Click and Release Chemistry for Activity-Based Purification of β-Lactam Targets

Article abstract of DOI:10.1002/chem.202100653

β-Lactams, the cornerstone of antibiotherapy, inhibit multiple and partially redundant targets referred to as transpeptidases or penicillin-binding proteins. These enzymes catalyze the essential cross-linking step of the polymerization of cell wall peptidoglycan. The understanding of the mechanisms of action of β-lactams and of resistance to these drugs requires the development of reliable methods to characterize their targets. Here, we describe an activity-based purification method of β-lactam targets based on click and release chemistry. We synthesized alkyne-carbapenems with suitable properties with respect to the kinetics of acylation of a model target, the Ldt_fm L,D-transpeptidase, the stability of the resulting acylenzyme, and the reactivity of the alkyne for the cycloaddition of an azido probe containing a biotin moiety for affinity purification and a bioorthogonal cleavable linker. The probe provided access to the fluorescent target in a single click and release step.

Full text of DOI:10.1002/chem.202100653

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Click and Release Chemistry for Activity-Based Purification
of β-Lactam Targets
Saidbakhrom Saidjalolov,[a] Emmanuelle Braud,[a] Zainab Edoo,[b] Laura Iannazzo,[a]
Filippo Rusconi,[c] Margaux Riomet,[d] Antoine Sallustrau,[d] Frédéric Taran,[d] Michel Arthur,*[b]
Matthieu Fonvielle,*[b] and Mélanie Etheve-Quelquejeu*[a]
Abstract: β-Lactams, the cornerstone of antibiotherapy,
inhibit multiple and partially redundant targets referred to as
transpeptidases or penicillin-binding proteins. These enzymes
catalyze the essential cross-linking step of the polymerization
of cell wall peptidoglycan. The understanding of the mecha-
nisms of action of β-lactams and of resistance to these drugs
requires the development of reliable methods to characterize
their targets. Here, we describe an activity-based purification
method of β-lactam targets based on click and release
chemistry. We synthesized alkyne-carbapenems with suitable
properties with respect to the kinetics of acylation of a model
target, the Ldtfm L,D-transpeptidase, the stability of the
resulting acylenzyme, and the reactivity of the alkyne for the
cycloaddition of an azido probe containing a biotin moiety
for affinity purification and a bioorthogonal cleavable linker.
The probe provided access to the fluorescent target in a
single click and release step.
Introduction
polymer is essential for sustaining the osmotic pressure of the
bacterial cytoplasm. It consists of glycan chains linked to each
other by short stem peptides (Figure 1A). The cross-linking
reaction catalyzed by HMW PBPs, the D,D-transpeptidases,
proceeds by the cleavage of the C-terminal D-Ala-D-Ala peptide
bond of a stem pentapeptide and linkage of the penultimate D-
Ala to the third residue of a stem peptide carried by an adjacent
glycan chain (4!3 cross-links) (Figure 1B). β-lactams are
structure analogues of the D-Ala-D-Ala extremity of peptidogly-
can precursors and act as suicide substrates of the HMW PBPs[3]
(Figure 1C). Acylation of the active site Ser of the PBPs results in
a long-lasting covalent adduct and irreversible inactivation of
the target at the time scale of a bacterial generation.
Bacterial resistance is a major threat to the efficacy of antibiotics
requiring innovative strategies for maintaining the efficacy of
antimicrobial chemotherapy. Identification and characterization
of the biological targets of drugs is a key step for their
development and the study of the mechanisms of resistance.
