In Vivo Delivery of M0, M1, and M2 Macrophage Subtypes via Genipin-Cross-Linked Collagen Biotextile

Abstract

Developing strategies to regulate the immune response poses significant challenges with respect to the clinical translation of tissue-engineered scaffolds. Prominent advancements have been made relating to macrophage-based therapies and biomaterials. Macrophages exhibit the potential to influence healing trajectory, and predominance of particular subtypes during early onset of healing influences repair outcomes. This study evaluated short- and long-term healing response and postoperative mechanical properties of genipin-cross-linked, electrochemically aligned collagen biotextiles with comparative administration of M0, M1, and M2 subtypes. Irrespective of macrophage subtype seeded, all the groups demonstrated existence of M2 macrophages at both time points as typified by arginase and Ym-1 expressions, and distinct absence of M1 macrophages, as indicated by lack of inducible nitric oxide synthase (iNOS) and interleukin-1β expression in all the groups for both time points. M2 macrophage-seeded collagen biotextiles revealed promising host tissue responses, such as reduced fibrous capsule thickness and minimal granulation tissue formation. Furthermore, the M2-seeded group displayed more abundant interstitial collagen deposition following degradation of the collagen threads. M2 macrophage supplementation improved structural and mechanical properties at the tissue and cellular level as indicated by increased modulus and stiffness. This study demonstrates improved biomechanical and histological outcomes following incorporation of M2 macrophages into genipin-cross-linked collagen biotextiles for tissue repair and offers future strategies focused on connective tissue regeneration.

Impact statement

Macrophages exhibit significant plasticity with complex phenotypes ranging from proinflammatory (M1) to proregenerative (M2). They release cytokines and chemokines governing immunological stability, inflammation resolution, and tissue healing and regeneration. However, utilization of macrophages as therapeutic tools for tissue engineering remains limited. In this study, genipin-cross-linked collagen biotextiles were employed to deliver M0, M1, and M2 macrophages and evaluate tissue responses and postsurgical mechanical properties in vivo. M2-seeded collagen biotextiles showed reduced fibrous capsule and favorable healing response. These outcomes shed new light on designing tissue-engineered constructs that offer a novel cell-based therapeutic approach for applications requiring structural augmentation.

Introduction

Controlling host immune responses remains a challenge regarding the clinical translation of tissue-engineered scaffolds.¹ Host immune response to any implanted material is multifaceted and relies on numerous factors and state of the host tissue.² Biomaterials designed to rebuild injured tissues are of critical importance in regenerative medicine. A better understanding of the complex interplay between innate immune response and biomaterials is an emerging core factor in engineering scaffolds that integrate with host tissues and support regeneration and recovery of function.^3,4 Therefore, managing initial inflammatory responses can be a useful tool for developing next-generation biomaterials that exert enhanced positive actions on implant integration and healing processes.⁴

Monocytes are precursor cells that migrate from bone marrow into peripheral tissues following injury.^5,6 They differentiate into macrophages and determine the host response trajectory and affect long-term success of biomaterial integration.⁷ Macrophages can be broadly divided into three separate categories: nonactivated (M0), proinflammatory (M1), and anti-inflammatory (M2) macrophages.⁸ M1 macrophages release potent proinflammatory cytokines (interleukin [IL]-12, IL-23, etc.) that contribute to chronic inflammation and formation of fibrous capsule. M2 macrophages are involved in orchestrating wound healing and resolution through release of anti-inflammatory cytokines (IL-13, IL-4, IL-10, etc.), for which controlled macrophage activation is essential.^9,10

Over the past decade, several innovations have arisen with respect to macrophage-based therapies and biomaterials. Activation of macrophages ex vivo, before injection or implantation, proved beneficial for skeletal muscle regeneration.^11,12 and modulation of local inflammatory responses in kidney disease.^13,14 However, utilization of macrophages as therapeutic tools for tissue engineering remains limited. Recent articles propose the use of macrophages for regenerative purposes either by way of modifying biomaterials to attract host macrophages or by direct delivery of macrophages.^15,16 Direct delivery ensures desired cell types are provided in sufficient quantity to the influence host response but attempts to realize macrophage-driven immunoregeneration in combination with biomaterials are sparse in the literature.

