Investigating Ugi/Passerini Multicomponent Reactions for the Site-Selective Conjugation of Native Trastuzumab**

Article abstract of DOI:10.1002/chem.202002432

Site-selective modification of proteins has been the object of intense studies over the past decades, especially in the therapeutic field. Prominent results have been obtained with recombinant proteins, for which site-specific conjugation is made possible by the incorporation of particular amino acid residues or peptide sequences. In parallel, methods for the site-selective and site-specific conjugation of native and natural proteins are starting to thrive, allowing the controlled functionalization of various types of amino acid residues. Pursuing the efforts in this field, we planned to develop a new type of site-selective method, aiming at the simultaneous conjugation of two amino acid residues. We reasoned that this should give higher chances of developing a site-selective strategy compared to the great majority of existing methods that solely target a single residue. We opted for the Ugi four-centre three-component reaction to implement this idea, with the aim of conjugating the side-chain amine and carboxylate groups of two neighbouring lysine and aspartate/glutamate. Herein, we show that this strategy can give access to valuable antibody conjugates bearing several different payloads; furthermore, the approach limits the potential conjugation sites to only six on the model antibody trastuzumab.

Full text of DOI:10.1002/chem.202002432

Accepted Article
Accepted Article
Title: Investigating Ugi / Passerini multicomponent reactions for the
Site-Selective Conjugation of Native Trastuzumab
Authors: Charlotte Sornay, Steve Hessmann, Stéphane Erb, Igor
Dovgan, Anthony Ehkirch, Thomas Botzanowski, Sarah
Cianférani, Alain Wagner, and Guilhem Chaubet
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202002432
Link to VoR: https://doi.org/10.1002/chem.202002432
0
1/2020

Chemistry - A European Journal
10.1002/chem.202002432
Investigating Ugi / Passerini multicomponent reactions for the Site-Selective
Conjugation of Native Trastuzumab
1
2
2
1
2
Charlotte Sornay , Steve Hessmann , Stéphane Erb , Igor Dovgan , Anthony Ehkirch , Thomas
2
2
1
1*
Botzanowski , Sarah Cianférani , Alain Wagner , Guilhem Chaubet
AUTHOR ADDRESS
1
Bio-Functional Chemistry (UMR 7199), LabEx Medalis, University of Strasbourg, 74 Route du Rhin,
6
7400 Illkirch-Graffenstaden, France
2
Laboratoire de Spectrométrie de Masse BioOrganique (LSMBO), LabEx Medalis, Université de
Strasbourg, CNRS, IPHC UMR 7178, 67000 Strasbourg, France
KEYWORDS
Bioconjugation; multicomponent reaction; antibodies; antibody-drug conjugates
ABSTRACT: Site-selective modification of proteins has been the object of intense studies over the past
decades, especially in the therapeutic field. Prominent results have been obtained with recombinant
proteins, for which site-specific conjugation is made possible by the incorporation of particular amino
acid residues or peptide sequences. In parallel, methods for the site-selective and site-specific
conjugation of native and natural proteins are starting to thrive, allowing the controlled functionalization
of various types of amino acid residues. Pursuing the efforts in this field, we planned to develop a new
type of site-selective method, aiming at the simultaneous conjugation of two amino acid residues. We
reasoned that this should give higher chances of developing a site-selective strategy compared to the
large majority of existing methods that solely target a single residue. We opted for the Ugi four-center
three-component reaction to implement this idea, with the aim of conjugating the side-chain amine and
carboxylate groups of two neighbouring lysine and aspartate/glutamate. Herein, we show that this
strategy can give access to valuable antibody conjugates bearing several different payloads, and limits
the potential conjugation sites to only six on the model antibody trastuzumab.
Proteins posttranslational modifications is Nature’s way of generating a rich and diverse proteome from
a more limited genetic coding capability. First occurrences of intentional, man-made – artificial – proteins
modifications using a defined chemical – thus excluding the food-related Maillard reaction for example
–
could be dated back to the use of formaldehyde in the tanning industry or for the production of
1,2
toxoids, which evolved later on to immunization studies using chemically-modified bovine serum
3,4
albumin in the 1900s and eventually led to Landsteiner’s synthetic haptenes studies. The field of
protein modification has since largely benefited from the understanding of proteins’ and amino acids’
structures coupled to the parallel appearance of more efficient and precise analytical tools. This finally
resulted in the development of bioconjugation reagents with excellent chemoselectivity towards various
amino acids’ side chains groups (i.e. residue-selectivity) that translated into major applications, notably
in the pharmaceutical field with the generation of protein-fluorophore adducts for trafficking studies, or
the polyethyleneglycol chains functionalization (PEGylation) of proteins to give less-immunogenic and
5,6
more plasma-stable conjugates.
However, site selectivity quickly emerged as the main limitation of chemoselective strategies, due to the
presence of multiple copies of each type of amino acid residue at the surface of proteins. Statistic
conjugation of surface-accessible lysine residues with amine-selective reagents typically results in
highly heterogeneous mixtures, containing up to millions of different adducts when large proteins such
as antibodies are utilised.7 Each of these adducts possessing distinct physicochemical properties, such
chemoselective conjugation necessarily leads to mixtures with different in-vivo pharmacokinetic
,8
9,10
properties along with virtually no reproducibility in batch-to-batch production. Regioselective (i.e., site-
specific) methods were thus developed and are currently dominated by the use of recombinant proteins,
incorporating exogenous amino acid residues – natural or unnatural – or peptide sequences that can be
11–13
specifically targeted by a tailored reagent or strategy.
In parallel, site-selective chemical strategies
for the conjugation of native and natural proteins have also flourished over the past few years, giving
rise to methods targeting various types of amino acids – e.g. lysine, cysteine, tryptophan, tyrosine – that
14–28
proved to be effective on proteins of all sorts of sizes, including antibodies.
