Stereo- and regioselective azide/alkyne cycloadditions in carbonic anhydrase II via tethering, monitored by crystallography and mass spectrometry

Article abstract of DOI:10.1002/chem.201002437

The carbonic anhydrase II mutant His64Cys was prepared and applied to tethered alkyne/azide cycloaddition reactions. The azide component could be tethered to the enzyme surface through a disulfide bridge, while the alkyne component was reversibly coordina

Full text of DOI:10.1002/chem.201002437

DOI: 10.1002/chem.201002437
Stereo- and Regioselective Azide/Alkyne Cycloadditions in Carbonic
Anhydrase II via Tethering, Monitored by Crystallography and Mass
Spectrometry
Johannes Schulze Wischeler,[a] Dong Sun,[b] Nicola U. Sandner,[b] Uwe Linne,[b]
Andreas Heine,[a] Ulrich Koert,*[b] and Gerhard Klebe*[a]
Abstract: The carbonic anhydrase II mutant His64Cys was prepared and applied
to tethered alkyne/azide cycloaddition reactions. The azide component could be
tethered to the enzyme surface through a disulfide bridge, while the alkyne com-
ponent was reversibly coordinated through a sulfonamide anchor to the zinc ion in
the original catalytic center of the enzyme. The incipient orientation of the reactants
in the binding site and of the formed triazole product were characterized by crystal-
lography. The reaction progression could be monitored by HPLC-MS analysis.
Keywords: carbonic anhydrase
cycloaddition · enzyme catalysis ·
·
Huisgen reaction
triazole
·
tethering
·
Introduction
ated by approximately 106-fold using catalytic amounts of
CuI, leading to 1,4-triazole products regioselectively.[4,5]
Most likely the Cu-catalyzed reaction does not proceed
through a simple one-step mechanism, as frequently dis-
cussed for cycloadditions. Instead, a more complex reaction
pathway is favored, as confirmed by kinetic data and molec-
ular modeling and supported by simulations of activation
barriers.[6] These studies suggest that up to two CuI ions are
involved in the catalytic cycle. In the first step, subsequent
to the deprotonation of the terminal alkyne carbon, the two
CuI ions and the acetylide form a complex. Thereafter the
innermost nitrogen of the azide displaces one of the coordi-
nating ligand atoms from the second CuI ion thus activating
the terminal nitrogen for nucleophilic attack towards the
alkyne. After cyclization and protonation, the catalyst re-
leases the 1,2,3-triazole product.[6–9] The isolation of a crys-
talline CuI triazolide strongly supports the existence of such
click intermediates.[10]
The synthesis of 1,2,3-triazoles by the Huisgen reaction has
been established as one of the best-suited transformations
out of the entire repertoire of organic reactions to meet the
criteria of click chemistry.[1] This cycloaddition of an azide
and an alkyne facilitates the generation of large triazole li-
braries, particularly as a huge variety of both reagent com-
ponents can be introduced into a broad range of molecular
skeletons. However, for more than 40 years the impact of
this reaction was diminished due to the low regioselectivity,
which yielded mixtures of 1,4- and 1,5-regioisomers.[2] As a
further drawback, these cycloadditions required high activa-
tion energy often resulting in rather slow turnover. There-
fore, either elevated temperatures and/or pressures were ap-
plied.[3] A significant improvement of the Huisgen reaction
was the discovery of the activating effect of transition-metal
ions.[3–5] The addition of RuII ions to the Huisgen reactions
leads to the regioselective synthesis of 1,5-disubstituted tria-
zoles.[3] CuI shows an even higher impact on the Huisgen re-
action. Sharpless and co-workers demonstrated that the rate
of cycloaddition between azides and alkynes can be acceler-
With respect to pharmaceutical research, click chemistry
provides novel opportunities, for example, for polymeric mi-
celles, nanoscale delivery systems, and DNA labeling.[11] Fur-
thermore, libraries produced by click chemistry can be
tested in high-throughput screening campaigns on pharma-
ceutically interesting protein targets.
It has recently been shown that the protein environment
of an enzyme can function as such an activator for the Huis-
gen reaction by catalyzing triazole formation without the
need for supplementary metal ions.[12,13] This opportunity is
particularly supported by the fact that alkynes and azides
are essentially inert and chemically orthogonal to most bio-
logical and organic conditions, including highly functional-
ized biological molecules, for example, as given in the bind-
ing pockets of enzymes. Cycloaddition reactions can be ac-
celerated significantly once the reacting building blocks are
brought into favorable spatial vicinity in the enzyme binding
[a] Dr. J. Schulze Wischeler, Dr. A. Heine, Prof. Dr. G. Klebe
Institut fꢀr Pharmazeutische Chemie, Philipps-Universitꢁt Marburg
Marbacher Weg 6, 35032 Marburg (Germany)
Fax : (+49)6421-282-8994
E-mail: klebe@mailer.uni-marburg.de
[b] D. Sun, N. U. Sandner, Dr. U. Linne, Prof. Dr. U. Koert
Fachbereich Chemie, Philipps-Universitꢁt Marburg
Hans-Meerwein-Strasse, 35032 Marburg (Germany)
Fax : (+49)6421-282-5677
E-mail: koert@chemie.uni-marburg.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002437.
5842
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5842 – 5851

FULL PAPER
pocket.[13,14] This strategy is exploited in the so-called target-
guided synthesis (TGS) using a protein environment as con-
straint. The enzyme is incubated with a mixture of diverse
building blocks that exhibit graded affinities to the site of
mutual interaction. This site will thus select mainly those
fragments that experience highest affinity; consequently, the
product with the strongest binding affinity will overwhelm-
ingly be found. The enzyme operates as a kind of blueprint,
filtering out the most potent inhibitor from a large number
of possible products. Applying this TGS approach, Sharpless
et al. showed in several studies that enzymes such as acetyl-
cholinesterase (AChE) or carbonic anhydrase II (CAII) can
be exploited successfully for the assembly of triazoles.[12,13]
The goal of the present study was to examine the azide/
alkyne cyloaddition in a protein environment by integrating
a tethering approach into the target-guided synthesis con-
cept.[15,16] The tethering method
Results and Discussion
By site-directed mutagenesis, we replaced His64 with a cys-
teine at the rim of the binding pocket of CAII. The general
concept for a tethered approach to study azide/alkyne cyclo-
addition reactions in the binding pocket is shown in
Scheme 1. An arylsulfonamide bearing a substituent with a
terminal alkyne group 1 and an azide-functionalized disul-
fide 2 react with CAII-H64C to form the complex 3. In 3
the alkyne component is bound to a lipophilic pocket and
reversibly coordinated through its sulfonamide anchor to
the zinc ion in the original catalytic center of CAII (His94,
His96, His119). The azide component is connected covalent-
ly through a disulfide bond to Cys64. After [2+3] cycloaddi-
tion the formed triazole remains covalently attached to
facilitates the introduction of
fragments with weak affinity
into the binding pocket of pro-
teins by covalently attaching
the ligand to a cysteine residue
close to or within the active
site. This system should be ap-
propriate to keep track of the
cycloaddition reaction by crys-
tallographic diffraction tech-
niques. For this purpose we se-
lected the binding pocket of
carbonic anhydrase II (CAII).
