Bi-layered Tubular Microfiber Scaffolds as Functional Templates for Engineering Human Intestinal Smooth Muscle Tissue

Abstract

Designing biomimetic scaffolds with in vivo-like microenvironments using biomaterials is an essential component of successful tissue engineering approaches. The intestinal smooth muscle layers exhibit a complex tubular structure consisting of two concentric muscle layers in which the inner circular layer is orthogonally oriented to the outer longitudinal layer. Here, we present a three-dimensional (3D) bi-layered tubular scaffold based on flexible, mechanically robust and well aligned silk protein microfibers to mimic native human intestinal smooth muscle structure. The scaffolds were seeded with primary human intestinal smooth muscle cells to replicate human intestinal muscle tissues in vitro. Characterization of the tissue constructs revealed good biocompatibility and support for cell alignment and elongation in the different scaffold layers to enhance cell differentiation and functions. Furthermore, the engineered smooth muscle constructs supported oriented neurite outgrowth, a requisite step to achieve functional innervation. These results suggested these microfiber scaffolds as functional templates for in vitro regeneration of human intestinal smooth muscle systems. The scaffolding provides a crucial step toward engineering functional human intestinal tissue in vitro, as well as for the engineering of many other types of smooth muscles in terms of their similar phenotypes. Such utility may lead to a better understanding of smooth muscle associated diseases and treatments.

Introduction

The human intestine is a tubular organ that facilitates food digestion, absorbs water, nutrients and minerals, and defends against pathogens^[1]. Intestinal diseases, such as inflammatory bowel disease, short bowel syndrome and cancers have affected over 60 million and cost over $100 billion/year in direct medical expenses in the United States alone ^[2]. Tissue engineering strategies using biomaterial scaffolds with human cells offer a potential approach to restore tissue functions and for the investigation of human diseases and treatments by building engineered tissue models in vitro ^[3]. Tissue-engineered intestine models provide systems to generate functional intestine tissue to study intestinal diseases and for utility in drug screening, exploiting different scaffolding approaches to recreate key structural and functional components of the intestine in vitro ^[4]. The majority of these efforts have focused on intestinal epithelium as a key location for transport and related functions ^[5]. However, the design of functional scaffolds for engineering intestinal smooth muscle remains relatively unexplored and challenging due to the unique and complex 3D architecture ^[6].

The intestinal wall surrounding the lumen consists of two layers of smooth muscles ^[6a]. The inner muscle layer is mostly composed of dense circumferentially oriented smooth muscle cells, while the outer layer is composed of longitudinally arranged smooth muscle cells. These two muscle layers provide the major structural support, mediate inflammation and immune activation, and contribute to the propulsive peristalsis functions (contraction and relaxation) along with enteric nervous system for the intestine ^[6a]. The dysfunction of the intestinal smooth muscle is involved in several diseases, including intestinal pseudo-obstruction, short bowel syndrome and Parkinson’s disease ^[7]. To engineer such well-organized tissue, natural and synthetic biomaterials have been fabricated as planar or hollow single-layered scaffold matrices for cell seeding, such as collagen sponges ^[8], composite scaffolds composed of collagen and chitosan ^[9], poly (glycolic acid) (PGA) scaffolds coated with growth factors^[10], and electrospun poly- ε-caprolactone (PCL) scaffolds ^[11]. From a fundamental perspective, the goal of biomaterial design in tissue engineering is to form 3D complex geometries and architectures for cell accommodation and to guide new tissue formation ^[12]. Therefore, a suitable scaffold template for intestinal smooth muscle engineering should permit the inner layer of the intestinal smooth muscle to align in the required circumferential orientation and form a hollow tube that can contract autonomously to propel luminal contents, while also allow for the alignment of the outer longitudinal smooth muscle in parallel sheets perpendicular to the inner circular muscle layer to support the mechanical and cell needs. In this regard, none of the current designs exhibit these native-like 3D architectures.

Here, our goal was to fabricate a new bilayer scaffold system to mimic the native tissue for the support of smooth muscle functions in our previously established tissue engineering intestine models in vitro ^[13]. To achieve this goal, we exploit silk, as a naturally derived protein-based biomaterial with excellent biocompatibility, biodegradability, and extraordinary mechanical properties ^[14]. Silk has been extensively exploited in various material formats for tissue engineering, including sponges, films, tubes, hydrogels and nanofibers, ^[15] and for use in 3D hierarchical scaffolds for mimicking the native extracellular matrices (ECM) for various tissues including bone ^[16], nerve^[17], small diameter vasculature ^[18], brain ^[19], fat ^[15a], lung ^[20], cornea ^[21] and others. We have previously adopted 3D silk sponges for the in vitro remodeling of the human intestinal epithelium ^[13], to support enteric innervation ^[22], to study inflammation ^[23], and for interactions related to bacterial and parasitic infections ^{[13a, 24]}.

