, R. Moore, M. Belov, D. Camp,
R. Smith, J. Proteome Res. 7 (2008) 3276.
4. Conclusions
Two rapid and efficient methods for protein digestion and 18O
incorporation during thermally enhanced trypsinization are com-
pared. Both heating methods are rapid – in most cases significant
digestion and incorporation of two 18O atoms into R-terminated
tryptic peptides is achieved in less than 5 min, thus considerably
reducing the time for bottom-up proteomics including quanti-
tation by 18O labeling. The number of 18O atoms incorporated
correlates well with the type of C-terminal amino acid. This
approach is not only useful for quantitative proteomics labeling
but suggests a simple method for discerning K- from R-terminated
tryptic peptides as well. These experiments provide additional
insights into the mechanism of rapid thermally enhanced trypsin
digestion of intact proteins. We demonstrate that up to two
18O atoms can be efficiently exchanged in R-terminated peptides
even when the exchange step is decoupled from the digestion
step. The described rapid methods for incorporation of 18O atoms
can also be useful in approaches for rapid de novo peptide and
protein sequencing. We plan to further optimize the methods
and to test them with other endoproteases as well as differ-
ent solvent systems. The methods are being modified with the
aim to develop reliable protocols for rapid, high-throughput bot-
tom up proteomics and accurate quantification of protein toxin
bio-agents.
[33] H.Y. Zhong, S.L.L. Marcus, L. Li, J. Am. Soc. Mass Spectrom. 16 (2005) 471.
[34] S. Swatkoski, S. Russell, N. Edwards, C. Fenselau, Anal. Chem. 79 (2007) 654.
[35] T.J. Cornish, M.D. Antoine, S.A. Ecelberger, P.A. Demirev, Anal. Chem. 77 (2005)
3954.
[36] N.A. Hagan, J.S. Lin, M.D. Antoine, T.J. Cornish, R.S. Quizon, B.F. Collins, A.B.
Feldman, P.A. Demirev, MALDI mass spectrometry for rapid detection and
characterization of biological threats, in: C. Fenselau, P. Demirev (Eds.), Rapid
Characterization of Microorganisms by Mass Spectrometry, American Chemical
Society, Washington, DC, 2011, pp. 211–224.
Acknowledgements
This work was supported by Johns Hopkins University Applied
Physics Laboratory (JHU/APL) independent research and devel-
opment funds. M.A. acknowledges partial support by the Stuart
S. Janney Sabbatical Fellowship Program at JHU/APL. We also
thank Charlene Bement, Aldo Nascimento, David Deglau and
Kate Kerbel (JHU/APL summer students) for technical help
in optimizing the microwave and PCR thermocycler digestion
experiments.
[37] M. Erdelyi, A. Gogoll, Synth. Stuttg. 11 (2002) 1592.
[38] J.W. Back, V. Notenboom, L.J. de Koning, A.O. Muijsers, T.K. Sixma, C.G. de Koster,
L. de Jong, Anal. Chem. 74 (2002) 4417.