This is particularly the case for antibiotics belonging to the β-
lactam family since these drugs have multiple targets in
bacterial cells with partially overlapping functions.[1,2] The
essential targets of β-lactams are the high-molecular weight
(HMW) penicillin-binding proteins (PBPs) that catalyze the last
cross-linking step of peptidoglycan synthesis. This cell wall
The cross-linking reaction is performed by multiple HMW
PBPs in most bacteria, typically 5 to 6 in the main human
pathogens.[1] In addition, β-lactams similarly inactivate low-
molecular weight (LMW) PBPs that hydrolyze the D-Ala-D-Ala
peptide bond of stem pentapeptides (D,D-carboxypeptidases)
or the cross-links formed by the HMW PBPs (endopeptidases). A
particular class of β-lactams, the carbapenems, also inactivate
L,D-transpeptidases (Ldts) that form unusual cross-links con-
necting the third position on both stem peptides (3!3 cross-
links)[4] (Figure 1B). Ldt inactivation results from acylation of
their active-site Cys residue by the β-lactam ring of carbape-
nems (Figure 1C). Ldts are the main cross-linking enzyme in
Mycobacterium tuberculosis[5] and can be responsible for in vitro
acquired resistance to β-lactams of the penam (penicillin) and
cephem (cephalosporin) classes in other bacteria.[6,7]
[a] S. Saidjalolov, Dr. E. Braud, Dr. L. Iannazzo, Prof. Dr. M. Etheve-Quelquejeu
Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques,
UMR 8601 CNRS
Université de Paris,
45, rue des saints-pères, Paris, 75006 (France)
E-mail: melanie.etheve-quelquejeu@parisdescartes.fr
[b] Dr. Z. Edoo, Dr. M. Arthur, Dr. M. Fonvielle
INSERM UMRS 1138
Sorbonne Universités, UPMC Univ Paris 06, Sorbonne Paris Cité, Université
de Paris
Centre de recherche des Cordeliers, Paris, 75006 (France)
E-mail: michel.arthur@crc.jussieu.fr
matthieu.fonvielle@crc.jussieu.fr
[c] Dr. F. Rusconi
PAPPSO,
Université Paris-Saclay, INRAE, CNRS, AgroParisTech
GQE - Le Moulon, Gif-sur-Yvette, 91190 (France)
Historically, the targets of β-lactams have been identified by
incubating membrane preparations with radiolabeled drugs
followed by denaturing polyacrylamide gel electrophoresis
(SDS-PAGE) and autoradiography.[8] This method typically
resolves a pattern of ca. 10 bands corresponding to proteins
covalently labeled by the drug. Competitive binding with
[d] Dr. M. Riomet, Dr. A. Sallustrau, Dr. F. Taran
Université Paris Saclay, CEA, INRAE
Département Médicaments et Technologies pour la Santé (DMTS), SCBM
Gif-sur-Yvette, 91191 (France)
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/chem.202100653
Chem. Eur. J. 2021, 27, 1–10
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Figure 1. (a) Peptidoglycan structure. (b) Formation of peptidoglycan cross-links by PBPs and Ldts. (c) Inhibition of Ldts by carbapenems. The efficacy of Ldt
inhibition depends upon slow hydrolysis of the acylenzyme. The inset shows the structure of carbapenems discussed in the text.
unlabeled drugs has been largely used to compare the
reactivity of various β-lactams for the PBPs. For this approach,
fluorescent β-lactams, such as bocillin, were developed to avoid
the use of radioactive reagents.[9–11] Further refinement of the
PBP labeling methods afforded single cell imaging of bacteria
exposed to the drugs.[12] The key element of this application is
the incorporation of a bioorthogonal active group in the β-
lactam with minimal impact on drug penetration and on
interaction of the drug with its multiple targets. Following
labeling of the targets in the bacterium, the drug-target
adducts were revealed by a bioorthogonal reaction with a
functionalized dye.[13,14] The most popular approach is the
incorporation of an alkyne group on the drug allowing copper-
catalyzed cycloaddition (CuAAC) with an azido-containing
dye.[13,14] This approach enables detection of known PBPs but
present limitations concerning the identification of unknown
targets since it only provides information on the electrophoretic
mobility of the protein-dye adduct. This limitation has been
addressed by the use of biotinylated azido-containing dyes for
streptavidin affinity pull-down purification followed by mass
spectrometry analysis.[15,16] Selective penicillin-binding protein
imaging probes have also been described using fluorescent β-
lactams,[17] such as fluorescent cephalosporins.[18] Of note,
fluorescent β-lactams were also used as fluorogenic molecular
probes for the detection of β-lactamases in vitro or in living
bacteria.[17,19,20]
In this study, we explore a strategy for combined affinity
purification and fluorescence labeling of carbapenem targets.
We report synthetic routes to alkynylated carbapenems that
effectively labeled a model L,D-transpeptidase (Ldtfm) both with
respect to the kinetics of acylation and the stability of the
resulting acylenzymes. The alkyne group of the enzyme-bound
carbapenem was subsequently exploited for CuAAC attachment
of a biotin-containing iminosydnone linker, recently developed
as a bioorthogonal cleavable spacer (Figure 2).[21,22] After bind-
ing to streptavidin magnetic beads and purification, the
iminosydnone mesoionic compound enabled the release and
labeling of the captured carbapenem targets from the strepta-
vidin support through a bioorthogonal click and release
reaction,[21] affording pure and fluorescent targets in a single
step.
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Figure 2. General strategy for activity-based purification of carbapenem targets.
Results and Discussion
Synthesis of alkynylated carbapenems
We started our synthesis by using a post-functionalization
strategy of the fully deprotected methyl-thienamycin (Route A,
Scheme 1). Methyl-thienamycin was obtained in two steps
(21% yield) from commercially available β-methyl vinyl-
phosphate.[23] Methyl-thienamycin is renowned for its unstable
chemical nature.[24] Indeed, strong acidic, basic, or nucleophilic
conditions as well as exposure to light, high temperature, or
prolonged storage in solution lead to β-lactam ring opening.