To the best of our knowledge, only Spiller et al. have investigated modified biomaterials with immunomodulatory factors aimed at harnessing host macrophages.¹⁷ They delivered IL-4 and interferon (IFN)-γ sequentially from demineralized trabecular bone with the intent of attracting host M1 and M2 macrophages, respectively. Again, to the best of our knowledge, work by Rybalko et al. represents the only cell delivery therapy wherein M1 macrophages were administered directly to ischemic muscles, which resulted in improved muscle function.¹² As macrophages have the potential to influence trajectory of healing, further studies are essential to understand how dominance of a specific macrophage subtype at the early onset of healing affects overall repair outcome.

Genipin-cross-linked collagen scaffolds were recently reported by our group to stimulate the proregenerative M2 macrophage subtype in vitro.¹⁸ We evaluated the impact of macrophage subtype–collagen scaffold interaction with respect to cell differentiation/elongation and reported polarization of M0 to the proregenerative M2 subtype following seeding on genipin-cross-linked collagen biotextiles. Establishing advanced biomaterials to induce early tissue remodeling and integration while preventing fibrous capsule formation may prove beneficial for long-term tissue regeneration strategies. Controlling or limiting the initial inflammatory response phase can provide a broad range of advancements in implanted biomaterials or tissue-engineered scaffolds for connective tissue engineering.

In this context, healing responses to delivery of M0, M1, and M2 subtypes have not been comparatively studied in vivo. Accordingly, the aims of this study were to evaluate short- and long-term tissue responses and postsurgical mechanical properties of M0-, M1-, and M2-seeded collagen scaffolds in a subcutaneous rat model. This comparative study may demonstrate whether there is a benefit in delivering macrophages in conjunction with a scaffold and reveal whether a particular subtype promotes tissue regeneration.

Materials and Methods

Scaffold fabrication

Electrochemical compaction was employed to convert collagen in solution (3 mg/mL, type I bovine collagen; Collagen Solutions, CA) to aligned threads (∼100 μm) of continuous lengths, as described previously.^18–20 Briefly, the collagen solution was dialyzed 24 h in deionized water at pH 7.0 at 4°C. A constant electrical voltage potential of 40 V was applied between the electrodes to form electrochemically aligned collagen threads, and 2-ply yarns were fabricated from individual threads.

A computer digital controlled machine (Sherline Inc., CA) was used to oversee the position controlled winding of 2-ply collagen yarns over a mandrel. Filament-wound yarns were then retrieved and compressed diametrically to obtain rectangular-shaped scaffolds (20 × 5 × 0.5 mm) wherein the yarns ran parallel to the longer axis of the rectangle (Supplementary Fig. S1). Collagen scaffolds were cross-linked with genipin (2% w/v in 90% ethanol, No. G4796; Sigma–Aldrich, Darmstadt, Germany) for 72 h at 37°C. Scaffolds were sterilized using 70% ethanol solution containing 0.01% peracetic acid for 4 h.

Allogeneic rat macrophage isolation and culture

Macrophages were isolated from bone marrow harvested from femurs and tibia of five 9-week-old female Sprague–Dawley (SD) rats (230–260 g), as reported previously.¹⁸ Adherent cells were cultured in standard monolayer conditions, provided fresh medium consisting of DMEM/F12 (No. 10565042; Thermo Fisher, Waltham, MA) supplemented with 20% macrophage colony-stimulating factor (No. 400-28; PeproTech), 10% fetal bovine serum (FBS), 1% penicillin/streptomycin, and l-glutamine 40 mM (No. 25030081; Thermo Fisher), yielding M0 macrophages.

For M1 polarization, M0 macrophages were incubated in 100 ng/mL lipopolysaccharide, Escherichia coli 0111: B4 (No. L2630; Sigma–Aldrich), and 25 ng/mL IFN-γ (No. 400-20; PeproTech). For M2 polarization, M0 macrophages were incubated in 10 ng/mL IL-4, IL-13, and TGF-β1 (PeproTech) for 24 h. Polarization efficiency of macrophages was determined by immunocytochemistry, Western blotting, and flow cytometry before seeding the cells onto scaffolds.