With the aim of pursuing the efforts in this field, we could not help but notice that the vast majority of
previously reported strategies for the site-selective conjugation of native proteins were focused on the
modification of a unique residue. We hypothesized that targeting two different amino acid side chains
simultaneously would lower the enormous subset of possibilities given by single-residue bioconjugation
techniques, thus increasing our chances of developing a site-selective method by minimising the number
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
of potentially reactive sites, a path that has also been successfully explored by others in the meantime.29
In order to do so, we elected the Ugi four-centre three-component reaction (U-4C-3CR) for the
concomitant conjugation of two spatially close lysine and aspartate (or glutamate) residues (Figure 1).
This variation of the classical Ugi four-component reaction (U-4CR) employs bifunctional reagents,
where two of the four components necessary to the reaction (i.e. an amine, a carboxylic acid, a carbonyl
and an isocyanide) are borne by the same molecule. It was thus anticipated that close amine/carboxylate
pairs at the surface of the protein would give the expected bis-amide final product in the presence of an
aldehyde and an isocyanide, while lysines that are not in the vicinity of a carboxylate moiety could form
3
0,31
imines but would not give any productive Ugi reaction.
A solution of hydroxylamine would then simply
be added at the end of the reaction to regenerate the amino groups of these lysine residues that did not
partake in the multicomponent reaction. While the Ugi reaction has been the object of extensive studies
for the conjugation and stapling of peptides, only a handful of publications reports the use of proteins as
coupling partners.3
2–36
Most notably, all these strategies use the four-component version of the Ugi
reaction (U-4CR) – i.e., they only target a single site at the protein surface, whether it be an amino acid
residue (lysines or aspartates / glutamates) or aldehydes from the glycan portion after periodate
oxidation – and necessitate excessive amounts of reagents (up to 4000 equivalents) and long reaction
times (up to 4 days). In addition, the absence of thorough analytical investigations precludes the
conclusion that the Ugi reaction, and not an uncontrolled side reaction, was indeed responsible for the
bioconjugation. This aspect is of prime importance in our case, due to the probable competition of the
Passerini reaction for the conjugation of aspartate / glutamate residues that could erode the potential
site-selectivity of the U-4C-3CR.
previous work
Ugi four-component
reaction for the conjugation
of one type of residue
this work
Ugi three-component four-center
reaction for the conjugation
of two neighbouring residues
Figure 1: application of the Ugi reaction to the bioconjugation of native proteins
As we ambition to develop a site-selective strategy that could be applied to the conjugation of large
proteins, we selected trastuzumab (~150 kDa) as our model protein, a FDA-approved antibody marketed
as Herceptin that is used for the treatment of breast cancer by targeting the overexpressed HER2
receptors.37 We embarked upon the study of the feasibility of our strategy by reacting it with various
amounts of the model TAMRA aldehyde 1a and tert-butyl isocyanide 2a (Figure 2, a). The reaction was
set up in phosphate-buffered saline (PBS) buffer (1x, pH 7.5, 137 mM NaCl) – to avoid the use of amine
or carboxylate buffers that could compete with the antibody for the Ugi reaction – and average degrees
of conjugation (DoC) and conversion rates were determined by native mass spectrometry (native MS)
by comparing peak intensities. We were delighted to see that even low concentrations of reagents –
from 1.4 mM (20 equiv.) to 2.7 mM (40 equiv.) – gave positive results in just 16 h at 25 °C without the
38,39
need of any catalyst or imine preformation.
We also observed a strong influence of the aldehyde-to-
isocyanide ratio on the reaction outcome, with better conversion rates and DoC obtained with equimolar
or near equimolar ratios. Increasing the amounts of both aldehyde and isocyanide led to higher DoC
and conversion rates, however, only to a certain extent: a plateau was reached when large excess were
used (i.e. ≥60:120 equiv. 1a/2a), possibly indicating a limit in the number of functionalization sites on
trastuzumab. This hypothesis was reinforced when looking at the conjugates distribution: for the highest
conversions (i.e. >80%), the maximal DoC value observed was only 6 and represented less than 5% of
all conjugates. Excellent conversion could be obtained with a narrower DoC distribution when 45
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
equivalents of both 1a and 2a were employed, with adducts with DoC > 3 (D4 and D5; Figure 2, b)
representing a negligible fraction of all conjugated species. Control experiments were also done at this
stage, where trastuzumab was reacted in the absence of either aldehyde 1a or isocyanide 2a, leading
to no conjugation in both cases and thus proving the multicomponent aspect of the reaction (see ESI
and Figure S2 for more details). Prolonging the reaction time up to 2.5 days or heating at 37 °C resulted
in improved conversions but extended DoC distribution – up to D4 when a 20:40 1a/2a ratio was utilised
–
while performing the reaction at 4 °C had the opposite effect. On the other hand, varying the NaCl
concentration or the pH of the buffer solution led to no noticeable change (see ESI and Figure S3 for
more details). Thus, optimal conditions for the conjugation of trastuzumab were determined to be 45
equivalents of both 1a and 2a in PBS buffer (1x, 137 mM of NaCl, pH 7.5), with a reaction time of 16
hours at 25 C, offering the best compromise between conversion and DoC distribution.
a
1a (equiv.)
trastuzumab
3aa
8
9%90%89%
3.0
2.0
1.0
0.0
8
0%
2.5
80%
60%
7
0%
2.4
2
.3
6
2%63%
5
7%
56%
54%
1
.6
4
1%
40%
1
.3
40%
3
1%
1.1 1.1
.0
1
0.9
0.8
0
.6
1
4%15% 0.5
20%
0%
9
%
0.4
0
.1 0.2
0
.1
1a/2a ratio
b
Zoom
1
45,9
D
35
0
±
3
Da
20:40 1a/2a
average DoC = 0.4
1% conversion
D
D1
1
45:45 1a/2a
1
00
100
average DoC = 1D.03=
3
70% conversion D1
=
=
=
=
=
D0
D2
D3
D4
D0
D1
6,473
1
4
D
1
±
2
Da
D2
D2
D5
%
%
D2
D3
D2
D3
1
47,016 ± 4 Da
D4
D4
D5
D5
0
0
6
050
6100
6150
6200
6050
6100
6150
6200
6250
m/z
m/z
Figure 2: Ugi four-centre three-component reaction on trastuzumab. (a) Influence of aldehyde 1a to isocyanide 2a ratio on
trastuzumab’s average DoC, determined by native MS. (b) Native mass spectra of the resulting functionalized trastuzumab after
deglycosylation (zoom on the 24+ charged state); reaction conditions: 1a (20 or 45 equiv.), 2a (40 or 45 equiv.), PBS 1x (137 mM
NaCl, pH 7.5), 25 °C, 16 h; conversion = 100 - %D0.