As a step toward engineering a functional intestine tissue in the lab, in the present study, we utilized a directional freezing technique ^[25] to fabricate silk into aligned microfibers to mimic the topographical features of muscle fibers in the intestine. The microfiber sheets were then trimmed and reformed into bi-layered tubular microfiber templates that resembled the native extracellular matrix (ECM) structures in the intestinal wall. These silk microfiber-based scaffold consisted of an inner layer composed of circumferentially aligned silk microfibers and an outer layer composed of longitudinally distributed silk microfibers. We then investigated the morphological, chemical, and mechanical properties of these bi-layered smooth muscle templates, as well as the in vitro affinity for primary human intestinal smooth muscle cells and human induced neural stem cells.

Experimental section

Preparation of aligned silk microfiber scaffold.

Bombyx mori (B. mori) cocoons were cut into small pieces and boiled in an aqueous 0.02 M Na₂CO₃ (Sigma-Aldrich, USA) for 30 min, followed by rinsing in distilled water to remove the Na₂CO₃ and sericin. The degummed silk was allowed to dry at room temperature for overnight. The dried silk was dissolved in 9.3 M LiBr solution at 60°C for 3–4 h. The solution was subsequently dialyzed for 3 days in distilled water using Slide-a-Lyzer dialysis cassettes (MWCO 3,500, Thermo Fisher Scientific). The water was changed five times during the dialysis (1h, 4h, 8h, 24h, 48h). After dialysis, the solution was centrifuged twice for 20 min at 9,000 rpm to remove insoluble impurities. The concentration of the final silk solution was determined by measuring a volume of solution and the final dried weight (~6 wt%). The silk solution was diluted to 2 wt% and then transferred to predesigned mold, followed by directional freezing using liquid nitrogen. The frozen silk solution was then lyophilized at −80°C. After that, the dry silk scaffolds composed of aligned silk microfibers were treated with methanol overnight and then washed with DI water several times to remove residual methanol. The hydrated scaffolds were then cut into predesigned shapes, followed by wrapping two aligned silk microfiber scaffolds concentrically around a rod sequentially to obtain the bi-layer scaffolds. The inner layer was wrapped along the parallel direction to the microfiber alignment while the outer layer was wrapped along the perpendicular direction to the microfiber alignment. Scaffolds were sterilized by autoclaving prior to cell seeding.

Morphological characterization.

The morphology of the scaffolds was characterized by Scanning Electron Microscopy (SEM) (EVO MA10 SEM and Ultra 55 field-emission SEM, ZEISS). The SEM images were collected with a voltage of 5 kV. The microfiber diameters were estimated by measuring 100 individual silk microfibers and gap distance between adjacent microfibers was determined by measuring 200 gap distances. The alignment of microfibers was analyzed by measuring the orientation angles of 100 individual silk microfibers in a single image.

Structural characterization.

Fourier-transform Infrared spectroscopy (FTIR) was carried out on a JASCO FTIR 6200 spectrometer (JASCO, Tokyo, Japan) equipped with a MIRacle attenuated total reflectance (ATR) Ge crystal cell in absorbance mode. For each measurement, the spectrum was recorded with 32 scans and a resolution of 4.0 cm⁻¹. The secondary structural analysis was carried out by performing peak deconvolution over the amide I region (1600−1700 cm⁻¹) with the following parameters: 1618 cm⁻¹ (β-sheet), 1646 cm⁻¹ (random coil/helix), 1685 cm⁻¹ (β-turn), and 1698 cm⁻¹ (β-sheet).

Mechanical properties.

The mechanical properties of the silk scaffolds were characterized using an Instron 3366 (Instron Inc.) Hydrated scaffolds with a rectangular shape were loaded onto the frame with different directions of microfiber alignment at room temperature. Tensile tests were carried out at a rate of 5 mm/min. At least n=3 samples were used to calculate average modulus with standard deviation.

Cell culture.

