Degradation rate affords a dynamic cue to regulate stem cells beyond varied matrix stiffness

Accepted Manuscript Degradation rate affords a dynamic cue to regulate stem cells beyond varied matrix stiffness Yuanmeng Peng, Qiongjie Liu, Tianlei He, Kai Ye, Xiang Yao, Jiandong Ding PII:

S0142-9612(18)30277-1

DOI:

10.1016/j.biomaterials.2018.04.021

Reference:

JBMT 18607

To appear in:

Biomaterials

Received Date: 17 January 2018 Revised Date:

31 March 2018

Accepted Date: 11 April 2018

Please cite this article as: Peng Y, Liu Q, He T, Ye K, Yao X, Ding J, Degradation rate affords a dynamic cue to regulate stem cells beyond varied matrix stiffness, Biomaterials (2018), doi: 10.1016/ j.biomaterials.2018.04.021. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Degradation rate affords a dynamic cue to regulate stem cells beyond

RI PT

varied matrix stiffness

Yuanmeng Peng, Qiongjie Liu, Tianlei He, Kai Ye, Xiang Yao, and Jiandong Ding* State Key Laboratory of Molecular Engineering of Polymers, Department of

M AN U

SC

Macromolecular Science, Fudan University, Shanghai 200433, CHINA

*

TE D

Correspondence should be addressed to J.D. DING. Email: [email protected]

Keywords: Degradation rate; dynamic cue; stem cell differentiation; surface patterning;

AC C

EP

PEG hydrogel

ACCEPTED MANUSCRIPT Abstract While various static cues such as matrix stiffness have been known to regulate stem cell differentiation, it is unclear whether or not dynamic cues such as degradation rate along with the change of material chemistry can influence cell behaviors beyond

RI PT

simple integration of static cues such as decreased matrix stiffness. The present research is aimed at examining effects of degradation rates on adhesion and differentiation of mesenchymal stem cells (MSCs) in vitro on well-defined synthetic hydrogel surfaces. Therefore, we synthesized macromers by extending both ends of

SC

poly(ethylene glycol) (PEG) with oligo(lactic acid) and then acryloyl, and the corresponding hydrogels that were obtained after photopolymerization of the

M AN U

macromers were biodegradable. Combining the unique techniques of block copolymer micelle nanolithography with transfer lithography, we prepared a nanoarray of cell-adhesive arginine-glycine-aspartate peptides on this nonfouling biodegradable hydrogel. The biodegradation is caused by hydrolysis of the ester bonds, and different degradation rates in the cell culture medium were achieved by different stages of

TE D

accelerated pre-hydrolysis in an acidic medium. For the following cell culture and induction, both the matrix stiffness and degradation rate varied among the examined groups. While adipogenic differentiation of MSCs can be understood by the lowered

EP

stiffness, the osteogenic differentiation was contradictory with common sense because we found enhanced osteogenesis on soft hydrogels. Higher degradation rates were suggested

to

account

for

this

interesting

phenomenon

in

the

sole

AC C

osteogenic/adipogenic induction and even more complicated trends in the co-induction. Hence, the degradation rate is a dynamic cue influencing cell behaviors, which should be paid attention to for degradable biomaterials.

ACCEPTED MANUSCRIPT 1. Introductions. Cell-material interactions are fundamental issues for the development of next-generation biomaterials [1-4]. Various chemical, physical and biological cues can influence cells [5-9]. For instance, functional groups [10-12] and surface topography [9] are known to regulate cell adhesion and differentiation; selective decoration with

RI PT

cell-adhesive ligands to restrict cell spreading area [13, 14] and using material techniques to control cell shape [15-19] can well adjust cell behaviors. Besides, cell cues like cell-cell contact [20, 21] and cell density [22] have also been found to tune

SC

lineage commitment of stem cells. A particularly important biomechanical cue revealed in the last decade is matrix stiffness [23-25]. To date, it has become common

M AN U

knowledge that a soft material favors adipogenic differentiation of mesenchymal stem cells (MSCs) and a stiff matrix favors osteogenic differentiation, especially in two-dimensional (2D) systems [26, 27].

Most of the known cues are static parameters. In a biomaterial-based system, these parameters are inevitably varied when material chemistry changes during cell

TE D

culture or tissue regeneration [28, 29]. It is critical to understand whether or not a change of one parameter can be understood simply from summation of a series of static cues, that is to say, whether or not the change rate of that quantity affords an additional parameter as a dynamic cue. Biodegradability becomes a key factor when

EP

designing new biomedical materials for potential applications in tissue engineering and tissue regeneration. Degradation of biomaterials has inspired much research

AC C

[30-35]. It is particularly meaningful to study cell-material interactions in a biodegradable microenvironment [36-41]. Degradation rate might be a candidate for a dynamic cue that influences cell

behaviors. Recently, some valuable efforts have been made to examine the effects of biodegradation on cells encapsulated in three-dimensional (3D) hydrogels. Anseth et al. [42] introduced a photodegradable hydrogel based on a poly(ethylene glycol) (PEG) derivative containing nitrobenzyl ether photolabile moieties, and found that human MSCs spread much better in hydrogels upon light exposure (and thus degraded) than in the control group that was not exposed to light. Burdick et al. [43] prepared

ACCEPTED MANUSCRIPT permissive/restricted hydrogel networks through distinct crosslinking protocols, and introduced matrix metalloproteinase (MMP)-sensitive peptides as degradable crosslinkers of hyaluronic acid (HA); they found that human MSCs preferred spreading and favored osteogenesis in permissive systems which were degraded by

RI PT

cells; when a later crosslinking changed the permissive system to a restricted one, they observed enhanced adipogenesis. Bian et al.[44] extended this study and further found that cell-mediated degradation regulates chondrogenesis of human MSCs and hypertrophy in MMP-sensitive HA hydrogels, independently of cell spreading. Some

SC

other dynamic material-related cues were also studied recently. For instance, Mooney et al. [45, 46] revealed that stress relaxation of alginate hydrogels enabled the

M AN U

encapsulated cells to remold the matrix and cluster the cell-adhesive ligands, resulting in more spreading.

While these elegant studies have revealed the roles of biodegradation on cells comparing biodegradable versus non-biodegradable or permissive versus restricted material systems, a study of the effects inflicted by the degradation rate on stem cells

TE D

would be more attractive. This kind of study should be based on well-controlled hydrogels with a series of degradation profiles in a cell culture medium. Herein, we pay attention to the effects of rates of biodegradation on adhesion and differentiation

EP

of stem cells. The present report is dealing with 2D cell behaviors based on well-designed nanopatterned surfaces of synthetic hydrogels of varied degradation rates. The nanopatterned surfaces are powerful 2D model material systems for research.

AC C

fundamental

Our

test

system

comprises

a

cell-adhesive

arginine-glycine-aspartate (RGD) nanoarray on an adhesion-resistant biodegradable hydrogel with nonfouling PEG as the major component and a hydrolysis-labile oligoester motif as the minor component. The idea is schematically shown in Fig. 1.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 1. Schematic summary of the fabrication process of an RGD nanoarray on a PEG-based yet hydrolysis-labile hydrogel and the underlying scientific question about cell-material interactions to be examined. (A) Synthesis routes of oligoLA-PEG-oligoLA macromers. In the first step, PEG with two free hydroxyl end groups triggered the ring-opening polymerization of D,L-lactide induced by catalytic

ACCEPTED MANUSCRIPT amounts of stannous octoate Sn(Oct)2; then, oligoLA-PEG-oligoLA was terminated with acryloyl chloride, using triethylamine (TEA) as the base. (B) Fabrication process of an RGD-grafted gold nanoarray on a hydrogel synthesized by polymerization of the above described macromers. The nanoarray was first formed on glass through block

RI PT

copolymer micelle nanolithography and then transferred to the synthesized hydrogel; finally, cyclic RGD with a thiol end group was grafted on each gold nanodot to form an RGD-grafted nanopattern on the biodegradable hydrogel. (C) Biodegradation of the hydrogel via hydrolysis of the oligoester segments. (D) Schematic illustration of

SC

the cell experiments and the corresponding key scientific question.