With this set of conditions in hands, we then turned our attention towards the development of a ‘plug-
15,40,41
and-play’ version of our strategy,
by first incorporating an azide group onto trastuzumab – via our
multicomponent strategy – that could then be easily functionalized by strain-promoted azide-alkyne
cycloaddition (SPAAC) with various payloads (Figure 3, a). For practical synthetic reasons, it was found
easier to have the azide group borne by the aldehyde component and azide-containing aldehyde 1b
was thus easily synthesized in three steps from commercially available tetraethylene glycol and reacted
with 2a and trastuzumab under our optimal conditions. The resulting conjugate 3ba was then purified
and subjected to SPAAC with the bicyclononyne (BCN) derivative 4a, equipped with a cyanine-5
fluorophore, delivering the functionalized adduct 5 whose average DoC was determined by native MS.
It is important to stress that partial precipitation was always observed due to the hydrophobic nature of
the cyanine-5 payload, thus slightly impacting the average DoC and conversion rate values.
Screening different aldehyde-to-isocyanide ratios, we found that conversion and average DoC values
were slightly lower with 1b than with 1a in general (Figure 3, b). However, DoC distribution turned out
to be better, with D4 being the highest conjugated species observed across all conducted experiments
and representing <8% of all conjugates. Stability studies of conjugate 5 (see ESI and Figure S4 for more
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
details), conducted for five days in different media, showed no apparent decomposition in human plasma
and a good stability in PBS at various pH.
a
trastuzumab
3ba
5
D0
145,865 ± 2 Da
D0
1
45,865 ± 2 Da
b
100
100
D0
86%
8
6
4
0%
0%
0%
7
2%
2
1
0
69%
1.7
5
8%
5
2%
D1
46,157 ± 2 Da
1
1
.2 1.2
4
1%
3
8%
%
%
2
8%
0.8
D1
0
.7
1
47,019 ± 3 Da
0
.5 0.5
0
.4
20%
0%
D2
*
146,433 ± 3 Da
*
*
*
**
0
0
6
050
6100
6150
6200
6
050
6100
6150
6200
1b / 2a ratio
m/z
m/z
Figure 3: ‘Plug-and-play’ strategy: (a) conjugation of trastuzumab with the azide-containing aldehyde 1b and isocyanide 2a
followed by functionalization with strained alkyne 4a and native mass spectra of the different adducts after deglycosylation (zoom
on the 24+ charged state – reaction conditions corresponding to the spectra: “plug step”: 1b (45 equiv.), tBuNC 2a (45 equiv.) in
PBS 1x (137 mM of NaCl, pH 7.5), 25 °C, 16 h; “play step”: 4a (10 equiv.) in PBS 1x (137 mM of NaCl, pH 7.5), 25 °C, 20 h); (b)
optimization of the “plug step” (average DoC and conversion rates determined by native MS analysis of the corresponding
conjugates 5 after deglycosylation).
Having developed a set of robust and modular conditions leading to stable constructs, the versatility of
our strategy at both the plug and the play stages was then evaluated by screening various isocyanides
and diversely functionalized strained alkynes. As shown in Figure 4, the reaction proceeded smoothly
and gave better average DoC values with commercially available isopropyl isocyanide 2b, cyclohexyl
isocyanide 2c and ethyl isocyanoacetate 2e, while diethyl isocyanomethylphosphonate 2f led to no
visible conjugation. Different isocyanides were also newly synthesized with the aim of incorporating
other reactive groups onto trastuzumab: 2g was accessed in a single step from 2e, while isocyanides
42
2
h-2m were easily generated from the known benchstable p-nitrophenol carbonate 2n. Mixed results
were obtained for the conjugation of trastuzumab with this set of compounds however, with 2i being the
only isocyanide to give acceptable conversion and average DoC values. Overall, small and apolar
isocyanides seem to work best for this transformation, which could possibly indicate that reactive
conjugation sites are located in hydrophobic pockets.
As for the play stage, starting from conjugate 3bc, complete conversions for the SPAAC reaction were
consistently observed with BCN motifs, allowing the introduction of various payloads onto the antibody:
mass spectrometry and NMR tags (4b and 4d), iminobiotin (4c), or 20-mer oligonucleotides (4e and 4f).
Switching to dibenzocyclooctyne (DBCO) groups, we successfully generated the antibody-drug
conjugate (ADC) 5g by incorporating the monomethyl auristatin E (MMAE) drug (4g), connected to the
strained alkyne scaffold by a cleavable valine-citruline linker. Analysis by native MS after purification by
gel filtration chromatography showed an incomplete conversion during the SPAAC reaction but an
average drug-to-antibody ratio (DAR) of 1.4 – a higher value than previously noted for 5, which could
be explained by an absence of precipitation during the SPAAC with 4g. The cytotoxicity of the ADC 5g
was assayed on the HER2-positive SK-BR-3 and the HER2-negative MDA-MB-231 cancer cell lines,
using the FDA-approved trastuzumab emtansine (T-DM1) as a benchmark. Gratifyingly, 5g was found
to be at least as active and selective as T-DM1, with comparable EC50 values on HER2-positive cells
and a lack of apparent toxicity on HER2-negative cells (Figure 4, e and Figure S6).
