Design, Synthesis, and Self-Assembly Studies of a Suite of Monodisperse, Facially Amphiphilic, Protein–Dendron Conjugates

Article abstract of DOI:10.1002/cbic.201900341

The custom design of protein–dendron amphiphilic macromolecules is at the forefront of macromolecular engineering. Macromolecules with this architecture are very interesting because of their ability to self-assemble into various biomimetic nanoscopic structures. However, to date, there are no reports on this concept due to technical challenges associated with the chemical synthesis. Towards that end, herein, a new chemical methodology for the modular synthesis of a suite of monodisperse, facially amphiphilic, protein–dendron bioconjugates is reported. Benzyl ether dendrons of different generations (G1–G4) are coupled to monodisperse cetyl ethylene glycol to form macromolecular amphiphilic activity-based probes (AABPs) with a single protein reactive functionality. Micelle-assisted protein labeling technology is utilized for site-specific conjugation of macromolecular AABPs to globular proteins to make monodisperse, facially amphiphilic, protein–dendron bioconjugates. These biohybrid conjugates have the ability to self-assemble into supramolecular protein nanoassemblies. Self-assembly is primarily mediated by strong hydrophobic interactions of the benzyl ether dendron domain. The size, surface charge, and oligomeric state of protein nanoassemblies could be systematically tuned by choosing an appropriate dendron or protein of interest. This chemical method discloses a new way to custom-make monodisperse, facially amphiphilic, protein–dendron bioconjugates.

Full text of DOI:10.1002/cbic.201900341

Accepted Article
Title: Design, Synthesis and Self-assembly Studies of Suite of
Monodisperse Facially Amphiphilic Protein-Dendron Conjugates
Authors: Britto Sandanaraj, Pavankumar Janardhan Bhandari,
Mullapudi Mohan Reddy, Akshay Bhagwan Lohote, and
Bankanidhi Sahoo
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: ChemBioChem 10.1002/cbic.201900341
Link to VoR: http://dx.doi.org/10.1002/cbic.201900341
A Journal of
www.chembiochem.org

10.1002/cbic.201900341
ChemBioChem
Design, Synthesis and Self-assembly Studies of Suite of Monodisperse Facially
Amphiphilic Protein-Dendron Conjugates
Britto S Sandanaraj1,2#, Pavankumar Janardhan Bhandari1*, Mullapudi Mohan Reddy1*, Akshay
Bhagwan Lohote1 and Bankanidhi Sahoo3
1Departments of Chemistry1 & Biology2, Indian Institute of Science Education and Research (IISER),
Pune, 411 008, India
3Tata Institute of Fundamental Research Hyderabad (TIFR Hyd), Hyderabad 500019, India
Abstract
Custom design of protein-dendron amphiphilic macromolecules is at the forefront of macromolecular
engineering. Macromolecules having this architecture are very interesting because of their ability to
self-assemble into various biomimetic nanoscopic structures. However, to date, there are no reports
based on this concept due to the technical challenges associated with the chemical synthesis. Towards
that end, herein, we disclose a novel chemical methodology for the modular synthesis of a suite of
monodisperse facially amphiphilic “protein-dendron” bioconjugates. Benzyl-ether dendrons of the
different generation (G1-G4) are coupled to monodisperse undeca(ethylene glycol) to make
macromolecular amphiphilic activity-based probe (AABP) having a single protein reactive
functionality. Micelle-assisted protein labeling technology is utilized for site-specific conjugation of
macromolecular AABPs to globular proteins to make monodisperse facially amphiphilic protein-
dendron bioconjugates. These bio-hybrid conjugates have the ability to self-assemble into
supramolecular protein nanoassemblies. The self-assembly is primarily mediated by the strong
hydrophobic interaction of benzyl-ether dendron domain. The size, surface charge, and oligomeric state
of protein nanoassemblies could be systematically tuned by choosing an appropriate dendron or protein
of interest. This chemical method discloses a new way to custom-make monodisperse facially
amphiphilic protein-dendron bioconjugates.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Introduction
Natural biomolecular assemblies such as viruses,1 molecular motors,2 protein cages,3 and bacterial
compartments4 are well defined, monodisperse, dynamic architectures that exhibit higher-order
functions. These natural biomolecular assemblies serve as an inspiration for the custom design of
synthetic supramolecular assemblies using nucleic acids,5,6 peptides,7 and proteins8 as building blocks.