Primary human intestinal smooth muscle cells (hiSMCs) were obtained from ScienCell and cultured in smooth muscle cell medium (SMCM, ScienCell). Cells were cultured in T-75 cm² tissue culture flasks (Corning), maintained at 37°C, 5% CO₂ humified atmosphere and harvested with 0.25% trypsin-EDTA (Gibco) prior to seeding. Cells at passage number 4–5 were used for the experiments. Human induced neural stem cells (hiNSCs) were generated and expanded as previously described ^[17]. Briefly, expanded hiNSCs were lifted off mouse embryonic fibroblasts (MEFs) using TrypLE Select and pelleted. The cell pellet was resuspended in hiNSC expansion media consisting of KnockOut Serum Replacement DMEM (Thermo Fisher), GlutaMax (Thermo Fisher), KnockOut SR (Thermo Fisher), Antibiotic-Antimycotic (Thermo Fisher) and 2-mercapto (Thermo Fisher), bFGF Basic (Thermo Fisher). The resuspended cell solution was vortexed and passed through a 40μm filter to achieve single cell suspension. The resulting single cells obtained from hiNSC colonies were then used for scaffold seeding. The media were changed every other day.

Cell seeding.

To produce cell-seeded intestinal smooth muscle equivalents, hiSMCs were trypsinized and resuspended in medium at approximately 5 million cells per mL and then delivered evenly over each inner and outer microfiber sheet separately. These seeded sheets were incubated for 1-hour to allow for cell attachment and then incubated in SMCM overnight for further cell settlement on the sheets. The second day, the inner and outer seeded microfiber sheets were assembled into bi-layered tubular microfiber scaffolds. The bi-layered tubular structure was created by rolling up an inner sheet around a Teflon coated stainless steel wire (Ø = 2 mm, McMaster-Carr) to produce the concentric structure followed by the rolling up of the outer sheet around the inner concentric structure. After the assembly of the scaffolds, the Teflon rods were removed. A small sterile needle tip was used to fix the outer layer of the scaffold in the first week. The bi-layered scaffolds were bound together without the aid of the needle tips a week post seeding during the rest of culture period. The medium was changed every other day over a 12-week period. For cell seeding, hiSMCs were pre-seeded on the scaffolds and cultured for 3 days, the outer layers were unwrapped, 1×10⁵ hiNSCs were evenly seeded on the surface between the inner and outer layers of the constructs containing hiSMCs, and the outer layers were rewrapped to reform tissues. The tissue constructs with hiSMCs and hiNSCs were then co-cultured for another 2 weeks for analysis.

Immunofluorescence and confocal imaging.

Tissue constructs were fixed with 4% paraformaldehyde (PFA, Santa Cruz). Silk scaffolds were unrolled in order to expose both inner and outer layers of the scaffolds to the blocking solutions and antibodies during the following incubation steps. All specimens were then permeabilized using 0.1% Triton x-100 (Sigma) in phosphate-buffered saline (PBS, Gibco), then blocked with 5% bovine serum albumin (BSA, Sigma) for 2 hours. These specimens were incubated overnight at 4°C with anti-Ki67 (BDbiosciences), anti-human-a-SMA (abcam), and anti-human-Calponin (abcam), and then immersed in Alexa Fluor 488 donkey anti-mouse and Alexa Fluor 546 goat-anti-rabbit secondary antibodies (Invitrogen) at a dilution of 1:100, respectively. Scaffolds were then counterstained with dihydrochloride (DAPI; Invitrogen) before being mounted using Vectashield mounting medium (Vector Laboratories). For live staining, calcein-AM (Invitrogen) was used at different time points, following manufacturer’s guidelines. The 3D scaffolds were scanned using a Leica SP2 confocal microscope (Leica Microsystems) and Nikon A1R (Nikon Instruments Inc.) with Z-series capability. Scaffolds were observed under a confocal microscope with a filter set for DAPI (Ex/Em: 350/470 nm), Texas Red (Ex/Em: 540/605 nm) and GFP/FITC (Ex/Em: 488/514 nm). 3D rendering images and confocal 3D maximum projection images were assembled with Leica confocal software (ver 2.61, Leica), NIS-Elements AR software package (ver 4.20.01, Nikon) and ImageJ.

Quantitative RT-PCR.

