Angewandte
Communications
Chemie
was also observed that these conditions resulted in significant
resulting peptide fragments were analyzed by LC-MS/MS.
protein loss (through degradation or precipitation) when
compared to the allyl selenocyanate reaction (mass yields of
This showed that modification with Py-Tz 2 occurs at Cys315
where the S-allyl handle has been chemically installed
(Figure 2d; results for C2Am are shown in the Supporting
Information). Labeling of AnxV S-allyl Cys with the tetrazine
fluorophores Tz-Rhod 4 and Tz-Cy3 5 was also studied. The
reactions were set up in PBS buffer at pH 7.6 with 10% DMF
in the presence of 200 equivalents of tetrazine dyes 4 and 5.
The labeling of AnxV was confirmed by SDS-PAGE (Fig-
ure 3b) and LC-ESI-MS (Figure 3c) analysis. Fluorescence
imaging of the gel showed a band corresponding to the
fluorescently labeled protein (Figure 3b), while LC-ESI-MS
indicated 40% conversion to the product of the reaction with
9
1% and 48% for allyl selenocyanate and allyl chloride
reactions, respectively, as assessed by Bradford protein assay).
S-allyl-tagged AnxV was then subjected to reaction with
tetrazine cores Py-Tz 2 and BnNH -Tz 3 (Figure 2c,e).
2
Product formation was detected ( ꢀ 40% conversion) by
LC-ESI-MS after 20 and 12 h of reaction, respectively
(
complete reaction was obtained after 96 h for 2 and 36 h
for 3; Supporting Information). Reactions were monitored
using Liquid Chromatography-Electrospray Ionization Mass
Spectrometry (LC-ESI-MS) analysis and Ellmanꢁs test (data
for C2Am shown in the Supporting Information).
LC-ESI-MS analysis of the reactions with both tetrazine
cores revealed that, besides the desired product, there was
also the formation of a secondary product (10–20%) corre-
sponding to the addition of two tetrazine moieties to the
proteins (identical results were obtained for both AnxV and
C2Am; Figure 2b,c and Supporting Information). Cross-
reactivity of reagents with other amino acids is a major
issue for the efficiency of bioorthogonal labeling. For
instance, it was recently reported that 2,5-diaryltetrazoles,
which were thought to selectively react with alkenes through
[
18]
light-induced 1,3-dipolar cycloadditions,
also react with
[
19]
native tryptophan (Trp) residues on proteins.
To verify
whether the observed addition of two tetrazine moieties was
the result of cross-reactivity with other unsaturated canonical
amino acids, tetrazine 2 was reacted with native proteins
AnxV and C2Am (that is, without the S-allyl handle) under
the same IEDDA reaction conditions (200 equivalents of
tetrazine, 378C for 96 h). These control experiments revealed
that there is no addition of 2 to either AnxV or C2Am when
the S-allyl handle is not present (Supporting Information).
Importantly, we showed that Trp and histidine (His), both
possessing potentially reactive double bonds, do not react
with 2 even under forcing conditions (Supporting Informa-
tion). Moreover, the specificity of the IEDDA reaction
towards the S-allyl handle was further demonstrated by
replacing it by dimethylallyl or propargyl groups that also
gave no addition reaction in the presence of a tetrazine
(
Supporting Information). Finally, the Boc- and methyl ester-
protected S-allyl Cys amino acid model was reacted with Py-
Tz 2 yielding the corresponding reduced and oxidized
dihydropyridazine and pyridazine species, as well as a third
compound resulting from the addition of a second tetrazine
moiety to one of the pyridazine species (Supporting Informa-
tion). These data on small molecules are consistent with our
experiments on the protein context, which demonstrate that
the formation of the bis-adduct is not a result of cross
reactivity with canonical amino acids but instead is the
product of the reaction of a second tetrazine moiety with the
first one. A detailed computational study and discussion of
Figure 3. a) Pre-targeting of apoptotic cells with AnxV S-allyl Cys
followed by IEDDA labeling with fluorogenic Tz-Cy3 5. b) SDS-PAGE of
AnxV labeled with Tz-Rhod 4 and Tz-Cy3 5 for 12 h at 378C. The
labeled protein was purified by dialysis. Left image shows Coomassie
blue-stained gels and the right image presents the fluorescent imaging
of the same gels before staining with Coomassie blue. c) ESI-MS
spectrum of the product of the reaction of AnxV S-allyl Cys with Tz-Cy3
5. d) Specific labeling of apoptotic cells (ii) pre-targeted with AnxV S-
allyl Cys and subsequently labeled with Tz-Cy3 5. Cells were imaged 3 h
after addition of the fluorogenic dye 5. Non-apoptotic cells (i) were
included as a control. The specificity of the labeling was confirmed
through blocking of cells with non-labeled AnxV (iii) prior to incubation
with AnxV S-allyl Cys or by treating cells only with the tetrazine probe
Tz-Cy3 5 (iv) without preincubation with the targeting protein AnxV S-
allyl Cys. Apoptotic cells are shown red, while the nuclei counter-
stained with Hoechst 33342 are shown blue. Scale bar=100 mm.
the whole reaction profile at the PCM(H O)/M06-2X/6-
2
3
1G(d) level of theory using complete models for both the
amino acid and tetrazine counterparts is available in the
Supporting Information (Figures S40–42).
To confirm the site of the modification, the protein
conjugates were subjected to tryptic digestion, and the
Angew. Chem. Int. Ed. 2016, 55, 1 – 6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!
Oliveira, Bruno L.
