Campus Monterrey
School of Engineering and Sciences
Biofabrication of nanoenhanced hydrogel fibers for muscle tissue engineering using surface chaotic flows: Chaotic 2D-printing
A thesis presented by
Ada Itzel Frías Sánchez
Submitted to the
School of Engineering and Sciences
in partial fulfillment of the requirements for the degree of
Master of Science In Nanotechnology
Monterrey Nuevo León, June 9th, 2020
Dedication
To my mother and my father, for the direct and indirect lessons for life and your undenying example
of hard work and integrity
To Ari and Fabián, for being my unconditional mental and
emotional support and for constantly encouraging me to pursue higher goals
3
I would like to express my deepest gratitude to the following people and institutions:
To Dr. Grissel and Dr. Mario, for accepting me in your team, investing huge amounts of time and effort in your students and volunteers, and for individually and publicly recognizing our work. Most importantly, thank you for cultivating a family-like environment in our group and motivating all of us to attempt to solve big problems.
To Dr. Ivonne, for all the expertise that was essential for the nanotechnological perspective of the whole project, but also for personal and professional advices offered along the way. Thank you for helping me remain positive and promoting a uniting attitude in the lab.
To Dr. Mohamadmahdi, for performing the CFD simulations and for the insights derived from them. Also, thank you for the kind and pertinent suggestions on this work.
To Dr. Fernando Ponz for your expertise and generosity in providing us with the “nano”
feature of this project, and for introducing us, through Dr. Ivonne, to the wonderful world of viruses.
To Dr. Anne-Sophie, for the straightforward and relevant feedback throughout the presentation of the advances of my thesis. Thank you for questioning the foundations of the project and helping me provide viable explanations for the collected evidence as a whole.
To M.Sc. Regina, for your support in the long SEM sessions that gave rise to the beautiful micrographs shown in this thesis.
To Diego Q, you are simply the best lab partner anyone can wish for. Thank you for your great attitude, hard work and creativity. Without a doubt, your talent and ideas made the journey particularly enriching and helped me grow as a researcher.
To Jorge T., for helping me answer every huge and tiny question I could think of. You are a truly wise and upright person and I am extremely grateful for having the opportunity to work with (and learn from) such a talented, professional, and committed researcher.
To Hugo, for enabling the production of great quality GelMA that allowed so many projects in the lab to function properly. Also, thank you for all the intellectual and well-founded opinions regarding the project and the different perspectives you brought to the discussions.
To Sara and Jorge O., you were my support throughout every academic step of the master’s program. Thank you for being patient with me and all the explanations of the most basic (and later complex) questions I had.
To all the members of the Alvarez-Trujillo Lab, especially Andrea, Beto, Brenda, Carlos, Caro, David, Gil, Johana, Miriam, Mónica, and Valeria. Your questions, suggestions, direct help, and even your mere company made my days in the lab better. Thank you all.
To Tecnológico de Monterrey for the financial support on tuition and CONACyT for the support for living.
4
______________________________________
Biofabrication of nanoenhanced hydrogel fibers for muscle tissue engineering using surface
chaotic flows: Chaotic 2D-printing
______________________________________
by
Ada Itzel Frías Sánchez
5
Multiple human tissues exhibit a fibrous nature. Therefore, the fabrication of hydrogel filaments for biomedical engineering applications is a trending topic. Current tissue models are made of materials that often require further enhancement for appropriate cell attachment, proliferation and differentiation. Here we present a simple strategy, based on the use of mathematically modelable surface chaotic flows, to fabricate continuous, long and thin filaments of gelatin methacryloyl (GelMA) added with Turnip mosaic virus (TuMV) for enhanced muscle tissue engineering. The fabrication of these filaments was achieved by chaotic advection in a finely controlled and miniaturized version of the journal bearing (JB) system. A drop of a pre-gel solution of GelMA was injected on a higher-density viscous fluid (glycerin) and a chaotic flow was applied through an iterative process. The hydrogel drop exponentially deformed and elongated to generate a fiber, which was then photocrosslinked under exposure to UV light.
Computational fluid dynamics (CFD) simulations were conducted for the design and prediction of our results. GelMA fibers were then used as scaffolds for C2C12 myoblast cells, and the effect of adding plant-based viral nanoparticles (VNP) to the hydrogel fibers as nano-scaffolds for cellular growth was evaluated.
Chaotic 2D-printing was proven to be a viable method for the fabrication of hydrogel fibers. CFD simulations accurately predicted the lengths of the printed fibers, and a correlation coefficient of R 2=0.9289 was determined from the experimental and simulated data of the first two cycles. The hydrogel fibers were effective scaffolds for muscle cells and show potential to be used as cost-effective models for muscle tissue engineering purposes. TuMV significantly increased the metabolic activity of the cell-seeded fibers (p<0.05), strengthened cell attachment throughout the first 28 days, improved cell alignment to ~50%, and promoted the generation of structures that resemble natural mammal muscle tissue.
6
List of Figures
Figure 1. Tissue engineering paradigm 5
Figure 2. Experimental strategy 12
Figure 3. MiniJB device 14
Figure 4. Schematic diagram of fiber fabrication process 15 Figure 5. Different types of fibers fabricated 16 Figure 6. GelMA fibers fabricated through regular and chaotic miniJB flows 20 Figure 7. Width diversity in brightfield micrographs of chaotically printed fibers 22 Figure 8. Progression of ink deformation produced by regular and chaotic flows 23 Figure 9. Graph of experimental and simulated fiber lengths 24 Figure 10. Effect of the initial ink position on elongation 25 Figure 11. Brightfield micrographs of C2C12 cells seeded on GelMA fibers 26 Figure 12. Metabolic activity of fiber samples throughout the first 24 post-seeding days 27 Figure 13. Fluorescence micrographs for cell morphology analysis 28 Figure 14. Cell nuclei density of GelMA and GelMA-TuMV fibers throughout time 29
Figure 15. Cell orientation analysis 31
Figure 16. Superficial and cross-sectional SEM characterization 32 Figure 17. Comparison between real and artificial muscle tissue 33
7
Table 1. Fiber biofabrication strategies 7
Table 2. Experimental fiber lengths 21
Table 3. CFD and experimental elongation comparison 23 Table 4. Prospective sequences for gene expression quantification 35
Table A1. List of abbreviations 37
Table A2. List of institutional acronyms 38
Table B1. Description of variables and units 39
Table B2. List of symbols 39
8
Contents
Abstract v
List of Figures vi
List of Tables vii
Introduction 1
Justification 2
Proposal 3
Theoretical Framework 4
Skeletal muscle tissue engineering (SMTE) 4
Current fiber biofabrication technologies 5
Chaotic printing 7
Hydrogels 8
Nanotechnology in tissue engineering 9
Objectives 11
General objective 11
Specific objectives 11
Experimental strategy 12
Materials and methods 13
Materials 13
Hydrogel preparation 13
MiniJB assembly 14
Fiber fabrication 14
Computational fluid dynamics (CFD) simulations 16
Cell culture studies 17
Scanning electron microscopy (SEM) 18
Image analysis 18
Statistical analysis 18
Results 19
Fiber fabrication 19
Computational fluid dynamics (CFD) simulations 22
Cell-laden fibers 25
Metabolic activity 26
Staining of actin filaments and nuclei 27
Image analysis 29
Scanning electron microscopy 31
9
Conclusions 34
Perspectives and future work 35
Appendix A 37
Abbreviations and acronyms 37
Appendix B 39
Variables and Symbols 39
Appendix C 40
Mini Journal Bearing (miniJB): 40
Bibliography 41
Published papers 50
Curriculum Vitae 51
10
Chapter 1
1. Introduction
Tremendous efforts have been made in the field of tissue engineering over the past few decades to develop tissue and organ substitutes for in vivo replacement and in vitro studies [1].