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
5
a-5g, R1 = cHex
trastuzumab
3ba-3bl
a
b 2
57%57%
60%
53%
51%
48%
4
5%
2a
2b
2c
2d
2e
2f
2g
41%
4
0%
0%
27%
1
0
0.9
0.9
0.8
0.7
19%21%20%
0.6
2h
2i
0.6
2
0.5
8
%
0.3
0.2 0.2 0.2
0
.1
0%
2m
2j
2k
2l
2a 2b 2c 2d 2e 2f 2g 2h 2i 2j 2k 2l
c
4a
4b
4c
4d
d
M
T
5e 5f
kDa
2
50
1
1
50
00
4e
7
5
5
3
0
7
4g
e
EC50 (pM)
g (DAR 1.4) T-DM1 (DAR 3.6)
2
2
5
0
5
4f
1
5
SK-BR-3
66 14
174 26
10
MDA-MB-231
nt
nt
Figure 4: Isocyanide and strained alkynes scope. (a) Isocyanide scope. Reaction conditions: i) 1b (45 equiv.), 2a-2l (45 equiv.),
PBS 1x (137 mM NaCl, pH 7.5), 25 °C, 16 h; ii) 4a (10 equiv.), PBS 1x (137 mM NaCl, pH 7.5), 25 °C, 20 h. (b) Conversion and
average DoC values obtained with 2a-2l. (c) Strained alkynes scope. Reaction conditions: i) 1b (45 equiv.), 2c (45 equiv.), PBS
1x (137 mM NaCl, pH 7.5), 25 °C, 16 h; ii) 4a-4g (10 equiv.), PBS 1x (137 mM NaCl, pH 7.5), 25 °C, 20 h. (d) Denaturing SDS
PAGE analysis of conjugates 5e and 5f (M: molecular weight marker; T: trastuzumab): the multiple bands correspond to the
different DoC. (e) Cytotoxicity studies of the ADC 5g on SK-BR-3 and MDA-MB-231 cancer cell lines (nt = non-toxic) – EC50 ± SD
values from two independent experiments.
While the plug-and-play strategy offers a simple and convenient way for the post-derivatization of
trastuzumab conjugate 3bc, the use of isocyanides 2g-2k, bearing another reactive group, gives an
excellent opportunity for dual, and potentially orthogonal, functionalization. We hypothesized that such
a double derivatization could be done by either: i) SPAAC followed by copper(I)-catalysed azide alkyne
cycloaddition (CuAAC), by combining aldehyde 1b with isocyanides 2g or 2h; ii) SPAAC followed by a
[
4+1] or a [4+2] cycloaddition reaction, with 1b and 2i or 2k respectively; or iii), in a non-orthogonal
manner, with a double SPAAC, using 1b and 2j, for the addition of two equivalents of the same strained
alkyne payload. In the latter case, double functionalization of conjugate 3bj with complete conversion
was successfully conducted with alkynes 4a and 4c, allowing the incorporation of two identical payloads
per conjugation site (Figure 5, b). However, orthogonal dual labelling, as detailed in i) and ii), was only
met with failure despite our numerous attempts (see ESI for more details), urging us to seek an
alternative route towards this goal. We thus elected aldehyde 1a and isocyanide 2j to first conduct the
multicomponent conjugation step, giving access to azide-containing conjugate 3aj. Subsequent SPAAC
reaction with alkyne 4c resulted in complete conversion to adduct 5j, successfully incorporating both a
fluorescent TAMRA probe and an iminobiotin (Figure S10). While only alkyne 4c was evaluated in this
case, the excellent results obtained with such a strategy should easily give access to a plethora of dually
functionalized conjugates by simply varying the alkyne component. This strategy thus joins the limited
existing group of site-selective dual conjugation methods applicable to native proteins, and most notably,
15,43,44
represents the first one to not target cysteine residues to the best of our knowledge.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
a
3bj
5h-5i
4a or 4c
PBS (1X, pH 7.5), 25 ꢀC, 20 h
5h,
= Cy 5
5i,
= iminobiotin
b
4c
PBS (1X, pH 7.5),
5 ꢀC, 20 h
3aj
5j
2
=
iminobiotin
Figure 5: Dual functionalization studies. (a) Homo functionalization by double SPAAC reaction; reaction conditions: 4a or 4c
(20 equiv.), PBS 1x (137 mM NaCl, pH 7.5), 25 °C, 20 h. (b) Hetero functionalization; reaction conditions: 4c (20 equiv.), PBS 1x
(137 mM NaCl, pH 7.5), 25 °C, 20 h.
