Protein enrichment by capture-release based on strain-promoted cycloaddition of azide with bicyclononyne (BCN)

Article abstract of DOI:10.1016/j.bmc.2011.07.049

An enrichment strategy was devised for azide derivatized macromolecules, based on strain-promoted alkyne-azide cycloaddition (SPAAC) and a cleavable linker. A ring-strained alkyne, bicyclo[6.1.0]non-4-yne (BCN), was covalently attached to agarose beads vi

Full text of DOI:10.1016/j.bmc.2011.07.049

Bioorganic & Medicinal Chemistry 20 (2012) 655–661
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Protein enrichment by capture–release based on strain-promoted
cycloaddition of azide with bicyclononyne (BCN)
Rinske P. Temming a, , Monique van Scherpenzeel b, , Esra te Brinke a, Sanne Schoffelen a, Jolein Gloerich b,
a,
Dirk J. Lefeber b,c, , Floris L. van Delft
⇑
⇑
a Institute for Molecules and Materials, Radboud University Nijmegen, Heijendaalseweg 135, 6525 AJ Nijmegen, The Netherlands
b Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Geert Grooteplein 10, 6525 GA Nijmegen, The Netherlands
c Department of Neurology, Radboud University Nijmegen Medical Centre, Geert Grooteplein 10, 6525 GA Nijmegen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 3 August 2011
An enrichment strategy was devised for azide derivatized macromolecules, based on strain-promoted
alkyne–azide cycloaddition (SPAAC) and a cleavable linker. A ring-strained alkyne, bicyclo[6.1.0]non-4-
yne (BCN), was covalently attached to agarose beads via a hydrazine-sensitive linker. Benchmark studies
of the resulting ‘azido-trap’ beads were performed with a fluorogenic coumarin derivative, leading to effi-
cient capture of the azidocoumarin with concomitant fluorescence staining of the beads via SPAAC. The
versatility of the beads for specific protein enrichment was shown by an effective and highly specific cap-
ture–release strategy for enrichment of azido-containing Candida antarctica lipase B (CalB) from a mix-
ture of proteins. This approach is suited for selective enrichment of (glyco)proteins after metabolic
incorporation of azides for subsequent (glyco)proteomics studies.
Keywords:
Protein enrichment
Cleavable linker
Strain-promoted cycloaddition
Bicyclononyne
Capture–release
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Similarly, efficient separation of glycan-containing macromolecules
from proteins and lipids in complex mixtures is required for their
An essential aspect in proteomics, metabolomics, glycomics and
other analytical life sciences is the identification, isolation, and
structural characterization of a specific biomolecule of interest. Sev-
eral strategies have been applied in the field of proteomics for the
global analysis of protein expression and function. Examples include
the shotgun analysis of protein expression and modification state by
liquid chromatography–mass spectrometry (LC–MS), the large-
scale mapping of protein–protein interactions by yeast two-hybrid
assays, and the proteome-wide analysis of biochemical activities
of proteins with protein microarrays. However, proteins and pep-
tides containing post-translational modifications like glycans are of-
ten difficult to detect in a background of (unmodified) biomolecules,
due to low ionization efficiencies and/or low abundance. An efficient
and specific enrichment procedure is therefore required to study de-
tailed structural aspects of a biomolecule of interest (peptide, pro-
tein, lipid or glycan) from complex mixtures like cell lysates. For
example, studying proteome dynamics by mass spectrometry re-
quires selective analysis of subsets of peptides that may be post-
translationally modified, cross-linked or newly synthesized.
efficient analysis. Recent advances have shown the possibilities of
selective labeling of glycans and proteins with azides, via metabolic
labeling withazidosugars1 like ManNAz, GalNAz and FucNAz or with
azidohomoalanine.2 Such procedures provide a selective handle to
isolate subsets of a proteome. However, mild and effective methods
for specific enrichment are still required.3
Several methodologies for protein enrichment have been devel-
oped over the years, such as activity-based protein profiling (ABPP)4
and photoaffinity labeling.5 For example, ABPP is an efficient method
for detection of the activity of proteins, rather than their mere
expression level. Similarly, photoaffinity labeling is a powerful strat-
egy to study the interactions between biologically active com-
pounds (ligands) and their target molecules. In both strategies, the
reactive probe requires the presence of an analytical handle, typi-
cally a fluorophore for visualization or an immune epitope tag for
binding and characterization. With respect to the molecular tag, bio-
tin is a popular choice because a biotinylated molecule can be easily
detected by immunological methods.6 Moreover, the presence of
biotin offers the possibility of isolation through binding to avidin
beads, although harsh conditions are subsequently required to liber-
ate the biotinylated molecule by disruption of the strong interaction
between biotin and avidin; typically boiling of the beads with deter-
gents. Clearly, such conditions may result in degradation of the tar-
get molecules as well as in contamination of other proteins that bind
to the beads in a non-specific manner.
⇑
Corresponding authors. Tel.: +31 (0)24 361 3953; fax: +31 (0)24 361 8900
(D.J.L.); tel.: +31 (0)24 365 2373; fax: +31 (0)24 365 3393 (F.L.v.D.).
E-mail addresses: D.Lefeber@neuro.umcn.nl (D.J. Lefeber), F.vanDelft@science.
ru.nl (F.L. van Delft).
These authors contributed equally to this manuscript.
