Effect of scaffold properties on adhesion and maintenance of boundary cap neural crest stem cells in vitro

Optimal combination of stem cells and biocompatible support material is a promising strategy for successful tissue engineering. The required differentiation of stem cells is crucial for functionality of engineered tissues and can be regulated by chemical and physical cues. Here we examined how boundary cap neural crest stem cells (bNCSCs) are affected when cultured in the same medium, but on collagen-or laminin-polyacrylamide (PAA) scaffolds of different stiffness (0.5, 1, or ~7 kPa). bNCSCs displayed marked differences in their ability to attach, maintain a large cell population and differentiate, depending on scaffold stiffness. These findings show that the design of physical cues is an important parameter to achieve optimal stem cell properties for tissue repair and engineering.


| INTRODUCTION
Stem cells are attractive for regenerative medicine as a source for cell replacement therapy, as well as for their potential to modify disease processes and promote tissue repair.The interplay between stem cells and their extracellular environment plays a crucial role in their survival, differentiation, and functionality (Burdick & Vunjak-Novakovic, 2009).This interplay occurs mainly through molecular interactions, but mechanical conditions such as extracellular matrix surface properties also play an important role.Understanding how chemical and physical cues cooperate in neural stem cell differentiation is therefore critical for optimal application of stem cell-based therapy.
In addition to their robust survival and broad differentiation potential, they display a remarkable ability to promote survival and support the function of different types of co-cultured cells.Thus, bNCSCs stimulate proliferation of co-cultured (Grouwels et al., 2012;Ngamjariyawat, Turpaev, Vasylovska, Kozlova, & Welsh, 2013;Ngamjariyawat, Turpaev, Welsh, & Kozlova, 2012;Wang et al., 2016) and co-implanted insulinproducing pancreatic beta-cells (Lau, Vasylovska, Kozlova, & Carlsson, 2015;Olerud et al., 2009).Co-transplantation of bNCSCs and human pancreatic islets increased islet vascularization and innervation, most likely mediated by diffusible factors released from the bNCSCs (Grapensparr et al., 2015;Lau et al., 2015).bNCSCs also promote survival of excitotoxically challenged spinal cord neurons in vitro (Schizas et al., 2018), showing that their beneficial effects are not limited to pancreatic beta-cells.bNCSC-derived astrocytes were able to protect motor neurons with a mutation associated with amyotrophic lateral sclerosis (ALS) from oxidative stress to a degree surpassing that of embryonic stem cell-derived astrocytes (Aggarwal, Hoeber, Ivert, Vasylovska, & Kozlova, 2017).At the same time, bNCSC-derived astrocytes have shown extreme resistance to oxidative stress (Aggarwal et al., 2017), and other environmental challenges (own, unpublished data).Recently, injection of bNCSCs into the spinal cord ventral horn of an ALS mouse model was shown to delay disease progression (Leyton-Jaimes et al, in preparation for publication).
It was previously demonstrated that neural stem cell differentiation depends on the stiffness of the scaffolds.Thus oligodendrocytes required material with high stiffness (~7 kPa), astrocytes medium (1-3 kPa) and neurons soft surfaces (<1 kPa;Thompson & Chan, 2016).These findings suggest that surfaces of certain stiffness can improve the specificity and homogeneity of differentiated neural stem cells.This interpretation is supported by recent findings showing that human induced pluripotent stem cell-derived neural crest stem cells implanted into the rat carotid artery preferentially differentiate to smooth muscle cells in a stiff environment, and to glial cells under softer conditions (Zhu et al., 2019).
Here we investigate how the stiffness of the material and the covering substrate affect bNCSC adhesion, maintenance, and differentiation and show that polyacrylamide (PAA) scaffolds of different stiffness covered with collagen or laminin influence these properties of bNCSCs in 2D cultures.

| MATERIALS AND METHODS
2.1 | Culture of bNCSCs bNCSCs were generated from transgenic mice harboring red fluorescent protein (RFP) under the universal actin promoter (Vintersten et al., 2004) according to previously published protocols (Aldskogius et al., 2009;Hjerling-Leffler et al., 2005).Briefly, the dorsal root ganglia along with boundary caps were mechanically separated from the isolated spinal cord and mechanoenzymatically dissociated using Collagenase/Dispase (1 mg/ml) and DNase (0.5 mg/ml) for 30 min at room temperature.Cells were plated at 0.5-1 × 10 5 cells/cm 2 in an N2 medium containing B27 (Gibco) as well as EGF and bFGF (R&D Systems, 20 ng/ml, respectively).After 12 hr, nonadherent cells were removed together with half of the medium before adding fresh medium.The medium was changed every other day (50% of the medium replaced with fresh medium) until neurospheres could be observed after approximately 2 weeks of culture.The cells were frozen and after thawing the dissociated cells were plated in 24 welldishes or 75 cm 2 flasks for the following experiments.