By covalently attaching the
azide component to a mutation-
ally introduced cysteine residue
next to the binding pocket and
linking the alkyne component
through an appropriate sulfona-
mide anchor to coordinate effi-
ciently the active site zinc ion,
we intended to monitor the re-
actions in the binding pocket by
denaturating HPLC-MS analysis
performed in parallel to protein
crystallography. In particular, we
were interested in the influence
of the enzyme binding pocket
on the yield and the rate of tria-
zole formation, the regioselec-
tivity, and the stereodiscrimina-
tion of chiral starting materials.
Our study provides the per-
spective to design selective tri-
ACHTUNGTRENNUNGazoles catalyzed by the protein
environment, since site-directed
mutagenesis allows the design
of a specific protein surround-
ing that drives triazole forma-
Scheme 1. Tethered enzyme mediated alkyne–azide cycloadditions. a) Reaction of the alkyne component 1
and the azide component 2 with CAII-H64C yields the complex 3, which results in the formation of the teth-
ered triazoles 4 and 5. b) By site-directed mutagenesis, histidine 64 is replaced by a cysteine residue. The azide
is covalently attached to this cysteine residue under mild reducing conditions. The alkyne coordinates with its
tion towards a desired product. sulfonamide anchor to the active site ZnII ion. The triazole is formed upon incubation.
Chem. Eur. J. 2011, 17, 5842 – 5851
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5843

G. Klebe, U. Koert et al.
Cys64. A priori, the two regioisomers, the 1,4-triazole 4 and
the 1,5-triazole 5 can be formed.
This strategy facilitates formation of a complex with both
reaction components in close vicinity and allows one to
record successive situations along the chemical transition by
crystallography. Furthermore, this setup allows tracing the
reaction turnover by HPLC-MS under denaturating condi-
tions.
Initial experiments: Molecular modeling considerations sug-
gested 1a and 2a as the first prospective candidates for
azide and alkyne starting materials (Scheme 2). Both com-
pounds were incubated at room temperature with CAII-
H64C. The formed ternary complex could be structurally
characterized by crystallography (Figure 1a,d,e). The alkyne
(IC50 value: 245 nm) immobilized through the coordination
of the sulfonamide to the zinc ion accommodates its ethynyl
substituent into a small hydrophobic crevice flanked by
Phe130, Val121, and Leu197 (Figure 1e). The azide compo-
nent attached covalently through a propylene disulfide
Scheme 2. Tethered cycloaddition (1a + 2a + CAII-H64C ! 4a).
linker to Cys64 occupies a surface-exposed hydrophilic
canyon next to the exit of the catalytic site of CAII.
Both components show well-defined electron density indi-
cating only minor residual mobility. The terminal group of
the azide component establishes an extensive hydrogen-
bonding network mediated by three water molecules with
Figure 1. Crystal structure of (1a + 2a + CAII-H64C ! 4a). Distances are shown in ꢃ. a) Ligand geometries of 1a and 2a (yellow) in the cocrystal
structure with CAII-H64C (PDB code: 3KIG). The FoꢀFc omit maps for the reactants are displayed in all Figures at a s level of 2.0 as green mesh
(before cyclization) and as blue mesh (after cyclization). The solvent-accessible surface of the protein is schematically represented in white in all images.
Both building blocks bind with distinct orientation from each other to CAII-H64C, not facing each other. b) Ligand geometry of the cyclization product
(green) formed out of 1a and 2a individually cocrystallized with CAII-H64C and incubated with CuI solution for 48 h. c) Ligand geometry of the click
product superimposed with the reactants 1a und 2a before cyclization. d) Ligand geometry of 2a in the cocrystal structure with CAII-H64C. The azide is
fixed to the surface by a hydrogen-bonding network to Tyr7, Asn11, Gly63, Lys169, Phe231, Glu235, and Glu238 mediated by three water molecules. e)
Ligand geometry of 1a in the cocrystal structure with CAII-H64C. The sulfonamide adopts the typical conformation, coordinating to the ZnII ion, while
the alkyne moiety lies within a lipophilic pocket formed by Phe130, Val121, and Leu197. f) Ligand geometry of the click product 4a seen from a close-
up view.
5844
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5842 – 5851

Azide/Alkyne Cycloadditions in Carbonic Anhydrase II
FULL PAPER
Tyr7, Asn11, Gly63, Lys169, Phe231, and one of the terminal
oxygens of Glu235 (Figure 1d). Furthermore, the azide
moiety lies in close vicinity to the terminal oxygens of
Glu238, indicating a possible interaction. More detailed con-
siderations about the established interaction pattern are im-
possible due to unresolved protonation states. With the re-
agent components 1a and 2a no cycloaddition leading to tri-
azole formation could be observed in the crystalline state at
room temperature, even after a time span of more than four
months crystallization.
We subsequently treated the protein crystals (CAII-H64C
+ 1a + 2a) with solutions containing CuI ions at 188C to
induce triazole formation. Crystal structure analysis per-
formed with a crystal from the accordingly treated batch
after 48 h revealed clear changes in the difference electron
density (Figure 1b, c, f). Even though the molecular ar-
rangement detected in the first structure clearly argues
against a geometry competent of executing the planned tri-
eral CuI-catalyzed Huisgen reaction. Within experimental
accuracy no indication for the formation of the 1,5-product
5a was found.
However, as already indicated in our crystallization ex-
periments, the reaction rate (1a + 2a ! 4a) seemed to be
rather slow, most likely resulting in an incomplete triazole
formation. Why is the reaction rate so low and why can the
initial crystal structure with the separated reaction compo-
nents prior to the triazole ring formation be trapped? Clear-
ly, both components are rather tightly immobilized in local
depressions of the protein surface additionally fixed through
hydrogen bonds. Furthermore, the supplemented CuI ions
appear to be essential to catalyze the triazole formation
with these reactants. Finally, the building blocks might
simply not feature the spatial requirement necessary for a
fast cycloaddition. The rather strained arrangement of the
formed product suggests that the geometry of the reaction
components has not yet been ideal for this transformation
(1a + 2a + CAII-H64C).
ACHTUNGTRENNUNGazole formation, the electron density in the crystal structure
obtained after CuI addition shows the former alkyne compo-
nent to be rotated by 1808 now facing toward the azide com-
ponent. The former azide component experiences even
larger rearrangements, now orienting its side chain towards
the rotated alkyne component. Even though the electron
density is only weakly defined next to the region where the
novel triazole ring should have assembled to the likely gen-
erated 1,4-click product 4a, the trace of the electron density
agrees with the geometry of this product. We assume that
enhanced residual mobility of the central part of the reac-
tion product that is not in close surface contact with the pro-
tein and that incomplete product formation across all unit
cells of the protein crystal are responsible for the partially
defined electron density.
