doi.org/10.26434/chemrxiv.8108105.v2

Biomimetic Composites with Enhanced Toughening Using Silk InspiredTriblock Proteins and Aligned Nanocellulose ReinforcementsPezhman Mohammadi, A. Sesilja Aranko, Christopher P. Landowski, Olli Ikkala, Wolfgang Wagermaier,Markus Linder

Submitted date: 12/07/2019 • Posted date: 12/07/2019Licence: CC BY-NC-ND 4.0Citation information: Mohammadi, Pezhman; Aranko, A. Sesilja; Landowski, Christopher P.; Ikkala, Olli;Wagermaier, Wolfgang; Linder, Markus (2019): Biomimetic Composites with Enhanced Toughening UsingSilk Inspired Triblock Proteins and Aligned Nanocellulose Reinforcements. ChemRxiv. Preprint.

Silk and cellulose are biopolymers that show a high potential as future sustainable materials.They also havecomplementary properties, suitable for combination in composite materials where cellulose would form thereinforcing component and silk the tough matrix. Therein, a major challenge concerns balancing structure andproperties in the assembly process. We used recombinant proteins with triblock architecture combiningstructurally modified spider silk with terminal cellulose affinity modules. Flow-alignment of cellulose nanofibrilsand triblock protein allowed a continuous fiber production.The protein assembly involved phase separationinto concentrated coacervates, with subsequent conformational switching from disordered structures to betasheets. This gave the matrix a tough adhesiveness, forming a new composite material with high strength andstiffness combined with increased toughness. We show that versatile design possibilities in proteinengineering enable new fully biological materials, and emphasize the key role of controlled assembly atmultiple length scales for realization.

File list (2)

download fileview on ChemRxivSupplementary _Mohammadi.pdf (2.75 MiB)

download fileview on ChemRxivManuscript_Mohammadi.pdf (1.11 MiB)

Supporting information

Biomimetic Composites with Enhanced Toughening Using Silk Inspired Triblock Proteins and Aligned

Nanocellulose Reinforcements

Pezhman Mohammadi*, A Sesilja Aranko, Christopher Paul Landowski, Wolfgang Wagermaier and Markus B Linder* P. Mohammadi, A.S. Aranko, M.B. Linder Department of Bioproducts and Biosystems, School of Chemical Engineering, Aalto University, P.O. Box 16100, FI-16100, Espoo, Finland. E-mail: pezhman. Mohammadi @aalto .fi and markus.linder@aalto.fi C.P Landowski VTT Technical Research Centre of Finland Ltd., Espoo, Finland. W. Wagermaier Department of Biomaterials, Max Planck Institute of Colloids and Interfaces, D 14424 Potsdam, Germany.

Cloning, expression & protein purification

Cloning, expression and protein purification carried out according to earlier publication 1. Briefly,

DNA sequences encoding bacterial family three cellulose binding module (CBM3) from

Ruminiclostridium Thermocellum (PDB Accession number: 1NBC) 2 and DNA sequence encoding

Aaraneus diadematus major ampulla gland (ADF3) and twelve time repeats of residues 325-368

(eADF3) 3–5 were synthesized and codon optimized by GeneArt gene synthesis (ThermoFisher

Scientific) for expression in E. coli . Similarly, eADF3 sequence also synthesized and codon optimized

by GeneArt gene synthesis for expression in Pichia pastoris. Constructs were either made by

conventional piece by piece cloning or golden gate cloning6–8. Resulting constructs named, CBM-

eADF3-CBM, CBM-ADF3-CBM, CBM-eADF3, CBM-CBM and CBM. XL1B (New England biolab® 5-

alpha F'Iq Competent E. coli (F´ proA+B+lacIq ∆(lacZ)M15 zzf::Tn10 (TetR) / fhu A2∆(argF-lacZ)U169

phoA glnV44 Φ80Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17)) and BL21 (F-ompT hsdSB

(rB-mB-) gal dcm (DE3)) (ThermoFisher Scientific) were used as cloning and expression vectors. In

general for all the expressions either EnPresso® B medium (BioSilta, Oulu, Finland) technology were

12

used or MagicMediaTM E.coli expression medium (ThermoFisher Scientific) according to instructions

with some changes otherwise states in the text. After 15-24 hours of induction cells were harvested

(Centrifugation at 16,000 ×g, 15min, 4°C), washed and lysed. To make eADF3 alone, codon

optimized sequence for expression in Pichia was inserted in pPICZ-α (Invitrogen) expression vectors

and then transferred into Top10 E. coli cells (ThermoFisher Scientific) and selected on zeocin low

salt LB plates. After initial screening assembled plasmids linearized and transformed into the X33

Pichia pastoris strain selected for on zeocin plates and then cultured overnight in BMGY expression

medium. The cells were harvested by centrifugation (1,000 ×g, 5 min) after the density reached an

OD600 value of 3-6. The protein production was initiated by diluting the cells to OD600=1.0 in 250 ml

of BMMY medium in 2-liter flasks, adding protease inhibitors chymostatin and pepstatin A, and

adding methanol daily to a final concentration of 2%. Proteins were purified either with HisTrap FF

crude columns (GE Healthcare Life Science) connected to a AKTA-Pure liquid chromatography. For

larger scale purification cells were lysed and supernatant heat shocked at 70-75 °C for 30min and

then centrifuged at 16,000 ×g for 80 min at 4°C and desalted using Econo-Pac® 10DG column.

Samples were then frozen in liquid nitrogen and stored at -80°C until use.