Intestinal smooth muscle constructs were homogenized. Total RNA was isolated using the Qiagen Mini mRNA Extraction kit. RNA was reverse-transcribed using High-Capacity cDNA Reverse Transcription Kit (Invitrogen) following the manufacturer’s instructions. Six nanograms of cDNA were used for real-time PCR amplification for each well, using following primer sequences. α-SMA: 5’-CCGACCGAATGCAGAAGGA-3’ (forward); 5’- ACAGAGTATTTGCGCTCCGAA-3’ (reverse): , VMT: 5’-TGTCCAAATCGATGTGGATGTTTC-3’ (forward); TTGTACCATTCTTCTGCCTCCTG-3’ (reverse), MYH11: 5’-AGGCGAACCTAGACAAGAATAAG-3’ (forward); 5’-CTGGATGTTGAGAGTGGAGATG-3’ (reverse), CNN1: 5’-ATGTCCTCTGCTCACTTCAAC-3’ (forward); 5’-CACGTTCACCTTGTTTCCTTTC- 3’ (reverse), COL1A1: 5’-ATCAGCCCAAACCCCAAGGAGA-3’ (forward); 5’-CGCAGGAAGGTCAGCTGGATAG-3’ (reverse), and GAPDH: 5’-AGCCACATCGCTCAGACAC-3’ (forward); 5’-GCCCAATACGACCAAATCC-3’ (reverse). For each gene tested we performed three experimental replicates and four biological replicates. Gene expression levels were normalized to the GAPDH mRNA level.

Statistical Analysis.

Data are presented as mean ± SEM (n = 5). A two tailed t-test was performed to compare means between two groups and Analysis of Variance (ANOVA) was performed to compare means of multiple groups. P-values ≤ 0.05 were considered significant.

Results and Discussion

Fabrication of silk microfiber scaffolds

Scaffolds represent critical components in tissue engineering as they provide a temporary structure and 3D environment to support and guide cell and tissue functions^[26] . The human intestine contains two continuous layers of uniformly distributed smooth muscle cells that are arranged in distinct orientations – inner circular layer is orthogonally oriented to the outer longitudinal layer. Therefore, designing a 3D scaffold that can control cellular orientation and organization in the different cell layers should be beneficial to generating functional intestinal tissues in vitro ^[5]. To address this challenge, we developed 3D silk scaffolds by rolling two silk microfiber sheets to form bi-layered cylindrical scaffolds for the engineering the intestinal smooth muscle. These silk microfiber scaffolds were prepared by directional freeze-casting and lyophilization (Figure 1A). Directional freeze-casting is a commonly used technique to fabricate aligned biomaterial structures, where the behavior of the ice-front propagation is controlled directionally ^[25]. Microstructural features obtained during freeze-drying have been controlled by exploiting the physics of ice formation, thus increasing the concentration of silk solution achieved scaffolds with aligned sheets ^[27]. Therefore, in this study, by lowering the concentration of silk solution (~2%), we were able to modulate the morphology of the ice crystals to generate anisotropic porous scaffolds with aligned microfibers. Briefly, the purified silk fibroin solution (referred to silk solution in this work) was obtained after degumming, dissolution, and purification. Then the concentration of silk solution was adjusted to 2 wt% for subsequent use in the freeze-casting to fabricate the silk microfiber scaffolds. To prepare the aligned silk microfiber scaffolds, a circular PTFE mold with a PTFE rod at the center was placed on an aluminum substrate (Figure 1A). Due to the low thermal conductivity of PTFE (0.25 W·m-1·K-1) and high thermal conductivity of aluminum (237 W·m-1·K-1), the directional ice crystal formation rate was rapid and formed from the aluminum substrate surface rather than the Teflon wall. After this directional freeze-casting, the lyophilized silk scaffold was treated with methanol overnight to introduce β-sheet structures, which results in more stable and stiffer scaffolds in aqueous solution. Finally, to recapitulate the ordered arrangement of fibrils of native intestinal smooth muscle layers, a pre-aligned outer microfiber sheet with a width of 1 cm in fiber direction and a length of 2.5 cm in the direction perpendicular to the fibers (Figure 1B) and a pre-aligned inner microfiber sheet with a width of 1 cm in a direction perpendicular to the fibers and a length of 2 cm in the fiber direction (Figure 1C) were separately trimmed. A bi-layered silk microfiber scaffold (Figure 1D) was prepared by wrapping the two trimmed silk microfiber sheets concentrically around a rod of 2 mm in diameter sequentially, where the inner layer was wrapped along the parallel direction to the microfiber alignment while the outer layer was wrapped along the perpendicular direction to the microfiber alignment.

Figure 1.

A.Schematic illustration of preparation process of the aligned bi-layered tubular silk microfiber scaffolds. Briefly, the purified silk fibroin solution (2 wt%) was obtained after degumming, dissolution, and purification. Then directional freeze-casting was used to prepare the aligned silk microfiber scaffolds. After lyophilization, the scaffolds were treated with methanol overnight to introduce β-sheet structures. A bi-layered silk microfiber scaffold was prepared by wrapping the two trimmed aligned silk microfiber scaffolds concentrically around a rod of 2 mm in diameter sequentially, where the inner layer was wrapped along the parallel direction to the microfiber alignment while the outer layer was wrapped along the perpendicular direction to the microfiber alignment. B-D. Photographs of aligned silk microfiber scaffolds and a bi-layer silk microfiber scaffold. Scale bar of D is 2 mm.