Immunocytochemistry

Polarization states of macrophages were surveyed by immunofluorescent staining. Induced macrophage subtypes (M0, M1, and M2) were fixed using 4% paraformaldehyde for 20 min and incubated with blocking solution containing 2% bovine serum albumin (BSA) for 30 min at room temperature. Permeabilized (0.2% Triton X-100) and blocked cells were then incubated by CD68 (1:100, No. ab31630), iNOS (1:300, No. ab15323), and arginase (1:200, No. AF5868) antibodies overnight at 4°C (n = 3). Cell nuclei were also stained 4′,6-diamidino-2-phenylindole (DAPI) to identify macrophages under fluorescent microscopy (Olympus BX50) (Supplementary Fig. S2).

Western blotting

Cell lysis buffer (RIPA, No. 89900; Thermo Fisher) was utilized for protein extraction supplemented with phosphatase and protease inhibitors (100 μg/mL phenylmethylsulfonyl fluoride, No. 36978; Thermo Fisher). Protein concentration in the cell lysis solution was determined using BCA Protein Assay Kit (No. 23225; Thermo Fisher), and 20 μg protein was loaded onto mini-protean gels (No. NP0321BOX; Thermo Fisher) and transferred to a nitrocellulose membrane (No. 88018; Thermo Fisher) (n = 3).

Membrane was blocked using 5% BSA 1 h at room temperature before incubation with primary antibodies against arginase for M2, iNOS for M1, and GAPDH (No. MAB5718; R&D systems) as a loading control overnight at 4°C. The membrane was incubated with conjugated secondary antibodies for 2 h at room temperature and detected by West Pico Chemiluminescent Substrate (No. 34580; Thermo Fisher). Original X-ray films with ladders are located within the supplemental material (Supplementary Fig. S3).

Flow cytometry

Polarized macrophages were detached using TrypLE (No. 12605010; Thermo Fisher), suspended in ice-cold staining buffer containing 10% FBS, 5% BSA, and sodium azide (0.09%). Fluorescent dye-conjugated antibodies against M0 (CD68), M1(CD86), and M2 (CD206) were utilized to confirm the polarized macrophage subtype presence (n = 3). Uniform histograms were generated to view frequency distribution of flow data with multiple events on the y-axis, and a single parameter on the x-axis were generated using FCSalyzer software v0.9.12-alpha (Supplementary Fig. S4A).

Cell seeding of collagen scaffolds

Before cell seeding, peracetic acid-sterilized scaffolds were rinsed thoroughly with sterile phosphate-buffered saline multiple times. Flow analyzed (M0) and polarized macrophages (M1 and M2) were seeded at a density of 1 × 10⁵ cells on collagen scaffolds and then incubated for 25 min at 37°C to allow for attachment. Cell-seeded scaffolds in six-well plates were cultured for 3 days in corresponding macrophage medium (M0, M1, and M2) before being transferred to the operating room. The cytoskeletal organization of macrophage subtypes was determined via actin staining to confirm attachment and uniform cell coverage over scaffold threads (Supplementary Fig. S4B). Total cell lysates were collected from cell-seeded scaffolds to characterize in vitro cell status before implantation at 72 h.

Animal surgeries

All animal procedures were conducted in accordance with rules and protocols established and approved by the Institutional Animal Care and Use Committee (Protocol No. 2014-0105) at Case Western Reserve University. Rats were anesthetized with 4% isoflurane using a laboratory animal anesthesia machine. The scaffolds were placed subcutaneously in the dorsal area of the animals by making two 2 cm long midline incisions. Each animal was implanted with four scaffolds.

A total of 20 animals were implanted with scaffolds. Specimens were divided into one of the four groups: (1) genipin scaffold (GES), (2) GES+M0, (3) GES+M1, and (4) GES+M2. Eight animals (n = 8, 32 scaffolds) were harvested at 3 weeks following implantation to evaluate biocompatibility, and 12 animals were sacrificed at 3 months following implantation for biocompatibility (n = 4, 16 scaffolds) and mechanical (n = 8, 32 scaffolds) assessments. Skin and muscles surrounding the implanted scaffolds were excised at the time of harvest.