| Cell adhesion assay
For cell adhesion analysis 1.0 × 10 4 cells/cm 2 were seeded in each well of a 24-well plate with laminin-or collagen-PAA substrates (1 μg/cm 2 ), and allowed to attach and grow for 48 hr.Cells were examined and counted in a fluorescent inverted phase-contrast microscope.Thereafter, the cultures were washed three times with PBS to remove nonadherent cells, and the remaining (attached) cells counted.

| Cell counts
To assess the ability of the different substrates to maintain the bNCSC population, the number of bNCSC cultured on PAA with different stiffness on collagen or laminin surfaces, was counted daily from Day 1 to Day 5. Cultures were photographed and the number of cells estimated by virtual measurement, using ImageJ.

| Cell differentiation
To estimate bNCSCs differentiation we counted the proportion of cells with an elongated profile (sign of dearly differentiation) in relation to the number of cells with a round profile (sign of undifferentiated state).The proportion of elongated and round cells was determined on Days 1, 2, and 3 on collagen-and laminin-PAA with different stiffness.After 5 days in culture, cells were fixed in 4% paraformaldehyde in PBS for 15 min, and processed for immunocytochemistry.Cells on glass coverslips were pre-incubated with blocking solution (1% bovine serum albumin, 0.3% Triton X-100, and 0.1% sodium azide [NaN3] in PBS) for 45 min at room temperature and then incubated overnight at 4 C with antibodies to glial fibrillary acidic protein (GFAP; mouse,1:400, Merck).Secondary antibodies were Alexa Fluor 488 goat anti-mouse IgG (H + L; 1:200, Life Technologies).Cells were washed twice with PBS, once in distilled water and embedded in 8 μl of mounting medium (50% glycerol in PBS and 100 mM propyl-gallate [Sigma]) on a glass slide.

| Statistical analysis
Results are presented as means ± SD.The difference among samples was considered statistically significant by the confidence interval of 95% (p < .05).

| Cell adhesion
To examine cell adhesion on collagen or laminin surfaces with different PAA stiffness, bNCSCs were cultured for 2 days before washing, and the total number of attached cells was counted.We found that the majority of cells attached to the same efficiency to laminin surfaces regardless of PAA stiffness (Figure 1).The lowest adhesion efficiency was registered on 0.5 kPa PAA covered with collagen.Collagen surface was also less efficient than laminin with a PAA stiffness of 1 kPa.These data show that surface properties (collagen or laminin) are most important for scaffolds with low stiffness, whereas, for higher stiffness, the type of surface is less important for cell adhesion.

| Cell number
Cell numbers on collagen or laminin surfaces with different PAA stiffness was assessed every 24 hr from Day 1-5 after attachment.On the collagen surface, a stiffness of ~7 kPa displayed the best ability to maintain bNCSCs (Figure 2a).Between Day 1 and 2, there was a significant increase in the number of cells on this substrate, indicating that the original cell population had grown through proliferation.This increase was abolished after washing the cultures on Day 2. The loss of cells after washing is presumably the result of poor adhesion by a proportion of living cells, which, in turn, could reflect their specific receptor pattern expression.A second increase in the bNCSC population on collagen with a stiffness of ~7 kPa occurred between Day 3 and 4.This rise was, however, not maintained on Day 5, when the bNCSC population was similar in size to Day 1. Collagen surfaces with lower stiffness were less favorable in terms of maintaining the bNCSC population (Figure 2a).On laminin surface with stiffness 0.5 kPa, the cell number was low 24 hr after plating, indicating an early loss of a large proportion of seeded cells (Figure 2b).From that stage, the cell number was maintained, however.For laminin substrate with a stiffness of 1 kPa, the cell number was essentially unchanged throughout the experiment.When the stiffness of laminin was increased to ~7 kPa it was able not only to maintain but also to support growth of the bNCSC population (Figure 2b).A significant increase occurred on this substrate from Day 1 to Day 2, and the loss of cells due to washing on Day 2 was followed by a second significant increase in bNCSC numbers.