Thus, the reaction conditions for functionalizing methyl-
°
thienamycin should be mild (6.0<pH<8.5; <25 C) and should
not involve prolonged reaction time (<1 hour). Despite these
limitations, three alkyne-carbapenems were obtained, CBA-1,
CBA-2, and CBA-3, using methyl-thienamycin in the presence of
alkyne linkers, which were commercially available (for CBA-1
and CBA-2) or synthesized in this study (Compound 1,
Scheme 1). Alkyne-carbapenems CBA-1, CBA-2, and CBA-3 were
obtained with variable yields (49%, 67%, and 14%, respectively;
Scheme 1, Route A).
Figure 3. (a) Acylation of Ldtfm by carbapenems. (b) Stability of Ldtfm
carbapenem adducts determined by mass spectrometry (Supporting
Information, Section 7.3).
-
Ring opening could not be fully prevented and subsequent
purification of the desired compounds free of their hydrolyzed
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Scheme 1. Two strategies for alkyne carbapenems synthesis. Reaction conditions for route A: i. 4-pentynoic acid succinimidyl ester or N-(prop-2-yn-1-yl)-1H-
imidazole-1-carboxamide[26] or compound 1. Reaction condition for route B: i. thiol alkyne-linkers 3, 6, 3-butynyl ethanethioate or 10 and DIPEA in DMF, ii. Zn
dust in THF:PB (1:2). Compounds were obtained by route A (CBA- 3), route B (CBA-4 and CBA-5), or both routes (CBA-1 and CBA-2).
products was cumbersome. For this reason, we explored an
alternative approach starting from a protected carbapenem for
direct addition of a thiol-alkyne onto the β-methyl vinyl-
phosphate (Scheme 1, route B). Syntheses of the thiol alkyne
linkers 3, 6, and 10 are described in the supporting information
(Section 2). This first step provided the intermediate carbape-
nems 11a, b, c, d in 97%, 77%, 78%[23] and 54% yields,
respectively. The critical step of the latter strategy was the final
selective deprotection of the para-nitrobenzyl group (PNB),
which was successfully achieved using zinc dust.[25] Minimal β-
lactam ring opening provided by this approach facilitated
purification of the final compounds (CBA-1, CBA-2, CBA-4, and
CBA-5 in 11%, 77%, 40%, and 43% yields, respectively).
Comparison of the NMR spectra of CBA-1 and CBA-2 obtained
by routes A and B (Supporting Scheme S5) supported the
superiority of route B with respect to the purity of the final
compound.
Acylation of L,D-transpeptidase Ldtfm by alkyne-carbapenems
The alkyne function of compounds CBA-1 to CBA-5 is
connected to the C2 position of the carbapenem core via
variable links, including an amide (CBA-1), a urea (CBA-2), a
squaramide (CBA-3), an alkyl (CBA-4), or a benzamide (CBA-5).
These compounds were synthesized to explore the impact of
the linkers on the kinetics of target acylation and the stability of
the resulting covalent adduct. This was first investigated by
mass spectrometry to detect Ldtfm-carbapenem adducts (Fig-
ure 3). Incubation of Ldtfm (6.5 μM) with alkyne carbapenems
°
CBA-1 to CBA-5 (39 μM) for 15 min at 25 C in 20 mM
ammonium acetate (pH 7.2) led to full enzyme acylation, except
for CBA-3 (ca. 30% acylation). Stability of the Ldtfm-carbapenem
adducts was also monitored by mass spectrometry.
For this purpose, the Ldtfm-carbapenem adducts were
purified by gel filtration to remove the excess of unreacted
°
carbapenem and incubated for 1 h and 3 h at 25 C. This
analysis showed that acylenzymes resulting from acylation of
Ldtfm by CBA-2, CBA-4, CBA-5, and the clinically-used carbape-
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Figure 4. Impact of the linker of alkyne carbapenems on the efficiency of CuAAC. (a) CuAAC reaction with 3-azido-7-hydroxycoumarin. CuAAC performed with
CBA-1, CBA-2, CBA-4, and CBA-5 provided compounds 13a, b, c, and d, respectively. (b) Upper panel: Fluorescence kinetics obtained with alkyne
carbapenems CBA-1, CBA-2, CBA-4, and CBA-5. Lower panel: Calibration curves obtained with purified triazole compounds, 13a to 13d. Synthesis and
characterization of these triazole compounds are reported in the Supporting Information, Section 4. Please note that the intrinsic fluorescence of 13d was
higher than that of 13b. For this reason, the efficiency of the CuAAC reaction was higher for CBA-2 than for CBA-5 although the rate of increase in the
fluorescence intensity was the highest for CBA-5. The efficiency of the reaction was evaluated based on the initial rate of the reaction, 2.00�0.02 μMminÀ 1
versus 0.90�0.03 μM min for CBA-2 and CBA-5, respectively. (c) The yield (%) of the reaction was evaluated based on the extent of the disappearance of the
alkynes after 30 min of incubation (determined by HPLC analysis), as exemplified for CBA-2 (see Supporting Figure S1, S2, and S3 for CBA-1, CBA-4, and CBA-
5). (d) Efficiency parameters. [i] Initial rate of CuAAC determined by fluorimetry (panel b). [ii] Conversion (%) was determined by HPLC (panel c).The conversion
was higher for CBA-2 (85%) than for CBA-5 (59%), CBA-1 (21%), and CBA-4 (21%).
