Spider Silk Nanostructuring and its Applications for Tissue Engineering

(1)

Doctoral Thesis in Electrical Engineering and Computer Science

Spider Silk Nanostructuring and its Applications for Tissue Engineering

LINNEA GUSTAFSSON

Stockholm, Sweden 2021

(2)

Spider Silk Nanostructuring and its Applications for Tissue Engineering

LINNEA GUSTAFSSON

Doctoral Thesis in Electrical Engineering and Computer Science KTH Royal Institute of Technology

Stockholm, Sweden 2021

Academic Dissertation which, with due permission of the KTH Royal Institute of Technology, is submitted for public defence for the Degree of Doctor of Philosophy on Friday the 26th March 2021, at 1:00 p.m. in Q2, Malvinas väg 10, Stockholm.

(3)

TRITA-EECS-AVL-2021:15 ISBN 978-91-7873-790-1

Printed by: Universitetsservice US-AB, Sweden 2021

(4)

iii

Abstract

This thesis introduces new ways to produce micro-and nanostructures of recombinant spider silk and explores ways to characterize their topography, mechanical properties, cell compatibility, and permeability. The suitability of the formed structures for applications within tissue engineering, primarily in vitro tissue modeling, is also investigated.

One big challenge in drug development is that many drug candidates fail to pass in vivo studies in humans. This is largely because the currently used animal models fail to emulate the full human condition. Therefore, researchers aim to develop in vitro models of various tissues using human cells. These new systems will allow studies of biological responses and mechanisms related to human health and disease. To accurately represent what happens in the body, the materials used for cell culture should as closely as possible mimic their in vivo counterparts. Many of the materials used today are made out of plastic and lack physiologically relevant properties, and do not replicate the micro-and nano dimensions present in the native cell environment.

Spider silk has been suggested as a suitable replacement material for cell culture. The usage of spider silk for medical purposes is not new; it was used already in ancient Greece and Rome to staunch wounds. However, the spider’s limited production has haltered the applicability. Lately, new doors have opened up through recombinant production of the base constituent of silk: the spider silk protein (spidroin). Recombinant spidroin production is not only scalable but also allows for facile integration of additional biofunctionality.

With this building material at hand, it is possible to produce other formats than spider silk fibers, i.e., coatings, films, membranes, hydrogels, porous scaffolds, and microparticles.

With the work presented in this thesis, the list is extended through the in- troduction of new methods to produce nanomembranes and uniformly shaped micro-and nanostructures by manipulating the liquid:air interface. Micropat- terned mm-sized films, microfilms, nanochains, and nanowires were produced by manipulating a droplet of soluble spidroin solution on a superhydrophobic surface. Alterations in the concentration of spidroins, the motion of the droplet, and the dimensions of the pillars allow for precise control of the silk formation. The formed silk structures retained their shape upon release from the surface, and the culture of mammalian cells showed good compatibility with the silk structures. Nanofibrillar spider silk membranes mimicking the dimensions of basal membranes (280 nm thick) were formed by letting spidroins self-assemble at the liquid:air interface of a standing solution. The assembly time, initial spidroin concentration, and beaker size are directly related to the membrane’s thickness and size. The thereby obtained membranes were stable, had an internal nanofibrillar structure, could stretch over 200%, and were permeable to human plasma proteins. An in vitro blood vessel model was established by growing human endothelial cells and smooth muscle cells on opposing sides of the membrane, showing the potential of using the mem- branes for further in vitro modeling.

(5)

Keywords: recombinant spider silk, nanostructures, microstruc- tures, nanowires, nanochains, nanodisks, nanomembranes, tissue engineering, in-vitro models, medical technology, health techno- logy, nanomedicine

Linnea Gustafsson, lingusta@kth.se Division of Micro- and Nanosystems

School of Electrical Engineering and Computer Science,

KTH Royal Institute of Technology, SE 100 44 Stockholm, Sweden.

(6)

v

Sammanfattning

Den här avhandling introducerar nya sätt att producera mikro- och nanostrukturer av rekombinant spindelsilke och utforskar sätt att karakterisera deras topografi, mekaniska egenskaper, cellkompatibilitet och permeabilitet.

Lämpligheten hos de formade strukturerna för applikationer inom vävnads- teknik, främst för in vitro vävnadsmodellering, undersöks också.

En stor utmaning i läkemedelsutveckling är att många kandidater inte uppvisar önskad effekt i in vivo studier i människor. Detta beror till stor del på att de djurmodeller som används i den primära utvärderingen inte efter- liknar den mänskliga kroppen tillräckligt bra. På grund av detta har forskare börjat utveckla metoder för att använda mänskliga celler i in vitro modeller av olika vävnader. Dessa nya system öppnar upp för möjligheten att studera biologiska reaktioner och mekanismer relaterade till människors hälsa. För att korrekt kunna modellera vad som händer i kroppen bör materialen som används för cellodling så nära som möjligt efterlikna deras motsvarigheter in vivo. Många av de material som används idag är gjorda av plast, saknar fysiologiskt relevanta egenskaper och replikerar inte de mikro- och nanodi- mensioner som finns i cellmiljön i kroppen.

Spindelsilke har föreslagits som ett lämpligt material för cellodling. An- vändningen av spindelsilke för medicinska ändamål är inte ny, utan det använ- des redan i det antika Grekland och Rom för att stoppa blödningar. Använd- barheten begränsas dock av att spindlar enbart producerar en liten mängd silke. På senare tid har nya dörrar öppnats genom rekombinant produktion av baskomponenten i silket: spindelsilksproteiner (spidroiner). Rekombinant produktion as spidroiner är inte bara skalbar utan möjliggör också enkel integration av biofunktionalitet. Med byggmaterialet till hands är det även möjligt att producera fler format än enbart spindelsilkesfibrer, dvs. beläggningar, filmer, membran, hydrogeler, porösa strukturer och mikropartiklar.

Arbetet som presenteras i den här avhandlingen fyller på listan genom att introducera nya metoder för att producera nanomembran och enhetligt formade mikro- och nanostrukturer genom att manipulera vätske:luftgränssnittet.

Mikromönstrade mm-filmer, mikrofilmer, nanokedjor och nanotrådar produ- cerades genom att manipulera en droppe spidroinlösning på en superhydrofob yta. Förändringar i spidroinernas koncentrationen, droppens rörelse och dimensionerna på pelarna möjliggör exakt kontroll av silkeformationen. De formade silkestrukturerna behöll sin form efter frisättning från ytan, och odling- en av mänskliga celler visade god kompatibilitet med silkesstrukturerna. 280 nm tjocka nanofibrillära spindelsilkesmembran, som imiterar dimensionerna hos basala membran, bildades genom att låta spidroiner självinteragera vid vätske:luftgränssnittet i en stillastående lösning. Tid, initial spidroinkoncent- ration och bägardimensioner är direkt relaterade till membranets tjocklek och storlek. Nanomembranen formade via denna metod var stabila, kunde sträcks över 200% och var permeabla för mänskliga plasmaproteiner. En in vitro- blodkärlsmodell upprättades genom att växa humana endotelceller och glatta muskelceller på motsatta sidor av membranet, vilket påvisar potentialen att använda membranen för vidare in vitro modellering.