These substitutes can potentially help shorten drug development timelines [2], reduce the amount of animals used for testing pharmaceuticals [3], and even function as transplants for patients that suffer the loss or deterioration of an organ [4]. Simple organs have successfully been developed and implanted in patients [5–7], however, the fabrication of more complex ones is still a challenge nowadays. More work needs to be done to improve the architecture, composition and interaction of materials within lab-fabricated tissue at finer and finer resolutions [8,9].
Most organs in the human body are composed of fiber-like structures. Nerves, tendons, blood vessels, bones and muscles are some examples of tissues exhibiting a fibrous or tubular nature [10]. This makes fibers a highly attractive architecture in the field of tissue engineering [11]. In particular, muscles make up over 40-45% of the human body mass of an adult [12] and are essential for locomotion and the whole metabolic activity of the body [12,13]. Even though muscles are capable of regenerating themselves to some extent from mild damages [14], when serious injuries or intrinsic defects occur, additional strategies need to be implemented in order to improve muscle quality and maintain functionality.
Tissue engineering offers a wide spectrum of alternatives for tissue repair, fabrication, and implantation. Among many possibilities, the interplay between biomaterials science and nanotechnology have contributed substantially to the development of nanoengineering solutions for tissue repair and regeneration [15]. Since cell behaviour may be regulated by nanoscale features, different nanotechnology approaches open the possibility for preventing, diagnosing and treating various diseases and medical conditions [16]. Furthermore, the use of nanoparticles, nanofibers, nanotopography and other nanotechnological strategies allow us to enhance the functioning of artificial tissue models by allowing us to control cell-cell and cell-scaffold interactions [17].
A wide variety of biofabrication methods are available nowadays. In this work, a novel, efficient, and cost-effective 3D-printing method is used for the biofabrication of nanoenhanced hydrogel fibers for muscle tissue engineering purposes.
1.1 Justification
Certain diseases, injuries, aging and even inactivity can cause irreversible damage to skeletal muscles [18–21]. This damage can exceed the self-healing capacity of muscles and lead to losses in function, which strongly impact the life quality of people who suddenly require assistance to perform daily tasks. To prevent this, some medical procedures, mainly autografts and heterologous implants, have been implemented to repair large-volume skeletal muscle defects. However, these surgeries are far from ideal, and often cause complications in patients and are greatly limited by tissue availability [22].
Nowadays, entire research fields are working synergistically to design and fabricate bioengineered tissues to prevent and treat these conditions. These artificial tissues can serve in future years as models for fundamental research, new therapy development, and as potential alternatives for organ and tissue transplants in a longer term [4]. To this aim, models have evolved from oversimplified 2D cultures (monolayers) to increasingly complex 3D structures that allow a more faithful recapitulation of the microarchitecture of native tissues [23] and can minimize ethical concerns regarding animal testing [24].
Producing muscle-tissue models is a non-trivial task; nonetheless, many useful fiber biofabrication techniques are available [11]. However, these methods often involve the use of extensive technical expertise, complex and often pricy equipment, and usually have a very limited portfolio of working materials. Although these technologies have reduced their costs throughout the years [25], the resolution obtained with the low-cost biofabrication methods to produce fibers is insufficient for building faithful 3D models of real tissues [26,27]. Furthermore, most state-of-the-art biofabrication systems involve complex multi-component equipment that may require high maintenance [27].
Several suitable materials for fiber fabrication lack essential components for cell attachment and proliferation. In order to enhance these materials, many nanotechnological
Chapter 1. Introduction
3
solutions can be implemented (e.g. addition of nanoparticles). However, the use of additional components can affect the biocompatibility and increase fabrication cost. [28]
1.2 Proposal
In this work, a simple, efficient and cost-effective surface-printing method is demonstrated for the biofabrication of three-dimensional fibers for muscle tissue engineering purposes. With this aim, we used a miniaturized version of the Journal Bearing (miniJB) flow system, previously used to fabricate highly packed 3D-microstructure within small hydrogel constructs through a technique called chaotic 3D-printing [29]. In a novel embodiment of this technique, two liquids with different densities, namely glycerin and gelatin methacryloyl (GelMA), can be used to fabricate hydrogel fibers in surface 2D-flows under laminar conditions. The miniJB works as a microfabrication wheel to elongate the water-based pre-gel drop in an exponential fashion into a long and thin filament, which is then gelled by photocrosslinking.
Two versions of hydrogel fibers were assessed as scaffolds to fabricate our muscle fibers: (1) Fibers made of pristine GelMA inks, and (2) those fabricated using GelMA/ Turnip mosaic virus(TuMV) nano-composites. TuMV nanoparticles were added to provide mechanical stability to the hydrogel fibers and to add a secondary nano-scaffolding structure that may serve as an amenable-to-functionalize platform to facilitate further increases of complexity. The performance of both types of fibers as scaffolds for muscle cells are tested.
The greatest challenge in this work is to achieve cell survival and growth in an organized fashion to obtain a model that can potentially perform similar functions as the real muscle tissue.
This proposal aims to provide the following benefits:
● Development of a simple and low-cost device for biofabrication of muscle fibers
● Implementation of computational fluid dynamic (CFD) simulations, which enable the prediction of features (and design) of the fibers
● Development of a protocol for fiber fabrication using chaotic flows
● Achievement of higher resolution structures than those obtained by popular low-cost biofabrication techniques
● Development of living-constructs that resemble the architecture of muscle fibers
2. Theoretical Framework
2.1 Skeletal muscle tissue engineering (SMTE)
Over the course of almost four decades, tissue engineering has become an increasingly relevant field in science. By combining cells, scaffolds and specific signals, new functional living tissues have been developed, with the objective of understanding the fundamental mechanisms of tissue generation and development, and ultimately generating functional replacements for lost or dysfunctional organs and tissues [30] (Figure 1). Certainly, the development of fibers is a fundamental part of tissue engineering and countless recent studies are devoted to this purpose. Fibrous scaffolds have been used for several tissue engineering applications, such as the fabrication or regeneration of bone, tendon, ligament, skeletal muscle, among others [11].