M AN U

The hydrogels that we utilized in this study were PEG-based derivatives. PEG hydrogels have been widely used in biomedical fields for their high fidelity towards extracellular matrices (ECMs) and their excellent biocompatibility [47-49]. Introducing hydrolysis-labile or enzyme-sensitive segments makes PEG-based materials degradable [50-52]. In this study, we used the block copolymer

TE D

PLA-b-PEG-b-PLA as a PEG derivative that combines the advantages of PEG and poly(lactic acid) (PLA) [48, 53]. Its admirable properties have rendered it a good candidate as a tissue engineering scaffold [54, 55] and for drug delivery [56-58]. PLA,

EP

a polyester with hydrolytically labile chemical bonds, has been generally considered to be a biodegradable material that mainly undergoes bulk erosion [59]. Incorporating oligomers of lactic acid (LA) as extension of PEG allows for tunable degradability by

AC C

changing the number of the LA units and the crosslinking density of the hydrogels [60-62]. Furthermore, introducing alkenes at both ends of the oligoLA-PEG-oligoLA chain through acryloyl functionalization can make the diacrylate (DA) macromer chemically cross-linkable. Thus, a hydrogel with tunable degradability can be achieved, as schematically indicated in Fig. 1C. Since PEG is chemically relatively inert, it is very difficult to generate a stable array of other chemicals on the surface of a PEG hydrogel. Nevertheless, we have combined block copolymer micelle nanolithography and transfer lithography to develop a unique patterning technique to generate a gold array on a PEG hydrogel [63,

ACCEPTED MANUSCRIPT 64]. In the present study, we first constructed a quasi-hexagonal nanoarray of gold on glass via self-assembly of micelles of block copolymers loaded with chloroauric acid upon dip-coating followed by plasma treatment. Then, a bifunctional linker with one thiol (-SH) end and one acryloyl end was grafted on the gold nanodots and joined in

RI PT

the photopolymerization of the macromers with acryloyl ends. Prior to cell experiments, RGD peptides were grafted on the gold nanodots, providing cell-adhesive sites.

It seems worthy to note that while the synthesis of the hydrolysis-labile

SC

macromers to prepare biodegradable hydrogels and the transfer strategy to prepare RGD nanoarrays on PEG hydrogels have been known [60, 63], the model material of

M AN U

RGD-nanoarray-decorated biodegradable hydrogels examined in the present study has never been reported. Most importantly, this article examines the cell behaviors on surfaces of biodegradable synthetic materials with a series of well-controlled degradation rates. In this context we used an acidic medium to speed up hydrolysis, and by controlling the accelerated hydrolysis stages we found that the obtained

TE D

samples exhibited different degradation rates in the cell culture medium, which enabled us to examine the potential effect of degradation rates on cell behaviors in a parallel way. The change of other parameters such as matrix stiffness should, of

experiments.

EP

course, also be taken into consideration to fully understand the results of the cell

MSCs from the bone marrow of Sprague Dawley (SD) rats were employed as the

AC C

cell model. Osteogenic and adipogenic induction were carried out in vitro. Both cell adhesion and lineage commitments were recorded on the surfaces of the hydrogels with well-controlled degradation rates leading to some interesting results.

2. Materials and methods 2.1 Synthesis and characterization of the biodegradable macromer The synthesis route of the macromers is showed in Fig. 1A. Before introducing the chemically crosslinkable alkenes into the PEG chains, oligoLA-PEG-oligoLA was synthesized via ring-opening polymerization of D,L-lactide (Zhejiang Medzone). PEG

ACCEPTED MANUSCRIPT with a number average molecular weight (Mn) of 2000 (Sigma) with hydroxyl (–OH) at both ends initiated the reaction, and stannous octoate (Sigma) was used as the catalyst. PEG was added into a three-neck flask and dried under vacuum at 120oC for 3 h. Then the reaction system was cooled down to 80°C, D,L-lactide and Sn(Oct)2

RI PT

were added to molten PEG. The reaction was allowed to proceed at 140°C under stirring. Twelve hours later, the acquired product was cooled to room temperature, diluted by chloroform (CHCl3) and purified by precipitation from diethyl ether. After filtration and lyophilization, the dry solid polymer was stored at -20°C for the next

SC

step.

The cross-linkable macromer was obtained via the nucleophilic substitution

M AN U

reaction of oligoLA-PEG-oligoLA and acryloyl chloride (Tokyo Chemical Industry). In order to avoid moisture, the reaction was carried out under argon using standard Schlenk techniques; dichloromethane (DCM) and TEA were obtained from commercial sources and were purified according to standard procedures before use. OligoLA-PEG-oligoLA was dissolved in DCM, and TEA was added at 0°C under

TE D

argon. After stirring in an ice bath for 30 min, the mixture was treated with acryloyl chloride through a quantitative injection pump with a constant addition rate. During the injection process, the reaction system was kept in an ice bath until the acryloyl

EP

chloride was completely added. The reaction was firstly run in an ice bath for 6 h, and stirring was continued at room temperature for 18 h. The product mixture was filtered through a diatomite column to remove the precipitated salts, and then eluted with

AC C

ethyl acetate (EtOAc). The filtrate was concentrated and gave a yellow mixture. The residue was then dissolved in a small amount of DCM and diluted by EtOAc to separate the residual salts. After filtering through a diatomite column, the filtrate was added

to

ice-cold

diethyl

ether

for

precipitation.

Finally,

acrylated

oligoLA-PEG-oligoLA was gained through filtration and lyophilization, and the white powder was stored at –20°C. The feed ratio of each step are listed in Table S1. The structures of the macromers obtained from the two-step protocol were determined by 1H nuclear magnetic resonance (1H NMR) in CDCl3 at 400 MHz (Bruker, AVANCE III). The

ACCEPTED MANUSCRIPT scan number of each sample was 16. The NMR spectra are shown in Fig. S1. 2.2 Photopolymerization of diacrylated oligoLA-PEG-oligoLA to form chemical hydrogels Diacrylated oligoLA-PEG-oligoLA powder was dissolved in Milli-Q water at

After

stirring

for

12

h,

0.1

2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone

RI PT

room temperature to prepare a macromer solution with a concentration of 45 % (w/w). wt%

(D2959,

photoinitiator

Sigma-Aldrich)

was added into the macromer solution. The mixture was stirred for 0.5 h. The whole

SC

process described above was run in the absence of light.

Photopolymerization was conducted under ultraviolet (UV) radiation (365 nm).

M AN U

The prepared solution was pipetted to tunnel in a mould which was covered by glass slides. The depth of each tunnel was 0.7 mm, and each mould was filled with around 500 µL of solution. Then the samples were submitted to 7 mW/cm2 UV irradiation under nitrogen atmosphere for 1 h. Finally, hydrogels with a thickness of approximately 0.7 mm were gained by peeling off from the glass.

hydrogels

TE D

2.3 Measurements of compressive moduli and swelling ratios of the biodegradable

Hydrogel slices (26 mm × 22 mm with thickness of 0.7 mm) obtained from

EP

photopolymerization were soaked in phosphate buffered saline (PBS, pH 7.4, Hyclone) and then moved to a 37°C water bath shaker. When the slices reached a swelling balance, the initial linear swelling ratio (SRinitial) was measured and calculated from

AC C

the specific length of the hydrogels. SRinitial was defined as the specific length of the hydrogel measured after swelling in a large amount of an aqueous medium such as PBS divided by the length after hydrogel synthesis yet before swelling (l0, 26 mm). Degradation experiments were conducted immediately after swelling. Besides

the normal yet slow hydrogel degradation in PBS, we also carried out the accelerated degradation by immersing the hydrogels in 2 mol/L hydrochloric acid (HCl) at 37°C for varied time periods (0 h, 2 h, 4 h, 6 h, 7 h, 8 h, 9 h, 10 h, 11 h, 12 h). Samples taken out at different points in time were washed with PBS for 5 times until the final pH of the solution was equal to 7.4. After washing, the hydrogel samples were

ACCEPTED MANUSCRIPT immersed in PBS for immediate use. Considering the envisioned use in cell induction experiments, hydrogels were then immersed into a cell culture medium in a 37°C incubator to explore their degradability in cell culture conditions. The linear swelling ratio (SR) was obtained by measuring the length of the

SR =  ×   

RI PT

hydrogels at different degradation stages and calculated by the equation

Here,    was defined as the final length after degradation for a certain

SC

time divided by the swelling length before degradation.

The volume swelling ratio was measured to evaluate the bulk swelling of the

M AN U

degraded hydrogels. Samples being washed with PBS were weighed and put into a vacuum drying oven for at least 24 h to fully evaporate the water in the hydrogels. Then the dry hydrogels were weighed, and the volume swelling ratio (QV) was calculated by the equation

 =

ρ

TE D

 = 1 +

ρ

 

 − 1

Here, Qm denotes the mass swelling ratio of a hydrogel, m1 refers to the mass of a hydrogel swollen in PBS, m0 refers to the mass of the corresponding dry hydrogel, ρp

EP

represents the density of the dry polymer, and ρs represents the density of PBS. A stress-controlled rheometer (Kinexus Pro, Malvern, UK) was used to measure

AC C

the compressive moduli of the hydrogels. The upper parallel plate attached to the rheometer slowly moved down with velocity v = 5×10-3 mm/s. When it contacted the top surface of the hydrogel slice, the normal force increased with the gap narrowing down. The original force-gap curve was collected in real time and converted into a stress-strain curve through calculation, as presented in Fig. S2. Compressive moduli of the hydrogels were obtained from the stress-strain curve at a low range of strain (ε ≤ 5%). 2.4 Fabrication of an RGD nanoarray on the biodegradable hydrogel The fabrication process of the nanopattern is depicted in Fig. 1B. Before

ACCEPTED MANUSCRIPT obtaining the nanoarray on the hydrogels, we needed to prepare a nanoarray on glass. Glass slides (26 mm × 22 mm with thickness of 0.5 mm) were washed thoroughly with piranha solution (H2SO4:H2O2 = 3:1) and immersed in Milli-Q water before use. Block copolymer poly(styrene-b-2-vinyl pyridine) (PS-b-P2VP, Polymer Source) was

RI PT

dissolved in toluene to form a micelle solution. The hydrophilic component P2VP constituted the micelle cores while the hydrophobic part PS constituted the coronas. Hydrophilic hydrogen tetrachloroaurate(III) trihydrate (HAuCl4⋅3H2O, Alfa Aesar) was added to the polymer solution as an Au precursor, and migrated into the cores of

SC

PS-b-P2VP micelles. Subsequently, glass slides were dipped in the micelle solution and pulled up. Thus, a self-assembled monolayer of micelles was coated on the glass

M AN U

surface. After treatment with oxygen plasma, the polymer segments of the micelles were thoroughly removed from glass, and the loaded gold acid was reduced to gold. Finally, a quasi-hexagonal array of Au nanodots was obtained. The preparation parameters are demonstrated in Table S2.