Liquid-like coacervation (LLC) preparation for the fiber spinning

The liquid-like coacervate formation carried out as described earlier 1. Briefly, Under controlled

ambient conditions never dried and dilute protein solution (0.01% w/v) of CBM-eADF3-CBM) and

CBM-ADF3-CBM suspended in Milli-Q water derived to under going liquid-liquid demixing and

coacervation by gradually concentrating the solution over the critical concertation of 0.8% w/v in

which micrometer size protein droplets formed. Samples were concentrated further in order to

reach to the final working and storage volume of 1ml and overall concentration of 4% w/v. This at

the end led to the formation of a protein reach phase of about 600-650 µl which was undisturbed

throughout the entire process. In order to homogeneously distribute the LLCs throughout the entire

13

1ml suspension of the dense LLC, condense phase containing LLCs where mixed gently and

immediately flash freezed in liquid nitrogen and stored at -80 to preserve the structure of LLCs. Each

1ml storage volume was thawed and used only once for the spinning. Unless otherwise stated

proteins were in Milli-Q water without the presence of any inorganic compound (for instance salts)

or organic compound (such as organic buffering reagent such as Tris, HEPES and etc) which would

interfere with the colloidal stability of the native cellulose nanofibrils. For other constructs (CBM,

CBM-CBM, AQ12, and CBM-AQ12) same procedure was done with differences that none of these

constructs undergo liquid-liquid phase separation with this conditions.

Plant cellulose CNF preparation

Never dried birch pulp (with 24 % hemicellulose content) was disintegrated by passing the pulp

suspension six-time (refer to as 6 passes) through fluidizer (Microfluidics Corp, Newton, MA, U.S.A.)

at 300kWh/t net specific energy with specific edge load of 0.5 J/m and refiner speed of 150, giving a

hydrogel with a consistency of approximately 2 % w/v in water without presence of any organic or

inorganic compounds. CNF suspension was stored in +4 °C the entire time. One hour prior to

experiments, required amount of CNF suspension equilibrated at room temperature.

Optical microscopy

Light microscopy was done using Axio observer inverted microscope 10-40×/1.6 objective and

AxioCam MRm camera (Carl Zeiss, Germany). Images were further processed with ImageJ 9 or

ImageJ Fiji (versions 1.47d) 10 .

Rheological measurement

To study shear-rate-dependence of the steady shear viscosity of the different CNF and

recombinantly produced fusion proteins suspensions, flow ramp with initial torque of 0.06µN.m

(Stress: 0.001Pa) to final torque of 6000µN.m (Stress: 100Pa) using a stress-controlled rheometer

(TA instrument ARG2 rheometer) using stainless steel serrated plate geometry (diameter 22 mm)

14

with a gap distance of 1mm using humidity control chamber All the measurements were carried out

at 23 ℃.

Instrumentation for extrusion of nano-composite silk-cellulose fibers (Cello-Silk)

For small-scale spinning of composite silk fusion protein and CNF fibers (Celo-Silk fiber), an in-house

custom-made experimental setup was used (Figure S1) consisting of four main parts; a pair of

pumps, a pair of sample containers, micro static mixing tee assembly, capillary tubings, a

coagulation bath (96% Ethanol). For larger scale spinning a motorized take up-spinning system

consisting of AC motors, pulleys, drying fans and collecting rollers were included downstream. Prior

to use the CNF and recombinantly produced protein solutions were loaded in each sample container

and then centrifuged (swing bucket rotor) for 10 min at 2000 RPM at 25 centigrade to avoid defects

caused by air bubbles and then connected to each pump. Using the high performance pumps with

the fine adjustment of flow rate (volumetric flow rate of 500 µL/min and linear flow rate of 150

cm/min), eluent was deliver to each sample containers separately and building a pressure behind a

movable seals forcing the CNF and protein solutions through separates capillary tubings (1/16) with

internal diameters 500 µm which intersect at a microstatic mixing tee for combining the two flow

streams. The outlet of the mixing tee connected to capillary tubings with the internal diameter of

500µm and length of 150cm as described before. 11 Recombinantly produced fusion protein-CNF

mixed solution then pushed through the capillary tubing and exit into a coagulation bath of ethanol

to precipitate out and form solid fibers.

15

Figure S1. The scheme of the double injection spinning apparatus used for the fiber spinning. Black

arrows represent direction of the flow and increasing velocity with developing laminar flow and fully

developed laminar flow depict parabolic velocity distribution at the cross section of capillary, with

the highest velocity and lowest shear at the center of capillary and lowest velocity and highest shear

close to the capillary wall

Instrumentation for making cellulose nanopapers

For the preparation of cellulose nanopaper films, a pair of high throughput in-house custom made

centrifugal filtrations apparatus made (Figure S2) consisting of three parts, a waste compartment,

the highly porous metallic substrate (0.1-0.3µm) and sample loading compartment assembling in

the same order. Prior to using porous metallic substrate was fixed on top of the waste container

with additional GVWP (0.22 µm, Millipore, U.S.A) collecting membrane on top sandwiched between

the porous substrate and sample loading compartment. CNF solutions (1% w/v) mixed and loaded

into any of six isolated sample container columns (with O-ring) and placed in a swing bucket rotor.

Two centimeters into two centimeters square, nano-composite silk-cellulose films made by applying

a centrifugal force of 2000 RPM for 30 min at 25 centigrade. To prevent any defect during the cutting

the films were cut while still wet using an in-house custom build metallic block with eight fixed single

16

edge razor blades into seven stripes. To prevent wrinkling a press with 300 g load for 10 min applied

and then changed to a 50 g weight overnight. By precise control over temperature, time and water

removal, we found nanopaper films made with this technique are more homogenous, have less

thickness variation and smoother in comparison to other techniques in the literature using vacuum

filtration apparatus.

Figure S2. Custom build instrumentation used for fabrication of the composite LLC-CNF and only

CNF films to study the effect of orientation of the CNF on the mechanical properties.