Having screened various isocyanides and investigated the diversity of payloads that could be
incorporated at the play stage of our ‘plug-and-play’ strategy, we set out to study the aldehyde scope of
this transformation, utilising 2i as a model isocyanide. The previously mentioned TAMRA aldehyde 1a
and commercially available aldehydes 1c-1r were thus evaluated and a strong influence of
stereoelectronic effects on average DoC and conversion values were observed (Figure 6, a). Degrees
and patterns of substitution were found to play an important role as exemplified by linear valeraldehyde
1c giving poor results compared to its branched analogue 1d, or by pivaldehyde 1g, showing a limited
reactivity as opposed to isobutyraldehyde 1e or methyl isopropyl ketone 1f. Interestingly,
cyclohexanecarboxaldehyde 1h yielded better conversion and narrower DoC distribution (D1 to D5) than
1
1
e, a feature already observed when isocyanides 2b and 2c were compared. More polar aldehydes 1i-
l gave contrasted results: while both glucose and glutaraldehyde failed to conjugate trastuzumab,
formaldehyde and glyceraldehyde performed very well, even though the latter gave mass spectra of
poor quality, possibly indicating side reactions. Screening aromatic aldehydes, the previously reported
19
2
-pyridinecarboxyaldehyde 1m was the only one to give positive results, the five others (1n-1r) leading
at best to mediocre conversions and DoC values. While this could be explained by a lack of solubility
for compounds 1q and 1r, the inefficacy of 1n and 1o is puzzling, especially considering their known
45
use for the bioconjugation of proteins. Taken together, these results tend to confirm our hypothesis
that the reactive bioconjugation sites could be located in hydrophobic clefts, explaining why small and
flexible apolar carbonyls and isocyanides seem to be the best fit for this reaction.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
R3
=
trastuzumab
3ai-3ri
9
7%97%
5
4
3
2
1
0
100%
80%
60%
40%
9
2%
8
4%
9%
1
c
1d
1e
1f
1j
88%
8
4.0
1a
3
.4
62%
1
g
1h
1ia
4
8%
4
6%
2
.2
2.0
2.1
1b
1
.7
1k
1l
1m
2
5%
0.9
17%
17%
20%
0.6
0.6
0.3
5%
0.2
0.2
0
.1
0%
1n
1o
1p
1q
1r
1a 1b 1c 1d 1e 1f 1g 1h 1i 1j 1k 1l 1m 1n 1o 1p 1q 1r
Figure 6: Aldehyde and protein scopes of the multicomponent bioconjugation reaction. (a) Aldehyde scope. Reaction conditions:
a
1
a-1r (45 equiv.), 2i (45 equiv.), PBS 1x (137 mM NaCl, pH 7.5), 25 °C, 16 h; : 5 equiv. used
Having developed an efficient, operationally simple and quick multicomponent bioconjugation strategy
that tolerates various reactive groups and substrates, we turned to LC-MS/MS studies to validate the
mechanism of the bioconjugation reaction and localize the exact conjugation sites in order to find out
whether our strategy could be coined site-specific. Peptide mapping was thus conducted on three
separate batches that were prepared and purified in different manners. A first batch of 5a with an
average DoC of 0.6 was generated from conjugate 3bc after SPAAC with the BCN-cyanine-5 4a and
analysed without further purification (protein sequence coverage: 100%). Due to the presence of
unconjugated trastuzumab yielding peptides that outcompeted the conjugated fragments for mass
detection, two enriched batches were produced by two similar but distinct routes (protein sequence
coverage: 31-48%). A first strategy utilized conjugate 5c – synthesized from 3bc and the BCN-biotin
derivative 4c. Trypsin digestion followed by affinity purification of the digestate on streptavidin column
gave a first enriched sample. A second strategy inverted the order of events: conjugate 3bc was
digested before the functionalization with strained alkyne 4c. The resulting mixture of peptides was then
enriched in the same manner as for the first strategy (see ESI for more details). Taken altogether, these
analyses highlighted six conjugation sites, only two of which being detected in all three batches: the N-
terminal glutamate (E1) of the heavy chain, and the N-terminal aspartate (D1) of the light chain.
Remarkably, both of these residues partook in two different types of multicomponent reaction: an
intramolecular U-4C-3R, yielding β- and γ-lactams (from D1 and E1 respectively), and a Passerini
reaction, involving only the side chains carboxylate functions (detected only in the two enriched samples
for D1, but in all three batches for E1). The four other conjugation sites were all detected in only a single
batch and could thus not be cross-validated. Three of them were found to correspond to glutamate
residues modified by a Passerini reaction: the light-chain E195, and the heavy-chain E236 and E348.
The expected inter-residue U-4C-3R was finally found to be responsible for the fourth conjugation site
detected, located on the light chain and involving the side chains of K126 and either D122 or E123
(
undistinguishable from each other). While such inter-chain conjugation could impact the protein’s
folding state, analyses of native trastuzumab and conjugate 3bc by circular dichroism spectroscopy
showed no influence of the chemical conjugation on the antibody’s secondary structure (see ESI).
Interestingly, the heavy-chain C-terminal lysine (K478) did not react in either a U-4C-3R or a Passerini
reaction, and was found to be unaffected. Finally, no intermolecular reactions (i.e. cross-linking between
two antibodies) could ever be detected, which could be explained by the high-dilution conditions coupled
with the prohibitive steric repulsion that is required to drive the reaction between two proteins in the
absence of a chemical spacer or linker.
While the number of modification sites on trastuzumab is certainly too high to call this strategy site-
specific, the N-terminal modification of both chains coupled to the excellent aspartate/glutamate
selectivity and the first example of a dual residue modification represent extremely interesting features
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
that could open new avenues towards the development of native antibodies site-specific conjugation
strategies.
D1
E195
E236
E1
K126
E123
E348
D122
single residue modification
dual residue modification
Q124-L125
R1
=
Passerini
U-4C-3R
D1 & E195 (LC)
E1, E236 & E348 (HC)
D1 (LC)
E1 (HC)
K126-D122 / E123 (LC)
Figure 7: Trastuzumab conjugation sites determined by peptide mapping
In conclusion, we reported the use of a new multicomponent strategy for the conjugation of trastuzumab.
By combining commercially available – or easily accessible – aldehydes and isocyanides, this modular
approach gives a facile and rapid access to functionalized antibodies incorporating various payloads.
Most notably, this strategy was used to synthesize an ADC that retained its selectivity towards HER2-
positive cancer cell lines and was found to be as active as the benchmark T-DM1. Two reactions were
proved to compete with each other in this conjugation process, the expected Ugi four-centre three-
component reaction (U-4C-3R) and the Passerini reaction, both of which occurred at the N-terminal
aspartate and glutamate residues. While not being fully effective at the present, such an approach could
be an excellent stepping stone towards the development of a yet unreported N-terminal-Asp/Glu
selective strategy. Moreover, this strategy represents the first example of an aspartate/glutamate
selective conjugation method that is not using activating agents for their intermediate conversion to
activated esters – strategies that often yield undesired protein cross-linking. Overall, these results offer
a new approach towards a site-specific bioconjugation strategy that could be applied to native proteins
and highlight the potential of multicomponent reactions for this purpose.
Supporting Information
The Supporting Information includes: Experimental procedures and characterization, additional
information, Figures S1–S12, Tables S1–S4, NMR and native MS spectra of newly synthesized
compounds.