To fabricate nanoarrays on the hydrogels, a transfer lithography strategy was

TE D

applied. Glass slides with nanopatterns were bathed in an ethanol solution (2 mM) of the linker N,N′-bis(acryloyl) cystamine (Sigma) at 35°C for 1 h and for further 12 h at room temperature to ensure complete grafting on Au nanodots through Au-S bonding.

EP

Then, we added a macromer aqueous solution and triggered photopolymerization. After peeling off the hydrogel, we obtained an array of gold nanodots on the biodegradable hydrogel.

AC C

Before the cell experiments, the cell-adhesive RGD sequence was introduced to

improve cell adhesion on the hydrogel. A cyclic peptide c(-RGDfK-)-thiol (f: D-phenylalanine, K: L-lysine; Synpeptide) was utilized to form an Au-S bond between

the

peptides

and

the

gold

nanodots.

The

gold-nanopatterned

oligoLA-PEG-oligoLA-DA hydrogels were soaked in 25 µM peptide solution at 4°C for 4 h. Excess peptides were removed by multiple rounds of PBS rinsing. Finally, the biodegradable hydrogel with an RGD nanoarray on the surface was obtained. 2.5 Observation of nanopatterns A transmission electron microscope (TEM, Tecnai G2 20 TWIN, FEI) was

ACCEPTED MANUSCRIPT employed to observe the dispersion and morphology of the micelles. The block copolymer solution was dropped on a copper grid and evaporated naturally. The accelerating voltage was set to 200 kV. The gold nanoarray on glass was observed with a field-emission scanning electron microscope (SEM, Ultra 55, Zeiss, Germany). The accelerating voltage was

RI PT

set at 1.5 kV, working distance at 3.5 mm, magnification at 8 × 104 and an aperture size of 30 µm. The secondary electron imaging mode was utilized. 2.6 Isolation and culturing of MSCs

SC

MSCs were obtained from 7-day-old SD rats. The femurs and tibias were isolated from muscle tissue of the rat´s hind legs. A syringe filled with low-glucose

M AN U

Dulbecco’s modified Eagle medium (DMEM, Gibco) was inserted into the bone marrow cavity to thoroughly flush out the marrow. Through centrifugation, primary cells were acquired and seeded in cell culture flasks. The cell growth medium consisted of 90% low-glucose DMEM with additional 10% fetal bovine serum (FBS, Gibco), 100 U/ml penicillin (Gibco), 100 µg/ml streptomycin (Gibco) and 2 mM

TE D

L-glutamine (Gibco). Cells grew and proliferated at 37°C in a humid condition with 5% CO2, and were passaged when the confluence of cells reached up to 70-80%. The second-passage cells were used in subsequent cell experiments.

EP

2.7 Seeding of stem cells on nanopatterned hydrogels The nanopatterned hydrogels were firstly fully swollen under wet environment and then punched into discs in a diameter of 20 mm to match the 12-well tissue

AC C

culture plates (TCPs). Before being seeded with cells, all hydrogel specimens were disinfected with a 75% ethanol solution for 30 min. Hydrogels were then washed with PBS for 6 times, each time for 30 min to completely replace the ethanol. MSCs were digested from cell culture flasks and seeded on hydrogels at a

density of 3000 cells per square centimeter. All the following cell experiments employed the same cell density. 2.8 Measurement of cell viability on nanopatterned hydrogels The hydrogels with nanopatterns were treated with 2 mol/L HCl and washed as mentioned in section 2.3. Viabilities of MSCs on the hydrogels obtained at different

ACCEPTED MANUSCRIPT degradation stages were quantitatively determined by Cell Counting Kit-8 (CCK-8, DOJINDO) assays. The cell viability referred to the activities of succinate dehydrogenase (SDH) on mitochondrial inner membranes of all cells in each sample. Cell viability on glass was chosen as a positive control. Wells with only DMEM and

RI PT

CCK-8 solution were selected as blanks to avoid background interference. After seeding the cells on surfaces for 1 day, the cell culture medium was replaced by a fresh medium with 10% CCK-8 solution for another incubation time of 2.5-3 h. Then the upper reacted medium was pipetted to 96-well TCPs. A microplate

SC

reader (Multiskan Mk 3) was used for characterizing solution absorbance of each well at 450 nm. The relative optical density (OD) was calculated to quantify the cell

M AN U

viability. The control group with the cells seeded on glass was defined as of 100% viability. The results of the experimental groups were calculated from OD values deducting the background absorbance values.

2.9 Immunofluorescence staining, observation and analysis of stained cells on nanopatterned biodegradable hydrogels

TE D

Filamentous actin (F-actin), vinculin and nuclei of cells were stained to observe cell adhesion on the hydrogels. After seeding of MSCs on the hydrogels in the cell growth medium for 1 day, all samples in 12-well TCPs were gently rinsed with PBS

EP

for once. Then samples were fixed by 4 % paraformaldehyde for 12 min and soaked in 0.1% Triton X-100 solution for 10 min to improve permeability of the cell membrane. Once a new reagent was used, samples needed to be rinsed with PBS in order to allow

AC C

the next reagent to function well. To label vinculin, specimens were firstly blocked with 5% bovine serum albumin (BSA) solution for 20 min to diminish nonspecific protein staining. Then a 1:100 dilution of vinculin primary antibody (2 µg/ml, Santa Cruz) was used to incubate the sample at 4°C overnight. After rinsing by PBS twice, the Alexa fluor 488-conjugated goat antimouse secondary antibody (Invitrogen) was added at a concentration of 10 µg/mL to conjugate with the primary antibody. The application conditions of the second antibody were at room temperature for 1 h. To visualize cytoskeleton and observe F-actin, samples were processed in 1 µg/mL phalloidin-tetramethyl rhodamine B isothiocyanate (phalloidin-TRITC, Sigma) for 30

ACCEPTED MANUSCRIPT min at room temperature. Subsequently, 2 µg/ml 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich) was added to the samples to stain the nuclei in a reaction time of 5 min. At last, all samples were rinsed with PBS for 6 times to fully remove the fluorescent reagents.

RI PT

The fluorescently stained cells were observed under a fluorescence microscope (Axiovert 200, Zeiss) and photographed by a charge coupled device (CCD, AxioCam HRC, Zeiss). In order to ensure quantitative analysis, the voltage of the excitation light source and exposure times of photographing were kept constant for all groups.

SC

To explore the cell adhesion on the surfaces at different biodegradation stages, single cells were chosen for comparison. For each fluorescence picture, cells were

M AN U

outlined using the software ImageJ to obtain cell spreading areas, circularity and aspect ratios. Photographs with stained nuclei were chosen to count the cells on the samples to obtain the cell density.

In order to semi-quantify the cytoskeleton tension, the integral intensity of F-actin per cell was obtained by calculation from the following equation

TE D

Intensity of F-actin = Scell × (< I > I0)

Here, I refers to the grey value of one cell in a pixel, and , to the mean grey value of the whole cell; I0 indicates the mean grey value of the background near the cell;

EP

Scell is the pixel number in a cell outlined with ImageJ. 2.10 Osteogenic, adipogenic and co-induction of MSCs on nanopatterned

AC C

biodegradable hydrogels

The second-passage MSCs were seeded on hydrogels that had undergone

different degrees of acid-promoted degradation. On the first day, cells were cultivated in the growth medium (low-glucose DMEM with 10% FBS) to guarantee their adhesion to the hydrogel surfaces. The growth medium was replaced by the induction medium in the following induction period. In terms of the osteogenic induction, cells were exposed to the osteogenic induction medium following a 1 d growth phase. The osteogenic medium contained high-glucose DMEM, 10% FBS, 50 µM ascorbic acid-2-phosphate (Sigma), 10 mM

ACCEPTED MANUSCRIPT β-glycerophosphate (Sigma) and 100 nM dexamethasone (Sigma). The medium was changed at the second, fourth and sixth days after cell seeding. For the adipogenic induction, cells were initially incubated in the adipogenic induction medium for 3 d, then in the maintenance medium for 2 d, and furthermore in the adipogenic induction medium for 2 d. The adipogenic induction medium

indomethacin

(Sigma),

10

µg/mL

insulin

RI PT

consisted of high-glucose DMEM, 10% FBS, 1 µM dexamethasone, 200 µM (Sigma)

and

0.5

mM

3-isobutyl-1-methylxanthine (Sigma). The adipogenic maintenance medium was

SC

composed of high-glucose DMEM in addition to 10% FBS and 10 µg/ml insulin.

