Synthesis of S-adenosyl-L-homocysteine capture compounds for selective photoinduced isolation of methyltransferases

Article abstract of DOI:10.1002/cbic.200900349

Understanding the interplay of different cellular proteins and their substrates is of major interest in the postgenomic era. For this purpose, selective isolation and identification of proteins from complex biological samples is necessary and targeted isolation of enzyme families is a challenging task. Over the last years, methods like activity-based protein profiling (ABPP) and capture compound mass spectrometry (CCMS) have been developed to reduce the complexity of the proteome by means of protein function in contrast to standard approaches, which utilize differences in physical properties for protein separation. To isolate and identify the subproteome consisting of S-adenosyl-L-methionine (SAM or AdoMet)-dependent methyltransferases (methylome), we developed and synthesized trifunctional capture compounds containing the chemically stable cofactor product S-adenosyl-L-homocysteine (SAH or AdoHcy) as selectivity function. SAH analogues with amino linkers at the N6 or C8 positions were synthesized and attached to scaffolds containing different photocrosslinking groups for covalent protein modification and biotin for affinity isolation. The utility of these SAH capture compounds for selective photoinduced protein isolation is demonstrated for various methyltransferases (MTases) acting on DNA, RNA and proteins as well as with Escherichia coli cell lysate. In addition, they can be used to determine dissociation constants for MTase-cofactor complexes.

Full text of DOI:10.1002/cbic.200900349

DOI: 10.1002/cbic.200900349
Synthesis of S-Adenosyl-l-homocysteine Capture
Compounds for Selective Photoinduced Isolation of
Methyltransferases
Christian Dalhoff,[b] Michael Hꢀben,[a] Thomas Lenz,[b] Peter Poot,[a] Eckhard Nordhoff,[c]
Hubert Kçster,*[b] and Elmar Weinhold*[a]
Understanding the interplay of different cellular proteins and
their substrates is of major interest in the postgenomic era. For
this purpose, selective isolation and identification of proteins
from complex biological samples is necessary and targeted iso-
lation of enzyme families is a challenging task. Over the last
years, methods like activity-based protein profiling (ABPP) and
capture compound mass spectrometry (CCMS) have been de-
veloped to reduce the complexity of the proteome by means
of protein function in contrast to standard approaches, which
utilize differences in physical properties for protein separation.
To isolate and identify the subproteome consisting of S-adeno-
syl-l-methionine (SAM or AdoMet)-dependent methyltransfer-
ases (methylome), we developed and synthesized trifunctional
capture compounds containing the chemically stable cofactor
product S-adenosyl-l-homocysteine (SAH or AdoHcy) as selec-
tivity function. SAH analogues with amino linkers at the N6 or
C8 positions were synthesized and attached to scaffolds con-
taining different photocrosslinking groups for covalent protein
modification and biotin for affinity isolation. The utility of
these SAH capture compounds for selective photoinduced pro-
tein isolation is demonstrated for various methyltransferases
(MTases) acting on DNA, RNA and proteins as well as with
Escherichia coli cell lysate. In addition, they can be used to de-
termine dissociation constants for MTase–cofactor complexes.
Introduction
The proteome, that is, the entire set of proteins expressed by a
genome, cell, tissue or organism at a given time under defined
conditions, is the prevalent bioanalytical challenge in the post-
genomic era. The vast number of proteins usually present in
cells or organisms preclude an entire direct proteomic analysis.
Protein separation by physical properties like molecular
weight, charge, and hydrophobicity can prefractionate the
complex protein mixtures, but thereafter time consuming and
extensive analyses of the single fractions might be necessary
to further separate and identify the protein(s) of interest. Alter-
natively, functional-directed separation of proteome subsets
from whole proteomes can reduce the complexity and allow
for identification of the protein(s) of interest by standard bio-
analytical methods. Affinity chromatography and chip technol-
ogy achieve functional separations by immobilized ligands, but
suffer from surface effects and steric hindrance. Additionally,
the polymeric network has to be swollen in appropriate sol-
vents, which have to be compatible with the solubility proper-
ties of the proteins. Since size and solubility of the targeted
proteins are a priori unknown, many proteins can remain
undetected. Thus, methods in which the critical affinity-based
interaction takes place in solution are desirable.
and the other one for detection (e.g., fluorophores) or isolation
(sorting function).
However, the design of ABPP compounds is not trivial be-
cause the reactivity function has to be incorporated into the
selectivity group, for example, a small molecule substrate,
without affecting its binding properties. Thus, a separation of
selectivity and reactivity function is often desirable.[2]
Capture compound mass spectrometry (CCMS) is based on
capture compounds (CC; Figure 1), which are small molecules
comprised of three different functionalities.[3] They contain:
1) a selectivity function for specific, reversible binding to target-
ed proteins, 2) a reactivity function for photoinduced covalent
attachment to the targeted proteins, and 3) a sorting function
for isolation of the covalently captured proteins on solid sup-
ports. The CCs offer two advantages compared to simple affini-
[a] Dr. M. Hꢀben, P. Poot, Prof. Dr. E. Weinhold
Institut fꢀr Organische Chemie, RWTH Aachen University
Landoltweg 1, 52056 Aachen (Germany)
Fax: (+49)241-80-92528
E-mail: elmar.weinhold@oc.rwth-aachen.de
[b] Dr. C. Dalhoff, T. Lenz, Prof. Dr. H. Kçster
Caprotec Bioanalytics, GmbH
Activity-based protein profiling (ABPP) has been used for tar-
geted isolation of many enzyme classes, including proteases,
kinases, phosphatases, glycosidases, hydrolases, histone deace-
tylases and oxidoreductases.[1] This approach is based on syn-
thetic compounds containing two functional moieties: one
moiety is responsible for binding (selectivity function) and co-
valent modification (reactivity function) of specific proteins,
Volmerstrasse 5, 12489 Berlin (Germany)
Fax: (+49)30-6392-3989
E-mail: hubert.koester@caprotec.com
[c] Dr. E. Nordhoff
Max-Planck-Institut fꢀr Molekulare Genetik
Ihnestrasse 63-73, 14195 Berlin (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.200900349.
256
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 256 – 265

SAH Capture Compounds
Figure 1. General composition of trifunctional capture compounds (CC). By
attaching different selectivity functions to a CC scaffold, which contains
both the reactivity and sorting function, CCs for a wide variety of protein
classes can be obtained.
