Rational Design of Supramolecular Dynamic Protein Assemblies by Using a Micelle-Assisted Activity-Based Protein-Labeling Technology

Article abstract of DOI:10.1002/chem.201802824

The self-assembly of proteins into higher-order superstructures is ubiquitous in biological systems. Genetic methods comprising both computational and rational design strategies are emerging as powerful methods for the design of synthetic protein complexes with high accuracy and fidelity. Although useful, most of the reported protein complexes lack a dynamic behavior, which may limit their potential applications. On the contrary, protein engineering by using chemical strategies offers excellent possibilities for the design of protein complexes with stimuli-responsive functions and adaptive behavior. However, designs based on chemical strategies are not accurate and therefore, yield polydisperse samples that are difficult to characterize. Here, we describe simple design principles for the construction of protein complexes through a supramolecular chemical strategy. A micelle-assisted activity-based protein-labeling technology has been developed to synthesize libraries of facially amphiphilic synthetic proteins, which self-assemble to form protein complexes through hydrophobic interaction. The proposed methodology is amenable for the synthesis of protein complex libraries with molecular weights and dimensions comparable to naturally occurring protein cages. The designed protein complexes display a rich structural diversity, oligomeric states, sizes, and surface charges that can be engineered through the macromolecular design. The broad utility of this method is demonstrated by the design of most sophisticated stimuli-responsive systems that can be programmed to assemble/disassemble in a reversible/irreversible fashion by using the pH or light as trigger.

Full text of DOI:10.1002/chem.201802824

A Journal of
Accepted Article
Title: Rational Design of Supramolecular Dynamic Protein Assemblies
Using a Micelle-Assisted Activity-based Protein Labeling
Technology
Authors: Britto Sandanaraj, Mohan Reddy, Pavankumar Bhandari,
Sugam Kumar, and Vinod Aswal
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.201802824
Link to VoR: http://dx.doi.org/10.1002/chem.201802824
Supported by

10.1002/chem.201802824
Chemistry - A European Journal
Rational Design of Supramolecular Dynamic Protein Assemblies Using a
Micelle-Assisted Activity-based Protein Labeling Technology
Britto S Sandanaraj1#, Mullapudi Mohan Reddy1* and Pavankumar Janardhan Bhandari1*, Sugam
Kumar2 and Vinod K. Aswal2
1Department of Chemistry & Biology, Indian Institute of Science Education and Research
(IISER), Pune, 411 008, India
2Solid State Physics Division, Bhabha Atomic Research Centre (BARC), Mumbai, 400085, India
* These authors contributed equally to this work
# e-mail: sandanaraj.britto@iiserpune.ac.in
Abstract
Self-assembly of proteins into higher order super-structures is ubiquitous in biological
systems. Genetic methods comprising both computational and rational design strategies are
emerging as powerful methods for design of synthetic protein complexes with high accuracy and
fidelity. Although useful, most of the reported protein complexes lack dynamic behavior that
may limit their potential applications. On the contrary, protein engineering using chemical
strategies offers excellent possibilities for the design of protein complexes with stimuli-
responsive functions and adaptive behavior. However, designs based on chemical strategies are
not accurate and therefore yield polydisperse samples that are difficult to characterize. Here, we
describe simple design principles for the construction of protein complexes through
supramolecular chemical strategy. Micelle-assisted activity-based protein labeling technology
has been developed to synthesize library of facially amphiphilic synthetic proteins, which self-
assemble to make protein complexes through hydrophobic interaction. The proposed
methodology is amenable for the synthesis of protein complex libraries with molecular weights
and dimensions comparable to naturally occurring protein cages. The designed protein
complexes display very rich structural diversity, oligomeric states, sizes and surface charges that
can be engineered through macromolecular design. The broad utility of this method is
demonstrated by design of most sophisticated stimuli-responsive systems that can be
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
programmed to assemble/disassemble in a reversible/irreversible fashion using pH or light as a
trigger.
Introduction
Design of protein super-structure through bottom-up approach has gained enormous interest
in the recent years1-5. Genetic engineering6-8 and supramolecular chemical strategies9-12 are two
complementary technologies that are widely used for the design strategy. The structures of
protein complexes made through genetic methods have been determined to a near-atomic
resolution using X-ray crystallography13 and cryo-electron microscopy (cryo-EM)14 . In addition,
small-angle X-ray scattering (SAXS) technique has also been used to determine low-resolution
structures15,16. Despite these advantages, genetic methods are prone to high failure rate and often
yield structures not intended by the design16-18. Most importantly, a majority of the designs
produced only static structures; integration of dynamic behavior has yet to be explored. In
addition, the standard genetic method is restricted to a small set of building blocks - 20 canonical
amino acids - that are used for the design purpose.
Natural protein complexes evolved to perform advanced functions in vivo. They achieve this
remarkable feat through several reversible/irreversible post-translational modifications, thus
converting the static structure into dynamic one and vice versa19. Hence, access to more building
blocks (chemical entities) is essential to realize dynamic/adaptive behavior. While the
incorporation of several unnatural amino acids onto protein through modified genetic method
was demonstrated long back20, to date there are only few reports on the use of unnatural amino
acids for design of de novo protein complex21. Alternatively, protein engineering using chemical
strategies provide opportunities for the design of stimuli-responsive protein complexes because
of access to innumerable building blocks22,23. Despite this advantage, most of the designs yield
polydisperse samples and therefore lack detailed biophysical characterization. Hence, the
underlying design rules are still elusive. Here, we report design and development of micelle-
assisted activity-based protein labeling technology and its utility in the synthesis of libraries of
protein complexes with defined size, surface charge and oligomeric state. In addition, the same
method is utilized for design of protein complex with stimuli-responsive function/adaptive
behavior.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Results
Macromolecular design
The inspiration for our design came from structural, biophysical and self-assembly studies of
naturally occurring facially amphiphilic hydrophobin proteins24 (Fig. 1a). In particular, a recent
report of the high-resolution structure of a bacterial hydrophobin reveals that it exists as a
decamer25 where the hydrophobic face of the protein is shielded away from the water molecules
by forming a dense hydrophobic core and hydrophilic face is presented outside, which mimics
protein cages26. This study gave us a clue for the “minimal design” of facially amphiphilic semi-
synthetic protein. The macromolecular design has three core structural elements: (i) hydrophilic
globular protein, (ii) flexible hydrophilic linker, and (iii) hydrophobic tail (Fig. 1a). We
envisioned that the designed protein would self-assemble into complex architecture, primarily
driven by strong hydrophobic interaction, which is one of the major driving force for the
formation of supramolecular protein assembly in natural27 and artificial systems28. The basic
tenet of our hypothesis is that the attributes of a protein complex (size, oligomeric state, surface
charge and shape) would be dictated by choice of a core structural element of macromolecular
design29.
The modular strategy for synthesis of libraries of facially amphiphilic semi-synthetic
protein is shown in Fig. 1f (Supplementary Results, Supplementary Scheme. 22). Activity-based
protein labelling method30 was chosen for following reasons: (i) site-specific modification of an
active-site residue of serine proteases; (ii) fast reaction kinetics and (iii) serine proteases
encompasses a large set of proteins and therefore provide opportunity to make large library
(library I, Fig. 1c, Supplementary Fig. 1a). In addition, monodisperse oligoethylene glycol
(defined repeating units) is used as spacer moiety because of its water solubility and protein
repellent properties and most importantly, spacer length could be systematically engineered to
generate a unique library of protein conjugates (library II, Fig. 1d, Supplementary Fig. 1b).
