Keywords
bioorthogonal chemistry - chemical biology - strained alkynes - tetrazines - drug
activation
1
Introduction
Figure 1 Timeline of developments in the field of bioorthogonal chemistry including contributions
from Radboud University.
Historically, chemical transformations have been carried out under strictly defined
conditions to ensure optimal reaction outcomes. These conditions often involve the
use of anhydrous solvents, high concentrations of reagents, prolonged reaction times,
and elevated temperatures. The growing significance of biotherapeutics and advancements
in chemical biology have urged chemists to develop chemical reactions that not only
exhibit a high level of chemoselectivity but also demonstrate biocompatibility. Carolyn
Bertozzi was the first to coin the term ‘bioorthogonal chemistry’, referring to reaction
systems that typically proceed at low micromolar concentrations with a high degree
of chemoselectivity while maintaining overall biocompatibility.[1] Ultimately, this allowed chemists to conduct synthetic chemistry within a biological
environment.
After years of intense research, an impressive repertoire of reactions has been reported
that can be considered as truly bioorthogonal, pushing the boundaries of synthetic
organic chemistry.[2] Our ‘Chemistry Institute’ at Radboud University – known as the Institute for Molecules
and Materials – has made significant contributions to these advancements. For example,
pioneering work on strained alkynes for strain-promoted azide–alkyne click chemistry
was developed at Radboud University in Nijmegen during the early stages.[3] This Account reflects on the developments that originated from Radboud University
and highlights their impact on the emerging field of bioorthogonal chemistry (Figure
[1]).
One can debate the exact origin of our institute’s focus on bioorthogonal chemistry,
but our strong emphasis on interdisciplinary research and education at Radboud University
certainly has played an essential role. This emphasis extends beyond traditional subdisciplines
of chemistry and includes research collaborations with biologists, clinicians, and
physicists, but also broader educational programmes such as the Bachelor’s programme
of Molecular Life Sciences. It was this interdisciplinary nature that paved the way
for early work in bioorthogonal chemistry, a discipline that emerged at the interface
of chemistry and biology with applications in biomedical research and medicine. Professors
Floris Rutjes and Floris van Delft conducted pioneering work on metal-free click chemistry
tools at our institute, laying the foundation for the modern field of bioorthogonal
chemistry.[4]
Since then, our institute has attracted numerous early career academics who have developed
into established group leaders with international recognition in the field of bioorthogonal
chemistry. The group led by Dr. Kim Bonger has not only developed vinyl boronic acids
as new chemical tools for bioorthogonal chemistry but has also reported metabolically
active amino acid derivatives with bioorthogonal functionalities, enabling the study
of protein synthesis at a molecular level.[5]
[6] Other faculty members expanded the use of bioorthogonal chemistry into the fields
of carbohydrate and RNA chemistry, namely Dr. Thomas Boltje and Dr. Willem Velema,
respectively.[7,8] Finally, new faculty member Dr. Kevin Neumann and his group exploit bioorthogonal
chemistry for applications in nanomedicine and biotherapeutics.[9] In this Account, we provide specific examples alongside a historical overview of
achievements associated with bioorthogonal chemistry at Radboud University, showcasing
the power of interdisciplinarity in modern chemistry for creating impact.
Providing BCN as a Robust Bioorthogonal Tool for Chemical Biology and Beyond (by Floris
Rutjes)
2
Providing BCN as a Robust Bioorthogonal Tool for Chemical Biology and Beyond (by Floris
Rutjes)
Since the initial introduction of bioorthogonal chemistry by Bertozzi and her colleagues,
based on copper click chemistry, researchers have sought to identify more biocompatible,
stable, and efficient reagents. Bertozzi’s group reported strain-promoted alkyne–azide
cycloaddition (SPAAC).[10] In brief, the ring strain alters the bond angles of the sp-hybridized carbons to
ca. 160°, thus changing the geometry towards the transition state of the cycloaddition.
It’s noteworthy that cyclooctynes and their rapid reactions with phenylazides were
reported in the 1950s by the teams around Blomquist and Prelog at Cornell University
and ETH Zurich, respectively.[11]
[12] At Radboud University, a team led by Floris van Delft introduced cyclooctynes as
suitable reagents for rapid strain-promoted cycloaddition with nitrones, referred
to as SPANC, in 2010.[13] At that time, available strained alkynes suffered from lengthy synthesis routes.
