Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Tunable Structural and Mechanical Properties of Cellulose Nanofiber Substrates in Aqueous Conditions for Stem Cell Culture Megan Smyth, Carole Fournier, Carlos Driemeier, Catherine Picart, E. Johan Foster, and Julien Bras Biomacromolecules, Just Accepted Manuscript • Publication Date (Web): 09 May 2017 Downloaded from http://pubs.acs.org on May 10, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Biomacromolecules is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
1
Tunable Structural and Mechanical Properties of Cellulose Nanofiber Substrates in Aqueous
2
Conditions for Stem Cell Culture
3
Megan Smyth1,2, Carole Fournier3,4, Carlos Driemeier5, Catherine Picart3,4 E. Johan Foster6*,
4
and Julien Bras1,2,7*
5
1. CNRS, LGP2, 461 Rue de la Papeterie, 38402, Saint-Martin-d'Hères, France
6
2. Université Grenoble Alpes, LGP2, 38000 Grenoble, France
7
3. CNRS, UMR 5628, LMGP, 38016 Grenoble, France
8
4. Université Grenoble Alpes, Grenoble Institute of Technology, 38016 Grenoble, France
9
5. Centro Nacional de Pesquisa em Energia e Materiais (CNPEM), Laboratório Nacional de
10
Ciência e Tecnologia do Bioetanol (CTBE), 13083-970 Campinas, São Paulo, Brazil
11
6. Virginia Polytechnic Institute and State University (Virginia Tech), Macromolecules
12
Innovation Institute (MII), Department of Materials Science and Engineering, Blacksburg,
13
Virginia 24061, USA
14
7. Institut Universitaire de France, 75005 Paris, France
15 16
KEYWORDS: Cell adhesion, cell viability, nanofibrillated cellulose, microfibrillated cellulose,
17
mechanical properties, material properties
18 19 20
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
21
ABSTRACT
22
Thin cellulose nanofiber (CNF) nanostructured substrates with varying roughness, stiffness
23
(Young’s modulus), porosity, and swelling properties were produced, by varying the conditions
24
used during fabrication. It was shown that with increased heat exposure, CNF substrate porosity
25
in an aqueous state decreased while Young’s modulus in a water submerged state increased. In
26
this study, the adhesion and viability of mesenchymal stem cells (MSCs) cultured on this CNF
27
substrate will be presented. Viability of D1/BALBc MSCs were assessed for 24 and 48 hours,
28
and it was shown that depending on the CNF substrate the viability varied significantly. The
29
adhesion of MSCs after 6 and 24 hours was conditional on material mechanical properties and
30
porosity of the CNF in cell culture conditions. These results suggest that material properties of
31
CNF nanostructured substrate within the aqueous state can be easily tuned with curing step
32
without any chemical modification to the CNF, and that these changes can affect MSC viability
33
in cell culture.
34
Table of Contents Graphic
35 36
INTRODUCTION
37
With being the most abundant polymer on earth, interest in the use of cellulose for a variety of
38
applications from paper or polymer composites to biomedical applications has increased annually
ACS Paragon Plus Environment
Page 2 of 37
Page 3 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
39
since the 1980s 1. Cellulose has favorable properties for biomedical researchers such as
40
biocompatibility, renewability and low cytotoxicity 2. The use of cellulose in biomedical
41
applications such as drug delivery, tissue engineering, and wound healing has become more of an
42
interest to researchers who want to use natural polymers 3. Since the 2000’s, a special focus on
43
nanoscale cellulose has been observed for such applications. For example, bacterial nanocellulose
44
(BNC) is commonly used in biomedical applications and is readily available on the market,
45
especially in the field of wound healing 4. Unfortunately, such material is expensive, its
46
production is difficult to upscale since such membranes are processed with difficulties. The use of
47
other forms of nanoscaled cellulose, cellulose nanocrystals (CNC) or microfibrillated cellulose
48
(MFC also called CNF), has gained greater interest with the announcement of their
49
industrialization the last 5 years. Respectively, discovered in the 50’s 5 and the 80’s 6 , they are
50
obtained from any cellulose fibers7 either by a chemical hydrolysis for CNC or by a mechanical
51
disintegration, mostly after a pre-treatment, for CNF. Books 8 and reviews 9-10 details their
52
production 11 and properties 12-14, but very recently they have been used in the field of biomedical
53
engineering with a focus on using cellulose as a scaffold for tissue engineering or cell culture 2, 15.