nem meropenem (Figure 1C) were stable (Figure 3). In addition,
native Ldtfm resulting from hydrolysis of the carbapenem-Ldtfm
thioester was not detected in the mass spectra. In contrast,
partial hydrolysis ca. 20% was detected for CBA-1 since the
ratio of peak intensities was in the order of 0.2 and 0.8 for
native Ldtfm and the Ldtfm-CBA-1 acylenzyme, respectively. Since
Ldtfm was only partially acylated by CBA-3 this alkyne
carbapenem was not selected for further studies.
Reactivity of alkyne-carbapenems in the CuI-catalyzed
cycloaddition
The success of our click and release strategy relies on both the
stability of the acylenzyme (above) and on the efficiency of the
cycloaddition reaction (step 2; Figure 2). Kinetics of CuAAC were
performed with
a fluorogenic azido partner (3-azido-7-
hydroxycoumarin)[27,28] to compare the impact of the linkers
present in the four alkyne carbapenems (Figure 4).
The rate of the cycloaddition reaction in the presence of
CuSO4 (at 25 μM), was higher for CBA-2 (2.00�0.02 μMminÀ 1)
than for CBA-1, CBA-4, and CBA-5 (0.56 to 0.90 μMminÀ 1)
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Figure 5. Kinetic parameters for Ldtfm acylation by carbapenems. R2 =H for CBA-2 and a diazirine moiety for CBA-6. (a) Reaction scheme for the two-step
acylation reaction. R2 =H for CBA-2 and a diazirine moiety for CBA-6. (b) Fluorescent kinetics obtained with three concentrations of carbapenem. (c) Kinetic
parameters for the acylation reaction. (d) Relative fluorescence of the three forms of Ldtfm. Biphasic fluorescence kinetics are accounted for by the low
fluorescence intensity of the amine anion intermediate.[29,30]
(Figure 4D). To evaluate the yield of the reaction for each alkyne
carbapenem, 3-azido-7-hydroxycoumarin was allowed to react
with alkyne carbapenems for 30 min and the progress of the
reaction was followed by HPLC. The yield was the highest for
CBA-2 (85%), intermediate for CBA-5 (59%), and the lowest for
CBA-1 and CBA-4 (21%). The high reactivity of CBA-2 might be
accounted for by the chelating effect of the urea linker on the
copper catalyst as observed for other chelating functions.[28]
This alkyne carbapenem was selected for further experiments as
the best candidate based on stability of the acylenzyme
(Figure 3) and its efficiency in the cycloaddition reaction (Fig-
ure 4).
CBA-2 proved to be superior to meropenem both with respect
to the rate constant k1 for formation of the amine anion
(0.426�0.009 versus 0.168�0.001 μMÀ 1 minÀ 1) and the rate
constant k2 for formation of the acylenzyme from this
intermediate (3.30�0.05 versus 2.24�0.03 minÀ 1) (Figure 5). In
agreement with mass spectrometry analysis (Figure 3), hydrol-
ysis kinetics obtained by spectrophotometry revealed that the
acylenzymes formed with CBA-2 and meropenem display a
similar stability (half-lives 323�2 and 504�4 min, respectively).
These values are higher than expected from the mass
spectrometry analysis reported in Figure 3, which revealed that
Ldtfm remained fully acylated after 3 h of incubation. This
discrepancy may indicate that the free unreacted carbapenems
were not fully eliminated by the gel filtration step used in the
mass spectrometry analysis (Figure 3) enabling re-acylation after
acylenzyme hydrolysis. These results indicate that we have
successfully functionalized the carbapenem core without
impairing the efficiency of Ldtfm acylation or the stability of the
resulting adduct.
Kinetics of Ldtfm acylation by alkyne carbapenem CBA-2
Our next objective was to evaluate the impact of the
incorporation of a urea-linked alkyne function in CBA-2 on the
inactivation of Ldtfm by this carbapenem.
Kinetics based on variation in the intrinsic fluorescence of
tryptophan residues of Ldtfm upon acylation were monitored for
the reference carbapenem meropenem and CBA-2 (Figure 5).
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
Assessing the specificity of the capture and release reactions
As a negative control, we used FemX, a tRNA-dependent
aminoacyl transferase that does not interact with carbapenems.
