Site-specifically phosphorylated lysine peptides

Article abstract of DOI:10.1021/ja507886s

Protein phosphorylation controls major processes in cells. Although phosphorylation of serine, threonine, and tyrosine and also recently histidine and arginine are well-established, the extent and biological significance of lysine phosphorylation has remained elusive. Research in this area has been particularly limited by the inaccessibility of peptides and proteins that are phosphorylated at specific lysine residues, which are incompatible with solid-phase peptide synthesis (SPPS) due to the intrinsic acid lability of the P(=O)-N phosphoramidate bond. To address this issue, we have developed a new synthetic route for the synthesis of site-specifically phospholysine (pLys)-containing peptides by employing the chemoselectivity of the Staudinger-phosphite reaction. Our synthetic approach relies on the SPPS of unprotected ε-azido lysine-containing peptides and their subsequent reaction to phosphoramidates with phosphite esters before they are converted into the natural modification via UV irradiation or basic deprotection. With these peptides in hand, we demonstrate that electron-transfer dissociation tandem mass spectrometry can be used for unambiguous assignment of phosphorylated-lysine residues within histone peptides and that these peptides can be detected in cell lysates using a bottom-up proteomic approach. This new tagging method is expected to be an essential tool for evaluating the biological relevance of lysine phosphorylation.

Full text of DOI:10.1021/ja507886s

Article
pubs.acs.org/JACS
Site-Specifically Phosphorylated Lysine Peptides
†
,§
§
†
†
†
Jordi Bertran-Vicente, Remigiusz A. Serwa, Michael Schumann, Peter Schmieder, Eberhard Krause,
̈
,
†,‡
and Christian P. R. Hackenberger*
†
Leibniz-Institut fu
̈
r Molekulare Pharmakologie (FMP), Robert-Ro
Department Chemie, Humboldt Universitat zu Berlin, Brook-Taylor-Strasse 2, 12489 Berlin, Germany
r Chemie und Biochemie, Freie Universitat Berlin, Takustrasse 3, 14195 Berlin, Germany
S Supporting Information
̈
ssle Str. 10, 13125 Berlin, Germany
‡
̈
§
Institut fu
̈
̈
*
ABSTRACT: Protein phosphorylation controls major processes in cells.
Although phosphorylation of serine, threonine, and tyrosine and also
recently histidine and arginine are well-established, the extent and
biological significance of lysine phosphorylation has remained elusive.
Research in this area has been particularly limited by the inaccessibility of
peptides and proteins that are phosphorylated at specific lysine residues,
which are incompatible with solid-phase peptide synthesis (SPPS) due to
the intrinsic acid lability of the P(O)−N phosphoramidate bond. To
address this issue, we have developed a new synthetic route for the
synthesis of site-specifically phospholysine (pLys)-containing peptides by
employing the chemoselectivity of the Staudinger-phosphite reaction. Our
synthetic approach relies on the SPPS of unprotected ε-azido lysine-
containing peptides and their subsequent reaction to phosphoramidates with phosphite esters before they are converted into the
natural modification via UV irradiation or basic deprotection. With these peptides in hand, we demonstrate that electron-transfer
dissociation tandem mass spectrometry can be used for unambiguous assignment of phosphorylated-lysine residues within
histone peptides and that these peptides can be detected in cell lysates using a bottom-up proteomic approach. This new tagging
method is expected to be an essential tool for evaluating the biological relevance of lysine phosphorylation.
INTRODUCTION
raises the question of the existence of the complementary
protein lysine phosphatase. Accordingly, Hiraishi et al. reported
the isolation of a phosphatase capable of hydrolyzing P(O)−
■
Reversible protein phosphorylation of specific amino acid side
chains is a key regulatory mark in most cellular processes. The
roles of this post-translational modification (PTM) have been
most widely studied in the context of serine, threonine, and
tyrosine, for which many synthetic, biochemical, and proteomic
1
1
6
N bonds with high specificity toward pLys residues. Although
these biochemical studies point toward the existence and
importance of Lys-phosphorylation, the analytical techniques
employed in those studies could not provide more detailed
information such as the detection of the particular phosphor-
ylation site or the analysis of specific binding partners. In
pursuing this aim, it is therefore essential to develop enhanced
phosphoproteomic tools that are well-suited for detection and
identification of labile Lys-phosphorylation sites in proteins.
The chemical properties of pLys, having a high-energy alkyl-
phosphoramidate bond, might suggest a role in phosphoryl
2
−5
tools are available to probe the biological significance.