Although the majority of naturally occurring biomolecular assemblies are built using proteins as
building blocks, biomimicry of this phenomenon is highly challenging. At the same time, nucleic acid5,6
and peptides7 are routinely used as building blocks for this purpose. However, recent developments in
computational protein design and rational approaches using genetic methods have made substantial
progress in this field.9-12 Although useful, these methods require sophisticated infrastructure, associated
with a high failure rate and sometimes yield structures not intended by design13. In parallel, chemical
strategies are also emerging as complementary methods for the design of protein with improved
functions.14,15
Pioneering work by Nolte and others have demonstrated novel chemical strategies for attachment of
hydrophobic synthetic polymers on to various proteins exploiting both covalent and non-covalent
chemistry.16-20 These amphiphilic bioconjugates have shown to self-assemble into various nanoscopic
structures driven via hydrophobic interactions. As a complementary method, polymers that exhibit
lower-critical solution temperature (LCST) property were used to construct protein-polymer
bioconjugates through the biosynthetic and/or chemical methodology. Below LCST, both the protein
and polymer domains are hydrophilic, however, above LCST, polymer domain becomes hydrophobic
and thereby induce global amphiphilicity, which imparts self-assembling property.21-28
Among the various bioconjugation methods available to make protein-polymer bioconjugates, chemical
modification of cysteine residue of a protein is a popular and straight-forward method.17,26-28 However,
this method has a little scope because it is applicable to a few proteins having a single cysteine residue
on their surface. Genetic incorporation of unnatural amino acids containing bioorthogonal functional
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
groups would expand the scope of site-specific bioconjugation chemistry to a large number of
proteins.29 However, the poor efficiency of unnatural amino acid incorporation on to protein and the
tedious process associated with this methodology limits the scope of this technology.30 In addition, most
of the studies reported so far exclusively use linear polymers to construct protein-polymer
bioconjugates. A significant disadvantage of this system is that the linear polymers are not
monodisperse and hence resultant nanoassemblies are polydisperse. A simple approach to synthesize
protein-polymer bioconjugates with polydispersity index (PDI) of 1 has been demonstrated by site-
specific attachment of hydrophilic monodisperse dendron to hydrophilic protein .31-36 This approch is
relatively easy as both dendrimer and protein are soluble in an aqueous medium and therefore
bioconjugation reaction is straight-forward. However, these bioconjugates lack self-assembling
properties as expected because of the lack of amphiphilic character.
Synthesis of facially amphiphilic biohybrid macromolecules containing both “hydrophilic protein” and
“hydrophobic dendron” domains would be an excellent scaffold. This synthetic methodology would
provide access to perfectly monodisperse protein-based semi-synthetic macromolecules similar to
artificial amphiphilic protein-polymers made through the biosynthetic pathway.24,25 In addition,
chemical diversity of the second domain is not limited in this approach as opposed to polymers made
through the biosynthetic pathway.24,25. However, the development of new chemical methodology to
accomplish this goal is a very challenging task for the following reasons; (i) developing site-specific
bioconjugation chemistry that can target a wide variety of protein is not easy, because proteins are
hetero-polymers made up of twenty amino acids having different functional groups, (iii) considering the
hydrophobic nature of dendron, performing bioconjugation reaction in complete aqueous medium is
highly challenging and, (iii) most importantly, purification and analytical characterization of protein-
dendron bioconjugate is very difficult as the resultant products are macromolecular/facially amphiphilic
in nature and resembles membrane proteins in terms of hydropathy index.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Our group has initiated a research program to develop chemical methodologies for the design of facially
amphiphilic semi-synthetic proteins. In our previous work, we have reported a chemical methodology
for the synthesis of novel protein amphiphiles utilizing a micelle-assisted protein labeling technology 37.
This work laid the foundation for the synthesis of monodisperse protein amphiphiles. However, the
reported methodology was only explored for the bioconjugation of small hydrophobic linear molecules
(<1000 Da). In addition, the synthesized protein nanoassemblies were relatively small in size (8-14 nm)
having low molecular mass (220 kDa – 660 kDa).37 In order to push the boundaries of our chemical
method. Herein, we disclose a novel method for chemical synthesis of a suite of monodisperse protein-
dendron conjugates and studied their self-assembling behavior in detail.