This protein was not recovered from streptavidin beads
indicating that the procedure was specific to the Ldtfm
carbapenem-target (Figure 6B). The procedure was also specific
in recovering Ldtfm from a reconstituted mixture containing a
whole protein extract from E. faecium (18.6 μg of protein) and
0.65 nmol of Ldtfm (18.8 μg) (Figure S5 in Supporting informa-
tion). Specificity was further evaluated by preincubating Ldtfm
with meropenem to block the active-site Cys and prevent
acylation by the CBA-2 alkyne carbapenem. This pretreatment
almost completely abolished recovery of fluorescent Ldtfm from
streptavidin beads (Figure 6C). In addition, we showed that
recovery of TAMRA-labeled Ldtfm required the native carbape-
nem core of CBA-2 since preincubation of this compound with
carbapenemase KPC-2, a hydrolytic enzyme of the β-lactamase
family, abolished detection of the fluorescent protein band at
35 kDa (Figure 6D). Recovery of the TAMRA-labeled-Ldtfm was
also dependent upon acylation of the active-site Cys of the
enzyme since no fluorescent band was observed for a Ldtfm
derivative devoid of the Cys nucleophile following substitution
of this residue by Ala.
Figure 6. SDS-PAGE analyses of CBA-2-Ldtfm adducts. (a) Click and release
reactions applied to the Ldtfm L,D-transpeptidase. Lane 1, 2 μg of native
Ldtfm; Lane 2, protein recovered from streptavidin beads. Proteins were
detected by fluorescence (top panel) and Coomassie blue staining (bottom
panel). (b) Assessing specificity by applying the click and release reactions to
a mixture of equimolar amounts of Ldtfm and FemX (Lane 1) and to FemX
alone (Lane 3). Lane 2, native Ldtfm and FemX proteins. (c) Assessing
specificity by applying the click and release reactions to Ldtfm (Lane 2) and to
Ldtfm preincubated with meropenem (Lane 3) to prevent acylation of the
protein by the CBA-2 alkyne carbapenem. Lane 1, native Ldtfm. (d) Assessing
specificity by applying the click and release reactions to native Ldtfm (Lane
1), to CBA-2 preincubated with carbapenemase KPC-2 (Lane 2) and to a
derivative of Ldtfm that cannot be acylated following replacement of the
catalytic cysteine by alanine (Lane 4). Lane 3, click and release reactions
applied to the Ldtfm L,D-transpeptidase (as performed for panel A).
Bi-functionalization of the carbapenem scaffold with
photoactivable and alkyne groups to improve acylenzyme
stability
One possible limitation of the click and release procedure is the
instability of the link connecting the drug to the target, a
thioester bond in the case of Ldtfm-carbapenem adducts.
Indeed, the link is exposed to multiple conditions during
the four steps of the click and release procedure (Figure 2),
which lasts for ca. 3 hours in its current version. Thus, our next
objective was to add a photoactivable group to the alkyne
carbapenem in order to stabilize the drug-target adduct (Fig-
ure 7). Diazirines have been successfully used for photoaffinity
labeling[31] and for studies aimed at detecting a variety of
interactions, such as ligand-receptor,[32] inhibitor-enzyme,[33] and
protein-protein,[34] or for fishing of drug targets.[35,36] The
combined advantages of diazirines over other photoactivable
cross-linking agents include their small size, long irradiation
wavelength, short period of irradiation, and high stability in the
presence of various nucleophiles.[33,37] Upon irradiation in the
330 to 380 nm range, diazirines form reactive carbenes, which
can insert into CÀ H, NÀ H, and OÀ H bonds.[38]
Capture of Ldtfm and release of the fluorescent protein
adduct
CBA-2 was used to explore the efficacy of step 3 (capture) and
step 4 (release) of our click and release strategy (Figure 2). Ldtfm
(0.65 nmol) was acylated with 6 equivalents of CBA-2 (Figure 2,
step 1). Excess alkyne-carbapenem was removed by gel
filtration (G25 microspin). The CuAAC reaction was performed
with 10.8 equivalents of the azido-platform[21,22] leading to the
CBA-2-Ldtfm adduct (Figure 2, step 2). The platform provided a
biotin handle for capture of the Ldtfm-CBA-2 adduct by
streptavidin-coated magnetic beads (Figure 2, step 3). Following
extensive washing, the click and release reaction was performed
with 60 equivalents of the TAMRA fluorescent cyclooctyne
(Figure 2, step 4). The latter reaction enabled elution and
labeling of the Ldtfm adduct in a single step. SDS-PAGE analysis
(Figure 6A) showed that TAMRA-labeled Ldtfm was successfully
eluted from the streptavidin beads.
In our approach, synthesis of the bi-functionalized carbape-
nem started from the protected methyl-thienamycin,[23] which
reacted with activated 3-(3-methyl-3H-diazirin-3-yl)propanoic
acid[39] to give compound 14 in 92% yield (Figure 7A). Removal
of the PNB group using hydrogenolysis catalyzed by platinum
on carbon, followed by the functionalization of compound 15
by N-(prop-2-yn-1-yl)-1H-imidazole-1-carboxamide, afforded
compound CBA-6.