In
addition to those, phosphorylation of N-containing amino
acids, e.g., Arg, His, and Lys, is also known to occur in nature,
but the acid lability of the P(O)−N bond has prevented this
modification from being studied extensively. Nevertheless,
recent studies have indicated the potential importance of
6
−8
9,10
phosphohistidine
and phosphoarginine
in different
biological signaling processes, pointing toward the necessity
11
transfer reactions to other more stable entities. Whether this
is being exploited in nature in signaling networks, as it is in the
to focus on the least studied modification from this group, i.e.,
1
1
phospholysine (pLys).
6
,17
case of pHis,
remains unknown.
To date, the biological role of pLys still remains unclear.
More than three decades ago, Smith et al. reported preliminary
evidence of Lys-phosphorylation by looking at nuclear extracts
from Walker-256 cells and regenerating rat liver. Those
contained acid-labile phosphates when they were supplemented
with Histone H1 and γ- P-ATP. The putative kinase was
isolated from the same nuclear extracts and assayed in vitro
against Histone H1 showing phosphorylation on lysine residues
Synthetic peptides are important probes for the study of
18
PTMs. Consequently, the accessibility of naturally modified
peptides is a key requirement. However, the lability of certain
post-translational modifications imposes a synthetic challenge
and is often incompatible with standard solid-phase peptide
synthesis (SPPS). Since P(O)−N bonds in pArg, pHis, and
12
3
2
3
2
after analyzing the protein digests by P labeling or paper
Received: January 20, 2014
Published: September 7, 2014
1
3−15
chromatography.
The detection of a potential lysine kinase
©
2014 American Chemical Society
13622
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
pLys are unstable under acidic conditions, adjustments to the
standard SPPS protocol are sought-after. Recently, pArg
peptides were prepared by SPPS with a trichloroethyl-protected
pArg building block, which increased the stability of the P(
O)−N bond under the acidic deprotection conditions required
our group has employed the Staudinger-phosphite reaction for
the site-specific peptide and protein PEGylation, glycosylation,
and biotinylation or the incorporation of a photocleavable aryl-
phosphoramidate to render pTyr analogues, which already
demonstrated the chemoselectivity of the reaction itself
1
9
22a−e
for SPPS. In another example labile pyrophosphopeptides
were prepared by reaction of a pSer peptides with a P(V)
(Scheme 1A).
Here, the reaction is applied in a similar
fashion but using an aliphatic azidonorleucine instead of the
previously applied aryl-azide moiety. Inspired by the out-
standing scope of the earlier reported reaction, we decided to
inspect the performance of the less reactive aliphatic azides in
the Staudinger-phosphite reaction to form the corresponding
phosphoramidates peptides 3.
20
phosphorimidazolide reagent in solution. To date the only
known synthesis of phosphorylated-lysine polypeptides relies
on the use of potassium phosphoramidate as a phosphorylation
2
1
reagent. This approach, however, does not only lead to the
phosphorylation of all (accessible) lysine residues but also to
the modification of other N-side chains like histidine, arginine,
or the N-terminus. In addition, the already mentioned acid
lability of the P(O)−N bond has hampered its synthesis by
standard SPPS using phospholysine as a building block due to
the required final acidic deprotection.
RESULTS AND DISCUSSION
■
Phosphite Synthesis. An important requirement for the
efficient synthesis of the phosphorylated peptide 4 is an
appropriate choice of deprotection conditions for the
phosphoramidate ester 3 under which the P(O)−N bond
is not cleaved (Scheme 1B). Consequently, the phosphites
employed for the Staudinger-phosphite reaction should already
include cleavable groups that meet these requirements.
Previously, our group and others have used 2-nitrobenzyl
esters-type caged groups to access free phosphoramidates or
Considering the lack of a synthetic route to access site-
specifically phosphorylated Lys peptides, we now present an
approach for their synthesis based on the Staudinger-phosphite
reaction and their subsequent characterization by tandem mass
spectrometry (MS/MS). Most importantly, our synthesis takes
advantage of the chemoselectivity of the Staudinger reaction in
solution, which is performed after the acidic deprotection of
amino acid side chains and the release of peptides from the
resin. Thereby, unprotected ε-azido lysine-containing peptides
2
2a,d,23a,b
phosphomonoesters by mild light-induced photolysis.