For instance, dibenzo-azacyclooctyne 1 (DIBAC, nowadays commonly termed DBCO) required nine steps, while difluorooctyne
2 (DIFO) needed eight synthetic steps (Figure [2]).[14]
[15] Additionally, these strained alkynes are asymmetrical, eventually resulting in several
isomers as products, complicating product isolation and characterization.
Figure 2 Chemical structures of strained alkenes and alkynes highlighted in chapter 2.
Prior to studying strained cyclooctynes, we developed the strained and electron-deficient
oxanorbornadiene 3 (OND), synthesized in one step from commercially available starting materials, as
a versatile molecule for (bio)conjugation.[16] OND readily reacts with azides in a cycloaddition reaction, forming the corresponding
triazole upon retro-Diels–Alder reaction and release of furan (Scheme [1]). Although the reaction rate is modest compared to cyclooctynes, its good water
solubility and ease of synthesis, have led to various applications.[17] This early work on bioorthogonal chemistry made us realize that an often-overlooked
key aspect is the synthetic availability of bioorthogonal reagents.
Scheme 1 a) Synthesis of OND in a one-step procedure from commercially available alkynes and
furan; b) reaction of OND with azides. Subsequent retro-Diels–Alder reaction provides
the triazole.
It was in the second half of 2010 that Dr. Floris van Delft and myself reported bicyclo[6.1.0]nonyne
4 (BCN) as a readily available and remarkably stable strained alkyne (Scheme [2]).[4a] Inspired by the fact that benzoannulation increases the reactivity of cyclooctynes
towards strain-promoted 1,3-dipolar cycloadditions as demonstrated for DIBO and DIBAC,
we envisioned that fusion with a cyclopropane should have a similar effect while limiting
lipophilicity. The synthesis of BCN was accomplished in four steps, yielding approximately
60% overall and started from commercially available 1,5-cyclooctadiene which was converted
into 5 by addition of ethyl diazoacetate in the presence of rhodium acetate. The resulting
diastereomeric mixture was readily separated by column chromatography.
Scheme 2 Synthesis of BCN 4 in four steps. The last three steps including reduction, bromination, and elimination
are performed sequentially and require only a final purification step.
Stability was assessed in the presence of glutathione without signs of degradation.
Initial experiments with benzyl azides revealed second-order rate constants of 0.29
and 0.19 M–1 s–1 for endo-BCN and exo-BCN, respectively, compared to second-order rate constants of 7.6·10–2 M–1 s–1 for DIFO.[10] Conveniently, BCN displays a C
s symmetry, overcoming challenges typically associated with its asymmetrical counterparts,
such as DIBAC/DBCO and DIFO. In the following years, BCN was not only applied in the
fields of chemical biology but also in the fields of materials science and supramolecular
chemistry. For example, in collaboration with the group of Prof. Jan van Hest, we
demonstrated that BCN-based crosslinkers enable the chemical triggering of shape transformations
in polymersomes.[18] Besides many applications in research, BCN is used in several antibody–drug conjugates
(ADCs) that are currently undergoing evaluation in phase 1 clinical trials, driven
by the company Synaffix (currently Lonza, Oss, Netherlands).
Thus, BCN has served us and many chemical biologists as a robust tool for bioorthogonal
chemistry. In particular, its feasible synthesis and stability are unmatched. One
disadvantage though that applies to most reaction partners for 1,3-dipolar cycloadditions
is the relatively low second-order rate constants, with k₂ ranging from 0.1 to 5 M–1 s–1. This has led us and others to turn towards alternative bioorthogonal tools.
Towards Readily Available Click-to-Release trans-Cyclooctenes (by Floris Rutjes)
3
Towards Readily Available Click-to-Release trans-Cyclooctenes (by Floris Rutjes)
Around the same time when we reported BCN as a readily available and stable reagent
for 1,3-dipolar additions, the group of Joseph Fox revisited the inverse electron-demand
Diels–Alder (IEDDA) reactions of tetrazines and reported their use for bioorthogonal
chemistry.[19] Reactions between tetrazines and dienophiles not only display high levels of chemoselectivity
but also high second-order rate constants in comparison to SPAAC. In particular, reactions
between trans-cyclooctene (TCO) and electron-deficient tetrazines exhibit unmatched reaction kinetics
with k₂ up to 3,300,000 M–1 s–1.[20] These unique properties have made the IEDDA cycloaddition the reaction of choice
when working under dilute biological conditions or even in vivo.[21]
Once again, one bottleneck of this powerful bioorthogonal tool was the synthesis of
the strained 8-membered ring, in this case, trans-cyclooctene. In particular, (di)functionalized trans-cyclooctenes possess synthetic challenges and often require multistep synthetic routes,
significantly impacting the field of bioorthogonal chemistry.