| Cell differentiation
We used transformation of bNCSCs from a round to an elongated shape as evidence of cell differentiation from Day 1 to Day 3. The proportion of elongated cells on collagen-covered ~7 kPa PAA was markedly higher than on collagen surfaces with lower PAA stiffness (Figure 3a).All three types of laminin-covered PAA support the F I G U R E 1 bNCSC adhesion rate on PAA scaffold with different stiffness (0.5, 1, and ~7 kPa) covered with collagen or laminin *p < .05;***p < .0014).Thus, on a PAA stiffness of 0.5 and ~7 kPa PAA the proportion of elongated cells reached a peak on Day 2, and then declined.However, the number of differentiated cells on laminin-covered 1 kPa PAA was larger than on the other two types of stiffness on Day 3. Labeling for the astroglia marker GFAP showed that on laminin a large majority of cells displayed an astroglial phenotype with the largest proportion on 1 kPa PAA stiffness (Figure 5).A few cells were also observed to be GFAP positive, in line with the expression of GFAP in early differentiating astrocytes (Bramanti, Tomassoni, Avitabile, Amenta, & Avola, 2010).

| DISCUSSION
Cells communicate with the environment and other cells through receptors and ion channels, which are located on the surface membrane.Therefore, their ability to attach and elongate on the surface is a crucial factor for their proliferation and differentiation capacity.This is particularly important in view of the broad cell supportive potential of bNCSC in 3D bio-printing for cell survival and support.Furthermore, the interaction cells and scaffold properties may also display a dynamic temporal pattern that needs to be considered in the design of optimal bioscaffold function.
Here, we investigated at different time points in vitro adhesion, maintenance, and differentiation capacity of bNCSCs on scaffolds with different stiffness covered with collagen or laminin.We show that the scaffold porosity, and its interaction with surface properties of the biomatrix (collagen or laminin), has a major impact on bNCSC attachment, maintenance, and differentiation.
Previous studies have shown that embryonic stem cells grown on stiff substrates maintain their proliferative capacity, but undergo  (Panciera, Azzolin, Cordenonsi, & Piccolo, 2017).However, there is not a straightforward relationship between mechanical conditions in the environment and stem cell differentiation, since, for example, mesenchymal stem cells (McBeath, Pirone, Nelson, Bhadriraju, & Chen, 2004;Panciera et al., 2017) and neural crest stem cells (Zhu et al., 2019) differentiate along different lineages depending on surrounding mechanical factors.Our findings show that the conditions that favor stem cell differentiation are also markedly affected by the molecular properties of the substrate itself.Thus, whereas soft collagen substrate is unfavorable for bNCSC adhesion, laminin substrates are compatible with a high degree of bNCSC adhesion regardless of scaffold stiffness.At the same time, we have detected that washing on Day 2 essentially reduced the cell number on soft collagen substrates and on stiff laminin substrate.This shows that the ability of bNCSCs to adhere to substrates with markedly different level of stiffness depends on the biomolecules of the substrate-collagen or laminin.Soft collagen substrate also shows inferior capacity to maintain the bNCSC population, indicating a mechanistic relationship between adhesion and bNCSC survival.Furthermore, collagen with 1 kPa PAA, a substrate that provided excellent conditions for bNCSC adhesion, failed to support their elongation, and differentiation.
Laminin provided good conditions for cell adhesion regardless of PAA stiffness, although with a moderate, but still significant, decline with the highest scaffold stiffness.Interestingly, this decline did not negatively affect bNCSC maintenance, which was, in fact, best on high stiffness laminin substrate.All laminin covered stiffness variants also supported differentiation of bNCSC, as evidenced by a large proportion of elongated cells during the culture period, and the presence of more than 90% GFAP expressing cells under all stiffness conditions at the termination of the experiment.

| CONCLUSION
Taken together, our finding demonstrates the complex interactions between chemical and physical cues for the outcome of neural stem cell properties and indicates that designing optimal substrate and bioscaffold interactions are fundamental for successful tissue engineering.
Cell number on scaffolds of different PAA stiffness with (a) collagen surface and (b) laminin surface from Day 1-5.n = 4; *p < .05;***p < .001emergence of a large proportion of elongated cells, although with a somewhat different temporal pattern (Figure

F
I G U R E 3 Differentiation of bNCSCs on (a,c) collagen and (b,d) laminin surfaces with different scaffold stiffness (0.5, 1, and ~7 kPa).The graphs display the proportion of elongated cells in relation to the total cell number at different time points F I G U R E 4 Images of elongated bNCSCs, expressing RFP on Day 3 in the scaffold with different stiffnesses.Bar = 25 μm

F
I G U R E 5 Cell differentiation on laminin covered PAA of different stiffness (0.5, 1, and ~7 kPa) for 5 days.(a) GFAP (green) expressing bNCSCs (red) growing on laminin-covered scaffolds.Elongated (arrow) and round (arrowhead) cells are present on all three stiffness variants.Blue-Hoechst nuclear stain.(b-d) Percentage of GFAP-positive cells on laminin-covered 0.5 kPa PAA (b), 1 kPa PAA (c), and ~7 kPa PAA (d).Bar = 25 μm differentiation when the strength of mechanical signals are reduced