Triazole formation with optimized building blocks: Accord-
ingly, we decided to replace 1a by an extended reagent pu-
tatively orienting its alkyne component with a more favora-
ble geometry. Therefore, we screened the known crystal
structures in the PDB for binding geometries of CAII inhib-
itors to find a more promising scaffold. We selected the
ligand 6, shown in Scheme 3, from an appropriate CAII
complex crystal structure (PDB-code: 1G1D) as a starting
point for the design of an improved alkyne component.[17]
Compound 1b is extended by an amide group, and the
carbon atom, attached to the amide nitrogen atom and sub-
stituted by a second phenyl moiety, suggests better spatial
positioning to hook up the alkyne substituent (Figure 2a).
To further characterize the anticipated product formation
of 4a, we applied HPLC-MS analysis under denaturating
conditions. Therefore, we treated the enzyme in solution
similarly to our crystallization protocol with DTT, 1a, and
2a. To increase the reaction rate we raised the temperature
from 188C applied for our crystallographic studies to 378C
in solution. After incubation for 48 h in the presence of CuI,
we subjected the protein to mass analysis. The enzyme mass
was determined after purification over a size-exclusion
column to remove any protein-unbound compounds and to
prevent nonspecific background reactions during denatura-
tion. Furthermore, any unreacted and noncovalently at-
tached sulfonamide component will be removed upon dena-
turation. The analysis of the turned-over protein variant
showed a mass peak in agreement with some amount of the
formed triazole–sulfonamide reaction product covalently at-
tached to the enzyme. Consequently, some of the covalently
bound azide reacted with the noncovalently coordinated
alkyne to the cyclization product resulting in the above-
mentioned enzyme mass modification. In conclusion, the
components 1a and 2a react in the presence of CAII-H64C
to give the triazole product only with CuI assistance. The
trace of the weak electron density indicated formation of
the 1,4-regioisomer 4a, which is in accordance with the gen-
Scheme 3. Tethered cycloaddition (rac-1b + 2c + CAII-H64C ! 5c)
Chem. Eur. J. 2011, 17, 5842 – 5851
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5845

G. Klebe, U. Koert et al.
Figure 2. Crystal structures of (rac-1b + 2c + CAII-H64C ! 5c - PDB code: 3KNE). Distances are shown in ꢃ. a) Crystal structure of 6 (PDB code
1G1D; yellow) used as template for the design of 1b superimposed with the subsequently determined crystal structure of the rac-1b (blue). The solvent-
accessible surface of the protein is schematically represented in white in all images. The two racemic alkynes place their triple bond into different direc-
tions, which has a significant influence on product formation. b) Crystal structure of rac-1b and 2c in complex with CAII-H64C. The FoꢀFc maps for the
reactants are displayed in (b) and (c) at a s level of 2.0 as blue mesh. c) Crystal structure of rac-1b and 2c in complex with CAII-H64C seen from close-
up. d) Crystal structure of the in situ formed 1,5-S-triazole 5c (yellow). The FoꢀFc omit map for the cyclization product is displayed in (d) and (e) at a s
level of 2.0 (green mesh). The product participates in hydrogen bonds to Thr200, Pro201, Asn67, and Gln92 partially mediated by two water molecules.
e) Crystal structure of the in situ formed triazole 5c seen from a close-up view. f) Superimposition of the crystal structure of 1b and the in situ formed
click product. The alkyne 1b adopts the appropriate orientation for triazole formation.
The stereocenter in the alkyne component 1b allows the in-
vestigation of the stereo-discriminating effect of the protein
binding pocket on the cycloaddition reaction. Parallel to the
replacement of 1a by 1b, we also replaced the likely too
short azide component 2a by 2b and 2c. Both elongated
azides with additional linker atoms should minimize the dis-
tance to the alkyne and avoid pronounced H-bond fixation
of the azide on the protein surface. This reagent component
should then experience the required mobility for reaction.
Thereafter, rac-1b (racemic mixture), 2b, and 2c were syn-
thesized and each cocrystallized with the enzyme. A first
crystal structure of CAII-H64C could be determined after
cocrystallization with 2c under mild reducing conditions and
subsequent soaking with the racemic mixture of 1b for 2 h.
Crystals were flash-cooled for data collection shortly after
soaking. The structure shows that the new reactants bind in
a favorable orientation for the cycloaddition reaction (Fig-
ure 2b, c). Compound 2c has been assigned to the visible
part of the electron density, which indicates the directional
orientation of the molecule. The obtained binding mode
shows that the azide is no longer immobilized by an exten-
sive H-bonding network (compared to 2a, cf. Figure 1a) on
the surface, instead it experiences pronounced residual mo-
bility. The soaked rac-1b populates the binding pocket in
terms of both enantiomeric forms, and correct positioning
for triazoles formation can be anticipated. Although both
enantiomers are present in the binding pockets, the density
for the alkyne side chain is only weakly defined (Figure 2c).
To evaluate whether the cycloaddition is executed in the
crystalline state, we allowed crystals of CAII-H64C, rac-1b,
and 2c to grow in our crystallization trial for two months.
With the thus obtained crystals we were able to observe a
difference electron density of the formed cycloaddition
product 5c (Figure 2d, e). The product 5c was formed with-
out any additional CuI soaking. The electron density of the
formed triazole is very well-defined indicating high cyclo-
AHCTUNGTREGaNNUN ddition yields and regioselective formation of the 1,5-regio-
isomer. The phenyl ring and aromatic sulfonamide anchor
remain virtually in the binding pose attributed to the 1b
alkyne prior to cyclization. In contrast, the azide component
has entirely rotated by 1808 to undergo the reaction, very
similar to the situation observed for the previously described
cyclization of 1a with 2a. The linker atoms of the azide
show some flexibility. They adopt two conformations in
5846
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5842 – 5851

Azide/Alkyne Cycloadditions in Carbonic Anhydrase II
FULL PAPER
agreement with the electron density and the conformation
of the formed triazole. The carboxyl oxygen of the distal
amide bond forms, due to a splitting in two conformations,
either a hydrogen bond to Asn 67 or to Thr 200 (Figure 2e).
In the latter orientation it is also involved in a hydrogen
bond to Pro 201 mediated through a water molecule. A
second interstitial water molecule forms a hydrogen bond to
the proximal amide oxygen atom of the ligand and to Gln
92. The crystals were grown at 188C, which indicates that
the cyclization can already take place at this temperature.
Even though we started with rac-1b and detected accommo-
dation of both enantiomers in the binding pocket prior to
cycloaddition, only one enantiomeric product (5c) with S-
configuration is observed. The difference electron density is
well defined around the stereocenter under consideration.
Even though we cannot fully exclude (due to limited sensi-
tivity in crystal structure analyses in general) that still some
percentages of the other stereoisomer are produced, the
stereodiscrimination of the process, assisted by the protein
environment, is remarkable and clearly indicates the syn-
thetic potential of our approach. The structure was obtained
after short-term soaking using racemic starting material. It
is likely that both enantiomers exhibit very similar binding
affinity to the enzyme as indicated by inhibition data mea-
sured in a bioassay (IC50: 1b 16 nm; (R)-1b 13 nm; (S)-1b
26 nm). Crystallographic experience shows that the solid
state is highly discriminative with respect to the binding of
stereoisomers, particularly, if the crystals were grown over a
sufficiently long period. Then only one isomer is usually
found, even though a racemic (or diastereomeric) mixture
of similarly potent ligands was exposed. Through equilibri-
um over the long period for crystal growth even small affini-
ty differences (13 nm vs. 26 nm!) are sufficient to finally pop-
ulate only the more potent ligand in the binding pocket.
