A Polymer Blend Substrate for Skeletal Muscle Cells Alignment and Photostimulation

Substrate engineering for steering cell growth is a wide and well-established area of research in the field of modern biotechnology. Here we introduce a micromachining technique to pattern an inert, transparent polymer matrix blended with a photoactive polymer. We demonstrate that the obtained scaffold combines the capability to align with that to photostimulate living cells. This technology can open up new and promising applications, especially where cell alignment is required to trigger specific biological functions, e.g. generate powerful and efficient muscle contractions following an external stimulus.


Introduction
In tissue engineering, cell-substrate coupling has a crucial role. Its mechanical, chemical, physical and morphological features are decoded by cells as stimuli that strongly affect their behavior. Often tools such as electrodes are further added to the cell culture in order to probe or actively stimulate the cellular processes. The use of electrodes has large practical success but it does bear a number of downsides. Alternatively, photostimulation provides advantages such as better spatial and temporal resolution, low toxicity and invasiveness. In addition, the stimulation map throughout the sample can be changed easily without the need of rewiring nor the limit of discrete points, owing to the wireless architecture of this optical strategy. The use of light to control living cells and organisms has a long history, but it is regaining interest due to the emerging of new techniques at the frontier between photonics and biology. Optogenetics, [1] NIR [2] and phototransducers based [3] stimulation are the most popular and promising techniques. In particular, the use of non-genetic phototransducers has the highest translational potential. The phototransducers may come in a variety of shapes and structure, such as plasmonic nanodots, [4,5] inorganic semiconductors devices, [6][7][8] nanostructures, [9][10][11][12] organic semiconductors [13][14][15][16][17] and small molecules [18,19] . Thanks to the similarity in composition and structure with biomolecules, organic semiconductors and molecular compounds express high biocompatibility and multifunctional actuation. Their peculiar photophysical properties allow the employment of different photostimulation mechanisms (i.e. electrical, mechanical or thermal) and at the same time, their mechanical flexibility favors conformational matching with biological environment. These properties make organic materials suitable for in vivo applications, where a soft yet efficient interaction with the biological counterpart is extremely important. [20,21] Furthermore, by using wet processes at room temperature, organic semiconductors are easily fabricated into a great number of forms whose morphology can mimic those of the extracellular matrix and improve both cell growth and cell coupling at the abiotic/biotic interface. [22,23] In particular, in muscular tissue engineering, in order to efficiently reproduce the natural muscle organization and achieve contraction ability, cell alignment on a suitably stiff substrate is required. This can be done by depositing extra matrix proteins [24][25][26][27] or polymer fibers that produce an aligned path to force cell's orientation. [28][29][30] Alternatively, a simpler approach relies on the fabrication of microgrooves onto a substrate that physically confine the cell. [31][32][33][34][35][36] There are no universal rules that define which pattern can lead to an optimal morphological stimulation, since it strongly depends on the target cell. Nevertheless, some guidelines can be identified, regarding both the width and depth of the pattern. Indeed, the depth should be ideally higher than the cell's height or at least of the order of micron. [37] The width can spun over different values depending on the application and the target cell. As a general rule in order to align a single cell, the width of a pattern should be lower than the cell's length that can be several times its height.
In this work we present a polymer substrate engineering designed to perform a double functionality: it drives cell alignment and allows the photostimulation of cell behavior. We chose Poly-(3hexylthiophene-2,5-diyl) (P3HT), a highly studied polymer, as photo-active material. Upon blending it with high density polyethylene (HDPE) [38] we obtained a free standing polymer film with good mechanical properties that preserves P3HT optoelectronic features. By using a maskless laser ablation approach we were able to inscribe a customizable guiding pattern in the films suitable for achieving aligned cell growth. The combination of blending and micromachining is used as a novel approach to achieve multifunctional substrate without the employment of complex and expensive fabrication methods such as photolithography. Moreover, the upscaling of this approach can be easily achieved in order to increase the complexity of the realized pattern. As a simple model system for preliminary biocompatibility and photostimulation studies we used Human embryonic kidney (HEK) cells. Once parameters had been optimized we focused on to C2C12, a murine myoblast cell line characterized by specific biophysical parameter, like myotube diameter ranging from 10 µm to 25 µm and length ranging from 130 µm to 520 µm. [39] C2C12 show many peculiar properties of skeletal muscle cells (i.e. the ability to differentiate toward multinucleated myotubes that can display contraction) thus making them a widely accepted in vitro model to validate our purpose.