Scheme 1. Methyltransferases (MTases) transfer the activated methyl group
from SAM (1) to nucleophilic positions (Nu) in various substrates, like DNA,
RNA, proteins and small molecules; this leads to the formation of SAH (2). In
addition, the N6 and C8 position for the attachment to capture compound
scaffolds are indicated.
ty-based separations on solid supports (pull-down experi-
ments). First, the critical selective affinity interaction takes
place in solution and avoids surface-associated problems, and
second, the transformation of weak noncovalent interactions
into a covalent bond allows stringent washing conditions to
remove unspecific (noncovalently bound) proteins.
based SAM analogues for labeling DNA using MTases,[8] X-ray
structures of MTases in complex with these cofactor ana-
logues[9] and analysis of PDB data,[10] the exocyclic N6 position
or C8 position of the adenine heterocycle (Scheme 1) are
much less recognized by MTases and more suitable for linker
introduction and subsequent attachment of CC scaffolds. To
target these positions we implemented new synthetic routes
to adenine-modified SAH derivatives.
In previous studies we used a hydrophobic CC scaffold with
lysine as trifunctional core to attach azidobenzoic acid (reactiv-
ity function for photocrosslinking) and biotin amidohexanoic
acid (sorting function for binding to avidin-coated beads) for
the isolation of carbonic anhydrase.[3] Paying attention to the
fact that most enzymes are located in the cytoplasm, we de-
signed a more hydrophilic CC scaffold using aspartic acid as
the core and a PEG linker for attaching biotin. The synthesis of
the CC scaffold is a straightforward sequence of amide-forming
reactions and the selectivity function can be introduced at the
end of the synthesis. This convergent synthesis gives the op-
portunity to target different protein classes by simply changing
the selectivity function.
Here we report the synthesis of four SAH-based CCs with
two different attachment positions for two scaffolds bearing
different photocrosslinking groups. These SAH-CCs are evaluat-
ed by using purified MTases and used to determine the dissoci-
ation constant KD for MTase–cofactor complexes. In addition,
we demonstrate that they can be used to isolate SAH-binding
enzymes from E. coli cell lysate.
The sulfonium compound S-adenosyl-l-methionine (1, SAM, Results and Discussion
AdoMet) is a ubiquitous cofactor and one of the most widely
Synthesis of N6 and C8 linker-functionalized SAH
used enzyme substrates, second only to adenosine triphos-
phate (ATP).[4] Due to the activation of the adjacent carbon
atoms by the pivotal sulfonium center, SAM (1) is involved not
only in several transfer reactions, including almost all methyl
transfer reactions in biological systems, but also in radical gen-
eration and other transformations.[5] SAM-dependent enzymes
are widely spread in all phyla of life and represent about 2%
of the total proteins in E. coli, making SAM-dependent en-
zymes an interesting target for proteomic research. After the
methyl group transfer reaction catalyzed by methyltransferases
(MTases), the demethylated cofactor product S-adenosyl-l-ho-
mocysteine (2, SAH, AdoHcy) is formed (Scheme 1). The result-
ing thioether function is chemically much more stable than
the activated sulfonium center in SAM, and SAH is a general
competitive product inhibitor of MTases.[6] This combination of
chemical stability and high binding affinity makes SAH, in con-
trast to SAM, a promising selectivity function for capturing
SAM-dependent MTases (the methylome).
derivatives
Chemical synthesis of SAH (2) and analogues is commonly per-
formed with C5’-activated adenosines, like 5’-O-tosyl or 5’-
deoxy-5’-halo derivatives, or by in situ activation of the 5’ posi-
tion and coupling with thiolates.[11] However, this approach
suffers from a competing intramolecular attack of N3 of the
adenine ring onto C5’ leading to irreversible cyclonucleoside
formation.[12] This intramolecular cyclization can be avoided by
reversing the reactivity as demonstrated by coupling of various
alkyl bromides with protected 5’-deoxy-5’-thioadenosine.[13]
This strategy was also used in the present work to incorporate
the full homoserine side chain into linker-functionalized SAH
derivatives.
The SAH derivative bearing an amino linker at N6 was syn-
thesized starting from 2’,3’-O-isopropylidene-protected 6-chlor-
opurine-b-d-riboside (3; Scheme 2A). Substitution with 1,4-di-
aminobutane gave the linker-functionalized adenosine deriva-
tive 4 and treatment with ethyl trifluoroacetate yielded the
amino-protected adenosine derivative 5.[8b] Introduction of
sulfur was performed with thioacetic acid under Mitsunobu
SAH (2) has been previously directly attached to solid sup-
ports via the amino or carboxylic acid group of the amino acid
side chain.[7] However, based on our experience with aziridine-
ChemBioChem 2010, 11, 256 – 265
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
257

H. Kçster, E. Weinhold et al.
Scheme 2. Synthesis of N6 and C8 linker-functionalized SAH derivatives. A) Experimental conditions for 8: a) 1,4-diaminobutane, NEt3, EtOH, 608C, 96%;
b) ethyl trifluoroacetate, NEt3, MeOH, room temperature, 92%; c) PPh3, DEAD, AcSH, THF, À108C then 08C, 94%; d) 1: NH3, MeOH, 08C; 2: N-Boc-g-tosyl-homo-
serine-tert-butylester, KOH, 18-crown-6, THF, À108C, 79% for both steps; e) NH3, H2O, MeOH, room temperature, 90%. B) Experimental conditions for 14: a) N-
(2-propinyl)-2,2,2-trifluoroacetamide, (PPh3)2PdCl2, CuI, NEt3, DMF, room temperature, 75%; b) H2, PtO2, MeOH, room temperature, 92%; c) PPh3, DEAD, AcSH,
THF, À108C then 08C, 87%; d) 1: NH3, MeOH, 08C; 2: N-Boc-g-tosyl-homoserine-tert-butylester, KOH, 18-crown-6, THF, À108C, 69% for both steps; e) NH3,
H2O, MeOH, room temperature, 65%. Details are given in the Supporting Information.
conditions[14] and led to the thioester 6. Selective cleavage of
the thioester was obtained in dry methanol saturated with gas-
eous ammonia whereas ammonolysis in methanol/water led to
additional removal of the TFA group. Since the resulting 5’-
thiol has a strong tendency for disulfide formation under
aerobe conditions the subsequent coupling step was per-
formed without purification and under rigorous exclusion of
oxygen. For introducing the amino acid side chain, commer-
cially available Boc-Asp-tBu was converted into the desired
homoserine derivative according to a literature procedure.[15]
During the reduction of a mixed anhydride intermediate in
methanol, cooling to À108C proved to be crucial, because oth-
erwise the corresponding methylether was obtained. The re-
sulting alcohol was activated as p-toluenesulfonic ester and
treated with freshly cleaved thioester 6 to yield the protected
SAH derivative 7 in good yield. Cleavage of the TFA group led
to the primary amine 8, which was coupled to the different CC
scaffolds (vide infra).
tion. Subsequent Mitsunobu reaction gave the desired thioest-
er 12, which was, after liberating the 5’-thiol, coupled with the
activated homoserine derivative. Finally, the resulting SAH
derivative 13 was treated with ammonia in methanol/water to
give the primary amine 14 for attachment to the CC scaffolds
(vide infra).
