Amino acid building blocks for efficient Fmoc solid-phase synthesis of peptides adenylylated at serine or threonine

Article abstract of DOI:10.1021/ol2024696

The first straightforward building block based (non-interassembly) synthesis of peptides containing adenylylated serine and threonine residues is described. Key features include final global acidolytic protective group removal as well as full compatibility with standard Fmoc solid-phase peptide synthesis (SPPS). The described Thr-AMP SPPS-building block has been employed in the synthesis of the Thr-adenylylated sequence of human GTPase CDC42 (Ac-SEYVP-T(AMP)-VFDNYGC-NH₂). Further, we demonstrate proof-of-concept for the synthesis of an Ser-adenylylated peptide (Ac-GSGA-S(AMP)-AGSGC-NH₂) from the corresponding adenylylated serine building block.

Full text of DOI:10.1021/ol2024696

ORGANIC
LETTERS
2011
Vol. 13, No. 22
6014–6017
Amino Acid Building Blocks for Efficient
Fmoc Solid-Phase Synthesis of Peptides
Adenylylated at Serine or Threonine
Michael F. Albers, Bart van Vliet, and Christian Hedberg*
Department of Chemical Biology, Max-Planck Institute of Molecular Physiology,
Otto-Hahn Strasse 11, D-44227, Dortmund, Germany
christian.hedberg@mpi-dortmund.mpg.de
Received September 12, 2011
ABSTRACT
The first straightforward building block based (non-interassembly) synthesis of peptides containing adenylylated serine and threonine residues is
described. Key features include final global acidolytic protective group removal as well as full compatibility with standard Fmoc solid-phase
peptide synthesis (SPPS). The described Thr-AMP SPPS-building block has been employed in the synthesis of the Thr-adenylylated sequence of
human GTPase CDC42 (Ac-SEYVP-T(AMP)-VFDNYGC-NH2). Further, we demonstrate proof-of-concept for the synthesis of an Ser-adenylylated
peptide (Ac-GSGA-S(AMP)-AGSGC-NH2) from the corresponding adenylylated serine building block.
Adenylylation, also known as AMPylation, or adenyla-
tion, is a recently rediscovered posttranslational modifica-
tion, of which much is still unknown.1 Enzymes that
mediate the posttranslational addition of adenosine mono-
phosphate (AMP) on to amino acid side chains of proteins
have gained interest as an important regulatory mecha-
nism, especially in a prokaryotic context.2 Adenosine phos-
phodiester transferases catalyze the addition of AMP onto
amino acid residues, like tyrosine or threonine, thus form-
ing a phosphodiester bond (Figure 1).1,2 So far, protein
adenylylation on a serine residue has not been reported but
is expected to occur, in analogy to intracellular phospho-
rylation and glycosylation.
Recently, it has been discovered that certain pathogenic
bacteria are able to influence eukaryotic cell signaling by
adenylylating small GTPases on either tyrosine or threo-
nine residues.3,4 GTPases act as molecular switches when
hydrolyzing GTP to GDP, thus regulating signaling at a
cellular level.5 The Vibrio parahemolyticus effector protein
VopS adenylylates a specific threonine in the Rho (Thr37
(3) Yarbrough, M. L.; Li, Y.; Kinch, L. N.; Grishin, N. V.; Ball,
H. L.; Orth, K. Science 2009, 323, 269–272.
(1) In recent literature, “adenylylation” and “adenylation” are both
used to describe the addition of adenosine phosphodiesters to proteins.
Previously, the terminology “adenylation” was primarily used in the
field of RNA chemistry. For a general introduction to protein adeny-
lylation, see: Itzen, A.; Blankenfeldt, W.; Goody, R. S. Trends Biochem.
Sci. 2011, 36 (4), 221–228 and references therein.
€
€
(4) Muller, M. P.; Peters, H.; Blumer, J.; Blankenfeldt, W.; Goody,
R. S.; Itzen, A. Science 2010, 329, 946–949.
(5) For a recent review covering the topic of small GTPases as
regulators of vesicular transport, see: Itzen, A.; Goody, R. S. Sem. Cell
Dev. Biol. 2011, 22, 48–56.
(2) Stadtman, E. R. J. Biol. Chem. 2001, 276, 44357–44364.
r
10.1021/ol2024696
Published on Web 10/26/2011
2011 American Chemical Society