54
CNCs are mostly used to impart specific chemical properties to promote cell adhesion and direct
55
cell growth 16,17 or to provide a structure for cell adhesion in a particular direction 18. CNF has
56
been used as a tissue engineering scaffold in as an additive to a matrix, and most commonly
57
hydrogels 19,20,21,22. Other processing techniques can be used to create celluose based substrates
58
such as electrospinning23-24 and areogels25. Both of these techniques have been employed in
59
biomedical applications. It is worth noting that most of these scientific papers are very recent
60
(less than 2-3 years), proving by itself the novelty of our study.
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 37
61
Indeed microfibrillated cellulose has been first extensively studied for its use in composites and
62
as barrier films 26,12. Compared to cellulose nanocrystals, CNF is less crystalline, flexible and has
63
diameter of approximately 20-60 nm and length of several microns 12. The morphological
64
properties of CNF can vary depending on the source of the CNF, and the means of production.
65
Enzymatic pre-treatment, mechanical fibrillation, homogenization, and chemical treatment are co
66
mmon ways to produce CNF. Each production technique imparts different characteristics to the
67
CNF, such as surface groups, charge, and crystallinity 11. CNF has numerous hydrogen bonds on
68
the surface of the fiber, and because of these bonds and the entanglement of these flexible
69
nanofibers, they can produce a nanostructured substrate, also called films or nanopaper. Such
70
substrates present very good barrier or mechanical properties and can also be transparent. There
71
are different means to produce such nanopaper 27-28, and their large scale production process is
72
still an issue in spite of announcement of pilot scale roll of such films from VTT. Depending on
73
the manufacturing procedure, their properties might change as recently shown for barriers
74
applications 29. Others have shown that the structural properties, which influence barrier
75
properties of the CNF film might differ based on drying techniques 29-31, but there has been no
76
research conducted to fine-tune the mechanical properties for biomedical applications.
77
The use of mesenchymal stem cells was first discovered by Friedenstein in the 1970s
78
research on the work of Friedenstein showed that cells isolated from bone marrow had the ability
79
to differentiate into osteoblasts, adipocytes, and chondrocytes
80
many factors including adhesion and proliferation. It has been shown that without adhesion stem
81
cells will undergo apoptosis, since stem cells are reliant on substrate attachment, or are anchorage
82
dependent
83
microenvironment leading to contractility and reorganization of the actin filaments within the
34-35
33
32
. Further
. MSC viability is dependent on
. This anchorage of the cells at the cellular level occurs when cells probe their
ACS Paragon Plus Environment
Page 5 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
35-36
84
cytoskeleton into filopodia
. The adhesion of stem cells can be dependent on various factors
85
such as surface chemistry 37, stiffness
86
Engler and colleagues it was shown that MSCs can present different morphologies and spreading
87
patterns dependent on substrate stiffness, and that these differences in morphology were indicative
88
of various differentiation pathways 39. The variation of the mechanical properties of polymer based
89
substrates such as polyacrylamide 39 and the effect on MSC adhesion and viability has been studied
90
extensively, but to the best of the authors’ knowledge never on nanocellulose based materials with
91
tunable material properties.
92
Culturing of MSCs onto cellulose based substrates has been performed by a variety of
93
researchers. Most of the work is recent and focuses on bacterial nanocellulose substrates, not
94
CNF produced from wood, which is the focus of this study. Krontiras and colleagues used murine
95
MSCs for adipogenic differentiation using bacterial nanocellulose and alginate scaffolds. It was
96
shown that after 4 weeks in culture, adipocytes displayed lipid formation 40. BNC has been used
97
alone as a scaffold in work by Favi et al. Favi showed that BNC scaffolds seeded with equine
98
MSCs and differentiated in vitro promoted cell viability, proliferation, adhesion, and supported
99
chondrogenic and osteogenic differentiation 41. For plant-based nanocellulose, limited work has
35
, and topography
34, 36, 38
. In seminal work conducted by
100
been completed using MSCs with both CNCs and CNF. CNCs were dispersed in a poly(lactic
101
acid), followed by electrospinning to create a scaffold, and tested with human MSCs for basic
102
cytocompatibility by Zhou et al. 42. Xing and colleagues created a CNF hydrogel with gelatin and
103
tested with human MSCs and determined that MSCs maintained differentiation potential and
104
secreted extracellular matrix, but did not form cellular colonies on the CNF 19. Even if others
105
have checked the cytotoxicity of purely CNF aerogels or other substrates 22, 43 or performed
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
106
human cell culture 25, to the best of the authors’ knowledge, there has been no reported research
107
on only CNF based substrates with tunable mechanical properties and mesenchymal stem cells.