Furthermore, relatively simple organs from integumentary, cardiovascular, respiratory and gastrointestinal systems have successfully been developed and clinically tested [5,31,32].
Likewise, studies on tissues and organs of growing complexity are becoming more frequent in scientific literature [33–35].
Muscles are highly abundant and complex. Between 500 and 1,000 skeletal muscles comprise the human body [36]. They can be classified into different types according to their mitochondrial and myoglobin content, motor protein activity, and ATP generation [37]. Each muscle has tens to hundreds of fiber bundles, which contain axially aligned packs of myofibrils that carry the contractile units, called sarcomeres [38]. One of the most critical structural properties of the muscle fibers is myofibril alignment, since it enables appropriate force transmission and contractile responses to the electrical stimuli provided by the motor neuron [38].
Great improvements have been made on cell orientation with the development of different methods for pre-aligning muscle cells and promoting muscle formation in vitro using several biodegradable and non-degradable scaffolds [39]. The benchmark cell line used in skeletal muscle tissue engineering is the mouse myoblast C2C12 [40], which is very responsive to the surface of the materials where it attaches. This fact has been leveraged to develop effective tools for myotube alignment by controlling the topography in scaffolds (i.e., by
Chapter 2. Theoretical Framework
5
incorporating grooves and ridges[41], as well as micro- and nanoscale patterns [42]). Additional stimuli (i.e., physical [43], electrical [44], mechanical [45] and magnetic [46]) have also been provided and shown great results for promoting myogenic differentiation [37,47].
Figure 1: Tissue engineering paradigm. Tissue engineering aims for the combination of cells, scaffolds and signals to build complex organs and tissues. Image adapted from Wikimedia. Modified from:
Blausen.com staff (2014) WikiJournal of Medicine1 (2) [48]
2.2 Current fiber biofabrication technologies
Biofabrication is the result of continuous interplay between engineering, biology and materials science, aiming to develop functional biological products [49]. 3D-bioprinting, as well as microfluidics and spinning techniques are popular options for the biofabrication of fibers, each of them with its advantages and limitations (Table 1).
Numerous techniques can be classified as 3D-bioprinting. The most popular ones are extrusion-based, laser-assisted and inkjet, with further subdivisions according to the specifics of their printing mechanisms [50]. Bioprinting allows the fabrication of three-dimensional structures with high precision on the spatial distribution of bioinks and can yield complex architectures.
Computer-aided design (CAD) is used to define the structure to be printed and a 3D-printer deposits the bioinks, generally in a layer-by-layer fashion, to fabricate such structure. Despite these valuable advantages, bioprinting still faces important challenges. For instance, the portfolio of biomaterials that possess the physicochemical properties required for the process are still limited [51]. Moreover, these bioprinters are multi-component and their resolutions are restricted by some key parameters of each printer [52,53].
Electrospinning is a particularly popular fiber-fabrication technique. Basically, a fiber is generated by applying strong electrical fields between a static or dynamic collector and a syringe that injects a polymer formulation. Topographies that could induce enhanced cell adhesion, proliferation and alignment can be generated with this technique, which is why electrospun scaffolds are extensively used in tissue engineering. However, these scaffolds are mainly dry polymeric structures that need other materials or further treatments to ensure cell survival. Commonly used electrospinning materials have to be modified to provide them with crucial cell-recognition signals [54] required for cell attachment. In addition to this, electrospun fibers are unsuitable for cell encapsulation, due to the harshness of the fabrication process and the fact that their dimensions are relatively small, typically ranging from a few micrometers to 100 nm or even less. Other spinning techniques, such as microfluidic spinning and wet-spinning are also available for fiber fabrication with different sets of materials [25].
Microfluidic-based fabrication uses fluid dynamics principles in microscale channels to manipulate liquids and generate fibers. Long and continuous filaments with various architectures can be accomplished, as well as multi-material structures [55]. In contrast with electrospinning, microfluidic spinning constitutes a milder process that allows the use of more sensitive materials, which is why the use of natural polymers is possible with this technique [54].
However, material selection for microfluidic-based fabrication is still limited and the development of additional materials is challenging, due to the fast solidification rate required for continuous fabrication [54].
Chapter 2. Theoretical Framework
7
Table 1. Fiber biofabrication strategies
Technique System
complexity Cost Speed Resolution Compatible
materials Cell viability References
Electrospinning Low Medium- high
Low-high (tunable)
100 nm - few
µm Limited Extremely
low ~0% [11,25,49,5 6–60]
Wet spinning Low Low Low-high
(tunable) ~30-600 µm
Suitable for natural polymers and
composites
Extremely low in some
sections of the fibers and high in others
[11,25,61]
Microfluidic
spinning Medium Low-
medium Low
Difficult to achieve
<100 µm
Suitable for natural polymers
High >80%
[25,54,59,6 2,63]
Extrusion-based
3D printing Medium Medium- high
Low- medium
200-1,000 µm
Wider range of viscosities
Low-high 40-95%
[25,64–66]
Laser-based 3D
printing High High Medium 10-100 µm
Restricted to low viscosities
1-200 mPa
High >95% [25,64–66]
Inkjet 3D printing Medium Low-
medium High 10-50 µm
Restricted to low viscosities
<10 mPa
High >85% [25,64–66]
2.3 Chaotic printing
In recent years, a novel biofabrication technique called chaotic printing was developed by Trujillo-de Santiago et al. [29]. This is a flow-based technology that can potentially achieve micro- and even nano-scale resolutions at exponentially fast rates and stands out due to its remarkable simplicity, portability and low cost. Briefly, a small cylindrical reservoir bears a hydrogel in its liquid state (pre-gel) and functions as an elongation wheel in which an ink drop located within is stretched and folded repeatedly until achieving pre-modeled inner lamellar structures. Hydrogel constructs were fabricated with GelMA and ink drops containing fluorescent particles, nanoparticles, or cells for the 3D microstructure generation. Since the flows used for printing are governed by the Navier-Stokes equations of motion, the process of deformation and elongation caused by the chaotic flow can be mathematically modeled. This characteristic is highly valuable, since it enables the use of computational fluid dynamics (CFD) to simulate and design the desired constructs.
Here, the printing process was accomplished by using a device called miniJB, basically a miniaturized version of the widely used journal bearing (JB) system for mixing highly viscous fluids at low speeds [67]. In the JB system, and any other chaotic system, the occurrence of
chaos depends on geometry factors and operational conditions. The miniJB consists of a cylindrical reservoir (outer cylinder) and an eccentrically located mixing shaft (inner cylinder); the reservoir and shaft rotate alternately in opposite directions according to a given mixing protocol or recipe. In the JB flow, the off-centered location of the inner cylinder is a mandatory requisite to produce chaotic flows. The counterclockwise angle of rotation of the internal cylinder (α), a momentary stop, and the clockwise angle of rotation of the external cylinder (β), fully define a mixing protocol: [α°, β°]. An Arduino UNO® microcontroller-based system (overall cost below 150 USD) was implemented for accurate and independent control of the speed and angle of rotation of the motors that drive the movement of each cylinder.