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.07.049

656
R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
To overcome the drawbacks associated with biotin, a variety of
opted to chain-extend 1 with a rigid aromatic moiety to preclude a
potential ring-closure. Therefore, we performed a copper(I)-cata-
lyzed alkyne–azide cycloaddition (CuAAC)21 of 1 with Fmoc-pro-
tected propargylamine, leading to triazole adduct 2 in excellent
yield. Next, in the reported two-step process, the tert-butyl ester
was smoothly converted into the hydroxysuccinimide-activated es-
ter 4. The latter ester was efficiently coupled to BCN-amino deriva-
tive 5 via a short hydrophilic spacer. Finally, Fmoc-removal was
accomplished by treatment of 6 with 20% piperidine in dichloro-
methane, leading to the effective and clean conversion of 6 into
the desired amino derivative 7.
linkers have been developed that can be cleaved under mild condi-
tions. Examples of such cleavable linkers are constructs based on
disulfides (reductive cleavage),7 on peptide substrates (enzymati-
cally cleavable),8 on acetals (cleavable with acid),9 on diazoben-
zene (cleavable with Na2S2O4),10 hydrazone-based linkers11 and
light-sensitive linkers.12 An ideal cleavable linker is stable towards
the various conditions to which the biological sample may be ex-
posed (acidic, basic, reductive, including generally applied buffer
systems), and can withstand the reactive (nucleophilic) species
that are present in a cell extract. At the same time, the linker
should be susceptible to mild cleavage conditions, compatible with
functional groups inherent to the biological sample.
2.2. Immobilization of BCN-Lev linker to agarose beads
Having obtained the desired amino derivative 7, coupling to an
insoluble bead was investigated. For this purpose, we selected cya-
nogenbromide-activated Sepharose™ 4B from GE Healthcare, a 4%
cross-linked spherical agarose which is known to react rapidly to
amines without the need for additional reagents. The resulting car-
bamate bonds are stable and resistant to a wide range of condi-
tions, including base and high temperatures. Therefore,
compound 7 was coupled at pH 8.3 to the Sepharose™ beads at
1, 2, and 10 mM concentration, thus affording the desired ‘azido-
trap’ beads.
2. Design and synthesis of resin-bound BCN
Recently, Nessen et al. disclosed a general method to sequester
peptides from complex mixtures by means of capture with an azi-
do-reactive cleavable cyclooctyne resin.13 Thus, metabolic incorpo-
ration of azidohomoalanine in the Escherichia coli proteome
allowed the characterization of numerous newly synthesized pep-
tides via trapping by strain-promoted alkyne–azide cycloaddition
(SPAAC),14 followed by reductive cleavage of the disulfide linker
employed. Despite the elegancy of the procedure, broad applica-
tion is unlikely because synthesis of the cyclooctyne probe is rather
lengthy and its reactivity in strain-promoted cycloadditions is
modest. Furthermore, the disulfide linker necessitates the strict
avoidance of reductive conditions during the enrichment process,
in order to avoid premature cleavage, while native protein disul-
fides are inevitably also reduced during cleavage.
We have recently reported on two cyclooctyne variants, that is,
dibenzoazacyclooctyne (DIBAC)15 and bicyclononyne (BCN),16
which display high reaction rates in cycloadditions with a range
of dipoles, including azides, nitrones,17 and nitrile oxides.18
Whereas both probes are synthetically readily accessible, BCN par-
ticularly stands out because it can be prepared from cheap starting
materials in a matter of days. Furthermore, BCN is stable, fairly
non-lipophilic and symmetrical in nature, which avoids the forma-
tion of regioisomers during cycloaddition. Based on these features,
we considered BCN as a promising probe for the development of a
novel enrichment tool. With respect to the choice of a linker, we
were drawn towards a report by Geurink et al. on a cleavable linker
inspired by the levulinoyl (Lev) protective group.19 By careful fine-
tuning of the phenolic leaving group, a construct was derived with
significant advantages over earlier published strategies. The Lev-
based linker was found to be robust enough to survive conditions
commonly applied to cell extracts in biochemical experiments,
including aqueous, acidic and basic media. Importantly, the linker
was found resistant towards disulfide reducing conditions but was
readily cleaved upon treatment with hydrazine. The versatility of
the linker was nicely demonstrated in enrichment of proteasome
active sites by activity-based profiling. Based on these arguments,
we set out to prepare a novel, readily accessible and reactive probe
for enrichment by a capture–release strategy based on a Lev-based
linker and bicyclo[6.1.0]non-4-yne (BCN).
3. Protein enrichment by capture–release
As delineated above, our strategy for protein enrichment in-
volved a selective pull-down from a mixture of proteins, or ideally
whole cell lysate, followed by filtration and chemical release. How-
ever, before the suitability of the agarose-BCN conjugate for cap-
ture–release of azido-containing proteins could be performed, it
was mandatory to validate the versatility of beads for such
purpose.
3.1. Evaluation of on-resin SPAAC and cleavage
The rapid and reagent-free reaction of ring-strained alkynes
with azides allows a straightforward validation of the versatility
of the BCN-modified beads with an azidocoumarin probe. The un-
ique aspect of specific azidocoumarins is that they are fluorogenic,
that is, they become fluorescent only upon conversion of the azide
moiety into a triazole.22 Therefore, the linker-coupled beads were
incubated with 7-azidocoumarin derivative 8 (1 mM) and aliquots
were taken at specific time intervals for analysis with fluorescence
microscopy (Fig. 1A). The reaction was carried out with beads ob-
tained by loading at three different linker concentrations (1, 2, and
10 mM).
A time-dependent increase of fluorescence staining of the beads
became apparent, which strongly indicated that the SPAAC reac-
tion between immobilized BCN and azidocoumarin had taken place
(Fig. 1A). We also found that fluorescent intensities of the beads in-
creased with higher loading concentrations of the linker and thus,
the highest signal was established with beads charged at 10 mM
(see Supplementary data). For this reason, the latter beads were se-
lected for forthcoming experiments.