Microstructure and mechanical properties of silk microfiber scaffolds

A functional scaffold for engineered tissue should possess appropriate structural and mechanical characteristics including morphology, porosity, stability, stiffness, and mechanical strength, to appropriately influence cell-matrix and cell-cell interactions ^{[12, 26]}.

SEM was employed to characterize the topography of the silk microfiber scaffolds. The images showed the morphology of directional fibers in the fiber sheets (Figure 2A) and confirmed the bilayer structure of the scaffolds, where the alignment of microfibers in the inner layer was perpendicular to the alignment in the outer layer (Figure 2C). In each layer, the silk microfibers showed good alignment with a narrow distribution of the orientation angles (Figure 2E). In addition, the average diameter of microfibers was 2.0 ± 0.3 μm (Figure 2D) and the average gap distance between adjacent silk microfibers was 5.9 ± 2.9 μm (Figure 2F). Fibers with an average diameter larger than 1 μm exhibited favorable geometry for cell attachment and growth ^[28]. In our scaffolds, except for the gap distance between fibers due to the inhomogeneous packing of silk microfibers, larger pores/gaps were randomly present in the scaffolds (indicated in red circles in Figure 2B). These micro-structural features made the scaffolds good substrates for cell proliferation/infiltration and nutrient/gas transport.

Figure 2.

A-C. SEM images of the bi-layer silk microfiber scaffolds. Scale bars are 10 μm, 10 μm, and 100 μm in A, B, and C, respectively. D-F, Histograms of silk microfiber diameters (D), silk microfiber orientations (E), and gap distances between adjacent silk microfibers (F). The microfiber diameters were estimated by measuring 100 individual silk microfibers and gap distance between adjacent microfibers was determined by measuring 200 gap distances. The alignment of microfibers was analyzed by measuring the orientation angles of 100 individual silk microfibers in a single image.

Biomaterial scaffolds for tissue culture generally do not dissolve during in vitro culture in aqueous media or prior to in vivo implantation. Silk protein is water-soluble ^[14]. Thus, to obtain a water stable structure the scaffolds were treated with methanol to induce β-sheet (crystalline) structure. FTIR spectroscopy was used to analyze the conformational states of the silk to confirm this structural transition. In deconvolution profiles of the Amide I bands in FTIR spectra, peaks at 1618 cm⁻¹ and 1698 cm⁻¹ were assigned to β-sheet structures, while peaks at 1646 cm⁻¹ and 1685 cm⁻¹ were assigned to random coil/helix structure and β-turn structure, respectively. The FTIR spectroscopy analysis showed that the silk presented predominantly random coil/helical structures (peaks at 1646 cm⁻¹ and 1685 cm⁻¹) in the microfibers prior to methanol treatment (Figure 3A–B). Upon methanol treatment, the frequency of β-sheet structure (peaks at 1618 cm⁻¹ and 1698 cm⁻¹) significantly increased (~33%, Figure 3B), supporting that the methanol treatment efficiently induced the formation of β-sheet structures in silk scaffolds as previously reported ^[29], leading to the stability of scaffolds in aqueous solution.

Figure 3.

A. FTIR spectra of silk microfiber scaffolds before and after methanol treatment. B. Analysis of secondary structures from the FTIR spectra. The secondary structural analysis was carried out by performing peak deconvolution over the amide I region (1600−1700 cm⁻¹) with the following parameters: 1618 cm⁻¹ (β-sheet), 1646 cm⁻¹ (random coil/helix), 1685 cm⁻¹ (β-turn), and 1698 cm⁻¹ (β-sheet). C. Tensile mechanical properties of the aligned silk microfiber scaffolds along two orthogonal directions. D. Comparison of tensile strength and tensile modulus of the aligned silk microfiber scaffolds along two orthogonal directions. At least n=3 samples were used to calculate average strengths and modulus with standard deviation.