The presence of this masking group provides a photoactivatable
approach, allowing spatial and temporal control over the release
of the active modification. In this aspect, a 2-nitrophenyl ethyl
(NPE) caging group is more suitable than 2-nitrobenzyl benzyl
1, which are easily accessed by standard SPPS using N-Fmoc-6-
azidonorleucine as a building block, can be reacted with a UV-
or base-cleavable phosphite triester 2, which delivers after
uncaging or alkaline deprotection of the phosphoramidate 3 the
corresponding pLys peptide 4 (Scheme 1B). In the recent past,
(
NB), because the photolysis byproduct nitrosoacetophenone
is less reactive than the corresponding nitrosobenzalde-
hyde.
23a,b
With this consideration in mind we synthesized
two different photocleavable phosphites, 5 and 6 (Scheme 2).
Scheme 1. Staudinger-Phosphite Reaction (A) of Aryl Azido-
Containing Peptides and (B) on ε-azido Lysine-Containing
Peptides
Scheme 2. Phosphite Synthesis
Those phosphites were accessible by the condensation of
hexamethylphosphorous triamide and the corresponding
24
alcohol. The stability of the water-soluble PEG-photo-
cleavable phosphite 6 in aqueous media allowed us to perform
2
2a
the reaction in aqueous buffers at physiological pH.
Additionally, we synthesized the base cleavable tris(2-
cyanoethyl) phosphite 7 (Scheme 2) to check whether basic
conditions release the free phosphoramidate but also whether
the P(O)−N bond remains stable in the process.
Synthesis of Phosporamidate Lysine-Containing Pep-
tides. At the outset of our studies we checked whether solid-
supported phosphoramidate esters 3 survive acidic deprotection
conditions necessary for standard SPPS. As a model peptide we
chose a short peptide sequence (8) from Histone H1.4
(
K23ARKSAGA30, Figure 1A), a linker Histone previously
11
reported to harbor phosphorylated lysines. The model
peptide was synthesized via standard Fmoc-based SPPS on a
Rink Amide resin and contained 4-chloro-7-nitrobenzofurazan
1
3623
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
Figure 1. Staudinger-phosphite reactions with ε-azido lysine peptides. (A) Staudinger-phosphite reactions on the solid support and in solution. (B)
HPLC-FLD of the reaction mixture of peptide 8 and phosphite 7 at t = 0 h (black) and after 36 h (red). (C) Phosphoramidate esters (11 and 12)
and phosphorylated dipeptide 21.
(
NBD) at the N-terminus to monitor the Staudinger-phosphite
azides relative to the previous reported aryl azides in the
22d
reaction products by fluorescence detection (FLD). N-Fmoc-6-
azidonorleucine was incorporated as a building block and
placed four residues from the N-terminus. After reaction with
commercially available trimethyl phosphite and phosphite 6
Staudinger-phosphite reaction.
At micromolar concentra-
tions, using a similar phosphite we observed that after 24 h and
at room temperature, only 19% conversion to phosphoramidate
10b was formed (Figure S12). This is in contrast to aryl azides,
which proceed in almost quantitative conversion to aryl
(Scheme S8), the peptides were cleaved from the resin with
22d
9
5% trifluoroacetic acid (TFA), 2.5% triisopropylsilane (TIS),
phosphoramidate in less than 8 h.
Therefore, we probed
and 2.5% thioanisole (TA). The cleaved peptides were then
analyzed by HPLC-UV-MS and HPLC-FLD (Figures S10 and
S11). Amine formation (9) was observed as the main product
the reactivity of peptide 8 at milimolar concentrations and at
higher temperature (45 °C) (Figure S15). After 24 h peptide 8
was nearly quantitatively transformed into phosphoramidate
10b using 50 equiv of phosphite 6 (Figure S15). Note that
using dimethylformamide (DMF) as a solvent the same trend
was observed (data not shown). In accordance with these data,
the unprotected peptide 8 was reacted in DMF with phosphite
7 (Figure 1A, Scheme S10) and the formation of the
phosphoramidate Lys peptide 10c was monitored by HPLC-
UV-MS and HPLC-FLD (Figure 1B). After 36 h full
conversion was achieved, and peptide 10c was isolated by
semiprepartive HPLC in good yield (Figure 1A, Table S3). As
shown in the HPLC-FLD trace (Figure 1B) the desired
phosphoramidate was the major product, and no amine (9) or
other peptidic compounds were observed. This result supports
our strategy to generate site-specifically phosphoramidate-Lys-
residues on peptides by the Staudinger-phosphite reaction
utilizing simple and easily accessible starting materials.