Polarized microscopy

For a qualitative observation to study birefringence of the wet-spun celo-silk fibers, polarized

microscopy imaging, done using a Leica DM4500 P LED polarized optical microscope (POM) as

described earlier. 11 Briefly, fibers placed in between two cross-polarizers at angles ±45 and

interference color resulting from the retardation between extraordinary and ordinary waves were

determined.

Scanning electron microscopy (SEM)

17

SEM imaging was performed via a Zeiss FE-SEM field emission microscope with variable pressure for

all the samples (Microscopy center, Aalto University, Espoo, Finland) Microscope was operating at

1–1.5kV for all the samples. A thin platinum coating (approximately 2-4 nm) sputtered onto the

samples prior to imaging of surfaces and cross-sections of the fibers. For further analysis and image

processing software package ImageJ 9 and ImageJ Fiji (versions 1.47d) 10 were used.

Mechanical measurement of the celo-silk fibers

For measuring mechanical properties of specimens, fibers fixed by gluing them between two pieces

of abrasive sand paper. Tensile testing performed on a 5kN tensile / compression module

(Kammrath & Weiss GmbH, Germany) using 100 N load cell with a nominal elongational speed of

8.35 µm/second and gauge length of 10 mm for all the samples as earlier study. 11 Fibers and films

stabilized at 50% relative humidity for at least 24 hours. To characterization the effect of humidity

on the mechanical properties, more specimens stabilized at 20% and 80% humidity. Cross section of

the materials was imaged with SEM for at least six samples at three different position and cross-

section areas measured with ImageJ Fiji software (area calculation function) and averaged. Data

processed using Matlab (MathWorks).

Synchrotron Wide-angle x-ray scattering (WAXS)

Wide-angle X-ray diffraction experiments carried out at the µSpot beamline at BESSY II (Berliner

Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung, Helmholtz-Zentrum Berlin,

Germany). The energy was 15 keV using a silicon 111 monochromator and a beam size of 100 µm.

Samples clamped in a sample holder unit in order to the positioning of the samples perpendicular

to the beam path. Five positions along the length of the fibers measured. After subtraction of dark

current and air scattering from the diffractogram, azimuthal intensity profiles at the (004) reflection

were extracted by sector-wise integration after masking the diffractogram to show only the (004)

18

diffraction pick. Hermans orientation parameters were then calculated according to equations 1

and 2.

𝑆𝑆=3

0𝐼𝐼(𝛷𝛷)𝑐𝑐𝑠𝑠𝑠𝑠𝛷𝛷𝑐𝑐𝑐𝑐𝑐𝑐2𝛷𝛷

∑𝜋𝜋𝐼𝐼(𝛷𝛷)𝑐𝑐𝑠𝑠𝑠𝑠𝛷𝛷

(2)

19

1

2

(1)

with

2𝛷𝛷⟩ − 2⟨𝑐𝑐𝑐𝑐

𝑐𝑐

∑𝜋𝜋

⟨𝑐𝑐𝑐𝑐𝑐𝑐2𝛷𝛷⟩ =

Figure S3. a) Phase contrast light microscopy images of concentrated solutions (4% w/v) of different

constructs showing coacervation of 3 block architecture protein (CBM-ADF3-CBM) versus 2 and

1 block architecture which doesn’t self-assemble into LLCs (scale bars are 10 µm). b) Infiltration of

CNF network by sticky LLC matrix protein of CBM-ADF3-CBM in comparison to 2 and 1 block

matrix protein architecture in free monomeric form (scale bars are 10 µm).

Figure S4. Zero-shear viscosity for only CNF and LLC (CBM-eADF3-CBM)-CNF at five different

blend ratios measured to identify suitable mixtures with a spinnable viscosity (ranging 500-5000

Pa.s). Previously we illustrated CNF with 2% w/v solid content to be the optimum dope concertation

during spinning which produced the best-performing fibers.11 Hence, we used 2% w/v CNF as the

starting concentration; however, it was then mixed with LLC in 1:3, 1:2, 1:1, 2:1 and 3:1 ratio (LLC

to CNF) (Figure S2). At the onset of shear (0.101 s-1), CNF suspension shows 1010 Pa.s viscosity.

Addition of the LLCs leads to increase in the viscosity of suspension dramatically, with values

ranging from 1350 Pa.s (1:3) to 6970 Pa.s (3:1). All the LLC-CNF suspension including pure

CNF showed typical shear thinning characteristic due to entanglement break up. This has

substantial importance as the viscosity decreases by an order of magnitude and thus reduces

the pressure-drop and facilities extrusion of the LLC-CNF suspension through the capillary. Out

of the entire mixing ratios tested, only 3:1 (LLC to CNF) LLC-CNF suspension eliminated from the

further study as the viscosity of the solution excised the spinnability range.

20

Figure S5. Mean value and standard deviations of mechanical properties for only CNF and composite

LLC (CBM-eADF3-CBM)-CNF and LLC (CBM-ADF3-CBM)-CNF fibers. a) Max-stress, b)

Young's modulus, c) Maximum strain, d) Work of fracture, e) Yield point, f) slope after the yield

point and g) Yield strain. h) Stress-strain curves from cyclic measurements of the only CNF and LLC

(CBM-eADF3- CBM)-CNF composite fibers.

21

Figure S6. Mean value and standard deviations of mechanical properties for only CNF, composite

LLC (CBM-eADF3-CBM)-CNF and LLC (CBM-ADF3-CBM)-CNF fibers at different mixing ratios.

a) Max- stress, b) Young's modulus, c) Maximum strain, d) Work of fracture, e) Yield point, f) slope

after the yield point and g) Yield strain.