Aside from 3D structural support, the biomaterials should also provide sufficient mechanical support for the cells to undergo spatial tissue organization and functions. Thus, mechanical tests were performed on the constructs along two different directions, parallel and perpendicular to the orientation of the microfibers in the plane of the scaffolds. The results showed that the tensile strength and modulus obtained from the parallel direction (strength: 779 ± 233 KPa, modulus: 2.05 ± 0.04 MPa) were higher than those obtained in the perpendicular direction (strength: 37± 2 KPa, modulus: 0.21 ± 0.01 MPa) (Figure 3C–D), indicating the anisotropic mechanical properties for the aligned silk microfiber scaffolds. The physical properties of the intestinal wall vary in the whole gastrointestinal (GI) tract, including different segments of the gut ^[30]. The microfiber scaffolds displayed a modulus in the range of the native GI tract ^[31] and the mechanical properties of the microfiber sheets can be tuned by altering layer thickness to meet the needs of engineering specific sites of the GI tract.

Overall, morphologically, the scaffolds were biomimetic, presenting two concentric layers with porous microfibers mimicking the configuration of the native intestinal muscle wall (Figure 1D and 2). Mechanically, the yield tensile strength and compressive modulus of the aligned silk microfiber composites complied with the biomechanical behavior of the native smooth muscle tissue (Figure 3). The use of 3D anisotropic scaffolds such as aligned porous scaffolds and micro/nano fibrous scaffolds can support myogenic differentiation and myotube formation to regenerate functional muscle fibers ^[32].

Smooth muscle cell alignment, proliferation and viability on individual silk scaffold layers

To assess the support for the smooth muscle cells, primary human intestinal smooth muscle cells (hiSMCs) were utilized to assess the cytocompatibility, cellular alignment and viability in these silk fiber scaffolds. Cells were seeded separately in the inner and outer microfiber sheets of the scaffolds and then the seeded microfiber sheets were assembled into the bi-layered tubular scaffolds as described in the methods and Figure 4A. After cell seeding, the tissue constructs were maintained in culture medium for up to 12 weeks, resulting in intrinsic cohesion between layers after about one week of culture. This feature was likely due to the deposition of ECM proteins like type I collagen (Figure 6E), and thus meant that adhesives or other exogenous treatments were not required to maintain the integrity of the bilayer scaffolds. Calcein-AM staining (live cell staining) on the scaffolds was used to visualize cell spatial distribution and survival in the engineered intestinal smooth muscle tissues at week 2 post cell seeding. The florescence images indicated that the design of the scaffolds (bi-layered, tubular, microfiber) and cell seeding strategy (layer by layer seeding and roll up) successfully localized hiSMCs within the scaffold compartments (inner circular and outer longitudinal layers) as designed (Figure 4B–D). The majority of cells (~90%, Figure 4E–G) were elongated and circumferentially orientated in the inner layers of the constructs, while they were longitudinally orientated in the outer layers. These results suggested that: (1) the bi-layered structure enabled the separate control of cell orientation in each layer, and (2) the microfibers in each layer provided cells with the positive effects of contact guidance, confirming that aligned topography guides cell alignment and organization along the arrangement direction of fibers ^[33]. It should be noted that alignment of smooth muscle cells contributes to the proper muscle peristaltic movement and therefore coordination of dynamic gut motility ^[30].

Figure 4.

A. Schematic representation of the general cell seeding strategy used for hiSMC-derived 3D intestinal smooth muscle constructs. B-D. Live staining of the tissue constructs showed the cell spatial distribution and survival in the scaffolds at week 2 post cell seeding. Scale bars are 2 mm, 500 μm and 50 μm for B, C and D, respectively. E-F. DAPI staining of the two muscle layers, circular (Cir, E) and longitudinal (Lon, F) layers at week 2 post seeding demonstrated the cell elongation and alignment in the microfibers. G. Quantification of nuclei alignment in the tissue constructs. Scale bars are 25 μm for E and F.

Figure 6.

A-E. qRT-PCR analyses showed the gene expression levels of different smooth muscle differentiation markers, α-SMA (A), VMT (B), MYH11 (C), CNN1 (D) and COL1A1 (E) over time (up to 12 weeks) in cultures, p<0.001. Both layers of the smooth muscle constructs showed significant upregulation (~5–75 fold) of all marker transcripts after 3 weeks of cultivation, with stable expression levels until week 12. No significant difference in marker gene expression was observed between the circular and longitudinal layers. F-R. Confocal images of the tissue constructs immunostained with a-SMA (red), CNN1 (green) and DAPI (blue). Scale bars are 50 μm for F and 25 μm for G-R.

For cell proliferation, the expression of Ki-67, a cell proliferation marker ^[34], was investigated in the inner circular and outer longitudinal layers separately. The percentage of Ki-67 positive cells (the Ki-67 labeling index) was quantified to evaluate cell growth up to 12 weeks. A high level of Ki-67 positive cells (~40–70%) was detected in both layers of the scaffolds harvested from day 3 to week 3, while the cell proliferative rate gradually declined after 3 weeks post cell seeding (~10–20%) (Figure 5A–K). This result may indicate that the hiSMCs became confluent in the scaffolds at about week 3, suppressing further cell proliferation and entering a more differentiated state; explaining why no significant changes in cell viability (AlamarBlue® assay) were observed after week 3 (Figure 5L).