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
°
Figure 7. Synthesis and use of photoactivable carbapenem CBA-6. (a) Synthesis of CBA-6. i) 3-(3-methyl-3H-diazirin-3-yl)propanoic acid, DMAP, EDC, À 20 C,
°
92%. ii) H2, 10% wt. Pt/C, THF/Buffer, 1 h, 25%. iii) N-(prop-2-yn-1-yl)-1H-imidazole-1-carboxamide, DMF/H2O, NaHCO3, 1 h, 25 C, 63%. (b) Acylation of Ldtfm
by CBA-6 and photoactivation leading to formation of a second covalent bond stabilizing the adduct. The Ldtfm-CBA-6 interaction remains covalent even after
hydrolysis of the Cys-CBA-6 thioester bond. (c) SDS-PAGE of the click and release procedure applied to the Ldtfm-CBA-6 acylenzyme.
Capture and release of labeled Ldtfm was performed as
described above (Figure 2) with or without an additional
photoactivation step between steps 1 and 2. Mass spectrometry
analysis showed formation of the additional cross-link upon
irradiation that resulted in the expected loss of two nitrogen
atoms (À 28.0 Da, see Table S1). CBA-6 was functional in the
click and release procedure (step 1 to 4 in Figure 2) but the
irradiation did not improve recovery of labeled Ldtfm (Figure 7C).
This result suggests that hydrolysis of the thioester bond
connecting CBA-6 to Ldtfm is not limiting to the recovery of
fluorescent Ldtfm after the final release step.
of the so-called “low-affinity” PBPs responsible for β-lactam
resistance. In addition, the complexity of these combinations
includes evolutionary unrelated enzymes,[5,41–43] including un-
identified ones,[44] in particular in M. tuberculosis.[42,45] This
context prompted us to develop here a new strategy for affinity
purification of β-lactam targets. We functionalized the carbape-
nem core for affinity purification of β-lactam targets using a
biotin-containing iminosydnone probe. First, we introduced at
the carbapenem C2 position an alkyne function connected by a
series of five linkers. A urea linker was identified as the best
candidate because (i) it did not impair the kinetics of target
acylation and the stability of the resulting acylenzyme and (ii) it
displayed the highest reactivity for copper-catalyzed cyclo-
addition of a fluorogenic probe. Second, we showed that the
reaction of biotinylated Ldtfm adducts captured on streptavidin
magnetic beads with a TAMRA-cyclooctyne enables release and
fluorescence labeling of Ldtfm in a single step. Third, we
synthesized bi-functionalized carbapenems containing both an
alkyne and a photoactivable diazirine to introduce a second
stabilizing covalent bond between Ldtfm and the carbapenem.
The strategy was compatible with mass-spectrometry detection
of protein adducts in addition to fluorescent labeling. These
tools should facilitate the characterization of the complex set of
carbapenem targets, including the detection of unknown
enzymes. Access to purification strategies based on acylation of
active-site nucleophiles and photoactivation, alone or in
combination offers specific advantages. Photoactivation pro-
Conclusion
The targets of β-lactams are highly diverse and complex since
they include both essential biosynthetic enzymes, the trans-
peptidases, as well as unessential hydrolases that participate in
peptidoglycan metabolism.[40] Assessing the essential nature of
these transpeptidases is critical for drug development since it
determines the killing activity of the drug. This task remains
highly challenging despite decades of analyses of the mode of
action and of the mechanisms of resistance to β-lactams. Since
the enzymes have partially redundant functions, essentiality of
the targets is defined by combinations of proteins rather than a
single enzyme complement. These combinations differ both
between and within bacterial species following the acquisition
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
8
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202100653
Chemistry—A European Journal
[20] J. Wang, W. Xu, S. Xue, T. Yu, H. Xie, Org. Biomol. Chem. 2020, 18, 4029–
4033.
[21] S. Bernard, D. Audisio, M. Riomet, S. Bregant, A. Sallustrau, L. Plougastel,
E. Decuypere, S. Gabillet, R. A. Kumar, J. Elyian, M. N. Trinh, O. Koniev, A.
Wagner, S. Kolodych, F. Taran, Angew. Chem. Int. Ed. 2017, 56, 15612–
15616; Angew. Chem. 2017, 129, 15818–15822.
vides very stable links but may decrease the specificity.
Acylation of the nucleophile active-site is highly specific but the
half-life of the adduct may compromise purification of certain
targets.
[22] E. Decuypere, L. Plougastel, D. Audisio, F. Taran, Chem. Commun. 2017,
53, 11515–11527.
[23] L. Iannazzo, D. Soroka, S. Triboulet, M. Fonvielle, F. Compain, V. Dubee,
J. L. Mainardi, J. E. Hugonnet, E. Braud, M. Arthur, M. Etheve-Quelquejeu,
J. Med. Chem. 2016, 59, 3427–3438.
[24] K. M. Papp-Wallace, A. Endimiani, M. A. Taracila, R. A. Bonomo, Anti-
microb. Agents Chemother. 2011, 55, 4943–4960.