Hydrophilic capture compound scaffolds
Synthesis of the new CC scaffolds (Scheme 3) is based on
amide bond forming reactions, starting from commercially
available trifunctional core Z-Asp(tBu)OSu (15) and linker N-
Boc-4,7,10-trioxa-1,13-tridecanediamine (16). After coupling
amino acid 15 with linker 16 the benzyloxycarbonyl group (Z
group) of the product 17 was removed by catalytic hydrogena-
tion and the resulting amine 18 was used to introduce two
photocrosslinking groups (reactivity functions). 4-Azidobenzoic
acid was incorporated by using the corresponding N-hydroxy-
succinimidylester 19, whereas 4-(3-(trifluoromethyl)-3H-diazir-
ine-3-yl)benzoic acid (20) was introduced by carbodiimide/
HOBt chemistry and yielded the coupling products 21a and
21b, respectively. Removal of the tert-butyloxycarbonyl group
(Boc group) and cleavage of the tert-butylester under acidic
conditions led to the amines 22a and 22b, which were cou-
pled with N-hydroxysuccinimidyl biotin (sorting function) to
yield the hydrophilic CC scaffolds 23a and 23b, respectively.
Synthesis of the SAH derivative with an amino linker at C8
started from 8-bromo-2’,3’-O-isopropylideneadenosine (9). Ini-
tially, we followed the synthetic route for the N6-modified SAH
derivative 8. Substitution with 1,4-diaminobutane and protec-
tion of the primary amine gave the corresponding adenosine
derivative, but attempts to introduce thioacetic acid at the
5’ position via the Mitsunobu reaction failed. Under these con-
dition an intramolecular cyclization occurred and led to an
N8,5’-anhydroadenosine derivative.[16] A successful approach to
C8-modified SAH was obtained by palladium-catalyzed Sono-
gashira cross-coupling[17] of 9 with TFA-protected propargyl
amine, which avoids a secondary amine at the 8 position
(Scheme 2B). The resulting alkyne 10 was reduced by catalytic
hydrogenation to form 11 with a flexible linker at the 8 posi-
Capture compounds with SAH as selectivity function
The primary amino groups of the linker-functionalized SAH
derivatives 8 and 14 were coupled with the carboxylic acids of
CC scaffolds 23a and 23b by using carbodiimide/HOBt
258
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 256 – 265

SAH Capture Compounds
Scheme 4. Synthesis of SAH-based trifunctional capture compounds with
different attachment positions and photocrosslinking groups. A) Experi-
mental conditions for N6-linked CCs 24a and 24b: a) diispropylcarbodi-
imide, HOBt, DMAP, dimethylacetamide, 508C for 1 h then room temperature
for 12 h; b) HCl in dioxan, DCM, 08C for 1 h, then add H2O, 08C for 1 h;
24a: 30% for both steps, 24b: 63% for both steps. B) Experimental condi-
tions for C8-linked CCs 25a and 25b: a) diispropylcarbodiimide, HOBt,
DMAP, dimethylacetamide, 508C for 1 h then room temperature for 12 h;
b) HCl in dioxan, DCM, 08C for 90 min, then add H2O, 08C for 90 min (25a)
or 60 min (25b); 25a: 28% for both steps; 25b: 43% for both steps. Details
are given in the Supporting Information.
Scheme 3. Synthesis of capture compound scaffolds. Experimental condi-
tions: a) NEt3, DCM, 08C then room temperature, 20 h, 95%; b) H2, Pd/C,
MeOH, room temperature, 48 h, 90%; c) NEt3, DCM, room temperature, 24 h,
94%; d) diisopropylcarbodiimide, HOBt, DMAP, NEt3, THF, 08C then room
temperature, 44 h, 95%; e) 21a, TFA, MeOH, room temperature, 48 h; 21b,
HCl/dioxane, DCM, 08C, 2 h; f) 22a, N-hydroxysuccinimidyl biotin, NEt3, THF,
room temperature, overnight, 39% for the last two steps; 22b, N-hydroxy-
succinimidyl biotin, NEt3, DMSO, 20% for the last two steps. Details are
given in the Supporting Information.
ity function was activated by irradiation with UV light and co-
valently crosslinked enzymes were isolated by using the biotin
sorting function for binding to streptavidin-coated magnetic
beads. Captured MTases were then analyzed and visualized by
SDS-PAGE and silver staining.
chemistry (Scheme 4). Afterwards all remaining protection
groups of the SAH fragments (2’,3’-O-isopropylidene, Boc and
tert-butyl group) were removed in one step under acidic condi-
tions. This yielded a set of four biotinylated CCs with two dif-
ferent attachment positions to SAH (N6: 24a and 24b, C8:
25a and 25b) and two different photocrosslinking groups (4-
azido-benzoic acid: a series, 4-(3-(trifluoromethyl)-3H-diazirine-
3-yl)benzoic acid: b series) which were used for comparative
capture studies with MTases.
First, we tested SAH-CC 24a carrying the aromatic azide
scaffold at the N6 position of SAH with the DNA adenine N6
MTase from Thermus aquaticus (M.TaqI). Analysis by SDS-PAGE
(Figure 2) showed that M.TaqI can be isolated by the CC tech-
nology (lane 3) and capture yields ranged between 5–10%
under saturating conditions. Importantly, several control ex-
periments established that capture is specific and requires co-
valent bond formation. No capture is observed in the presence
of the competitor SAH (2; lane 4); this indicates that SAH-CC
24a binds to the cofactor binding site of M.TaqI and prebind-
ing to the protein is required for photocrosslinking. In addition,
the CC scaffold 23a lacking the selectivity function does not
lead to the isolation of M.TaqI (lane 5), which demonstrates
Capturing various MTases
The functionality and specificity of the new SAH-CCs 24a, 24b,
25a and 25b are demonstrated and compared by isolation of
MTases acting on different substrates. Purified enzymes were
incubated with an excess of the respective SAH-CC, the reactiv-
ChemBioChem 2010, 11, 256 – 265
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
259

H. Kçster, E. Weinhold et al.
From a practical view this small improvement in capture yield
might not balance the much higher costs (or much longer syn-
thesis) for the 4-(3-(trifluoromethyl)-3H-diazirine-3-yl)benzoic
acid (20) compared to 4-azidobenzoic acid. In addition, the N6-
linked SAH-CCs 24a and 24b (lanes 2 and 5) gave somewhat
higher capture yields than the C8-linked SAH-CCs 25a and
25b (lanes 3 and 6). Control experiments with the CC scaffolds
23a and 23b lacking the SAH selectivity functions did not lead
to the isolation of M.TaqI in appreciable amounts. This again
demonstrates that the selectivity function is required to direct
the probes to the protein and enables photocrosslinking
within the protein–probe complex.