Finally, tail hydrophobicity can be systematically varied by modifying tail length/branching
utilizing click chemistry strategy31 (library III, Fig. 1e, Supplementary Fig. 1c)
As expected, the solubility of the probe (FP-OEG-C12-3T, Supplementary Fig. 2a) posed a
challenge in terms of choosing an optimum condition to carry out protein labelling reaction in
100% aqueous medium. Several reaction conditions were attempted but none of them was
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
successful. Considering the amphiphilic nature of the probe, we hypothesized that it could be
solubilized in the presence of a small-molecule micelle. A reasonable assumption is that the
amphiphilic probe would make a mixed micelle in the presence of an another suitable surfactant
(Fig. 1b), where the tail component of a probe would be part of the hydrophobic micellar interior
and hydrophilic end (oligoethylene glycol with a terminal fluorophosphonate group) would be
available for active-site labelling of a serine protease32 (Fig. 1b). To test this hypothesis,
extensive surfactant screening was carried out. Triton X-100 was found to be best, as it satisfies
all the requirements for protein labelling reaction. As a test case, micelle-assisted labelling of
trypsin (serine protease) by FP-OEG-C12-3T was first attempted (Fig. 1b and 1f). The reaction
was carried out in the presence of 10X critical micelle concentration (CMC) of triton X-100 at
room temperature pH 7.4. The extent of protein modification was monitored using MALDI-ToF
at different time points. After 12 h, a major peak for trypsin conjugate (Try-OEG-C12-3T) at
24,453 Da and a minor peak at 23,272 Da for native trypsin were observed (Fig. 2a). Motivated
by these results, we set out to test whether this method could be widely applicable to other serine
proteases as well (chymotrypsin, proteinase K, and subtilisin) to further validate the method (Fig.
1c). FP-OEG-C12-3T labeled all the serine proteases as evident from MALDI-ToF results (Fig.
2a). Encouraged by these results, we focused our efforts on the design and synthesis of 4
additional AABPs (AABP library I, Supplementary Fig. 2a, Supplementary Scheme. 22) having
different linker length (di-, tetra-, dodeca- and cetyl-ethylene glycol). All the probes labelled
trypsin except FP-DEG-C12-3T because of sub-optimal linker length (Fig. 2b, Supplementary
Fig. 2a). Finally, to demonstrate the systematic variation of hydrophobic tail, 7 new AABPs were
synthesized (AABP library II, Supplementary Fig. 2b, Fig. 1e, Supplementary Scheme. 22) with
varying alkyl chain length (hexyl, dodecyl, and octadecyl abbreviated as C6/C12/C18) and
branching (1-tail/2-tails/3-tails abbreviated as 1T/2T/3T). All the probes including a most
hydrophobic probe FP-OEG-C18-3T (Fig. 2c, Supplementary Fig. 2b) were reactive as evident
from MALDI-ToF data.
The next major challenge in this study is the scale-up and purification of a protein
conjugate as it resembles natural hydrophobin/membrane proteins (Fig. 1a) in terms of
hydropathy index24. Synthesis of a Try-OEG-C12-3T conjugate was performed in large scale
(200 mg scale). Triton X-100 and unreacted probe in the reaction mixture were successfully
removed using ion-exchange chromatography (IEX) method (Fig. 2d). Since Try-OEG-C12-3T
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
is facially amphiphilic, we hypothesized that increasing ionic strength of the solution would
promote self-sorting of Try-OEG-C12-3T through strong hydrophobic interaction but not the
native trypsin, which would aid in purification by the size-exclusion chromatography (SEC)
method33. As expected, elution of protein conjugate Try-OEG-C12-3T (1 M NaCl phosphate
buffer pH 7.4) in void volume suggests formation of higher-order complex, while monomeric
native trypsin eluted at 65 mL in SEC experiment (Fig. 2e). Finally, NaCl was removed by
standard desalting method (Fig. 2f). After three-step purification, a single peak for the purified
conjugate at 24,453 Da with a very high purity was obtained with an overall yield of 20% (Fig.
1g). Most importantly, the same method was used for synthesis and purification of all the protein
conjugates in the library I to III except for Try-OEG-C18-3T (Fig. 1b, 1f, 2d-2f, 2g-2i,
Supplementary Scheme. 1-16).
Programmed self-assembly
In order to test the self-assembling ability of designed protein conjugate (Fig. 1g-i),
detailed self-assembly studies were carried out using a variety of techniques which include
analytical SEC, dynamic light scattering (DLS), small-angle neutron scattering (SANS), size-
exclusion chromatography coupled with multi-angle laser light scattering (SEC-MALS) and zeta
potential measurements. First, the effect of protein structure (Fig. 1c, Supplementary Fig. 7),
size (Supplementary Fig. 4, Supplementary Table. 2a) and surface charge (Fig. 3m,
Supplementary Fig. 6) on self-assembly was studied. For this purpose, we focused our attention
on protein library I (Fig. 1c and 1g, Supplementary Fig. 1a). All the protein conjugates in this
library have an identical linker (OEG) and a hydrophobic tail (C12-3T) except the protein is
systematically varied in the macromolecular design. Analytical SEC studies revealed that all four
conjugates eluted at lower elution volumes (11.5 to 10.9 mL) indicative of large protein
complexes (Fig. 3a, Supplementary Fig. 3 and Supplementary Table. 1). In order to further
understand various attributes of protein complexes such as the radius of gyration (Rg), oligomeric
state and shape (σ), detailed SANS studies were carried out34. Both trypsin and chymotrypsin
complexes exist as a 10-mer with molecular mass of 245 kDa whereas proteinase K conjugate
self-assemble into a 22-mer with molecular mass of 661 kDa (Fig. 3d, Supplementary Fig. 4 and
Supplementary Table. 2b). However, SANS experiment of subtilisin complex failed as sample
precipitated during the course of SANS measurement (12 h). Nevertheless, it was evident from
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
the SEC and DLS results that subtilisin conjugate is also capable of forming higher-order protein
complex. The Rg of protein complexes were obtained from Guinier region of SANS data. The Rg
of trypsin, chymotrypsin, and proteinase K complexes are 41 Ǻ, 40 Ǻ and 53 Ǻ, respectively
(Table. 1, Supplementary Table. 2b). The hydrodynamic radii (Rh) were determined by DLS
studies (Fig. 3g, Supplementary Fig. 5 and Supplementary Table. 3). Rh data showed an
interesting trend similar to Rg data (Table. 1, Supplementary Table. 3). The shape factor σ
(Rg/Rh) were obtained from SANS and DLS data. The obtained σ values of all of the protein
complexes are in the range of 0.60-0.79 (Table. 1, Supplementary Table. 3), indicating that the
designed protein complexes exist as a spherical object similar to both naturally occurring26,35 and
synthetic protein cages7,14. To confirm SANS results, the molecular mass of Try-OEG-C12-3T
was verified using SEC-MALS and value of 276 kDa (Fig. 3j) obtained was in agreement with
SANS results (245 kDa, Table. 1). Zeta potential measurements were carried out to investigate
the surface charge of designed protein complexes (Supplementary Data Fig. 6 and Table. 1).
Both trypsin and chymotrypsin complexes are highly cationic (+16.5 mV and +8.7 mV),
proteinase K complex is slightly anionic (-3.2 mV) and subtilisin complex is close to neutral (-
0.7 mV).