Result and Discussion
Synthesis of Monodisperse Hydrophilic Linker having a Protein Reactive Group
From our previous experience, we knew that the direct site-specific attachment of benzyl-ether dendron
to a hydrophilic protein is difficult to achieve because of the steric hindrance and orthogonal solubility
properties of hydrophilic protein and hydrophobic dendron (Figure 1). To overcome this limitation, we
hypothesized that bioconjugation of an amphiphilic macromolecular probe containing a hydrophilic
monodisperse oligoethylene glycol equipped with the protein reactive functional group covalently
attached to hydrophobic dendron block would be a viable approach. To do that, we chose
undeca(ethylene glycol) as the linker molecule for the following reasons; (i) it is sufficiently long,
flexible and therefore circumvents steric hindrance problem that may arise during bioconjugation
reaction of higher-generation dendrons, (ii) it is hydrophilic and has protein-repellent property, thus
would not denature the protein. To synthesize monodisperse alkyne-terminated diphosphonate derivative
of undeca(ethylene glycol), a strategy by French et al. was adopted with a slight modification (Scheme
1).38 The multi-step synthesis of the target molecule (11) were carried out in the following way. Tosyl
derivative of tetraethylene glycol (TEG) (1) on reaction with monopropargyl TEG (2) using sodium
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
hydride (NaH) in tetrahydrofuran (THF) gave trityl protected octaethylene glycol (3). The compound 3
on trityl deprotection using p-toluenesulfonic acid (TsOH) in methanol (MeOH) produced alcohol 4;
which further on treatment with compound 1 in the presence of NaH in THF yielded trityl protected
dodecaethylene glycol (5) using standard condition. Subsequently, trityl deprotection of compound 5
followed by treatment with compound 1 gave trityl protected undeca(ethylene glycol) (7). The obtained
compound on trityl deprotection furnished corresponding alcohol (8); which on tosylation gave
compound 9. Further, the obtained tosyl derivative (9) on reflux in acetone in the presence of potassium
iodide afforded corresponding iodide derivative (10). Finally, diphosphonate ester (11) was synthesized
by refluxing iodide compound (10) with triethyl phosphite (Scheme 1). The diphosphonate ester of
monodisperse undeca(ethylene glycol) containing sixteen repeating units of ethylene glycol with
terminal alkyne group (11) was synthesized in a total of 12 steps with an overall yield of 2.4% (Scheme
1).
Synthesis of macromolecular Amphiphilic Activity-based Probes (AABPs)
We chose activity-based protein labeling method for bioconjugation because this method provides
remarkable advantages over other bioconjugation methods such as fast reaction kinetics, mild reaction
conditions, chemo- and site-selective modification of the active-site of protein.39 All these features assist
in preserving the native structure of protein during bioconjugation reaction and provides an opportunity
to design monodisperse protein conjugate with little structural perturbation. Synthetic route to make
macromolecular amphiphilic activity-based probe (AABP) is shown in Scheme S9 and S10. We choose
Frechet type benzyl-ether dendron as the hydrophobic domain for the following reasons; (i)
monodisperse benzyl-ether dendrons having high molecular weight can be easily synthesized, (ii) most
importantly, dendrimer of higher-generation adopts the 3-dimensional architecture, and therefore would
have pronounced effect on self-assembling behavior. The benzyl-ether dendrons (G1-G4) with an azide
at the focal point were synthesized through the convergent dendrimer synthesis (Scheme S9, S10).40,41
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Further, hydrophobic G1- and G2-dendritic azide were coupled to alkyne-terminated undeca(ethylene
glycol) derivative using click chemistry strategy in the presence of THF/H2O (1:1) mixture (Scheme
2a).42 However, when we attempted the same reaction condition for coupling of G3- and G4-azides,
those reactions did not work. We suspect the extreme hydrophobic character G3- and G4-azide may be
the reason for the failure. To mitigate this problem, the polarity of the solvent system was decreased by
adding dichloromethane and the reaction was attempted again in the new solvent system
DCM/THF/H2O (2:2:1). We were delighted to see that the reaction went smoothly as evident from
MALDI-ToF results. However, the reaction was not quantitative and some unreacted compound (11)
can be seen in MALDI-ToF results. To separate the unreacted compound 11, we attempted normal
phase thin layer chromatography (TLC) experiment. However, Rf value difference between the product
and the linker molecule same in various solvent combinations. Discouraged by the above results, we
again attempted TLC experiment in reverse phase conditions. This time appreciable Rf difference was
observed for all the clicked products due to their amphiphilic nature. Hence, the purification of all the
clicked products was achieved using a reverse phase chromatography method. The resultant pure
diphosphonate esters were then deprotected using an oxalyl chloride followed by treatment with a
diethylaminosulfur trifluoride (DAST) reagent to make macromolecular AABPs with a fluoro-
phosphonate (FP) as a protein reactive group (Scheme 2a, 2b). Synthesis of monodisperse
macromolecular AABPs containing undeca(ethylene glycol) coupled of dendron of different generation
(FP-CEG-G1, FP-CEG-G2, FP-CEG-G3, and FP-CEG-G4 – simply represented as FP-
CEG/G1/G2/G3/G4) were achieved in 8, 10, 12, and 14, steps with an overall yield of 1.4%, 1.5%,
1.3%, and 0.8%, respectively. The FP was chosen as a protein reactive functional group because it can
target a wide range of serine proteases39, therefore provides an opportunity to study the effect of protein
on the self-assembly of protein-dendron bioconjugates.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Micelle-Assisted Protein Labeling
The bioconjugation of macromolecular AABPs with chymotrypsin was attempted in 50 mM phosphate
buffer pH 7.4. Considering the amphiphilic nature of macromolecular AABPs, i.e., (FP-CEG-
G1/G2/G3/G4), we first attempted the bioconjugation reaction in neat phosphate buffer pH 7.4.