Biomechanical testing

Hydrated explants were subjected to biomechanical testing under uniaxial tension (Universal Testing Machine, Test Resources Inc., Minnesota) with a calibrated 50 lb load cell (Omega, Norwalk, CT). Samples were affixed to grips using cyanoacrylate on both sides of explants to prevent the sample from slipping during testing. Samples were monotonically loaded with a displacement rate of 10 mm/min, and load and displacement data were recorded. Sample width and thickness were measured by caliper under minimal pressure. Cross-sectional area and gage length were used to calculate stiffness, modulus, and stress values for each sample.

Histology and immunohistochemistry

Explanted scaffolds were fixed in 10% neutral buffer formalin for 7 days, dehydrated through a series of graded alcohols, embedded in paraffin, sectioned (5 μm thickness), and mounted onto microscopy slides. Slides were then stained with hematoxylin and eosin or Masson's trichrome (MT) to detect and interpret morphological characteristics and collagen deposition, respectively. Slides were appraised in a blinded manner by a senior pathologist (J.M.A.) to assess material biocompatibility in accordance with ISO 10993, Biological evaluation of medical devices—Part 6: Tests for local effects after implantation.²¹

A semi-quantitative scale rating from 0 to 3 for minimal to extensive was utilized for the host response and tissue integration parameters. Host reaction at implant surfaces was scored for foreign body reaction, fibrous encapsulation, and granulation tissue (defined as multinucleated giant cells, fibroblasts, and blood vessels). Tissue integration was scored for cellular infiltration (indicated by the density of monocytes and macrophages) and new collagen deposition.

MT-stained slides were evaluated with stereological point counting^22,23 to ascertain interstitial collagen and collagen scaffold thread content. A grid (25 × 25 μm) was overlaid onto images collected from slides, and the number of points containing interstitial collagen and collagen scaffold threads was tabulated. Thickness measurements of fibrous capsule that formed around the exterior of the scaffolds were also assessed using ImageJ (n = 8 slides were used for each group). For point counting, a total of 3 slides from the continuum of the scaffold regions were collected from each specimen giving a total of 12 sections scored for each group.

Immunohistochemistry (IHC) was performed for M1 macrophages (iNOS, catalog No. sc-7271; IL-1β, catalog No. ab9722), M2 macrophages (arginase 1, catalog No. sc-271430; Ym-1, catalog No. 60130), collagen type I (COL1A, catalog No. sc-59772), collagen type III (COL3A1, catalog No. NB600594), and alpha-smooth muscle actin (α-SMA, catalog No. A2547). Secondary antibodies (mouse anti-rabbit IgG HRP, catalog No. 5127S; mouse IgG kappa binding protein, catalog No. sc-516102) were employed with the unconjugated primary antibodies.

The number of positive cells (arginase/Ym-1-M2 markers, α-SMA-myofibroblast marker) within a high-power field from 5 regions within the continuum of the collagen scaffold was collected from 3 slides of 4 specimens from each group (GES, GES+M0, GES+M1, and GES+M2) for a total of 12 slides analyzed per group. IHC slides were assessed in a binary manner for staining of arginase/Ym-1, iNOS/IL-1β, collagen type I, collagen type III, and α-SMA (n = 12 slides were evaluated for each group).

Statistical analyses

Mechanical data and quantitative histological measurements were assessed for significant differences by using nonparametric Mann–Whitney U-tests with statistical significance defined as p < 0.05. All statistical analyses were performed using GraphPad Prism (version 5.01).

Results

Characterization of preimplant macrophage subtypes

Polarization efficiency of macrophages was confirmed by immunocytochemistry, western blotting, and flow cytometry. Cells cultured under M0, M1, and M2 conditions were positive for CD68 (green), iNOS (red), arginase 1 (red), respectively (nuclei stained blue, DAPI) (Supplementary Fig. S2). Western blotting confirmed protein expression of iNOS (M1) and Arg1 (M2) macrophage subtypes (Supplementary Fig. S3).

In addition to immunocytochemistry and western blotting, flow cytometry confirmed the total number of cells in a sample expressing markers from a gated population (Supplementary Fig. S4A). M0 and M2 macrophages exhibited elongation after being seeded on scaffolds, and M1 cells remained round. M0 and M2 macrophages showed expression of Arg1 (M2 marker); however, M1 macrophages exhibited expression of iNOS (M1 marker) (Supplementary Fig. S5).