As for the osteogenic and adipogenic co-induction, we used a 1:1 mixture of the

M AN U

osteogenic and adipogenic medium. The change of the induction medium was consistent with the sole induction process.

The total induction period was 7 days. In order to avoid interference with cell proliferation in the cell differentiation experiments, 0.5 µg/ml aphidicolin (Sigma) in dimethylsulfoxide (DMSO, Sigma) was added to the medium at the fourth to fifth day

TE D

of induction.

2.11 Determination of the extents of osteogenesis and adipogenesis After 7 d of induction, the osteoblasts and adipocytes were stained to be

EP

distinguished from nondifferentiated cells. We applied Fast Blue RR salt (Sigma) supplemented with naphthol AS-MX phosphate (Sigma) to stain the alkaline phosphatase (ALP) in cells. Cells positive to ALP would be stained in blue, and the

AC C

grey levels reflected the osteogenic differentiation extent. For the osteogenic induction, samples were fixed with 4% paraformaldehyde for 20 min after gently rinsed with PBS. Then samples were soaked in fast blue/naphthol solution for 30 min. The oil droplets in adipocytes were chosen as the marker to determine the adipogenic extent. Oil Red O (Sigma-Aldrich) was utilized to stain the adipogenically induced cells. Samples were treated with 3 mg/mL Oil Red O in 60% isopropanol after rinsing with PBS, fixing with 4% paraformaldehyde and rinsing with 60% isopropanol solution. Nuclei were also labeled by DAPI in order to count the cell numbers. As for co-induction samples, the ALP, oil droplets, and nuclei in the cells were stained in

ACCEPTED MANUSCRIPT turn. The stained cells were photographed through a microscope for statistical analysis. Cells possessing oil droplets and being stained in red were set as adipocytes. Cells with ALP being stained in deep blue were defined as osteoblasts. In the co-induction

RI PT

groups, the differentiated cells were expressed as percentages. In osteogenic induction, the relative ALP activity per cell was chosen as the criteria [20]. Bright-field images were converted to 8-bit greyscale images, and only the green channel images were put into grey value analysis. Within the ALP stained

SC

region, the maximum absorbance of visible light was at 574 nm which fell into the scale of green spectrum. Thus, the ALP activity per cell AALP was calculated from !"

= #

× lg< ' >⁄< ' >/+

M AN U



Here, Scell denotes the area of a single cell or cell groups outlined by the software ImageJ; refers to the mean grey value of the background near the outlined cells with the same area; refers to the mean grey value of the cells; N is the number of cells in the outlined region.

TE D

2.12 Statistical analysis

Four independent experiments in each group were performed and averaged in the induction experiments (n = 4). In cell adhesion analysis, six independent experiments

EP

were carried out (n = 6), and for each sample, approximately 200 cells were measured. Data from above experiments among different groups were analyzed by One-way ANOVA tests. A significant difference was denoted as “*” (p < 0.05), “**” (p < 0.01)

AC C

and “***” (p < 0.001, very significant difference). 3. Results

3.1 Synthesis of oligoLA-PEG-oligoLA and diacrylated macromers Block copolymer oligoLA-PEG-oligoLA was obtained via ring-opening polymerization of lactide triggered by the hydroxyl end groups of PEG and catalyzed by Sn(Oct)2. The product was a white sticky solid at room temperature and easily dissolved in DCM. The chemical structure of the resultant polymer with LA units was confirmed by 1H NMR, as demonstrated in Fig. S1. Three different protons existed in

ACCEPTED MANUSCRIPT oligoLA-PEG-oligoLA: protons from the methine (-CH) and methyl (-CH3) of the LA units had chemical shifts of 5.2 ppm and 1.4-1.6 ppm, respectively; the characteristic peaks in the ranges of 3.45-3.81 ppm and 4.25-4.35 ppm belong to the methylene (-CH2-CH2) protons of the PEG segment. The two shifts indicate that the two kinds of

RI PT

protons in the PEG segment are not chemically equivalent. The average number of LA units at each side of PEG chains is about 2.5, as shown in Table S1.

Diacrylated macromer oligoLA-PEG-oligoLA DA was gained by treating the oligo-block copolymer oligoLA-PEG-oligoLA with acryloyl chloride, using TEA as

SC

an acid acceptor during the reaction process. The resulting product was also a white solid, but with powder-like appearance possibly caused by the alkene part of the

M AN U

acryloyl groups. The presence of alkenes was confirmed by 1H NMR, as shown in Fig. S1. The plane structure of the alkene’s double bond lay in three different protons (peak e and two f peaks). The –CH= proton peak can be found at 6.1-6.2 ppm and the proton peaks from =CH2 are at 5.89-5.92 ppm and 6.45-6.5 ppm. The modified ratio of double bonds is approximately 70%.

TE D

3.2 Photopolymerization of oligoLA-PEG-oligoLA diacrylate macromers and degradability of the resultant hydrogels

The as-synthesized macromers were photopolymerized to form hydrogels. The

EP

resulting chemically crosslinked hydrogels were transparent, resilient and with a certain strength. After swelling, the equilibrium linear swelling ratio of the hydrogel in 37°C PBS was determined to be 1.13, owing to retained water after

AC C

photopolymerization.

Hydrogels were first treated with 2 mol/L HCl solution to quickly observe

changes during degradation. After soaked in HCl, hydrogels were taken out at given time points and rinsed with PBS. We then detected moduli and swelling ratios of hydrogels in cell culture medium. The approach to measure the compressive modulus of the hydrogel is demonstrated in Fig. 2A and Fig. S2. For the compressive modulus, an obvious decrease (approximately 35 times) appeared after 12 h treatment.

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

EP

Fig. 2. Compressive moduli of hydrogels during biodegradation and varied degradation rates characterized by modulus change as a function of time at

AC C

different degradation stages. (A) Illustration of the measurement of compressive modulus and the modulus evolution relative to the time of treatment by the indicated acid. (B) Compressive moduli of the hydrogels with pre-treatment by acid for indicated times, being washed by PBS five times, and then immersed in cell culture medium for 7 days. The normalized data in the right illustrate more clearly the different degradation rates of the three hydrogel samples with the basic characters summarized at the bottom.

Untreated samples (0 h) and samples after 2 h and 6 h of acid treatments were

ACCEPTED MANUSCRIPT selected as three groups (samples A, B and C) in the later cell experiments. In order to monitor degradation behaviors under in vitro biomimetic conditions, all of the samples were washed with PBS for five times, and then put into a cell culture medium (DMEM with 10 % FBS, 37°C) for 7 days. The three groups exhibited the same trend

RI PT

of declining moduli during biodegradation but different degradation rates, as illustrated in Fig. 2B. The moduli and degradation rates of different samples are summarized at the bottom of Fig. 2B. The stiffer hydrogel (sample A) showed a small change of modulus which was defined as slow degradation, while the softer sample C

SC

showed a fast degradation; hydrogels from sample B held moderate properties.

Swelling ratios varied also at different acid-treatment stages. Nevertheless, the

M AN U

linear swelling ratios (SR) didn’t change so much as moduli. Among the three groups used in the later cell experiments, the softest sample C changed the most albeit still less than 1.6 (Fig. S3).

3.3 Preparation of nanopatterns on the biodegradable hydrogels Biodegradable hydrogels with an RGD array on their surfaces were acquired via

TE D

block copolymer micelle nanolithography to prepare a gold nanoarray on glass, and transfer lithography to transfer the gold nanoarray from the glass to the hydrogels, which was formed by photopolymerization of the as-synthesized biodegradable

EP

macromer. Amphipathic block polymer chains were self-assembled into micelles in a selective solvent, as shown in a TEM observation (Fig. 3A). The polymeric micelles loaded HAuCl4 in their cores. After plasma treatment, the polymer was removed, the

AC C

precursor of gold was reduced to gold nanodots, and the micelle array formed after dip-coating served as an ordered template of the resultant gold nanoarray. The array of gold nanodots on glass was observed by SEM, as shown in Fig. 3A. The spacing between nearest-neighbor nanodots was around 30 nm. The mean and standard deviation of nanospacing are presented in Table S2.

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

EP

Fig. 3. Fabricated nanopatterns and cells on the patterned surfaces. (A) A TEM image of micelles of block copolymer PS-b-P2VP in toluene (left) and an SEM image of a self-assembled quasi-hexagonal gold nanoarray on glass after dip-coating and treatment

(right).

AC C

plasma

(B)

Fluorescence

micrographs

of

MSCs

on

RGD-nanopatterned biodegradable hydrogels at 1 d (left) and 7 d (right) after cell seeding. Cells were stained with F-actin in red, vinculin in green, and nuclei in blue. The dashed lines indicate diplines.