In order to understand the effect of linker length on protein self-assembly, we focused our
attention on protein library II (Fig. 1d and 1h, Supplementary Fig. 1b). All the protein complexes
are monodisperse as evident from SEC and SANS results (Fig. 3b and 3e, Supplementary Fig. 3
and 4). As predicted, Rg of protein complex increased with increase in spacer length. The size of
Try-TEG-C12-3T is 32 Ǻ whereas Try-CEG-C12-3T is 49 Ǻ (Table. 1). SANS studies also
revealed that molecular mass of Try-TEG-C12-3T with the shortest linker (TEG) is 243 kDa
whereas Try-CEG-C12-3T with the longest linker (CEG) is 372 kDa (Fig. 3e, Supplementary
Fig. 4 and Supplementary Table. 2b). DLS results also showed a similar trend (Fig. 3h,
Supplementary Fig. 5 and Table. 1). Molecular mass of Try-CEG-C12-3T was independently
verified by SEC-MALS. The obtained value (400 kDa) matches with SANS results (Fig. 3k,
Table. 1). As expected, all the protein complexes are cationic and zeta potential values are
between +4.6 to +16.5 mV (Supplementary Fig. 6 and Table. 1).
Finally, the influence of tail hydrophobicity on protein conjugate self-assembly was
systematically studied (protein library III, Fig. 1e and 1i, Supplementary Fig. 1c). As predicted,
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
negative design Try-OEG-C6-3T with short hydrophobic tail predominantly exists in monomeric
form at 200 µM concentration (Supplementary Fig. 3). However, Try-OEG-C6-3T self-
assembled to trimer at higher concentration (800 µM) as evident from SANS and DLS results
(Supplementary Fig. 4 and 5, Table.1 and Supplementary Table. 2b and 3). Similarly, Try-OEG-
C12-1T predominantly exists as monomer during SEC experiment, whereas same conjugate
behaved as higher-order complex when analyzed by SANS and DLS experiments. This is
primarily attributed to result of dilution effect during SEC experiment (Supplementary Fig. 4
and 5, Table.1 and Supplementary Table. 2b and 3). As expected, protein conjugates with
sufficient tail length and hydrophobicity (Try-OEG-C12-2T/3T and Try-OEG-C18-2T) self-
assemble into large protein complexes (Fig. 3c, 3f and 3i, Table.1, Supplementary Fig. 3, 4 and
5, Supplementary Table. 1, 2b and 3). Remarkably, protein conjugates with just 1 hydrophobic
tail, Try-OEG-C18-1T, self-assemble into large protein complex as evident from SEC studies
(Fig. 3c, Supplementary Fig. 3). To further corroborate SANS results, the molecular mass of
Try-OEG-C12-2T was independently verified using SEC-MALS and the result obtained (261
kDa) matches with SANS results (Fig. 3l, Table. 1)
In order to probe the secondary and tertiary structure of designed protein complexes, detailed
biophysical studies were carried out using fluorescence36 and CD spectroscopy37. Fluorescence
studies reveal there was no shift in tryptophan emission maximum which indicates the 3D
structure of a protein is preserved in the complex state (Supplementary Fig. 7a-7c). However,
CD indicates a small perturbation in the secondary structure of some conjugates (Supplementary
Fig. 7d-7f). A detailed investigation is underway and the results will be reported in due course.
Design of stimuli-responsive supramolecular protein complex
In order to test whether change in pH has any effect on self-assembly of designed protein
conjugate7,14,35, SEC studies of Chy-OEG-C12-3T were carried out at different pH. The elution
volume of protein complex was shifted to lower side by 0.4 mL at pH 10. This result indicates
the size of protein complex enlarged with increase in pH. Surprisingly, decreasing solution pH
to 5 resulted in the shrinkage of the size with respect to dimension of protein complex at pH 7.4
(Supplementary Fig. 8).
In order to demonstrate the versatility of our method, photo-responsive semi-synthetic
protein was designed (Fig. 4a, Supplementary Fig. 1d and 2c). Design principle is same as
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
previous design except a photo-responsive group (2-nitrobenzyl derivative) is installed between
octaethylene glycol (hydrophilic spacer) and a hydrophobic tail (C12-2T) (Fig. 4a,
Supplementary Fig. 2c). Our hypothesis is upon photo-irradiation; cleavage of a 2-nitrobenzyl
group38 would lead to a separation of hydrophobic domain from the rest of protein conjugate
(hydrophilic globular domain) that would result in dis-assembly of a protein complex into
constitutive monomers (Fig. 4a), as the assembly is kept intact by strong hydrophobic
interaction. Next, time-dependent photo-chemical reactions were carried out on the photo-
responsive protein conjugate (50 µM) and the samples were analyzed by analytical SEC. As
expected, at time zero, the protein conjugate eluted at 11.5 mL indicative of higher-order
complex (8-mer). However, at intermediate time intervals (2 mins to 45 mins), two peaks were
observed, first peak at 11.5 mL and the second peak at 17.0 mL corresponding to a protein
complex and a cleaved protein conjugate, respectively (Fig. 4b). The concentration of protein
complex gradually decreased with a concomitant increase in concentration of a cleaved
monomeric protein as function of sample exposure time to UV light (Fig. 4b). In parallel,
detailed MALDI-ToF analyses were carried out for reaction mixtures at different time points
(Supplementary Fig. 9). At time zero, a single peak at 24,362 Da for the full conjugate Try-
OEG-NB-C12-3T was observed (Fig. 4c). At 2 minutes, two peaks were seen, one major peak
for the full conjugate and a very small peak for a cleaved protein Try-OEG-NH2 at 23,723 Da
(Fig. 4c). With increase in UV light exposure, an appreciable decrease in peak height for the full
conjugate and concomitant increase for a cleaved protein were observed. At 90 mins, complete
conversion of full conjugate into a cleaved protein was achieved (Fig. 4c and Supplementary Fig.
9). The peak at 10.5 mL did not vanish completely even after prolonged irradiation time (90
mins). MALDI-ToF results revealed that sample did not contain trace amount of full conjugate
or a cleaved protein. The width of the peak at 11.5 mL was slightly increased and elution volume
moved to lower side (11.3 mL), which suggests that the released amphiphilic small molecule (2-
nitrosobenzaldehyde derivative) forms a micelle-like structure. Attempt to characterize this
1
compound by H-NMR and mass spec was not successful. We have also carried out the control
experiment by exposing Try-OEG-NB-C12-3T to the visible light, as expected there was no
disassembly (Supplementary Fig. X). Similarly, protein complex Try-OEG-C12-3T is stable
upon exposure to UV light. These control experiments prove that the disassembly of Try-OEG-
NB-C12-3T protein complex is because of selective cleavage of nitro-benzyl group.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Discussion
Accurate design of supramolecular protein complex through bottom-up approach is at the
forefront of bio-molecular engineering efforts6-8. Protein engineering utilizing chemical
technologies offer remarkable advantages over genetic method as they are not restricted to
standard 20 canonical amino acids9-12. However, most of the reported chemical methods yield
polydisperse sample, restricted to few proteins and lacks modular approach9. Our results
demonstrate that intelligent rational design based on chemical method can yield supramolecular
protein complex with narrow polydispersity. The protein self-assembly is mediated by strong
favorable interaction between hydrophobic tails opposed by electrostatic repulsion between
protein head group containing point charges (presence of acidic and basic amino acid at
physiological pH and ionic strength)39. This mechanism is further supported by the fact that
protein conjugate with short hydrophobic tail (Try-OEG-C6-2T) failed to self-assemble. This is
because electrostatic repulsion between proteins would be much higher compared to weak
hydrophobic interaction in the assembled state, which would be energetically unfavorable39. In
addition, oligomeric state of ProK-OEG-C12-3T is quiet different compared to Try-OEG-C12-
3T or Chy-OEG-C12-3T. This result can be explained based on the surface charge of proteins.