In our case the crystals had two months at 188C to accom-
plish the cycloaddition reaction, whereas for the complexes
with the reagents where we could detect the racemic mix-
ture of the starting material, a short soaking over 2 h has
only been allowed. Most likely, as the racemic alkyne com-
ponent is only bound to the enzyme reversibly, (S)-1b could
be gradually exchanged by (R)-1b present in excess in the
crystallization buffer. The latter R enantiomer adopts a
more favorable orientation for the cyclization reaction (Fig-
ure 2 f) with better suited spatial orientation of the cycliza-
tion step, thus after two months exclusively the 1,5-S-prod-
uct 5c is revealed.
component rac-1b was added to this mixture and the reac-
tion batch was incubated for 12 to 144 h at 378C. As long as
the noncovalently attached alkyne has not reacted, the
enzyme mass is not altered, since under denaturating condi-
tions the noncovalent component will be removed. Once the
reaction gradually forms the triazole product (regio- and ste-
reochemistry remained unresolved), an additional mass
peak is detected (29902 Da + 314 Da=30216 Da). As time
proceeded the mass spectra showed a shift in the ratio of
enzyme–azide to enzyme–product peaks (Figure 3b). Rela-
tive quantitative rating could be revealed by integration
over the peaks. Analyzing the triazole formation after 30 h
at 378C, we observed a yield of approximately 20% for
either the reaction of rac-1b with 2a, 2b, and 2c, respective-
ly. In the case of CAII+1b+2c, the [2+3] cycloaddition
was monitored by this approach for 12 to 144 h taking sam-
ples every 12 h. Figure 4 shows the cycloaddition rates as a
function of reaction time. Subsequent to an induction time,
the triazole formation occurs with similar reaction rate until
the reaction components are fully transformed to the prod-
uct. Pachꢄn and colleagues have described previously that
for different copper catalysts an incubation time is required,
for example, for Cu2+/ascorbate of up to 5 h, before any
click product can be detected; however, no satisfactory ex-
planation for this observation has yet been provided.[18] The
MS analysis established here is a first step towards future ki-
netic studies of the enzyme-tethered azide/alkyne cycloaddi-
tion. Preliminary experiments showed elevated reaction
rates at augmented temperatures as well as upon the supple-
mentation with CuI and CuII ions. Addition of RuII
ACHTUNGTRENNUNG(dmso)4,
RuIII, or ZnII had no effects on the cycloaddition product
formation rates.
Control experiments: Up to now, it could only be evidenced
indirectly in the literature that triazole formation occurs
inside the active site of a protein.[19] Triazoles formed in
background reactions in solution might give the impression
of product formation inside the active site of the protein. To
verify that in our case the click reaction has actually been
executed inside the binding pocket, we performed several
control experiments. First, we added the potent CAII inhibi-
tor ethoxzolamide to our reaction batch of rac-1b, 2c, and
CAII-H64C. After exposure at 378C for 72 h, the HPLC-
MS spectra did not indicate any product formation. Only
the mass shift for the tethered 2c bound to the enzyme
(29729 Da ! 29902 Da) could be recorded. Clearly, ethox-
zolamide displaces or prevents binding of the less potent
sulfonamide 1b. If a background reaction occurs in solution,
the generated triazole would be tethered to Cys 64, leading
to the product mass shift. Similarly any reaction of the azide
component covalently tethered to the protein and the un-
bound sulfonamide-type alkyne can be excluded. Conse-
quently, the alkyne is only suitable for the cyclization once
bound to the proteinꢅs active site. To validate this conclusion
we tried to react the azide 2c with different alkynes lacking
a sulfonamide anchor. Also under these conditions, howev-
er, no product formation could be observed.
In-solution studies of the triazole formation: Our elaborate
crystallographic analysis only defines the start and end point
of the reaction. It hardly allows one to record details about
the reaction progression. Thus, we decided to monitor the
triazole formation in solution by mass spectrometric experi-
ments parallel to the crystallographic studies. The HPLC-
MS analysis of CAII-H64C incubated with the azide compo-
nent 2c showed a mass peak in agreement with the covalent
CAII-H64C azide complex (with 2c: 29729 Da + 173 Da=
29902 Da; Figure 3a). Subsequently, the racemic alkyne
Chem. Eur. J. 2011, 17, 5842 – 5851
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5847

G. Klebe, U. Koert et al.
Figure 3. Monitoring reaction progress by HPLC-MS. a) Reaction pathway for (CAII-H64C + 2c + rac-1b ! 5c) monitored by HPLC-MS experi-
ments. CAII-H64C shows a mass of 29729 Da. After addition of 1b, 2c, and 100 mm DTT to reduce the azide internal disulfide bridge, the HPLC-MS
analysis shows one definite mass peak of 29902 Da, representing the mass of the enzyme with the covalently attached azide. Upon incubation with the
alkyne at 378C the triazole formation is induced by the enzyme and an additional mass peak of 30216 Da appears in the mass spectrum. b) Incubating
the setup mentioned in (a) for 12 to 144 h taking samples every 12 h we could monitor the rate of triazole formation by comparing the peaks correspond-
ing to the enzyme–azide with those of the enzyme–triazole products. Within 132 h full conversion of enzyme-bound azide and alkyne components can be
observed.
Finally, in another control experiment the enzyme was pu-
rified over a size-exclusion column after short incubation at
room temperature with the azide 2c and the alkyne rac-1b.
Thereby the enzyme was purified from all unbound reac-
tants. Only the tethered and zinc-coordinated reagent com-
ponents were expected to remain in the samples. After ex-
posure at 378C for 48 h and HPLC-MS analysis of these
samples, product formation could be observed with the
same rate as observed for all the other test series. Further-
more, we analyzed the solutions from which the protein had
been extracted. If any solvated reactants would have been
present, a background reaction might have occurred. How-
ever, no triazole product was detected. These control experi-
Figure 4. Time-dependent [2+3] cycloaddition (CAII-H64C
+ 2c +
ments strongly support our hypothesis that the cycloaddition
reaction occurs through the tethered target-guided synthesis
concept shown in Scheme 1.
rac-1b ! 5c); the triazole formation rate of the CAII-H64C azide to
CAII-H64C triazole complex is shown in percent with respect to the re-
action time in hours.
5848
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5842 – 5851

Azide/Alkyne Cycloadditions in Carbonic Anhydrase II
FULL PAPER
ences) with the following deviations. A first 100 mL culture was incubat-
ed in the presence of 100 mgmLꢀ1 amp and 34 mgmLꢀ1 cam overnight at
378C. A 10 mL portion of this solution was used to inoculate a 1000 mL
expression culture with 100 mgmLꢀ1 amp. Expression of the CAII gluta-
thione-S-transferase fusion protein was induced at an optical density at
600 nm (OD600) of 0.6–0.8 with 2 mm IPTG, and incubation was contin-
ued for 4 h at 328C.
Conclusion
For the first time, a carbonic anhydrase II mutant was used
to perform a tethered alkyne/azide cycloaddition reaction.