Result and Discussion
It has been widely reported in literature that P3HT can be blended with a large variety of insulating polymers preserving its optoelectronics properties. [40][41][42][43][44] The blend strategy is employed to introduce desired processing and mechanical properties of commodity polymers (i.e. common plastics) in organic electronics applications. Plastic "additives" are ductile, display high elongation at break, while in contrast semiconducting polymers are typically rather brittle owing to their low molecular weight. [43] Moreover, insulating polymers have been reported to be beneficial in a number of applications leading for instance to improved charge carriers mobilities, increased device stability, [45][46][47] tailored optical properties, [48,49] decreased energetic disorder. [50] We focus here on HDPE, a semicrystalline insulating polymer that has been already employed in combination with P3HT to produce both thin and thick films for a large variety of applications. Remarkably, the blend system retains semiconducting optoelectronic properties even at very low concentrations of the photoactive material in the insulating matrix, while showing improved mechanical properties. Indeed, the improved viscoelastic properties of the blend system allow the production of robust self-standing films that can be used directly in tissue engineering, while films comprising neat semiconducting polymers would be too brittle to do so. Beside the combination of mechanical and optoelectronic properties, blending has an economic advantage for HDPE being much cheaper than P3HT, which is particularly useful when producing thick films (tens of microns).

Fabrication Process
The fabrication process is designed to obtain a photo-excitable substrate suitable for laser pattern ablation. We assumed that a surface pattern would be able to induce a satisfactory cell alignment if it had a characteristic depth higher than the cell diameter and a width shorter than the cell length, as we discussed in the previous section (for a muscle cell height ≈ 15 µm and length ≈ 200 µm). [39] Following these requirements, we selected a pattern geometry formed by parallel lines with a width of 60 µm and a depth of at least 15 µm, thus requiring a film thickness >20 µm. The fabrication process ( Figure 1) is based on a pre-compounding step performed in solution, followed by solid-state processing of the self-standing film and final laser patterning. In principle, the solution processing could be avoided and the polymer:polymer compounding could be achieved upon melt mixing P3HT and HDPE, leading to a solvent free process. However, the latter requires the employment of a large volume of materials and high temperature of processing owing to the high melting point of P3HT (ca. T = 238°C). [51] Consequently, the solution-based method is more practical. Once a homogeneous solution is achieved, it is drop cast to obtain a solid state deposit. Despite this deposit exhibits the required self-standing and mechanical properties, it lacks the necessary thickness and homogeneity. In order to reach a suitable film thickness, around 30 µm, we delaminated, folded and pressed the polymer blend into a uniform film. The blend strategy is particularly useful, not only to tune the mechanical properties of the films, but also to control the optical density. Indeed, HDPE does not absorb in the UV-vis range (see Figure S1) and when intermixed with the semiconductor it enhances the transparency of the otherwise opaque P3HT thick film. Films obtained with a concentration of P3HT in the range 1-5 wt% show an optical density suitable for photostimulation experiments at the required thickness of 30 µm.