Having successfully tested the four CCs with the adenine-
specific DNA MTase M.TaqI, we were interested in testing fur-
ther MTases acting on various other substrates. Capture experi-
ments were performed with the DNA cytosine-C5 MTase
M.HhaI, the RNA cytosine-C5 MTase Trm4 and the protein his-
tone H3 lysine 9 MTase Clr4. All three MTases were isolated by
using SAH-CC 24a and capture yields ranged between 2–10%
(Figure 4). Two control experiments were performed to demon-
Figure 2. Capturing the DNA MTase M.TaqI (1 mm) with SAH-CC 24a (10 mm).
Samples were analyzed by SDS-PAGE and silver staining. Lane 1: M.TaqI
standard; lane 2: molecular weight (MW) marker with indicated molecular
weights in kDa; lane 3: capture experiment; lane 4: control in the presence
of SAH (2; 6.8 mm) as competitor; lane 5: control with the CC scaffold 23a
(10 mm); lane 6: control without UV irradiation; lane 7: control in the absence
of CC.
that the selectivity function is essential. No capturing occurs
without UV irradiation (lane 6) showing that covalent bond for-
mation is required for protein isolation. Finally, no capturing
occurs in the absence of the SAH-CC (lane 7).
We then compared all four SAH-CCs 24a, 24b, 25a and 25b
containing different photocrosslinking functions (a=aromatic
azide, b=aromatic diazirine) and different attachment posi-
tions (24=adenine-N6, 25=adenine-C8) in their ability to cap-
ture M.TaqI (Figure 3). Slightly more intense bands were ob-
tained with the aromatic diazirines 24b and 25b (lanes 5 and
6) than with the aromatic azides 24a and 25a (lanes 2 and 3).
Figure 4. Capturing the DNA MTase M.HhaI (1 mm), the RNA MTase Trm4
(0.5 mm) and the protein MTase Clr4 (1 mm) with SAH-CC 24a (5 mm for
M.HhaI, 20 mm for Trm4 and 10 mm for Clr4). Samples were analyzed by SDS-
PAGE and silver staining. Lanes 1, 5 and 9: molecular weight (MW) marker
ranging from 15 to 200 kDa; lanes 2, 6, and 10: capture experiments with
the three MTases (bands corresponding to the isolated MTases are circled in
gray); lanes 3, 7 and 11: control experiments in the presence of SAH (2;
7.3 mm for M.HhaI, 4.0 mm for Trm4 and 6.8 mm Clr4); lanes 4, 8 and 12:
control experiments with the CC scaffold 23a (5 mm for M.HhaI, 20 mm for
Trm4 and 10 mm for Clr4).
strate that initial noncovalent interactions between the SAH se-
lectivity function and proteins are required for photocrosslink-
ing. Competition by an excess of SAH (2) completely (lanes 3
and 7) or strongly (lane 11) blocks photocrosslinking and no
(lanes 4 and 8) or almost no (lane 12) proteins are detected
when using the CC scaffold 23a, which lacks the selectivity
function.
In addition to SDS-PAGE, we analyzed M.TaqI captured with
SAH-CC 24a by matrix-assisted laser desorption/ionization
mass spectrometry (MALDI-MS). A mixture of M.TaqI and cross-
linked M.TaqI yielded two strong signals (Figure 5). Photocross-
linking by aromatic azides is initiated by loss of nitrogen (N2)
and the resulting singlet nitrene can rapidly decay into various
activated species for crosslinking.[18] The experimental mass dif-
Figure 3. Capturing the DNA MTase M.TaqI (1 mm) with all four SAH-CCs 24a,
25a, 24b, or 25b (2 mm each; nonsaturating conditions). Samples were ana-
lyzed by SDS-PAGE and silver staining. Lane 1: molecular weight (MW) marker
with indicated molecular weights in kDa; lanes 2–4: capture experiments
with SAH-CCs 24a, 25a or CC scaffold 23a carrying an aromatic azide for
photocrosslinking; lanes 5–7: capture experiments with SAH-CCs 24b, 25b
or CC scaffold 23b carrying an aromatic diazirine for photocrosslinking;
lane 8: M.TaqI standard (100 ng).
260
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 256 – 265

SAH Capture Compounds
centration of SAH-CC 24b (5 mm) and increasing initial M.TaqI
concentrations (up to 500 nm) showed a linear increase in inte-
grated band intensities (the coefficient of determination of a
linear fit to the data points was R2 =0.998) with respect to the
initial M.TaqI concentration (Figure S1 in the Supporting Infor-
mation). This result verifies that the yield of isolated M.TaqI–
SAH-CC crosslink and hence the crosslinking yield is constant
for the applied concentration range. Additionally, from the
same experiment, the detection limit for M.TaqI in the capture
assay was determined to be about 5 pmol (50 nm in 100 mL).
The detection limit for assayed M.TaqI by direct MS/MS analysis
of peptides obtained after capturing and tryptic fragmentation
was approximately 100 fmol (1 nm in 100 mL; data not shown),
which demonstrates the higher sensitivity of CCMS compared
to capturing and simple SDS-PAGE analysis.
In a second set of experiments (Figure 6B) we determined
the dissociation constant KD for M.TaqI and SAH (2) by “com-
petitive capture titration”. Constant concentrations of M.TaqI
(250 nm) and SAH-CC 24b (1 mm) were treated with increasing
concentrations of the natural cofactor product SAH (2; 0–
125 mm), which led to an almost complete inhibition of cross-
linking. This result is best explained by binding of SAH to the
cofactor binding site of M.TaqI and hindrance of complex for-
mation with SAH-CC 24b and, thus, crosslinking. Fitting Equa-
tion (16) given in the Supporting Information, which describes
analytically the competitive binding of two ligands to the
same binding site, to the obtained data points yielded a KD of
2.3 mm for the complex of M.TaqI and SAH (2). The obtained KD
value is in excellent agreement with the published KD of (2.4Æ
0.5) mm, which was determined in a fluorescence assay.[20] Al-
though the accuracy of quantification by SDS-PAGE and silver
staining might be lower than by spectroscopic or radioactive
methods, capture titrations only require a UV lamp, a strong
magnet and standard biochemical laboratory equipment,
which makes them a low-cost alternative to other binding
assays.
Figure 5. MALDI-MS spectrum with magnification in the 46–51 kDa range of
a mixture of the DNA MTase M.TaqI and M.TaqI crosslinked with SAH-CC
24a. The experimental (found) mass difference of the two main signals is in-
dicated and matches the calculated (calcd) mass difference. Satellite signals
mainly originate from MALDI-specific adducts with salts and matrix mole-
cules.
ference between unmodified and modified M.TaqI was 1114,
which is in very good agreement with the calculated mass in-
crease resulting from covalent attachment of SAH-CC 24a (m/z
1115.5 for [MÀN2]). In addition, MALDI-MS analysis of peptides
obtained after tryptic fragmentation identified M.TaqI with a
sequence coverage of 65%.