A naturally formed dipline divided the glass surface into two parts: one side with nanopattern and the other without nanopattern. In addition to the successful transfer of the gold nanopattern to the oligoLA-PEG-oligoLA hydrogel, the dipline would also be transferred to the hydrogel. Cell seeding on the hydrogel with the

ACCEPTED MANUSCRIPT properties mentioned above achieved an adhesion contrast on both sides of the dipline (Fig. 3B). The labeled vinculin (stained in green) of the cells was distributed at one side of the dipline. Even after 7 days of cell culturing, the dipline could be well observed, which illustrated both the persistently nonfouling property of the hydrogel

RI PT

background and the cell-adhesive potency of the RGD nanoarray. 3.4 Cell adhesion on nanopatterned hydrogels at different degradation stages

MSCs were cultured on the hydrogels of different degradation stages with varied stiffness and degradation rates (samples A, B and C). Hydrogels were seeded with

SC

cells of similar densities. Non-adherent cells were washed away before cell staining. Representative fluorescence micrographs of the stained cells and corresponding

M AN U

statistics after 1 day of cell adhesion are shown in Fig. 4. While cell density was lower in soft hydrogels, slightly complicated trends were observed for spreading area and shape (Fig. 4A). It is very interesting that although the softest group (sample C) led to relatively more confined cell spreading, the cytoskeleton was not weaker, as judged

AC C

EP

TE D

from the F-actin staining images shown in the upper row of Fig. 4B.

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 4. Cell adhesion on hydrogels with different degradation rates. (A) Statistic

AC C

results of cell adhesion parameters after 1-day culture on indicated samples. (B) Corresponding fluorescence micrographs of stained MSCs with F-actin in red, vinculin in green, and nuclei in blue.

3.5 Cell viabilities on nanopatterned surfaces at different degradation stages We also did some statistics on the cell cytoskeleton. The F-actin intensity of single cells was measured. After being seeded on hydrogels for 1 day, the cytoskeleton of the cells was fluorescently stained and photographed with consistent exposure time. For semi-quantitatively analyzing cell tension on different samples, the integral

ACCEPTED MANUSCRIPT intensity of F-actin per cells was calculated. According to the graph demonstrated in Fig. 5A, cells on softer sample B showed a decrease of intensity compared to on sample A; but cells on the softest sample C presented a different trend. The intensity of cells on sample C surpassed that on sample B, which implied that cell tension

M AN U

SC

RI PT

increased on matrices of a higher degradation rate.

Fig. 5. Statistics concerning the cytoskeleton of single cells and cell viabilities of

TE D

the whole cells on indicated hydrogels 1 day after seeding of the MSCs. (A) Statistic results of integral intensity of F-actin per cell adhering on hydrogels. Intensity of F-actin was normalized to the mean value of sample B. (B) CCK-8 results of cell viabilities on different hydrogel samples, using culture on glass as a control.

EP

Cell viability was obtained from the measurements of activities of the enzyme SDH in

AC C

all cells on each sample and normalized to the mean of the control group.

As for the overall activities of cells adhering on those hydrogels, a CCK-8 kit

was utilized to quantify the whole cell viability. After seeding of the cells for 1 day, cells adhered and spread well on glass which constituted the maximum cell viability (Fig. 5B). Among the three samples, cells on degrading hydrogels in samples B and C showed increased viabilities; and the moderate and high degradation rates of the hydrogels activated cells much more than the low degradation rate in sample A. 3.6 Sole osteogenic or adipogenic induction of MSCs on samples at different degradation stages

ACCEPTED MANUSCRIPT To evaluate the effects of biodegradation on stem cell differentiation, MSCs were seeded on samples A, B and C, and underwent osteogenic or adipogenic induction. Osteogenically induced cells were stained by Fast Blue RR Salt. The osteogenesis extents among all samples were compared by ALP activities, which were

RI PT

quantitatively calculated by the relative grey values of stained cells. Meanwhile, red-stained oil droplets that appeared in the cells were considered as a marker of adipocytes. The results are presented in Figs. 6A and 6B.

Both osteogenesis and adipogenesis exhibited enhancement in the sequence of

SC

samples A, B and C in Figs. 6A and 6B. Such trends on degrading hydrogels are very interesting and abnormal, because the previous common knowledge is that a stiff

M AN U

hydrogel favors osteogenesis but a soft hydrogel favors adipogenesis on static matrices. It seems also worthy of noting that the same trends do not mean no selectivity. The difference between adipogenesis and osteogenesis was raised in our well designed co-induction experiments, with the results presented in the next

AC C

EP

TE D

subsection.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 6. Bright-field images and statistics of MSC differentiation on samples at

ACCEPTED MANUSCRIPT different degradation stages. (A) Osteogenesis of MSCs after osteogenic induction for 7 d. Cells experienced ALP staining and nucleus staining prior to the optical microscopic observations. (B) Adipogenesis of MSCs after adipogenic induction for 7 d. Cells were submitted to oil-droplet staining and nucleus staining prior to

RI PT

observations. (C) Co-induction of MSCs in a mixed induction medium for 7 days. Cells experienced ALP staining, oil-droplet staining and nucleus staining prior to the microscopic observations.

SC

3.7 Complicated lineage commitments in co-induction of MSCs on samples at different degradation stages

M AN U

We further examined stem cell differentiation in a mixed osteogenic/adipogenic medium. The co-induction results are very complicated owing to competition between the two lineage commitments. As shown in Fig. 6C, the optimal adipogenesis occurred for sample B with a moderate matrix stiffness and degradation rate while the osteogenic extent was at a minimum. The softest hydrogel showing the fastest

TE D

degradation (sample C) led to more significant osteogenic differentiation than the hardest hydrogel that show the slowest degradation (sample A).

EP

4. Discussion

It is very important to reveal various cues to regulate stem cells [7, 9, 11, 13, 65]. Certain cues might be predominant over others in some cases of lineage commitment.

AC C

In the present study, we prepared RGD nanoarrays on PEG-based yet biodegradable hydrogels by combining the techniques of extension of PEG by oligoester and then acryloyl to synthesize a biodegradable macromer, block copolymer micelle nanolithography to obtain a gold nanoarray on glass, transfer lithography assisted by a bifunctional linker to transfer the gold nanoarray to the biodegradable hydrogel, and eventually, linking RGD peptides using a self-assembled monolayer. The preparation principles and the basic steps are schematically presented in Fig. 1. In this way, we generated a well-controlled nanopatterned surface with PEG as a nonfouling background and RGD peptides affording cell-adhesive sites. The excellent control of

ACCEPTED MANUSCRIPT cell adhesion is demonstrated in Fig. 3. Another key new issue of material chemistry comes from the introduction of oligoester into the macromer prior to the nanopatterning, which enables the biodegradability of the synthesized hydrogels. The biodegradation made the hydrogels softer, as demonstrated in Fig. 2. We also

RI PT

found that the three samples which were prepared at different accelerated hydrolysis stages led to different biodegradation rates in the cell culture medium. So, we eventually obtained three well-controlled samples with both varied stiffness and degradation rates, as highlighted in the bottom of Fig. 2.

SC

As the stiffness effects on cell adhesion and differentiation are concerned, it seems necessary to indicate that real differences in stem cell differentiation are

M AN U

typically observed at moduli in the low kPa (1-2 kPa) to 20-30 kPa range as reported in the literature [28, 46]. But they are for cells encapsulated in some 3D hydrogels. Different material systems have different appropriate modulus ranges with significant stiffness effects, and we found that the stiffness range for cells on 2D surfaces decorated with a nanogold array might shift to higher moduli. According to our

TE D

previous publications about the significant stiffness effects on adhesion of MSCs and the following adipogenic/osteogenic induction in nanopatterned 2D systems [24, 25], the relatively soft and stiff static PEG hydrogels were of moduli about 3200 kPa and

EP

130 kPa, respectively. The present studies of dynamic hydrogels of PEG derivatives during biodegradation in cell culture medium exhibited moduli among 70-1800 kPa, which is sufficient to reflect the underlying stiffness effect if any.

AC C

It is very interesting that cell behaviors on our degrading hydrogel cannot be

fully understood by those on static hydrogels with also varied stiffness. Our adipogenic induction experiments agree very well with the normal tend that soft matrices benefit to adipogenesis, as shown in Fig. 6A. However, the osteogenic induction results are contradictory with the normal assumption, which implies the impact of some unknown material cues. Meanwhile, we found that cell adhesion behaviors on our degradable hydrogels did not completely follow those on static hydrogels with different matrix stiffness either. MSCs exhibited smaller adhesion density and spreading area (Fig. 4A), which are consistent with the previous

ACCEPTED MANUSCRIPT publications on nonbiodegradable PEG hydrogels of two stiffness values [24]. However, the cytoskeleton on the softest sample (sample C) seemed more stringent (Fig. 4B). Such an unexpected phenomenon was further confirmed by the relative integral F-actin intensity per cell (Fig. 5A). More interestingly, in spite of less cell

detected via CCK-8 experiments (Fig. 5B).