108
Thin CNF substrate mechanical and material properties were easily tuned by exposing the solvent
109
casted CNF films to elevated temperatures. This facile curing step created a change in the
110
material properties without any complicated chemical processes, or the addition of a polymer
111
matrix or filler. Mechanical and material measurements were analyzed in cell culture like
112
conditions in order to have a better understanding how the properties of CNF thin films are
113
effected by hydration and how these properties could impact future cell based studies. The
114
influence of surface roughness, porosity, and Young’s modulus on MSC adhesion and viability
115
will be reported.
116 117
MATERIALS AND METHODS
118
Materials
119
A 2 % wt suspension of enzymatically pre-treated CNF from wood pulp (Domsjo) was purchased
120
from Centre Technique du Papier (CTP) (France). Phosphate buffer solution (PBS), Minimum
121
Essential Medium Eagle (αMEM), 4,6-diaminidino-2-phenylindole-dilactate (DAPI),
122
phalloidintetramethylrhod B isothiocyanate dyes (Phalloidin), glycerol, mowiol 4-88, tris buffer
123
saline (TBS), thiazolyl blue tetrazolium bromide (MTT), and dimethyl sulfoxide (DMSO) were
124
purchased from Sigma-Aldrich (USA). BALB/c D1 murine bone marrow mesenchymal stem
125
cells were purchased from American Type Culture Collection (ATCC, France). 10% heat-
126
inactivated fetal bovine serum (FBS) was acquired from PAA Laboratories (Austria), and
ACS Paragon Plus Environment
Page 6 of 37
Page 7 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
127
penicillin streptomycin mixture and L-Glutamine were purchased from Invitrogen (USA). All
128
materials were used as delivered without any chemical modification.
129
Production of CNF Films
130
CNF was characterized to determine the cellulose content following the protocol of Rowell et
131
al.44, and zeta potential was measured a Malvern Zetasizer 2000 (Malvern, United Kingdom). 10
132
mL of approximately 0.001% suspension of CNF was measured in triplicate. Results are reported
133
in mean ± standard deviation. CNF membranes were produced by casting 1% wt solution of CNF
134
onto teflon mold and allowed to dry at ambient conditions for 5 days (AC-CNF) followed by
135
curing for 1 (1hC-CNF) or 2 (2hC-CNF) hours at 150°C following a modified protocol by Bardet
136
et al. 29, 31. Basis weight was determined by the following equation:
137
𝑔
𝑀𝑎𝑠𝑠 𝑜𝑓 𝐶𝑁𝐹 𝑓𝑖𝑙𝑚
𝐵𝑎𝑠𝑖𝑠 𝑊𝑒𝑖𝑔ℎ𝑡 (𝑚2 ) = 𝐴𝑟𝑒𝑎 𝑜𝑓 𝐶𝑁𝐹 𝑓𝑖𝑙𝑚 (Eq. 1)
138
Atomic Force Microscopy (AFM) Imaging
139
Images of the surface of CNF films were acquired using tapping mode with a Nanoscope IIA
140
microscope from Vecco Instruments (USA) at a frequency of ~265-340 Hz. Silicon cantilevers
141
with a curvature of 10-15 nm were used (OTESPA®, Bruker, USA). For surface roughness
142
measurements 10x10 µm images were acquired to analyze the surface roughness. All
143
measurements were analyzed using Nanoscope Analysis (USA). ImageJ software (NIH) was used
144
to analyze materials dimensions and at least 50 measurements have been performed on at least 5
145
different images and averages are presented.