Figure 5.

A-E. Confocal images of Ki-67 staining (Ki-67: green, DAPI: blue) of the circular layer of the tissue constructs. F-J. Confocal images of Ki-67 staining (Ki-67: green, DAPI: blue) of the longitudinal layer of the tissue constructs. K. Quantification of Ki-67 positive cells (the Ki-67 labeling index) in each layer of the tissues. The percentage of Ki-67 positive cells was quantified to evaluate cell growth up to 12 weeks. L. Quantification of cell viability over time using AlamarBlue™ assay revealed that the tissues survived at least up to 12 weeks. Data presented as mean±SEM, p<0.001.

The above findings indicated that the silk was cytocompatible to hiSMCs and the design of the scaffolds successfully guided cell adhesion, orientation, and spindle-like morphology in the inner circular layer and outer longitudinal layer. In previous studies, aligned silk-based lamellar scaffolds were prepared using directional freezing and were also used in adipogenic and chondrogenic differentiation of human mesenchymal stem cells as well as to recapitulate form and function of the intervertebral disc ^[35].

Intestine smooth muscle cell differentiation

Functional biomaterial scaffolds should support cell growth as well as support lineage-specific differentiation for the restoration of tissue function. Therefore, we explored the smooth muscle phenotypes of the hiSMCs in the scaffolds at the transcriptional and translational levels by examining smooth muscle differentiation markers ^[36]. mRNA expression analysis was adopted to quantify gene expression levels of known intestinal smooth muscle markers over time (up to at least 12 weeks). The markers included α-smooth muscle actin (a-SMA), Vimentin (VMT), Calponin 1 (CNN1), and Myosin heavy chain 11 (MYH-11). Both layers of the smooth muscle constructs showed significant upregulation (~5–75 fold) of all marker transcripts after 3 weeks of cultivation, with stable expression levels until week 12 (Figure 6A–E). No significant difference in marker gene expression was observed between the circular and longitudinal layers. αSMA and VMT are known as common differentiation markers of smooth muscle cells, while MYH11 and CNN1 are late differentiation markers and more specific to contractile smooth muscle cells ^[37]. Expression of MYH11 and CNN1 in the tissues over time (Figure 6C, D) revealed that hiSMCs were differentiated into contractile phenotype in the aligned scaffolds and cytologically capable of performing potential contraction/relaxation. We also detected the expression of Type I collagen (COL1A1), as the differentiation of SMCs induces the deposition of ECM proteins including collagen ^[38]. The mRNA level of COL1A1 was significantly upregulated in both inner and outer layers, which suggested an increasing collagen content. We then performed confocal immunofluorescence on the scaffolds to confirm protein expression of selective marker genes, a-SMA and CNN1. Immunostaining of scaffolds showed significant positive staining of α-SMA and VMT in both microfiber layers from week 1 to week 12 of culture (Figure 6 F–R). The elongated cell/nuclei morphology along the fiber orientation were also further confirmed in the microfiber scaffolds (Figure 6 F–R).

One of the main functions of the smooth muscle is contraction. In this study we recorded calcium signals in the hiSMCs seeded on the scaffolds at week 2 post-seeding as a direct indicator of the contractile phenotype of the smooth muscles ^[39]. Calcium staining displayed significant changes in fluorescence intensity of Ca²⁺ over time in the aligned smooth muscle cells (Sup Video 1) implicating contractile properties found in smooth muscles. We were unable to detect significant force (tension) generated by the engineered tissues when using isometric transducers (data not shown). As aligned fibers have significantly higher modulus compared to nonaligned fibers ^[40], it is possible that the modulus of the aligned microfiber scaffolds was not comparable with the modulus of hiSMC-produced ECM, which restricted the extent of the contraction and relaxation of the tissues. Balancing the stiffness and modulus of the scaffolds for mimicking 3D tissue biostructures and concurrently replicating global tissue contraction has been a recurrent challenge while studying smooth muscle engineering ^[41]. In our system, this challenge could be potentially overcome by optimizing the gap distance between the aligned microfibers, introducing silk nanofibers into the microfiber scaffolds, or combining other biomaterials with better elastic modulus, such as collagen and gelatin, into the scaffolds.