followed by completion of the peptide sequence, phos-
phitylation on the solid phase, coupling of the adenosine
moiety by H-phosphonate activation (Pybop), and sub-
sequent oxidation. The peptide is then removed from the
solid support by acidolytic cleavage. The interassembly
approach to adenylylated Ser- and Thr-containing pep-
tides carriesa numberof inherentproblems, suchaslimited
applicability to oxidation-sensitive peptide sequences
(methionine and tryptophanoxidation), pooroverallyield,
and low functional group compatibility.9 Although limited
in synthetic scope, the interassembly protocol was recently
used to prepare peptides containing adenylylated Thr,
which were successfully employed as antigens for raising
of antibodies.10 With our applications in mind, we found
this protocol incompatible with our demands on synthesis
flexibility, as well as functional group tolerance.
Initially, we investigated serine and threonine building
blocks in the form of phosphotriesters masked by
β-cyanoethyl protection (CNE) in analogy to caged phospho-
amino acid building blocks.11 In our previous work on
tyrosine adenylylation, we successfully employed β-cya-
noethyl protection, which was cleaved under the first
piperidine treatment after AMP-building block coupling,
thus leading to clean conversion into the unprotected
phosphodiester.9 In the case of fully protected adenyly-
lated threonine and serine (1, 2), β-cyanoethyl-protection
led to a large amount of β-elimination product under
standard coupling conditions (Figure 2, A). In this case,
peptide coupling could only be carried out under base-free
conditions, which resulted in incomplete couplings, as well
as prolonged reaction times, not compatible with efficient
parallel peptide synthesis. Therefore, we envisioned a more
general strategy, based on unprotected phosphodiester
intermediates such as 3 and 4, as the phosphodiester is a
poor leaving group in monoanionic form, it should sup-
press β-elimination product formation (Figure 2, B).
The general strategy for the synthesis of fully protected
adenylylated building blocks of serine and threonine is
outlined in Scheme 1. NR-Fmoc-serine (5) and NR-Fmoc-
threonine (6) were converted to corresponding allyl esters
(7, 8) by treatment with allyl bromide under standard
conditions.12 Next, N6,N6-bis-Boc-20,30-isopropylidenea-
denosine (9) was treated with 1-(allyloxy)-1-chloro-N,N-
diisopropylphosphinamine (10), under basic conditions
(DIPEA) yielding building block 11.13,14 We found it
essential to purify 11 over a short silica gel column in order
to achieve high yields in the subsequent steps. Compound
11 is highly unstable and should be used immediately after
Figure 1. Adenylylation of protein substrate at tyrosine and
threonine residues by adenylyl transferase.
in RhoA, Thr35 in Cdc42 and Rac1) subfamily of small
GTPases,3 whereas the protein IbpA of Histophilus somni
adenylylates a tyrosine residue of the same proteins (Tyr34
in RhoA, Tyr32 in Cdc42 and Rac1).6 The substrates of the
effector protein DrrA/SidM from the human pathogen
Legionella pneumophila are GTPases from the Rab-sub-
family, with the modified tyrosine being in the switch
II-region (Tyr77 in Rab1b).4 VopS and IbpA belong to the
FIC family, a protein family withmore than2700 members
of sequentially homologous proteins.7 Although the sub-
strates of VopS and IbpA have been identified, the physi-
ological protein substrates of DrrA and the remaining
members of the FIC family proteins are less clear. We
hypothesizedthat identificationofphysiological substrates
of adenylylating proteins (e.g., FIC domains) could be
aided by antibodies that specifically recognize adenylyl-
ated proteins in eukaryotic cells or cell lysates. Recently,
we reported a convenient synthesis of an AMP-tyrosine
building block for standard Fmoc solid-phase peptide
synthesis, as well as generation of anti-adenylyl-Tyr
antibodies.8 Herein, we present the extension from adenylyl-
ated tyrosine residues to adenylylated threonine and serine
containing peptides by solid-phase peptide synthesis, ac-
cording to the standard Fmoc-protocol, with preformed
stable building blocks. Until now, peptides adenyl-
ylated at serine or threonine have only been avaliable
through synthesis via the interassembly approach. In the
interassembly protocol, the peptide is constructed under
SPPS conditions until the amino acid to be adenylylated (Ser,
Thr), which is incorporated without side-chain protection,
(9) Al-Eryania, R. A.; Yan, L.; Ball, H. L. Tetrahedron Lett. 2010,
51 (13), 1730–1731.
(10) Hao, Y. H.; Chuang, T.; Ball, H. L.; Luong, P.; Li, Y.; Flores-
Saaib, R. D.; Orth, K. J. Biotechnol. 2011, 151 (3), 251–254.
(11) Rothman, D. M.; Vazquez, M. E.; Vogel, E. M.; Imperiali, B.
J. Org. Chem. 2003, 68 (17), 6795–6798.
(6) Worby, C. A.; Mattoo, S.; Kruger, R. P.; Corbeil, L. B.; Koller,
A.; Mendez, J. C.; Zekarias, B.; Lazar, C.; Dixon, J. E. Mol. Cell 2009,
34 (1), 93–103.
(12) Ficht, S.; Payne, R. J.; Guy, R. T.; Wong, C.-H. Chem.;Eur. J.
2008, 14 (12), 3620–3629.
(7) Kinch, L. N.; Yarbrough, M. L.; Orth, K.; Grishin, N. V. PLoS
ONE 2009, 4 (6), e5818.
(13) Ikeuchi, H.; Meyer, M. E.; Ding, Y.; Hiratake, J.; Richards,
N. G. J. Bioorg. Med. Chem. 2009, 17 (18), 6641–6650.
(14) Wippoa, H.; Recka, F.; Kudicka, R.; Ramaseshanb, M.;
Ceulemansa, G.; Bollia, M.; Krishnamurthy, R.; Eschenmoser, A.
Bioorg. Med. Chem. 2001, 9 (9), 2411–2428.
€
€
(8) Smit, C.; Blumer, J.; Eerland, M. F.; Albers, M. F.; Muller, M. P.;
Goody, R. S.; Itzen, A.; Hedberg, C. Angew. Chem., Int. Ed. 2011, 50,
9200–9204.
Org. Lett., Vol. 13, No. 22, 2011
6015