Determination of dissociation constants
The dissociation constant KD for M.TaqI and SAH-CC 24b was
determined by “capture titration” by using a constant concen-
tration of M.TaqI (250 nm) and increasing concentrations of
SAH-CC 24b (0–5 mm) as shown in Figure 6A. This approach is
based on the very fast time scale of photolytic reactions (nano-
seconds to milliseconds) and under ideal conditions the photo-
lytic crosslinking reaction can take a snapshot of an enzyme–
ligand complex in equilibrium. In the experiment, however, the
time interval for photoactivation and subsequent crosslinking
is not infinitely small. Ligand exchange can take place during
the irradiation process, which could lead to deviations from
the true solution equilibrium. The yield of the photolytic cross-
linking reaction within the enzyme–ligand complex is assumed
not to be quantitative but constant, so that the amount of iso-
lated enzyme–ligand crosslink is proportional to the concentra-
tion of the enzyme–ligand complex in solution before irradia-
tion. For measuring the amount of covalently crosslinked
enzyme–ligand complex, the M.TaqI–SAH-CC crosslink was iso-
lated by using streptavidin-coated magnetic beads, subjected
to SDS-PAGE and subsequently silver stained. The bands were
quantified by integration by using the software ImageJ,[19] and
fitting of the quadratic binding equation for one binding site
(Equation (9) given in the Supporting Information) to the data
resulted in a KD of 1.4 mm for the complex of M.TaqI and SAH-
CC 24b. A control experiment with a constant saturating con-
Capturing SAH-binding proteins from E. coli lysate
The utility of our SAH-CCs to isolate proteins from complex
mixtures was tested by using a cell lysate from E. coli. Analysis
by SDS-PAGE (Figure 7) showed clear functional enrichment of
proteins compared to the control with added SAH (2). Most
importantly, we were able to identify proteins from the gel by
treating gel slices with trypsin, analyzing peptides by MALDI-
MS/MS and interrogating protein sequence databases. Besides
streptavidin from the magnetic beads and naturally biotinylat-
ed biotin carboxyl carrier protein of acetyl-CoA carboxylase
(BCCP), we found two SAH-binding proteins from E. coli: trans-
aconitate 2-MTase (tam)[21] and 5’-methylthioadenosine/S-ade-
nosylhomocysteine (MTA/SAH) nucleosidase (MtnN).[22] SAH
binds with micromolar affinity to trans-aconitate 2-MTase (72%
inhibition at 0.38 mm SAH)[21] and to MTA/SAH nucleosidase
(KM for SAH is 4.3 mm).[22] This demonstrates that the SAH-CC
allows isolation of MTases and other SAH-binding proteins
from complex protein mixtures like cell lysates in a form that
is suitable for identification by MALDI-MS/MS. Since direct
ChemBioChem 2010, 11, 256 – 265
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
261

H. Kçster, E. Weinhold et al.
MTases for follow-up analysis by
SDS-PAGE and mass spectrome-
try (CCMS). Importantly, the SAH-
CCs also allow for isolation and
identification of SAH-binding
proteins from complex protein
mixtures like E. coli lysate. This
paves the way for isolating func-
tionally defined subproteomes,
like the methylome. In addition,
the linear relationship between
protein
amount of captured protein, as
demonstrated with M.TaqI,
concentration
and
should be useful for quantitative
comparison of expression pat-
terns in different cell states
(functional protein expression
profiling).
Experimental Section
General: All solvents and chemi-
cals were of analytical grade unless
otherwise noted. HPLC analysis
and purification was performed
with a Waters Breeze system and
NMR spectra were recorded by
using a Mercury 300 spectrometer
(Varian) or an Ultrashield 400 spec-
trometer (Bruker). Mass spectro-
metric analysis was performed on
a Scout MTP Ultraflex II MALDI-
TOF-MS/MS equipped with a solid-
state SmartBeam laser (Bruker Dal-
tonics, Bremen, Germany).
Figure 6. Determination of dissociation constants (KD) for the complexes of M.TaqI with SAH-CC: A) 24b, or B) SAH
(2) by “capture titration”. After irradiation, crosslinked M.TaqI was isolated by using streptavidin-coated magnetic
beads, and subjected to SDS-PAGE with silver staining. Crosslinked M.TaqI corresponding to the bands in the
middle of the two gels was quantified by using the software ImageJ (integrated band intensities are shown
above the protein bands). Plots of the integrated M.TaqI band intensities versus the concentration of SAH-CC 24b
*
or SAH ( ) and least-square fits (black lines) of the corresponding equation to the data points are shown below
the gels. A) The KD for the complex of SAH-CC 24b with M.TaqI was determined by varying the initial SAH-CC 24b
concentration (detailed above the gel) at a constant initial M.TaqI concentration of 250 nm. Fitting of Equation (9)
(see the Supporting Information) to the data points gave a KD of 1.4 mm. B) The KD for the complex of SAH with
M.TaqI was determined by varying the initial SAH concentration (detailed above the gel) at constant initial M.TaqI
and SAH-CC 24b concentrations of 250 nm and 1 mm, respectively. Fitting of Equation (16) (see the Supporting In-
formation) to the data points gave a KD of 2.3 mm.
The MTases M.TaqI,[23] M.HhaI[24]
and Clr4[25] were over-expressed
and purified as described before.
Trm4 was kindly provided by Priv.-Doz. Dr. Mark Helm (Institut fꢀr
Pharmazie und Molekulare Biotechnologie, Universitꢂt Heidelberg,
Germany). Silver staining of gels was performed with the ProteoSil-
ver Plus silver stain kit (Sigma). The peptide calibration standards
angiotensin I and ACTH 18-39 (corticotropin-like intermediate lobe
peptide) were purchased from Bachem and porcine trypsin was
from Promega. All synthesis procedures and equations for deter-
mining dissociation constants KD are given in detail in the Support-
ing Information.
protein identification from the magnetic beads is much more
sensitive than SDS-PAGE (see above results with M.TaqI), we
expect to find many more MTases by CCMS.
Conclusions
We have established synthesis routes to previously unknown
N6- and C8-modified SAH derivatives suitable for attachment
to CC scaffolds. However, the protected intermediates 8 and
14 containing primary amines should also be useful for cou-
pling to other chemical entities or solid supports leading to
various applications. Our design of SAH-CC proved to be viable
and all four SAH-CCs 24a, 24b, 25a and 25b were successfully
used for the isolation of the DNA MTase M.TaqI. In fact, the
binding affinity of SAH-CC 24b to M.TaqI is even slightly in-
creased compared to the binding affinity of the natural cofac-
tor product SAH (2) as directly determined by capture titration.