RI PT

density on sample C (Fig. 4A), a higher cell viability of the whole adhesive cells was

Besides matrix stiffness, another parameter that changed during hydrogel degradation in our experiments is the enlarged RGD nanospacing along with hydrogel

SC

swelling. RGD is a ligand in ECM to promote specific cell adhesion after conjugating with its receptor, integrin, on the cell membrane [66, 67] . The critical nanospacing is

M AN U

about 70 nm: cells can only adhere well on surfaces with an RGD nanospacing < 70 nm [68-72]. We have reported that the enlarged RGD nanospacing can enhance both osteogenesis and adipogenesis when the spacing changes from < 70 nm to > 70 nm [65, 73]. The initial RGD nanospacing on glass in the present experiment was 30.9 nm as listed in Table S2. Considering the linear swelling ratio of the initial hydrogel

TE D

obtained from photopolymerization, the RGD nanospacing on our biodegradable hydrogel was about 34.9 nm. Even with the highest linear swelling ratio of 1.6 for sample C after degradation for 7 days as shown in Fig. S3, the resulting RGD

EP

nanospacing might still be less than 70 nm. A more important evidence to rule out the interference of the changed RGD nanospacing comes from the facts that larger RGD nanospacings always lead to significantly decreased cell viability and weaker

AC C

cytoskeletons [24, 65] and in contrast higher cell viability and stronger cytoskeleton could be observed with degradation of our hydrogels. So, our present experiments implied a new cue beyond both decreased matrix stiffness and different RGD nanospacing.

Material degradation itself could contribute to the cell results. The degradation byproducts such as lactic acid afforded new components in culture medium. Our previous research on the byproducts of hydrolysis of polyesters illustrates that only a large amount of lactic acid could affect cells and it is the decreased pH instead of the lactate ion that influences the differentiation of MSCs [74]. In the present hydrogel

ACCEPTED MANUSCRIPT system, the oligoLA modification to PEG chains resulted in only a little amount of degradation byproducts of lactic acid, and the buffered cell culture medium would stable pH around neutral according to our experimental record. Hence, the byproducts cannot be the main factor of the hydrogel degradation to influence the cell behaviors

RI PT

in this study. So, the rate of biodegradation, another factor closely related to the dynamic process of degradation should be taken into consideration. The highest degradation rate in sample C exerted a dynamic stimulus to cells. Such a stimulus was so strong

SC

that shaded the otherwise weakened adhesion of cells on soft matrices, and sometimes a more stringent cytoskeleton was raised and even the cell viability was increased, as

M AN U

reflected in Figs. 4 and 5. This dynamic cue well accounts for more osteogenesis of MSCs on sample C in Fig. 6. It has been known that osteogenesis is very sensitive to cell tension [13, 15, 17, 19, 65, 75]. Our previous publication based on micropatterned nonbiodegradable surfaces has indicated that the extent of osteogenesis is related to cell tension while that of adipogenesis is not [17]. So, the sensitivities of these two

TE D

lineage commitments to the cytoskeleton and cell viability are quite different. We thus interpret our sole induction results as follows: in the adipogenic induction on samples A, B and C, the matrix stiffness is predominant over the

EP

degradation rate; and in the osteogenic induction, the degradation rate is predominant

AC C

over the matrix stiffness. These two trends are schematically presented in Fig. 7A.

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 7. Schematic presentation of the fate decision for stem cells on hydrogels during biodegradation. (A) In the cases of sole osteogenic or adipogenic induction

EP

of MSCs on the material surfaces; (B) In the case of co-induction in a bipotential mixed medium.

AC C

More complicated results were found in co-induction, as seen in Fig. 6C. Here

competition should be taken into consideration. The competition between osteogenesis and adipogenesis in the mixed induction medium has already been studied [13, 22]. The two kinds of sole induction led to the same increased differentiation trends (Figs. 6A and 6B), and thus it is not unreasonable that a peak might, under co-induction, appear at sample B in Fig. 6C. Since osteogenesis is more sensitive to the dynamic stimuli caused by material degradation, sample C with the highest degradation rate exhibited a higher extent of osteogenesis than sample A with the slowest biodegradation. The combined effects in the co-induction experiments are

ACCEPTED MANUSCRIPT also schematically shown in Fig. 7. The present report is in agreement with the previous pertinent studies. Burdick’s group found that cells extended more, showed a stronger cell tension, generated more traction force to the HA matrix around cells, and exhibited osteogenesis lineage

RI PT

tendency in cell-mediated degradation of enzyme-degradable and HA-based hydrogel [43]; they pointed out that the key factor to influence differentiation was the degradation rather than the initial cell spreading. Bian et al. confirmed and extended the viewpoint, and reported a more significant differentiation under the chondrogenic

SC

induction medium in enzyme-degradable HA hydrogels than in non-degradable hydrogels [44]. Anseth’s group found that osteogenesis and chondrogenesis were

M AN U

more sensitive to the degradation extent than the adipogenesis of MSCs in MMP-degradable and PEG-based hydrogel [52]. So, our experiments of cells on 2D patterned surfaces are consistent with those 3D studies of MSCs encapsulated in hydrogels.

The degradation rate affords a new cue to regulate cell adhesion and

TE D

differentiation. Such a dynamic cue should be considered during both 2D culturing and 3D encapsulation, although dimensionality might influence signaling of MSCs in hydrogel environments [76]. The main difference of a cell in 3D or 2D systems comes

EP

from the presence or absence of steric hindrance. Our in vitro experiments on 2D surfaces are helpful for understanding various experiments in cell culture when the matrix has a dynamic characteristic. We can also reasonably anticipate that

AC C

degradation rates can influence cell fate significantly during tissue engineering and tissue regeneration after implanting biodegradable materials. Even in the case of nonbiodegradable implants, ECM remodeling might also alter matrix stiffness [77], and thus the change rate of the matrix stiffness should be considered as a dynamic cue. Recently, matrix remolding during degradation was found to support the cell-cell contact and spreading, thus promoting stemness of neural progenitor cells (NPCs) [78]. Different from NPCs, more contractile lineages of MSCs showed the dependence upon cell-adhesive ligand clustering and the traction force generated by cells in the degradation process [79]. The present study illustrated that under fast

ACCEPTED MANUSCRIPT degradation cells might sense the impact of changes of the hydrogel and turn out to be activated along with the reconstruction of the cytoskeleton and a higher traction force.

5. Conclusions

RI PT

In order to examine stem cells on well-controlled biomaterials during biodegradation, we built nanoarrays of cell-adhesive RGD peptides on hydrolysable hydrogels synthesized from biodegradable diacrylated oligoLA-PEG-oligoLA macromers. The accelerated hydrolysis in an acidic medium led to a series of samples

SC

with varied matrix stiffness and degradation rates during the subsequent treatment in the cell culture medium for several days. The cell density and spreading area declined

M AN U

on samples at the late stage of acidic hydrolysis, but cytoskeleton and cell viability were found to be enhanced on softer samples with a higher degradation rate. While more adipogenesis happened on softer samples, osteogenesis was also more significant, which cannot be attributed to the well-known stiffness effect. We thus put forward that it was the degradation rate that stimulated cells and then enhanced

TE D

osteogenesis of MSCs. The cell adhesion and differentiation on the biodegradable materials were found to be determined by the combinatory effects of degradation rate, matrix stiffness, competition of commitment between different lineages, etc. The

EP

present study reveals that degradation rates cannot be interpreted simply by the changed static cues. Such a dynamic cue should be paid much attention to in the understanding of cell-matrix interactions and development of next-generation

AC C

biomaterials.

Acknowledgements

The authors are grateful for the financial supports from NSF of China (grants No. 51533002), National Key R&D Program of China (grant No. 2016YFC1100300), and Science and Technology Commission of Shanghai Municipality (grant No. 17JC1400200).

ACCEPTED MANUSCRIPT Supplementary Information Supplementary information associated with this article can be found, in the online

AC C

EP

TE D

M AN U

SC

RI PT

version, at doi: ***.