Author Information
Corresponding Author:
*chaubet@unistra.fr
Author Contributions
The work was conceived by G.C. All reactions were carried out by C.S. Cytotoxic studies were done by
I.D. Native mass spectrometry analyses and peptide mapping were performed by S.H., S.E, A. E. and
T. B. The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
Funding Sources
This work was supported by the CNRS, the University of Strasbourg and the French Proteomic
Infrastructure (ProFI; ANR-10-INBS-08-03). The authors would like to thank GIS IBiSA and Rꢀgion
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
Alsace for financial support in purchasing Synapt G2 HDMS and Orbitrap Exactive Plus EMR
instruments, respectively. C.S. thanks the “Région Grand-Est” and “LabEx MEDALIS” for a PhD
studentship. A.E. acknowledges the “Association Nationale de la Recherche et de la Technologie”
(
ANRT) and Syndivia for funding his PhD fellowship.
Acknowledgment
We thank Marc Nothisen for cell cultures, Dr Nicolas Humbert for circular dichroism studies, and Dr
Oleksandr Koniev and Dr Sergii Kolodych for insightful discussions.
Conflicts of interest
There are no conflicts of interest to declare.
References
(1)
Harlan, J. W.; Feairheller, S. H. Chemistry of the Crosslinking of Collagen during Tanning. In Protein
Crosslinking; Friedman, M., Ed.; Springer US: Boston, MA, 1977; Vol. 86A, pp 425–440.
https://doi.org/10.1007/978-1-4684-3282-4_27.
(2)
(3)
(4)
(5)
(6)
(7)
Means, G. E.; Feeney, R. E. Chemical Modifications of Proteins: History and Applications. Bioconjug. Chem.
1
990, 1 (1), 2–12. https://doi.org/10.1021/bc00001a001.
Rees, A. R. The Antibody Molecule: From Antitoxins to Therapeutic Antibodies ; Oxford University Press,
014. https://doi.org/10.1093/med/9780199646579.001.0001.
Shreder, K. Synthetic Haptens as Probes of Antibody Response and Immunorecognition. Methods 2000,
0 (3), 372–379. https://doi.org/10.1006/meth.1999.0929.
2
2
Spicer, C. D.; Davis, B. G. Selective Chemical Protein Modification. Nat. Commun. 2014, 5, 4740.
https://doi.org/10.1038/ncomms5740.
deGruyter, J. N.; Malins, L. R.; Baran, P. S. Residue-Specific Peptide Modification: A Chemist’s Guide.
Biochemistry (Mosc.) 2017, 56 (30), 3863–3873. https://doi.org/10.1021/acs.biochem.7b00536.
Kim, M. T.; Chen, Y.; Marhoul, J.; Jacobson, F. Statistical Modeling of the Drug Load Distribution on
Trastuzumab Emtansine (Kadcyla), a Lysine-Linked Antibody Drug Conjugate. Bioconjug. Chem. 2014, 25
(7), 1223–1232. https://doi.org/10.1021/bc5000109.
(
8)
9)
Gautier, V.; Boumeester, A. J.; Lössl, P.; Heck, A. J. R. Lysine Conjugation Properties in Human IgGs
Studied by Integrating High-Resolution Native Mass Spectrometry and Bottom-up Proteomics.
PROTEOMICS 2015, 15 (16), 2756–2765. https://doi.org/10.1002/pmic.201400462.
Junutula, J. R.; Raab, H.; Clark, S.; Bhakta, S.; Leipold, D. D.; Weir, S.; Chen, Y.; Simpson, M.; Tsai, S. P.;
Dennis, M. S.; Lu, Y.; Meng, Y. G.; Ng, C.; Yang, J.; Lee, C. C.; Duenas, E.; Gorrell, J.; Katta, V.; Kim, A.;
McDorman, K.; Flagella, K.; Venook, R.; Ross, S.; Spencer, S. D.; Lee Wong, W.; Lowman, H. B.; Vandlen,
R.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Mallet, W. Site-Specific Conjugation of a Cytotoxic Drug
to an Antibody Improves the Therapeutic Index. Nat. Biotechnol. 2008, 26 (8), 925–932.
https://doi.org/10.1038/nbt.1480.
(
(
(
(
10)
11)
12)
Schumacher, D.; Hackenberger, C. P. R.; Leonhardt, H.; Helma, J. Current Status: Site-Specific Antibody
Drug Conjugates. J. Clin. Immunol. 2016, 36 (1), 100–107. https://doi.org/10.1007/s10875-016-0265-6.
Rabuka, D. Chemoenzymatic Methods for Site-Specific Protein Modification. Curr. Opin. Chem. Biol. 2010,
1
4 (6), 790–796. https://doi.org/10.1016/j.cbpa.2010.09.020.
Krall, N.; da Cruz, F. P.; Boutureira, O.; Bernardes, G. J. L. Site-Selective Protein-Modification Chemistry
for Basic Biology and Drug Development. Nat. Chem. 2016, (2), 103–113.
https://doi.org/10.1038/nchem.2393.
Stephanopoulos, N.; Francis, M. B. Choosing an Effective Protein Bioconjugation Strategy. Nat. Chem. Biol.
011, 7 (12), 876–884. https://doi.org/10.1038/nchembio.720.
8
(
13)
14)
2
(
Walsh, S. J.; Omarjee, S.; Galloway, W. R. J. D.; Kwan, T. T.-L.; Sore, H. F.; Parker, J. S.; Hyvönen, M.;
Carroll, J. S.; Spring, D. R. A General Approach for the Site-Selective Modification of Native Proteins,
Enabling the Generation of Stable and Functional Antibody–Drug Conjugates. Chem. Sci. 2019, 10 (3),
6
94–700. https://doi.org/10.1039/C8SC04645J.
(15)
(16)
(17)
(18)
Maruani, A.; Smith, M. E. B.; Miranda, E.; Chester, K. A.; Chudasama, V.; Caddick, S. A Plug-and-Play
Approach to Antibody-Based Therapeutics via a Chemoselective Dual Click Strategy. Nat. Commun. 2015,
6
(1). https://doi.org/10.1038/ncomms7645.