2.1. Attachment of BCN to cleavable linker
The fluorophore-attached beads also enable a straightforward
evaluation of the release concept by cleavage of the levulinoyl lin-
ker. To this end, the fluorescent beads were incubated with
100 mM hydrazine, leading to a rapid and irreversible decrease
in fluorescence (Fig. 1B).
In order to quantitate the loading of BCN on the Sepharose beads,
the beads were treated, after fluorophore attachment, with 100 mM
hydrazine, with the objective to directly correlate fluorescence in
the resulting supernatant to beads loading. Unfortunately, it was
Our work started from the known levulinoyl-based 2,6-di-iso-
propylphenolic ester 119,20 (Scheme 1). We reasoned that the azido
moiety of 1 could be readily reduced to an amine and therefore serve
as a useful functionality for attachment to activated agarose beads.
However, without taking appropriate precautions, an amino group
at that position rapidly cyclizes onto the d-keto function, leading
to a non-nucleophilic dihydropyrrole. Based on this rationale, we

R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
657
O
O
OtBu
OR
O
O
a
N
N
O
N3
O
N
O
O
1
FmocHN
t
2
R = Bu
b
c
3 R = H
4
R = Su
H
d
H
N
O
O
H2N
O
H
H
O
H
N
O
O
O
O
N
O
5
O
H
H
O
OH
O
N
N
O
N
O
RHN
6 R = Fmoc
7
R = H
N3
O
e
8
Scheme 1. Reagents and conditions: (a) tert-BuOH, FmocNHCH2C„CH, Cu(OAc)2, Na-ascorbate, TBAT (quant), (b) CF3CO2H, CH2Cl2 (1:1), 1 h, then (c) CH2Cl2, EDC, HOSu, 40 h
(88%, two steps), (d) 5, DIPEA, DMF, 16 h (63%), (e) piperidine, CH2Cl2 (1:4), 10 min (95%).
S1 showed only minimal degradation (data not shown). After
correction for background hydrolysis, it was determined that load-
ing of the beads with BCN amounts to 0.39
(Fig. S1C), which correlates nicely to the loading capacity as indi-
cated by the manufacturer (25–60 mg
-chymotrypsinogen mLÀ1).
l
mol mLÀ1 gel
a
Taken together, these finding established the suitability of the
BCN-modified ‘azido-trap’ beads for an enrichment protocol of azi-
do-containing molecules.
3.2. Selective enrichment of an azido-modified protein from a
protein mixture
To test whether the ‘azido-trap’ beads are useful for selective
enrichment of a specific protein of interest from a mixture of pro-
teins, we selected azido-modified Candida antartica lipase B (CalB)
as a model protein. This modified CalB, obtained by expression in
methionine auxotrophic E. coli in the presence of azidohomoala-
nine (AHA), contains five azides, of which typically only one to
two are accessible for reaction with an alkyne probe.15,23 Thus,
AHA–CalB was mixed with proteins lacking an azido function, a
mixture consisting of
a-lactalbumin, trypsin inhibitor, carbonic
anhydrase, ovalbumin, albumin, and phosphorylase B.
The mixture of proteins, including azido-CalB, was incubated
with the BCN derivatized beads at three different temperatures
and two different reaction times. Of each sample, the supernatant
from the beads after the SPAAC reaction was resolved on SDS–
PAGE (Fig. 2A). After SPAAC, the initial amount of CalB was clearly
reduced in all experiments, best conditions involving a prolonged
reaction time (18 h) at 37 °C. Under these conditions, a near quan-
titative removal of CalB from the mixture was observed, whereas
no detectable change of the other proteins became apparent.
Next, cleavage of the linker with hydrazine was performed at 4
and 37 °C for 18 h. At 4 °C, a protein with a similar mass as AHA–CalB
appeared in the gels, although at a slightly higher molecular weight.
A logical explanation for the latter observation lies in the presence of
the residual part of the linker, covalently attached to the protein. At
37 °C, the optimal temperature reported by Geurink et al., we ob-
served some protein precipitation (Fig. S2A). This indicates that, in
contrast to the capture conditions, a lower reaction temperature
(4 °C) is advantageous for the isolation of CalB from the mixture.
Nevertheless, some non-specific binding of phosphorylase B and
albumin during SPAAC and subsequent release was observed. To re-
duce the non-specific binding to the beads, we compared washing
Figure 1. (A) Treatment of BCN-modified Sepharose™ beads with the fluorogenic
substrate leads to time-dependent fluorescent staining of the beads. (B)
Cleavage of the linker on the beads using 100 mM hydrazine releases the
fluorophore from the beads.
8
a
found that the fluorophore underwent rapid degradation in 100 mM
hydrazine, leading to complete disappearance of fluorescent signal.
Therefore, a BCN derivative lacking the Lev-linker but including a
base-sensitive ester (S1, see Supplementary data) was coupled to
CNBr-Sepharose, followed by SPAAC with azido-ester-coumarin S2
under the same conditions as described for the levulinoyl linker
(10 mM). As expected, following the SPAAC reaction by fluorescence
microscopy indicated comparable fluorescent staining of the beads.