Surgical postoperative and harvest

No perioperative or postoperative complications were noted. All rats survived to the end of the survival period, and all filament–wound collagen scaffold specimens were found to be intact at the time of recovery (Supplementary Fig. S6).

Biomechanical testing

No significant differences in max load and stress were observed between treatment groups (Fig. 1A and B). The GES+M2 group demonstrated a statistically significant increase in stiffness relative to GES (p = 0.0003) and GES+M0 groups (p = 0.0011) (Fig. 1C). However, no significant difference was found between GES+M2 and GES+M1 (p = 0.328). An increased modulus was found in GES+M2 compared with GES (p < 0.05) (Fig. 1D).

FIG. 1.

Graphs illustrating biomechanical properties for specimens harvested following 3 months of implantation in vivo. (A) Maximum load, (B) stress to failure, (C) stiffness, and (D) modulus for scaffolds at 3 months (n = 8). Significance was determined using the Mann–Whitney U tests (p < 0.05). Letters above groups in the plots indicate overall grouping in terms of significance (i.e., “a” were not significantly different from one another, “ab” were not significantly different from “a” or “b,” “a” were significantly different from “b”). Error bars indicate standard deviation in each graph. GES, genipin scaffold; GES+M0, M0 cells seeded on genipin scaffold; GES+M1, M1 cells seeded on genipin scaffold; GES+M2, M2 cells seeded on genipin scaffold.

Scoring for host response and tissue integration

Histological scoring highlighted minimal granulation tissue, fibrous encapsulation, and foreign body reaction at 3 months for the GES+M2 group, whereas an increase in the same parameters was noted for the M1+GES group at 3 months (Table 1). Cellular infiltration and collagen deposition scored moderate to extensive for all groups at 3 months (Fig. 2 and Table 2).

Table 1.

Host Response Evaluation

	Granulation tissue		Fibrous encapsulation		Foreign body reaction
	3 Weeks	3 Months	3 Weeks	3 Months	3 Weeks	3 Months
GES	2 ± 0	2 ± 0	1 ± 0	2 ± 0	2 ± 0	2 ± 0
GES+M0	2 ± 0	2 ± 0	1 ± 0	2 ± 0	2 ± 0	2 ± 0
GES+M1	2 ± 0	3 ± 0	2 ± 0	3 ± 0	3 ± 0	3 ± 0
GES+M2	2 ± 0	1.5 ± 0.57	2 ± 0	0.25 ± 0.5	2 ± 0	1 ± 0

Scoring for host response parameters reported as the mean ± SD of scores at 3 weeks and 3 months. Scores were given on a scale from 0 to 3; 0, no presence of host response characteristic; 1, minimal presence; 2, moderate presence; 3, extensive presence. Parameters were measured within the scaffolds (n = 8).

GES, genipin scaffold; GES+M0, M0 cells seeded on genipin scaffold; GES+M1, M1 cells seeded on genipin scaffold; GES+M2, M2 cells seeded on genipin scaffold; SD, standard deviation.

FIG. 2.

Histology micrographs of H&E-stained explanted scaffolds. Genipin-cross-linked scaffold fibers are marked with an asterisk (*). The scale bar is 50 μm. H&E, hematoxylin and eosin.

Table 2.

Tissue Integration Evaluation

	Cellular infiltration		Collagen deposition
	3 Weeks	3 Months	3 Weeks	3 Months
GES	2 ± 0	3 ± 0	1 ± 0	2.2 5 ± 0.5
GES+M0	3 ± 0	3 ± 0	2 ± 0	2.25 ± 0.5
GES+M1	3 ± 0	3 ± 0	2 ± 0	2.75 ± 0.5
GES+M2	3 ± 0	3 ± 0	2 ± 0	2.75 ± 0.5

Scoring for tissue integration parameters reported as the mean ± SD at 3 weeks and 3 months. Scores were given on a scale from 0 to 3; 0, no presence of host response characteristic; 1, minimal presence; 2, moderate presence; 3, extensive presence. Parameters were measured within the scaffolds (n = 8).