146
Scanning Electron Microscopy (SEM) Imaging
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
147
SEM images of the CNF films for thickness measurements were obtained using a Quanta200®
148
(Netherlands). Samples were attached to a holder, coated with a thin layer of Au/Pb to create a
149
conductive charge, and scanned using an accelerated voltage of 10 kV and a working distance of
150
10.3 mm. An Everhart Thornley-Secondary electron detector (ETD) was employed. For thickness
151
measurements at least 25 measurements were taken from each sample using ImageJ software
152
(U.S National Institutes of Health, USA).
153
Thermoporometry
154
Calorimetric thermoporometry was performed following the protocol developed by Driemeier et
155
al. 45. Thermoporometry employed a differential scanning calorimeter DSC-Q200 (TA
156
Instruments, USA) with a RCS90 (TA Instruments, USA) cooling unit and an autosampler. CNF
157
samples were pre-conditioned by soaking in deionized water (DI H2O) for 24 hours. The water-
158
saturated samples were inserted into aluminum Tzero® pans later sealed with hermetic lids. The
159
DSC temperature program freezes samples to -70°C followed by 18 heating steps, each one made
160
of a heating ramp (1°C/min) followed by an equilibrating isotherm. Measured calorimetric signal
161
is analyzed to quantify the amount of ice melting below 0°C, termed Freezing Bound Water
162
(FBW). The temperature depression ΔT of ice melting is related to pore diameter d through the
163
Gibbs-Thomson equation, d = -2 Kc/ΔT, with Kc = 19.8 nm/K. The reported thermoporometry
164
profiles are the cumulative distributions of FBW (given in units of g water per g dry matter) as
165
function of d in the 1-200 nm range of pore size.
166
Mechanical Tests
167
Tensile tests and dynamic mechanical analysis (DMA) were measured using a TA Instruments
168
RSA 3 (USA) with a submersion clamp system designed internally. CNF films were tested in
ACS Paragon Plus Environment
Page 8 of 37
Page 9 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
169
physiological-like conditions, i.e. in aqueous environment. For physiological-like conditions,
170
CNF films were conditioned for 24 hours in PBS prior to mechanical tests. Samples underwent
171
rectangular tension measurements and compression testing. The distance between the clamps for
172
the tension was 10 mm, and the films had a width of approximately 5 mm and thickness of 0.2
173
mm. Samples underwent rectangular tension measurements at a rate of 0.001 mm/s at 37°C. For
174
compression, films were tested with an angular frequency from 0-20 Hz at 37°C with a constant
175
strain of 10%. Samples had dimensions of 15 mm in diameter and thickness of approximately 0.2
176
mm. The stress vs strain and frequency dependent curves were measured and analyzed using TA
177
Orchestrator (USA). Results were done in triplicate.
178
Cell Culture
179
Cell culture studies were performed with BALB/c D1 murine bone marrow mesenchymal stem
180
cells (D1 MSCs) (American Type Culture Collection, ATCC, France). Cells were cultured in
181
αMEM supplemented with 10% heat-inactivated fetal bovine serum, 1% penicillin streptomycin
182
mixture and 2 mM L-Glutamine. CNF films were sterilized with UV exposure prior to cell
183
seeding and rinsed twice with DI H2O. Cells were seeded at a density of 20,000 cells/cm2 for 6
184
hours and 24 hours for DAPI-phalloidin assay, and 24 hours and 48 hours for MTT analysis in a
185
humidified atmosphere of 5% CO2 at 37 °C.