Chaotic printing is compatible with numerous appealing materials for biofabrication. The system allows the use of any Newtonian fluid under laminar conditions. This enables the implementation of highly viscous fluids, which are commonly difficult to use in many biofabrication techniques (Table 1).
2.4 Hydrogels
Hydrogels have strongly supported the transition from 2D to 3D cell culture [68]. These polymeric networks have an impressive swelling capacity and closely resemble the extracellular matrix (ECM), which is the set of non-cellular components of tissues and organs that provide physical support and function as regulators of critical cellular processes [69]. Hydrogels have been a popular subject of study in biomaterials engineering. This has led to noteworthy improvements on their physicochemical properties over the years[70]. Ideally, hydrogels should be fully biocompatible, provide enough mechanical support for cells at initial stages and be biodegradable in a reasonable time span for cells to be able to produce their own ECM.
Common naturally-derived hydrogels are based on alginate [71,72], collagen [73,74], gelatin [68], hyaluronic acid [75], and hybrids [76–78], and their gelation can be achieved through several physical and chemical methods that enable their use as bioinks in many biofabrication processes.
GelMA hydrogels are a popular choice because of their biocompatibility, tunability and low cost [68]. Since gelatin is derived from collagen, the main component of the ECM of mammals, the immune response associated to their implantation is relatively low [79]. The
Chapter 2. Theoretical Framework
9
mechanical properties of GelMA inks can be easily customized for specific applications by modifying the degree of methacrylation (DoM), the concentration, and UV light intensity and exposure time. Altering the amount of methacrylic acid incorporated in the synthesis process varies the DoM; GelMA pregels are generally used in concentrations in the range of 5%-15%
(w/v), depending on the specific application [80]. By adding photoinitiators such as 2-hydroxy-1-[4-(2-hydroxyethoxy) phenyl]-2-methyl-1-propanone (Irgacure 2959) and lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) to the liquid pregel, the gelation can be triggered by UV light exposure. GelMA is a versatile option for biofabrication and has been applied in numerous tissue regeneration approaches [80].
2.5 Nanotechnology in tissue engineering
A number of additional tools can be used in hydrogel scaffolds for providing a more faithful replication of natural biosystems. To this aim, nanotechnology has provided a wide range of sophisticated solutions to create smart hydrogels by incorporating nanoparticles, micelles, colloids, among others. Nanotechnology has contributed to many areas of tissue engineering and regenerative medicine, for instance, the modeling and repair of bones, nerves, skin, muscle, etc [81]. One of the many applications of these nanotools is the enhancement of cell alignment, differentiation, and adhesion to scaffolds [28].
Viral nanoparticles (VNPs) and virus-like particles (VLPs) are versatile nanobiotechnological platforms that can be directed to many different biomedical applications, such as sensing devices, drug delivery systems, and imaging tools[82,83]. These nanotools are mainly made of protein, and are thus biodegradable. In addition to this, plant-viruses are harmless for mammals, can be extracted through processes that reach high yields, and can be produced relatively fast and at a low cost [84].
Some virions, like the Turnip mosaic virus (TuMV) are especially attractive for tissue engineering purposes because their dimensions and flexibility are similar to some of the components of the ECM. These properties could provide cells with morphological cues that may improve their development. TuMV is a filamentous virus from the Potyvirus genus, composed of around 2000 identical coat protein subunits of 33 kDa. Cell-TuMV nanoparticle interactions are currently being studied through flow cytometry and confocal microscopy, and the virions have
shown indications of membrane affinity with several cell lines (i.e., Caco-2, HEK, HeLa, MCF-7, and MCF10a) [85]. Furthermore, TuMV can be functionalized with one or multiple compounds [86] and can potentially be aligned by shear flows in hydrogels, as other filamentous nanomaterials[87]. Altogether, these characteristics make these nanomaterials highly attractive for exploration in the biomedical field.
Chapter 3
3. Objectives
3.1 General objective
To produce nanoenhanced hydrogel fibers using chaotic 2D-printing and evaluate their performance as scaffolds for muscle tissue biofabrication
3.2 Specific objectives
● To establish and standardize a printing protocol for fiber fabrication using chaotic flows
● To fabricate hydrogel fibers using chaotic 2D-printing and compare their length and width to the one predicted by the CFD simulations
● To use chaotically 2D-printed nanoenhanced hydrogel fibers as muscle tissue scaffolds and evaluate their biological performance
4. Experimental strategy
Figure 2. Experimental strategy. Standardization of a chaotic printing protocol, fabrication of fibers using the miniJB device and evaluation of the biological performance of hydrogel scaffolds.
Chapter 5
5. Materials and methods
5.1 Materials
Type A porcine skin gelatin, methacrylic anhydride (MA), Dulbecco's phosphate-buffered saline (DPBS), 4′,6-diamidino-2-phenylindole (DAPI), paraformaldehyde, formaldehyde and penicillin-streptomycin (pen/strep) were purchased from Sigma-Aldrich. Phosphate-buffered saline (PBS), fetal bovine serum (FBS), and Dulbecco's modified Eagle's medium (DMEM) were purchased from Gibco. Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) was purchased from Allevi, glycerin was purchased from a local supplier, murine C2C12 myoblasts were purchased from ATCC. Actin-Phalloidin iFluor 647 was purchased from AbCam, PrestoBlue®
was purchased from Invitrogen, and ethanol (EtOH) was purchased from Hycel de México.
CreateX fluorescent particles were purchased at an online store. TuMV virions were kindly provided by Fernando Ponz's lab at Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), Madrid, Spain.
5.2 Hydrogel preparation
5.2.1 GelMA synthesis
Type A porcine skin gelatin was mixed at 10% (w/v) in DPBS AT 60°C. After being fully dissolved, 20% (w/v) of MA was incorporated dropwise in order to introduce the methacrylate substitution groups in the gelatin. The reaction was stopped by adding a 4 times dilution of warm DPBS. In order to remove the residues of methacrylic acid, the mixture was continuously dialyzed for 5 days at 40°C using a custom-made dialysis system and later lyophilized for 5 days and stored at -80°C until use.
5.2.2 GelMA inks
All pregel inks were prepared with 10% (w/v) GelMA and 0.067% (w/v) LAP photoinitiator dissolved in DPBS at 60°C. Fluorescent particles were first washed by diluting 500 µl of fluorescent magenta CreateX particles in 14 ml of distilled water, followed by a thorough
homogenization and a 5-minute centrifugation process at 1500 rpm. Excess water was replaced with new distilled water and the whole process was repeated twice. The remaining fluorescent particles were homogenized in 1 ml GelMA pregel. Virus-embedded inks included a 10 µg/ml concentration of TuMV virions, isolated from Indian mustard plants as described in the literature [88].