In the study, the scaffolds maintained cell viability and phenotype of hiSMCs for at least 12 weeks. This result is in contrast to the differentiation of hiSMCs on conventional 2D tissue culture plates or other 3D substrates, where the cells in these systems usually did not survive over 4 weeks ^{[9, 42]}. Therefore, the tissue constructs described in the present study may be more appropriate for long-term or chronic studies with intestinal models.

Neurite extension within the intestinal smooth muscle tissue constructs

Intestinal smooth muscle layers are innervated by different populations of motor neurons to affect gut mobility. Within the gut wall, autonomic nerves are found in two locations: the myenteric plexus and the submucosal plexus. The myenteric plexus is located between the longitudinal and circular muscle layers. During development, enteric neural crest-derived cells migrate radially from the muscularis externa (which encompasses longitudinal muscle, myenteric plexus, and circular muscle) to populate the submucosal plexus before extending radial and longitudinal projections that connect the plexuses and form functional circuits with the surrounding tissue ^[43]. Neurite outgrowth represents a key step of neuronal migration and differentiation ^[44]. Therefore, we tested if neurons would differentiate and grow extensions into the engineered bi-layered muscle constructs in culture. The muscle constructs were prepared as described above and cultured for 3 days. To mimic enteric nerve development in gut, human induced neural stem cells (hiNSCs) ^[17] were evenly delivered in between the inner and outer layers of the constructs pre-seeded with hiSMCs as summarized in Figure 7A. The tissue constructs were then cultured for 2 weeks and immunostained for a-SMA and a neuronal marker TUJ1 to observe the spatial pattern of hiNSCs in the bi-layered muscle constructs. The hiNSCs proliferated, differentiated and extended neurites across the muscle constructs, including into the inner circular (Figure 7B, B’) and outer longitudinal (Figure 7C, C’) layers of the tissue constructs. In addition, the neurites preferentially extended along the direction of the filaments of smooth muscle cells in both layers, which morphologically matched the appearance of normal neurite outgrowth in muscles ^[45].These data suggest that during co-culture, topographical cues presented by the microfibers supported neurite extension via the pre-seeded smooth muscle cells.

Figure 7.

A. Schematic illustration of the cell seeding of hiNSCs in the engineered intestinal smooth muscle tissues. B. Confocal images of circular layers of the constructs immunostained with α-SMA and TUJ1. B’ is the enlarged area in B. Scale bars are 50 μm and 20 μm. C. Confocal images of longitudinal layers of the constructs immunostained with α-SMA and TUJ1. C’ is the enlarged area in C. Scale bars are 50μm for B and C and 15μm for B’ and C’. The hiNSCs proliferated and extended neurites across the muscle constructs, including into the inner circular (B, B’) and outer longitudinal (C, C’) layers of the tissue constructs.

Innervation is a key to successful tissue engineering and regenerative medicine with an end goal of the restoration of tissue function ^[44]. Different biomaterials have been investigated for their in vitro ability to support neuronal outgrowth and differentiation ^[46]. However, only a few studies have investigated innervation of tissue engineered muscle tissues and their functional integration in vitro ^[47]. While the present study showed evidence that neurons can migrate and grow neurite extensions in our tissue engineered smooth muscle constructs, it is not clear whether the submucous plexus would form and whether the nervous tissues were functional, a process that requires long time frames to achieve functional innervation depending on cell sources and in vitro co-culture conditions ^[48]. In this regard, although the current study did not further evaluate function, it is plausible that the many active neurite extensions observed in tissue constructs could eventually regenerate functional connections. Future studies are needed to understand the mechanisms underlying the process of innervation in the constructs, such as how the aligned scaffolds provide constructive signaling cues to direct the formation of the submucous plexus resulting in neuromuscular regeneration. By understanding the mechanisms involved in the promotion of innervation and controlling the spatio-temporal patterns of innervation, it may be possible to ultimately support the remodeling of functional innervated intestinal smooth muscle tissues.

Conclusions

The complete reproduction of the natural structure of human intestinal smooth muscle is a challenging step in functional intestine engineering approaches. To address this need, biomaterial scaffolds were designed in a tubular bilayer system to mimic the native architecture and to engineer smooth muscle tissue formation, as well as innervation. The results allow us to move a step closer to mimicking a fully functional human intestinal tissue in vitro. In human body, many smooth muscle-containing hollow organs and tubes have multiple smooth muscle layers ^[49], such as the urinary bladder, gall bladder, uterus, bile ducts, ureter, and arteries. Thus, this type of composite scaffold could also be useful as candidates for engineering numerous other types of smooth muscles.