Smith, M. E. B.; Schumacher, F. F.; Ryan, C. P.; Tedaldi, L. M.; Papaioannou, D.; Waksman, G.; Caddick,
S.; Baker, J. R. Protein Modification, Bioconjugation, and Disulfide Bridging Using Bromomaleimides. J.
Am. Chem. Soc. 2010, 132 (6), 1960–1965. https://doi.org/10.1021/ja908610s.
Lee, M. T. W.; Maruani, A.; Baker, J. R.; Caddick, S.; Chudasama, V. Next-Generation Disulfide Stapling:
Reduction and Functional Re-Bridging All in One. Chem. Sci. 2015,
7
(1), 799–802.
https://doi.org/10.1039/C5SC02666K.
Gilmore, J. M.; Scheck, R. A.; Esser‐Kahn, A. P.; Joshi, N. S.; Francis, M. B. N-Terminal Protein Modification
through a Biomimetic Transamination Reaction. Angew. Chem. Int. Ed. 2006, 45 (32), 5307–5311.
https://doi.org/10.1002/anie.200600368.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
(19)
(20)
(21)
(22)
MacDonald, J. I.; Munch, H. K.; Moore, T.; Francis, M. B. One-Step Site-Specific Modification of Native
Proteins with 2-Pyridinecarboxyaldehydes. Nat. Chem. Biol. 2015, 11 (5), 326–331.
https://doi.org/10.1038/nchembio.1792.
Kodal, A. L. B.; Rosen, C. B.; Mortensen, M. R.; Tørring, T.; Gothelf, K. V. DNA-Templated Introduction of
an
https://doi.org/10.1002/cbic.201600254.
Yu, C.; Tang, J.; Loredo, A.; Chen, Y.; Jung, S. Y.; Jain, A.; Gordon, A.; Xiao, H. Proximity-Induced Site-
Specific Antibody Conjugation. Bioconjug. Chem. 2018, 29 (11), 3522–3526.
https://doi.org/10.1021/acs.bioconjchem.8b00680.
Aldehyde
Handle
in
Proteins.
ChemBioChem
2016,
17
(14),
1338–1342.
Matos, M. J.; Oliveira, B. L.; Martínez-Sáez, N.; Guerreiro, A.; Cal, P. M. S. D.; Bertoldo, J.; Maneiro, M.;
Perkins, E.; Howard, J.; Deery, M. J.; Chalker, J. M.; Corzana, F.; Jiménez-Osés, G.; Bernardes, G. J. L.
Chemo- and Regioselective Lysine Modification on Native Proteins. J. Am. Chem. Soc. 2018, 140 (11),
4
004–4017. https://doi.org/10.1021/jacs.7b12874.
(23)
(24)
(25)
Adusumalli, S. R.; Rawale, D. G.; Singh, U.; Tripathi, P.; Paul, R.; Kalra, N.; Mishra, R. K.; Shukla, S.; Rai,
V. Single-Site Labeling of Native Proteins Enabled by a Chemoselective and Site-Selective Chemical
Technology. J. Am. Chem. Soc. 2018, 140 (44), 15114–15123. https://doi.org/10.1021/jacs.8b10490.
Sato, S.; Matsumura, M.; Kadonosono, T.; Abe, S.; Ueno, T.; Ueda, H.; Nakamura, H. Site-Selective Protein
Chemical Modification of Exposed Tyrosine Residues Using Tyrosine Click Reaction. Bioconjug. Chem.
2
020. https://doi.org/10.1021/acs.bioconjchem.0c00120.
Lee, M. T. W.; Maruani, A.; Richards, D. A.; Baker, J. R.; Caddick, S.; Chudasama, V. Enabling the
Controlled Assembly of Antibody Conjugates with a Loading of Two Modules without Antibody Engineering.
Chem. Sci. 2017, 8 (3), 2056–2060. https://doi.org/10.1039/C6SC03655D.
(
26)
27)
Hoyt, E. A.; Cal, P. M. S. D.; Oliveira, B. L.; Bernardes, G. J. L. Contemporary Approaches to Site-Selective
Protein Modification. Nat. Rev. Chem. 2019, 3 (3), 147–171. https://doi.org/10.1038/s41570-019-0079-1.
Adusumalli, S. R.; Rawale, D. G.; Thakur, K.; Purushottam, L.; Reddy, N. C.; Kalra, N.; Shukla, S.; Rai, V.
Chemoselective and Site‐Selective Lysine‐Directed Lysine Modification Enables Single‐Site Labeling of
Native Proteins. Angew. Chem. Int. Ed. 2020. https://doi.org/10.1002/anie.202000062.
(
(28)
(29)
(30)
(31)
Chilamari, M.; Kalra, N.; Shukla, S.; Rai, V. Single-Site Labeling of Lysine in Proteins through a Metal-Free
Multicomponent
Approach.
Chem.
Commun.
2018,
54
(53),
7302–7305.
https://doi.org/10.1039/C8CC03311K.
Hansen, B. K.; Loveridge, C. J.; Thyssen, S.; Wørmer, G. J.; Nielsen, A. D.; Palmfeldt, J.; Johannsen, M.;
Poulsen, T. B. STEFs: Activated Vinylogous Protein-Reactive Electrophiles. Angew. Chem. Int. Ed. 2019,
5
8 (11), 3533–3537. https://doi.org/10.1002/anie.201814073.
Medeiros, G. A.; Silva, W. A. da; Bataglion, G. A.; Ferreira, D. A. C.; Oliveira, H. C. B. de; Eberlin, M. N.;
Neto, B. A. D. Probing the Mechanism of the Ugi Four-Component Reaction with Charge-Tagged Reagents
by ESI-MS(/MS). Chem. Commun. 2013, 50 (3), 338–340. https://doi.org/10.1039/C3CC47156J.