Scheme 1. Synthesis of Protected Adenylylated Amino Acid
Building Blocks 14 and 15
Figure 2. (A) CNE-protected Ser/Thr-adenosine phosphotries-
ters (1 and 2) cause β-elimination upon application to standard
Fmoc-SPPS-protocol. (B) Analogous unprotected phosphodie-
ster building blocks (3, 4) are applicable to the standard Fmoc-
SPPS protocol.
preparation. Coupling between 11 with amino acid esters 7
and 8 under optimized conditions, employing tetrazole (2
equiv) as activator, followed by in situ oxidation (TBHP, 5
M in decane, 2 equiv) led to the corresponding phospho-
triesters 12 and 13 in excellent isolated yields. Attempts to
perform the coupling in a reversed manner (phosphoro-
amidite on amino acid, coupling with 9) did not result in
significant product formation and were not considered
further. Column purification resulted in analytically pure
allyl phosphotriesters 12 and 13 as 1:1 diastereomeric mix-
tures at phosphorus. Simultaneous deallylation of the
carboxylic acid allyl ester functionality and the allyl phos-
photriester was carried out with tetrakis(triphenylphos-
phine)palladium (Pd(PPh3)4, 5 mol %) as catalyst, phe-
nylsilane (PhSiH3, 3 equiv) as nucleophile, and 2,6-lutidine
(2.5 equiv) as acid scavenger.15 We found the addition of
2,6-lutidine tobeessentialfor achieving high isolated yields
of stable products. Omitting the amine base, or employing
a stronger base (DIPEA), resulted in lower yields and
inferior purity of isolated products. The resulting crude
deallylation reactions were evaporated to dryness, dis-
solved in minimum amounts of acetonitrile, and applied
to reversed-phase (C18) Sep-pak cartridges, subsequently
eluted with a waterꢀacetonitrile gradient. Lyophilization
of the product-containing fractions resulted in final build-
ing blocks 14 and 15 in high yields as mono-2,6-lutidine
salts. As a guideline, we typically employed 10 g of Sep-pak
synthesis of 14 and 15 has been carried out on a multigram
scale with consistent results.
Next, we evaluated building blocks 14 and 15 in stan-
dard solid-phase peptide synthesis according to the Fmoc
protocol. We chose to prepare human CDC42 sequence
SEYVP-T(AMP)-VFDNYG, which was assembled on
Tentagel resin functionalized with a RAM-anchored
Fmoc-(Trt)-cysteine amide (16, Scheme 2), resulting in
Ac-SEYVP-T(AMP)-VFDNYGC-NH2 (17) suitable
for immobilization in various applications through the
C-terminal cysteine amide.16 Coupling of the Fmoc-amino
acids (10 equiv) was carried out using standard HBTU/
HOBt activation on an automated peptide synthesizer,
except for the protected adenylylated building blocks 14
and 15 (2.5 equiv), which were coupled manually employ-
ing HATU/HOAt as activating reagent, followed by cou-
pling of residual Fmoc-amino acids according to the
automated protocol.17ꢀ19 The N-terminal Fmoc-group
was removed, and the resulting free N-terminus was cap-
ped with an N-acetyl group. Subsequent detachment from
(16) Acquired from Rapp Polymere GmbH, www.rapp-polymere.
com, accessed October 24, 2011.
C18 material per gram of crude deallylation product. The
(17) Dourtoglou, V.; Gross, B.; Lambropoulou, V.; Ziodrou, C.
Synthesis 1984, 572–574.
(18) Automated peptide synthesis was carried out on a Applied
Biosystems peptide synthesizer.
(19) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397–4398.
(15) Dessolin, M.; Guillerrez, M. G.; Thieriet, N.; Guibe, F.; Loffet,
A. Tetrahedron Lett. 1995, 36, 5741–5744.
6016
Org. Lett., Vol. 13, No. 22, 2011

