Incorporation of resident macrophages in engineered tissues: Multiple cell type response to microenvironment controlled macrophage‐laden gelatine hydrogels

The success of tissue engineering strategy is strongly related to the inflammatory response, mainly through the activity of macrophages that are key cells in initial immune response to implants. For engineered tissues, the presence of resident macrophages can be beneficial for maintenance of homeostasis and healing. Thus, incorporation of macrophages in engineered tissues can facilitate the integration upon implantation. In this study, an in‐vitro model of interaction was developed between encapsulated naive monocytes, macrophages induced with M1/M2 stimulation and incoming cells for immune assisted tissue engineering applications. To mimic the wound healing cascade, naive THP‐1 monocytes, endothelial cells and fibroblasts were seeded on the gels as incoming cells. The interaction was first monitored in the absence of the gels. To mimic resident macrophages, THP‐1 cells were encapsulated in the presence or absence of IL‐4 to control their phenotype and then these hydrogels were seeded with incoming cells. Without encapsulation, activated macrophages induce apoptosis in endothelial cells. Once encapsulated no adverse effects were seen. Macrophage‐laden hydrogels attracted more endothelial cells and fibroblasts compared to monocytes‐laden hydrogels. The induction (M2 stimulation) of encapsulated macrophages did not change the overall number of attracted cells; but significantly affected their morphology. M1 stimulation by a defined media resulted in more secretion of both pro‐ and anti‐inflammatory cytokines compared to M2 stimulation. It was demonstrated that there is a distinct effect of encapsulated macrophages on the behaviour of the incoming cells; this effect can be harnessed to establish a microenvironment more prone to regeneration upon implantation.


| INTRODUCTION
Once a material is implanted into the body, adverse immune reactions can be triggered. Cells receive a diverse range of signals from their surrounding microenvironment and adjust their responses accordingly (Kzhyshkowska et al., 2015). The immune system recognizes the material as foreign, initiating a macrophage-mediated acute inflammatory phase (Sridharan, Cameron, Kelly, Kearney, & O'Brien, 2015).
Inflammatory cells initially are recruited to the site of inflammation due to the endothelial response to injury. Monocytes respond to monocyte chemoattractant protein-1 (MCP-1), which is secreted by cytokine-stimulated endothelial and smooth muscle cells in the intima.
In response to cytokines secreted by endothelial cells, fibroblasts and other cells, the incoming monocytes at the site of injury differentiates into macrophages (Takahara, Kashiwagi, Maegawa, & Shigeta, 1996).
This chain of events makes macrophages an important factor in the acceptance or rejection of implanted biomaterials.
Another important aspect of the fate of engineered tissues is their integration with the host vasculature; as angiogenesis is involved in wound healing by promoting the outgrowth of new blood vessels from pre-existing vasculature (Fritsche et al., 2009). Inflammation also plays a major role in physiological angiogenesis. Inflammatory cells such as macrophages secrete a multitude of inflammatory mediators which may influence angiogenesis (Coussens & Werb, 2002). Thus, it is important to establish a strong control over the cross-talk between endothelial cells and macrophages in the vicinity of engineered tissues.
The success of the tissue engineering strategy is strongly related to the inflammatory response, mainly through the activity of macrophages. They can be polarized either into M1 or M2 phenotype, depending on the activation signals (Kzhyshkowska et al., 2015). M1 macrophages are considered to be proinflammatory, while M2 macrophages promote tissue regeneration. However, the exploitation of these cells to control engineered tissue integration has not been widely studied (Vrana, 2016). Resident macrophages are cells observed in most of tissues with a heterogenous phenotype (Davies et al., 2013).
They are involved in wound healing and resolution of inflammation; thus, encapsulation of monocytes or predifferentiated macrophages in hydrogels can mimic this function in tissue engineering applications.
Some examples of resident macrophages are microglia, Kupffer cells, alveolar macrophages and osteoclasts (van de Laar et al., 2016).
One of the main means for macrophages to control the microenvironment around the implanted biomaterials is the cytokine production, M1 macrophages produce larger amounts of the proinflammatory cytokines: tumour necrosis factor-α (TNF-α), interleukin (IL)-12 and IL-1β. TNF-α is a master proinflammatory cytokine involved in chronic inflammation (Munker, Gasson, Ogawa, & Koeffler, 1986). IL-12 promotes T H 1 cell-mediated responses, which have been implicated in the pathogenesis of a number of inflammatory and autoimmune diseases (Teng et al., 2015). IL-1β is an important mediator of the inflammatory response (Dinarello, 2009). M2 macrophages produce anti-inflammatory cytokines such as IL-1RA, CCL18. IL-1RA is a natural inhibitor of the proinflammatory effect of IL-1β. IL-1RA is used to treat autoinflammatory diseases such as rheumatoid arthritis and juvenile idiopathic arthritis (Dinarello, 2009). CCL18 is a chemokinea cytokine with chemotactic activity (Struyf et al., 2003). CCL18 is produced by myeloid cells and induces chemotaxis of lymphocytes and immature DCs, as well as collagen deposition by fibroblasts (Schutyser, Richmond, & Van Damme, 2005). Although it is not one of the main components of macrophage secretome, IL-4 is one of the main factors in macrophage polarization. It drives the differentiation of naive T