ACCEPTED MANUSCRIPT References

AC C

EP

TE D

M AN U

SC

RI PT

[1] L.L. Hench, J.M. Polak, Third-generation biomedical materials, Science 295 (2002) 1014-1017. [2] P. Roach, D. Eglin, K. Rohde, C.C. Perry, Modern biomaterials: a review—bulk properties and implications of surface modifications, J. Mater. Sci.: Mater. Med. 18 (2007) 1263-1277. [3] B. Trappmann, J.E. Gautrot, J.T. Connelly, D.G. Strange, Y. Li, M.L. Oyen, M.A.C. Stuart, H. Boehm, B. Li, V. Vogel, Extracellular-matrix tethering regulates stem-cell fate, Nat. Mater. 11 (2012) 642-649. [4] X. Yao, R. Peng, J.D. Ding, Cell–material interactions revealed via material techniques of surface patterning, Adv. Mater. 25 (2013) 5257-5286. [5] M. Lutolf, J. Hubbell, Synthetic biomaterials as instructive extracellular microenvironments for morphogenesis in tissue engineering, Nat. Biotechnol. 23 (2005) 47-55. [6] T. Muraoka, C.Y. Koh, H. Cui, S.I. Stupp, Light triggered bioactivity in three dimensions, Angew. Chem. 121 (2009) 6060-6063. [7] A. Higuchi, Q.D. Ling, Y. Chang, S.T. Hsu, A. Umezawa, Physical cues of biomaterials guide stem cell differentiation fate, Chem. Rev. 113 (2013) 3297-3328. [8] S. Moeinzadeh, S.R.P. Shariati, E. Jabbari, Comparative effect of physicomechanical and biomolecular cues on zone-specific chondrogenic differentiation of mesenchymal stem cells, Biomaterials 92 (2016) 57-70. [9] X.N. Liu, R.L. Liu, B. Cao, K. Ye, S.Y. Li, Y.X. Gu, Z. Pan, J.D. Ding, Subcellular cell geometry on micropillars regulates stem cell differentiation, Biomaterials 111 (2016) 27-39. [10] J.M. Curran, R. Chen, J.A. Hunt, The guidance of human mesenchymal stem cell differentiation in vitro by controlled modifications to the cell substrate, Biomaterials 27 (2006) 4783-4793. [11] D.S. Benoit, M.P. Schwartz, A.R. Durney, K.S. Anseth, Small functional groups for controlled differentiation of hydrogel-encapsulated human mesenchymal stem cells, Nat. Mater. 7 (2008) 816-823. [12] B. Cao, Y.M. Peng, X.N. Liu, J.D. Ding, Effects of functional groups of materials on nonspecific adhesion and chondrogenic induction of mesenchymal stem cells on free and micropatterned surfaces, ACS Appl. Mater. Interfaces 9 (2017) 23574-23585. [13] R. McBeath, D.M. Pirone, C.M. Nelson, K. Bhadriraju, C.S. Chen, Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment, Dev. Cell 6 (2004) 483-495. [14] B. Cao, R. Peng, Z.H. Li, J.D. Ding, Effects of spreading areas and aspect ratios of single cells on dedifferentiation of chondrocytes, Biomaterials 35 (2014) 6871-6881. [15] K.A. Kilian, B. Bugarija, B.T. Lahn, M. Mrksich, Geometric cues for directing the differentiation of mesenchymal stem cells, Proc. Natl. Acad. Sci. U.S.A. 107 (2010) 4872-4877. [16] R. Peng, X. Yao, J.D. Ding, Effect of cell anisotropy on differentiation of stem

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

cells on micropatterned surfaces through the controlled single cell adhesion, Biomaterials 32 (2011) 8048-8057. [17] X. Yao, R. Peng, J.D. Ding, Effects of aspect ratios of stem cells on lineage commitments with and without induction media, Biomaterials 34 (2013) 930-939. [18] S. Zhang, X. Liu, S.F. Barreto-Ortiz, Y. Yu, B.P. Ginn, N.A. DeSantis, D.L. Hutton, W.L. Grayson, F.-Z. Cui, B.A. Korgel, S. Gerecht, H.-Q. Mao, Creating polymer hydrogel microfibres with internal alignment via electrical and mechanical stretching, Biomaterials 35 (2014) 3243-3251. [19] M. Bao, J. Xie, A. Piruska, W.T. Huck, 3D microniches reveal the importance of cell size and shape, Nat. Commun. 8 (2017) 1962. [20] J. Tang, R. Peng, J.D. Ding, The regulation of stem cell differentiation by cell-cell contact on micropatterned material surfaces, Biomaterials 31 (2010) 2470-2476. [21] B. Cao, Z.H. Li, R. Peng, J.D. Ding, Effects of cell–cell contact and oxygen tension on chondrogenic differentiation of stem cells, Biomaterials 64 (2015) 21-32. [22] R. Peng, X. Yao, B. Cao, J. Tang, J.D. Ding, The effect of culture conditions on the adipogenic and osteogenic inductions of mesenchymal stem cells on micropatterned surfaces, Biomaterials 33 (2012) 6008-6019. [23] D.E. Discher, P. Janmey, Y.-L. Wang, Tissue cells feel and respond to the stiffness of their substrate, Science 310 (2005) 1139-1143. [24] K. Ye, X. Wang, L.P. Cao, S.Y. Li, Z.H. Li, L. Yu, J.D. Ding, Matrix stiffness and nanoscale spatial organization of cell-adhesive ligands direct stem cell fate, Nano Lett. 15 (2015) 4720-4729. [25] K. Ye, L.P. Cao, S.Y. Li, L. Yu, J.D. Ding, Interplay of matrix stiffness and cell– cell contact in regulating differentiation of stem cells, ACS Appl. Mater. Interfaces 8 (2016) 21903-21913. [26] A.J. Engler, S. Sen, H.L. Sweeney, D.E. Discher, Matrix elasticity directs stem cell lineage specification, Cell 126 (2006) 677-689. [27] J.H. Wen, L.G. Vincent, A. Fuhrmann, Y.S. Choi, K.C. Hribar, H. Taylor-Weiner, S. Chen, A.J. Engler, Interplay of matrix stiffness and protein tethering in stem cell differentiation, Nat. Mater. 13 (2014) 979-987. [28] K.M. Mabry, R.L. Lawrence, K.S. Anseth, Dynamic stiffening of poly(ethylene glycol)-based hydrogels to direct valvular interstitial cell phenotype in a three-dimensional environment, Biomaterials 49 (2015) 47-56. [29] H.Y. Liu, T. Greene, T.Y. Lin, C.S. Dawes, M. Korc, C.C. Lin, Enzyme-mediated stiffening hydrogels for probing activation of pancreatic stellate cells, Acta Biomater. 48 (2017) 258-269. [30] F. von Burkersroda, L. Schedl, A. Göpferich, Why degradable polymers undergo surface erosion or bulk erosion, Biomaterials 23 (2002) 4221-4231. [31] L.B. Wu, J.D. Ding, In vitro degradation of three-dimensional porous poly (D, L-lactide-co-glycolide) scaffolds for tissue engineering, Biomaterials 25 (2004) 5821-5830. [32] Z. Zhang, R. Kuijer, S.K. Bulstra, D.W. Grijpma, J. Feijen, The in vivo and in vitro degradation behavior of poly(trimethylene carbonate), Biomaterials 27 (2006)