Functional studies with other purified MTases acting on differ-
ent substrates confirmed the ability of our SAH-CCs to isolate
Preparation of E. coli lysate: Four Erlenmeyer flasks (5 L) contain-
ing LB medium (2.5 L each) were inoculated with an E. coli DH5a
culture (50 mL each) grown to exponential phase (OD600 of 0.8)
and incubated at 378C and 166 rpm until an OD600 of 0.8 was
reached. The cells were harvested by centrifugation (4200 g, 48C,
30 min) and the cell pellet (25 g) was stored at À208C for 14 h.
The cells were resuspended in lysis buffer (100 mL, 6.7 mm MES,
6.7 mm HEPES, 6.7 mm NaOAc, 200 mm NaCl, 1 mm EDTA, 10 mm
b-mercaptoethanol, 0.2 mm PMSF, 10% (v/v) glycerol, pH 7.5), were
cooled on ice and treated in aliquots with ultrasound (Brason Soni-
fier 250, big tip, duty cycle 10%, output control 10). Cell debris was
removed by centrifugation (35000 g, 48C, 60 min) and the com-
262
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 256 – 265

SAH Capture Compounds
transferred into a new tube, incubated on the orbital shaker at 48C
for 5 min and magneto-precipitated. This washing step was repeat-
ed twice. After removal of the supernatants the beads were
washed twice with basic wash buffer (180 mL, 50 mm NH4OAc,
0.05% octyl-b-d-glucopyranoside, pH 9.0). Then the beads were
washed again two times with wash buffer containing SAH
(100 mm; 180 mL for each sample) and two times with water
(180 mL, Milli-Q grade).
SDS-PAGE analysis: For SDS-PAGE (15% gel), the supernatants
were removed, the beads were suspended in SDS sample buffer
(14 mL, 50 mm Tris-HCl, 2.5% SDS, 10% glycerol, 320 mm b-mercap-
toethanol, 0.05% bromphenole blue, pH 6.8) and heated to 958C
for 10 min. After electrophoresis protein bands were visualized by
silver staining.
In gel trypsinolysis: Gel slices were treated twice with aqueous
acetonitrile (400 mL, 50%) for 30 min while shaking, once with ace-
tonitrile (400 mL) for 5 min and once with reducing buffer (50 mL,
10 mm DTT, 50 mm NH4HCO3) at 378C for 30 min. After an addi-
tional treatment with acetonitrile (400 mL) for 5 min, iodoacetamide
solution (50 mL, 20 mm in 50 mm NH4HCO3) was added and the gel
samples were incubated at room temperature in the dark for
30 min. After solvent removal a solution of 2,2’-thiodiethanol
(50 mL, 20 mm) was added and incubation was continued at room
temperature for 10 min. The liquid was removed and the gel sam-
ples were treated twice with aqueous acetonitrile (400 mL, 50%) for
5 min and once with acetonitrile (400 mL) for 10 min. After remov-
ing the liquid the gel samples were dried in a vacuum centrifuge
for 30 min. A solution of trypsin (12 mL, 2.5 ngmLÀ1 in 50 mm
NH4HCO3, pH 7.8) was added to the gel pieces, and after 10 min
incubation on ice NH4HCO3 solution (50 mm, pH 7.8), sufficient to
just cover the gel pieces, was added and followed by incubation,
overnight, at 378C. For peptide extraction, an aqueous solution
(20 mL) of trifloroacetic acid (TFA; 0.1%) and octyl-b-d-glucopyrano-
side (0.5 mm) was added and the gel samples were placed on a
shaker for 20 min. Of each sample, two aliquots of the supernatant
(0.5 mL) were aspirated and directly prepared for MALDI.
Figure 7. Capturing SAH-binding proteins with SAH-CC 24a (5 mm) from
E. coli lysate (3.5 mgmLÀ1). Samples were analyzed by SDS-PAGE and silver
staining. In addition, gel slices containing protein bands were excised, treat-
ed with trypsin, and resulting peptides were analyzed by MALDI-MS/MS.
Lane 1: molecular weight (MW) marker with indicated molecular weights in
kDa; lane 2: capture experiment; lane 3: control in the presence of SAH (2)
as competitor. A clear enrichment of proteins as compared to the control is
observed in the molecular weight range between 20 and 50 kDa (dashed
box) and two SAH-binding proteins (*) were identified.
bined supernatants were concentrated (to 17.5 mL) by ultrafiltra-
tion (Jumbosep 10 K, Pall, 500 g, 48C, overnight). Small molecules
were removed by gel filtration (NAP5 columns, GE Healthcare;
equilibrated with lysis buffer). Aliquots (500 mL) were applied to
the columns, lysis buffer (200 mL) was added and proteins were
eluted with lysis buffer (650 mL). The combined fractions (23 mL)
were supplemented with lysis buffer (37 mL) and concentrated (to
11 mL) by ultrafiltration (Jumbosep 10 K, Pall, 350 g, 48C, over-
night). Glycerol (9 mL) was added and the cell lysate (total protein
concentration 35 mgmLÀ1) stored at À208C.
General procedure for capture experiments: The capture proce-
dure is described for M.TaqI and SAH-CC 24a. For the other MTases
or E. coli lysate and SAH-CCs, capturing was performed in an analo-
gous manner and details regarding concentrations are given in the
respective figure legend. M.TaqI (1 mm) was incubated with SAH-CC
24a (10 mm) in capture buffer (100 mL, 20 mm HEPES, 10 mm
Mg(OAc)2, 50 mm KOAc, 10% (v/v) glycerol, pH 7.9) in thin-walled
PCR tubes (0.5 mL, Eppendorf) at 48C for 5 min. The tubes were
placed into a reflector bowl (8.2 cm diameter) filled with ice water
and the samples were irradiated (ten times) with a studio flash
lamp (Alienbees B1600 640 WS without UV filter, 2 cm distance be-
tween flashtube and samples). Controls in the presence of CC scaf-
fold or additionally containing SAH (2) were treated in the same
way. After exposure to UV light, all samples were supplemented
with SAH (100 mm final concentration) and centrifuged (16000 g,
48C, 10 min). The samples were applied to ZebaTM gel filtration col-
umns (0.5 mL, Pierce) previously equilibrated with wash buffer
(50 mm Tris-HCl, 1m NaCl, 1 mm EDTA, 0.05% octyl-b-d-glucopyra-
noside, pH 7.5) containing SAH (100 mm) and proteins were directly
eluted to streptavidin-coated magnetic beads (200 mg, MyOneꢃ
Streptavidin C1 Dynabeadsꢄ, Invitrogen, prepared according to the
instruction of the manufacturer) in 10 mL wash buffer containing
SAH (100 mm). Then the samples were incubated on an orbital
shaker (Thermomixer 5433, Eppendorf) at 48C for 40 min. Beads
were collected with a strong neodymium magnet for 1 min and
the supernatants were removed. The beads were suspended in
wash buffer containing SAH (100 mm; 180 mL for each sample),
MALDI mass analysis: Thin microcrystalline layers of a-cyano-4-hy-
droxycinnamic acid were prepared on AnchorChips 384/600 as de-
scribed before.[26] Each aliquot was deposited on a separate matrix-
coated sample anchor and left to dry at ambient conditions. The
dried samples were washed with TFA solution (3 mL, 0.1%) and an-
alyzed by MALDI-MS/MS. Positively charged ions in the m/z range
700–4000 Da were detected and 4000 single-shot spectra were ac-
cumulated for each analysis. Optimal values for fixed laser attenua-
tion were determined prior to analysis by evaluation of a few frac-
tions. Spectrum processing was performed automatically with the
software FlexAnalysis 3.0. Automatic detection of peptide monoiso-
topic signals was performed by using the algorithm SNAP with a
signal-to-noise threshold of six. Internal mass correction was per-
formed by using the signals of two reference peptides (angioten-
sin I, [M+H]+ 1296.6853 and ACTH 18–39, [M+H]+ 2465.1989) in-
cluded in the MALDI matrix solution.