(Table S2) in both cases, which indicated hydrolysis of the
P(O)−N bond and confirmed that pLys peptides are not
accessible by reaction with protected solid-supported peptides.
Thus, we decide to study the stability of the phosphor-
amidate ester bond under different conditions with two
different substrates (11 and 12) (Figure 1C, Schemes S3 and
S4). As observed previously, the P(O)−N bond does not
survive harsh acidic conditions (Figure S1, Table S1); however
the phosphoramidate bond is highly stable at 0.1% TFA
concentration (Figure S1), a widely used condition to analyze
and purify peptides by HPLC.
As mentioned before, the Staudinger-phosphite reaction is a
chemoselective reaction and therefore compatible with
22d
unprotected peptide side chains.
Therefore, we aimed for
an approach in solution to generate phosphoramidate-
containing peptides, for which the same ε-azido lysine-
containing H1.4 peptide was cleaved from the resin and
purified by semipreparative HPLC (Figure 1A, Scheme S5).
First, we tested the reactivity of peptide 8 with phosphite 6
using Tris-HCl buffer (pH 8.2) as a solvent (Scheme S9). As
mentioned before, we expected lower reactivity of the aliphatic
Encouraged by these results, we aimed to synthesize a larger
peptide sequence containing more diverse functionalities and
specifically other basic amino acids. Consequently, we prepared
a peptide (13) belonging to the N-terminal domain of Histone
H1 (H1.0 isoform) of the sequence K17ASKK(N )-
3
STDHPKYS29, in which the N-Fmoc-6-azidonorleucine
1
3624
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
building block was placed instead of Lys21. To test the
compatibility of the reaction with cysteine-containing peptides,
we also prepared the same peptide sequence (14) containing a
cysteine residue among others amino acids (R15AKASKK-
observed for 15a is in agreement with the already published
photolysis of caged pSer.
25
No other peptidic products were observed after photolysis.
Note that for the cysteine-containing peptide 16, dithiothreitol
(DTT) was added to prevent side reactions of the released
nitrosoacetophenone with the peptide or phosphate during UV
(
N )STDHPKYSDC31). Analogous to peptide 8, peptides 13
3
and 14 were cleaved from the resin and successfully isolated by
semipreparative HPLC (Schemes S6 and S7, Figures S7 and
S9). Peptide 13 was then reacted with phosphites 5 and 7 in
DMF and with phosphite 6 in Tris-HCl buffer, while peptide
25
irradiation (Figure S24). To our knowledge, these results
show, for the first time, the site-specific phosphorylation of Lys
residues on a peptide level.
1
4 was reacted with phosphite 5 in DMF (Table 1). The
Basic Deprotection of bis(2-Cyanoethyl) Phosphor-
amidate Lysine-Containing Peptides. The base-labile 2-
cyanoethyl group has been widely used as a protecting group
(PG) in phosphoramidites for solid-phase synthesis of
Table 1. Synthesis of Phosphoramidate Peptides 15a−c and
16.
26
oligonucleotides. This base-labile PG can be removed quickly
in concentrated solutions of hydroxide anions via β-elimination.
Nevertheless, we were concerned about the reactive electro-
philic acrylonitrile that is formed after the deprotection, which
could react with the nucleophilic side chains of the unprotected
histone peptide.
Incubation of peptide 15c in 250 mM sodium hydroxide
solution revealed full deprotection after 20 min (Figure 2A,C,
Scheme S16). The main observed product was the desired pLys
peptide 17, whereas no other peptidic side-products were
observed. Remarkably, only a very minimal amount of P(
O)−N hydrolysis was detected (<7% by LC-UV) (Figure S25),
which was in accordance with the higher stability of N-alkyl
phosphoramidates under basic as opposed to acidic con-
11
ditions.
Stability of the Phosphoramidate Bond in Unpro-
tected Phosphorylated Lys-Peptides. A first attempt to
isolate pLys peptide 17 by semipreparative HPLC using an
acetonitrile/water gradient in 0.1% TFA led to the
dephosphorylation of peptide 17 (data not shown). We
therefore decided to check the stability of the P(O)−N
bond under different conditions. To accomplish this we
synthesized a phosphorylated-lysine dipeptide 21 (Figure 1C)
obtained by basic deprotection of the bis(2-cyanoethyl)
phosphoramidate 20 (Scheme S17). The crude dipeptide 21
was incubated under neutral, slightly acidic and acidic
conditions (Tables S4 and S5). The decay of the P(O)−N
a
Overall peptide synthesis yield (incl. Fmoc-SPPS) in brackets.