The azide component was tethered to the enzyme surface
through a disulfide bridge, while the alkyne component was
reversibly coordinated through a sulfonamide anchor to the
zinc ion in the literal catalytic center of the enzyme. The in-
cipient orientation of both reactants in the active site and
that of the formed triazole product were analyzed by crys-
tallography. In addition, HPLC-MS studies allowed the pro-
gression of the reaction to be monitored in solution. The re-
action of 1a with 2a required the presence of CuI to form
the 1,4-triazole product, otherwise no product formation
was observed over a time span of several months. The
alkyne rac-1b and the azide 2c yield under similar condi-
tions, apart of the absence of copper-(I), the 1,5-triazole 5c
regioselectively in the binding pocket of the protein. Re-
markable is the protein-induced stereodiscrimination, ob-
served in the crystal, as only the slightly more potent (R)-1b
reacted to give the 1,5-S-product.
Lysis: Cells were harvested by centrifugation at 5000 rpm for 15 min at
48C. The cell pellet was re-suspended with 100 mL PBS buffer (140 mm
NaCl, 2.7 mm KCl, 10 mm Na2HPO4, 1.8 mm KH2PO4 (VWR), pH 7.3),
which contained a protease inhibitor cocktail tablet (Roche) and approxi-
mately 100 mg of lysozyme (Merck). After chilling for 30 min on ice, the
cells were lysed by ultrasonic sonication using a Branson Sonifier 250, ap-
plying seven cycles of 2 min with duty cycle set to 70 and output control
to 4. This solution was centrifuged at 20000 rpm for 45 min at 48C and
filtered (0.45 mm) to separate the soluble fractions from the insoluble ma-
terial.
Protein purification: The flow rate along all purification steps was adjust-
ed to a maximum pressure of 0.25 MPa. The sample was applied to a
HiPrep GSTPrep FF 16/10 column (GE Healthcare) at a flow rate of
1 mLminꢀ1 to ensure that all the fusion protein binds to the sepharose.
Afterwards the column was washed with 10 column-volumes of PBS
buffer at a flow rate of 3 mLminꢀ1. To induce the cutting between the
GST tag and the CAII enzyme, approximately 500 I.U. of thrombin
(from Beriplast, CSL Behring) were added at 1 mLminꢀ1 and the column
was slightly shaken for 20 h at room temperature. Following incubation,
the target protein was eluted from the column with approximately three
column-volumes of PBS at a flow rate of 1 mLminꢀ1. The elution volume
was mixed with the same volume of 100 mm Tris-HCl pH 7.4, 1m NaCl
buffer. This mixture was purified over a HiTrap benzamidine column
(Amersham Bioscience) to separate the thrombin from CAII. As a final
purification step, the probe was applied on a HiLoad 26/60 Superdex200
size-exclusion column (GE Healthcare) to purify the target enzyme from
any last impurity and change the buffer to 50 mm Tris pH 7.8. To monitor
the different purification steps, the probes were controlled by SDS-
PAGE.[20] The final purification of 1 L expression culture yielded approxi-
mately 25 mg CAII. Finally the GSTPrep column was washed with re-
duced gluthathion buffer to elute the GST tag from the column and
check for full thrombin cleavage. The purified enzyme was concentrated
to a final concentration of 10 mgmLꢀ1, which was used for crystallization
experiments. For further measurements the enzyme was diluted to the set
concentration using a 50 mm Tris-HCl pH 7.8 buffer.
The proposed protein-tethered system provides the per-
spective to design novel reactions in sites exposed on pro-
tein surfaces. In a more sophisticatedly designed system, the
next step would be to omit the tether and to proceed via
multiple turnovers towards catalytic applications.
Experimental Section
Cloning, expression, purification, and mutagenesis of CAII: Sodium chlo-
ride, yeast extract, peptone, ampicillin (amp), chloramphenicol (cam), di-
thiothreitol (DTT), isopropyl b-d-1-thiogalactopyranoside (IPTG), and
disodium hydrogen phosphate were purchased from Roth. Diammonium
sulfate, agarose, potassium chloride, and ascorbic acid were purchased
from Fluka. Dimethyl sulfoxide (DMSO), copper sulfate, and zinc sulfate
were obtained from Sigma–Aldrich. All other chemicals were obtained
as mentioned. Vivaspin20 (10000) columns from Sartorius Stedium Bio-
tech were used to concentrate protein samples. HPLC analyses were per-
formed by using an ꢆkta FPLC (Amersham Pharmacia biotech), consist-
ing of the following components: UPC-900, U-900, Frac-900, INV-907,
and M-925. UV analyses were done using a BioRad Smartspec 3000. For
centrifugation a Beckmann Coulter Avanta J-25 centrifuge was used.
Crystallization of CAII complexes: The CAII crystals were grown using
the sitting drop vapor diffusion method at 188C by mixing the protein so-
lution (5 mL: ca. 10 mgmLꢀ1) with well solution (5 mL: 2.75m (NH4)2SO4,
50 mm Tris-HCl pH 8.0). After a few days the crystallization drops were
seeded with a seeding solution from old CAII crystals. Crystals appeared
within 2–4 weeks in space group P21 and took up to two months to reach
their maximum size. Complex structures were obtained by cocrystalliza-
tion or soaking of the enzyme with the building blocks (1 mm). For cryo-
protection, crystals were briefly soaked in mother liquor containing 25%
glycerol.
Cloning and mutagenesis: The gene coding for CAII was amplified by
PCR from a construct, generously provided by the group of Pastorekova
(Slovak Academy of Sciences, Bratislava), and cloned into the expression
vector pGEX-4T1 (GE Healthcare) between the restriction sites XhoI
and EcoRI. The following primers were used: XhoIC2a: 5’-AC-
CGCTCGAGTTATTTGAAGGAAGCTTTGATTTGC-3’ and EcoRI-
CA2s: 5’-CGGAATTCATGTCCCATCACTGGGGGTA-3’. The plasmid
was transformed into E.coli XL-2Blue competent cells and incubated for
plasmid multiplication. A sequencing experiment showed that the se-
quence was identical to the original copy (data not shown). The mutants
H64C of CAII were made by site-directed mutagenesis using the expres-
sion vector containing the wild-type CAII coding region. The point muta-
tions were made using the QuikChange II Kit (Quiagen) for site-directed
mutagenesis. The following primers were used: CAIIH64Cf 5’-GAG-
GATCCTCAACAATGGTTGCGCTTTCAACGTGGAGTTTG-3’ and
Data collection, phasing, and refinement: The data sets were collected at
the synchrotron BESSY II in Berlin/Germany on PSF beamline 14.2
(PDB codes: 3 KIG, 3 KNE). Data were processed and scaled with
Denzo and Scalepack as implemented in HKL2000.[21] The structures
were determined by the molecular replacement method with Phaser[22]
with our first inhouse CAII-H64C structure as the search model. Refine-
ment was continued with CNS[23] and SHELXL-97.[24] For each refine-
ment step at least ten cycles of conjugate gradient minimization were per-
formed, with restraints on bond lengths, angles, and B values. Intermit-
tent cycles of model building were done with the program COOT.[25] The
coordinates have been deposited in the PDB (http://www.rcsb.org/pdb/)
with the following access codes: 3KIG, 3KNE. The statistics are shown in
Table 1.