Histology and IHC

MT staining indicated abundant collagen deposition (blue) in the interstitial regions of collagen constructs at all time points and all groups (Fig. 3A). Fibrous capsule thickness was significantly reduced in the GES+M2 group at 3 weeks and 3 months (p < 0.0001). Conversely, an increased capsule thickness was observed in the GES+M1 group compared with the GES and GES+M0 groups at 3 months (Fig. 3C). High magnification images of MT-stained specimens can be found in Supplementary Figure S7.

FIG. 3.

Masson's trichrome staining of explanted images. (A) Representative specimens from each group demonstrate the differences between groups in terms of fibrous capsule structure. Scaffolds with surrounding tissue were indicated by a dotted yellow line. (B and C) Quantification of thickness of fibrous capsule at 3-week and 3-month time points, respectively. Significance was determined using the Mann–Whitney U tests (p < 0.05). Letters above groups in the plots indicate overall grouping in terms of significance (i.e., “a” were not significantly different from one another, “a” were significantly different from “b”). Error bars indicate standard deviation in each graph (n = 8). Collagen fibers are marked with an asterisk (*). Scale bar = 200 μm.

Collagen threads degraded significantly between 3 weeks and 3 months (p < 0.001; Fig. 4B and C) accompanied by a significantly greater amount of interstitial collagen (p < 0.001; Fig. 4D and E), implying that M2 macrophage-seeded collagen scaffolds were gradually remodeled and replaced by host connective tissue.

FIG. 4.

Fraction of total area occupied by collagen thread and interstitial collagen. (A) An example of grid (25 × 25 μm) utilized for point counting assessment of M2 cells seeded histological slides. (B and C) Quantification of collagen threads at 3-week and 3-month time points, respectively. (D and E) Quantification of interstitial collagen at 3-week and 3-month time points, respectively. Significantly more interstitial collagen within the continuum of the scaffold structure was noted for the GES+M2 group and less of the total area was occupied by GES+M2. Significance was determined using the Mann–Whitney U tests (p < 0.05). Letters above groups in the plots indicate overall grouping in terms of significance (i.e., “a” were not significantly different from one another, “a” were significantly different from “b”). Error bars indicate standard deviation in each graph.

Immunohistochemical staining confirmed type I collagen to be present in all groups, with more prominent staining in the GES+M2 group compared with the other treatment groups (Figs. 5 and 6). At 3 weeks and 3 months, a significantly greater number of Arg1/Ym-1-positive cells were observed in GES+M2 when compared with all the groups. The M2+GES group demonstrated low α-SMA expression at both time points in comparison to all the groups (p < 0.001) (Fig. 7).

FIG. 5.

Immunohistochemistry results of IL-1β (M1 marker), Ym-1 (M2 marker), iNOS (M1 marker), arginase (M2 marker), collagen type III, collagen type I, and α-SMA (myofibroblast marker) at 3 weeks. An asterisk (*) indicates collagen threads of the scaffold. Scale bar is 150 μm.

FIG. 6.

Immunohistochemistry results of IL-1β (M1 marker), Ym-1 (M2 marker), iNOS (M1 marker), arginase (M2 marker), collagen type III, collagen type I, and α-SMA (myofibroblast marker) at 3 months. An asterisk (*) indicates collagen threads of the collagen scaffold. Scale bar is 150 μm.

FIG. 7.

Immunohistochemical identification of M2 macrophage (Arg1/Ym-1) and α-SMA (myofibroblast) markers at 3-week and 3-month time points. Significance was determined using the Mann–Whitney U tests (p < 0.05). Letters above groups in the plots indicate overall grouping in terms of significance (i.e., “a” were not significantly different from one another, “a” were significantly different from “b”). Error bars indicate standard deviation in each graph.

Semiquantitative analysis demonstrated that M2 markers, arginase and Ym-1, were present in all the groups, with greater staining noted in the GES+M2 group (Table 3). M1 markers, iNOS (positive control is shown in Supplementary Fig. S8) and IL-1β, were notably absent in all the groups at both time points. Collagen type III and α-SMA were present in all repair groups, with abundant staining observed in the GES+M1 group and minimal staining present in the GES+M2 group at both time points (Table 3). Collagen type 1 was also present in all the groups with more abundant staining in the GES+M2 group.