helper (T H ) cells to T H 2 effector cells and monocytes and macrophages
to an M2 (or alternatively) activated phenotype (Martinez & Gordon, 2014). As can be seen from the aforementioned activities of these cytokines, the cytokine microenvironment around an engineered tissue would significantly affect its remodelling and integration and needs to be controlled (Knopf-Marques et al., 2016).
The wound healing cascade can be divided into three steps. The initial inflammatory step is characterized first by the formation of blood clot and the mobilization of neutrophils and then macrophages to the site of injury (Sinno & Prakash, 2013). Then there is a proliferative phase with the formation of a granulation tissue. At this stage, endothelial cells are crucial because neovascularization will enable the procurement of the nutrients to the wound and this will lead to the formation of a new stroma. The last stage is the maturation phase, which is characterized by the transition from granulation tissue to scar formation (Jun, Kim, & Lau, 2015;Levenson et al., 1965). In this step, fibroblasts are of interest because they will secrete collagen and they will be the key players to remodel extracellular matrix. In all these three stages and especially in the last two, macrophages will continuously secrete growth factors that will stimulate angiogenesis and collagen deposition and that is why these three cell types are the main actors for tissue remodelling after injury.
One way to understand and control these complex interactions is to develop simplified in vitro models (Mahler, Esch, Glahn, & Shuler, 2009). However, currently there are no models that directly deal with the interaction of immune cells with connective tissue cells in tissue engineering context. A recent example of monitoring of cell-cell interactions is an intestine organ-on-chip study by Kim, Li, Collins, and Ingber (2016). The cells are exposed to physiological peristalsis-like motions and fluid flow (Kim, Huh, Hamilton, & Ingber, 2012). The system involves a coculture of human intestinal endothelial and immune cells, in the presence of bacteria to elucidate how they contribute to host tolerance of infection and disease inflammation (Kim et al., 2016). However, to assess regeneration in a multi-cellular environment a model should integrate extracellular matrix component.
One way to achieve this is to encapsulate cells in ECM such as hydrogels as the three-dimensional (3D) cell culture in vitro environment is a crucial key for acquiring phenotypes and for responding to stimuli analogous to in vivo biological systems (Szot, Buchanan, Freeman, & Rylander, 2011).
The main objective of this study is to develop an in vitro model of interaction between incoming cells and encapsulated macrophages as a model of inclusion of resident macrophages in engineered tissues.
Fibroblasts and endothelial cells have been selected in coculture condition with macrophages due to their key role in wound healing process.
Thus, in a tissue engineering setting it is important to control the incoming cells to have a precise control over the remodelling of the implanted engineered tissue. To this end, the attachment, morphology and cytokine profile were observed in different settings pertaining to wound healing cascade in the presence of macrophages in M1/M2 phenotype inducing microenvironments. Through this model, the aim was to define the optimal inclusion conditions for macrophages in engineered tissues. Bacterial transglutaminase was kindly provided by Ajinomoto Inc (Tokyo, Japan).
HUVECs were used at passages between 4 and 8. The culture media used were endothelial cell growth medium supplemented with Supplement Mix C-39215 (mainly composed of heparin, hydrocortisone, fetal calf serum, basic fibroblast growth factor, epithelial growth factor and endothelial cell growth supplement) and 1% penicillin/ streptomycin. Prior to seeding, cells were harvested using 0.05% trypsin/0.02% EDTA, centrifuged and resuspended in media.
M1 and M2 phenotypes are generally characterized by specific cytokine inductions of a variety of phenotype marker expression, such as cytokines, surface markers, bioactive molecules such as ROS, NO and chemokines. The commercially available M1 and M2 generation media from PromoCell (Heidelberg, Germany) contains the necessary cytokines to induce M1 or M2 differentiation. In this configuration, they were used to induce a specific stimulation microenvironment.
The cells were pretreated with the differentiation medium prior to the seeding of incoming cells (Gordon, 2003).