(7)

Nyckelord: rekombinant spindelsilke, nanostrukturer, mikrostruk- turer, nanotrådar, nanokedjor, nanodiskar, nanomembran, vävnads- teknik, in vitro-modeller, medicinsk teknik, hälsoteknik, nanome- dicin

(8)

L IST OF P UBLICATIONS

This thesis is based on the following papers in peer-reviewed, interna- tional journals and conferences:

I "Structuring of Functional Spider Silk Wires, Coatings, and Sheets by Self- Assembly on Superhydrophobic Pillar Surfaces"

Gustafsson, L., Jansson, R., Hedhammar, M. & van der Wijngaart, W.

Advanced Materials, 30, 1704325, Dec 2017.

II "Formation of a Thin-walled Spider Silk Tube on a Micromachined Scaffold"

Guo, W.^∗, Gustafsson, L.^∗, Jansson, R., Hedhammar, M., & van der Wijn- gaart, W.

2018 IEEE 31st International Conference on Micro Electro Mechanical Sys- tems (MEMS), Institute of Electrical and Electronics Engineers (IEEE), 2018, Vol. 2018, p. 83-85.

∗Shared first authorship & Presenting authors

III "Recombinant Spider Silk Forms Tough and Elastic Nanomembranes that are Protein-Permeable and Support Cell Attachment and Growth"

Gustafsson, L., Tasiopoulos, C. P., Jansson, R., Kvick, M., Duursma, T., Gasser, T. C., van der Wijngaart, W., & Hedhammar, M.

Advanced Functional Materials, 2002982, Aug 2020.

IV "Fibrillar nanomembranes of recombinant spider silk protein support cell co- culture in an in vitro blood vessel wall model"

Tasiopoulos, C. P., Gustafsson, L., van der Wijngaart, W., & Hedhammar, M.

Manuscript submitted.

V "Scalable Synthesis of Monodisperse Bioactive Spider Silk Nanowires to Im- prove Cell Culture Conditions"

Gustafsson, L., Dorka, N., Åstrand, C., Ponsteen, N., Svanberg, S., Hegrova, V., Jansson, R., Horak, J., Hedhammar, M. & van der Wijngaart, W.

Manuscript.

xi

(13)

The contributions of Linnea Gustafsson to each publication listed above, major (• • •), partial (• •), or minor (•):

Design Experiments Analysis Writing

I • • • • • • • • • • • •

II • • • • • • • • • •

III • • • • • • • • • • • •

IV • • • • • • • • • • •

V • • • • • • • • • • •

The work has also been presented in the following international peer- reviewed conferences (not included in paper reprints):

"Controlled Formation of Spider Silk Nanowires, Localized Surface Coatings, and Sheets Using Superhydrophobic Surfaces"

Gustafsson, L., Jansson, R., Hedhammar, M. & van der Wijngaart, W.

The 21st International Conference on Miniaturized Systems for Chemistry and Life Sciences, MicroTAS 2017, 22-26 October 2017, Savannah, GA, USA.

"Mechanical Characterization of Spider Silk Nanomembranes"

Gustafsson, L., Tasiopoulos, C. P., Duursma, T., Jansson, R., Gasser, T. C., Hedhammar, M. & van der Wijngaart, W.

The 24th International Conference on Miniaturized Systems for Chemistry and Life Sciences, MicroTAS 2020, 4-9 October 2020, online.

"Spider silk nanomembranes support cell co-cultures"

Tasiopoulos, C. P.^∗, Gustafsson, L., Jansson, R., van der Wijngaart, W. & Hed- hammar, M.

11TH World Biomaterials Congress, WBC 2020, 11-15 December 2020, online.

∗Presenting author

Other peer- reviewed international conference papers not included in this thesis:

"Bend-and-bond Polymer Microfluidic Origami"

Guo, W.^∗, Hansson, J. Gustafsson, L., & van der Wijngaart, W.

2021 IEEE 34th International Conference on Micro Electro Mechanical Systems (MEMS), Institute of Electrical and Electronics Engineers (IEEE), 2021, Vol. 2021, p. 222-225.

∗Presenting author

(14)

A BBREVIATIONS

AFM Atomic Force Microscope

AMP Antimicrobial peptide B. Mori Bombyx Mori

BAM Brewster Angle Microscopy

BSA Bovine Serum Albumin

COL-I Type Collagen I

E. australis Euprosthenops australis E. coli Escherichia coli

EnC Endothelial Cell

ECM Extracellular Matrix

FITC Fluorescein isothiocyanate

IgG Immunglobulin G

MSC Mesenchymal Stem Cells

LBIA Liquid Bridge Induced Assemlby MaSp Major Ampullate Spidroin) PDMS Polydimethylsiloxane

PM-IRRAS Polarization Modulation Infrared Reflection Absorption Spectroscopy

PRINT Particle Replication In Nonwetting Templates

PVA Polyvinyl-Alcohol

SEM Scanning Electron Microscope SHS Superhydrophobic Surface

SMC Smooth Muscle Cell

Spidroin Spider Silk Protein

TEER Transendothelial Electrical Resistance TEM Transmission Electron Microscopy

xiii

(15)

(16)

A IMS AND STRUCTURE OF THIS THESIS

P REAMBLE

It is a material that has been used since ancient times as fishing nets and wound dressings. It is tougher than Kevlar, stronger than steel, lighter than carbon fibers, and extends to extraordinary lengths before breaking. It is predicted to be used in slim bulletproof vests, durable winter jackets, car seats, to repair nerve damage, replace tissue, and, clearly, to make unbreakable socks. The sky is the limit for this extraordinary material that glimmers with dew in the morning, gets stuck in our hair when walking through the woods, and hides in the corners of our homes. That material is spider silk.

Most of us probably have some spider silk in our home and would be happy to collect and donate it for research. Nonetheless, hearing that the same silk has been used for medical purposes is less appealing. Having spiders physically in the production environment making long threads of silk is not suitable either. Beyond the ethical aspects, spiders in captivity do not produce silk of the same quality as spiders in nature. This, in combination with the spider’s territorial and sometimes cannibalistic nature, makes it impractical to farm them. Instead, ways to produce the building block of spider silk, spider silk proteins i.e. spidroins, have been developed. The spidroins can, in turn, be spun into silk threads, which can be used for the applications introduced above. However, having the base material at hand also opens up for developing other spider silk formats such as foams, gels, coatings, or, as within the scope of this thesis, spider silk micro-and nanostructures for tissue engineering applications.

xv

(17)

A IM AND OBJECTIVES

This thesis aimed to explore new ways to turn soluble spider silk proteins (spidroins) into silk structures, under mild conditions by utilizing the liquid:air interface, as well as determine ways to handle and characterize the formed structures, with the intent to reveal their suitability for applications in tissue engineering.