Unfortunately, none of the macromolecular AABPs was soluble (Scheme 2b). Use of water-miscible
organic solvent is an alternate approach; however, we consciously decided not to use any organic
solvent as it might denature the protein structure and also pose a problem during down-stream
purification. Alternatively, we adopted micelle-assisted bioconjugation method reported by our group.37
Remarkably, all of the macromolecular AABPs are soluble in the presence of 10X critical micelle
concentration (CMC) of triton X-100, including the most hydrophobic probe FP-CEG-G4, presumably
due to the formation of a mixed-micelle. As a test reaction, labeling of chymotrypsin protein (serine
protease) by FP-CEG-G1 probe was attempted in the presence of triton X-100 in 50 mM sodium
phosphate buffer at room temperature (Scheme 2c). The progress of the reaction (formation of protein-
dendron bioconjugate) was monitored using MALDI-ToF at different time points. To our delight, we
observed two peaks after 12 h; the first peak corresponds to unreacted native protein and second peak
for protein-dendron bioconjugate at a higher molecular weight which matches with a theoretical
molecular weight of bioconjugates (Figure 2a). Using the similar procedure, all of the macromolecular
AABPs were tested. The MALDI-ToF results indicate that all of them, including the most hydrophobic
and highly branched FP-CEG-G4 were reactive towards chymotrypsin.
Further, to investigate the effect of protein head on protein-dendron self-assembly and to broaden the
scope of methodology; conjugation with three other serine protease, i.e., trypsin, proteinase K, subtilisin
with FP-CEG-G1 was performed under similar conditions (Scheme 2c). As expected, FP-CEG-G1
labeled all the serine proteases as evident from MALDI-ToF results (Figure. 2c). This is understandable
because all the proteins used in this study have the same catalytic triad (serine, aspartic acid, and
histidine residues) and hence capable of reacting with an FP group in a site-specific manner.39 Unlike
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
previously reported methods, our technology is applicable to wide variety of proteins and hence
amenable for the custom design of protein nanoassemblies with very rich structural diversity.
Multi-step Purification and Characterization of Protein-Dendron Bioconjugates
We anticipated that the purification and analytical characterization of protein-dendron bioconjugates
(Figure 1) would be highly challenging, as they are facially amphiphilic bio-hybrid macromolecules and
resemble membrane proteins in terms of hydropathy index. For the purification, the strategy pioneered
by our group was followed.37 Triton X-100 and unreacted macromolecular AABPs from the reaction
mixture were removed by ion-exchange chromatography (IEX) (Figure S1-S6). IEX purification of all
the bioconjugates was successful except Chy-CEG-G4 presumably due to its extreme hydrophobicity
(Figure S7). The second step involves the removal of native protein from protein-dendron bioconjugate,
which is highly challenging because the molecular weight difference between the native protein and
protein-dendron bioconjugate is very small. In addition, the surface charge of protein-dendron
biconjugate is same as the native protein. Considering these issues, traditional SEC and ion-exchange
chromatography cannot be employed for the purification. In order to solve this problem, we
hypothesized that increasing the ionic strength of solution would promote self-sorting of protein-
dendron bioconjugate but not the native protein. This is a reasonable assumption because protein-
dendron bioconjugate contains two distinct domains i.e. hydrophilic protein and hydrophobic dendron,
whereas native protein does not contain the hydrophobic domain. To do that, SEC studies were carried
out in 200 mM NaCl phosphate buffer pH 7.4. To our delight, we observed two well-resolved peaks
with a substantial difference in elution volume. MALDI-ToF analysis of first peak revealed that the first
fraction is exclusively contains protein-dendron bioconjugate and the second fraction contains only
native protein. As expected, with increased ionic strength, protein-dendron conjugate self-assembled to
form higher-order protein assemblies whereas highly hydrophilic native protein failed to self-assemble.