Scheme 3 a) Structural features displayed by difunctionalized trans-cyclooctenes; b) key step in the synthetic route toward TCO 8 is acetylated iodohydrin 6.
Many applications of TCO rely on its ability to be either pretargeted or used for
‘click-to-release’ chemistry, describing the release of cargos in the allylic position
from TCOs upon IEDDA and subsequent isomerization.[22] Notable, the so-called ‘click-to-release mechanism’ was developed by the team of
Marc Robillard associated to Tagworks Pharmaceuticals which are nowadays also based
in Nijmegen. Inspired by our earlier success employing a fused cyclopropane, we sought
to explore the possibility of providing readily available strained TCOs susceptible
to further modifications and ‘click-to-release’ chemistries. Eventually, we and others
reported a five-step synthesis that provides TCOs displaying competitive rate constants
in IEDDA.[23] Our synthesis started from commercially available cis,cis-1,5-cyclooctadiene which was transformed into compound 5 (Scheme [3]). During this synthesis, a key intermediate was the acetylated iodohydrin 6 and its subsequent elimination toward cyclooctene 7. Over the years, our group has gained significant expertise in flow chemistry and
reported, already in 2018, a continuous-flow protocol that enabled efficient photoisomerization
toward trans-cyclooctenes 8.[24]
[25] In brief, substituting the traditionally employed silica gel column with a liquid–liquid
extraction module allowed the production of up to 2.2 g/h of specific TCOs. A similar
setup was employed for the isomerization of our difunctionalized TCO. Altogether,
we have been able to provide a robust synthetic route that allowed access to difunctional
TCOs in only four steps, including the widely employed releasable one. The future
will show if this TCO as a bioorthogonal tool holds similar success compared to our
previously reported BCN.
Giving Molecules Guidance (by Kimberly Bonger)
4
Giving Molecules Guidance (by Kimberly Bonger)
Figure 3 The coordination of VBA with the pyridyl substituent induces proximity of the reagents
and improved reaction rates. This unique feature allowed the orthogonal bioorthogonal
reaction of two proteins labeled. Figure adapted from ref. 27.
The development of the IEDDA reaction with tetrazines and dienophiles have revolutionized
the field of bioorthogonal chemistry. The superior reaction rates allowed researchers
to perform reactions in very dilute conditions needed for in vivo applications. Especially
the cycloaddition with strained alkenes, such as TCO, proved suitable and several
groups have applied these reagents for imaging and targeted drug delivery. Yet, the
stability of these dienophiles in vivo remains challenging.[26] In our group, we were interested to see if we could obtain more stable, yet reactive
dienophiles. As the IEDDA proceeds faster with electron-rich dienophiles, we envisioned
that the reactivity of the otherwise unreactive linear alkenes may be improved by
introducing electron-donating substituents. While alkoxy substituents resulted only
in slight improved reactivity, we were very excited to see that the boronic acid substituents
showed unexpectedly high reaction rates that were order of magnitude faster than the
unsubstituted alkenes.[5] We envisioned that this unique reactivity was partly attributed to the formation
of a charged boronate in aqueous environment. While this may explain part of the reactivity,
we additionally found that the vinyl boronic acids were especially reactive to tetrazines
with a pyridyl substituent, while they render inactive to tetrazines containing a
phenyl or a pyrimidyl substituent. We hypothesized that the boronic acid coordinates
to the pyridyl substituent, thereby inducing proximity of the reagents to facilitate
the cycloaddition reaction. The more acidic pyrimidyl substituent or the noncoordinating
phenyl substituent are less preferred coordinating substituents and therefor are not
that reactive with VBAs.
Figure 4 a) Second-order reaction rates of a panel of tetrazines with VBA and norbornene.
Rates are measured in 50% methanol/PBS at 20 °C. b) Measured absorption of the tetrazines
in 5% DMSO/PBS at 20 °C over time. Figure adapted from ref 28.
The coordination-induced cycloaddition of tetrazines with VBAs and the large difference
in reaction rates with tetrazines containing coordinating- or noncoordinating tetrazines,
allowed us to introduce orthogonality within the iEDDA with other strained alkenes
for dual orthogonal protein modification. In this example, we modified two proteins
containing either a VBA or norbornene dienophile. We were able to react the norbornene
selectively and fully with a pyrimidyl tetrazine after which the VBA was reacted with
a pyridyl tetrazine in a sequential reaction step (Figure [3]).[27]
It is generally accepted that tetrazines containing electron-withdrawing substituents
react fastest in IEDDA reactions, but they are also rather unstable in aqueous conditions.