Specific objectives of this thesis are to:

• Develop ways to manipulate the liquid:air interface to control the formation of new conformations of spider silk structures (Paper I and II)

• Reveal governing factors influencing the formation of spider silk structures at the liquid:air interface (Paper III)

• Develop ways to handle spider silk structures formed at the liquid:air interface (Paper I, II, III, IV, and V)

• Assess retained biofunctionality of the produced spider silk structures (Paper I and II)

• Characterize relevant material properties of the produced spider silk struc- tures (Paper III, IV, and V)

• Evaluate the cell compatibility of the produced spider silk structures (Paper III, IV, and V)

• Evaluate the suitability of the produced spider silk structures for applications within tissue engineering (Paper III, IV, and V)

The design and manufacturing of spidroins are outside this thesis’s scope, which focuses on using soluble spidroins to engineer new silk structures and evaluate their properties and applicability.

T HESIS STRUCTURE

Chapter 1: Introduces why spider silk is an interesting material and how the con- stituent, the spidroin, can be obtained and turned into fibers both naturally and artificially.

Chapter 2: Focuses on the formation and handling of spider silk micro-and nanos- tructures. Some conventional formation approaches are introduced before the self- assembly of spider silk at interfaces is presented together with various ways to manipulate the interfaces. The chapter finishes by describing ways to handle micro-and nanostructures post-formation, both on substrates and in solution.

Chapter 3: Focuses on the material properties of silk micro-and nanostructures.

The source of natural spider silk toughness is introduced, followed by various ways to characterize the mechanical properties of spider silk micro-and nanostructures.

(18)

THESIS STRUCTURE xvii

The chapter ends by presenting ways to evaluate cell compatibility, biofunctionality, and permeability.

Chapter 4: Focuses on biomedical applications where spider silk micro-and nanos- tructures can come of use. It begins by introducing tissue engineering and the interest for micro-and nanostructures before presenting three potential applications for spider silk micro-and nanostructures. The chapter ends by discussing three other biomedical applications.

Chapter 5: Summarizes the key outcomes of this thesis before doing a critical analysis of the methods developed to produce spider silk micro-and nanostructures in this thesis. The chapter ends with a short conclusion.

(19)

(20)

C ^HAPTER 1

I ^T ’ S ALL ABOUT SPIDER SILK

"One of nature’s wondrous chemical structures is being dissected so that it can be used in human inventions" - Unraveling the Weave of Spider Silk, Ricki Lewis This chapter introduces the current natural and state-of-the-art techniques used to produce spider silk proteins (spidroins) and fibers. It is, in short, all about unraveling the weave that is spider silk. After providing a brief history of spider silk usage through the ages, it introduces the base material of all spider silks, the spidroin.

Focus is put on the spidroin structure and routes to obtain it, both naturally and synthetically. The chapter ends by describing how spiders and humans convert spidroins into silk fibers.

1.1 S PIDER SILK - A MULTIFUNCTIONAL MATERIAL

Spiders and their webs have always intrigued humans, as is apparent through their presence in both ancient Greek mythology [1] and present-day blockbuster movies.

What makes spider silk so fascinating is its unique combination of material properties. It is not only strong and tough but also soft and extendable. On top of that, it is also a purely protein-based material, making it both biocompatible and biodegradable [2, 3]. As a result, spider silk has not only become part of human culture, but it has also found its way into several day-to-day applications.

In ancient times, spider webs were collected by the natives in New Guinea for much the same purpose as spiders themselves use them, to capture food [4]. In ancient Greece and Rome webs were used to staunch bleeding wounds [4, 5]. This practice was still in place in the 1300s when French troops carried cobwebs with them in battle [5]. Purely collecting the webs have quite limited use, leading to that attempts were made to harvest silk fibers directly from spiders [4]. This practice

1

(21)

was already in place for the domesticated silkwormBombyx mori (B. mori) which had been used to produce textiles for thousands of years [6]. The first references concerning using spider silk to make stockings and gloves date back to the 1700s, where it quickly was realized that too many spiders were needed for this application [4]. Soon after, it also became evident that raising spiders was far more complicated than farming silkworms as their cannibalistic nature prevented them from being kept in close quarters. The process was further complicated as spider had to be immobilized during silk extraction. In the mid-1800s, devices had been developed purely for this purpose. However, scientists estimated that about 50 000 spiders would be needed to make enough silk for one dress [5]. Textile artist Simon Peers and design-entrepreneur Nicholas Godley put this to the test in the mid-2000s, where they spent eight years and 1.2 million golden orb spiders to make one naturally golden cape out of pure spider silk [7].

As is apparent, making clothing out of spider silk collected from spiders is not scalable. Nevertheless, the material remained in use for medical purposes in Colonial America [5] and as crosshairs in optical devices such as microscopes, telescopes, and guns during World War II. In fact, up until the mid-1990s some military facilities still kept spiders to repair crosshairs in old instruments [4]. In the last decade, natural spider silk has been braided into sutures for flexor tendon repair [8] and implemented to repair damage of long nerves [9]. Still, the limited production and variations in silk quality due to differences between individual spiders, their diets, and harvesting conditions remain an issue [3, 6, 10]. Consequently, focus has turned to producing spider silk synthetically, where the minimum requirement is access to the base material of all spider silks [1], the fibrous proteins containing highly repetitive sequences of amino acids, the spidroins [11].

1.2 O BTAINING SPIDROINS

There are three ways to obtain any protein: extraction from the living organism, chemical synthetization, or production using recombinant DNA technology [4].

When it comes to spiders, protein extraction is not only expensive and time- consuming [12], but it also suffers from the same drawbacks as the fiber harvesting introduced in Section 1.1 [3, 6, 10]. Chemical synthesis and genetic engineering of spidroins were explored in the mid-90s. At the time, not enough was known about the silk’s constituents to make full chemical synthesis successful, and the approach was abandoned in favor of genetic engineering [4]. While recombinant production comes with its own set of challenges, as detailed below, the method is attractive as it not only opens up for high quantities and quality [2] but also programmed sequences, secondary structures, architectures, and precise molecular weight [13].

(22)

1.2. OBTAINING SPIDROINS 3

1.2.1 N

ATURAL SPIDROINS

All spidroins have the same basic three-part structure: a non-repetitive globular N-terminal, an extensive repetitive central part, and a non-repetitive globular C-terminal (Figure 1.1a) [3]. The N- and C-terminals are conserved between all spidroins and are thought to play key roles in fiber formation and spidroin solubility [14]. The long, central repetitive domain consists of hundreds of repeats of the nonessential amino acids glycine and alanine, varies greatly, and determines the mechanical properties of the fiber [3, 15]. By altering the length and sequence of the repetitive part, spiders can produce up to seven different types of silk that have distinctly different properties and applications (Figure 1.1d) [2, 3, 15]. In the final fiber, alanine-rich segments form crystalline β-sheets while glycine-rich segments have no ordered structure. Thus having a high alanine content makes the final fibers stronger, but less extendable [7, 14].

The strongest and most well-studied silk is the major ampullate silk, also known as dragline silk. It consists of at least two different spidroins called MaSp1 and MaSp2 (Ma = major ampullate, Sp = spidroin) [2, 3] that are around 200-720 kDa [14].

These two spidroins have served as the main source of inspiration for recombinant production.