22

Figure S7. a) Synchrotron WAXS measurement on oriented fibers and not oriented films made from

only CNF and LLC (CBM-eADF3-CBM)-CNF. Mean value and standard deviations of

mechanical properties for oriented only CNF and composite LLC (CBM-eADF3-CBM)-CNF fibers

and not oriented corresponding films. b) Max-stress, c) Young's modulus, d) Maximum strain, e)

Work of fracture, f) Yield point, g) slope after the yield point and h) Yield strain. Only CNF film

demonstrated highest strain, lowest stiffness and strength among all four specimens with a slope

after plastic deformation illustrating typical strain hardening. This is a tendency of isotopic CNF

fibrils network during the pulling forces in which fibrils mainly dislocate and orient themselves in

the direction of stretching after the yield point and during plastic deformation. This dramatically

can change, if CNFs become oriented with direct influences on material mechanical properties.

Orientation in the fibers uplifted almost any of the tensile properties as a result of more advanced

fibril-fibril interaction and increased hydrogen bonding between oriented CNFs. Moreover,

noticeable change in CNF fiber was that strain hardening almost entirely disappeared during plastic

deformation which was correlating strongly to orientation parameter. Infiltration of the both

isotropic and anisotropic CNF network by anisotropic LLC lead to greater mechanical properties.

However, the effect was great if the materials fabricated with isotropic LLC and anisotropic LLC

(fiber) over anisotropic LLC and anisotropic LLC (film).

23

Figure S8. a) Stress-strain curve of only CNF and LLC (CBM-eADF3-CBM) measured at different relative humidity. Mean value and standard deviations of mechanical properties for only CNF, composite LLC (CBM-eADF3-CBM)-CNF and LLC (CBM-ADF3-CBM)-CNF fibers measured at different relative humidity. b) Max-stress, c) Young's modulus, d) Maximum strain, e) Work of fracture, f) Yield point, g) slope after the yield point and h) Yield strain. Distinctly at 80 % RH, pure CNF based fibers showed a stress-strain curve with negligible yield point (15 MPa) and lower stiffness (5 GPa), however with increased plastic deformation (17 %). This can be explained with water acting as a lubricant on the surface of CNFs, in which facilitating the CNFs pass each other during mechanical stretching.

24

Figure S9. Fourier transform infrared spectroscopy (FTIR) of the LLCs before (Dark blue spectra)

and after anti-solvent precipitation, (Red spectra precipitated with 96% ethanol same as the

coagulation bath used during fiber spinning and pink spectra precipitated LLC with 95.5%

isopropanol). FTIR spectra of LLCs in water showed a major amide I band at 1630 cm-1 (associated

with C=O vibration). However, aggregation of LLC in 96 % Ethanol (same as in the coagulation

bath for aggregating the extruded fibers) resulted in a shift of the major amide I band and higher

intensity peak (at 1642 cm- 1). Moreover, aggregated LLC showed prominent amide II bands at 1542

cm-1 ν (associated with N-

H vibration) which was absent from non-aggregated LLC. Additionally, LLC showed amide III band

at 1243 cm-1 ν (associated with C-N vibration), this was slightly shifted and appearing with higher

intensity for aggregated LLCs 1248 cm-1 ν. In order to confidently interpret observed peak associated with conformational transition and crystallization in FTIR, we repeated the experiment however

instead of Ethanol and Isopropanol we precipitated the protein using potassium phosphate (strong

kosmotropic salt) which is shown to induce conformation of transition and crystallization in

engineered spider silk protein and reconstituted silk proteins (light blue spectra).

25

Figure S10. Mean value and standard deviations of mechanical properties for only CNF and

composite fibers made from 1, 2 and 3 block protein variants. a) Max-stress, b) Young's modulus, c)

Maximum strain, d) Work of fracture, e) Yield point, f) slope after the yield point and g) Yield strain.

26

Figure S11. Effect of adhesive matrix protein variant and their effects on the morphology of the fibers studied

by polarized optical microscopy, WAXS, and SEM. a) Only CNF spun fiber. b) CBM-eADF3-CBM spun with

CNF. c) CBM-ADF3-CBM spun with CNF. d) CBM-eADF3 spun with CNF. e) CBM-CBM with CNF. f) CBM spun

with CNF. Alignment of CNFs can be qualitatively seen as intense colors product of variation in

thickness and the birefringent throughout the length of the fibers when images between crossed

polarizers. SEM images show the preferential alignment of the surface texture of the fibers and a

cross-sectional structure of the fibers showing nearly circular with a diameter of about 50 µm.

Altogether; we noticed similar morphological characteristic shared among all fibers spun from either

dispersed monomeric matrix protein, self-assembled LLC matrix protein or even only CNF fibers.

27

Figure S12. Calculated full-width half max (FWHM) and Hermans orientation parameter for all the

fiber. Quantitative measurement of the fibrillar alignment using synchrotron WAXS and calculated

Herman’s orientation parameters from the azimuthal intensity profile of the (004) reflection showed

high fibrillar alignment with more or less similar values ranging from 0.68-0.74.

28

Figure S13. SEM fractography for all the fibers after tensile measurement test.

Figure S14. Demonstration of potential application for the LLC-CNF fibers. a) Photograph from the

role of fiber, b) SEM image of yarns made by twisting two to seven individual fibers, c) photograph

of surface-hydrophobized (silanization via chemical vapor deposition) yarn holding a weight while

immersed in water and d) SEM images illustrating textile made from the fibers.

Movie S1. Infiltration of cellulose nanofibril network by an adhesive liquid-like coacervate matrix

protein (CBM-eADF3-CBM).

(1) Mohammadi, P.; Aranko, A. S.; Lemetti, L.; Cenev, Z.; Zhou, Q.; Virtanen, S.; Landowski, C. P.; Penttilä, M.; Fischer, W. J.; Wagermaier, W.; et al. Phase Transitions as Intermediate Steps in the Formation of Molecularly Engineered Protein Fibers. Commun. Biol. 2018, 1, 86.