186
DAPI-Phalloidin Assay
187
4,6-Diaminidino-2-phenylindole-dilactate (DAPI, 20 mg/mL, Sigma-Aldrich, USA) and
188
phalloidintetramethylrhod B isothiocyanate dyes (phalloidin, 10 mg/mL, Sigma-Aldrich,USA)
189
were used to determine the adhesion of cells at 6 hours and 24 hours following seeding on CNF
190
substrates and a glass control. At each time point, the culture media was moved and the samples
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 37
191
were fixed with 3.7% paraformaldeyde. The paraformaldehyde was removed after 20 min at room
192
temperature and washed twice with PBS. The phalloidin/DAPI solution was added for 45 min at
193
room temperature (12.5 µg/mL phalloidin, 10 µg/mLDAPI). The samples were then washed three
194
times with PBS, mounted with anti-fade reagent, which is 24% glycerol, 9.6% mowiol 4-88 in
195
tris buffered saline (TBS). Followed by analysis by inverted fluorescent microscopy (Zeiss,
196
Germany) or confocal laser scanning microscopy (Zeiss, Germany) 24 hours later. The cell
197
surface area and aspect ratio were analyzed using ImageJ with a sample size of 150 cells per
198
condition using at least 12 different microscopy images from 4 samples from two independent
199
experiments. Cell surface area and aspect ratio measurements were done by measuring clearly
200
defined individual cells using the tracing tool in ImageJ after calibrating the scale and
201
thresholding each image.
202
MTT Assay
203
Thiazolyl blue tetrazolium bromide (MTT, 2.5 mg/mL, Sigma Aldrich, USA) was used to
204
perform a MTT assay on five independent experiments to measure the mitochondrial activity of
205
D1 MSCs after 24 hours and 48 hours post seeding on CNF substrates or tissue culture polysterne
206
(TCPS) control. MTT solution was added to the cells and incubated in media for 1 hour. After 1
207
hour, the growth media was removed and 500 µL of dimethyl sulfoxide (DMSO) was added to
208
solubilize the formazan crystals formed in the reduction of MTT. 100 µL of the formazan
209
solution was added to a microplate reader and diluted with 100 µL of DMSO. The absorbance of
210
each sample was read at 550 nm. The samples were imaged using optical microscopy (Ziess,
211
Germany).
212
Statistical Analysis
ACS Paragon Plus Environment
Page 11 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
213
All experiments were carried out in triplicate unless otherwise noted. All results are presented
214
with mean ± standard deviation unless otherwise noted. The R2 value of thermoporosity vs.
215
Young’s modulus was performed by Originlab using a linear-regression test. Statistical analysis
216
of cell studies (DAPI-Phalloidin Assay/MTT Assay) were performed in Originlab using a
217
Kolmogorov-Smirnov test for normality. For the cell area and aspect ratio results following the
218
DAPI-Phalloidin assay, the samples did not follow a normal distribution and were normalized
219
using a Box-Cox transformation. Following the normalization transformation, a Kolmogorov-
220
Smirnov test was repeated. Normalized results were analyzed by two-way ANOVA with post-hoc
221
testing using a Bonferroni test.
222 223
RESULTS AND DISCUSSION
224 225
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
226
Figure 1. Schematic of the experimental set-up. From raw material of CNF suspension to CNF
227
films and sequential curing of films and eventual cellular testing of CNF films for cell attachment
228
and viability
229 230
Proof of Concept - CNF to Cells
231
CNF used in this study had a diameter of 14 ± 5 nm as reported by Bardet et al.29, a cellulose
232
content of 86%, which is in accordance with literature 25, and a zeta potential of -17.0 ± 0.7 mV.
233
To determine the optimal properties of the CNF films before cell adhesion and viability tests,
234
several CNF films with varying basis weight was produced and either left as-castor dried at
235
150°C for 2 hours. Based on literature29, such curing treatment modifies the structures of this
236
entangled network of CNF. Others called it densification by hornification due to the increase in
237
number of strong hydrogen bonds and by the elimination of water. The mechanical properties of
238
these films were tested using traction testing and compression testing in aqueous conditions, i.e.
239
cell culture like conditions, 37°C submerged in PBS. The results from this work are presented in
240
Figure 2. As shown in the table, there are different mechanical properties of the films depending
241
on the basis weight and the Young’s modulus varies in regards to curing the films. The general
242
trend that with curing the films had a higher modulus compared to the as casted films.
ACS Paragon Plus Environment
Page 12 of 37
Page 13 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
243 244
Figure 2. A) Basis weight (g/m2), elongation at break (%), Young’s modulus (MPa), and storage
245
modulus (MPa) of CNF films as-cast(AC) or after 2h curing (2hC) following mechanical testing
246
at physiological conditions at 37°C in PBS. D1 MSCs interactions with the CNF with different
247
basis weight were assessed after 24 h of culture by staining the cell nucleus and cytoskeleton
248
(DAPI and phalloidin). B)1-AC-CNF, C) 2-2hC-CNF, D) 3-2hC-CNF, and E) 4-2hC-CNF as
249
measured by confocal laser scanning microscopy.