CAIIH64Cr
5’-CAAACTCCACGTTGAAAGCGCAACCATTGTT-
GAGGATCCTC-3’.
HPLC-MS analysis: For HPLC-MS measurements 10 mL of the CAII-
H64C mutant (75 mm), 5 mL DMSO, 10 mL DTT solution (500 mm), 0.5 mL
azide (100 mm), and 0.5 mL alkyne (100 mm) were diluted to a final
Gene expression: CAII and CAII mutants were expressed and purified
according to the GST gene fusion system handbook (Amersham Biosci-
Chem. Eur. J. 2011, 17, 5842 – 5851
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5849

G. Klebe, U. Koert et al.
Table 1. Summary of data sets collected.
Compound synthesis: Chemicals from
commercial sources were purchased at
the highest commercial quality used
without further purifications. E. Merck
silica gel (60, particle size 0.040–
0.063 mm) was used for flash column
chromatography. 1H and 13C NMR
spectra were recorded at 258C on a
Bruker ARX 300 instrument relative
to tetramethylsilane or deuterated sol-
vent as internal standard. The follow-
ing abbreviations were used to explain
the signals: s=singlet, d=doublet, t=
triplet, q=quartet, qi=quintet, m=
multiplet. IR spectra were recorded on
a Bruker Alpha-P FT-IR spectrome-
ter, and values are given in units of
Dataset
3KIG[a]
3KNE
CAII-H64C
click product 1a+2a
CAII-H64C
rac-1b+2c
resolution [ꢃ]
space group
25.00–1.39
P21
25.00–1.35
P21
30.00–1.79
P21
50.0–1.30
P21
unit cell parameters [ꢃ]
a [ꢃ]
42.2
42.4
41.5
72.0
104.3
1.37–1.35
3.0 (2.3)
52457
42.4
41.5
72.0
103.9
1.82–1.79
3.4 (2.5)
21988
42.4
41.5
72.4
104.2
1.32–1.30
2.9 (1.8)
58654
b [ꢃ]
41.4
c [ꢃ]
72.3
b [8]
104.1
highest resolution shell [ꢃ]
redundancy[b]
no. of unique reflections[b]
1.41–1.39
2.7 (2.5)
47302
A
(2456)
A
ACHTUNGTRNE(NUNG 2329)
completeness[b]
96.7 (95.7)
19.9 (3.3)
3.8 (24.8)
13.4; 12.8
19.0; 18.0
15.5
98.2 (93.1)
13.8 (2.8)
7.4 (33.5)
12.8; 12.0
18.3; 17.3
15.3
95.0 (86.4)
16.0 (5.3)
6.1 (16.1)
16.2; 15.3
24.0; 22.4
18.9
97.0 (78.0)
17.7 (2.2)
5.5 (33.7)
24.2; 23.1
26.6; 25.3
15.0
I/s[b]
Rsym [%][b]
cmꢀ1
were used to explain the multiplicities:
s=strong, m=middle, b=broad.
.
The following abbreviations
Rcryst (Fo; F>4sFo)
Rfree (Fo; F> 4sFo)
mean B-factor [ꢃ2] protein
mean B-factor [ꢃ2] ligands
High-resolution mass spectra (HR-
MS) were recorded on a Finnigan Agi-
lent MAT 95S mass spectrometer.
Melting points were measured on a
Stuart SMP10 melting points appara-
tus and are uncorrected.
azide 31.5[d]
alkyne 28.3[d]
26.6
19.6[d]
–
–
mean B-factor [ꢃ2] water
occupancy ligands
27.9
–
–
–
–
1.0
0.29–1.0[c]
88.9
most favored geometry [%]
additionally allowed [%]
generously allowed [%]
RMSD bonds [ꢃ]
89.4
10.2
0.4
0.010
89.0
10.0
0.9
0.007
1.803
87.9
11.6
0.5
0.009
2.1
10.6
0.5
0.011
2.2
Synthesis of alkynes 1
1a:
12.0 mmol),
0.6 mmol),
Diisopropylamine
(1.70 mL,
(0.15 mL,
(0.12 g,
PtBu3
PdCl2A(PhCN)2
RMSD angles [8]
2.1
CTHUNGTRENNUNG
[a] PDB code. [b] Highest resolution shell. [c] Sulfonamide 1.0; triazole (N) 0.68; chains 0.28 and 0.40. [d] Re-
actants and product only.
0.3 mmol), CuI (38 mg, 0.2 mmol), and
ethynyltrimethylsilane (1.70 mL,
12.0 mmol) was added at room tem-
perature to a solution of the 3-bromo-
volume of 50 mL with buffer (50 mm Tris-HCl pH 7.8). The samples were
incubated at 378C between 12 and 144 h. After incubation, the sample
solutions were pre-purified by size-exclusion chromatography (Illustra
AutoSeq G-50 columns - GE Healthcare) to separate them from excess
substrates to avoid undesired side reactions. Complete desalting of the
protein was achieved by HPLC using an Agilent 1100 system; samples
were applied to a monolithic 50/1 ProSwift RP-4H column (Dionex). De-
salted protein samples were eluted by the following gradient of buffer A
(water/0.05% formic acid) and buffer B (acetonitrile/0.045% formic
benzenesulfonamide (2.00 g, 10.0 mmol) in dioxane (10 mL). The result-
ing dark reaction mixture was stirred for 16 h. It was diluted with aque-
ous saturated NaHCO3 (20 mL) and extracted with EtOAc (60 mL). The
organic layer was washed with brine, dried over Na2SO4, filtered, and
concentrated under reduced pressure. The crude product was purified by
flash column chromatography on silica gel (ethyl acetate/pentane, 1:4) to
give 3-[(trimethylsilyl)ethynyl]benzenesulfonamide (2.18 g, 8.6 mmol,
86%). 1H NMR: (300 MHz, CDCl3): d=0.25 (s, 9H), 5.07 (s, 2H), 7.45
(t, J=7.8 Hz, 1H), 7.63 (dt, J=7.7, 1.3 Hz, 1H), 7.84 (ddd, J=7.9, 1.9,
1.1 Hz, 1H), 8.01 ppm (t, J=1.5 Hz, 1H); 13C NMR: (75 MHz, CDCl3,
258C): d=ꢀ0.08, 97.21, 102.97, 124.75, 126.10, 129.24, 129.96, 135.88,
142.37 ppm.
acid) at a column temperature of 408C and a flow rate of 0.2 mLminꢀ1
:
Isocratic elution with A (5%) for 2 min, followed by a linear gradient to
B (95%) within 8 min and holding B (95%) for additional 4 min. Online
mass spectrometric analysis was done with a Qstar Pulsar I mass spec-
trometer (Applied Biosystems) equipped with an ESI source. Measure-
ment parameters were as follows: DP1 75, FP 265, DP2 15, CAD 2, GS1
35, and CUR 25. The applied voltage was 5000 V. Positive ions within the
mass range of 500–2000 m/z were detected. For better performance, the
“enhance all” mode was activated.