Using the same strategy, all the bioconjugates in library I and II were purified and characterized.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Self-assembly Studies
The ability of protein-dendron bioconjugates to self-assemble into nanoassemblies was first tested by
doing size exclusion chromatography (SEC) studies. All the bioconjugates in the library I and II eluted
at lower elution volume indicative of the formation of higher-order protein complexes(Figure 3a, 3b). .
In addition to SEC studies, we have also carried out dynamic light scattering (DLS) experiments in
parallel to get information on hydrodynamic diameter (Dh) of protein complexes. The Dh of
bioconjugate containing Chy-CEG-G1 self-assemble into a protein complex of ~10 nm. Increase in
dendron volume in the case of Chy-CEG-G2 resulted in an overall increase in the size of a protein
complex by ~2 nm. Interestingly, Dh of Chy-CEG-G3 protein complex was found to be 18 nm, which
approximately double the size of protein complex Chy-CEG-G1. The observed size difference between
Chy-CEG-G1 and Chy-CEG-G3 protein complexes is mainly attributed to the increase in hydrophobic
dendron volume.43 Further, we set out to examine the effect of protein size and surface charge on Dh of
protein complexes. All the bioconjugates in the library I have the identical linker (CEG) and G1
dendron domain but the protein is systematically varied in the macromolecular design. DLS results
revealed that bioconjugates Try-CEG-G1, Sub-CEG-G1, and ProK-CEG-G1, self-assembled to make
protein complexes of sizes 10, 12, and 13 nm respectively. (Figure 3d, Table 1).
In order to further to investigate the polydispersity, molecular weight and oligomeric state of protein
complexes, we performed size-exclusion chromatography coupled with multi-angle laser light scattering
(SEC-MALS). We were gratified to note that the polydispersity index of protein complexes obtained
from SEC-MALS was in the range of 1.001 to 1.003. These results indicate all the protein complexes
are nearly monodisperse, which is an unexpected finding considering the very flexible nature of linker
and highly branched dendrimer domain. We believe that this behavior may be attributed to the
monodisperse nature of protein-dendron bioconjugates. (Figure 4, Table 1). The above results indicate
that the synthesis of monodisperse protein nanoassemblies is feasible provided all the components of
macromolecular design is perfectly monodisperse as exemplified in this study. The oligomeric state and
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
the molecular weight of protein complexes revealed the generation-dependent behavior (library II).
Bioconjugate Chy-CEG-G1 forms ~500 kDa protein complex with an average oligomeric state of 18,
whereas Chy-CEG-G2 self-assembles into ~520 kDa complex with an oligomeric state of 20 (Figure 4a,
4b, Table 1). Remarkably, the bioconjugate having highest dendron volume in a library II, i.e., Chy-
CEG-G3 self-assemble into megadalton-scale protein complex having a molecular weight of ~ 1.3 MDa
with an average oligomeric state of 44 (Figure 4c, Table 1). The obtained SEC-MALS results are in
agreement with DLS data and manifest the effect of dendron size on the oligomeric state and size of
protein complexes in the library I . Next, SEC-MALS results of library II revealed that the Try-CEG-G1
also formed 480 kDa protein complex as expected with an average oligomeric state of 18, which is
similar to the results obtained with Chy-CEG-G1 (Figure 4c, Table 1). The SEC-MALS studies of Sub-
CEG-G1 could not be performed due to the instability problem right after ion-exchange purification. We
do not know the reason for this behavior, a detailed investigation is underway and the results will be
reported in due course. ProK-CEG-G1 also self-assembled into 840 kDa (13 nm) protein complex with
an average oligomeric state of 26 (Figure 4c, Table 1).These results indicates the oligomeric state of
protein complex strongly depends on the surface charge of native proteins (try = 30.5, chy = 19.5, sub =
-6.0, pro = -8.5 mV) .