(2) Tormo, J.; Lamed, R.; Chirino, A. J.; Morag, E.; Bayer, E. A.; Shoham, Y.; Steitz, T. A. Crystal Structure of a Bacterial Family-III Cellulose-Binding Domain: A General Mechanism for Attachment to Cellulose. EMBO J. 1996, 15, 5739.

(3) Huemmerich, D.; Helsen, C. W.; Quedzuweit, S.; Oschmann, J.; Rudolph, R.; Scheibel, T. Primary Structure Elements of Spider Dragline Silks and Their Contribution to Protein Solubility. Biochemistry 2004, 43, 13604–13612.

(4) Guerette, P. a; Ginzinger, D. G.; Weber, B. H.; Gosline, J. M. Silk Properties Determined by Gland-Specific Expression of a Spider Fibroin Gene Family. Science 1996, 272, 112–115.

(5) Gosline, J. M.; Guerette, P. a; Ortlepp, C. S.; Savage, K. N. The Mechanical Design of Spider Silks: From Fibroin Sequence to Mechanical Function. J. Exp. Biol. 1999, 202, 3295–3303.

(6) Engler, C.; Kandzia, R.; Marillonnet, S. A One Pot, One Step, Precision Cloning Method with High Throughput Capability. PLoS One 2008, 3, e3647.

29

(7) Sanjana, N. E.; Cong, L.; Zhou, Y.; Cunniff, M. M.; Feng, G.; Zhang, F. A Transcription Activator-like Effector Toolbox for Genome Engineering. Nat. Protoc. 2012, 7, 171–192.

(8) Yan, P.; Gao, X.; Shen, W.; Zhou, P.; Duan, J. Parallel Assembly for Multiple Site-Directed Mutagenesis of Plasmids. Anal. Biochem. 2012, 430, 65–67.

(9) Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W. NIH Image to ImageJ: 25 Years of Image Analysis. Nat. Methods 2012, 9, 671–675.

(10) Schindelin, J.; Arganda-Carreras, I.; Frise, E.; Kaynig, V.; Longair, M.; Pietzsch, T.; Preibisch, S.; Rueden, C.; Saalfeld, S.; Schmid, B. Fiji: An Open-Source Platform for Biological-Image Analysis. Nat. Methods 2012, 9, 676–682.

(11) Mohammadi, P.; Toivonen, M. S.; Ikkala, O.; Wagermaier, W.; Linder, M. B. Aligning Cellulose Nanofibril Dispersions for Tougher Fibers. Sci. Rep. 2017, 7.

30

download fileview on ChemRxivSupplementary _Mohammadi.pdf (2.75 MiB)

Biomimetic Composites with Enhanced Toughening Using Silk Inspired Triblock Proteins and Aligned

Nanocellulose Reinforcements

Pezhman Mohammadi*, A Sesilja Aranko, Christopher Paul Landowski, Wolfgang Wagermaier and Markus B Linder* P. Mohammadi, A.S. Aranko, M.B. Linder Department of Bioproducts and Biosystems, School of Chemical Engineering, Aalto University, P.O. Box 16100, FI-16100, Espoo, Finland. E-mail: pezhman. Mohammadi @aalto .fi and markus.linder@aalto.fi C.P Landowski VTT Technical Research Centre of Finland Ltd., Espoo, Finland. W. Wagermaier Department of Biomaterials, Max Planck Institute of Colloids and Interfaces, D 14424 Potsdam, Germany.

Abstract

Silk and cellulose are biopolymers that show a high potential as future sustainable materials.They also have

complementary properties, suitable for combination in composite materials where cellulose would form the

reinforcing component and silk the tough matrix. Therein, a major challenge concerns balancing structure

and properties in the assembly process. We used recombinant proteins with triblock architecture combining

structurally modified spider silk with terminal cellulose affinity modules. Flow-alignment of cellulose

nanofibrils and triblock protein allowed a continuous fiber production.The protein assembly involved phase

separation into concentrated coacervates, with subsequent conformational switching from disordered

structures to beta sheets. This gave the matrix a tough adhesiveness, forming a new composite material with

high strength and stiffness combined with increased toughness. We show that versatile design possibilities in

protein engineering enable new fully biological materials, and emphasize the key role of controlled assembly

at multiple length scales for realization..

Biological approaches to materials includes biomimetics, i.e. inspiration from biological structures to achieve functionality, as well as actually using biologically produced components in novel ways. One of the materials science problems that is approached by biomimetics is the question of how high stiffness and strength can be combined with a high toughness. Typically, the property combination of toughness and strength is challenging to achieve, but since finding solutions to this problem is evolutionarily advantageous, many biological materials have evolved to overcome these difficulties. Several strategies have been found, and on a general level these mechanisms can be grouped into two categories, intrinsic and extrinsic. The intrinsic mechanisms function ahead of the crack, by dissipating energy at the crack tip by plastic deformations or other molecular mechanisms. The extrinsic mechanisms function behind the crack tip, by providing for example bridges that span the forming crack, creation of sacrificial bonds, or by energy dissipation by sliding and pull-out of platelets[1–3].

Here we explore how to combine two highly promising biological materials, namely recombinant silk-protein and nanofibrillated cellulose, to solve the problem of making composites to achieve high toughness in combination with strength and stiffness. Cellulose shows many possibilities for new

1

materials, for example, using bacterial production a material with 836 MPa in strength and 65 GPa in stiffness was produced[4]. One widely used approach is to disintegrate plant material to obtain high aspect ratio fibrils of cellulose, so-called cellulose nanofibrils (CNF), which has been widely explored for use as building blocks in new materials[5]. Composites such as sheets and fibers can be made, but overall it still remains a challenge to increasing stiffness, strength and toughens together.