1.2.2 R

ECOMBINANT PRODUCTION OF SPIDROINS

The fundamental process for recombinant production of proteins has become rela- tively standardized over the last 20 years and begins with determining the natural DNA sequence of the protein that is to be produced [2, 11]. This sequence is used as inspiration when designing the recombinant DNA sequence. At this point, it is possible to alter the sequence to one that can be produced more efficiently or incorporate new functionalities [11]. After designing and optimizing the DNA sequence, a cloning vector into which the new DNA can be inserted is selected. The cloning vector is inserted into the host organism, which produces the protein. Last, the protein is extracted and purified before use [2, 11]. Using this process to produce spidroins has encountered several challenges, primarily associated with the long repetitive central gene sequence, which complicates sequencing, cloning, and expression [16]. There are several ways to overcome this, for example, through host selection, codon optimization, and DNA sequence optimization [11, 16].

Several different hosts have been evaluated for spidroin production, for example, bacteria, yeasts, mammalian cell lines, and transgenic animals such as silkworms and goats [7,11,17]. Escherichia coli (E. coli) quickly became the most widely used as it is well-controlled, cost-efficient, and suitable for large scale production [3].

However, two main issues causing low production rates made early attempts highly inefficient [7]. First, more than one codon can encode the amino acids that make up the spidroin’s central region, and the bacteria could not recognize the codons in the original spidroin genetic sequence [4]. This issue was resolved by optimizing the genetic information to the bacteria’s codon [2]. Second, bacteria have an inherent

(23)

Figure 1.1. Representative illustration of a) a natural spidroin, composed of a nonrepetitive N-terminal (green), a central repetitive region, and a nonrepetitive C-terminal (red), b) the recombinantly produced spidroin 4RepCT and c) the RGD-motif functionalized 4RepCT, FN-4RepCT. d) Illustration of a spider with its different spinning glands, different silks, and applications. d) is adapted with permission from [3], Nature.

tendency to remove repetitive sequences, and the long repetitive part of the original spidroin DNA would cause DNA deletion, as well as transcription and translation errors during production [2, 7]. This problem implies that a shorter and less repetitive region would be better suited for recombinant spidroin production [3]. As a result, most current work focuses on producing engineered variants with varying, but short, repeats of the central core structure and both with and without the terminal domains [2]. Many of the current recombinantly produced spidroins were recently reviewed here [11].

After optimizing the production of spidroins, another issue appeared: keeping the spidroins in soluble form as they tend to self-assemble into insoluble complexes.

The loss of protein can be mitigated by incorporating a solubilization step prior to use [14]. Solubilization generally involves adding various chemicals such as urea, lithium bromide, or formic acid. However, to ensure that the spidroins are suitable for biological applications, chemical residues must be removed [16]. The process

(24)

1.3. TURNING SPIDROINS INTO SILK FIBERS 5

also destroys the native spidroin structure, and consequently their natural self- assembly capability [18]. Strategies to overcome this involve introducing methion- ine residues for controlled oxidation/reduction to prevent (or promote) assembly, kinase recognition motifs for controlled assembly via enzymatic phosphorylation, or solubility-enhancing fusion partners that can be released enzymatically [16]. The last of the three approaches have successfully been used to produce the spidroin used in this thesis, 4RepCT [14].

1.2.3 4R

^EP

CT

AND VARIANTS THEREOF

4RepCT is a miniaturized spidroin, about 10% of the length of the original sequence taken from the MaSp1 protein fromEuprosthenops australis (E. australis).

As indicated by the naming convention, 4RepCT consists of four repetitions of al- ternated polyalanine and glycine regions (4Rep) flanked by a globular C-terminal (CT) domain (Figure 1.1b) [14]. Beyond being shorter than the original sequence, 4RepCT also lacks an N-terminal. It was excluded as the N-terminal fromE. aus- tralis was identified as a pH-dependent regulator and not a crucial component for silk-assembly nor the properties of the final silk product. Both the CT- and 4Rep- domain are key for stable fiber formation. The CT-domain can form fibers on its own, but these are not chemically stable and can easily be disassembled using, for example, urea [19]. The 4Rep-domain can only form shorter fibers on its own [18].

Together the two domains make up the first recombinantly produced spidroin that spontaneously forms chemically and thermally stable silk fibersin vitro, using a physiological buffer and ambient conditions (a process described in Section 1.3.2).

4RepCT has been functionalized by fusion to a wide range of biofunctional motifs such as a cell-adhesion integrin-binding motif RGD (FN-4RepCT) (Figure 1.1c) [20]

and an antibody (Immunoglobulin G (IgG)) binding Z domain (Z-4RepCT) [21].

These two modifications and the unmodified 4RepCT were used in this thesis. Other functionalizations include, but are not limited to, Xylanese which degrades xylan (Xyl-4RepCT) [22], SAL-1 which degrades bacterial cell walls (SAL-1-4RepCT) [23], the antimicrobial motif Magainin I (Mag-4RepcT) [24], and the biofilm degrading Dispersin B (DspB-4RepCT) [23]). The various functionalization make the materials formed out of the spidroins interesting for several biomedical applications such as antibacterial coatings and scaffolds for cell growth [22–24], further discussed in Chapter 4.

1.3 T URNING SPIDROINS INTO SILK FIBERS

"How they [spiders] originally succeeded in converting food into a natural glue which on entering the air instantly becomes silk, we shall never be able to understand"

British Encyclopedia 1928 [4].

(25)

It would be nice to write that all the time and effort scientists have put into proving the previous statement wrong, would have been fruitful. However, there are still aspects of the spider’s spinning process that are not fully understood [18]. Note that spinning does not refer to the action of rotating something but rather to the process of making insoluble filaments from an aqueous protein solution. Key in this process is the alignment and change in the secondary structure of the spidroins [15].

More general information and details about protein structure and sequences can be found here [25].

1.3.1 H

OW SPIDERS SPIN SILK

In spiders, spidroins are produced by specialized cells lining the epithelium of the specific gland (Figure 1.1d). They are stored at high concentrations (up to 50 wt%) in what is referred to as dope [26]. The fact that the spidroins do not aggregate upon storage is quite extraordinary and is thought to be due to the spidroin’s secondary through quaternary structure [3, 17].

There are currently two, not mutually exclusive, models describing the spidroin’s tertiary and quaternary structures. According to these, the spidroins are stored either (1) as micelles where the terminal domains form a hydrophobic outer shell shielding the repetitive region or (2) as a liquid crystalline feedstock [3]. As of late, it has been shown that spidroins tend to prefer micelle formation at lower concentrations and the ordered hexagonal columns associated with crystalline feedstock at high concentrations [17]. While there still is some uncertainty regarding the more fundamental secondary structure of the spidroin during storage, various studies indicate that they take on a random-coil and polyproline-II helix-like secondary structure [3, 17]. The latter structure helps maintain solubility by preventing the formation of intramolecular hydrogen bonds by favoring the formation of hydrogen bonds with the solvent [17]. During spinning, these structures are converted into polyalanine β-sheet crystals flanked by amorphous glycine-rich repeats, where the latter contains both 31-helical and type II β-turns [3].