b
Subsequent addition of phosphite (2 × 5 equiv) over 24−48 h,
c
4
4
5 °C. Subsequent addition of phosphite (3 × 5 equiv) over 9 h,
5 °C.
reactions were followed by HPLC-UV-MS (Figure S16), and
the reaction mixtures were left until depletion of the ε-azido
lysine-containing peptides 13 and 14. Within the range of 9 and
1
31
bond was monitored by H, P-HMBC NMR. In <2 h, half of
the phosphoramidate bond in dipeptide 21 was hydrolyzed
both in 0.05% formic acid and in 0.1% trifluoroacetic acid
(Table S5, Figure S31). In contrast, after 18 h, the
phosphoramidate dipeptide 21 was still the major product at
pH 4.5 (59%) and pH 7.5 (69%) (Table S4, Figure S30).
Although these data confirm the instability of the P(O)−N
bond at very acidic pH, it also suggests that Lys-phosphorylated
peptides could be analyzed by short-gradient HPLC runs
employing acidic solvents.
4
8 h the corresponding phosphoramidates 15a−c and 16 were
obtained in excellent conversions, and the reaction mixtures
were purified again by semipreparative HPLC. Peptides 15a−c
and 16 were isolated in good yields (Table 1, Figures S18−
S21) and characterized by high-resolution (HR)-MS and 31
P
NMR (Figures S57−S60). As mentioned before, the
phosphoramidate ester bond was stable during analysis and
purification by HPLC under standard conditions.
Uncaging of Photocleavable Phosphoramidates
Lysine-Containing Peptides. After the isolation of the
caged peptides 15a, 15b, and 16, our goal was to cleave the
phosphoramidate ester under mild conditions to release the
natural pLys modification without harming the acid-labile
phosphoramidate bond. Following our reported UV-depro-
tection of nitrobenzyl ester-type caged aryl phosphor-
We then performed the purification with an acetonitrile/
water gradient in alkaline aqueous buffer (pH 9.2) as a mobile
phase (Figure S33). Under these conditions, no hydrolysis of
the P(O)−N bond was observed, and peptides 17 and 18
were successfully isolated in 42% and 38% yield, respectively
(Figures S33 and S34). Similar conditions have been used
previously for the purification of phosphorylated-arginine
22a,d
amidates,
we were able to uncage all peptides, 15a, 15b,
19
and 16, thus providing the pLys peptides 17 and 18 as main
products (Schemes S13−S15, Figures S22−S24). Interestingly,
irradiation of 15a at 355 nm for 15 s rendered full deprotection
as observed by HPLC-UV (Figure 2A,B). Phosphoramidate
peptides without indication of hydrolysis. With peptide 17
in hand, we also studied the thermal stability of the
phosphorylated-lysine peptide (Scheme S18, Figure S32). As
expected, higher temperatures (60−85 °C) caused rapid
hydrolysis of the P(O)−N bond, whereas lower temper-
atures (2 °C) seemed to be the best conditions to store the
1
5b was fully deprotected after irradiation under the same
conditions for 30 s (Figure S23). The faster uncaging rate
1
3625
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
Figure 2. (A) Uncaging of peptide 15a (left) and basic deprotection of peptide 15c (right). (B and C) HPLC-UV traces before and after
deprotection.
phosphorylated-lysine peptides. At room temperature and pH
.4, phosphoramidate hydrolysis in 17 occurred (31% by LC-
UV), similarly to the dipeptide 21 under the same conditions
technique for modified peptides, specifically for those
28
7
containing O-phosphorylated residues. Although CID has
been applied successfully in many studies, the analysis is often
hampered by a neutral loss of the phosphate group, thus
preventing the unambiguous assignment of the modification
(Figures S30 and S32).
Localization of Phosphorylation by Use of Electron-
29
Transfer Dissociation Mass Spectrometry (ETD-MS).
Having succeeded in establishing a synthetic route to access
site-specifically phosphorylated-lysine peptides, we focused on
the characterization of the corresponding phosphorylated-lysine
peptides. First, we used NMR to characterize the 31P NMR
signal for peptides 17 and 18. A unique signal peak was
observed in the 3 P NMR spectra for both peptides (Figures
site. This complication is even more severe in the case of the
more labile phosphoramidate bond for N-phosphorylated
30
peptides.