Silk forms a large family of proteins that are found in the familiar silkworm cocoons, as spider draglines, and also as a matrix material in biological composites[6]. Silk, and especially the spider spidroins have been prepared by bacterial using recombinant DNA techniques, with excellent results in for example reproducing its high toughness[7]. The promise of using recombinant DNA systems for producing silk-like materials is that proteins can be molecularly engineered in new combinations, allowing for example to tailor their structures to fit new types of applications and assembly processes[8]. Especially, it is of interest to engineer proteins to fit practical processing methods for continuous and scalable production.

However, even if components such as CNF and silk are becoming easier to produce, one main challenge is still how to control the assembly process in time and space to give the optimal composition and structure. For the assembly process we were inspired by the case of the Humboldt squid beak to build a composite material where a carbohydrate scaffold is impregnated by protein to create a stiff and tough material[9]. In particular, this model suggests that the assembly process for the proteins goes through an intermediate stage of condensed protein that is achieved through phase separation of the protein. In this approach phase separation results in a condensed state referred to as coacervation. Coacervated states of protein are also found in mussel threads and sandcastle worm cement. The condensed coacervated state results in relatively weak polymer interactions that lead to subsequently stronger interactions and new assembled structures[10].

2

Figure 1. a) Schematic representation of the block-architecture of proteins used in this study. b)

Phase contrast light microscopy and phase atomic force microscopy (AFM) images of the liquid-liquid

phase separated coacervates (scale bars are 20 µm for both images). c) Scanning electron

microscopy and height AFM images of cellulose nanofibril network (scale bars are 200 nm for SEM

micrograph and 1 µm both images). d) Infiltration of CNF network by coacervate at various time

point (scale bars are 1 µm). e) Light microscope image of deformation of coacervated during shear

stress (scale bars are 20 µm). f) The extruded fiber in a coagulation bath (scale bars are 5 mm).

We used a block-polymer approach to design proteins (Figure 1a), with an overall 3-block architecture. At each terminus, a cellulose anchoring block was added. For this, we used a cellulose binding module (CBM) from the Clostridium thermocellum cellulosome[11]. It is known from previous studies that the 156-amino acid CBM has a high affinity towards cellulose and that this binding is mediated by a set of aromatic resides on one face of the protein. It has a compact globular fold and jelly roll beta barrel conformation. As self-assembling and cohesive mid-blocks we used sequences from the major ampulla ADF3 dragline silk sequence spidroin from Araneus diadematus. Two versions were used, either a part of the wild-type spidroin sequence (called ADF3) or then an engineered version having a repeating consensus sequence (called eADF3). Both the ADF3 and eADF3 sequence blocks contain 12 repeats of stretches of hydrophobic A-residues with stretches of GPQ-rich motives in between[12]. Different combinations of the blocks were made. The CBM-ADF3- CBM and CBM-eADF3-CBM variants had CBMs on both sides of the mid-regions of spidroin protein. The 2-block variant CBM-eADF3 had only one CBM block and the 1-block variant eADF3 had no CBM

3

blocks. In the 2-block variant CBM-CBM the spidroin sequence block was missing. A 1-block single CBM was also used a reference.

Coacervates of the 3-block CBM-eADF3-CBM and CBM-ADF3-CBM formed when protein solutions were concentrated up to approximately 4% w/v in pure water (Figure 1b), as previously described for these proteins[13]. We noted that CNF fibrils (Figure 1c) were easily wetted by the coacervates. Coacervates penetrated into networks and wetted and embedded cellulose fibers readily, as seen in Movie S1 and the still frames in Figure 1d. The coacervate droplets showed very easy deformability under shear, with a high elongation of individual drops (Figure 1e) allowing coacervates easily to fill space between CNF fibrils. The other proteins did not form coacervates at 4% and hence did not show the above properties.

Composite fibers were assembled by extrusion of CNF through capillary tubing (150 cm in length and 500 µm ID) using a pump to align CNF fibrils, essentially as described previously[14]. The system was modified by introducing a mixing tee before the capillary tube so that the protein solution could efficiently be mixed with the CNF immediately prior to fibril alignment in the capillary tube (Figure S1). The fibers were directly extruded into a bath containing 96% ethanol to induce rapid coagulation. In Figure 1f, an example of a typical fiber in the coagulation bath is shown. All fibers, with protein and controls without protein had initially an indistinguishable appearance, except the ones containing the variant eADF3 without terminal CBMs. In this case the fibers fell easily apart and there was a clear release of protein solution that spread out into the surroundings of the fibers, forming loose whitish flocs as it coagulated. These were not studied further as this combination did not allow making fibers in a continuous manner. We attribute the low cohesiveness of these fibers to the eADF protein preventing favorable internal CNF interactions, as all proteins containing CBMs as well as the CNF alone resulted in cohesive fibers.

4

Figure 2. a) Representative stress-strain curves of CNF fibers with and without 3-block

spidroin proteins. b) Representative stress-strain curves of fibers spun at different protein to CNF

ratios. c) Representative stress-strain curves of composite fiber spun with different protein variants.

d) Spider- chart representing the performance space for the composite fibers spun from different

protein variants. e) Toughness values for all the composite fibers with different protein variants

(standard deviations shown for all). f) Representative stress-strain curves of oriented (fiber) and

non-oriented (film) composite and only CNF.

Functional properties of the protein composite structures.