Photophysical Properties
The optical properties and elementary excitation dynamics of the P3HT:HDPE films were investigated by measuring absorbance, photoluminescence ( Figure S1) and transient photoinduced absorption (Figure 2) in a spin-cast P3HT thin film (P3HT-TF) and for three different HDPE:P3HT pressed blends (with P3HT concentration equal to 1, 3 and 5 wt%). Note that HDPE is virtually transparent in the considered spectral range (400 -800 nm) as shown in Figure S1, thus spectroscopy essentially addresses P3HT properties. [52][53][54][55][56] The optical properties of the three blend samples are very similar to that of the neat film suggesting that: a) the fabrication process has a minor effect onto the P3HT photophysics and b) P3HT successfully phase separates from HDPE in the blend (suggested by the highly visible vibronic structure). Indeed, it has been well established how the processing conditions, and in particular the deposition temperature, play a critical role in insulating:semiconducting polymer blends not to hamper the solid state optoelectronic properties of the semiconductor. [57][58][59] Here, we selected a deposition temperature of 100 °C, in a range in which P3HT is known to phase separate into micro-crystallites prior the HDPE crystallization during the solvent evaporation, leading to an unperturbed UV-vis absorption of the semiconductor. The ordered nature of P3HT is confirmed by the highly structured absorption spectra, typical of crystalline aggregates of this polymer. By and large the transient absorption data confirm this picture, even though quantitative differences are present depending on the blend composition. In particular, in the 1% blend the photoinduced absorption (PA) band ( T < 0) appears much wider than in the other spectra, suggesting a more efficient charge separation process. In the 3% and 5% samples the first vibronic resonance of the bleaching band ( T > 0) is enhanced suggesting a higher degree of order in the P3HT domains. This hypothesis is enforced by the red shift of the emission spectra as a function of the P3HT concentration that suggests an enhanced aggregation as well as a planarization of the P3HT molecules (Figure S1). Despite a more detailed analysis is beyond the scope of this work, we can speculate that the crystal packing of P3HT chains secluded in the blend might favor singlet fission into triplets. Overall, we conclude that P3HT properties in the blend are similar to those in the neat film, possibly with higher degree of planarization that enhances inter chain coupling and in turn the crystal versus amorphous contribution.

Substrate Biocompatibility
The biocompatibility of the substrate and of the patterning process were tested with two distinct assays. For substrate biocompatibility the alamarBlue proliferation test was performed with HEK cells (Figure 3b). These cells were preferred for this kind of methodology due their dimension, which is lower than C2C12 and therefore more suitable for acquiring data for longer proliferation times and for choosing a higher seeding density, a crucial aspect for this test. The biocompatibility of the patterning process was also tested, to be sure that laser ablation did not produce toxic fabrication debris.

Conversely, C2C12 viability was evaluated with the HOECHST 33342/NucGreen Dead 488
ReadyProbes assay (Figure 3c). Both the film and the patterning process highlight very good biocompatibility showing proliferation and viability rates similar to those obtained seeding the cells on glass, that represents our standard control condition.

Cell Photostimulation
Having established that the blend films retain their optoelectronic properties and have suitable biocompatibility, we went on studying the performance in cell photo-stimulation. The electrophysiological properties were measured by carrying out patch clamp experiments in whole-cell configuration and current clamp mode (I = 0). Using this approach, changes of the cell membrane potential are recorded during light excitation. As we showed in previous works, [60,61]   We detected a transient depolarization upon illumination with short pulses, quickly evolving into hyperpolarization when the light was switched off. The membrane potential recovered its resting value approximately after 300 ms (Figure 4a). By using longer pulses the initial depolarization turned into hyperpolarization during illumination. At the light offset there was again an overshoot increase of the hyperpolarization, followed by a recovery to the initial resting potential in about 300 ms (Figure   4b). A very good reproducibility of the data is observed, as it can be noticed from the low dispersion of the data in both the cell lines and the pulse length shown in Figure 4 b-d. We performed measurements on a control sample (black line Figure 4) without observing any signal during measurement, ruling out a direct interaction between cell or HDPE with light. Once the photoinduced effect has been demonstrated, we performed further studies on the biophysical mechanisms. In previous works, the photoinduced membrane potential variation, measured in cells seeded on bare P3HT films and subjected to light stimulation protocol similar to the one adopted here, was assigned to a thermal effect. In particular, the initial depolarization is reported to be due to the increase of capacitance of the cell membrane, while the following hyperpolarization, still under light exposure, is assigned to the change in the cell baseline. [60] The dark evolution towards equilibrium is ascribed to the recovery of capacitance and cooling off. In principle our blend could have a different thermodynamic behavior under illumination when compared with the neat thin film. We therefore measured the temperature variation in close proximity of the substrate, using the calibrated pipette method (Figure 5, a-b) described by Martino et al. [54] Effectively, upon irradiation with a 200 ms light pulse, the temperature rose up to 33°C, with ΔTBlend = +9°C, which is comparable to that observed on neat P3HT thin films, ΔTP3HT-TF = +7°C with a comparable power density (≈40 mW/mm 2 ). Apparently, absorption in the blend is large enough to give rise to a change in temperature that slightly overcomes that observed in the neat film, despite the low P3HT density in the blend. This could probably be due to the different thickness and the different thermal conductivity of the two films. To confirm the mechanism reported above we also measured the capacitance in dark condition and upon light excitation by applying the double sinusoidal technique. Figure 5-c reports the results.
As expected, the capacitance increases due to light excitation in both the investigated cell lines. The percentage increase, ΔC, is 4.87 ± 1.62% for the HEK cells and 6.40 ± 0.85% for C2C12 cells.