TBAF hydrate (0.75 g, 2.9 mmol) was added at room temperature to a
solution of the 3-[(trimethylsilyl)ethynyl]benzenesulfonamide (2.18 g,
8.6 mmol) in THF (30 mL). The resulting reaction mixture was stirred for
2 h. It was diluted with water (40 mL) and extracted with EtOAc
(100 mL). The organic layer was washed with brine, dried over Na2SO4,
filtered, and concentrated under reduced pressure. The crude product
was purified by flash column chromatography on silica gel (ethyl acetate/
pentane, 1:1) to give the 3-ethynylbenzenesulfonamide 1a (1.42 g,
Metal-ion incubation: In some cases a solution of CuI, CuII (CuSO4), ZnII
(ZnSO4), or RuII (RuCl ·
2 ACTHNUTRGNE(UGN dmso)4) of 20–40 mm were added before incuba-
1
5.5 mmol, 91%). H NMR: (300 MHz, [D6]DMSO): d=4.38 (s, 1H), 7.46
tion. CuI was obtained by adding a fivefold excess of ascorbic acid to a
CuII solution. We applied metal-ion concentrations ranging from 10 mm to
1 mm. At concentrations above 40 mm the samples often showed spectra
that were difficult to interpret most probably due to denaturation. To an-
alyze the impact the metal ions develop on the mutant without any build-
ing block, we performed an initial measurement by incubating CAII-
H64C with each metal ion solution. In the case of CuI, we observed in
several samples an additional mass peak that shows the size of the
mutant reduced by a mass of 35 Da. This byproduct is dependent on the
reaction time and the applied temperature, and its occurrence increased
with raised concentrations along solvent evaporation. We assign these
changes to some redox instabilities of the protein upon exposure to CuI
ions.
(s, 2H), 7.60 (t, J=7.8 Hz, 1H), 7.71 (dt, J=7.7, 1.3 Hz, 1H), 7.84 (dt,
J=7.7, 1.5 Hz, 1H), 7.89 ppm (t, J=1.5 Hz, 1H); 13C NMR: (75 MHz,
[D6]DMSO): d=82.04, 82.38, 122.42, 125.93, 128.65, 129.62, 134.76,
144.60 ppm; IR: n˜ =3362 (m), 3260 (s), 1324 (s), 1304 (s), 1140 (s), 1087
(s), 916 (s), 802 (s), 689 (s), 650 (s), 586 (s), 495 cmꢀ1 (s); HR-MS (ESI):
calculated for C8H7NO2SNa [M+Na]+ 204.0095; found: 204.0090; m.p.:
139–1408C.
rac-1b: HOBt (0.58 g, 3.8 mmol), EDAC (0.73 g, 3.8 mmol), and NEt3
(1.10 mL, 7.6 mmol) was added at room temperature to a solution of rac-
emic 3-phenylpropargylamine[23] (0.50 g, 3.8 mmol) and 4-sulfamoylben-
zoic acid (0.77 g, 3.8 mmol) in CH2Cl2 (5.00 mL) and DMF (0.50 mL).
The resulting solution was stirred for 16 h. It was diluted with water
(20 mL) and extracted with CH2Cl2/MeOH (10:1) (40 mL). The organic
5850
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 5842 – 5851

Azide/Alkyne Cycloadditions in Carbonic Anhydrase II
FULL PAPER
layer was washed with brine, dried over Na2SO4, filtered, and concentrat-
ed under reduced pressure. The crude product was purified by flash
column chromatography on silica gel (CHCl3/MeOH, 10:1) to give the
alkyne 1b (0.85 g, 2.7 mmol, 71%). 1H NMR: (300 MHz, [D6]DMSO):
d=3.57 (d, J=2.5 Hz, 1H), 6.11 (dd, J=2.5, 8.2 Hz, 1H), 7.33–7.42 (m,
3H), 7.51 (s, 2H), 7.47–7.55 (m, 2H), 7.90 (d, J=8.5 Hz, 2H), 8.05 (d, J=
8.7 Hz, 2H), 9.64 ppm (d, J=8.3 Hz, 1H); 13C NMR: (75 MHz,
[D6]DMSO): d=44.05, 75.45, 82.35, 125.59, 126.91, 127.79, 128.27, 128.50,
136.46, 138.91, 146.56, 164.53 ppm; IR n˜ =3280 (m), 3188 (b), 1639 (s),
1535 (s), 1333 (s), 1158 (s), 920 (m), 858 (m), 606 (b) 535 cmꢀ1 (m); HR-
MS (ESI): calculated for C16H14N2O3SNa [M+Na]+ 337.0623; found:
337.0617; m.p.: 179–1808C.
available to us and giving J. Schulze Wischeler an introduction to the ex-
pression of the protein.
[1] R. Huisgen, G. Szeimies, L. Mçbius, Chem. Ber. 1967, 100, 2494–
2507.
[2] G. C. Tron, T. Pirali, R. A. Billington, P. L. Canonico, G. Sorba,
A. A. Genazzani, Med. Res. Rev. 2008, 28, 278–308.
[3] L. Zhang, X. Chen, P. Xue, H. H. Y. Sun, I. D. Williams, K. B. Sharp-
less, V. V. Fokin, G. Jia, J. Am. Chem. Soc. 2005, 127, 15998–15999.
[4] V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless, Angew.
Chem. 2002, 114, 2708–2711; Angew. Chem. Int. Ed. 2002, 41, 2596–
2599.
Synthesis of azides 2
[5] C. W. Tornøe, C. Christensen, M. Meldal, J. Org. Chem. 2002, 67,
3057–3064.
[6] F. Himo, T. Lovell, R. Hilgraf, V. V. Rostovtsev, L. Noodleman,
K. B. Sharpless, V. V. Fokin, J. Am. Chem. Soc. 2005, 127, 210–216.
[7] V. O. Rodionov, S. I. Presolski, D. Diaz Diaz, V. V. Fokin, M. G.
Finn, J. Am. Chem. Soc. 2007, 129, 12705–12712.
[8] V. D. Bock, H. Hiemstra, J. H. van Maarseveen, Eur. J. Org. Chem.
2006, 51–68.
[9] V. O. Rodionov, V. V. Fokin, M. G. Finn, Angew. Chem. 2005, 117,
2250–2255; Angew. Chem. Int. Ed. 2005, 44, 2210–2215.
[10] C. Nolte, P. Mayer, B. F. Straub, Angew. Chem. 2007, 119, 2147–
2149; Angew. Chem. Int. Ed. 2007, 46, 2101–2103.
[11] C. D. Hein, X.-M. Liu, D. Wang, Pharm. Res. 2008, 25, 2216–2230.
[12] A. Krasinski, Z. Radic, R. Manetsch, J. Raushel, P. Taylor, K. B.
Sharpless, H. C. Kolb, J. Am. Chem. Soc. 2005, 127, 6686–6692.
[13] V. P. Mocharla, B. Colasson, L. V. Lee, S. Rçper, K. B. Sharpless, C.-
H. Wong, H. C. Kolb, Angew. Chem. 2005, 117, 118–122; Angew.
Chem. Int. Ed. 2005, 44, 116–120.
[14] W. L. Mock, T. A. Irra, J. P. Wepsiec, T. L. Manimaran, J. Org.