Protein identification was performed by using the Mascot soft-
ware 2.2 (Matrixscience, London, UK) searching the UniProt/Swiss-
Prot and UniProt/Trembl sequence databases. The following set-
tings were used for the searches. Mass error tolerance for peptide
masses: 15 ppm; mass error tolerance for the fragment ions:
0.35 Da; fixed modification: carbamidomethylation; variable modi-
fication: methionine oxidation; number of missed cleavage sites: 1;
type of instrument: MALDI-TOF-PSD. Identifications based on pep-
tide mass fingerprinting were considered very likely to be correct if
ChemBioChem 2010, 11, 256 – 265
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
263

H. Kçster, E. Weinhold et al.
the score was greater than 80. For identification by MS/MS a differ-
ent score is calculated and in this case the threshold for identifica-
tion was set to 50.
Keywords: CCMS · photoaffinity labeling · proteomics · SAH ·
transferases
Determination of dissociation constants by capture titration:
M.TaqI (250 nm) was incubated with varying concentrations of
SAH-CC 24b (0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.3, 1.6, 2.0, 3.0, and 5.0 mm)
in a slightly modified capture buffer (100 mL; 20 mm HEPES, 10 mm
Mg(OAc)2, 50 mm KOAc, 10% (v/v) glycerol, 0.02% Triton X-100,
pH 7.9) placed in thin-walled tubes (0.2 mL) of a 12-tube strip
(Thermo-Strip, Thermo Scientific) at 28C in the sample holder of a
CaproBoxTM (Caprotec Bioanalytics GmbH) for 10 min. The sample
tubes (open lids) were irradiated with UV light (310 nm,
10 mWcmÀ1) within the CaproBoxTM at 28C for 10 min. Each
sample was supplemented with SAH (2; 20 mL, 10 mm), 5ꢅwash
buffer (25 mL, 250 mm Tris-HCl, 5m NaCl, 5 mm EDTA, 0.25% octyl-
b-d-glucopyranoside, pH 7.5) and streptavidin-coated magnetic
beads (50 mL, 10 mgmLÀ1) by carefully inverting the tube with
closed lid after each addition. The beads were kept in suspension
by rotating the samples on a rotator wheel at 48C for 1 h and then
collected by using a caproMagTM magnetic particle collector (Cap-
rotec Bioanalytics, GmbH) containing a strong N52 neodymium
magnet. The supernatants were discarded and the beads were
washed three times with wash buffer (200 mL), three times with
basic wash buffer (200 mL) and once with water (200 mL) by repeat-
ed resuspension, magneto-precipitation and removal of the super-
natants. Finally, the collected beads were resuspended in SDS
sample buffer (20 mL), heated to 958C for 10 min and analyzed by
SDS-PAGE (OLSꢄ ProPage 4–20% Tris-glycine precast gel and
25 mm Tris base, 200 mm glycine, 0.1% SDS, pH 8.3 running buffer)
with subsequent silver staining. To determine the dissociation con-
stant KD of the M.TaqI–SAH complex, a mixture of M.TaqI (250 nm)
and SAH-CC 24b (1 mm) was incubated with varying concentra-
tions of SAH (2; 0, 0.12, 0.24, 0.49, 0.98, 2.0, 3.9, 7.8, 15.6, 31.3,
62.5, and 125 mm) and the samples were processed as described
above.
[1] a) M. J. Evans, B. F. Cravatt, Chem. Rev. 2006, 106, 3279–3301; b) K. T.
Barglow, B. F. Cravatt, Nat. Methods 2007, 4, 822–827; c) M. Fonovic, M.
Bogyo, Expert Rev. Proteomics 2008, 5, 721–730; d) M. Uttamchandani,
J. Li, H. Sun, S. Q. Yao, ChemBioChem 2008, 9, 667–675.
[2] a) A. Saghatelian, N. Jessani, A. Joseph, M. Humphrey, B. F. Cravatt, Proc.
Natl. Acad. Sci. USA 2004, 101, 10000–10005; b) E. W. Chan, S. Chatto-
padhaya, R. C. Panicker, X. Huang, S. Q. Yao, J. Am. Chem. Soc. 2004,
126, 14435–14446; c) S. Chattopadhaya, E. W. S. Chan, S. Q. Yao, Tetra-
hedron Lett. 2005, 46, 4053–4056; d) W.-W. Qiu, J. Xu, D.-Z. Liu, J.-Y. Li,
Y. Ye, X.-Z. Zhu, J. Li, F.-J. Nan, Bioorg. Med. Chem. Lett. 2006, 16, 3306–
3309; e) M. Wiegand, T. K. Lindhorst, Eur. J. Org. Chem. 2006, 4841–
4851; f) C. M. Salisbury, B. F. Cravatt, Proc. Natl. Acad. Sci. USA 2007, 104,
1171–1176; g) W.-W. Qiu, J. Xu, J.-Y. Li, J. Li, F.-J. Nan, ChemBioChem
2007, 8, 1351–1358; h) X. Li, J.-H. Cao, Y. Li, P. Rondard, Y. Zhang, P. Yi,
J.-F. Liu, F.-J. Nan, J. Med. Chem. 2008, 51, 3057–3060; i) S. Tsukiji, M.
Miyagawa, Y. Takaoka, T. Tamura, I. Hamachi, Nat. Chem. Biol. 2009, 5,
341–343.
[3] H. Kçster, D. P. Little, P. Luan, R. Mꢀller, S. M. Siddiqi, S. Marappan, P. Yip,
Assay Drug Dev. Technol. 2007, 5, 381–390.