Both physical and biochemical processes facilitate the structural conversion of the spidroin’s secondary structure [26]. During spinning, the dope moves distally through the gland and to the spigot through a long progressively narrowing s- shaped duct [18]. The dope’s high viscosity prevents capillary break up [15] and the slow elongational flow in combination with the exerted sheer forces causes the proteins to align. As the dope flows through the duct, it is also exposed to changes in pH and ion concentration, which facilitates protein conversion [18]. For example, lowering the pH throughout the channels is thought to neutralize acidic side chains, allowing tighter interactions. Changes in the ionic composition enhance intramolecular interaction, and towards the end of the duct, water is reabsorbed, increasing the spidroin concentration leading to increased interactions [26].

The final step of transition from high-density liquid to solid starts in the duct’s distal part, where rapid water removal initiates liquid-solid phase transition. A semi-solid, premature fiber is moved through the duct and finally exits the spigot.

(26)

1.3. TURNING SPIDROINS INTO SILK FIBERS 7

Liquid-solid transition is completed after the fiber enters the air through a combination of drawing and water evaporation. However, the latter is not essential as there are silks spun in aqueous conditions [17]. The full process of forming a fiber takes a fraction of a second [3] and begins with some environmental cues where the spider ejects a droplet of spidroin solution on a substrate and then draws away from the droplet, leaving a solidified β-sheet rich fiber in its path [15].

1.3.2 A

RTIFICIAL SPINNING

As described above, the spider’s natural spinning process is a complex combination of biochemical exchanges, extrusion, and drawing [17]. A full imitation of the process is not possible, however, there are several different ways in which the two key components of the spinning process, alignment and change in secondary structure, can be reached to produce spider silk fibers artificially. A few well-developed approaches are introduced below.

Wet spinning: the dope is extruded from a syringe into a coagulation bath, and a fiber is formed through precipitation in a non-solvent such as monohydric alcohols (methanol, ethanol, isopropanol) or ammonium sulfate [27, 28]. The formed fiber is then rolled up and generally subjected to post-processing, for example, immersion in another coagulant, post-drawing, or steam-annealing [27]. The post-processing is critical for the alignment of the β-sheet crystals and influences the mechanical properties of the final fiber, meaning that the properties can be tuned for the desired applications [27, 29].

Dry spinning: the dope is ejected from a syringe into the air, where fiber forma- tion occurs due to evaporation for a volatile solvent. While the method has been used successfully for reconstitutedB. mori dope [27], it has so far not been used to form mechanically stable fibers out of recombinant spidroins [28].

Microfluidic chip: the dope is ejected into microchannels and subjected to dif- ferent flows and biochemical processes. This process comes the closest to mimic aspects of the natural spinning process such as ion exchange, pH change, and elongational flow [28]. As this method allows for fine-tuning these parameters, it has been used to gain further insight into the kinetics and sequence of silk assembly in the spider [17, 30]. One critical aspect of fiber formation in microfluidic chips is the concentration of the dope. While native silk dope behaves like a molten polymer at high concentration, reconstituted does not. Consequently, high flow rates are necessary for fiber formation in microfluidic chips [17]. Note that it also is possible to combine wet-spinning and microfluidics such that the microfluidic chip is used to create shear forces on the spidroin solution before extrusion into a coagulant bath [31].

Electrospinning: the dope is extruded into an electric field, which yields repul- sive forces on the solution, resulting in the eruption of a thin jet that stretches towards the collector (i.e., a counter electrode). Solvent evaporation leads to fiber formation. Here the β-sheet transformation is assisted by electric field interactions

(27)

with hydrogen bond dipoles of the protein, which stabilizes helical segments and initiates the formation of β-sheets. However, post-treatment in an organic solvent is generally necessary to render the fiber water-insoluble [28].

Phase separation: the spidroin dope undergoes liquid-liquid phase separation, and fibers can be hand-pulled from the high-density phase of the solution. In short, by altering conditions such as concentrations, pH, or ionic composition, macro- molecule interactions become stronger and overcome the spidroins entropic tendency to remain homogeneously mixed. This leads to the formation of a highly viscous phase with a concentration of protein assemblies nearing 75%. Fibers are pulled by either inserting a pipette tip and moving it away from the solution [32,33]

or capturing a droplet in a pair of tweezers and pulling them apart. The fiber extension induces α-helical to β-sheet transition, and strain can be used to induce orientation and alignment [34].

Self-assembly: the spidroins in the dope self-assemble at the liquid-air interface, upon which pulling [35], or gentle wagging initiates fiber formation [14, 36]. The latter has been used to make fibers out of 4RepCT (introduced in Section 1.2.3). In short, a horizontal tube is filled with spidroin solution, and the liquid:air interface initiates self-assembly at both ends of the tube. Wagging the tube lengthwise causes the solution to slosh back and forth, circulating the liquid, resulting in further interactions with the liquid-air interface. The spidroins begin to self-assemble into a film at the interface, and the compression and extension of the film caused by the wagging make the film wrinkle/fold, respectively. The motion causes ruptures in the film, which heal through continued self-assembly and the process repeats until a fiber has formed. This method has recently been further developed using a vertical oscillation set up driven by a syringe pump programmed to perform vertical motion with a defined period and amplitude [36].

(28)

C ^HAPTER 2

F ORMATION AND HANDLING OF SPI -

DER SILK MICRO - AND NANOSTRUC -

TURES

This chapter explores the opportunities and challenges associated with forming and handling spider silk structures on the micro-and nanoscale. It begins by describing how new silk formats can be formed by altering the techniques used to make fibers and why this is of interest. After this, the mechanisms governing spidroin self-assembly into silk at interfaces are introduced, followed by a few different ways to manipulate the liquid:air interface to generate new spider silk structures. The chapter ends by discussing various ways to handle and concentrate micro-and nanostructures for further use.

2.1 M OVING BEYOND FIBERS

The artificial production of spider silk fibers was initially of interest to further understand the spidroin assembly process and produce materials with superior mechanical properties [37]. While fibers are of interest for a wide range of applications, the thread format is quite limited and generally requires post-processing into other shapes (such as textiles) before usage [38]. As spider silk has such a unique combination of interesting material properties (biodegradable, biocompatible, mechanically robust, and elastic), researchers began exploring other ways to form silk structures, both on the macroscale in the form of coatings, films [38], and hydrogels [2,28,38] as well as on the microscale for applications within, for example, drug delivery [37, 39, 40].

These new silk formats can be formed by altering the techniques developed to form fibers (Section 1.3.2). For example, spherical microparticles can be formed

9

(29)

by mixing the dope with a kosmotropic salt and changing the ion concentration [37, 39, 40] or by initiating capillary break up in the stream in a microfluidic chip. Concentration-dependent gelation can be used to form hydrogels [2,28] and 3D foams can be formed by letting spidroins self-assemble around salt crystals [41]. It is also possible to form films and coatings out of spider silk by casting the dope and letting the solvent evaporate [38], further discussed in the next section. Beyond these basic techniques inspired by the natural assembly process, state-of-the-art nanomanufacturing techniques such as electron-beam (e-beam) and ion-beam patterning can be used to initiate (or break up) silk formation. Using this method, films with complex 2D and 3D topographies can be formed [42–45].

Several of these formation methods have been reviewed in various combinations [2, 28, 40, 46]. As always, all methods come with their own set of advantages and disadvantages depending on the intended application. It is advan- tageous for biological and medical applications to use all aqueous processes as this eliminates the risk of toxic chemical residues in the final product [28]. One method suitable for this is the usage of interfaces to trigger silk-assembly [19].