5.3 MiniJB assembly
An inner mixing shaft of radius rin and an external reservoir of radius routare the main cylindrical components that comprise the miniJB. The designed geometry of our system is based on two dimensionless parameters shown in Figure 3A): cylinder radii ratio, r=rin/rout, and the ratio of the eccentricity ( e) to the outer cylinder radius, ϵ=e/rout. Two arrangements were used: the regular (ϵ=0) and the chaotic (ϵ=1/3), and for both r=1/3. The actual diameters of the external and internal cylinders were 15 mm and 5 mm, respectively (Figure 3B). The process was carried out in cycles, composed of two mixing actions (Figure 3C): first, the counterclockwise rotation of the inner cylinder ( ⍺), followed by the clockwise rotation of the outer one (β). Of the many mixing protocols [⍺°, β°] available in the literature, we selected the [270°, 810°]. In order to adequately control the movement of each cylinder, we used an Arduino platform and 2 stepper motors. Our device was made of materials that can be easily found in hardware or online stores, and its overall cost is below 150.00 USD.
Figure 3. MiniJB device. A) Dimensionless parameters, B) actual dimensions, C) representation of one cycle of the process.
5.4 Fiber fabrication
We fabricated hydrogel fibers in a batch process using chaotic surface flows induced by the miniJB system (Figure 4). After the assembly of the device, 400 µl of glycerin (⍴=1.13 g/cm 3)
Chapter 5. Materials and methods
15
was placed inside the reservoir, and a 1 µl drop of GelMA pregel ( ⍴=1.01 g/cm3) was deposited on top. The device was later activated for a given amount of cycles, followed by a 15 s UV light exposure at 365 nm, glycerin drainage, and PBS was used to wash any fluid remainders (Figure 4). In these experiments, the angular velocity for both cylinders, the mixing shaft and the reservoir of the system, was set to 26.5 rpm. All the process was carried out in sterile conditions inside a biosafety hood. The lengths of the fibers were measured with the Zen Blue Edition software (Carl Zeiss Microscopy GmbH, Germany) and plotted according to their cycle number.
The collected data was used to determine the experimental Lyapunovexponent by calculating the slope of the linear portion from the plot of equation (1):
n(L/L ) n
l 0 = Λ (1)
where n is the number of flow cycles, L is the length of the fiber after n cycles, and L0is the original length of the filament (in this case, the diameter of the drop.
Figure 4. Schematic diagram of fiber fabrication process. Long and thin hydrogels fibers were fabricated using surface 2D chaotic flows generated by the miniJB.
All types of fibers produced for the validation of the printing process and the biological experiments are shown in Figure 5.
5.4.1 TuMV coating: superficial virus
Hydrogel fibers, produced with a 1-cycle process, were placed in 24-well ultra-low attachment well plates and covered with a 10 µg/ml suspension of TuMV in PBS and incubated for 24 hours at 4°C prior to the cell-seeding process.
5.4.2 Cell seeding
A 400 µl C2C12 cell suspension in growth medium (4x10 4cells ml-1) was seeded on all the chaotically-printed hydrogel fibers (1 cycle) in 24-well ultra-low attachment well plates.
Fibers were incubated at 37°C in a humidified 5% CO 2 atmosphere for 3 h to allow cell attachment on the fiber surface. The cell suspension was then replaced by DMEM growth medium with FBS 10% (v/v). The metabolic activity, as well as the cell morphology and alignment were monitored at different days.
Figure 5. Different types of fibers fabricated. A) Fibers loaded with fluorescent particles, Bi) GelMA and Bii) superficial TuMV GelMA fibers with C2C12 cells, and C) TuMV-embedded GelMA fibers with
C2C12 cells.
5.5 Computational fluid dynamics (CFD) simulations
Numerical modelling was implemented to determine relevant mathematical properties of the chaotic flow. First, the Navier-Stokes equations, formulated for incompressible laminar flows in a rotating coordinate system, were solved in the well-known journal bearing (JB) geometry.
Two rotating domains were used to describe the independent rotation of the internal and external cylinders. The dynamics on the transition points were neglected in our simulations, meaning that we assumed that convection at each cycle ceased immediately after stopping each mixing action (i.e., inertial effects were neglected). Simulations were performed in
Chapter 5. Materials and methods
17
COMSOL Multiphysics® 5.4 (Comsol Inc., MA, USA). Two velocity fields were solved for inner and outer rotations using “Rotating Machinery, Laminar Flow” physics and “Frozen Rotor”
stationary solvers. A fluid viscosity of 1 Pa·s and density of 1.12 g/ml were considered for the material properties of glycerin at room temperature. In each rotation, a non-slip boundary condition was considered for the fixed walls, and a rotating wall condition was applied to the wall of the moving cylinder. Subsequently, a particle tracking algorithm with massless formulation was implemented in the pre-calculated velocity fields to evaluate the dynamics and the position of all points composing the GelMA solution interface. A freeze boundary condition was considered in particle tracking and a time-dependent solver was used with multiple steps corresponding to multiple rotations to track the position of up to 50,000 particles in the system.
For the estimation of the Lyapunov exponent in our 2D-chaotic flow protocols, we simulated the deformation of 1 mm lines (i.e., collections of points) located at different radial positions within the JB surface and calculated the value as we did with the fabricated fibers.
5.6 Cell culture studies
5.6.1 Metabolic activity analysis
Cell viability was indirectly assessed by analyzing cell metabolic activity on days 0, 1, 2, 3, 6, 9, 12, 18, and 24 using the PrestoBlue ® assay. Briefly, fibers were covered in DMEM culture medium with 10% v/v PrestoBlue® reagent and incubated for 2 h at 37°C. In a 96-well plate, 100 µl of medium were dispensed, and the fluorescence was measured in a microplate reader (Synergy™ HTX Multi-Mode Microplate Reader, BioTek Winooski, VT, USA) at 530/570 nm excitation and emission wavelengths.
5.6.2 Staining of actin filaments and cell nuclei
The cell morphology was analyzed using fluorescent staining methods on days 1, 7, 14, 21, and 28. Briefly, the cell-laden fiber was fully covered with 4% paraformaldehyde solution in PBS for 30 min at room temperature to fix the cells. Samples were later stained with 1X Phalloidin iFluor 647 for filamentous F-actin, and 1 µg/ml DAPI-PBS for cell nuclei. After 90 min of incubation at room temperature, the samples were examined in an inverted microscope (Axio Observer Z1, Carl Zeiss Microscopy GmbH, Germany), using the Zen Blue edition software.