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Conclusion
In summary, we have successfully synthesized suite of monodisperse facially amphiphilic protein-
dendron conjugates using a micelle-assisted protein labeling method. For th first time, we have
demonstrated that the monodisperse three-dimensional hydrophobic dendrons could be site-specifically
conjugated to a globular protein. The designed protein complexes exhibit both dendrimer generation-
dependent as well as protein-dependent self-assembly properties. Most importantly, this method
provides exquisite control over size, oligomeric state and the molecular weight of the protein
nanoassemblies by either choosing an appropriate dendron or protein of interest and thus open up the
possibility to design nearly monodisperse, megadalton-scale protein nanoassemblies through a chemical
method. The fact that this method could be applicable to a wide variety of proteins and dendrimers
opens up enormous opportunities for this method in the area of vaccine design, targeted drug delivery,
and diagnostic imaging.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Figures
Figure 1. Chemical structures of protein-dendron bioconjugates composed of globular water-soluble
protein (green = trypsin, blue = chymotrypsin, cyan = subtilisin, magenta = proteinase K), hydrophilic
PEG (blue), and hydrophobic dendron block (red). Structures of (A) Try-CEG-G1 (B) Chy-CEG-G1
(C) Sub-CEG-G1 (D) ProK-CEG-G1 (E) Chy-CEG-G1 (F) Chy-CEG-G2 (G) Chy-CEG-G3 (H)
Chy-CEG-G4.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Scheme 1 | Synthesis of monodisperse diphosphonate esters of undeca(ethylene glycol) containing
terminal alkyne functional group.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Scheme 2 | (A) Scheme for synthesis of macromolecular AABPs. (B) Chemical structures of PEG-
dendron macromolecular AABP composed of monodisperse hydrophilic PEG, and hydrophobic
dendrons block. Chemical structures of (i) FP-CEG-G1, (ii) FP-CEG-G2, (iii) FP-CEG-G3 and, (iv) FP-
CEG-G4. (C). Scheme for synthesis of protein-dendron bioconjugates.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
1.2
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Chy-CEG-G1
Chy-CEG-G2
Chy-CEG-G3
Chy-CEG-G4
Chy-CEG-G1
Chy-CEG-G2
Chy-CEG-G3
27,018
27,858
29,542
29539
27916
27065
0.8
0.4
0.0
32,945
20000 25000 30000 35000 40000
Mass (m/z)
20000 25000 30000 35000 40000
Mass (m/z)
1.2
1.2
Try-CEG-G1
Try-CEG-G1
23325
27018
30487
28876
27,065 28,859
30,478
24,922
Chy-CEG-G1
Sub-CEG-G1
ProK-CEG-G1
Chy-CEG-G1
Sub-CEG-G1
ProK-CEG-G1
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
27310
24907
25402
20000
25000
30000
Mass (m/z)
35000
40000
20000
25000
30000
35000
Mass (m/z)
Figure 2 | MALDI-ToF characterization of protein-dendron bioconjugates. (A) Reaction mixture
dendron variants (Library I) (B) after purification. (C) Reaction mixture (Library II) protein variants,
(D) after purification.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
1.2
1.0
0.8
0.6
0.4
0.2
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Chy-CEG-G1
Chy-CEG-G2
Chy-CEG-G3
Try-CEG-G1
Chy-CEG-G1
Sub-CEG-G1
ProK-CEG-G1
0.0
4
6
8
10
12
14
16
4
6
8
10
12
14
16
Elution Volume (mL)
Elution Volume (mL)
1.2
1.0
0.8
0.6
0.4
0.2
1.2
1.0
0.8
0.6
0.4
0.2
Chy-CEG-G1
Chy-CEG-G2
Chy-CEG-G3
Try-CEG-G1
Chy-CEG-G1
Sub-CEG-G1
ProK-CEG-G1
0.0
1
0.0
1
10
Dh / nm
100
10
Dh / nm
100
Figure 3 | Self-assembly data of protein-dendron conjugates. SEC data for (A) Dendron variants, and
(B) Protein varients. Experiments were carried out on a Superdex-200 10/300 GL column and 500 µL