Proteinase K surface charge is close to neutral and hence the complex can accommodate more
subunits without much electrostatic repulsion. The most striking features of our method are (i)
simple design, (ii) high success rate, (iii) rapid generation of library of supramolecular protein
complexes with unique structural diversity, and (iv) the size, oligomeric state and surface charge
of protein complex can be precisely engineered through macromolecular design.
Accurate design of photo-responsive monodisperse protein complexes is a fascinating
idea. However, to date, there are no reports based on this concept. This kind of design is very
difficult to achieve through standard genetic method as 20 naturally occurring amino acids have
not evolved to respond to light1,2. Amber-codon suppression method could be used, however,
that method is cumbersome and engineering-intensive20. Our results demonstrate that accurate
design of most sophisticated photo-responsive molecular machine can be achieved through
chemical method. The programmed dis-assembly of protein complex is achieved upon exposure
to light. These results also emphasize the proposed mechanism that favorable interaction
between the hydrophobic tails is the major driving force for protein conjugate to self-assemble.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
The cleavage of hydrophobic tail upon photo-irradiation leads to complex destabilization
because of loss of strong hydrophobic interactions.
The simple design principle described here makes it easy to extend this methodology to
large number of enzyme families and proteins, for example amphiphilic activity-based probe
equipped with appropriate reactive groups40 can be used to label large number to enzymes
(cysteine, aspartic and metallo proteases, etc) using similar approach described here. Similarly,
this methodology can be extended for labeling of N-terminal amino acid of any protein41. In
addition, modular fashion of our synthetic strategy provides opportunity to introduce chemically
diverse linkers to make stimuli-responsive protein complexes which can respond to ionic
strength, temperature, enzyme, pH, redox, photo and electro-trigger. Most importantly,
hydrophobic tail of a protein conjugate can be systematically varied using monodisperse linear
polymers and dendrimers. These modifications provide opportunity for custom-design of protein
complexes in the size range 20-100 nm and molecular weights in mega-dalton range. Therefore,
this method has immense potential to push the boundaries of bio-molecular engineering efforts42-
46. We firmly believe this work would inspire genetic methods (both computational and rational
approach) for construction of stimuli-responsive protein complexes using unnatural amino acids
as building blocks through bottom-up approach. The ability to engineer small molecule coupled
with technology to make synthetic protein complex opens up new avenues for accurate design of
“smart protein complex”. This technology would have huge impact in the area of vaccine
design47, targeted drug delivery48, in vivo diagnostics49 and synthetic biology50.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
References
1
2
3
4
5
6
Huang, P. S., Boyken, S. E. & Baker, D. The coming of age of de novo protein design.
Nature 537, 320-327 (2016).
Yeates, T. O. Geometric principles for designing highly symmetric self-assembling protein
nanomaterials. Annu. Rev. Biophys. 46, 23-42 (2017).
Salgado, E. N., Radford, R. J. & Tezcan, F. A. Metal-directed protein self-assembly. Acc.
Chem. Res. 43, 661-672 (2010).
Reynhout, I. C., Cornelissen, J. J. L. M. & Nolte, R. J. M Synthesis of polymer-biohybrids:
from small to giant surfactants. Acc. Chem. Res. 42, 681-692 (2009).
Luo, Q., Hou, C., Bai, Y., Wang, R. & Liu, J. Protein Assembly: versatile approaches to
construct highly ordered nanostructures. Chem. Rev. 116, 13571-13632 (2016).
Lai, Y. T., Tsai, K. L., Sawaya, M. R., Asturias, F. J. & Yeates, T. O. Structure and
flexibility of nanoscale protein cages designed by symmetric self-assembly. J. Am. Chem.
Soc. 135, 7738-7743 (2013).
7
8
9
King, N. P. et al. Accurate design of co-assembling multi-component protein
nanomaterials. Nature 510, 103-108 (2014).
Wörsdörfer, B., Woycechowsky, K. J. & Hilvert, D. Directed evolution of a protein
container. Science 331, 589-592 (2011).
Velonia, K., Rowan, A. E. & Nolte, R. J. Lipase polystyrene giant amphiphiles. J. Am.
Chem. Soc. 124, 4224-4225 (2002).
10 Heredia, K. L. et al. In situ preparation of protein-“smart” polymer conjugates with
retention of bioactivity. J. Am. Chem. Soc. 127, 16955-16960 (2005).
11 Biswas, S. et al. A tubular biocontainer: metal ion-induced 1D assembly of a molecularly
engineered chaperonin. J. Am. Chem. Soc. 131, 7556-7557 (2009).
12 Sim, S., Niwa, T., Taguchi, H. & Aida, T. Supramolecular nanotube of chaperonin groel:
length control for cellular uptake using single-ring groel mutant as end-capper. J. Am.
Chem. Soc. 138, 11152-11155 (2016).
13 King, N. P. et al. Computational design of self-assembling protein nanomaterials with
atomic level accuracy. Science 336, 1171-1174 (2012).
14 Bale, J. B. et al. Accurate design of megadalton-scale two-component icosahedral protein
complexes. Science 353, 389-394 (2016).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
15 Lai, Y. T. et al. Designing and defining dynamic protein cage nanoassemblies in solution.
Sci. Adv. 2, e1501855 (2016).
16 Kobayashi, N. et al. Self-assembling nano-architectures created from a protein nano-
building block using an intermolecularly folded dimeric de novo protein. J. Am. Chem.
Soc. 137, 11285-11293 (2015).
17 Padilla, J. E., Colovos, C. & Yeates, T. O. Nanohedra: using symmetry to design self
assembling protein cages, layers, crystals, and filaments. Proc. Natl Acad. Sci. USA 98,
2217-2221 (2001).
18 Sciore, A. et al. Flexible, symmetry-directed approach to assembling protein cages. Proc.
Natl Acad. Sci. USA 8681–8686 (2016).
19 Bowman, G. D. & Poirier, M. G. Post-translational modifications of histones that influence
nucleosome dynamics. Chem. rev. 115, 2274-2295 (2014).
20 Cornish, V. W., Mendel, D. & Schultz, P. G. Probing protein structure and function with
an expanded genetic code. Angew. Chem. Int. Edn. Engl. 34, 621-633 (1995).
21 Mills, J. H. et al. Computational design of a homotrimeric metalloprotein with a
trisbipyridyl core. Proc. Natl Acad. Sci. USA 113, 15012–15017 (2016).
22 Lavigueur, C. et al. Thermoresponsive giant biohybrid amphiphiles. Polym. Chem. 2, 333-
340 (2011).
23 Xia, Y., Tang, S. & Olsen, B. D. Site-specific conjugation of RAFT polymers to proteins
via expressed protein ligation. Chem. Commun. 49, 2566-2568 (2013).
24 Ren, Q., Kwan, A. H. & Sunde, M. Two forms and two faces, multiple states and multiple
uses: properties and applications of the self assembling fungal hydrophobins.