Chem. 1983, 48, 3619–3620.
[15] D. A. Erlanson, A. C. Braisted, D. R. Raphael, M. Randal, R. M.
Stroud, E. M. Gordon, J. A. Wells, Proc. Natl. Acad. Sci. USA 2000,
97, 9367–9372.
2a: Chloroacetyl chloride (3.7 mL, 46.6 mmol) was added at 08C to a sus-
pension of the cystamine dihydrochloride (5.0 g, 22.2 mmol) in CH2Cl2
(100 mL) and NEt3 (12.5 mL, 88.8 mmol). The reaction mixture was
stirred for 15 h at room temperature. It was diluted in CH2Cl2 (100 mL)
and washed subsequently with water, aqueous NaHCO3, water, 2n HCl,
and aqueous saturated NaCl (100 mL each). The organic layer was dried
with Na2SO4, filtered, and concentrated under reduced pressure. The re-
sulting solid was suspended in MeOH/Et2O (100 mL, 1:10), filtered, and
washed by Et2O to give the corresponding chloride. The chloride (1.15 g,
3.8 mmol) was dissolved in DMF (16 mL), and NaN3 (0.51 g, 7.9 mmol)
was added. The resulting solution was stirred for 16 h at 608C. It was
cooled to room temperature and diluted with CH2Cl2 and water. The or-
ganic layer was dried with Na2SO4, filtered, and concentrated under re-
duced pressure. Precipitation in cool MeOH/Et2O (1:25) gave the azide
2a (1.64 g, 3.3 mmol, 87%). 1H NMR: (300 MHz, [D6]DMSO): d=2.80
(t, J=6.8 Hz, 4H), 3.39 (q, J=6.4 Hz, 4H), 3.83 (s, 4H), 8.31 ppm (t, J=
5.4 Hz, 2H); 13C NMR: (75 MHz, [D6]DMSO): d=36.88, 37.96, 50.76,
167.41 ppm; IR: n˜ =3252 (b), 2102 (s, N3), 1644 (s), 1546 (s), 1273 (s), 737
(m), 574 cmꢀ1 (m); HR-MS (ESI): calculated for C8H14N8O2S2Na
[M+Na]+ 341.0579; found: 341.0573.
The azides 2b and 2c where prepared along the same procedure as for
2a using cystamine/2-chloropropionyl chloride for 2b and homocysta-
mine/chloroacetyl chloride for 2c.
2b: 1H NMR: (300 MHz, [D6]DMSO): d=2.38 (t, J=6.4 Hz, 4H), 2.77
(t, J=6.7 Hz, 4H), 3.35 (q, J=6.4 Hz, 4H), 3.50 (t, J=6.3 Hz, 4H),
8.21 ppm (t, J=5.3 Hz, 2H); 13C NMR: (75 MHz, [D6]DMSO): d=34.50,
37.09, 37.96, 46.90, 169.63 ppm; IR: n˜ =3240 (b), 2090 (s, N3), 1631 (s),
1546 (s), 1271 (s), 1184 (m), 580 cmꢀ1 (m); HR-MS (ESI): calculated for
C10H18N8O2S2Na [M+Na]+ 369.0892; found: 369.0886.
[16] D. A. Erlanson, R. S. McDowell, M. M. He, M. Randal, R. L. Sim-
mons, J. Am. Chem. Soc. 2003, 125, 5602.
[17] C.-Y. Kim, J. S. Chang, J. B. Doyon, T. T. Baird, C. A. Fierke, A.
Jain, D. W. Christianson, J. Am. Chem. Soc. 2000, 122, 12125–12134.
[18] L. D. Pachꢄn, J. H. van Maarseveen, G. Rothenberg, Adv. Synth.
Catal. 2005, 347, 811–815.
[19] W. G. Lewis, L. G. Green, F. Grynszpan, Z. Radic, P. R. Carlier, P.
Taylor, M. G. Finn, K. B. Sharpless, Angew. Chem. 2002, 114, 1095–
1099; Angew. Chem. Int. Ed. 2002, 41, 1053–1057.
1
2c: H NMR: (300 MHz, CDCl3, 258C): d=1.95 (qi, J=7.0 Hz, 4H), 2.72
(t, J=7.1 Hz, 4H), 3.42 (q, J=6.6 Hz, 4H), 3.99 (s, 4H), 6.57 ppm (s,
2H); 13C NMR: (75 MHz, CDCl3, 258C): d=29.05, 36.22, 38.33, 52.90,
166.90 ppm; IR: n˜ =3297 (b), 2099 (s, N3), 1642 (s), 1543 (s), 1272 (s), 752
(m), 570 cmꢀ1 (m); HR-MS (ESI): calculated for C10H18N8O2S2Na
[M+Na]+ 369.0892; found: 369.0886.
[20] U. K. Laemmli, Nature 1970, 227, 680–685.
[21] Z. Otwinowski, W. Minor, C. W. Carter, Jr., Methods in Enzymol.,
Vol. 276, Academic Press, New York, 1997, pp. 307–326.
[22] L. C. Storoni, A. J. McCoy, R. J. Read, Acta Crystallogr. Sect. A
2004, 60, 432–438.
[23] A. T. Brꢀnger, P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros,
R. W. Grosse-Kunstleve, J.-S. Jiang, J. Kuszewski, M. Nilges, N. S.
Pannu, R. J. Read, L. M. Rice, T. Simonson, G. L. Warren, Acta
Crystallogr. Sect. A 1998, 54, 905–921.
IC50 measurements: The IC50 values were determined photometrically
with p-nitrophenol acetate (pNPA) as substrate as described by Pocker
and Stone.[27] Acetazolamide was measured as a reference with an IC50 of
(2.9ꢁ0.3) nm (Lit.: 3.4 nm[28]). A plate reader Labsystem iEMS Reader
MF was used for the photometric measurement. The buffer solution con-
tains 50 mm Na2HPO4, 150 mm NaCl, 70 nm hCAII, and 3 mm pNPA.
[24] G. M. Sheldrick, T. R. Schneider, W. C. Charles, Jr., M. S. Robert,
Methods in Enzymol., Vol. 277, Academic Press, New York, 1997,
pp. 319–343.
[25] P. Emsley, K. Cowtan, Acta Crystallogr. Sect. A 2004, 60, 2126–2132.
[26] F. Messina, M. Botta, F. Corelli, M. P. Schneider, F. Fazio, J. Org.
Chem. 1999, 64, 3767–3769.
[27] Y. Pocker, J. T. Stone, Biochemistry 1967, 6, 668–678.
[28] M. F. Surgrue, A. Harris, I. Adamsons, Drugs Today 1997, 33, 283–
298.
Acknowledgements
The present research study was financially supported by the DFG (grant
FO 594 KL1204/10-1; KO 1349/9-1). We acknowledge support from the
beamline staff at Bessy II, Berlin and a travel grant from the Helmholtz-
zentrum Berlin. We thank Prof. S. Pastorekova, Institute of Virology—
Slovak Academy of Sciences (Bratislava), for making a clone of CAII
Received: August 23, 2010
Revised: December 23, 2010
Published online: April 19, 2011
Chem. Eur. J. 2011, 17, 5842 – 5851
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5851