B. Contact cell coculture
For all cell experiments, the same protocol of seeding was used.
Gelatine type A solution was prepared in cell culture medium (RPMI 1640 GlutaMAX for THP-1) under sterile condition. All the

| Metabolic activity
To assess metabolic activity, samples were incubated with 10% v/v

| Scanning electron microscopy
The samples were fixed with 4% glutaraldehyde. The specimen were washed with Dulbecco's PBS prior to following a dehydration protocol using an alcohol series of increasing concentration (70%, 95% and 2 × 100%), with incubation periods of 5 min for each. Subsequently, samples were incubated in 100% ethanol/hexamethyldisilazane (1:1) for 5 min, then only in hexamethyldisilazane for 2 × 5 min and dried overnight. Samples were adhered onto titanium discs using a carbon tape and coated with gold/palladium in a sputter coater. The samples were sputtered at 7.5 mA for 3 min under argon atmosphere and images were acquired using a scanning electron microscope (Hitachi TM1000).

| Statistical analysis
The statistical significance of the obtained data was assessed using the t test or Mann-Whitney test (n ≥ 3). The error bars are representative of standard deviation. Differences at p ≥ 0.05 were considered statistically not significant.

| RESULTS AND DISCUSSION
Any implanted cell-containing material in vivo will be in contact with several incoming cell types including monocytes/macrophages, endothelial cells and connective tissue cells such as fibroblasts. The working hypothesis was that addition of a macrophage component in such systems to mimic resident macrophages and their role in wound healing will help the regeneration process. As macrophage function is tightly linked to their physicochemical niche, encapsulation in hydrogels might provide the necessary signal for encapsulated macrophages to assume a more resident macrophage-like phenotype.
To assess the effects of the macrophage presence, a system has been developed where prelabelled cells of different types can be directly added onto macrophage-laden hydrogels to assess their interaction with the encapsulated macrophages. To determine the effect of encapsulation, the interaction was first monitored in the absence of the gels with a Transwell system (paracrine interactions).

A. Transwell system
To see the extent of the reaction of HUVECs to the presence of THP-1 cells without cell to cell contacts; naive and activated THP-1 cells were put into coculture conditions with HUVECs with the help of a Transwell inserts. In the absence of THP-1 cells, HUVECs formed a monolayer by 5 days with an increase in metabolic activity ( Figure 1b); in the presence of THP-1 cells, however, particularly with activated THP-1 cells, HUVECs had significantly lower metabolic activity (Figure 1b), with a substantial number of cells going through apoptosis as evidenced by positive Annexin-V staining (Figure 1c, d). To see whether this effect has any direct relation with the macrophage phenotype; the macrophages were put in contact with IL-4 (M2 inducer; Figure S1); aside from M2 inducing capacity of IL-4; it has also been shown that IL-4 dependent proliferation can be seen in resident macrophages (Davies et al., 2013). Addition of IL-4 did not have a significant effect on metabolic activity of HUVECs; however, the number of alive attached cells were significantly higher in the presence of IL-4 ( Figure S1B, C). The IL-4 has been shown to induce V-CAM expression by vascular endothelial cells (Iademarco, Barks, & Dean, 1995)  where Pettet, Byrne, McElwain, and Norbury (1996)  In regenerative medicine, the final aim is the complete integration of the implanted artificial tissue over time. This implies that the artificial tissue should integrate with the host circulatory system, nervous system and also the immune system. The immune system integration can be triggered by addition of macrophages in the artificial tissue formulation, in the manner described in this study, which can facilitate the establishment of a resident macrophage population. This can also facilitate the healing process and the resolution of inflammation induced by implantation.

| CONCLUSION
The design and maturation of an engineered tissue is important for its integration within the host. Here we demonstrated in an in vitro model that the encapsulated macrophages, mimicking resident macrophages, have an impact on the behaviour of incoming cells in a cell-type specific manner. The presence of pro-or anti-inflammatory cytokines can be impacted by the conditioning of the encapsulated macrophages.
In the future, the incorporation of phenotype controlled macrophages in engineered tissues has an important potential to control the initial immune reaction upon implantation. Controlled delivery of polarization inducing cytokines could achieve partially a similar effect; but incorporation of macrophages would provide an active source of secretion for a variety of cytokines and bioactive agents for a longer term. Such a model can be used to better understand possible immunoengineering strategies using macrophages and also studies of immune response to xenogenic and allogenic cells in a controlled manner. Future studies will focus on the elucidation of the encapsulation conditions on macrophage polarization for potential regenerative medicine applications. Vrana, N. E. (2016). Immunomodulatory biomaterials and regenerative immunology. Future Science OA, 2(4). FSO146

SUPPORTING INFORMATION
Additional Supporting Information may be found online in the supporting information tab for this article.