5.7 Scanning electron microscopy (SEM)
5.7.1 Biofabricated fiber sample preparation
The microarchitecture of our samples was analyzed on days 1 and 14 by SEM. The sample preparation involved a cell-fixation process and a dehydration process. The former consisted of a 4% paraformaldehyde followed by a 4% formaldehyde sample submersion, and the latter by a 4-hour ethanol-based graded dehydration process. Once samples were fully dry, they were covered with a 5 nm gold layer using a Q150R ES rotating sputter machine (Quorum, United Kingdom) to make the surface electrically conductive. These samples were analyzed using an EVO® MA 25 SEM (Carl Zeiss Microscopy GmbH, Germany).
5.7.2 Real muscle tissue sample preparation
Calf gluteobiceps and gracilis muscles were dissected and submerged in PBS with 1%
pen/strep, cut in 1 cm 3cubes and gradually frozen until reaching -80 °C. The fixation process for SEM analysis of these samples was the same as the previously described. Real tissue samples were dehydrated for 8 hours and gold-coated with a layer of 10 nm thickness.
5.8 Image analysis
The number of nuclei and the cell orientation were determined from fluorescent microscopy images using Fiji/ImageJ (National Institutes of Health, AZ, USA) software. In order to determine the nuclei density, 15 sections of 100x100 µm were analyzed per micrograph. For the orientation analysis, cells whose direction was within ±10° with respect to the fiber longitudinal axis were considered aligned. Angles were normalized to 0° and plotted in polar graphs using a custom-made program in MATLAB (MathWorks, MA, USA).
5.9 Statistical analysis
All the experimental data regarding fiber dimensions was reported as mean ± standard deviation (SD). The statistical analysis for the metabolic activity assay data was performed using one-way ANOVA, and the nuclei density statistical analysis was assessed with two-way ANOVA, with significance levels of (*) p<0.05, (**) for p<0.01, and (***) for p<0.001.
Chapter 6
6. Results
6.1 Fiber fabrication
Long GelMA fibers were fabricated using surface 2D flows generated by a lab-made miniJB system. The JB flow was first described by Swanson and Ottino [67] as a simple flow system capable of producing chaotic flows. Here we have miniaturized this system to enable the creation of hydrogel fibers of thicknesses in the range of hundreds to tens of micrometers. To that aim, we generated a 3D flow with a viscous fluid (glycerin), on top of which we placed a small GelMA pregel drop of approximately 2.5 mm in diameter. The difference in density between both fluids is key for this method, since it allows the pregel drop to remain on the surface of the glycerin during the whole process and enables a 2D deformation of the hydrogel.
In this work we used two miniJB setups: the concentric and the eccentric. When the two cylinders are concentric, the flows generated are regular and cause a linear elongation, while the eccentric arrangement generates chaotic flows that cause an exponential stretching.
Representative fluorescence micrographs in Figure 6 display the resulting fibers achieved through the regular and chaotic setups for the first two cycles. Evidently, the fibers fabricated using regular flows were substantially shorter than those produced by chaotic flows at the second cycle. Additionally, the thicknesses observed in linearly stretched fibers also display more consistent widths along the structure. In this particular approach, the glycerin is used exclusively as a vehicle for elongation of the GelMA drop. Glycerin subsequent drainage, as well as the removal of the mixing shaft, cause the fiber to lose the shape achieved at the end of the printing process. However, the dimensions of the fibers were accurately described by our CFD simulations.
Figure 6. GelMA fibers fabricated through regular and chaotic miniJB flows. Fluorescence micrographs of GelMA fibers fabricated through regular and chaotic arrangements and at different
numbers of cycles.
Table 2 shows the lengths of fibers fabricated using a different number of cycles in regular and chaotic flows generated by our miniJB system. High standard deviations can be attributed to initial differences in the drop placement, accidental losses of fiber portions due to experimental manipulation (these hydrogel fibers are extremely fragile), and intrinsic limitations of the microscopy equipment. The quantitative determination of fiber dimensions was performed using the Zen Blue software (Carl Zeiss Microscopy GmbH, Germany). Results reveal that our single initial drops were turned into filaments in average 56x longer after only two cycles of exposure to chaotic flow. In contrast, filaments stretched through the regular process only increased their length by a 28-fold in average (Table 3).
Conventional 3D-bioprinters generally generate structures at a linear rate, while our chaotic 3D-printer does it exponentially fast and at a noticeably lower cost. Therefore, chaotic 2D-printing can fabricate fibers of a given length much faster than a 3D bioprinter. Interestingly, the time economy achieved by chaotic printing is more evident as the number of cycles is increased, since the process of elongation of the hydrogel filament is exponential in chaotic flows. Note that the volume of the initial drop of GelMA is preserved during the process of elongation. Therefore, as the fiber elongates, its thickness is reduced. After a few cycles, we
Chapter 6. Results
21
can potentially achieve fiber thicknesses in the nanoscale range [29]; however, handling hydrogel fibers of these dimensions is a challenging task. Furthermore, the resolution of the currently employed methods for the analysis of the length and width of our samples will soon be insufficient to document remarkably thin filaments.
Table 2. Experimental fiber lengths. Experimental measurements from lengths of fibers fabricated through regular and chaotic flows at different numbers of cycles.
Cycle n Chaotic length (µm) Regular length (µm)
0.0 2,684.60 ± 12.14 2,684.60 ± 12.14
0.5 13,564.55 ± 447.86 14,519.65 ± 871.58
1.0 46,632.92 ± 4,426.09 38,253.77 ± 1,116.01
1.5 102,507.78 ± 1,786.41 58,385.91 ± 1,528.96
2.0 152,246.01 ± 21,121.50 75,972.51 ± 643.49
Relevant fiber dimensions for SMTE can be achieved since the first cycle. Chaotic 2D-printing can produce fibers with varying diameters (Figure 7), as a result of the different levels of stretching that specific sections of the fiber undergo. These variations can be beneficial for the development of muscle-fiber models that can be used for specific tissue engineering applications. For instance, the depressor anguli oris, a facial muscle that has a maximum length of 48 mm and width ranging from 8.8 to 35 mm [89]. This muscle could be built with chaotically printed fibers in a bottom-up approach. Fibers with even diameters can also be useful and can easily be achieved through the regular arrangement in the same miniJB system. Since we can design appropriate protocols to achieve specifically aimed dimensions, we could adapt our printing process for the fabrication of different types of fibers.
Figure 7. Width diversity in brightfield micrographs of chaotically printed fibers. Different field-of-view showing an important diversity of widths in single printed fibers, ranging from 16.8 to 386.4
µm.
6.2 Computational fluid dynamics (CFD) simulations
Deterministic flows, whose dynamics are governed by the Navier-Stokes equations of motion, were used for the chaotic printing process. Therefore, the deformation and elongation caused by the chaotic flow can be mathematically modeled. The geometry of the system, the physical properties of the materials used, and the low-speed angular velocity involve a laminar regime that can be modeled as a 2D flow, which allows a numerical calculation of the velocity fields using CFD. Chaotic flows are capable of generating microstructure at an exponential rate [90,91] due to their inherent properties (i.e., existence of a unique flow intrinsic skeleton and asymptotic directionality).