250 251
It is worth noting that with similar raw material, a significant difference in mechanical properties
252
(E’ in PBS is between 6 and 90 MPa) can result. This kind of ability to vary the properties is not
253
so common, even within the CNF field, with only few scientific papers dealing with such an
254
approach29, 31.
255
It is already recognized that mechanical values are crucial for cell growth and differences might
256
be interesting for stem cell differentiation39 which lead to the preliminary tests to determine if it
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
257
was possible to culture D1 MSCs with these different CNF substrates. D1 MSCs were cultured
258
for 24 hours, stained with Phalloidin/DAPI and observed using confocal microscopy as shown in
259
Figure 2. Thinner films allowed imaging and showed cell attachment for the high mechanical
260
property substrates, but low attachment for the as-cast films with the same fiber amount. This
261
result is promising because of the apparent attachment of the cells to the CNF, but it is difficult to
262
say if this qualitative difference is due to mechanical properties or to other substrate properties.
263 264
Tuning the Nanostructure Organization of CNF Substrate
265
Starting from this proof of concept, new CNF films were produced with same mass of fiber, but
266
the drying time was varied for tuning the intrinsic properties of such CNF films. The table below
267
summarizes the properties of such substrates.
ACS Paragon Plus Environment
Page 14 of 37
Page 15 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Biomacromolecules
1
Sample
Transmittance Surface Freezing Bound Elongation Water Content PBS Uptake Basis Weight (%) 300-800 Roughness Water at < at Break 3 4 2 5 (%) (w/w) (%) (w/w%) (g/m ) nm1 (RMS) (nm)2 200 nm (g/g)6 (%)7
Young’s Modulus (E) (MPa)7
Storage Modulus (E’) (MPa) from 0.05-20 Hz8
0.022± 0.003 – AC-CNF
4.0-10.1
4.71±1.55
13.1±4.0
103.5±2.8
38.6±1.1
0.44± 0.006 4.2 ± 1.3
5.6 ± 0.6 0.305±0.120
1hC-CNF
3.8 -11.5
4.71±1.53
6.6±1.8
70.6±9.5
29.5±3.1
0.38± 0.02
52.2 ±
0.026 ± 0.011 –
26.3
0.329 ± 0.085
3.4 ± 1.3
0.050± 0.028 – 2hC-CNF
2.0 -8.2
4.34±0.48
3.4±1.4
65.3±3.7
28.1±2.8
0.36± 0.02
2.5 ± 0.6
116 ± 29.1 0.342 ± 0.194
2
Table 1. Summary of CNF films properties as measured by 1) UV spectroscopy (S1), 2) Root Mean Squared Roughness values from
3
AFM images of CNF films, as measured by nanoscope analysis AFM, 3) drying of films at 105°C for 24 hours (S2), 4) submersion of
4
films for 24 hours in PBS at 37°C (S2), 5) Basis Weight of CNF Films, 6) thermoporometry, 7) tensile testing at 37°C in PBS, 8)
5
compression DMA at 37°C in PBS.
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
6
The transmittance of the CNF films was measured using UV spectroscopy with a film holder. As
7
shown in Table 1, there is no clear relationship between the drying time and transmittance of the
8
CNF films. Further images of the films and transmittance graphs are reported in the supporting
9
information (S1). The transmittance of the films is an important measurement in regards to
Page 16 of 37
10
imaging and cell staining; if the transmittance of the films is too low, it is difficult to acquire
11
fluorescent microscopy images and reproducibility of measurement might become a concern.
12
Even though there are slight differences in the transmittance of the samples, the transmittance is
13
high enough to allow for staining and microscope imaging following cell culture.
14
The structural characteristics of the CNF substrates were measured using AFM, SEM, water
15
content, PBS uptake, and theromoporometry. It was important to analyze the structure of the CNF
16
substrates to investigate if there is any relationship between structural changes caused by curing
17
at high temperature and the mechanical properties of these films.