Laser Ablation and Cell Orientation
Once the cell photostimulation ability of the plain pressed blend was assessed, we further aimed at micromachining its surface in order to force cell alignment. The manufacturing process creates a grooved line pattern on our blend substrate taking advantage from the ultra-short pulse laser direct surface ablation technique. This innovative and maskless fabrication technique gives a high degree of freedom in designing both the shape and the wall dimension of such patterns, allowing geometries that are more complex. By changing the repetition rate, the pulse energy or the position of the laser beam focal spot it is possible to tune the wall depth, the geometry (which can also be curvilinear) and control the spatial resolution of the chosen pattern (as shown in Figure S3). Furthermore, direct laser ablation provides a quicker and cheaper manufacturing technique, compared to classical lithographic methods, that well matches the cost-effectiveness of the blended substrates presented in this work. We selected a simple square cross-section target pattern with a wall depth greater than 20 µm and a width of 60 µm (Figure 6). We then assessed the ability of the realized pattern to align cells with Scanning electron microscopy (SEM), by acquiring images of C2C12 cells seeded directly on the substrates. To better visualized the effect of the pattern we first selected the experimental conditions that allowed the growth of isolated cells. A representative image acquired in this condition is shown in Figure 7-a, where it is possible to appreciate how the substrate pattern effectively steers the cells orientation. In  The mean orientation angle is 12.95°±1.52°. Moreover, a good alignment efficiency is observed, indeed 77% of the cell's directions are included in the angular interval 0°-15° and 90.3% in 0°-30°.
Having established that cell growth anisotropy is achieved in the low-density cell regime, we carried out also high cell density experiments trying to achieve a muscular tissue (see supporting materials).Despite it is hard to obtain a quantitative evaluation of the alignment; nevertheless it is still possible to observe that the cell growth follows the pattern direction (see Figure S5). The effect is particularly evident at the edge of the pattern where is possible to appreciate an evolution of the cells from a random to an aligned orientation condition canalized by the pattern.

Conclusion and Perspective
Substrate mediated cell engineering for both growth control and photostimulation is a wide and wellestablished area within biotechnology research. Even though the two sub-fields share common application paths and are focused on a similar biophysical interaction (cell-substrate coupling), they are normally studied separately by researchers with different expertise. Here we combine, for the first time, the ability to align and photostimulate cells within the same substrate. Thanks to the features of the fs-micromachining facility it is possible to obtain patterning of any shape and size, without mask, easily and quickly adaptable to each series of different cells. The same setup also allows a highresolution soft-cut of the material to give a macro customized shape to the patterned areas, if and when necessary. This is an enabling technology that can open up new applications, especially when cell alignment is mandatory to carry out specific functions, e.g. muscle contraction. We demonstrated both successful alignment and photostimulation. Ideally, exploiting this novel and multitasking approach will permit to fabricate systems in which fully developed skeletal or cardiac muscle cells can grow in an organized array that allows carrying out specific functions under light control. Further developments of this work regard replacement of both the photoactive and the insulator material. The former aims to move the optical absorption spectra towards wavelengths that can penetrate deeper into the tissue in order to broaden the spectrum of possible biological applications. The insulator instead can be replaced with hydrogel or elastomeric material to well reproduce biological stiffness.
Moreover, the obtained blend, in its solution step, and the mentioned possible ameliorations could be optimized to exhibit ink-properties compatible with 3D-printing techniques, thus aiming to reproduce the structural complexity of living tissue.