For the analysis of labile PTMs such as arginine and lysine-
phosphorylation radical-driven fragmentation techniques such
31
as electron capture dissociation (ECD) and electron-transfer
1
32
dissociation (ETD) were found to be more useful than CID
3
1
9,30
S61 and S62). To demonstrate that this P NMR signal was
due to the reduced elimination of phosphoric acid.
In fact,
1
31
attached to a lysine residue, H− P HMBC was performed.
The ε-methylene hydrogen atoms from the lysine side-chain
showed coupling to the phosphorus signal peak, thus
confirming that the phosphate was attached to a lysine side
chain (Figures S61 and S62). However, localization of the
phosphorylated-lysine residue using NMR would require a full
sequential assignment of all peptide resonances. We therefore
decided to use mass spectrometry (MS), with the intention of
unambiguous localization of the site of phosphorylation.
Tandem mass spectrometry (MS/MS) has become the method
of choice for analysis of PTMs of peptides or proteins because
of the high sensitivity and high-throughput capabilities when
ECD was applied to determine the phosphorylation site of
different lysine-phosphorylated peptides, which were synthe-
21
sized by using a nonsite-specific approach. Compared to CID,
ECD showed fragment ion spectra, allowing the localization of
N-phosphorylation in the corresponding pLys peptides.
However, the lack of a synthetic route to access site-specifically
phosphorylated-lysine peptides together with the high instabilty
of the P(O)−N bond in the gas phase suggested that the
phospho-site assignment by ECD should be done with
21
caution.
With these considerations in mind, we proceeded with the
analytical characterization of peptides 17 and 18 (Figures 2A
and 3A). As expected, CID showed complete neutral loss of
phosphate (data not shown). No phosphorylated fragment ions
2
7
coupled to nanoliquid chromatography (nLC). Collision
induced-dissociation (CID) is the most common fragmentation
1
3626
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
Figure 3. (A) ESI-MS spectra showing the triply charged precursor ion
of 17. (B) ETD fragment ion spectra of peptide 17. Individual
fragments are labeled according to the c- or z-ion nomenclature. The
characteristic mass difference of the phosphorylated Lys21 is shown in
the Figures S38 and S39.
Figure 4. (A) Detection workflow of chymotryptic peptide 22 in a cell
lysate context. (B) Base peak intensity chromatogram of chymotryptic
peptide mixture. Extracted ion chromatogram of the doubly/triply/
quadriply charged precursor ion of phosphorylated-lysine peptide 22.
for the identification of phosphorylated lysine could be found in
any tandem MS fragment preventing the localization of
phosphorylation with this method. In contrast, ETD
fragmentation verified phosphorylation at the lysine side
chain in position 21 (Figures 3B, S38, S39, S43, and S44).
Fragment ions indicating phosphorylation of other amino acid
side-chains such as Arg15, Lys17, Lys20, Lys27, Ser19, Ser22,
Thr23, His25, Tyr28, Ser29, Asp30, or Cys31 could not be
observed. This result confirms the chemoselectivity of the
Staudinger-phosphite reaction and indicates that migration of
the phosphate in the gas phase does not play a role during ETD
analysis of peptides 17 and 18.
confirms the phosphorylation site. However, despite the
addition of a phosphatase inhibitor (Na VO ), a significant
3
4
loss of phosphate was observed which cannot be explained
based on the intrinsic instability of phosphorylated-lysine
peptides at neutral pH (Figures S30 and S32). Possibly other
enzymes might have an influence on the stability of
phosphorylated-lysine peptides in a cell lysate context.
These results confirm that with the applied UV light and
basic deprotection conditions the phosphorylation remains on
the desired modification site to a larger extent and no
phosphate migration to other phosphate acceptors occurs.
Detection of a Synthetic Phosphorylated-Lysine
Peptide in a Proteome Context. Having established both
synthetic and analytical tools to study Lys-phosphorylation, we
aimed to detect phosphorylated-lysine peptides in a more
complex sample. Our goal was not only to detect phosphory-
lated-lysine peptides in a cell lysate but also to ensure that
phosphorylated-lysine peptides survive the conditions of a
bottom-up proteomic analysis. We therefore incubated
synthetic peptide 17 at various concentrations in a whole-cell
lysate of HEK293 cells (Figure 4A, Scheme S19, and Table S6).