The unique coordinating reactivity of VBAs prompted us to also explore other potential
coordinating tetrazines. We envisioned that the more electron-rich phenolic substituents
would provide more stable tetrazines, while still allowing fast reaction with VBAs.
Indeed, a striking difference in reaction rate was observed with o-phenol-substituted tetrazines that showed more than 5 orders of magnitude improved
reaction rates compared to m-phenol-substituted tetrazines in the reaction with VBA (Figure [4]).[28] Additionally, the tetrazines proved fully stable in aqueous conditions for at least
12 h. Expectedly, the more electron-rich phenol-substituted tetrazines react poorly
with norbornene, providing the additional possibility to perform orthogonal bioorthogonal
IEDDA reactions with noncoordinating- and more electron-withdrawing tetrazines.
To further explore the scope of the VBA–tetrazine IEDDA reaction we additionally explored
VBAs for click-to-release chemistry. We envisioned that VBAs are especially useful
for this application as they are water-soluble and stable under aqueous conditions.
We designed a VBA containing a self-immolative linker to cage a doxorubicin toxin
(Figure [5]).[29] We were excited to observe cell death only when applying the construct and an uncaging
phenol-containing coordinating tetrazine, while no cell death was observed when subjecting
the cells to the caged doxorubicin alone. The more electron-rich phenol tetrazine
allowed selective reaction with the VBA cage, while no reaction was observed with
vinyl ethers, a caging modality that was reported before by the groups of Devaraj
and Bernardes.[30] While this reaction provides an orthogonal decaging strategy, we observed that the
reaction rates were rather low to be used in living systems, likely due to the increased
electron density present on the boronic acid due to the alkoxy substituent.
Figure 5 A) VBA-based click-to-release strategy. Doxorubicin was caged with a VBA connected
to a self-immolative linker. B) Cell death was only observed upon addition of decaging
tetrazine. Figure adapted from ref 29.
In our experience, we have observed great benefits of using VBAs in bioorthogonal
reactions. The presence of high amounts of free thiols present in cells challenges
many bioorthogonal reactions as they perform nucleophilic addition with strained alkynes
or alkenes. VBAs are inert and stable under aqueous conditions. We have used activity-based
probes containing VBA handles for labeling and imaging the proteasome.[31] Indeed, using this probe we could image the catalytic activity of proteasome subunits
inside living cells. Additionally, VBAs are also hydrophilic and very soluble which
is a great benefit when the use of more lipophilic reactants, such as DBCO, proved
challenging due to unfavorable physicochemical properties.
Next Generation of Bioconjugation Strategies: Dynamic Click Chemistry (by Kevin Neumann)
5
Next Generation of Bioconjugation Strategies: Dynamic Click Chemistry (by Kevin Neumann)
A longstanding challenge in synthetic chemistry is the modification of complex (bio)molecules
with temporal control. At the very beginning of my independent career, we envisioned
that the use of a reversible click reaction with an on-demand switch towards irreversibility
offers numerous applications in synthetic chemistry and chemical biology. For example,
one can imagine that this chemistry may find applications in proteome-wide screening
of cysteines. In principle, such a system requires i) a rapid reversible click transformation,
ii) a (bio)orthogonal reaction that induces irreversibility at any given point, and
iii) the possibility to use the reactivity handles to incorporate chemical functionality.
Thiol–maleimide chemistry, in principle, fulfils many of these aspects, namely reversibility
and the possibility to incorporate complexity.[32] Yet, while the switch to irreversibility is possible via hydrolysis, this process
is typically difficult to control precisely. Instead, K. Gavriel, a talented PhD candidate
in my group, sought to employ 1,2,4,5-tetrazines and to take advantage of their ability
to undergo two forms of click transformations, namely aromatic nucleophilic substitutions
and IEDDA chemistries.[33] This work was inspired by early efforts related to the aromatic nucleophilic substitution
of 1,2,4,5-tetrazines, allowing access to stapled peptides and crosslinked polymeric
networks.[34] The reversibility of the reported bis-sulfide-functionalized tetrazines under biological
relevant conditions, however, remained elusive, which we attributed to the high electron
density displayed by the typically employed bis-sulfide tetrazine. We hypothesized
that an asymmetric sulfide tetrazine would not only enable conjugation of complex
cargos such as biotin or drug molecules but, importantly, also tailor the reactivity
towards aromatic nucleophilic substitution.