of the sample (5 mg/mL) were injected. DLS data for (C) Dendron variants, and (D) Protein variants.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
0.25
0.24
0.23
0.22
0.21
0.20
0.25
0.219
0.219
0.216
0.213
0.210
0.207
0.204
Chy-CEG-G1
Chy-CEG-G2
900.0k
0.24
0.216
600.0k
0.213
0.23
600.0k
400.0k
0.210
0.22
0.207
300.0k
0.21
200.0k
0.204
0.201
0.201
40
0.20
60
45
50
55
60
40
45
50
55
Time (min)
Time (min)
2.0M
Chy-CEG-G3
0.30
0.25
1.6M
1.2M
800.0k
400.0k
0.20
35
40
45
50
55
Time (min)
0.28
0.26
0.24
0.22
0.20
Try-CEG-G1
ProK-CEG-G1
600.0k
0.30
0.25
0.20
900.0k
600.0k
300.0k
400.0k
200.0k
40
45
50
Time (min)
55
60
40
45
50
55
60
Time (min)
Figure 4 | SEC-MALS data of protein-dendron bioconjugates (A) Chy-CEG-G1 (B) Chy-CEG-G2 (C)
Chy-CEG-G3, (D) Try-CEG-G1 (E) ProK-CEG-G1, respectively.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Protein-
dendron
bioconjugate
Elution Vol
SEC (mL)
Mol Wt SEC
–MALS (kDa)
Oligomeric
State
SEC-MALS
(mer)
Polydispersity Hydrodynamic
Index (PDI )
Diameter (Dh)
DLS (nm)
Chy-CEG-G1
Chy-CEG-G2
Chy-CEG-G3
11.6
11.2
10.6
521
521
20
20
44
1.025
1.003
1.003
10.1
11.7
18.1
1320
Try-CEG-G1
Sub-CEG-G1
Prok-CEG-G1
10.9
10.5
10.0
458
-----
841
18
-----
26
1.001
-----
10.0
11.7
13.4
1.002
Table 1 | Summary of self-assembly studies TBPs
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
Associated content
Supporting Information
Corresponding Author # e-mail: sandanaraj.britto@iiserpune.ac.in
Author Contributions * These authors contributed equally to this work
Acknowledgements
This work was supported by IISER-Pune Start-up Grant and DST-SEB Early Career Award to BSS
(ECR/2015/000253) DBT grant (BT/PRI1450/BRB/I0/1370/2015). PJB and MMR thanks CSIR and
UGC for the Scholarship. We thank Arjun Gontala for synthesizing few intermediate compounds and
Rasika Avatade for assisting in conjugation and purification of TBPs. We also thank all our lab
members for valuable discussions.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
References
1. Mateu, M. G. Arch. Biochem. Biophys. 2013, 531, 65−79.
2. Stock, D.; Leslie, A. G. W.; Walker, J. E. Science 1999, 286, 1700-1705
3. Massover, W. H. Micron 1993, 24, 389–437.
4. Kerfeld, C. A.; Heinhorst, S.; Cannon, G. C. Annu. Rev. Microbiol. 2010, 64, 391−408
5. Seeman, N. C. Nature 2003, 421, 427−431.
6. Guo, P. X. Nat. Nanotechnol 2010, 5, 833−842.
7. Ulijn, R. V.; Smith, A. M. Chem. Soc. Rev. 2008, 37, 664–675.
8. Luo, Q.; Hou, C.; Bai, Y.; Wang, R.; Liu, J. Chem. Rev. 2016, 116, 13571-13632.
9. Lai, Y.-T.; Tsai, K.-L.; Sawaya, M. R.; Asturias, F. J.; Yeates, T. O. J. Am. Chem. Soc. 2013,
135, 7738−7743
10. King, N. P.; Bale, J. B.; Sheffler, W.; McNamara, D. E.; Gonen, S.; Gonen, T.; Yeates, T. O.;
Baker, D. Nature 2014, 510(7503), 103−108.
11. Worsdorfer, B.; Woycechowsky, K. J.; Hilvert, D. Science 2011, 331, 589– 592.
12. Yeates, T. O. Annu. Rev. Biophys. 2017, 46, 23 −42
13. Huang, P. S.; Boyken, S. E.; Baker, D. Nature 2016, 537, 320-327.
14. Abuchowski, A.; van Es, T.; Palczuk, N. C.; Davis, F. F. J. Biol. Chem. 1977, 252, 3578−3581.
15. Abuchowski, A.; McCoy, J. R.; Palczuk, N. C.; van Es, T.; Davis, F. F. J. Biol. Chem. 1977,
252, 3582−3586.
16. Hannink, J. M.; Cornelissen, J. J. L. M.; Farrera, J. A.; Foubert, P.; De Schryver, F. C.;
Sommerdijk, N. A. J. M.; Nolte, R. J. M. Angew. Chem., Int. Ed. 2001, 40, 4732-4734.
17. Velonia, K.; Rowan, A. E.; Nolte, R. J. J. Am. Chem. Soc. 2002, 124, 4224-4225.
18. Boerakker, M. J.; Hannink, J. M.; Bomans, P. H. H.; Frederik, P. M.; Nolte, R. J. M.; Meijer, E.
M.; Sommerdijk, N. A. J. M. Angew. Chem., Int. Ed. 2002, 41, 4239-4241.