After incubation for 1 h, proteins were digested with
chymotrypsin. Because of the basic character of the histone-
derived Lys-phosphorylated peptide, chymotrypsin instead of
trypsin was used for enzymatic digestion. The resulting peptide
mixture was analyzed using a two-dimensional reversed-phase
CONCLUSION
■
In summary, we have developed the first synthesis of site-
specifically phosphorylated-lysine peptides. Our synthetic route
takes advantage of the chemoselectivity of the Staudinger-
phosphite reaction of straightforwardly accessible unprotected
ε-azido lysine-containing peptides with phosphites. The
method circumvents the final acidic treatment of the acid-
sensitive alkyl-phosphoramidate moiety. Furthermore, we
employed phosphoramidate esters as precursors for the natural
modification, which can be liberated by photolysis or applying
basic conditions. Finally, we were able to demonstrate the site-
specific localization of phosphorylated lysine by ETD tandem
MS. Referring to previous ECD results, our data indicate that
the stability of the phosphoramidate bond during the ETD
process is high, keeping the pLys side-chain completely intact
during fragmentation. This makes ETD mass spectrometry
particularly suitable for analysis of phospholysine peptides.
Consequently, ETD can also be used for proteomic studies of
Lys-phosphorylated proteins. With this chemical approach in
hand, we expect to contribute in the near future toward
understanding the role of lysine phosphorylation and its
21
33
(
2-D) LC-MS/MS approach. The expected chymotryptic
peptide 22 was detected at a concentration as low as 0.4 μM
Figure 4B, S45, and Table S6). Characterization by ETD
(
1
3627
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628

Journal of the American Chemical Society
Article
biological impact. Further studies in that direction are currently
underway in our laboratories.
(21) Kowalewska, K.; Stefanowicz, P.; Ruman, T.; Fraczyk, T.; Rode,
W.; Szewczuk, Z. Biosci. Rep. 2010, 30, 433−443.
(22) (a) Serwa, R.; Horstmann, B.; Majkut, P.; Swiecicki, J.-M.;
Gerrits, M.; Krause, E.; Hackenberger, C. P. R. Chem. Sci. 2010, 1,
96−602. (b) Bohrsch, V.; Mathew, T.; Zieringer, M.; Vallee, R. M. J.;
Artner, L. M.; Dernedde, J.; Haag, R.; Hackenberger, C. P. R. Org.
Biomol. Chem. 2012, 10, 6211−6216. (c) Bohrsch, V.; Serwa, R.;
Majkut, P.; Krause, E.; Hackenberger, C. P. R. Chem. Commun. 2010,
6, 3176−3178. (d) Serwa, R.; Wilkening, I.; Del Signore, G.;
Muhlberg, M.; Claussnitzer, I.; Weise, C.; Gerrits, M.; Hackenberger,
ASSOCIATED CONTENT
Supporting Information
Experimental procedures including characterization data for all
new compounds, NMR spectra, HPLC-UV or FLD traces as
well as MS and tandem MS spectra. This material is available
free of charge via the Internet at http://pubs.acs.org.
■
5
̈
́
*
S
̈
4
̈
C. P. R. Angew. Chem., Int. Ed. 2009, 48, 8234−8239. (e) Nischan, N.;
Chakrabarti, A.; Serwa, R. A.; Bovee-Geurts, P. H. M.; Brock, R.;
Hackenberger, C. P. R. Angew. Chem., Int. Ed. 2013, 52, 11920−11924.
(23) (a) Goguen, B. N.; Aemissegger, A.; Imperiali, B. J. Am. Chem.
Soc. 2011, 133, 11038−11041. (b) Givens, R. S.; Kueper, L. W., III
Chem. Rev. 1993, 93, 55−66.
AUTHOR INFORMATION
Corresponding Author
hackenbe@fmp-berlin.de
■
Notes
(
24) Skopek, K.; Gladysz, J. A. J. Organomet. Chem. 2008, 693, 857−
66.
25) Nguyen, A.; Rothman, D. M.; Stehn, J.; Imperiali, B.; Yaffe, M.
B. Nat. Biotechnol. 2004, 22 (8), 993−1000.
26) Sinha, N. D.; Biernat, J.; McManus, J.; Ko
The authors declare no competing financial interest.
8
(
ACKNOWLEDGMENTS
■
(
̈
ster, H. Nucleic Acids
The authors acknowledge support from the DFG (SFB 765 and
SPP 1623), the Fonds der Chemischen Industrie, the Einstein
Foundation, and the Boehringer-Ingelheim Foundation (Plus 3
award). We thank Dr. Gerrits, Dr. Majkut, and Iris Claußnitzer
the use of the laser (Spectron Laser Systems). We would like to
dedicate the article to Prof. Hans Reißig on the occasion of his
Res. 1984, 12, 4539−4557.