[25] T. Kumagai, T. Abe, Y. Fujimoto, T. Hayashi, Y. Inoue, Y. Nagao,
Heterocycles 1993, 36, 1729–1734.
Acknowledgements
The authors thank the “Ministère de l’Enseignement supérieur,
de la Recherche et de l’Innovation” for PhD grant to S.S.
[26] C. D. Rillahan, E. Schwartz, C. Rademacher, R. McBride, J. Rangarajan,
V. V. Fokin, J. C. Paulson, ACS Chem. Biol. 2013, 8, 1417–1422.
[27] A. A. Kislukhin, V. P. Hong, K. E. Breitenkamp, M. G. Finn, Bioconjugate
Chem. 2013, 24, 684–689.
[28] V. Bevilacqua, M. King, M. Chaumontet, M. Nothisen, S. Gabillet, D.
Buisson, C. Puente, A. Wagner, F. Taran, Angew. Chem. Int. Ed. 2014, 53,
5872–5876; Angew. Chem. 2014,126, 5982–598.
Conflict of Interest
The authors declare no conflict of interest.
Keywords: Activity-based probe · Antibiotic · Click and release ·
Iminosydnone · Peptidoglycan
[29] N. Bhattacharjee, S. Triboulet, V. Dubee, M. Fonvielle, Z. Edoo, J. E.
Hugonnet, M. Etheve-Quelquejeu, J. P. Simorre, M. J. Field, M. Arthur,
C. M. Bougault, Antimicrob. Agents Chemother. 2019, 63, e02039–18.
[30] S. Triboulet, Z. Edoo, F. Compain, C. Ourghanlian, A. Dupuis, V. Dubee,
L. Sutterlin, H. Atze, M. Etheve-Quelquejeu, J. E. Hugonnet, M. Arthur,
ACS Infect. Dis. 2019, 5, 1169–1176.
[31] A. Blencowe, W. Hayes, Soft Matter 2005, 1, 178–205.
[32] L. Dubinsky, L. M. Jarosz, N. Amara, P. Krief, V. V. Kravchenko, B. P. Krom,
M. M. Meijler, Chem. Commun. 2009, 47, 7378–7380.
[33] L. Dubinsky, B. P. Krom, M. M. Meijler, Bioorg. Med. Chem. 2012, 20, 554–
570.
[34] T. Durek, J. Zhang, C. He, S. B. Kent, Org. Lett. 2007, 9, 5497–5500.
[35] J. R. Hill, A. A. B. Robertson, J. Med. Chem. 2018, 61, 6945–6963.
[36] J. J. Hulce, A. B. Cognetta, M. J. Niphakis, S. E. Tully, B. F. Cravatt, Nat.
Methods 2013, 10, 259–264.
[37] Y. Hatanaka, Y. Sadakane, Curr. Top. Med. Chem. 2002, 2, 271–288.
[38] J. Brunner, Annu. Rev. Biochem. 1993, 62, 483–514.
[39] A. M. Ahad, S. M. Jensen, J. C. Jewett, Org. Lett. 2013, 15, 5060–5063.
[40] A. Zapun, C. Contreras-Martel, T. Vernet, FEMS Microbiol. Rev. 2008, 32,
361–385.
[41] J. L. Mainardi, J. E. Hugonnet, F. Rusconi, M. Fourgeaud, L. Dubost, A. N.
Moumi, V. Delfosse, C. Mayer, L. Gutmann, L. B. Rice, M. Arthur, J. Biol.
Chem. 2007, 282, 30414–30422.
[42] R. Gupta, M. Lavollay, J. L. Mainardi, M. Arthur, W. R. Bishai, G.
Lamichhane, Nat. Med. 2010, 16, 466–469.
[43] S. Magnet, S. Bellais, L. Dubost, M. Fourgeaud, J. L. Mainardi, S. Petit-
Frere, A. Marie, D. Mengin-Lecreulx, M. Arthur, L. Gutmann, J. Bacteriol.
2007, 189, 3927–3931.
[1] A. Zapun, C. Contreras-Martel, T. Vernet, FEMS Microbiol. Rev. 2008, 32,
361–385.
[2] B. Gold, R. Smith, Q. Nguyen, J. Roberts, Y. Ling, L. Lopez Quezada, S.
Somersan, T. Warrier, D. Little, M. Pingle, D. Zhang, E. Ballinger, M.
Zimmerman, V. Dartois, P. Hanson, L. A. Mitscher, P. Porubsky, S. Rogers,
F. J. Schoenen, C. Nathan, J. Aube, J. Med. Chem. 2016, 59, 6027–6044.
[3] D. J. Tipper, J. L. Strominger, Prot. Natl. Acad. Sci. USA 1965, 54, 1133.