Biopolymers 100, 601-612 (2013).
25 Hobley, L. et al. BslA is a self-assembling bacterial hydrophobin that coats the Bacillus
subtilis biofilm. Proc. Natl Acad. Sci. USA 110, 13600-13605 (2013).
26 Theil, E. C. Ferritin: structure, gene regulation, and cellular function in animals, plants, and
microorganisms. Annu. Rev. Biochem. 56, 289-315 (1987).
27 Simons, K., Helenius, A., Leonard, K., Sarvas, M. & Gething, M. J. Formation of protein
micelles from amphiphilic membrane proteins. Proc. Natl Acad. Sci. USA 75, 5306–5310
(1978).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
28 Hsia, Y. et al. Design of a hyperstable 60-subunit protein icosahedron. Nature 535, 136-
139 (2016).
29 Israelachvili, J. N. Intermolecular and surface forces. (Academic press, 2011).
30 Liu, Y., Patricelli, M. P. & Cravatt, B. F. Activity-based protein profiling: the serine
hydrolases. Proc. Natl Acad. Sci. USA 96, 14694-14699 (1999).
31 Rostovtsev, V. V., Green, L. G., Fokin, V. V. & Sharpless, K. B. A stepwise huisgen
cycloaddition process: copper (I) catalyzed regioselective “ligation” of azides and
terminal alkynes. Angew. Chem. Int. Edn. Engl. 114, 2708-2711 (2002).
32 Bader, B. et al. Bioorganic synthesis of lipid-modified proteins for the study of signal
transduction. Nature 403, 223-226 (2000).
33 Ceres, P. & Zlotnick, A. Weak protein− protein interactions are sufficient to drive
assembly of hepatitis B virus capsids. Biochemistry 41, 11525-11531 (2002).
34 Jacques, D. A. & Trewhella, J. Small angle scattering for structural biology-Expanding
the frontier while avoiding the pitfalls. Protein Sci. 19, 642-657 (2010).
35 Speir, J. A., Munshi, S., Wang, G., Baker, T. S. & Johnson, J. E. Structures of the native
and swollen forms of cowpea chlorotic mottle virus determined by X-ray crystallography
and cryo-electron microscopy. Structure 3, 63-78 (1995).
36 Desie, G., Boens, N. & Schryver F. C. D. Study of the time-resolved tryptophan
fluorescence of crystalline. alpha.-chymotrypsin. Biochemistry 25, 8301-8308 (1986).
37 Provencher, S. W. & Gloeckner, J. Estimation of globular protein secondary structure from
circular dichroism. Biochemistry 20, 33-37 (1981).
38 Cameron, J. F. & Frechet, J. M. Photogeneration of organic bases from o-nitrobenzyl-
derived carbamates. J. Am. Chem. Soc. 113, 4303-4313 (1991).
39 Kegel, W. K. & van der Schoot, P. Competing hydrophobic and screened-coulomb
interactions in hepatitis B virus capsid assembly. Biophys. J. 86, 3905-3913 (2004).
40 Evans, M. J. & Cravatt, B. F. Mechanism-based profiling of enzyme families. Chem. Rev.
106, 3279-3301 (2006).
41 Rosen, C. B. & Francis, M. B. Targeting the N terminus for site-selective protein
modification. Nat. Chem. Biol. 13, 1094–1101 (2017).
42 Wagenbauer, K. F., Sigl, C. & Dietz, H. Gigadalton-scale shape-programmable DNA
assemblies. Nature 552, 78-83 (2017).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
43 Ljubetič, A. et al. Design of coiled-coil protein-origami cages that self-assemble in vitro
and in vivo. Nature Biotechnol. 35, 1094–1101 (2017).
44 Mou, Y., Yu, J. Y., Wannier, T. M., Guo, C. L. & Mayo, S. L. Computational design of co-
assembling protein-DNA nanowires. Nature 525, 230-233 (2015).
45 Mosayebi, M. et al. Beyond icosahedral symmetry in packings of proteins in spherical
shells. Proc. Natl Acad. Sci. USA 114, 9014-9019 (2017).
46 Garcia-Seisdedos, H., Empereur-Mot, C., Elad, N. & Levy, E. D. Proteins evolve on the
edge of supramolecular self-assembly. Nature 548, 244–247 (2017).
47 Smith, M. L. et al. Modified tobacco mosaic virus particles as scaffolds for display of
protein antigens for vaccine applications. Virology 348, 475-488 (2006).
48 Ma, Y., Nolte, R. J. & Cornelissen, J. J. Virus-based nanocarriers for drug delivery. Adv.
Drug Deliv. Rev. 64, 811-825 (2012).
49 Weissleder, R., Nahrendorf, M. & Pittet, M. J. Imaging macrophages with nanoparticles.
Nat. Mater. 13, 125-138 (2014).
50 Smanski, M. J. et al. Synthetic biology to access and expand nature's chemical diversity.
Nat. Rev. Microbiol. 14, 135-149 (2016).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Figure 1 | Overview of the technology. a. Crystal structure of fungal and bacterial
hydrophobins (PDB accession number 2B97 and 4BHU) and designed synthetic amphiphilic
protein - consists of globular water soluble protein and hydrophobic tail connected by
hydrophilic linker (hydrophilic and hydrophobic faces are shown in blue and red, respectively).