Next, loadingof the beadswas determinedby measuringthe fluores-
cent emission of the coumarin adduct (ex 325 nm, em 452 nm) from
a beads sample (25
lL), relative to a standard curve of 7-hydroxy-
coumarin in 10 mM NaOH. Under these conditions, coumarin ester

658
R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
Figure 2. Enrichment of azido-modified CalB from a protein mixture. CalB (arrows) in a mixture of standard proteins M (from top to bottom phosphorylase B, albumin,
ovalbumin, carbonic anhydrase, trypsin inhibitor and -lactalbumin) was used for SPAAC mediated enrichment. (A) Coomassie stained SDS–PAGE gel. Lanes 1–3 show the
a
supernatants after SPAAC reaction. At 37 °C for 18 h (lane 3), almost all azido-modified protein was captured. Lanes 4–6 show the supernatant of the cleavage reaction. CalB is
clearly enriched, although there is some non-specific binding as well. The cleavage product runs at a higher molecular weight than CalB, likely due to the part of the linker,
that is, covalently bound to the protein. Lane 7 contains azido-modified CalB,22 lane 8 contains protein mixture M. (B) Silver stained SDS–PAGE gel. Lanes 1 and 8 are CalB in
protein mix M before capture. Lanes 2–4 are supernatants after the SPAAC reaction. Lanes 2 and 4 are comparable. From lane 3 it is clear that a reaction time of 2 h is not
sufficient to capture all CalB. Lanes 5–7 show the supernatants after cleavage, indicating clean enrichment with only marginal non-specific binding in lane 7. Lanes 9–11 show
the supernatants of the beads treated with denaturing buffer after specific release. It shows that non-specific binding is comparable for these samples.
the beads in PBS with washing in 0.5% SDS in PBS, and washing the
beads an additional three times at 37 °C with larger volumes while
shaking for 2 h instead of just washing three times with a larger vol-
ume. It turned out that the effect of addition of SDS to the washing
buffer was negligible, while more extensive washing steps seemed
to decrease the recovery of modified CalB (Fig. S2B and C). Varying
the amount of beads per sample indicated that use of 50% of the
beads as used in Figure 2A resulted in slightly less non-specific bind-
ing, while CalB recovery remained similar (Fig. S2A). Apparently, the
lower amount of beads is sufficient to capture CalB specifically,
while higher amounts lead to more non-specific binding of other
proteins. The SDS–PAGE profile of hydrazine-released proteins un-
der optimized conditions (Fig. 2B) showed a single band at the posi-
tion of CalB without the presence of other proteins, indicating a
highly specific enrichment of modified CalB from the protein
mixture.
Finally, the specificity of the azido-trap capture was unequivo-
cally established by mass spectrometry of bead-enriched CalB.
Hydrazine-released CalB was analyzed by nanoLC–MS/MS after
tryptic digestion and compared to unbound CalB. The N-terminal
peptide, known to harbor an accessible azide moiety,23 was shown
to contain the expected triazole adduct of BCN and azide (Fig. S3B),
which was absent in the unreacted CalB sample (Fig. S3A). Other
tryptic fragments without (accessible) azidohomoalanine residues
were comparable in both CalB samples. These results prove the
mechanism for selective binding and hydrazine-mediated release
of azido-modified proteins using BCN-beads.
effectively cleaved from the beads by treatment with hydrazine. By
optimization of conditions, highly enriched CalB could be recovered
from the protein mixture in two simple and straightforward steps.
Based on these findings, it can be concluded that a versatile con-
struct has been prepared for enrichment of azido-containing mol-
ecules, including proteins. The strategy presented here has several
distinct advantages over the earlier reported capture–release ap-
proach involving strain-promoted cycloadditions. First of all, the
cyclic alkyne applied here (BCN) is more readily obtained by syn-
thesis (and commercially available) and displays much higher
reactivity to azides. Secondly, we have exploited a stable, levuli-
noyl-based cleavable linker that does not preclude reductive con-
ditions during the capture step, yet the linker is readily cleaved
upon treatment with hydrazine. Thirdly, it must be noted that
the symmetrical nature of BCN (versus all other cyclooctynes em-
ployed for SPAAC published to date) will facilitate detailed struc-
tural analysis of the enriched products after cleavage, due to the
absence of regioisomeric triazole products.
Taken together, the successful capture–release of CalB from the
protein mixture constitutes a promising general strategy for
enrichment of azido-containing (glyco)proteins from complex bio-
logical samples after metabolic incorporation.
5. Experimental section
5.1. General methods
1H NMR spectra were recorded in CDCl3 on Bruker DMX 300 or
Varian Inova-400 spectrometers at 300 K. TMS (dH 0.00) was used
as the internal reference. 13C NMR spectra were recorded in CDCl3
at 75 MHz on a Bruker DMX 300 spectrometer, using the central
resonance of CDCl3 (dC 77.0) as the internal reference. Mass spectra
of synthesized compounds were obtained on a JEOL AccuToF.
Chemicals were purchased from Aldrich and used without further
purification. CH2Cl2, acetonitrile, THF, Et2O and toluene were ob-
tained dry from a MBRAUN SPS-800 solvent purification system;
and CH3OH was distilled from magnesium and iodine. Aqueous
solutions are saturated unless otherwise specified. All reactions
4. Conclusions
We have synthesised an effective ‘azido-trap’ bead based on a
conjugate of Sepharose™ beads, a hydrazine-sensitive cleavable lin-
ker, and a strained alkyne (BCN). We demonstrated the concept by
fluorogenic staining of the beads with azidocoumarin, which dis-
played a clear time and concentration dependence. Next, we suc-
ceeded in the selective pull-down of an azido-modified protein
(AHA–CalB) from a mixture of proteins by covalent attachment to
thebeads. Finally, boththecoumarinderivativeandtheproteinwere

R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
659
were performed under anhydrous conditions under argon and
monitored by TLC on Kieselgel 60 F254 (Merck). Detection was
by examination under UV light (254 nm) and by charring with
10% sulfuric acid in methanol or with aqueous KMnO4. Silica gel
(Acros 0.035–0.070 mm) was used for chromatography. Bicy-
clo[6.1.0]non-4-yne was a kind gift from SynAffix B.V. 7-Azido-
coumarin-4-acetic acid was prepared according to a literature
procedure.24 Fluorescence microscopy was done on a Zeiss Axio-
skop 20, with an HB050 mercury lamp.