Multicolored lines shown in Figure 8 were tracked both for the regular and chaotic simulations of the miniJB system flows throughout the first 5 cycles. This simple exercise, with relatively low computing investment, provided an accurate prediction of the structure observed
Chapter 6. Results
23
at the top layer of the working fluid and enabled the estimation of the Lyapunovexponent (Λ) of the surface chaotic flow. A profound physical meaning is conveyed through the Λ value; this parameter describes the overall potential (and rate) of a chaotic mixing protocol to produce elongation and generate structure. CFD simulations determined a Λ value of 1.97 cycle -1for the first two cycles, which closely matched the one calculated from the experimental data (2.01 cycle-1).
Figure 8. Progression of ink deformation produced by regular and chaotic flows. CFD simulations of the deformation of 1 mm lines caused by regular and chaotic flows throughout 5 cycles.
In order to make a direct comparison, the data gathered from the CFD simulations and the experiments are shown in Table 3. The experimental length has been normalized with respect to the diameter of the initial drop located in the system. The correlation between the experimental and simulated data was R2=0.9289, which indicates a high level of accuracy in the CFD predictions. A direct comparison of the simulated and experimental results is presented in Figure 9.
Table 3. CFD and experimental elongation comparison. Simulated and experimental elongation coefficients of fibers fabricated by chaotic 2D-printing throughout different numbers of cycles.
Cycle n CFDC ExperimentalC CFDR ExperimentalR
0.0 1.00 1.00 ± 0.01 1.00 1.00 ± 0.01
0.5 3.64 5.05 ± 0.17 5.45 5.41 ± 0.33
1.0 14.41 17.37 ± 1.65 10.75 14.25 ± 0.42
1.5 17.60 38.18 ± 0.67 16.09 21.75 ± 0.57
2.0 65.52 56.71 ± 7.87 21.42 28.30 ± 0.24
C: chaotic; R: regular
Figure 9. Graph of experimental and simulated fiber lengths. Normalized length of experimental fibers with lengths predicted through CFD simulations.
Altogether, results demonstrate that we can reproduce the behavior of the actual experimental system using CFD simulations; therefore, our simulations enable the prediction (and design) of the filaments to be produced. However, these simulations also highlighted low movement areas, which indicate that the initial drop position can induce important variations in the length of the fiber. In order to determine the level of stretching that each part of the centerline undergoes throughout the printing process, a material line composed of 50,000 particles was tracked throughout four cycles. Figure 10 shows the distances between these particles at an initial stage and after 2 and 4 cycles. Particles located between particles 1 and 25,000 exhibit relevant increases in the distance between neighboring particles both after the second and fourth cycles in most sections, in comparison to their original position. Meanwhile, particles in the 25,000-34,000 range display very similar distances after two cycles and some get even closer to each other after the fourth cycle (i.e., 27,000-28,000). Locating an initial drop at the position of the particle 30,000 will therefore yield a different fiber length than placing it in the position of particle 20,000.
Chapter 6. Results
25
Figure 10. Effect of the initial ink position on elongation. Distances between neighboring particles along the centerline at initial condition (yellow), and after the second (purple) and fourth (blue) cycles for
the protocol [270°,810°]. Initially all particles are located 0.1 µm apart from each other. Three different initial drop positions in blue, green, and orange, are highlighted, as well as their evolution after 10 cycles.
6.3 Cell-laden fibers
Once a robust fabrication process for GelMA fibers was established, we tested different fabrication strategies for developing fibers containing actual cells (Figure 5A). We adopted C2C12 myoblasts as our cellular model. A fist strategy, possibly the most intuitive, consisted in formulating a cell-laden bioink (i.e., suspending a cellular pellet in 10% GelMA) and depositing a drop of this bioink as initial condition for our chaotic elongation process. Preliminary assays with cell-laden hydrogel drops indicated low post printing cell viability, possibly due to the shear stress that cells undergo during the elongating process. In any location of a chaotic flow, the stretching intensity can vary by several orders of magnitude, and cells are sensitive to these conditions [92]. Consequently, we opted to use the produced hydrogel as cell-scaffolds and seeding C2C12 cells on the surface of the GelMA fibers soon after fabrication as seen in Figure 11. Several variants of this strategy were investigated (Figure 5B,C). In the first one, cells were surface seeded on pristine 10% GelMA (Figure 5Bi). In the second, GelMA was surface treated
with TuMV particles (Figure 5Bii). In a third variation, TuMV VNPs were suspended in the hydrogel prior to the printing process (Figure 5C).
Figure 11. Brightfield micrographs of C2C12 cells seeded on GelMA fibers. A) GelMA and B) GelMA-TuMV fibers at days i) 1 and ii) 3.
6.4 Metabolic activity
Figure 12 shows the PrestoBlue® assay results, conducted to measure the metabolic activity of our cell-seeded fibers. We observed a gradual and substantial increase in activity over the first 12 days in all cases (i.e., GelMA fibers with or without TuMV), which is expected at proliferation stages and has also been reported for C2C12 cells seeded on GelMA scaffolds [93,94]. All samples followed approximately the same trend the first 6 days, but in the following days they exhibited significantly different behaviors (p<0.05). Fibers with superficial TuMV displayed a higher metabolic activity on days 9 and 12, which could be the result of interactions between the cells and the readily exposed VNPs. On the other hand, fibers with embedded TuMV had a steadier and slower metabolic activity increase, with a peak of activity on day 18, unlike the other samples, which had an earlier metabolic activity peak on day 12. This could possibly be explained by the idea of cells taking longer to be in contact with the TuMV nanoparticles.
Chapter 6. Results
27
Figure 12. Metabolic activity of fiber samples throughout the first 24 post-seeding days. A) Fluorescence readings from PrestoBlue assay *(p<0.05). Schematic representation of C2C12 cells
interacting with B) embedded, and C) superficial TuMV.
6.5 Staining of actin filaments and nuclei
Fluorescence microscopy images in Figure 13 strongly suggest that cell alignment occurs as cells populate the fibers. As the myoblasts start elongating and fusing with each other, the resulting multinucleated cells self-align with respect to the longitudinal axis of the fiber. The attachment of the cells decreases over time on fibers made with pristine GelMA.
However, the cells that remain adhered to the hydrogel seem to keep their orientation and form a highly packed and aligned fiber. Moreover, multinucleated cells on the TuMV-containing fibers appear to have a better attachment that is increasingly noticeable over time. For example, at day 21 and 28, a greater number of cells remain attached to the hydrogel fibers added with TuMV than to pristine GelMA fibers. At day 28, we observed highly dense tissue constructs that exhibit high degree of cell alignment in filaments containing TuMV at their surface.
Figure 13. Fluorescence micrographs for cell morphology analysis. Fluorescence stainings of C2C12 cells cultured on A) GelMA and B) GelMA fibers containing TuMV. F-actin shown in red and nuclei
in blue.