b. Summary of chemical methodology to make synthetic amphiphilic proteins. i. The
amphiphilic probe was solubilised in the presence of triton X-100, ii. Micelle-assisted protein
labelling rection. iii. Three-step purification to remove triton X-100 and the unreacted protein
using IEX, SEC and standard desalting. iv. Programmed protein self-assembly. c. Crystal
structures of serine proteases used in this study (trypsin, chymotrypsin, subtilisin, proteinase K -
PDB accession number 4MTB, 1GLO, 4C3V, 5B1D respectively) d, e. Chemical structures of
hydrophilic linker and hydrophobic tails. f. General scheme for making libraries of synthetic
amphiphilic proteins. g, h, i. Cartoon representation of protein complexes. g. Protein variants
(Library I). h. Spacer variants (Library II). i. Tail variants (Library III).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
1.2
- Try-CEG-C12-3T
- Try-DDEG-C12-3T
- Try-OEG-C12-3T
- Try-TEG-C12-3T
- Try-DEG-C12-3T
24187 24280
24219 24456
- Try-OEG-C18-3T
- Try-OEG-C18-2T
- Try-OEG-C18-1T
- Try-OEG-C12-3T
- Try-OEG-C12-2T
- Try-OEG-C12-1T
- Try-OEG-C6-3T
- Try-OEG-C6-2T
24109
24101
24453
- Try-OEG-C12-3T
- Chy-OEG-C12-3T
- Sub-OEG-C12-3T
- Pro-OEG-C12-3T
24648
24453
26579
30068
24824
27255
24453
24698
24282
25419
0.8
0.4
0.0
28425
23272
23272
23272
28897
20000
25000
Mass (m/z)
30000
20000
25000
30000
20000
30000
40000
Mass (m/z)
Mass (m/z)
1.2
0.8
0.4
0.0
1.2
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
Protein Conjugate
Protein Conjugate
NaCl
Protein Mixture
Triton X-100
Native Protein
0.8
0.4
0.0
0
20
40
60
80
25
50
75
0
40
80
120
160
Elution Volume (mL)
Elution Volume (mL)
Elution Volume (mL)
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
26579
28425
- Try-OEG-C12-3T
- Chy-OEG-C12-3T
- Sub-OEG-C12-3T
- Pro-OEG-C12-3T
- Try-CEG-C12-3T
- Try-DDEG-C12-3T
- Try-OEG-C12-3T
- Try-TEG-C12-3T
- Try-OEG-C18-2T
24648
24453
24280
24456
24280
24219
30068
24453
24101
24187
- Try-OEG-C18-1T
- Try-OEG-C12-3T
- Try-OEG-C12-2T
- Try-OEG-C12-1T
- Try-OEG-C6-3T
24824
24453
20000
25000
30000
20000
30000
Mass (m/z)
40000
20000
25000
Mass (m/z)
30000
Mass (m/z)
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Figure 2 | MALDI-ToF characterization of amphiphilic proteins before and after
purification. a, g. Molecular weights of protein variants (Library I). b, h. spacer variants
(Library II) and c, i. tail variants (Library III) before and after purification respectively. To
determine molecular weights and to monitor the extent of protein modification, the samples were
directly withdrawn from reaction mixture using a pipette. To check the purity of protein
conjugate, after size exclusion, 98 µL of the sample from fractions containing the modified
proteins, were mixed with 2 µL of triton X-100 (2% or 100 times the CMC) in a separate
microcentrifuge tube and vortexed for 4 h. After that, 1:1:1 ratio of protein sample, 2% TFA and
final matrix mixture were mixed and monitored using MALDI-ToF MS. d, e, f. Represents three-
step purification for separating protein conjugate from reaction mixtures. d. Representative IEX
chromatogram showing the separation of of triton X-100 form protein mixture (native and
modified protein). e. Representative SEC chromatogram showing the separation of native and
modified proteins. The modified protein appears larger in size as evident from elution volume,
because of complex formation, f. Representative desalting chromatogram, shows the removal of
salt from the purified conjugates.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
1.2
- Try-CEG-C12-3T
- Try-DDEG-C12-3T
- Try-OEG-C12-3T
- Try-TEG-C12-3T
- Try-OEG-C18-2T
- Try-OEG-C12-3T
- Chy-OEG-C12-3T
- Sub-OEG-C12-3T
- Pro-OEG-C12-3T
11.8mL
11.6mL
10.7mL
11.7mL
11.4mL
- Try-OEG-C18-1T
- Try-OEG-C12-3T
- Try-OEG-C12-2T
11.2mL
10.9mL
11.4mL
12mL
10.6mL
11.5mL
11.4mL
0.8
0.4
0.0
7.5
10.0
12.5
15.0
7.5
10.0
12.5
15.0
7.5
10.0
12.5
15.0
Elution Volume (mL)
Elution Volume (mL)
Elution Volume (mL)
5
1
5
1
Try-OEG-C18-2T
- Try-CEG-C12-3T
- Try-OEG-C12-3T
- Try-OEG-C18-1T
- Try-OEG-C12-3T
- Try-OEG-C12-2T
- Try-OEG-C12-1T
- Try-DDEG-C12-3T
- Try-OEG-C12-3T
- Try-TEG-C12-3T
- Chy-OEG-C12-3T
- Pro-OEG-C12-3T
1
0.1
0.1
0.01
1E-3
0.1
0.01
1E-3
0.01
1E-3
0.015
0.015
0.1
0.35
0.015
0.1
0.35
0.1
0.35
Q (Å-1)
(Å-1)
Q (Å-1
)
Q
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
- Try-CEG-C12-3T
- Try-DDEG-C12-3T
- Try-OEG-C12-3T
- Try-TEG-C12-3T
- Try-OEG-C12-3T
- Chy-OEG-C12-3T
- Sub-OEG-C12-3T
- Pro-OEG-C12-3T
- Try-OEG-C18-2T
- Try-OEG-C18-1T
- Try-OEG-C12-3T
- Try-OEG-C12-2T
- Try-OEG-C12-1T
- Try-OEG-C6-3T
1
10
Dh / nm
100
1
10
Dh / nm
100
1
10
Dh / nm
100
108
107
106
105
104
103
108
107
106
105
104
103
1.2
1.2
0.8
0.4
0.0
1.2
0.8
0.4
0.0
108
107
106
105
104
- Try-OEG-C12-2T
- Try-CEG-C12-3T
- Try-OEG-C12-3T
261 KDa
276 KDa
400 KDa
0.8
0.4
0.0
7.5
10.0
12.5
15.0
7.5
10.0
Elution Volume (mL)
12.5
15.0
7.5
10.0
12.5
15.0
Elution Volume (mL)
Elution Volume (mL)
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Figure 3 | Self-assembly studies of protein complexes. a, b, c. SEC data - The molecular
weights of the protein complexes were calculated from elution volumes of standard proteins.
SEC runs were performed in Superdex-200 10/300 GL in 50 mM sodium phosphate pH 7.4, 1 M
NaCl with 0.25 mL/min as the flow rate. The concentration and volume of protein samples were
5 mg/mL and 0.5 mL respectively. d, e, f. SANS data - For SANS measurements, the samples (5
mg/mL) were dissolved in 50 mM sodium phosphate pH 7.4 prepared in D2O, in order to
minimize the incoherent scattering and to increase contrast. The temperature was kept fixed at 30
C during the measurements. g, h, i. DLS data - Measurements were carried out at 5 mg/mL
concentration in 50 mM sodium phosphate pH 7.4. 1 mL of sample was taken in disposable
polystyrene cells and the mean size of the complexes was measured at 90º scattering angle. j, k,
l. SEC-MALS Data - Molecular weight determination of protein complexes using SEC-MALS
was performed on a Superdex-200 10/300 GL column connected to an Agilent HPLC system
equipped with the 18-angle light scattering detector and a refractive index detector. 100 µL of the
protein samples at concentrations (5 mg/mL) were injected, and the molecular weight was
calculated using ASTRA software. m. Electrostatic surface potential of native proteins.
Negatively charged residues like Asp and Glu are colored red. Positively charged residues like
Arg and Lys are colored blue. His residues are colored green and active site serine alone is
represented in magenta (site of conjugation).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Table 1 | Summary of bio-physical data of protein complexes
Protein
Conjugate
Elution
Vol SEC
(mL)
Mol Wt -
SEC
(kDa)
Mol Wt
SANS (kDa)
Oligomeric Rg from P(r)
Rh
Shape
Factor
(σ = Rg/Rh)
Zeta
potential
(mV)
state
DLS (nm)
SANS (Å)
SANS
(mer)
Library I Protein Variants
Try-OEG-C12-3T
Chy-OEG-C12-3T
Pro-OEG-C12-3T
Sub-OEG-C12-3T
11.4
11.5
11.2
10.9
276
258
300
339
245
266
661
----
10
10
22
----
41
40
53
---
5.7
5.2
7.0
8.8
0.72
0.77
0.76
----
16.5
8.7
-3.2
-0.7
Library II Spacer Variants
Try-TEG-C12-3T
*Try-OEG-C12-3T
Try-DDEG-C12-3T
Try-CEG-C12-3T
11.7
11.4
10.7
10.6
241
276
375
388
243
245
320
372
10
10
13
15
32
41
49
49
5.3
5.7
7.3
6.7
0.60
0.72
0.67
0.73
10.3
16.5
9.8
4.6
Library III Tail Variants
Try-OEG-C6-3T
Try-OEG-C12-1T
Try-OEG-C12-2T
*Try-OEG-C12-3T
Try-OEG-C18-1T
Try-OEG-C18-2T
17.3
17.3
11.8
11.4
12.0
11.7
23
73
3
35
37
42
41
43
47
4.2
5.0
5.5
5.7
5.5
5.9
-----
0.74
0.76
0.72
0.78
0.79
6.5
5.5
23
289
267
245
411
440
12
11
10
17
18
225
276
210
241
7.6
16.5
7.1
12.6
* Included again for the sake of comparison.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Figure 4 | Programmed dis-assembly of protein complex. a. Schematic representation of
irreversible dis-assembly of protein complex upon photo-irradiation. The photo-sensitive
conjugate was dissolved in 50 mM sodium phosphate pH 7.4 in a beaker and the temperature
was maintained using ice bath during entire reaction. The sample was exposed to the UV lamp
(range: 305-315 nm). Samples were withdrawn at different time points using a pipette and
analysis of the same was carried out using SEC and MALDI-ToF MS. b. SEC chromatogram of
time-dependent photo-chemical irreversible dis-assembly. c. MALDI-ToF data of protein
conjugate before and after cleavage.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
METHODS
Selection of proteins used in this work
We chose 4 serine proteins for our study i.e. trypsin (23 kDa), chymotrypsin (25 kDa),
subtilisn (27 kDa) and proteinase K (29 kDa). Proteins were purchased from commercial vendors
and molecular weights were determined prior to modification to track the molecular weight
changes accurately.