(EtOAc). IR (neat) 2958, 1735, 1524, 1442, 1364, 1208, 1161, 1143,
1070, 906, 733, 651, 616 cmÀ1 1H NMR (CDCl3, 400 MHz): d 7.75
.
(2H, d, J = 7.6 Hz), 7.57 (2H, d, J = 7.2 Hz), 7.47 (1H, s), 7.40 (2H, t,
J = 7.4 Hz), 7.30 (2H, t, J = 7.6 Hz), 6.98 (2H, s), 5.40 (1H, br s),
4.45 (2H, d, J = 6.0 Hz), 4.40 (2H, d, J = 7.2 Hz), 4.34 (2H, t,
J = 6.6 Hz), 4.21 (1H, t, J = 7.0 Hz), 3.04–3.00 (2H, m), 2.93–2.77
(11H, m), 2.51 (2H, t, J = 6.8), 2.21–2.14 (3H, m), 1.17 (12H, d,
J = 6.8). 13C NMR (CDCl3, 75 MHz): d 207.0, 171.6, 169.0, 167.9,
156.3, 145.0, 144.2, 143.8, 141.3, 140.6, 137.2, 127.7, 127.0,
125.0, 123.8, 122.1, 120.0, 66.8, 49.1, 47.2, 38.8, 37.0, 36.5, 32.7,
30.4, 27.7, 27.5, 25.6, 24.2. HRMS (ESI+) m/z calcd for C44H50N5O9
(M+H)+: 792.3609, found: 792.3655.
5.2. Synthesis of the BCN-cleavable linker conjugate
5.2.1. 4-(3-(tert-Butoxy)-3-oxopropyl)-2,6-diisopropylphenyl 7-
(4-(((((9H-fluoren-9-yl)methoxy)carbonyl)amino)methyl)-1H-
1,2,3-triazol-1-yl)-4-oxoheptanoate (2)
5.2.3. N-Fmoc-protected levulinoyl-BCN conjugate (6)
Activated ester 4 (30 mg, 37.9
(1 mL) before the addition of DIPEA (13
BCN–(POE)3–NH2 (5, 15 mg, 46.2 mole). The mixture was stirred
lmole) was dissolved in DMF
To a solution of sodium ascorbate (3.5 mg, 17.8
BuOH (1 mL) was added successively tris[(1-benzyl-1H-1,2,3-tria-
zol-4-yl)methyl]amine (TBTA, 4.7 mg, 8.9 mole), azide 1 (42 mg,
88.6 mole)
mole) and Fmoc-N-propargylamine25 (27 mg, 97.6
and the mixture was gently heated until clear. Then 250 L from
lmole) in tert-
l
L, 75.8 mole) and
l
l
l
overnight, after which time TLC analysis (EtOAc) indicated com-
plete conversion of 6 into a more polar product (RF 0.13). The mix-
ture was concentrated, applied onto a column of silica gel and
eluted with 4% MeOH in CH2Cl2. Title compound 6 was obtained
as a colorless oil. Yield 24 mg (63%). RF 0.13 (EtOAc). IR (neat)
3320, 2959, 2923, 1714, 1667, 1533, 1450, 1364, 1257, 1142,
l
l
l
a stock solution of CuSO4 (11 mg in 2.5 mL H2O) was added to
the mixture and the reaction was stirred at room temperature until
TLC analysis (EtOAc/heptane, v/v, 1:1) after 16 h indicated the
complete consumption of 1 and the formation of a more polar
product (RF 0.1). The reaction was diluted with EtOAc (20 mL)
and washed successively with H2O (4 mL) and brine (4 mL). The
organic layer was dried (MgSO4), filtered, concentrated and
purified by silica gel column chromatography (EtOAc/heptane,
v/v, 1/1 ? 2/1 ? 3/1) to give title compound 2 as a colorless oil.
Yield 67 mg (quant). RF 0.1 (EtOAc/heptane, v/v, 1:1). IR (neat)
1102, 1047, 800, 762, 741, 619 cmÀ1 1H NMR (CDCl3, 400 MHz):
.
d 7.76 (2H, d, J = 7.2 Hz), 7.57 (2H, d, J = 7.2 Hz), 7.47 (1H, s), 7.40
(2H, t, J = 7.4 Hz), 7.30 (2H, t, J = 7.2 Hz), 6.96 (2H, s), 6.02 (1H, br
s), 5.49 (1H, br s), 5.20 (1H, br s), 4.45–4.32 (6H, m), 4.22–4.12
(3H, m), 3.63–3.49 (8H, m), 3.47–3.40 (2H, m), 3.39–3.31 (2H,
m), 2.96–2.86 (6H, m) 2.80–2.77 (2H, m), 2.53–2.46 (4H, m),
2.27–2.17 (8H, m), 1.61–1.52 (2H, m), 1.36–1.31 (1H, m), 1.16
(12H, d, J = 6.8 Hz), 0.96–0.88 (2H, m). 13C NMR (CDCl3, 75 MHz):
d 207.0, 172.2, 171.7, 162.5, 156.4, 145.0, 143.8, 141.3, 140.2,
139.0, 127.7, 127.0, 125.0, 123.8, 122.1, 120.0, 98.8, 70.2, 70.1,
69.9, 66.8, 62.8, 49.1, 47.1, 40.7, 39.1, 38.7, 38.6, 36.9, 36.5, 31.7,
31.4, 29.7, 29.0, 27.7, 27.4, 24.1, 21.4, 20.1, 17.7. HRMS (ESI+) m/z
calcd for C57H72N6O10Na1 (M+Na)+: 1023.5208, found: 1023.5207.