Sample Preparation
Regio-regular P3HT (99.995% purity, 20000-45000 molecular weight) was purchased from Sigma Aldrich and used without any further purification. High Density Polyethylene (125000 molecular weight) was purchased from Alfa Aesar and used without any further purification.
Solutions of the two material (P3HT and HDPE) were prepared in di-chlorobenzene at concentration of 15 g/l, mixed in a specific ratio (1-99%; 3-97% and 5-95%) and stirred for at least 30 minutes. The obtained mixing solution was drop casted on a hot plate at 100°C. Once the solvent was evaporated the obtained film was folded and pressed with a hydraulic press with several step to arrive to the maximum pressure of 7 MPa held for 8 minutes. Although the solution step is not required and the exact amount of powder can be directly pressed to obtain the film, in this work we decided to go through solution to be more precise in the mixing step. The obtained film was circular with a diameter of 18 mm and thickness of 30 µm.

PL measurements
The PL measurements were carried out with a Horiba Nanolog Fluorimeter equipped with two detectors (photomultiplier and InGaAs). The sample was excited at 530 nm using the same LED lamp and power employed for transmission measurements (see above).

Ultrafast time-resolved spectroscopy
For the femtosecond TA measurements, we used a Ti:Sapphire laser with a repetition rate of 1 kHz and a pulse width of 100-150 fs. We excited the films at 530 nm. The beam was generated by means of an optical parameter amplifier and probed with a white light beam generated by a CaF2 plate. The excitation energy was of ≈ 30 µJ and the beam spot size of ≈ 200 µm in diameter.

Laser Pattern
The micromachining setup consists of an amplified Yb:KGW Excess fibronectin was then removed by rinses with PBS.

Viability assay
To preliminarily evaluate the substrate cytotoxicity, alamarBlue proliferation assay, that allows a continuous monitoring of cells in culture, was performed on HEK cells. Briefly, the alamarBlue Reagent (Invitrogen DAL 1100) was diluted 11:1 with DMEM without phenol red. 500 µl of the obtained solution were added on each well. The solution was incubated for 3 hours at physiological condition. Once removed from the well, the emission at 590 nm of three aliquots per each well was measured. The emission value was acquired 5 times per each aliquot to obtain a reliable measure.
Furthermore, the biocompatibility of the patterning process was accomplished on the more relevant biological model C2C12 via HOECHST 33342/NucGreen Dead 488 ReadyProbes assay. The substrates were incubated in extracellular containing the two dyes (HOECHST 33342 (10 µg ml -1 ) and NucGreen Dead 488 ReadyProbes Reagent (2 drops ml -1 )) for 5 minutes protected from ambient light. The samples were then washed with extracellular solution and multiple images were acquired with a Nikon Eclipse Ti-S epifluorescence inverted microscope. Standard DAPI and FITC filter sets were employed for HOECHST and NucGreen respectively. The percentage of viable cells was estimated by counting the total number of cells nuclei (stained by HOECHST) and the total number of dead cells nuclei (stained by NucGreen).
All SEM images were acquired by using a TESCAN MIRA III scanning electron microscope (operating voltage 3 kV, working distance 7 mm, stage tilt angle 0°). Only single cells were selected for recordings. Acquisition was performed with pClamp-10 software (Axon Instruments). Membrane currents were low pass filtered at 2 kHz and digitized with a sampling rate of 10 kHz (Digidata 1440 A, Molecular Devices). Data were analyzed with Origin 9.0 (OriginLab Corporation) and with Matlab software.

Electrophysiology
The light source for excitation was provided by a green LED coupled to the fluorescence port of the microscope and characterized by maximum emission wavelength at 530 nm; the illuminated spot on the sample has an area of 0.23 mm 2 and a photoexcitation density of 40 mW/mm 2 , as measured at the output of the microscope objective (Pobj).

Cell Capacitance Measurement
A double sinusoidal voltage-clamp signal is applied to the cell in whole-cell configuration. The response current signal is acquired and membrane capacitance, membrane resistance and access resistance are then extracted fitting the current with a custom Matlab program.

Statistical analysis
Data are represented as mean ± standard error of the mean (s.e.m.). Normal distribution was assessed