Matrix preparation and molecular weight determination
Apart from molecular weight determination of native proteins, all stages of protein
modification and purification were also followed by MALDI-ToF MS. The samples were
analysed in Linear High Mass mode in AB Sciex 4800 plus MALDI-ToF/ToF analyzer with
4000 Series Explorer as software. Mass was scanned between 10,000 Da and 40,000 Da with
focus mass at 25,000-29,000 Da depending on the protein analysed.
4 mg of DHAP and 4.5 mg of DAHC were weighed separately in microcentrifuge tubes.
150 µL of ethanol and 200 µL of milli-Q water were added to each tube, respectively. Both the
solutions were sonicated using bath sonicator for 1 minute and vortexed for another minute.
Then 50 µL of DAHC aqueous solution was transferred to DHAP solution and the resulting
solution was vortexed for one more minute to yield matrix mixture. The sample preparation was
carried out in a 500 µL microcentrifuge tube, 1:1:1 ratio of protein sample, 2% trifluoroacetic
acid (TFA) and matrix mixture were mixed and vortexed for 1 minute. The samples were then
kept undisturbed. When the crystallization observed in centrifuge tubes, 0.5 µL of the samples
were spotted on the MALDI plate and air dried for 15 minutes. The plate was then loaded and
fired to get accurate molecular weight both in +1 and +2 states. 100 µM protein concentration
was found to be optimum for MALDI-ToF MS analysis.
Protein modification and monitoring
Protein modification was carried out at the concentration of 100 µM, which was found to
be optimum for MALDI-ToF MS monitoring. Triton X-100 was used to solubilize the AABPs at
concentrations 100 times (20 mM) more than CMC or 2% of total volume of reaction mixture.
Typically, for test reactions, final volume of reaction mixture is 1 mL. Proteins were weighed
(2.3 to 2.9 mg, depending on protein) in microcentrifuge tubes and 500 µL 50 mM sodium
phosphate pH 7.4 was added and mixed gently with a pipette to make 200 µM solutions. Then
AABPs (1 or 2 equivalents) were weighed in a different microcentrifuge tube, followed by
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
addition of 20 µL triton X-100 and 480 µL of 50 mM sodium phosphate pH 7.4 and vortexed for
15 minutes. When the AABP solution becomes clear, the protein solution was added into AABP
solution to get 100 µM (1 mL) protein solutions and allowed to react for 24 h on rotospin at 20
rpm at 25 ºC. For probes FP-OEG-C18-2T and FP-OEG-C18-3T, 10 equivalents were weighed
in a microcentrifuge tube and 4% triton X-100 was added and sonicated till the mixture became
homogenous (please note that we used bath sonicator to solubilize these two AABPs). Protein
modifications were carried out in falcon tubes at 200 mg scale following the linear scale up of
above mentioned procedure for understanding the self-assembly behavior.
To monitor the extent of protein modification, the samples were directly withdrawn from
reaction mixture using a pipette, and analyzed. In brief, 2 µL of reaction mixture was mixed with
2 µL of 2% TFA and 2 µL of matrix mixture as previously stated, vortexed and spotted on
MALDI-ToF MS plate. Please note that the matrix preparation and sample preparation
procedures remained same here.
Purification of modified proteins
All the conjugates were purified by three-step purification i.e IEX, SEC and desalting,
performed using either Aktaprime or Aktaprime plus or Akta Explorer or Akta Pure. IEX was
performed to remove triton X-100 using either SP sepharose or Q sepharose resins (GE)
depending on isoelectric point (pI) and surface charges of proteins (Try: 30.4, Chy: 19.5, Pro K: -
8.4, Sub: -6.2). For example, to purify the reaction mixture of trypsin or chymotrypsin, we used
SP sepharose, a cation-exchange resin at pH 7.4 and to purify subtilisin and proteinase K we
used Q sepharose, an anion-exchange resin at pH 10.
During cation-exchange chromatography, the column was pre-equilibrated with the same
buffer (50 mM sodium phosphate pH 7.4) which was used for modification and then sample was
injected followed by post injection equilibration for at least 2 Column Volumes (CVs) or until
the complete removal of triton X-100 for large scale reactions. The elution of native protein and
its corresponding conjugate together as mixture was later achieved using 50 mM sodium
phosphate pH 7.4, 1 M NaCl as elution buffer.
In the case of anion-exchange chromatography, the protein reaction mixtures were buffer-
exchanged first to 50 mM tris base pH 10 (since the buffer used for modification was 50 mM
sodium phosphate pH 7.4). Then the anion-exchange column was pre-equilibrated using same
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
buffer (50 mM tris base pH 10) which was used for buffer-exchange and then sample was
injected followed by post injection equilibration for at least 2 CVs or until the complete removal
of triton X-100 for large scale reactions. The elution of the native protein and its conjugate
together as mixture was later achieved using 50 mM tris base pH 10, 1 M NaCl as elution buffer.
The obtained IEX fractions were subjected to SEC immediately to remove the native proteins
from the protein conjugates. For the separation of native protein from protein conjugate, 50 mM
sodium phosphate pH 7.4, 1 M NaCl was used as buffer using either Sephacryl S-100 HR 16/60
or Sephacryl S-200 HR 16/60 or Sephacryl S-300 HR 16/60. The NaCl was later removed by
Sephadex-G25 desalting column. The desalted fractions were quickly lyophilized and later
dissolved in required buffer when needed.
Monitoring of IEX, SEC and desalted fractions were carried out using the same
procedure mentioned for native proteins and reaction mixtures except for the addition of triton
X-100 in protein fractions. In short, 98 µL of samples from each fraction were mixed with 2 µL
of triton X-100 (2% or 100 times the CMC) in a separate microcentrifuge tube and vortexed for 4
h. The samples were then analysed using the same procedure as mentioned above. Please note
that the matrix and sample preparation procedure remained same.