Tensile testing was used to study the effect of engineered proteins on the CNF fibers. The 3-block spidroins gave the most noticeable result with a combined increase in stiffness, yield point, strength, and work of fracture. This occurred at the expense of only a relatively small reduction in ultimate strain (Figure 2a). With increasing ratios of protein to cellulose there was a clear increase in the stiffness of the fibers (Figure 2b). A high ratio of 2:1 CBM-ADF3-CBM showed a remarkably high stiffness of up to 35 GPa, while a 1:3 ratio gave a stiffness of about 20 GPa. This shows that the ratio

5

of protein to cellulose is a way to regulate stiffness over a broad range. However, these stiffest fibers showed highly reduced toughness. For maximal toughness in combination with strength a 1:2 ratio gave the best results. This combination is remarkable since cross-linking by itself did not previously lead to increased ultimate strength [15,16], as also noted in Figure 2 c, d and e, with the addition of CBM-CBM. The CBM-CBM shows a strong cross linking, but with a simultaneous large decrease in the work of fracture (Figure2c), demonstrating the expected mutual exclusion of toughness and strength. The isolated CBM and the CBM-eADF3 showed some increased cross-linking but to a significantly lower extent than the crosslinking CBM-CBM. CBM-ADF3-CBM and CBM-eADF3-CBM were the only proteins able to improve multiple properties simultaneously (Figure 2d).

WAXS showed that the capillary extruded fibers had a high degree of orientation in direction of fiber axis, showing four-quadrant symmetry with meridian and equatorial arcs for the (004) and (200) diffractions. To gain more insight in the role of structural arrangements in the composite we made 2D films with coacervate and CNF using the same components as for the extruded fibers (Figure 2f). For the films, we note clear ring patterns for (004) and (200) as the main diffraction, showing a distributed orientation of fibrils in the plane of the film. The stress-strain curves show that the coacervate has a similar effect on both films and fibers, but the magnitude of the effect is greater for the highly aligned fibers.

Biological materials such as silk and cellulose are typically highly responsive to humidity. We therefore measured the 1:2 CBM-ADF3-CBM and CNF-only fibers at 20 % and 80 % RH (Figure S7). Increasing the water content (at 80 % RH) had plasticizing effect, whereas decreasing the water content (at 20 % RT) had stiffening effect in comparison to fibers measured at 50 % RH. The protein showed a significant effect, making the fibers significantly less responsive to humidity than the CNF- only fibers. The high effect of humidity of CNF-only fibers is predicted as hydrogen bonds on the surface of fibers are easily replaced by water molecules, reducing the inter fiber hydrogen bonding. While hydrogen bonding also plays a significant role in protein-cross linking of CNF fibers, these are less efficiently competed out by water molecules. This is an expected property for proteins since their macromolecular structures give rise to cooperative weak bonding, in internal secondary structures and the large interaction interfaces of for example CBMs, that can efficiently compete with water for hydrogen bonding.

6

Figure 3. FTIR spectra of 3-block spidroin CBM-eADF-CBM protein in the coacervated state and

after coagulation with ethanol. The amide bands show a characteristic switch from -helix towards

high -sheet content by the treatment.

A significant aspect of infiltrating the nanocellulose network with coacervate is that it undergoes a conformational transition as a result of the ethanol treatment, which acts as a curing step. Fourier transform infrared spectroscopy (FTIR) of the coacervate before and after ethanol treatment in range of 1000 cm-1 to 1700 cm-1 wavenumber (conformation- sensitive amide-I, II and III bands), reveled conformational transitions from high α-helix/random coil and low β-sheet content towards more β-sheet and less α-helix/random coil conformation (Figure 3a). The observation that proteins with Ala-rich sequences with initial α-helix rich conformations undergo conformational transitions to lower-energy β-sheet structures over time has been generally observed in several different systems[17,18]. This conformational switch provides an explanation for a curing mechanism after the initial infiltration by the highly fluid and easily deformable coacervates to a cured end state which provides stronger cohesion and lower molecular motility.

7

Figure 4. SEM fractography of the fibers. a) Fracture cross-section of the all the fibers after tensile

measurement test. b) High magnification images of fracture surfaces of the CNF control, CBM- CBM CNF and CBM-eADF3-CBM -CNF composite fibers. At this magnification, the fibril tips of the CBM- eADF3-CBM show distinguishable rounded tips and show tight bundles with blunt ends.

For obtaining insight into the effects of infiltrated coacervate we used high-magnification SEM to image fracture surfaces. Fibers consisting of CNF alone showed surfaces with uneven creep with pull-out of fibrils and large voids. These features are expected for cellulose-based materials while undergoing catastrophic failure and is due to fibrils moving and sliding past each other, forming micron-scale voids that gradually propagate throughout the system (Figure 4a). The coacervate infiltrated fibers showed formation of voids to a clearly lesser extent. Pull-out lengths were also significantly shorter. The 1-and 2 block-proteins with CBM, double CBM, and CBM-eADF3 showed fracture surfaces similar to the CNF-only samples. Although the proteins gave a stiffer material it seems that fracture mechanisms were still essentially unchanged, with large pull-outs and the formation of voids. High magnification of fibrils in the fracture surfaces (Figure 4b) show that the fractured tips of individual fibers are rounded, often having a curled structure. These rounded tips are found in all samples, but to a much higher extent in the 3-block spidroin-infiltrated ones.

8

Figure 5. The reduced fiber pullout seen in the fracture surfaces could be explained by cohesiveness

that the silk material provides and allows efficient mechanisms for hindering crack propagation.