2958, 1718, 1524, 1450, 1368, 1256, 1145, 741, 612 cmÀ1 1H
.
NMR (CDCl3, 400 MHz): d 7.75 (2H, d, J = 7.2 Hz), 7.57 (2H, d,
J = 7.6 Hz), 7.48 (s, 1H), 7.39 (2H, t, J = 7.6 Hz), 7.29 (2H, t,
J = 7.4 Hz), 6.95 (2H, s), 5.50 (1H, br s), 4.45 (2H, d, J = 6.0 Hz),
4.40 (2H, d, J = 6.8 Hz), 4.34 (2H, t, J = 6.8 Hz), 4.20 (1H, t,
J = 7.0 Hz), 2.92–2.84 (6H, m), 2.84–2.80 (2H, m), 2.55–2.49 (4H,
m), 2.21–2.14 (2H, m), 1.42 (9H, s), 1.16 (12H, d, J = 6.8 Hz). 13C
NMR (CDCl3, 75 MHz): d 207.0, 172.3, 171.6, 156.4, 145.0, 143.8,
143.7, 141.3, 140.1, 138.8, 127.7, 127.0, 125.0, 123.8, 122.1,
119.9, 80.3, 66.8, 49.1, 47.2, 38.7, 37.0, 36.9, 36.4, 31.0, 29.7,
28.1, 27.7, 27.4, 24.1. HRMS (ESI+) m/z calcd for C44H55N4O7
(M+H)+: 751.4071, found: 751.4096.
5.2.4. Amino-levulinoyl-BCN conjugate (7)
To a solution of Fmoc-protected conjugate 6 (11 mg, 10.8
lmole)
in CH2Cl2 (800 L) was added piperidine (200 L) and the solution
l
l
was stirred for 10 min. The mixture was concentrated and purified
by silica gel column chromatography (CH2Cl2/MeOH/Et3N, v/v/v,
94:5:1 ? 89:10:1) to give title compound 7 as a white solid. Yield
8 mg (95%). RF 0.1 (CH2Cl2/MeOH, v/v, 9:1). IR (neat) 3320, 2959,
2923, 1714, 1667, 1533, 1450, 1364, 1257, 1142, 1102, 1047, 800,
5.2.2. 4-(3-((2,5-Dioxopyrrolidin-1-yl)oxy)-3-oxopropyl)-2,6-di-
isopropylphenyl 7-(4-(((((9H-fluoren-9-yl)methoxy)carbonyl)-
amino)methyl)-1H-1,2,3-triazol-1-yl)-4-oxoheptanoate (4)
Trifluoroacetic acid (1 mL) was added to a solution of tert-butyl
762, 741, 619 cmÀ1 1H NMR (CDCl3, 400 MHz): d 7.76 (2H, d,
.
J = 7.2 Hz), 7.57 (2H, d, J = 7.2 Hz), 7.47 (1H, s), 7.40 (2H, t,
J = 7.4 Hz), 7.30 (2H, t, J = 7.2 Hz), 6.96 (2H, s), 6.02 (1H, br s), 5.49
(1H, br s), 5.20 (1H, br s), 4.45–4.32 (6H, m), 4.22–4.12 (3H, m),
3.63–3.49 (8H, m), 3.47–3.40 (2H, m), 3.39–3.31 (2H, m), 2.96–
2.86 (6H, m) 2.80–2.77 (2H, m), 2.53–2.46 (4H, m), 2.27–2.17 (8H,
m), 1.61–1.52 (2H, m), 1.36–1.31 (1H, m), 1.16 (12H, d, J = 6.8 Hz),
0.96–0.88 (2H, m). 13C NMR (CDCl3, 75 MHz): d 207.0, 172.2,
171.7, 162.5, 156.4, 145.0, 143.8, 141.3, 140.2, 139.0, 127.7, 127.0,
125.0, 123.8, 122.1, 120.0, 98.8, 70.2, 70.1, 69.9, 66.8, 62.8, 49.1,
47.1, 40.7, 39.1, 38.7, 38.6, 36.9, 36.5, 31.7, 31.4, 29.7, 29.0, 27.7,
27.4, 24.1, 21.4, 20.1, 17.7. HRMS (ESI+) m/z calcd for C42H63N6O8
(M+H)+: 779.47074, found: 779.47329.
ester 2 (67 mg, 88.6 lmole) in CH2Cl2 (1 mL) and the mixture was
stirred until TLC analysis (EtOAc/heptane, v/v, 3:1) showed com-
plete consumption of starting material (30 min). Toluene (3 mL)
was added and the mixture was concentrated under reduced pres-
sure, which was repeated twice to give carboxylic acid 3 as an oil
(crude). RF 0.25 (EtOAc). Without further purification, compound
3 was dissolved in CH2Cl2 (2 mL) and to this solution was added
N-hydroxysuccinimide (15 mg, 0.133 mmole) and EDC (25 mg,
0.133 mmole). The mixture was stirred overnight after which time
TLC (EtOAc) indicated the incomplete conversion of 3 into a more
lipophilic product (RF 0.35, EtOAc). To the mixture was again added
N-hydroxysuccinimide (7.5 mg, 66.5
lmole) and EDC (12.5 mg,
66.5 mole) and the mixture was stirred overnight again, leading
l
5.3. Immobilization of BCN-Lev linker to agarose beads
to the complete disappearance of starting material. The mixture
was diluted with CH2Cl2 (10 mL), and washed with 1 N HCl-solu-
tion (2 Â 1 mL) and H2O (1 mL). The organic layer was dried
(MgSO4), filtered, concentrated and purified by silica gel column
chromatography (CH2Cl2/acetone, v/v, 90/10 ? 85/15) to give title
compound 4 as a colorless oil. Yield 62 mg (88%, two steps). RF 0.35
Coupling linker-BCN conjugate 7 to agarose was performed
with Sepharose™ beads from GE Healthcare (formerly Amersham,
cat. no. 17-0430) according to the instructions of the manufacturer.