Size exclusion-based molecular weight determination
To determine the molecular weights of soluble protein complexes, we performed SEC in
Superdex-200 10/300 GL (GE healthcare). For this purpose, we chose standard proteins (GE
Healthcare) i.e. blue dextran (2000 kDa), to determine the void volume of this column,
thyroglobulin (660 kDa), ferritin (440 kDa), aldolase (158 kDa), conalbumin (75 kDa),
ovalbumin (44 kDa), carbonic anhydrase (29 kDa) and trypsin (23 kDa).
A series of size exclusion runs were performed with these proteins in 50 mM sodium
phosphate pH 7.4, 1 M NaCl with 0.25 mL/min as the flow rate. All the proteins were dissolved
individually in milli Q water at 3 mg/mL and 0.5 mL was injected. With the obtained elution
volumes, the partition coefficients (Kav) for standard proteins were calculated according to
standard protocol (GE Healthcare) using the formula Kav = Ve-Vo / Vc-Vo where, Vo = column
void volume (8.5 mL, from the elution volume of blue dextran), Ve = elution volume of
particular sample, and Vc = geometric CV (23.5 mL). Calibration curve was plotted for Kav of
standards against relative molecular weights (Mr) of standard proteins. (r2 = 0.99, please refer to
the calibration curve). Now, to determine the molecular weights of soluble protein complexes,
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
the protein conjugates were dissolved individually in milli Q water at 5 mg/mL and 0.5 mL was
injected.
Dynamic light scattering
Analysis of protein complexes was performed using DLS using Zetasizer Nano 2590
(Malvern, UK). Protein samples (5 mg/mL) were prepared in 50 mM sodium phosphate pH 7.4.
1 mL of sample was taken in disposable polystyrene cells and then the mean size of the
complexes was measured at 90° scattering angle.
SANS measurements
SANS experiments were carried out at SANS spectrometer operating at Dhruva Reactor, Bhabha
Atomic Research Centre (BARC), Mumbai, India51. The facility employs a neutron velocity
selector to provide a monochromatic neutron beam with a mean wavelength (λ) of 5.2 Å and
wavelength spread (Δλ/λ) approximately 15%. The scattered neutrons were detected using a one
3
dimensional He position sensitive detector (PSD) in an angular range of 0.5-15 corresponding
to wave vector transfer (Q = 4πsin(θ/2)/λ, where θ is scattering angle) range of 0.015-0.3 Å-1.
The measured SANS data were corrected for the background, the empty cell contribution, and
the transmission. The corrected data were presented on an absolute scale using the standard
protocols. For SANS measurements, the samples (5 mg/mL) were prepared in sodium phosphate
pH 7.4 buffer in D2O in order to minimize the incoherent scattering and to increase contrast. The
temperature was maintained 30°C during the measurements.
SANS Analysis
SANS is a very promising technique to investigate self assembled structures in nanometer
length scales. In SANS experiment, one usually measures the differential scattering cross section
(dΣ/dΩ) per unit volume as a function of wave vector transfer (Q). For monodispersed particles
in a medium, it can be expressed as52
dΣ
(1)
= n P Q S Q + B
( ) ( )
dΩ
where n is the particle number density. P(Q) is the intra-particle structure factor (square of the
form factor) and S(Q) is the inter-particle structure factor. P(Q) provides the specific size and
shape of the scatterer whereas S(Q) decides the spatial arrangement of the particles and thereby
gives the information about the inter-particle interaction. For a dilute system, S(Q) may be
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
approximated to unity. B is a constant term representing the incoherent background arising
mainly from the hydrogen atoms present in the sample.
P(Q) for spherical particles of radius R is given by53
2
16π 2
(
)
(
⎡
⎢
⎢
⎣
⎤
⎥
⎥
⎦
2
)
sin QR − QR cos QR
) ( )
(2)
P Q =
( )
ρ − ρ R6 3
(
p
s
3
9
QR
(
)
where ρp and ρs are the scattering length densities of particle and solvent respectively.
For prolate ellipsoidal particles, P(Q) can be expressed as53
1
16π 2
2
2
(3)
P Q =
( )
ρ − ρ
ab2
⎡F Q,µ ⎤2 dµ
⎣ ⎦
(
)
(
)
(
)
∫
0
p
s
9
where the functions are given by
x = Q[a2µ2 +b2(1− µ2 )]1
and
3(sinx − xcos x )
2
F(Q,µ ) =
x3
where a and b=c are, respectively semi-major and semi-minor axes of prolate ellipsoid (a>b=c).
The variable µ is the cosine of the angle between the directions of a and Q.
The following expression provides P(Q) for a spherical core-shell structure53
16π2
sinQ(R + dR)−Q(R + dR)cosQ(R + dR)
Q3(R + dR)3
sinQR −QRcosQR
2
P(Q) =
[(R + dR)3 ρ − ρ
− R3 ρshell − ρcore
]
(
)
(
)
shell
s
9
Q3R3
where R is core radius and dR is shell thickness. ρcore and ρshell represents the scattering length
densities of the core and shell, respectively.
Throughout the data analysis corrections were made for instrumental smearing. The calculated
scattering profiles were smeared by the appropriate resolution function to compare with the
measured data. The parameters in the analysis were optimized by means of a nonlinear least-
square fitting program54.
SEC-MALS
SEC-MALS analyses of the protein complexes were performed on a Superdex-200
10/300 GL column (GE Healthcare) connected to an Agilent HPLC system equipped with the
18-angle light scattering detector (Wyatt Dawn HELIOS II) and a refractive index detector
(Wyatt Optilab TrEX). The system was calibrated with BSA at a concentration of 2 mg/mL; 100
µL of the protein conjugates at concentrations 5 mg/mL were injected, and the molecular weights
were calculated using ASTRA software (Wyatt Technologies).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802824
Chemistry - A European Journal
Fluorescence Spectroscopy
The fluorescence of native proteins and protein conjugates were measured using Xenon
arc lamp on Fluorolog spectrofluorimeter (HORIBA, Jobin-Yvon). Protein samples (1 mg/mL)
were prepared in 50 mM sodium phosphate pH 7.4. Measurements were carried out at tryptophan
specific excitation wavelength of 295 nm in a 10 mm path length cuvette. The samples remained
clear during the meseaurement.
Circular Dichroism (CD) Spectroscopy
The CD spectra of native proteins and protein conjugates were recorded on Jasco J-815
spectrometer. Protein samples (0.25 mg/mL) were prepared in 50 mM sodium phosphate pH 7.4
and measurements were performed in a 1 mm path length cuvette. The spectra (background
corrected) were obtained at scanning speed of 50 nm/minute within range of 195-250 nm. Mean
residue elasticity was converted from accumulated spectra of five measurements for each
sample.
Zeta potential measurement
The zeta potential for native proteins and protein conjugates was measured using
Zetasizer Nano 2590 (Malvern, UK). Protein samples (1 mg/mL) were prepared in milli-Q water.
Programmed Dis-assemby of photo-sensitive supramolecular protein complex
The photo-sensitive protein conjugate was dissolved in 50 mM sodium phosphate pH 7.4
in a 50 mL beaker and temperature was maintained at 0° C using ice bath over the entire
reaction. The sample was exposed to the Sankyo Denki G8T5E UVB linear lamp (8 W output)
which emits ultraviolet rays between 280 to 360 nm (peak at 305-315 nm). Samples were
withdrawn at different time points using a pipette and analysis of the same was carried out using
SEC and MALDI-ToF MS.
This article is protected by copyright. All rights reserved.