2.2 S PIDROIN SELF - ASSEMBLY AT INTERFACES

At the most fundamental stage, the presence of an interface will cause proteins to adsorb, refold, and assemble [19, 47]. For metastable proteins such as spidroins, this leads to structural rearrangements [19]. As a result, when the initially disordered spidroins in solution adsorb to an interface, they immediately transition to β-sheets [19, 47, 48]. The formed β-sheets interact with each other and form cross- links through intermolecular interactions, forming a stable and cohesive film [48].

For the interested reader, more information and details about protein structure and folding can be found here [25].

There are three potential interfaces for spidroins in solution: liquid:air, liquid:solid, and liquid:liquid. The liquid:air interface has primarily been used to directly study the self-assembly mechanisms [19, 24, 48, 49] while the liquid:solid and liquid:liquid has been used to form silk, and then use the properties of the formed silk to further the understanding of the self-assembly process [47, 50, 51]. In Paper III, this was changed by developing a novel method to lift the interfacial film from the liquid:air interface, enabling new studies of the formed silk. The material studies revealed that the film formed at the liquid:air interface was permeable (further discussed in Section 3.4.2). Based on this, when referring to the work in Paper III, the interfacial film will be called a membrane.

2.2.1 S

PIDROIN SELF

-

ASSEMBLY AT LIQUID

:

AIR INTERFACES There are several techniques available to study the adsorption of proteins at the liquid:air interface. For example, ellipsometry can be used to determine the amount

(30)

2.2. SPIDROIN SELF-ASSEMBLY AT INTERFACES 11

of proteins at the interface, surface pressure to characterize lateral interactions, and surface rheology to reveal if a rigid network has formed. Using these techniques to look at spidroin assembly has revealed that the spidroins rapidly anchor to the interface and immediately begin to form lateral interactions [49]. The rapid adsorption is due to the spidroin’s amphiphilic nature [47]. Both the adsorption of molecules and development of intermolecular forces slow down and level of with time [48]. Eventually, the initially formed rigid layer develops into a thicker, viscous film [24].

The film’s thickness depends on the initial spidroin concentration, where a higher concentration results in a thicker silk film [24, 48]. Interfacial studies also revealed that the relative β-sheet content in the film increases with time (Figure 2.1). The β-sheet content can be determined using, for example, polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS) [48]. This method involves shining polarized light onto a film and recording the reflected light, where the adsorbed wavelengths provide information about the chemical identity, geometry, and coordination of the species in the film [52]. Figure 2.1 shows such a spectrum for the formation of an interfacial spider silk film at various time points. The spectrum is dominated by the amide I region (1626 and 1697 cm⁻¹) associated with intermolecular β-sheets. The peak around 1655 cm⁻¹ is associated with disordered secondary structures, and the peak at 1516 − 1537 cm⁻¹ with amide II vibrations consistent with β-sheet structures. Over time, the amide I region increases while the shape remains constant, suggesting that spidroins continuously gather at the interface without conformational change. The increase in the amide II/amide I upon continued adsorption suggests a reorientation of the proteins at the interface [48].

Figure 2.1. PM-IRRAS spectra at the airwater interface recorded as a function of time showing an increase in β-sheet content. Reprinted with permission from [48], copyright 2013 American Chemical Society

These two effects were also observed in Paper III, where it was shown that a stable membrane formed more rapidly when a higher spidroin concentration was used and that the final thickness depended linearly on the initial concentration. In

(31)

Paper III, a comparatively longer time frame was used for formation observations, which showed that the β-sheet content continued to increase even after the growth had stopped. The discrepancy to previous work [48] could be due to the longer time used allowed for evaporation, resulting in more β-sheet transitions.

When it comes to the formed film’s internal structure, a few different observations have been made. Adsorbing the interfacial film formed by the recombinant spidroin 4RepCT after a few minutes of assembly, revealed that the spidroins form nanofibrils at the interface (Figure 2.2a) [19]. When looking at a film’s internal structure formed from native spidroins through transmission electron microscopy, spherical aggregates were observed in an otherwise homogeneous film (Figure 2.2b- d). Cutting of the film and directly imaging it while at the liquid:air interface using Brewster angle microscopy (BAM) confirmed the presence of spherical aggregates (Figure 2.2e) [48]. In Paper III, observations confirmed the existence of both fibrillar and spherical structures. The development of a method (further discussed in Section 2.4.3) to lift the intact FN-4RepCT membrane allowed further studies of the internal and interfacial appearance. Scanning Electron Microscopy (SEM) revealed that the film had a smooth top (air) side, an internal nanofibrillar structure, and a bottom (water) side textured by spherical aggregates (Figure 2.2f-g).

Figure 2.2. Various imaging techniques used to visualize the self-assembled silk at the liquid:air interface. a) AFM image 4RepCT nanofibrils formed at the liquidair interface. TEM images of an interfacial native spidroin film at b)5 keV (scale bar = 2 µm) c) 300 keV (scale bar = 50 nm) d) 5 keV (scale bar = 2 µm) e) BAM image of the interfacial film at the airwater interface after being cut with tweezers. The white area corresponds to the film, and the black area, to water (scale bar = 100 µm.) SEM images of the f) smooth top side g) the nanofibrillar cross-section h) the textured bottom side of the membrane lifted from the liquid- air interface in Paper III (scale bars = 2 µm). a) is reprinted with permission from [19], copyright 2018 American Chemical Society, b-e) are reprinted with permission from [48], copyright 2013 American Chemical Society, f-h) are reprinted with permission from Paper III, Wiley under the Creative Commons Attribution License.

(32)

2.2. SPIDROIN SELF-ASSEMBLY AT INTERFACES 13

2.2.2 S

PIDROIN SELF

-

:

SOLID INTERFACES Most of the silk formation studies on solid substrates focus on studying the formed silk products and using their properties to reveal details about the self-assembly mechanism. In essence, soluble spidroins cast on a substrate form a film through self-assembly at the liquid:air and liquid:solid interface in combination with evaporation [38]. The final appearance and the structure of the formed film are affected by the surface’s properties, the cast layer’s thickness, and the solvent used.

Starting with the surface itself, if the spidroin solution is cast on a hydrophilic substrate such as glass, the film becomes more hydrophobic than if it is cast on a hydrophobic substrate [53]. This effect is due to the spidroins’ amphiphilic nature, which causes the protein to maximize interactions between similar blocks and min- imize interactions between dissimilar blocks [51, 53]. On a hydrophilic template, the hydrophilic regions will orient towards the surface, and the hydrophobic will orient to form micellar structures, away from the surface and the bulk as well as towards the liquid:air interface (Figure 2.3). The effect is reversed on a hydrophobic substrate [53]. This outcome was also observed in the membrane formed at the liquid:air interface presented in Paper III. In this work, the top surface of the membrane, facing air during formation, is more hydrophobic than the bottom side, which was in constant contact with the aqueous solution.