hydrolysis of the isopropylidene acetal functionality of the
adenosine moiety.8 Concentration in vacuo at ambient
temperature and trituration with diethyl ether yielded the
crude peptide. Only traces of depurinated peptide could be
detected by LCꢀESIꢀMS. After preparative reversed-
phase HPLC purification (C18) and subsequent lyophiliza-
tion, 17 was isolated in 41% yield (from resin loading).
To verify the generality of the methodology, we synthe-
sized one additional peptide sequence carrying adenyly-
lated threonine (Ac-GSGA-T(AMP)-AGSGC-NH2 (18),
37%) as well asa peptide carrying an adenylylated serine in
the identical sequence(Ac-GSGA-S(AMP)-AGSGC-NH2
(19), 47%).
Scheme 2. Solid-Phase Peptide Synthesis of Adenylylated
Peptides 17ꢀ19 According to the Standard Fmoc Protocol
In conclusion, we have developed a facile and efficient
synthesis of peptides containing adenylylated serine and
threonine amino acids, which were introduced by standard
Fmoc-based SPPS. The general availability of such build-
ing blocks will greatly simplify the synthesis of peptide
sequence incorporating adenylylated serine or threonine
amino acids, thus extending our previous work on adenylyl-
ated tyrosine peptides. The potential biological applica-
tions of the complete toolbox of adenylylated peptides are
numerous, such as standards for proteomics analysis,
antigens for immunization, as well as pull-down probes.
A number of serine, threonine, and tyrosine peptides have
been synthesized in our laboratory, which are currently
under investigation and will be reported in due course.
Acknowledgment. We are grateful to Prof. Herbert
Waldmann, Department of Chemical Biology, Max
Planck Institute of Molecular Physiology, and Prof. Roger
S. Goody, Department of Biophysical Chemistry, Max
Planck Institute of Molecular Physiology, for their gener-
ous and unconditional support.
the resin, as well as global deprotection, was carried out by
applying a mixture of trifluoroacetic acid (TFA)/triisopro-
pylsilane (TIPS)/H2O (90:5:5). After filtering from the
resin, an additional 10% H2O (v/v) was added to the cleav-
age mixture, which was aged for 30 min to ensure complete
Supporting Information Available. Experimental pro-
cedures and full spectroscopic data for all new com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
Org. Lett., Vol. 13, No. 22, 2011
6017