Chapter 6. Results
29
6.6 Image analysis
One obvious signal of cell population growth is an increase in nuclei density, which can be easily measured through image analysis techniques. We analyzed the fluorescence micrographs using Fiji/ImageJ (open source software) and found that the nuclei density was initially increasing rapidly, and decreasing for GelMA fibers in a later stage. Previous studies have reported the importance of the stiffness of scaffolds for an appropriate development of C2C12 cells. Our working concentration of GelMA exhibits a stiffness (compressive modulus of
~15 kPa [95]), which is similar to native conditions of muscles (10-15 kPa [95]). Gatazzo et al.
used gelatin-based hydrogels and concluded that reproducing a native-resembling stiffness of scaffolds for muscle tissue engineering increases proliferation and differentiation of C2C12 cells [96], and reported an equivalent increase in cell nuclei to the one shown in Figure 14.
Interestingly, the cell nuclei density graphs are consistent with the trend observed throughout the whole duration of the metabolic activity assay, which confirms a fast initial cell population growth during the first 14 post-seeding days. Moreover, these results support the hypothesis of cell death in the pristine-GelMA scaffolds during the last stage of the experiments (from day 21 to 28), by displaying a drastic decrease in nuclei density in the pristine-GelMA fibers. In contrast, TuMV-containing fibers seem to achieve stability on this very same stage.
Figure 14. Cell nuclei density of GelMA and GelMA-TuMV fibers throughout time. P-values *<0.05,
***<0.001
As mentioned previously, cell alignment is one of the key characteristics that need to be achieved in order to have an artificial muscle that has potential to be functional. The geometrical confinement where myocytes are enclosed, as well as the topography of the substrates where cells are seeded, have been proven to be relevant in their resulting orientation. Several techniques have been implemented in hydrogel scaffolds to improve cell orientation. Ebrahimi et al. fabricated micropatterned GelMA fibers (30% w/v) of ~500 µm through microfluidic spinning and demonstrated that the topographical cues, coupled with an agrin (a proteoglycan) treatment, have a beneficial and synergistic effect in cell orientation, backed up by an outstanding ~90% alignment of cells seeded on the surface [97]. In another recent study, Chen et al. provided mechanical stimulation to C2C12 cell-laden GelMA fibers (10% w/v) and reported that ~80% of cells aligned within ±10° [98]. Various studies have found that wider structures hinder C2C12 cell orientation [99–102], and plain hydrogels have been widely used as control groups to show that the lack of micro- and nanopatterns on large scaffolds results in very low degrees of alignment (~10%) [42,102–104]. In our study, the nanoenhanced GelMA fibers were seemingly randomly oriented at the beginning, as opposed to the GelMA ones that showed a relatively higher organization (Figure 15). However, a progressively increasing degree of alignment was henceforth observed on our fiber samples, both for GelMA and GelMA-TuMV.
Our orientation analysis indicates that after day 7, in average, ~50% of cells located in nanoenhanced samples successfully aligned, while only ~40% of the cells in pristine-hydrogel fibers did. Both types of fibers showed considerably higher degrees of alignment than the aforementioned control groups from the literature. Remarkably, this was accomplished in the absence of any sort of external stimulation. We can thus suggest that the elongated and narrow architecture of our fibers induces cell alignment.
Chapter 6. Results
31
Figure 15. Cell orientation analysis. A) Schematic representation of axial reference used for the determination of angles, and B) polar graphs comparing cell-orientation for C2C12 cells cultured on
GelMA and GelMA-TuMV fibers.
Cooperative approaches have also been performed by combining a diverse set of stimuli (i.e., topological, thermal, electrical, or mechanical) to enhance cell alignment [62,98,100,102,104–106]. Ahadian et al. improved cell orientation from ~50% to ~80% by electrically stimulating micropatterned hydrogel samples with an interdigitated array of electrodes. Therefore, we anticipate that by providing further stimulation to our fibers, we can achieve higher degrees of cell alignment. All in all, our experimental data confirms that C2C12 cells attach to the fiber-scaffold, successfully spread, proliferate, and align over time.
6.7 Scanning electron microscopy
In order to study the topography and internal microstructure of our fibers, we conducted SEM analysis. Figure 16 presents superficial and cross-sectional SEM micrographs from the chaotically printed fibers. These images reveal that GelMA and GelMA-TuMV fibers are sparsely populated at day 1 after seeding, and do not exhibit any cell organization. On day 14, striations can be seen on the surface of both types of fibers, which are similar to those observed in the
musculoskeletal hydrogel constructs fabricated by Heher et al. in a bioreactor that provides mechanical stimulation [104]. Cross-sectional views of our samples allow visualization of inner granular structures of similar dimensions to dehydrated myotubes from real muscle, which could suggest cell penetration into the hydrogel scaffold. More repetitions of staining assessments need to be performed to further support our hypothesis of “geometry-induced” cell migration. We plan to perform the same cell-seeding strategy in considerably thicker fibers and analyze their cross-section.
Figure 16. Superficial and cross-sectional SEM characterization. Cross-sectional and superficial SEM micrographs of A) GelMA and B) GelMA-TuMV fibers.
Cell migration can be directed by biochemical and biophysical interactions. Every cell type has its own mode of migration; C2C12 dynamic rearrangement depends on anchoring sites and inner actomyosin contractile forces [107]. In three-dimensional arrangements, myoblast infiltration occurs at porosities above 80 µm [108]. Ma et. al. worked with relatively large
Chapter 6. Results
33
composite collagen scaffolds of ~100-200 µm porosity and observed spontaneous cell penetration under dynamic culture conditions [109]. In case our cells fail to penetrate the scaffold, TuMV can be exploited as a migration-inducing platform. Previous reports show that these VNPs have been successfully functionalized with different compounds (i.e., fluorescent dyes, epitopes, peptides, biotinylated and polyphenolic compounds[86,110,111]). The presence of lysine and cysteine residues in the TuMV capsid enables the incorporation of growth factors that induce cell-specific responses. Consequently, we can guide one or multiple cell lines to migrate into our hydrogel through different combinations of biochemical stimulation.
Flow-induced organization of VNPs has also been accomplished previously by Wu et al. in rod-like nanoparticles [87]. Our flow-based technique can potentially align the TuMV, which would be beneficial for our purpose.
Several factors in the microenvironment affect the morphology of tissues. In Figure 17 , a straightforward comparison between our nanoenhanced muscle tissue-engineered fibers and a real mammal muscle tissue is made. Characteristic fibrillar-like structures observed in the real tissues can also be found in our lab-made fibers within similar scales, which suggests a structural and biological relevance. Their current resemblance, just like the cell alignment, can arguably be improved by providing further stimulation (i.e., mechanical, electrical), in order for the tissue to mature in a more physiologically relevant manner.
Figure 17. Comparison between real and artificial muscle tissue. SEM micrographs from A) native mammal muscle tissue and B) lab-made fiber (GelMA-TuMV) after 14 days of culture.