The linker-BCN conjugate 7 was coupled to the beads at 1, 2, and

660
R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
10 mM concentrations, for 16 h by end-over-end rotation. Beads
were stored in 20% ethanol at 4 °C.
was performed for 18 h at 37 °C. The beads were washed three
times with 200 L PBS and cleavage was performed for 18 h at 4 °C.
l
After running, the gels were stained with Coomassie brilliant
blue or silver stain (PlusOne/PhastGel SDS, fixing solution and sil-
ver solution 30 min instead of 10 min.
5.4. SPAAC of linker-coupled beads with 7-azidocoumarin
Bead solutions obtained above were individually incubated
with 1 mM azidocoumarin in 50% methanol (1:1, v/v) at 20 °C in
5.8. NanoLC–MS/MS analyzes of tryptic protein digests
the dark, by shaking at 400 rpm. After 0.25, 0.5, 1, and 4 h, 1
of the mixture was used for fluorescence microscopy.
lL
Proteins were in-solution digested using LysC and trypsin as de-
scribed elsewhere.26 Tryptic peptides were desalted and concen-
trated using stop and go elution (STAGE) tips according to
Rappsilber et al.27 NanoLC–MS/MS analyzes were performed using
an Easy-nLC (Proxeon) liquid chromatograph coupled online to a
LTQ FT Ultra (Thermo Fisher Scientific) mass spectrometer as de-
scribed previously.26 Chromatographic separations were per-
formed using an in-house packed reversed-phase C18 column.
Instrument settings are depicted in Supplementary Table A.
For peptide identification purposes, mass spectrometric data
files were searched using the database search program Mascot
and Mascot Distiller (Matrix Science Inc., USA, version 2.2 and ver-
sion 2.3.2.0, respectively). A combined database was used for the
searches containing the exact sequence of CalB, the E. coli Refseq
database (release 29) and sequences of known contaminants such
as human keratins, trypsin and Lys-C. Validation criteria are de-
scribed in Supplementary Table A.
5.5. Cleavage of the linker
To follow the cleavage in time, samples of 15
10 mM linker concentration and 15 L 1 mM azidocoumarin were
incubated for 2 h. Cleavage was carried out by addition of 30
lL beads of the
l
lL
200 mM hydrazine and shaking at 400 rpm at either 4 or 37 °C in
the dark. At each time point, a sample was centrifuged, and
50 lL of the supernatant was removed. The remaining 10 lL was
used for fluorescence microscopy.
5.6. Determination of the loading
BCN–(POE)3–NH2 S1 (10 mM in DMSO/H2O 1:9 v/v) was cou-
pled to CNBr-activated Sepharose beads (0.1 g freeze-dried beads,
GE Healthcare, cat. no. 17-0430-01) according to the instructions
of the manufacturer. The beads were stored in EtOH/H2O 1:4
Acknowledgments
(v/v) at 4 °C. Beads (100
lL) were incubated with azido-ester-cou-
Prof. Hermen Overkleeft, Paul Geurink and Rian van den Nie-
uwendijk (Leiden Institute of Chemistry, Leiden University, the
Netherlands) are kindly acknowledged for a generous gift of com-
pound 1. This research has been financially supported (in part) by
the Council for Chemical Sciences of the Netherlands Organization
for Scientific Research (NWO-CW, to F.L.D and NWO-Middelgroot,
to D.J.L.).
marin S2 (100 L of a 1 mM solution in DMSO/H2O 1:99 v/v) for 4 h
l
at rt, shaking in the dark at 600 rpm. Fluorescence of the beads was
checked with fluorescence microscopy (Zeiss Axioskop 20, serial
no. 451487, with an HB050 mercury lamp, using the UV filter).
Cleavage of the ester linkage was effected by incubation of
25 lL beads with 100 lL 10 mM NaOH at rt. The reaction was fol-
lowed in time until fluorescence (ex 325 nm, em 452 nm) in solu-
tion showed saturation (about 3 h). A standard curve was prepared
of 7-hydroxycoumarin in H2O, which was followed in time to cor-
rect for coumarin degradation by NaOH.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2011.07.049.
5.7. Isolation of Candida antarctica lipase B from a protein
mixture
References and notes
1. (a) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007; Agard, N. J.; Baskin, J. M.;
Prescher, J. A.; Lo, A.; Bertozzi, C. R. ACS Chem. Biol. 2006, 1, 644.
2. (a) Brustad, E. M.; Lemke, E. A.; Schultz, P. G.; Deniz, A. A. J. Am. Chem. Soc. 2008,
130, 17664; (b) Link, A. J.; Tirrell, D. A. J. Am. Chem. Soc. 2003, 125, 11164.
3. Rowland, M. M.; Best, M. D. Chem. Biol. 2010, 17, 1166.
4. (a) Evans, M. J.; Cravatt, B. F. Chem. Rev. 2006, 106, 3279; (b) Uttamchandani,
M.; Li, J.; Sun, H.; Yao, S. Q. ChemBioChem 2008, 9, 667.
5. Tomohiro, T.; Hashimoto, M.; Hatanaka, Y. Chem. Rec. 2005, 5, 385.
6. Hofmann, K.; Kiso, Y. Proc. Natl. Acad. Sci. U.S.A. 1976, 73, 3516.
7. Gartner, C. A.; Elias, J. E.; Bakalarski, C. E.; Gygi, S. P. J. Proteome Res. 2007, 6,
1482.