Table 3.

Semiqualitative Immunohistochemical Assessment

	ARG1/Ym-1		iNOS/IL-1β		Collagen I		Collagen III		α-SMA
	3 Weeks	3 Months	3 Weeks	3 Months	3 Weeks	3 Months	3 Weeks	3 Months	3 Weeks	3 Months
GES	2/3	1/3	0/3	0/3	2/3	2/3	1/3	1/3	2/3	1/3
GES+M0	2/3	2/3	0/3	0/3	2/3	2/3	1/3	1/3	1/3	1/3
GES+M1	2/3	1/3	0/3	0/3	1/3	1/3	3/3	2/3	3/3	2/3
GES+M2	3/3	3/3	0/3	0/3	2/3	3/3	1/3	1/3	1/3	1/3

Discussion

With a better understanding of the role macrophages play in the tissue healing processes, particularly with respect to early events following biomaterial implantation, it is feasible to design novel tissue-engineered constructs capable of harnessing immune responses for enhanced regeneration. As far as we are aware, this is the first report assessing histological and biomechanical properties of collagen scaffolds seeded with different macrophage subtypes in vivo. Regardless of seeded macrophage subtype, all the groups showed the presence of M2 macrophages after 3 months in vivo as indicated by arginase and Ym-1 expression.

However, M2 macrophage-seeded collagen scaffolds revealed encouraging host responses, such as decreased fibrous capsule thickness and reduced granulation tissue formation. Furthermore, the GES+M2 group displayed greater deposition of interstitial collagen following collagen thread resorption. These findings imply that the proposed genipin-cross-linked collagen biomaterials supplemented with M2 macrophages could play an essential role in developing future strategies aimed at regenerating connective tissues.

The end stage of the foreign body reaction and healing response to a biomaterial involves the formation of a collagenous and vascular fibrous capsule that confines the implant and precludes implant integration with the surrounding tissue. Fibrous encapsulation of an implanted medical device, commonly referred to as scar formation, can be a major impediment to device function. Various implant modification methods have been adopted in attempts to diminish fibrous capsule formation.^24–26 Our study showed a significant decline in the fibrous capsule thickness for the M2-seeded group.

The M2 phenotype of macrophages, referred to as “alternatively activated,” consistently releases anti-inflammatory cytokines that aid in tissue remodeling. The presence of anti-inflammatory cytokines assisting in tissue remodeling greatly improves biomaterial integration and enables it to fulfill its intended function by inhibiting fibrous capsule formation. Reportedly, IL-4 from coated mesh promotes shifted early-stage macrophage polarization and was associated with positive long-term effects. An IL-4-eluting coating increased the percentage of M2 macrophages at the implant surface and diminished fibrotic capsule formation.²⁷

In another study, IL-10 delivery via a poly(lactide-co-glycolide) scaffold significantly decreased the foreign body reaction by suppressing the implantation-induced inflammatory reaction.²⁸ However, M1 macrophages are believed to increase the thickness of the fibrous capsule by upregulating inflammatory responses, similar to the results of this study.¹⁷ This is congruent with the previously reported literature in that direct delivery of M2 macrophages reduced fibrous capsule formation. This implies that M2 macrophages can be exploited to achieve reduced or, potentially, scar-free healing. At this point, precise mechanisms underlying the reduction of fibrous capsule formation associated with M2 macrophages remain to be elucidated.

The biomechanical properties of tissue-engineered scaffolds are crucial to aid in repair and regeneration of connective tissue defects. The increase in apparent modulus for the GES+M2 group when compared with the GES group highlights that M2 macrophage supplementation improved structural and mechanical properties at the tissue and cellular level. Attaining an increased modulus with a decreased fibrous capsule thickness indicates that mediation improved both quality and quantity of tissue deposition within the scaffolds seeded with the M2 subtype.