On understanding the mechanisms of the observed toughening we note, on the one hand, that the highly aligned fiber system gave much more significant effects than when we used a flat isotropic sheet, i.e. the applied toughening mechanisms provide a better function in highly aligned systems. On the other hand, modelling has suggested that a significant mechanism in cellulose nanomaterial toughening is an extrinsic one due to the bridging effect by fibers behind the crack tip[19]. This modelling result suggests that bridging is most efficient with alignment in the stress direction. It seems therefore that our 3-block spidroin proteins were able to enhance this effect, possibly by locking fibers efficiently in place, which modelling also suggests should increase toughness [20]. This prediction fits with the observation that fibers that were toughened by the 3-block spidroin proteins showed relatively blunt fracture surfaces of fibril bundles, with much less tear-out than the other samples (Figure 4). High resolution images also showed small bundles of fibrils with blunt ends. In these the tips of the individual CNF fibrils showed a rounded shape, not typically observed in other samples. The observation of individual tips suggest that they had been broken at the fracture surface as would be expected by an extrinsic fiber bridging mechanism. We summarize in Figure 5 how the 3-block spidroin infiltration could enhance the cohesiveness of the structures leaving less possibilities for tear-outs at weakly connected regions that result in cavities during fracture. In this model, the cross-linking by CBMs without the mid spidroin was too short range for locking of CNF fibrils and allow this mechanism to function although stiffness did increase. Simple cross-linking did not give increased toughness in previous studies even when using much longer 30kDa resilin protein in between CBMs [16].

On considering why the spidroin protein is efficient as a matrix cross-linker we note that spidroin

infiltration was possible when the protein was in the -helix form, and that the spidroin switched

9

conformation to -sheet form during ethanol coagulation (Figure 3). This switch in conformation has been predicted by modelling to significantly reduce molecular mobility, because the -sheet structure allows more efficient chain crosslinking by hydrogen bonding than is possible for hydrophobic interactions between Ala-containing -helices [21]. This prediction is supported by or findings that the high- forms shows a much slower molecular diffusivity than the -form. The 3- block spidroin in the coagulated form would thus form a connected network having a strong cellulose interface due to the terminal CBMs. It is unclear how to compare the 3-block spidroins to the 2-block CBM-eADF3, since this protein did not show the characteristic transition shown in figure 3. We also note that overall the material consists of 50% protein and therefore the bulk of the soft matrix can retard crack propagation by an intrinsic mechanism involving plastic deformations at the crack tip, thereby blunting the tip and dissipating energy.

As a main conclusion, this work brings out the essential role of the multi-level molecular to macro level control of assembly for functional materials properties to emerge. We combine molecular to micro-scale protein assembly through coacervation and molecular recognition together with nano to unlimited macro-scale assembly of nanocellulose by flow-alignment. Together this multi-level assembly led to property combination that are typically mutually exclusive, but achievable by biomimetics. The works shows how protein engineering opens a new route for molecular materials design that is a highly promising route for future sustainable materials.

[1] R. O. Ritchie, Nat. Mater. 2011, 10, 817–822. [2] U. G. K. Wegst, H. Bai, E. Saiz, A. P. Tomsia, R. O. Ritchie, Nat. Mater. 2014, 14, 23–36. [3] M. E. Launey, R. O. Ritchie, Adv. Mater. 2009, 21, 2103–2110. [4] S. Wang, F. Jiang, X. Xu, Y. Kuang, K. Fu, E. Hitz, L. Hu, Adv. Mater. 2017, 1702498, 1702498. [5] E. Kontturi, P. Laaksonen, M. B. Linder, A. H. Nonappa, A. H. Gröschel, O. J. Rojas, O. Ikkala,

Adv. Mater. 2018, 1703779, DOI 10.1002/adma.201703779. [6] B. E. Pennisi, Science 2017, 358, 288–291. [7] A. Heidebrecht, L. Eisoldt, J. Diehl, A. Schmidt, M. Geffers, G. Lang, T. Scheibel, Adv. Mater.

2015, 27, 2189–2194. [8] A. Rising, J. Johansson, Nat. Chem. Biol. 2015, 11, 309–15. [9] Y. Tan, S. Hoon, P. A. Guerette, W. Wei, A. Ghadban, C. Hao, A. Miserez, J. H. Waite, Nat.

Chem. Biol. 2015, 11, 488–495. [10] Y. Shin, C. P. Brangwynne, Science 2017, 357, eaaf4382. [11] J. Tormo, R. Lamed, A. J. Chirino, E. Morag, E. A. Bayer, Y. Shoham, T. A. Steitz, EMBO J.

1996, 15, 5739. [12] D. Huemmerich, C. W. Helsen, S. Quedzuweit, J. Oschmann, J. Oschmann, R. Rudolph, T.

Scheibel, Biochemistry 2004, 43, 13604–13612. [13] P. Mohammadi, A. S. Aranko, L. Lemetti, Z. Cenev, Q. Zhou, S. Virtanen, C. P. Landowski, M.

Penttilä, W. J. Fischer, W. Wagermaier, et al., Commun. Biol. 2018, 1, 86. [14] P. Mohammadi, M. S. Toivonen, O. Ikkala, W. Wagermaier, M. B. Linder, Sci. Rep. 2017, 7,

11860. [15] J. M. Malho, S. Arola, P. Laaksonen, G. R. Szilvay, O. Ikkala, M. B. Linder, Angew. Chemie -

Int. Ed. 2015, 54, 12025–12028.

10

[16] W. Fang, A. Paananen, M. Vitikainen, S. Koskela, A. Westerholm-Parvinen, J. J. Joensuu, C. P. Landowski, M. Penttilä, M. B. Linder, P. Laaksonen, Biomacromolecules 2017, acs.biomac.7b00294.

[17] Z. Lin, W. Huang, J. Zhang, J. Fan, D. Yang, Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 8906–

8911. [18] T. Giesa, C. C. Perry, M. J. Buehler, Biomacromolecules 2016, 17, 427–436. [19] Q. Meng, B. Li, T. Li, X. Feng, Eng. Fract. Mech. 2018, 194, 350–361. [20] Q. Meng, B. Li, T. Li, X. Feng, J. Mech. Phys. Solids 2017, 103, 22–39. [21] I. Su, M. J. Buehler, Nanotechnology 2016, 27, 302001.

11

download fileview on ChemRxivManuscript_Mohammadi.pdf (1.11 MiB)