8. Speers, A. E.; Cravatt, B. F. J. Am. Chem. Soc. 2005, 127, 10018.
9. (a) van der Veken, P.; Dirksen, E. H. C.; Ruijter, E.; Elgersma, E. H. C.; Heck, A. J.
R.; Rijkers, D. T. S.; Slijper, M.; Liskamp, R. M. J. ChemBioChem 2005, 6, 2271; (b)
Fauq, A. H.; Kache, R.; Khan, M. A.; Vega, I. E. Bioconjugate Chem. 2006, 17, 248.
10. Verhelst, S. V.; Fonovic´, M.; Bogyo, M. Angew. Chem., Int. Ed. 2007, 46, 1284.
11. (a) Park, K. D.; Liu, R.; Kohn, H. Chem. Biol. 2009, 16, 763; (b) Dirksen, A.;
Yegneswaran, S.; Dawson, P. E. Angew. Chem., Int. Ed. 2010, 49, 2023.
12. (a) Fodor, S. P. A.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Lu, A. T.; Solas, D. Science
1991, 251, 767; (b) Holmes, C. P.; Adams, C. L.; Kochersperger, L. M.;
Mortensen, R. B.; Aldwin, L. A. Biopolymers 1995, 37, 199.
13. Nessen, M. A.; Kramer, G.-J.; Back, J.-W.; Baskin, J. M.; Smeenk, L. E. J.; de
Koning, L. J.; van Maarsseveen, J. H.; De Jon, L.; Bertozzi, C. R.; Hiemstra, H.; de
Koster, C. G. J. Proteome Res. 2009, 8, 3702.
14. Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004, 126, 15046.
15. Debets, M. F.; van Berkel, S. S.; Schoffelen, S.; Rutjes, F. P. J. T.; van Hest, J. C. M.;
van Delft, F. L. Chem. Commun. 2010, 46, 97.
For all experiments, a protein mixture of 10
CalB and 10 L 1 g/ L low molecular weight marker (LMW,
Amersham, 17-0446-01) was used. To the mixture, 10 or 20
beads of the 10 mM linker concentration was added. This SPAAC
mixture was shaken at 600 rpm for 2 or 18 h at 4, 20 or 37 °C, cen-
trifuged, and the supernatant was transferred to a new tube and
stored at 4 °C until loading it on SDS–PAGE. The beads were
lL 1 lg/lL AHA–
l
l l
lL
washed three or five times with 30 or 200
lL PBS with or without
0.5% SDS, possibly followed by three washing steps with 200
lL
PBS, 2 h at 37 °C, shaking at 600 rpm. Cleavage was carried out
for 18 h at 4 or 37 °C, shaking at 600 rpm. After cleavage, the sam-
ples were centrifuged and the supernatant was transferred to a
new tube. 15
gether with 5
l
l
L of both the SPAAC and cleavage supernatants, to-
L 4Â Laemmi sample buffer was used for SDS–PAGE
(BioRad mini Protean 3). Markers were made by using 1 or 2
CalB and/or LMW, adding MQ up to 15 L and 5
L 4Â sample buf-
fer. To the remaining beads, 10
L 4Â sample buffer was added. All
samples and markers were boiled at 95 °C for 5 min before loading
them on the gel. Of the bead samples, 10 L supernatant was
L) were loaded
lL
l
l
l
l
loaded on gel, the other samples and markers (20
l
completely.
16. Dommerholt, J.; Schmidt, S.; Temming, R. P.; Hendriks, L. J. A.; Rutjes, F. P. J. T.;
van Hest, J. C. M.; Lefeber, D. J.; Friedl, P.; van Delft, F. L. Angew. Chem., Int. Ed.
2010, 49, 9422.
Optimized conditions: to the protein mixture was added 10
lL
beads of the 10 mM linker concentration. Subsequently, SPAAC

R. P. Temming et al. / Bioorg. Med. Chem. 20 (2012) 655–661
661
17. Ning, X.; Temming, R. P.; Dommerholt, J.; Guo, J.; Ania, D. B.; Debets, M. F.;
Wolfert, M. A.; Boons, G.-J. P. H.; van Delft, F. L. Angew. Chem., Int. Ed. 2010, 49,
3065.
22. Le Droumaguet, C.; Wang, C.; Wang, Q. Chem. Soc. Rev. 2010, 39, 1233.
23. Schoffelen, S.; Lambermon, M. H. L.; van Eldijk, M. B.; van Hest, J. C. M.
Bioconjugate Chem. 2008, 19, 1127.
18. Jawalekar, A. M.; Reubsaat, E.; Rutjes, F. P. J. T.; van Delft, F. L. Chem. Commun.
2011, 47, 3198.
19. Geurink, P. P.; Florea, B. I.; Li, N.; Witte, M. D.; Verasdonck, J.; Kuo, C.-L.; van der
Marel, G. A.; Overkleeft, H. S. Angew. Chem., Int. Ed. 2010, 49, 6954.
20. Compound 1 was a kind gift from prof. Hermen Overkleeft, Leiden Institute of
Chemistry, Leiden University, the Netherlands.
24. Pianowski, Z. L.; Winssinger, N. Chem. Commun. 2007, 3820.
25. Seth Horne, W.; David Stout, C.; Reza Ghadiri, M. J. Am. Chem. Soc. 2003, 125,
9372.
26. Wessels, H. J.; Gloerich, J.; Van der Biezen, B. E.; Jetten, M. S.; Kartal, B. Methods
Enzymol. 2011, 486, 465.
27. Rappsilber, J.; Ishihama, Y.; Mann, M. Anal. Chem. 2003, 75, 663.
21. (a) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057; (b)
Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed.
2002, 41, 2596.