[4] J. L. Mainardi, J. E. Hugonnet, F. Rusconi, M. Fourgeaud, L. Dubost, A. N.
Moumi, V. Delfosse, C. Mayer, L. Gutmann, L. B. Rice, M. Arthur, J. Biol.
Chem. 2007, 282, 30414–30422.
[5] M. Lavollay, M. Arthur, M. Fourgeaud, L. Dubost, A. Marie, N. Veziris, D.
Blanot, L. Gutmann, J. L. Mainardi, J. Bacteriol. 2008, 190, 4360–4366.
[6] J. E. Hugonnet, D. Mengin-Lecreulx, A. Monton, T. den Blaauwen, E.
Carbonnelle, C. Veckerle, Y. V. Brun, M. van Nieuwenhze, C. Bouchier, K.
Tu, L. B. Rice, M. Arthur, Elife 2016, 5.
[7] J. L. Mainardi, R. Legrand, M. Arthur, B. Schoot, J. van Heijenoort, L.
Gutmann, J. Biol. Chem. 2000, 275, 16490–16496.
[8] D. A. Preston, C. Y. E. Wu, L. C. Blaszczak, D. E. Seitz, N. G. Halligan,
Antimicrob. Agents Chemother. 1990, 34, 718–721.
[9] G. S. Zhao, T. I. Meier, S. D. Kahl, K. R. Gee, L. C. Blaszczak, Antimicrob.
Agents Chemother. 1999, 43, 1124–1128.
[10] C. M. June, R. M. Vaughan, L. S. Ulberg, R. A. Bonomo, L. A. Witucki, D. A.
Leonard, Anal. Biochem. 2014, 463, 70–74.
[11] K. E. Beatty, Curr. Opin. Chem. Biol. 2020, 57, 50–57.
[44] L. Lopez Quezada, R. Smith, T. J. Lupoli, Z. Edoo, X. Li, B. Gold, J. Roberts,
Y. Ling, S. W. Park, Q. Nguyen, F. J. Schoenen, K. Li, J. E. Hugonnet, M.
Arthur, J. C. Sacchettini, C. Nathan, J. Aube, Front. Microbiol. 2020, 11,
1248.
[45] P. Kumar, K. Arora, J. R. Lloyd, I. Y. Lee, V. Nair, E. Fischer, H. I. Boshoff,
C. E. Barry, Mol. Microbiol. 2012, 86, 367–381.
[12] M. H. Wright, S. A. Sieber, Nat. Prod. Rep. 2016, 33, 681–708.
[13] I. Staub, S. A. Sieber, J. Am. Chem. Soc. 2008, 130, 13400–13409.
[14] I. Staub, S. A. Sieber, J. Am. Chem. Soc. 2009, 131, 6271–6276.
[15] E. Romeo, S. Ponzano, A. Armirotti, M. Summa, F. Bertozzi, G. Garau, T.
Bandiera, D. Piomelli, ACS Chem. Biol. 2015, 10, 2057–2064.
[16] X. Chen, Y. Wang, N. Ma, J. Tian, Y. Shao, B. Zhu, Y. K. Wong, Z. Liang, C.
Zou, J. Wang, Signal Transduct. Target Ther. 2020, 5, 72.
[17] Y. Ding, Z. Li, C. Xu, W. Qin, Q. Wu, X. Wang, X. Cheng, L. Li, W. Huang,
Angew. Chem. Int. Ed. 2021, 60, 24–40. Angew. Chem. 2021,133, 24–40.
[18] O. Kocaoglu, R. A. Calvo, L. T. Sham, L. M. Cozy, B. R. Lanning, S. Francis,
M. E. Winkler, D. B. Kearns, E. E. Carlson, ACS Chem. Biol. 2012, 7, 1746–
1753.
Manuscript received: February 21, 2021
Accepted manuscript online: April 1, 2021
Version of record online: ■■■, ■■■■
[19] S. Das, J. Ihssen, L. Wick, U. Spitz, D. Shabat, Chem. Eur. J. 2020, 26,
3647–3652.
Chem. Eur. J. 2021, 27, 1–10
www.chemeurj.org
9
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
In this study, we functionalized carba-
penem β-lactams for the develop-
ment of an original activity-based pu-
rification method of the complex
targets of these drugs. The alkyne
drug covalently bound to the target
specifically reacted with an azido
probe containing a biotin moiety for
affinity purification and an iminosyd-
none for single-step elution and fluo-
rescent labeling of the targets.
S. Saidjalolov, Dr. E. Braud, Dr. Z. Edoo,
Dr. L. Iannazzo, Dr. F. Rusconi, Dr. M.
Riomet, Dr. A. Sallustrau, Dr. F. Taran,
Dr. M. Arthur*, Dr. M. Fonvielle*,
Prof. Dr. M. Etheve-Quelquejeu*
1 – 10
Click and Release Chemistry for
Activity-Based Purification of β-
Lactam Targets