19. Reynhout, I. C.; Cornelissen, J. J. L. M.; Nolte, R. J. M. J. Am. Chem. Soc. 2007, 129, 2327
20. Dirks, A. J. T.; Berkel, S. S. V.; Hatzakis, N. S.; Opsteen, J. A.; Delft, F. L.V.; Cornelissen, J. J.
L. M.; Rowan, A. E.; Hest, J. C. M. V.; Rutjes, F. P. J. T.; Nolte, R. J. M. Chem. Commun. 2005,
41, 4172–4174.
21. Chen, G.; Hoffman, A. S. Bioconjugate Chem. 1993, 4, 509-514.
22. Kulkarni, S.; Schilli, C.; Müller, A. H.; Hoffman, A. S.; Stayton, P. S. Bioconjugate Chem.
2004, 15, 747-753.
23. Wan, X. J.; Liu, S. Y. Macromol. Rapid Commun. 2010, 31, 2070-2076.
24. Park, W. M.; Champion, J. A. J. Am. Chem. Soc. 2014, 136, 17906-17909.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201900341
ChemBioChem
25. Hassouneh, W.; Fischer, K.; MacEwan, S.R.; Branscheid, R.; Fu, C.L.; Liu, R.; Schmidt, M.;
Chilkoti, A. Biomacromolecules 2012, 13, 15981605..
26. Thomas, C. S.; Glassman, M. J.; Olsen, B. D. ACS Nano 2011, 5, 5697−5707
27. Lam, C. N.; Olsen, B. D. Soft Matter 2013, 9, 2393−2402.
28. Huang, A.; Qin, G.; Olsen, B. D. ACS Appl. Mater. Interfaces 2015, 7(27), 14660−14669.
29. Averick, S.; Karácsony, O.; Mohin, J.; Yong, X.; Moellers, N. M.; Woodman, B. F.; Zhu, W.,
Mehl, R. A.; Balazs, A. C.; Kowalewski, T. Angew. Chem., Int. Ed. 2014, 53, 8050−8055.
30. Li, X., and Liu, C. C. ChemBioChem. 2014, 15, 2335−2341
31. Kluger, R.; Zhang, J.; J. Am. Chem. Soc. 2003, 125, 6070-6071 .
32. Van Baal, I.; Malda, H.; Synowsky, S. A.; van Dongen, J. L. J.; Hackeng, T. M.; Merkx, M.;
Meijer, E. W. Angew. Chem., Int. Ed. 2005, 44, 5052-5057.
33. Rendle, P. M.; Seger, A.; Rodrigues, J.; Oldham, N. J.; Bott, R. R.; Jones, J. B.; Cowan, M. M.;
Davis, B. G. J. Am. Chem. Soc. 2004, 126, 4750.
34. Sato, M.; Furuike, T.; Sadamoto, R.; Fujitani, N.; Nakahara, T.; Niikura, K.; Monde, K.; Kondo,
H.; Nishimura, S. I. J. Am. Chem. Soc. 2004, 126, 14013–14022.
35. Kostiainen, M. A.; Szilvay, G. R.; Smith, D. K.; Linder, M. B.; Ikkala, O. Angew. Chem., Int.
Ed. 2006, 45, 3538–3542.
36. Kostiainen, M. A.; Szilvay, G. R.; Lehtinen, J.; Smith, D. K.; Linder, M. B.; Urtti, A.; Ikkala, O.
ACS Nano 2007, 1, 103–113I
37. Sandanaraj, B. S.; Reddy, M. M.; Bhandari, P. J.; Kumar, S. and Aswal, V. K. Chem. Eur. J.
2018, 24(60), 16085-16096.
38. French, A. C.; Thompson, A. L.; Davis, B. G. Angew. Chem., Int. Ed. 2009, 121, 1274-1278.
39. Liu, Y., Patricelli, M. P. & Cravatt, B. F. Proc. Natl Acad. Sci. USA, 1999, 96, 14694-14699.
40. Hawker, C. J.; Frechet, J. M. J. Am. Chem. Soc. 1990, 112, 76387647.
41. Wooley, K. L.; Hawker, C. J.; Fréchet, J. M. J. J. Am. Chem. Soc. 1991, 113, 4252-4261.
42. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed. 2002,
114, 2708-2711.
43. Mourey, T. H.; Turner, S. R.; Rubinstein, M.; Frechet, J. M. J.; Hawker, C. J.; Wooley, K. L.
Macromolecules 1992, 25, 2401-2406 Macromolecules 1992, 25.90, 2401-2406.
This article is protected by copyright. All rights reserved.