18
Atomic force microscopy was performed on the CNF nanopaper to characterize their surface
19
roughness. Figure 3 shows the AFM images of the film surface acquired following drying and
20
curing. The reported roughness values from root-mean-square analysis of AFM images of AC-
21
CNF, 1hC-CNF, and 2hC-CNF (Table 1.) shows that for micro scale roughness there is no
22
significant difference between the roughness. Compared to glass, which was used as a control in
23
the cell aspect ratio and surface area measurements, CNF has higher surface roughness (~4.7
24
nm), while glass is about 0.5 nm as reported by Domke et al. with the same scan size by AFM 46.
ACS Paragon Plus Environment
Page 17 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
25 26 27
Figure 3. Atomic force microscopy images (10x10 µm) of the surface of the CNF films A) AC-CNF, B) 1hC-CNF, and C) 2hC-CNF taken by tapping mode in air.
28 29
Scanning electron microscopy was performed to determine the thickness and image the porosity
30
of each CNF film as shown in Figure 4. As shown in the figure, the thicknesses of the CNF
31
substrates are similar, but there is a noticeable difference in porosity within the samples, most
32
notably porosity in micrometric scale that can be visualized in the SEM images.
33
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
34
Figure 4. SEM images of the thickness with different magnifications A,B,C) 600x and D,E,F)
35
5000x A,D) AC-CNF, B,E) 1hC-CNF, and C,F) 2hC-CNF with different magnifications and G)
36
measurements of thickness using Image J (mean + SD of 35 membrane section).
Page 18 of 37
37 38
These differences in layer formation could influence the cell adhesion and viability on CNF. The
39
direct nanoscale porosities of the CNF films were measured using thermoporometry, which will
40
be reported later, and the CNF images supports those results with pore size and number
41
decreasing with increased curing time at 150°C. AC-CNF and 1hC-CNF have a more porous
42
structure, while 2hC-CNF is denser with less pores, but still a similar thickness. The direct
43
measurement of this qualitative observation will be discussed later.
44
The water content of the CNF films was determined following conditioning and evaporation of
45
the films at 105°C for 24 hours as outlined in the methods section. This method is commonly
46
used to determine the water content within paper pulp and was adapted for CNF films. As shown
47
in Table 1 curing of CNF films at 150°C for 1 or 2 hours lessens the water content within the film
48
compared to casted CNF (AC-CNF), with maximum loss experienced by 2hC-CNF. This loss of
49
water causes a change in the porosity and the basis weight of the CNF films but have not
50
influenced mechanical properties of the CNF films which are obtained in an aqueous state. PBS
51
absorption of CNF films is dependent on the morphology and structure of the film. Also since the
52
CNF will be tested in liquid conditions and culturing requires the films to be exposed to media,
53
the differences in PBS uptake between the samples can change cell attachment and viability.
54
Since 2hC-CNF has lower porosity compared to AC-CNF and 1hC-CNF the PBS solution cannot
55
easily infiltrate the film and cause swelling. Water content and PBS uptake results are reported in
56
the supporting information (S2)
57
ACS Paragon Plus Environment
Page 19 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
58 59
Nanoscale porosity of the CNF films in aqueous conditions were determined using
60
thermoporometry. This technique has been employed to characterize cellulosic fibers.47,48 The
61
method was recently improved by Driemeier et al. 45 and is here for the first time used to analyze
62
the nanoscale porosity of CNF films. As expected, with curing at 150°C for either 1 or 2 hours
63
the nanoscale porosity of the films decreased compared to AC-CNF, and further decreased with
64
increased curing time (AC-CNF > 1hC-CNF > 2hC-CNF). The observed loss of porosity is
65
distributed in the measured 1-200 nm pore size range, without special behavior in any specific
66
window of pore size. The maxima of the profiles (FBW read at 200 nm in Figure 5) is also
67
reported in Table 1. The changes in porosity are caused by water evaporation during curing,
68
which causes a decrease in the basis weight of the CNF films. Noteworthy, the changes are
69
irreversible, since the AC-CNF porosity does not recover after 1hC-CNF and 2hC-CNF rehydrate
70
in aqueous conditions. The measured reduction of nanoscale porosity (Figure 5) manifests the
71
cohesive action of drying, which likely impacts other properties, such as CNF film mechanics
72
and PBS uptake, as well as cell growth and viability.