To confirm our hypothesis, we accessed a range of tetrazines that displayed varying
electronic properties (Scheme [4]).[33] Initial experiments not only confirmed our hypothesis that the tetrazine–thiol exchange
(TeTEx) on asymmetric tetrazines rapidly occurs in aqueous environment but also displays
a reversible nature. Conveniently, by employing methylthiol-substituted tetrazines,[35] the reaction proceeds to full conversion, while other substitutes provided an equilibrium.
Scheme 4 A) The exchange between thiols and sulfide-bearing tetrazines is reversible. Upon
exposure to a dienophile and subsequent IEDDA reaction, the reaction is locked and
irreversible. B) TeTEx kinetics depend strongly on electronic properties of the tetrazine.
C) TeTEx is chemoselective and allows modification of peptides and proteins.
This is because methanethiol can be removed from the reaction mixture by simply saturating
the solution with inert gas such as nitrogen. As hypothesized, second-order rate constants
depend on the electronic properties, with electron-deficient tetrazines such as pyrimidine-substituted
tetrazines displaying rate constants higher than 20 M–1 s–1. In stark contrast to existing cysteine modifications, TeTEx allows switching from
reversible to irreversible modifications of complex biomolecules enabled by the reaction
of tetrazines with dienophiles towards pyridazines via IEDDA chemistry. We demonstrated
this on-demand switch on representative peptide scaffolds by employing different dienophiles,
with BCN being the most efficient lock. Our results revealed that those dienophiles
afford locked products that initially provide pyridazines, while dienophiles that
afford dihydropyridazines are prone to hydrolysis. Independently of us, the group
around Joseph Fox elegantly employed the rapid exchange of tetrazine sulfides and
thiols as a click reaction for applications in proteome-wide screening of cysteines.[36] In this case, the capability of locking the reaction partners was employed for pull-down
assays.
Scheme 5 The N-terminal installation of methylsulfide tetrazine enabled traceless cyclization
with internal and C-terminal cysteines in the absence of additional coupling reagents
or extensive protecting group reshuffling. Instead, the cyclization is triggered by
buffered aqueous solutions. Figure adapted from ref 37.
Our group envisioned similar usage but, in addition, was keen on further applying
dynamic chemistry for synthetic applications in the field of peptide and protein chemistry.
In particular, the cyclization of peptides often suffers from undesired side reactions,
such as the formation of dimers and trimers. By utilizing the exchange between tetrazine
sulfides and thiols as a dynamic click reaction, we anticipated that higher concentrations
could be tolerated during peptide cyclization. Initial concerns related to the compatibility
of the tetrazine sulfide scaffold during solid-phase peptide synthesis (SPPS) have
proven unjustified, and a small library of tetrazine-terminated peptides was accessed
(Scheme [5]).[37] Interestingly, we observed that silanes, as scavengers, provide better results than
pure water. Cyclization is triggered by exposure to buffered aqueous environments,
which might prove powerful in the future when applied in high-throughput screenings.
Ultimately, TeTEx was employed for the cyclization of peptide scaffolds with a selection
of representative amino acids, with the hope of sparking interest in other groups
to use this cyclization strategy that avoids extensive protecting group reshuffling
or exotic activation reagents.
6
Conclusions
Since the first report by Bertozzi and co-workers over 20 years ago, the field of
bioorthogonal chemistry has witnessed many exciting advances, profoundly impacting
other fields such as chemical biology and drug delivery. Naturally, this progress
is driven by a collaborative effort from both chemists and biologists, making interdisciplinary
research indispensable. In this Account, we aim to illustrate this essential aspect
by reflecting on the developments related to bioorthogonal chemistry within our own
chemistry institute at Radboud University. A closer look at the research described
herein exemplifies that chemists can contribute in many ways to a diverse field such
as bioorthogonal chemistry. For example, the expertise in synthetic organic chemistry
provided by Rutjes and co-workers has resulted in the design of new click tools and
conceptually new synthetic routes, allowing scalable synthesis. The group of Kimberly
Bonger is enhancing the field of bioorthogonal chemistry by employing a chemical biology
perspective. In another example, the groups of Neumann and Bonger demonstrate that
bioorthogonal chemistry is a unique chemical tool for developing new advances in medicine.
Both, our research lines and that of others highlight the importance of interdisciplinary
and collaborative efforts. While individual contributions are valuable, ultimately,
one can conclude that teamwork and collaborative efforts are essential.