Iacobucci, C.; Reale, S.; Gal, J.-F.; Angelis, F. D. Insight into the Mechanisms of the Multicomponent Ugi
and Ugi–Smiles Reactions by ESI-MS(/MS). Eur. J. Org. Chem. 2014, 2014 (32), 7087–7090.
https://doi.org/10.1002/ejoc.201403179.
(
32)
33)
Marek, M.; Jarý, J.; Valentová, O.; Vodrážka, Z. Immobilization of Glycoenzymes by Means of Their
Glycosidic Components. Biotechnol. Lett. 1983, 5 (10), 653–658. https://doi.org/10.1007/BF01386357.
Ziegler, T.; Gerling, S.; Lang, M. Preparation of Bioconjugates through an Ugi Reaction. Angew. Chem. Int.
Ed. 2000, 39 (12), 2109–2112. https://doi.org/10.1002/1521-3773(20000616)39:12<2109::AID-
ANIE2109>3.0.CO;2-9.
(
(
34)
Méndez, Y.; Chang, J.; Humpierre, A. R.; Zanuy, A.; Garrido, R.; Vasco, A. V.; Pedroso, J.; Santana, D.;
Rodríguez, L. M.; García-Rivera, D.; Valdés, Y.; Vérez-Bencomo, V.; Rivera, D. G. Multicomponent
Polysaccharide–Protein Bioconjugation in the Development of Antibacterial Glycoconjugate Vaccine
Candidates. Chem. Sci. 2018, 9 (9), 2581–2588. https://doi.org/10.1039/C7SC05467J.
García, A.; Hernández, K.; Chico, B.; García, D.; Villalonga, M. L.; Villalonga, R. Preparation of
Thermostable Trypsin–Polysaccharide Neoglycoenzymes through Ugi Multicomponent Reaction. J. Mol.
Catal. B Enzym. 2009, 59 (1–3), 126–130. https://doi.org/10.1016/j.molcatb.2009.02.001.
Camacho, C.; Matías, J. C.; García, D.; Simpson, B. K.; Villalonga, R. Amperometric Enzyme Biosensor for
Hydrogen Peroxide via Ugi Multicomponent Reaction. Electrochem. Commun. 2007, 9 (7), 1655–1660.
https://doi.org/10.1016/j.elecom.2007.03.013.
(
35)
36)
(
(37)
(38)
(39)
(40)
Nahta, R.; Esteva, F. J. Trastuzumab: Triumphs and Tribulations. Oncogene 2007, 26 (25), 3637–3643.
https://doi.org/10.1038/sj.onc.1210379.
Marcaccini, S.; Torroba, T. The Use of the Ugi Four-Component Condensation. Nat. Protoc. 2007, 2 (3),
6
32–639. https://doi.org/10.1038/nprot.2007.71.
Dömling, A.; Ugi, I. Multicomponent Reactions with Isocyanides. Angew. Chem. Int. Ed. 2000, 39 (18),
168–3210. https://doi.org/10.1002/1521-3773(20000915)39:18<3168::AID-ANIE3168>3.0.CO;2-U.
3
Dovgan, I.; Ursuegui, S.; Erb, S.; Michel, C.; Kolodych, S.; Cianférani, S.; Wagner, A. Acyl Fluorides: Fast,
Efficient, and Versatile Lysine-Based Protein Conjugation via Plug-and-Play Strategy. Bioconjug. Chem.
2
017, 28 (5), 1452–1457. https://doi.org/10.1021/acs.bioconjchem.7b00141.
(41)
Dovgan, I.; Erb, S.; Hessmann, S.; Ursuegui, S.; Michel, C.; Muller, C.; Chaubet, G.; Cianférani, S.; Wagner,
A. Arginine-Selective Bioconjugation with 4-Azidophenyl Glyoxal: Application to the Single and Dual
Functionalisation of Native Antibodies. Org. Biomol. Chem. 2018, 16 (8), 1305–1311.
https://doi.org/10.1039/C7OB02844J.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
(42)
Stöckmann, H.; Neves, A. A.; Stairs, S.; Brindle, K. M.; Leeper, F. J. Exploring Isonitrile-Based Click
Chemistry for Ligation with Biomolecules. Org. Biomol. Chem. 2011,
9
(21), 7303–7305.
https://doi.org/10.1039/C1OB06424J.
(
43)
44)
Maruani, A.; Richards, D. A.; Chudasama, V. Dual Modification of Biomolecules. Org. Biomol. Chem. 2016,
1
4 (26), 6165–6178. https://doi.org/10.1039/C6OB01010E.
(
Agrawalla, B. K.; Wang, T.; Riegger, A.; Domogalla, M. P.; Steinbrink, K.; Dörfler, T.; Chen, X.; Boldt, F.;
Lamla, M.; Michaelis, J.; Kuan, S. L.; Weil, T. Chemoselective Dual Labeling of Native and Recombinant
Proteins. Bioconjug. Chem. 2018, 29 (1), 29–34. https://doi.org/10.1021/acs.bioconjchem.7b00675.
Cambray, S.; Gao, J. Versatile Bioconjugation Chemistries of Ortho-Boronyl Aryl Ketones and Aldehydes.
Acc. Chem. Res. 2018, 51 (9), 2198–2206. https://doi.org/10.1021/acs.accounts.8b00154.
(45)
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.202002432
Investigating Ugi / Passerini multicomponent reactions for the Site-Selective
Conjugation of Native Trastuzumab
1
2
2
1
2
Charlotte Sornay , Steve Hessmann , Stéphane Erb , Igor Dovgan , Anthony Ehkirch , Thomas
2
2
1
1*
Botzanowski , Sarah Cianférani , Alain Wagner , Guilhem Chaubet
Abstract. Ugi and Passerini reactions were found
to be effective strategies for the conjugation of
native trastuzumab. The use of functionalized
aldehydes and isocyanides allows for the smooth
introduction of various payloads onto the
antibody, including the cytotoxic drug MMAE.
Ugi and Passerini reactions for the conjugation of native antibody
broad scope of aldehydes and isonitriles
doubly labelled conjugates
This article is protected by copyright. All rights reserved.