73
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
74
Figure 5. Cumulative distribution of Freezing Bound Water (FBW) (g/g) vs. pore diameter (nm)
75
as measured by thermoporometry for CNF films. For 1hC-CNF and 2hC-CNF, respectively the
76
low and the high branches of the symmetric error bars are omitted for better visualization. With
77
curing the amount of FBW in pores between 1-200 nm decreases.
78 79
In order to attribute the changes in porosity and mechanical properties of the film only to the
80
evaporation of water within the CNF alone, TGA measurements were conducted from 30°C until
81
900°C. Changes in residual mass percentage above 150°C for the samples would indicate a
82
chemical modification of the CNF. As shown in the thermograms, which are reported in the
83
supporting information (S3), there are no differences. This lack of change or shift in the
84
thermograms indicates that the curing of CNF does not produce any chemical change. Therefore,
85
it can be noted that the mechanical properties and porosity changes are only dependent on the
86
curing of the CNF with no chemical modification of the nanocellulose.
87 88
Tuning Mechanical Properties of CNF Substrate
89
Mechanical properties of the CNF films, as well as the viscoelastic properties were determined
90
using tensile tests and DMA with a submersion clamp system. DMA is commonly used to
91
determine if a material used for biomedical applications is suitable, while applying conditions
92
that mimic a physiological environment49. Tensile tests and compression testing were conducted
93
in a reservoir filled with PBS at 37°C to assess the mechanical properties in cell culture
94
conditions. The stress-strain curve and the storage modulus (E’) for the films were measured in
95
hydrated conditions. The stress-strain curves show a significant increase in the Young’s modulus
96
(E) with increase in drying time, and a decrease in elongation at break. These variations in values
ACS Paragon Plus Environment
Page 20 of 37
Page 21 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biomacromolecules
97
show that the samples become more brittle following drying at elevated temperature. The storage
98
modulus was measured using compression from 0.05-20 Hz, which mimics physiological
99
conditions. As shown in Table 1, there was an increase in the storage modulus from 0.05 Hz to 20
100
Hz from AC-CNF to 2hC-CNF with a 56% increase in E’ at 0.05 Hz and 10% increase at 20 Hz
101
for AC-CNF vs. 2hC-CNF. The change in mechanical properties can be attributed to the new
102
nanostructures (described previously) of the CNF substrate since the fiber content and
103
preparation of the films were the same. The only differences occurring in the same trend of
104
increasing mechanical properties is the decrease of porosity with increased drying time, which
105
implies that mechanical performance of CNF films and porosity are related with an R2 value of
106
0.88.
107 108
Figure 6. A) Stress-Strain curves for CNF films tested in phosphate buffer solution at 37°C. The
109
films were either tested as casted (AC-CNF) or after curing (1hC-CNF,2hC-CNF) with (N=3).
110
ACS Paragon Plus Environment
Biomacromolecules
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
111
The results of mechanical testing in conjunction with CNF film structure characterization (i.e.
112
surface roughness, thickness, porosity) shows that there is a direct relationship between the
113
drying of films for extended periods of time and the material properties. Such results are
114
innovative in regards to CNF substrate material properties, and this tenability has not been
115
studied in details in other scientific paper up to our knowledge. These properties can be tuned
116
depending on the amount of drying time, without any further chemical modification.
117 118
Influence of Tunable CNF substrate on Cell Adhesion and Morphology
119
Since cell adherence and spreading is the first essential step to study the biocompatibility of a
120
material, our next step was to study the adhesion and morphology of D1 MSCs on the tunable
121
CNF substrates. The cell morphology, surface area, and aspect ratio were determined by DAPI-
122
Phalloidin staining of D1 MSCs at 6 h and 24 h following cell seeding. Cells cultured on CNF
123
with different curing conditions displayed different morphologies (Figure 7). Independently the
124
culture time, D1 MSCs seeded on CNF show more elongation and cytoplasmic extension when
125
compared to when seeded on glass (Figure 7 A-H). While cell area at 6 h was not affected by
126
culturing D1 MSCs on any CNF, at 24h only 1hC-CNF and 2hC-CNF showed a higher value
127
when compared to glass (+33%, +28% p