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

1741-1748. [33] Y.-A. Lin, Y.-C. Ou, A.G. Cheetham, H. Cui, Rational design of MMP degradable peptide-based supramolecular filaments, Biomacromolecules 15 (2014) 1419-1427. [34] Y.L. Qi, H. Qi, Y. He, W. Lin, P.Z. Li, L. Qin, Y.W. Hu, L. Chen, Q.S. Liu, H. Sun, Q. Liu, G. Zhang, S.Q. Cui, J. Hu, L. Yu, D.Y. Zhang, J.D. Ding, Strategy of metal– polymer composite stent to accelerate biodegradation of iron-based biomaterials, ACS Appl. Mater. Interfaces 10 (2018) 182−192. [35] M.A. Washington, D.J. Swiner, K.R. Bell, M.V. Fedorchak, S.R. Little, T.Y. Meyer, The impact of monomer sequence and stereochemistry on the swelling and erosion of biodegradable poly (lactic-co-glycolic acid) matrices, Biomaterials 117 (2017) 66-76. [36] C. Adelöw, T. Segura, J.A. Hubbell, P. Frey, The effect of enzymatically degradable poly (ethylene glycol) hydrogels on smooth muscle cell phenotype, Biomaterials 29 (2008) 314-326. [37] K.B. Fonseca, D.B. Gomes, K. Lee, S.G. Santos, A. Sousa, E.A. Silva, D.J. Mooney, P.L. Granja, C.C. Barrias, Injectable MMP-sensitive alginate hydrogels as hMSC delivery systems, Biomacromolecules 15 (2013) 380-390. [38] D.S. Ferreira, Y.-A. Lin, H. Cui, J.A. Hubbell, R.L. Reis, H.S. Azevedo, Molecularly engineered self-assembling membranes for cell-mediated degradation, Adv. Healthcare Mater. 4 (2015) 602-612. [39] X. Li, S.Y. Tzeng, X. Liu, M. Tammia, Y.-H. Cheng, A. Rolfe, D. Sun, N. Zhang, J.J. Green, X. Wen, H.-Q. Mao, Nanoparticle-mediated transcriptional modification enhances neuronal differentiation of human neural stem cells following transplantation in rat brain, Biomaterials 84 (2016) 157-166. [40] E.B. Peters, N. Christoforou, K.W. Leong, G.A. Truskey, J.L. West, Poly(ethylene glycol) hydrogel scaffolds containing cell-adhesive and protease-sensitive peptides support microvessel formation by endothelial progenitor cells, Cell. Mol. Bioeng. 9 (2016) 38-54. [41] D. Gvaramia, E. Müller, K. Müller, P. Atallah, M. Tsurkan, U. Freudenberg, M. Bornhäuser, C. Werner, Combined influence of biophysical and biochemical cues on maintenance and proliferation of hematopoietic stem cells, Biomaterials (2017) 108-117. [42] A.M. Kloxin, M.W. Tibbitt, A.M. Kasko, J.A. Fairbairn, K.S. Anseth, Tunable hydrogels for external manipulation of cellular microenvironments through controlled photodegradation, Adv. Mater. 22 (2010) 61-66. [43] S. Khetan, M. Guvendiren, W.R. Legant, D.M. Cohen, C.S. Chen, J.A. Burdick, Degradation-mediated cellular traction directs stem cell fate in covalently crosslinked three-dimensional hydrogels, Nat. Mater. 12 (2013) 458-465. [44] Q. Feng, M. Zhu, K. Wei, L. Bian, Cell-mediated degradation regulates human mesenchymal stem cell chondrogenesis and hypertrophy in MMP-sensitive hyaluronic acid hydrogels, PLoS One 9 (2014) e99587. [45] O. Chaudhuri, L. Gu, M. Darnell, D. Klumpers, S.A. Bencherif, J.C. Weaver, N. Huebsch, D.J. Mooney, Substrate stress relaxation regulates cell spreading, Nat. Commun. 6 (2015) 6365.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[46] O. Chaudhuri, L. Gu, D. Klumpers, M. Darnell, S.A. Bencherif, J.C. Weaver, N. Huebsch, H.P. Lee, E. Lippens, G.N. Duda, D.J. Mooney, Hydrogels with tunable stress relaxation regulate stem cell fate and activity, Nat. Mater. 15 (2016) 326-334. [47] J.L. West, J.A. Hubbell, Photopolymerized hydrogel materials for drug delivery applications, React. Polym. 25 (1995) 139-147. [48] A. Metters, K. Anseth, C. Bowman, Fundamental studies of a novel, biodegradable PEG-b-PLA hydrogel, Polymer 41 (2000) 3993-4004. [49] J. Zhu, Bioactive modification of poly (ethylene glycol) hydrogels for tissue engineering, Biomaterials 31 (2010) 4639-4656. [50] J. Patterson, J.A. Hubbell, Enhanced proteolytic degradation of molecularly engineered PEG hydrogels in response to MMP-1 and MMP-2, Biomaterials 31 (2010) 7836-7845. [51] S.P. Zustiak, J.B. Leach, Hydrolytically degradable poly (ethylene glycol) hydrogel scaffolds with tunable degradation and mechanical properties, Biomacromolecules 11 (2010) 1348-1357. [52] S.B. Anderson, C.-C. Lin, D.V. Kuntzler, K.S. Anseth, The performance of human mesenchymal stem cells encapsulated in cell-degradable polymer-peptide hydrogels, Biomaterials 32 (2011) 3564-3574. [53] H. Peng, S. Varanasi, D.K. Wang, I. Blakey, F. Rasoul, A. Symons, D.J. Hill, A.K. Whittaker, Synthesis, swelling, degradation and cytocompatibility of crosslinked PLLA-PEG-PLLA networks with short PLLA blocks, Eur. Polym. J. 84 (2016) 448-464. [54] S.J. Bryant, K.S. Anseth, Hydrogel properties influence ECM production by chondrocytes photoencapsulated in poly (ethylene glycol) hydrogels, J. Biomed. Mater. Res., Part A 59 (2002) 63-72. [55] S.J. Bryant, K.S. Anseth, Controlling the spatial distribution of ECM components in degradable PEG hydrogels for tissue engineering cartilage, J. Biomed. Mater. Res., Part A 64 (2003) 70-79. [56] J.A. Burdick, M.N. Mason, A.D. Hinman, K. Thorne, K.S. Anseth, Delivery of osteoinductive growth factors from degradable PEG hydrogels influences osteoblast differentiation and mineralization, J. Control. Release 83 (2002) 53-63. [57] Y. Zhang, W. Zhu, B.B. Wang, J.D. Ding, A novel microgel and associated post-fabrication encapsulation technique of proteins, J. Control. Release 105 (2005) 260-268. [58] W. Zhu, J.D. Ding, Synthesis and characterization of a redox-initiated, injectable, biodegradable hydrogel, J. Appl. Polym. Sci. 99 (2006) 2375-2383. [59] L.S. Nair, C.T. Laurencin, Biodegradable polymers as biomaterials, Prog. Polym. Sci. 32 (2007) 762-798. [60] A.S. Sawhney, C.P. Pathak, J.A. Hubbell, Bioerodible hydrogels based on photopolymerized poly (ethylene glycol)-co-poly (. alpha.-hydroxy acid) diacrylate macromers, Macromolecules 26 (1993) 581-587. [61] B.B. Wang, W. Zhu, Y. Zhang, Z.G. Yang, J.D. Ding, Synthesis of a chemically-crosslinked thermo-sensitive hydrogel film and in situ encapsulation of model protein drugs, React. Funct. Polym. 66 (2006) 509-518.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[62] J.J. Roberts, G.D. Nicodemus, E.C. Greenwald, S.J. Bryant, Degradation improves tissue formation in (un) loaded chondrocyte-laden hydrogels, Clin. Orthop. Relat. Res. 469 (2011) 2725-2734. [63] S.V. Graeter, J.H. Huang, N. Perschmann, M. López-García, H. Kessler, J.D. Ding, J.P. Spatz, Mimicking cellular environments by nanostructured soft interfaces, Nano Lett. 7 (2007) 1413-1418. [64] J.G. Sun, S.V. Graeter, L. Yu, S.F. Duan, J.P. Spatz, J.D. Ding, Technique of surface modification of a cell-adhesion-resistant hydrogel by a cell-adhesion-available inorganic microarray, Biomacromolecules 9 (2008) 2569-2572. [65] X. Wang, S.Y. Li, C. Yan, P. Liu, J.D. Ding, Fabrication of RGD micro/nanopattern and corresponding study of stem cell differentiation, Nano Lett. 15 (2015) 1457-1467. [66] E. Ruoslahti, M.D. Pierschbacher, New perspectives in cell adhesion: RGD and integrins, Science 238 (1987) 491-498. [67] M. Kato, M. Mrksich, Using model substrates to study the dependence of focal adhesion formation on the affinity of integrin− ligand complexes, Biochemistry 43 (2004) 2699-2707. [68] M. Arnold, E.A. Cavalcanti Adam, R. Glass, J. Blümmel, W. Eck, M. Kantlehner, H. Kessler, J.P. Spatz, Activation of integrin function by nanopatterned adhesive interfaces, ChemPhysChem 5 (2004) 383-388. [69] J.H. Huang, S.V. Grater, F. Corbellini, S. Rinck, E. Bock, R. Kemkemer, H. Kessler, J.D. Ding, J.P. Spatz, Impact of order and disorder in RGD nanopatterns on cell adhesion, Nano Lett. 9 (2009) 1111-1116. [70] S.Y. Li, X. Wang, B. Cao, K. Ye, Z.H. Li, J.D. Ding, Effects of nanoscale spatial arrangement of arginine–glycine–aspartate peptides on dedifferentiation of chondrocytes, Nano Lett. 15 (2015) 7755-7765. [71] Z.H. Li, B. Cao, X. Wang, K. Ye, S.Y. Li, J.D. Ding, Effects of RGD nanospacing on chondrogenic differentiation of mesenchymal stem cells, J. Mater. Chem. B 3 (2015) 5197-5209. [72] J. Deng, C. Zhao, J.P. Spatz, Q. Wei, Nanopatterned adhesive, stretchable hydrogel to control ligand spacing and regulate cell spreading and migration, ACS Nano 11 (2017) 8282-8291. [73] X. Wang, C. Yan, K. Ye, Y. He, Z.H. Li, J.D. Ding, Effect of RGD nanospacing on differentiation of stem cells, Biomaterials 34 (2013) 2865-2874. [74] Y. He, W.R. Wang, J.D. Ding, Effects of L-lactic acid and D, L-lactic acid on viability and osteogenic differentiation of mesenchymal stem cells, Chinese Science Bulletin 58 (2013) 2404-2411. [75] S.A. Ruiz, C.S. Chen, Emergence of patterned stem cell differentiation within multicellular structures, Stem Cells 26 (2008) 2921-2927. [76] S.R. Caliari, S.L. Vega, M. Kwon, E.M. Soulas, J.A. Burdick, Dimensionality and spreading influence MSC YAP/TAZ signaling in hydrogel environments, Biomaterials 103 (2016) 314-323. [77] P. Schedin, P.J. Keely, Mammary gland ECM remodeling, stiffness, and mechanosignaling in normal development and tumor progression, Cold Spring Harbor

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Perspect. Biol. 3 (2011) a003228. [78] C.M. Madl, B.L. LeSavage, R.E. Dewi, C.B. Dinh, R.S. Stowers, M. Khariton, K.J. Lampe, D. Nguyen, O. Chaudhuri, A. Enejder, Maintenance of neural progenitor cell stemness in 3D hydrogels requires matrix remodelling, Nat. Mater. 16 (2017) 1233-1242. [79] N. Huebsch, P.R. Arany, A.S. Mao, D. Shvartsman, O.A. Ali, S.A. Bencherif, J. Rivera-Feliciano, D.J. Mooney, Harnessing traction-mediated manipulation of the cell/matrix interface to control stem-cell fate, Nat. Mater. 9 (2010) 518-526.