Figure 2.3. Illustrations showing the influence of the substrate on the secondary structure of a cast spider silk film. A hydrophilic substrate (left) leads to overall less β-sheet structures, but β-sheet exposure at the interface, making the surface of the film hydrophobic. A hydrophobic substrate (right) has the reverse effect with a higher β-sheet content within the film but a more hydrophilic surface. Reprinted with permission from [53], The Royal Society of Chemistry

Just as was the case for the film formed at the liquid:air interface, the thickness of formed silk film cast on a substrate depends on the initial concentration of the spidroin solution [51]. However, the thickness of the cast solution also affects the surface properties of the film. If the layer is very thin (< 10 nm), the spidroins can only interact with the substrate in a monolayer like manner, and the proteins at this solid:liquid:air interface remain unstructured. This is because the only possible interaction for the spidroins is with the surface itself, and there is no chance to form the interactions necessary for β-sheet formation. As the thickness of the

(33)

solution layer increases, there is more room for interplay and folding and inter-and intramolecular interactions increase, leading to the formation of larger hydrophobic and hydrophilic blocks. The larger β-sheet patches will arrange towards the air-interface and away from the bulk and form a film with a higher contact angle (i.e., more hydrophobic). As a result of this behavior, the β-sheet content is not uniform throughout the thick films. The majority of the β-sheets, about 85%, are located at the surface. The remaining 15% arrange in a micellular like fashion in the bulk phase [51]. This behavior was also observed in Paper I, further discussed in Section 2.3.1.

The final aspect that significantly affects the properties of the formed film is the solvent used. Using organic solvents which tend to evaporate quickly freezes the spidroins in a metastable, unstructured, state as they do not have sufficient time to fold into β-sheets [51]. This tends to produce films with holes. Slowing down the evaporation rate by reducing the airflow or changing to an all aqueous process can resolve this. Using a slower evaporation process allows proper folding of the proteins, which eliminates the need for post-treatment [38].

2.2.3 S

PIDROIN SELF

-

:

LIQUID INTERFACES A liquid:liquid interface can be generated using two immiscible liquids, such as oil and water. Here, the self-assembly effects are in-line with the observations made for the liquid:air and liquid:solid interfaces. The liquid:liquid interface has primarily been used to form emulsions so that the spidroins assemble around the spherical droplets in solution to form microcapsules [47] and to support fiber formation in microfluidic chips [54].

2.3 M ANIPULATING THE LIQUID : AIR AND LIQUID : SOLID INTERFACES

Most of the current work on silk formation at the liquid:air and liquid:solid interfaces have focused on furthering the understanding of the self-assembly mechanisms or finding ways to form stable films/coatings and evaluate their properties.

However, it is also possible to manipulate the interfaces to make the spidroins self- assemble into new silk formats on the micro-and nanoscale. This can be done by manipulating a spidroin droplet on a superhydrophobic surface (SHS) or introducing an interlocked pillar scaffold or air bubbles into a spidroin solution.

2.3.1 U

TILIZING SUPERHYDROPHOBIC SURFACES

Two standard definitions in surface science, introduced as common knowledge in the previous section, are hydrophobicity and hydrophilicity, loosely translated to water-repelling and water-loving [55]. In the ideal case, once a droplet is placed on

(34)

2.3. MANIPULATING THE INTERFACES 15

a surface, it either completely wets the surface or forms a sphere. The first surface would be classified as superhydrophilic or super wetting, while the latter would be superhydrophobic or anti-wetting. However, most surfaces end up somewhere between these two cases, and their wettability is classified using the contact angle.

If a droplet is placed on a horizontal surface (Figure 2.4), the contact angle is defined as the angle formed between the solid, liquid, and gas interfaces [56]. If the liquid is distributed over the surface so that the contact angle is less than 90^◦ the surface is considered hydrophilic (Figure 2.4a) [55, 56]. If the liquid retains a small contact area with the surface so that the contact angle is larger than 90^◦ the surface is considered hydrophobic (2.4b). A surface is considered superhydrophobic if the contact angle is over 150^◦ (2.4c) [57].

Figure 2.4. Illustration of the contact angle on a a) hydrophilic, b) hydrophobic and c) superhydrophobic surface.

How a droplet of water interacts with a surface depends mainly on the material itself and the surface roughness [56,58]. The maximum contact angle for water on a flat surface is about 120^◦. By adding surface roughness, this value can be increased to 170^◦ [56]. Surface roughness can be introduced in various ways, commonly by introducing micropillars. When a droplet is placed on such a patterned substrate, it can take on one of two states: Wenzel or Cassie-Baxter (Figure 2.5a-b). In short, in the Wenzel state, the liquid penetrates between the pillars, pinning the droplet, causing it to stick. In the Cassie-Baxter state, the liquid does not penetrate between the pillars but remains in a so-called non-wet-contact mode. In this state, the droplet can easily roll off the surface. Note that a droplet can move from the Cassie-Baxter state to the Wenzel state or remain somewhere in between due to, for example, evaporation [59]. More detailed information about the different states, applications of SHS, and production methods are found here [56].

In general, which state a droplet ends up in depends upon a combination of the surface material, the liquid used, as well as the size, interspacing, and shape of the microstructured pillars [60]. Introducing, for example, nanostructures ontop of micropillars enhances the surface hydrophobicity [59]. It is also possible to reduce the importance of the surface material by making the pillars re-entrant or even

(35)

doubly re-entrant (shaped like a T) (Figure 2.5c-e). In the latter case, the liquid first has to wet the top surface, then continue down the sides of the T, upon which the surface tension begins to point upwards, preventing the liquid from penetrating further. Note that the liquid:solid fraction, that is, the relationship between the pillar size and interspacing, still is a critical factor in suspending the droplet even when the micropillars’ structure is changed [60]. In Paper I and V, pillars of the re-entrant design were used.

Figure 2.5. Illustration of a droplet in a) the Wenzel state and b) the Cassie- Baxter state followed by illustration of c) a regular pillar, d) a re-entrant pillar (design used in Paper I and V), and e) a double re-entrant pillar.

SHSs are of interest for a wide range of applications such as anti-bacterial surfaces, anti-fog coating, anti-freeze surfaces [56], and self-cleaning surfaces [57]. As of late SHSs have also been used to guide evaporation for concentrating low-concentrate solutions [61], to locally deposit salt crystals [58], and to form nanowires of various polymers and biological materials [62–64]. Adding spidroins into the solution further allows for the formation of silk in the format of patterned films, microfilms, nanowires, and various combinations there-off (as detailed below).

MM-SIZED,MICRO-PATTERNED FILMS

SHSs can be used to manipulate soluble spidroins into forming patterned, insoluble films in a rapid, all aqueous, one-step process (Paper I). Patterned spider silk films are of interest for applications such as guided cell growth. They have previously been formed using two-step processes involving the formation of a film followed by either through capillary filling of a polydimethylsiloxane (PDMS) mold [65] or solvent assisted microcontact molding [66]. Beyond the fact that multiple steps were required to generate the film, both methods involved organic solvents (formic acid [65] and toluene [66]) which can leave behind toxic chemical residues [28] and destroy the native protein structure and thereby its function. All aqueous processes have been used to make patternedB. mori films. However, long formation times or post-treatments are generally needed to render the films water-insoluble [67, 68].

(36)

A patterned film can be formed using a SHS by placing a droplet of spidroin solution on the surface and letting it evaporate under ambient conditions (Paper I).