(27) Pandey, A.; Mann, M. Nature 2000, 405, 837−846.
(28) Garcia, B. A.; Busby, S. A.; Barber, C. M.; Shabanowitz, J.; Allis,
C. D.; Hunt, D. F. J. Proteome Res. 2004, 3, 1219−4557.
(29) Boersema, P. J.; Mohammed, S. J.; Heck, A. J. R. J. Mass
Spectrom. 2009, 44, 861−878.
(30) Schmidt, A.; Ammerer, G.; Mechtler, K. Proteomics 2013, 13,
6
5th birthday.
9
45−954.
31) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem.
Soc. 1998, 120, 3265−3266.
32) 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.
33) Stephanowitz, H.; Lange, S.; Lang, D.; Freund, C.; Krause, E. J.
Proteome Res. 2012, 11, 1175−1183.
(
REFERENCES
■
(
(
(
1) Hunter, T. Cell 2000, 100, 113−127.
(
2) Johnson, S. A.; Hunter, T. Nat. Biotechnol. 2004, 22, 1093−1094.
3) Olsen, J. V.; Blagoev, B.; Gnad, F.; Macek, B.; Kumar, C.;
(
Mortensen, P.; Mann, M. Cell 2006, 127, 635−648.
4) Kalume, D. E.; Molina, H.; Pandey, A. Curr. Opin. Chem. Biol.
003, 7, 64−69.
5) Tarrant, M. K.; Cole, P. A. Annu. Rev. Biochem. 2009, 78, 797−
25.
6) Attwood, P. V.; Piggott, M. J.; Zu, X. L.; Besant, P. G. Amino
Acids 2007, 32, 145−156.
7) Kee, J.-M.; Oslund, R. C.; Perlman, D. H.; Muir, T. W. Nat.
Chem. Biol. 2013, 9, 416−421.
8) Kee, J.-M.; Villani, B.; Carpenter, L. R.; Muir, T. W. J. Am. Chem.
Soc. 2010, 132, 14327−14329.
9) Fuhrmann, J.; Schmidt, A.; Spiess, S.; Lehner, A.; Turgay, K.;
Mechtler, K.; Charpentier, E.; Clausen, T. Science 2009, 324, 1323−
327.
10) Elsholz, A. K. W.; Turgay, K.; Michalik, S.; Hessling, B.; Gronau,
K.; Oertel, D.; Maderd, U.; Bernhardt, J.; Becher, D.; Hecker, M.;
Gerth, U. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 7451−7456.
11) Besant, P. G.; Attwood, P. V.; Piggott, M. J. Curr. Protein Pept.
Sci. 2009, 10, 536−550.
12) Smith, D. L.; Bruegger, B. B.; Halpern, R. M.; Smith, R. A.
Nature 1973, 246, 103−104.
13) Smith, D. L.; Chen, C. C.; Bruegger, B. B.; Holtz, S. L.; Halpern,
R. M.; Smith, R. A. Biochemistry 1974, 13, 3780−3785.
14) Chen, C. C.; Smith, D. L.; Bruegger, B. B.; Halpern, R. M.;
Smith, R. A. Biochemistry 1974, 13, 3785−3789.
15) Chen, C. C.; Bruegger, B. B.; Kern, C. W.; Lin, Y. C.; Halpern,
R. M.; Smith, R. A. Biochemistry 1977, 16, 4852−4855.
16) Hiraishi, H.; Yokoi, F.; Kumon, A. Arch. Biochem. Biophys. 1998,
(
2
(
8
(
(
(
(
1
(
̈
(
(
(
(
(
(
3
(
(
(
49, 381−387.
17) Robinson, V. L.; Stock, A. M. Structure 1999, 7, 47−53.
18) Schwarzer, D. J. Pept. Sci. 2010, 16, 530−537.
19) Hofmann, F. T.; Lindemann, C.; Salia, H.; Adamitzki, P.;
Karanicolas, J.; Seebeck, P. F. Chem. Commun. 2011, 47, 10335−
0337.
20) Marmelstein, A. M.; Yates, L. M.; Conway, J. H.; Fiedler, D. J.
Am. Chem. Soc. 2014, 136, 108−111.
1
(
1
3628
dx.doi.org/10.1021/ja507886s | J. Am. Chem. Soc. 2014, 136, 13622−13628