[4] a) M. Fontecave, M. Atta, E. Mulliez, Trends Biochem. Sci. 2004, 29, 243–
249; b) G. L. Cantoni, Annu. Rev. Biochem. 1975, 44, 435–451.
[5] C. Dalhoff, E. Weinhold in Modified Nucleosides in Biochemistry Biotech-
nology and Medicine (Ed.: P. Herdewijn), Wiley-VCH, Weinheim, 2008,
pp. 223–242.
[6] a) T. Deguchi, J. Barchas, J. Biol. Chem. 1971, 246, 3175–3181; b) R. T.
Borchardt in The Biochemistry of S-Adenosylmethionine (Eds.: F. Salva-
tore, E. Borek, V. Zappia, H. G. Williams-Ashman, F. Schlenk), Columbia
University Press, New York, 1977, pp. 151–171; c) P. K. Chiang, Pharma-
col. Ther. 1998, 77, 115–134.
[7] a) S. Kim, S. Nochumson, W. Chin, W. K. Paik, Anal. Biochem. 1978, 84,
415–422; b) S. K. Sharma, S. A. Brown, J. Chromatogr. 1978, 157, 427–
431.
[8] a) G. Pljevaljcic, M. Pignot, E. Weinhold, J. Am. Chem. Soc. 2003, 125,
3486–3492; b) F. H.-G. Schmidt, M. Hꢀben, B. Gider, F. Renault, M.-P. Teu-
lade-Fichou, E. Weinhold, Bioorg. Med. Chem. 2008, 16, 40–48; c) S. Wil-
kinson, M. Diechtierow, A. E. Estabrook, F. Schmidt, M. Hꢀben, E. Wein-
hold, N. O. Reich, Bioconjugate Chem. 2008, 19, 470–475; d) G. Braun,
M. Diechtierow, S. Wilkinson, F. Schmidt, M. Hꢀben, E. Weinhold, N. O.
Reich, Bioconjugate Chem. 2008, 19, 476–479.
Intensities of protein bands were integrated from pictures of silver-
stained SDS-PAGE gels by using the software ImageJ[19] and plotted
against the initial concentrations of SAH-CC 24b (Figure 6A) or
SAH (Figure 6B). The dissociation constant KD1 of the complex
between M.TaqI and SAH-CC 24b was determined by varying the
parameters C and KD1 in a least-square fit of Equation (9) (see the
Supporting Information) to the data points by using the solver
module of MS Excel. Analogously, the dissociation constant KD2 of
the complex between M.TaqI and SAH was determined by a least-
square fit of Equation (16) (see the Supporting Information) to the
corresponding data points by using the solver module of MS Excel.
The KD1 from above was used as a constant and the parameters C
and KD2 were varied.
[9] G. Pljevaljcic, F. Schmidt, A. J. Scheidig, R. Lurz, E. Weinhold, ChemBio-
Chem 2007, 8, 1516–1519.
[10] G. Pljevaljcic, F. Schmidt, E. Weinhold, ChemBioChem 2004, 5, 265–269.
[11] a) J. Baddiley, G. A. Jamieson, J. Chem. Soc. 1955, 1085–1089; b) R. T.
Borchardt, J. A. Huber, Y. S. Wu, J. Org. Chem. 1976, 41, 565–567; c) K.
Ramalingam, R. W. Woodard, J. Org. Chem. 1984, 49, 1291–1293; d) P.
Serafinowski, Synthesis 1985, 926–928.
[12] V. M. Clark, A. R. Todd, J. Zussman, J. Chem. Soc. 1951, 2952–2958.
[13] M. Pignot, G. Pljevaljcic, E. Weinhold, Eur. J. Org. Chem. 2000, 549–555.
[14] R. P. Volante, Tetrahedron Lett. 1981, 22, 3119–3122.
[15] G. Tang, J. W. Perich, R. B. Johns, Aust. J. Chem. 1992, 45, 1225–1240.
[16] T. Sasaki, K. Minamoto, Y. Fujiki, Chem. Lett. 1983, 12, 1017–1020.
[17] a) L. A. Agrofoglio, I. Gillaizeau, Y. Saito, Chem. Rev. 2003, 103, 1875–
1916; b) M. Hocek, Eur. J. Org. Chem. 2003, 245–254.
[18] K. L. Buchmueller, B. T. Hill, M. S. Platz, K. M. Weeks, J. Am. Chem. Soc.
2003, 125, 10850–10861.
Competing financial interest: C.D., T.L. and H.K. are employees of
Caprotec Bioanalytics GmbH (Berlin, Germany), a company that
produces and markets a SAH caproKit containing capture com-
pound 24a. C.D. and T.L. own stock options and H.K. and E.W. own
stocks.
[19] M. D. Abramoff, P. J. Magelhaes, S. J. Ram, Biophotonics International
2004, 11, 36–42.
[20] G. Schluckebier, M. Kozak, N. Bleimling, E. Weinhold, W. Saenger, J. Mol.
Biol. 1997, 265, 56–67.
Acknowledgements
[21] H. Cai, S. Clarke, J. Biol. Chem. 1999, 274, 13470–13479.
[22] F. Della Ragione, M. Porcelli, M. Cartenꢆ-Farina, V. Zappia, A. E. Pegg, Bio-
chem. J. 1985, 232, 335–341.
[23] a) B. Holz, S. Klimasauskas, S. Serva, E. Weinhold, Nucleic Acids Res.
1998, 26, 1076–1083; b) K. Goedecke, M. Pignot, R. S. Goody, A. J.
Scheidig, E. Weinhold, Nat. Struct. Biol. 2001, 8, 121–125.
We thank Kerstin Glensk and Dr. Wibke Peters for the preparation
of M.TaqI, M.HhaI and Clr4 and Priv.-Doz. Dr. Mark Helm for pro-
viding Trm4. This work was supported by the Human Frontier Sci-
ence Program Organization (HFSP Award 2007, RGP0058/2007-C).
264
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 256 – 265

SAH Capture Compounds
[24] S. Klimasauskas, T. Szyperski, S. Serva, K. Wꢀthrich, EMBO J. 1998, 17,
[26] a) E. Mirgorodskaya, C. Braeuer, P. Fucini, H. Lehrach, J. Gobom, Proteo-
mics 2005, 5, 399–408; b) J. Gobom, M. Schuerenberg, M. Mueller, D.
Theiss, H. Lehrach, E. Nordhoff, Anal. Chem. 2001, 73, 434–438.
317–324.
[25] J. Min, X. Zhang, X. Cheng, S. I. S. Grewal, R.-M. Xu, Nat. Struct. Biol.
2002, 9, 828–832.
Received: June 3, 2009
Published online on January 4, 2010
ChemBioChem 2010, 11, 256 – 265
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
265