The droplet retains its spherical shape, which is characteristic for the Cassie-Baxter state, and the spidroins self-assemble at the liquid:air interfaces and solid pillar heads (Figure 2.6). The silk formation between the pillar heads prevents the solution from moving into the Wenzel state during evaporation and the solution remains suspended. As a result, the formed film encapsulates the bulk, where the spidroins orient into micelles as described in Section 2.2.2. As the solution evaporates, a double-layered film encapsulating spherical aggregates is formed (Figure 2.6). The top side of the film, facing the surroundings, becomes wrinkled, while the bottom side is smooth with a regular pillar imprint corresponding to the surface patterning.

The formation time depends on the size of the droplet in combination with the hu- midity. A micropatterned sheet from a 20 µL droplet forms in ambient conditions within 5 hours.

Figure 2.6. a) Illustration of the formation of a patterned spider silk film on a superhydrophobic surface: i) Droplet with spidroins is placed on the SHS, ii) the droplet evaporates while silk formation takes place at the liquid:air and liquid:solid (pillar) interfaces and spherical aggregates form within the solution, and iii) the formed film. SEM images of b) the top side and c) the patterned bottom side of a 4RepCT silk film and d) top view and e) side view of a cracked film showing the internal spherical aggregates (scale bars = 10 µm). b-c) are reprinted with permission from Paper I, Wiley.

The film’s diameter depends on the surface structure, the droplet’s size, and the spidroin concentration. The surface structure is key in determining the contact angle. A higher contact angle will lead to a smaller interface for the film to form on and vice versa. The film begins to form right after droplet deposition, pinning

(37)

it to the substrate pending a sufficiently high spidroin concentration (for 4RepCT above 0.3 mg/mL). As a result, the formed film will correspond to the initial droplet size. At lower concentrations, evaporation will occur before film formation can take place, and the formed film will be smaller than the initial droplet size. This behavior is in line with observations of deposition of particles due to evaporation on superhydrophobic surfaces [69].

MICRON-SIZED FILMS

The formation of spider silk films has mainly been done through casting and spin coating, generating mm-sized films [38,51]. Using SHSs, it is possible to scale down the formed films’ size to µm in diameter and nm in thickness, and shape them into any desired shape in a one-step, rapid, all aqueous process (Paper I and IV).

Previous work on producing spider silk microstructures has at large either focused on spherical particles in solution [37, 39, 40, 47] or surface patterning using complex techniques such as e-beam and ion-beam [42–45].

Miniaturized spider silk films can be formed using a simple, one-step technique introduced in Paper I as "touch-and-release," previously referred to as the "clinging- microdroplet" technique [58]. The method involves bringing the SHS close to a liquid, upon which the solution adheres to the pillars through van der Walls forces.

As the surface is moved away from the solution, liquid bridges are formed between the solution and the surface. Eventually, the liquid bridge breaks, depositing a droplet on top of each pillar (Figure 2.7). The process can be observed visually as the droplet does not retain a spherical shape but rather elongates upon removal from the substrate. Introducing a nanoscale pattern on the pillars prevents the adhesion of the solution [58].

Figure 2.7. Illustration showing the "clinging-microdroplet" technique. a) The solution is brought into contact with the surface and adheres to the top of the pillars, b) the solution is moved away, and liquid bridges are formed between the solution and the pillars, c) as the solution is moved further away, the liquid bridges break and droplets are deposited on the pillar heads. Reprinted with permission from [58], Wiley

In previous work, deposition of microdroplets was achieved by lowering the surface onto the solution, covering the entire area [58]. In the work presented in Paper I, the process was reversed, and a small droplet was brought into contact with the surface, allowing for selective, uniform patterning of a localized area (Figure 2.8a).

(38)

Figure 2.8. a) Fluorescent image of a superhydrophobic surface, selectively patterned with spider silk, using post fluorescent staining for visualization (scale bar

= 100 µm). SEM images of micro-sized spider silk films after being released from a superhydrophobic substrate with b) round pillars (scale bar = 2 µm) and c) elongated hexagonal pillars (scale bar = 10 µm).a) is reprinted with permission from Paper I, Wiley.

Spidroin self-assembly on micropillars occurs in the same way as film formation (Section 2.2.2), both towards the liquid:air and liquid:solid interfaces. As such, the deposition of a droplet results in the formation of a micronsized-film identical to the pillars’ shape and size. The formed micron-sized silk films also retain the shape of the pillars upon release (a process described in Section 2.4.1) (Figure 2.8b-c) (Paper V). This method enables the simple production of silk microstructures of a wide range of precise shapes and sizes, as these are directly related to the pillar’s appearance. Note that this feature is unique for silk, in previous work, deposited particles and salts have not been influenced by the shape of the pillars [58].

NANOWIRES

Spider silk nanofibrils have previously been formed through self-assembly at the liquid:air interface of a standing solution (Figure 2.2a) [19] as well as through salt initiated assembly in solution [70]. The uncontrolled self-assembly does not allow for precise control over the sizes of the nanofibrils [19,70]. When using electrospinning, it is possible to control the formed fibers’ diameter, but precise control over the length and orientation remains a challenge [28]. Superhydrophobic surfaces can be utilized to form uniformly sized nanowires of controlled length through Liquid Bridge Induced Assembly (LBIA) (Figure 2.9). This method has previously been used to form nanowires from for example polyvinyl fluoride (PVF) [71], calcein [62], DNA [63], and fibrinogen [72]. Current progress and developments were recently reviewed [64].

The LBIA process is fairly simple; the solution containing the molecule that is to form the nanowire is rolled over the superhydrophobic surface. Moving the droplet over the surface results in the formation of liquid bridges between the pillars. If

Spider Silk Nanostructuring and its Applications for Tissue Engineering

Spider Silk Nanostructuring and its Applications for Tissue Engineering

LINNEA GUSTAFSSON

Spider Silk Nanostructuring and its Applications for Tissue Engineering

LINNEA GUSTAFSSON

Contents

L IST OF P UBLICATIONS

A BBREVIATIONS

A IMS AND STRUCTURE OF THIS THESIS

P REAMBLE

A IM AND OBJECTIVES

T HESIS STRUCTURE

C HAPTER 1

I T ’ S ALL ABOUT SPIDER SILK

1.1 S PIDER SILK - A MULTIFUNCTIONAL MATERIAL

1.2 O BTAINING SPIDROINS

1.2.1 N

1.2.2 R

1.2.3 4R

CT

1.3 T URNING SPIDROINS INTO SILK FIBERS

1.3.1 H

1.3.2 A

C HAPTER 2

F ORMATION AND HANDLING OF SPI -

DER SILK MICRO - AND NANOSTRUC -

TURES

2.1 M OVING BEYOND FIBERS

2.2 S PIDROIN SELF - ASSEMBLY AT INTERFACES

2.2.1 S

-

:

2.2.2 S

-

:

2.2.3 S

-

:

2.3 M ANIPULATING THE LIQUID : AIR AND LIQUID : SOLID INTERFACES

2.3.1 U

C ^HAPTER 1

I ^T ’ S ALL ABOUT SPIDER SILK

C ^HAPTER 2