Genipin-cross-linked collagen threads are significantly stiff, and there is a large pore volume in the basic scaffold between the collagen threads. In implanted scaffolds, the pore volume between the threads is filled with de novo collagen, which increases the effective stiffness of the tissue-integrated implant. Biomechanical testing data also highlighted possible synergistic effects of combining M2 macrophages and genipin-cross-linked collagen scaffolds concerning increasing stiffness of the tissue healing.

Despite notable degradation of implanted GES+M2 scaffolds at 3 months, findings indicated the presence of M2 macrophages between threads with induction of a significantly higher degree of alignment of newly deposited host tissue. Histological findings are notable in that the tissue regenerated within and around the scaffolds was composed primarily of collagen type I, and the addition of M2 macrophages on genipin-cross-linked collagen scaffolds could further contribute to the robustness of material. Overall, results suggest that M2-seeded scaffold group experienced a greater degree of remodeling of originally seeded threads with de novo regenerated tissue.

Many efforts are undertaken to avoid cytotoxicity in the use of biomaterials. In this study, collagen scaffolds were cross-linked utilizing genipin. The degree of cytotoxicity of a cross-linking agent and the fates of macrophage subtypes on genipin-cross-linked collagen scaffolds were assessed in a recent publication from our group.¹⁸ Therefore, data were interpreted based on histological analysis as a whole, as the results of macrophage characterization on genipin-cross-linked collagen scaffolds in vitro were investigated and reported previously. Genipin, the cross-linker used in our study, is derived from the fruit of Gardenia jasminoides. Genipin is superior to other chemical cross-linking agents because of its lower toxicity and anti-inflammatory properties, which have been used widely in Asian medicine.^29,30

Immunohistochemical staining in our previous study indicated a greater amount of proregenerative M2 macrophages accumulated around genipin-cross-linked scaffolds.³¹ Our recent study in vitro showed polarization of M0 macrophages toward an M2 phenotype following seeding on genipin-cross-linked collagen scaffolds.¹⁸ Our recent study in vitro also showed that M2 polarization of M0 macrophages occurs when cell culture media was treated with genipin. M0-seeded treatment group may have been guided initially by the scaffold to behave similarly to M2 because of the direct effect of the genipin scaffold.

However, the effect of the genipin scaffold alone and M0-seeded groups were not as effective as the M2-seeded treatment group. Therefore, we propose that the maximum regenerative effect requires direct initial induction and seeding of M2 cells. Our past study in vitro also indicated that M1-macrophages maintained their phenotype when seeded on genipin-cross-linked collagen threads. Thus, we do not believe that the scaffold model altered the effects of delivered M1 macrophages.

It should also be noted that M2 cells could express iNOS, and M1 cells could express Arg1, but the expression level is relatively low. Western blot data showed that rat bone marrow-derived M2 macrophages did not express iNOS, and M1 cells did not express Arg1, as reported by similar studies.^18,32,33 Ym-1 and arginase are considered M2 markers, which were abundant in GES+M2 and contribute to tissue repair. M1 markers, iNOS and IL-1β, were absent in all the groups.

Histological examinations of the GES+M2 group showed robust interstitial collagen production and the presence of collagen type I within the newly deposited tissue within the continuum of the scaffold structure. Collagen type I is a significant extracellular matrix (ECM) component and plays an essential role in ECM reorganization during tissue regeneration.³⁴ Immunohistochemical analyses also revealed that GES+M1 exhibited a substantial presence of collagen type III, a strong indicator of fibrous tissue or scar formation at the implant site.

M2 macrophages are known to promote collagen deposition.^35,36 Our study showed abundant collagen accumulation around implants supplemented with M2 macrophages. The greater degree of thread degradation in the GES+M2 group suggests accelerated scaffold resorption and tissue remodeling. Increased collagen type I/III ratio in the interstitial matrix implies the formation of robust connective tissue compared with the scar-like formation observed in the GES+M1 group. Besides, a diminished ratio of collagen type III over collagen type I was observed in fibrotic capsule of the GES+M2 group indicating the absence of prolonged inflammation based on semiqualitative histological assessment. Some studies report opposite outcomes regarding collagen accumulation and induction by M2 cells.^37,38

The GES+M2 group exhibited higher cell density of the M2 macrophages within the continuum of the scaffold in comparison to other groups. Initial administration of M2 could possibly contribute to shortening the duration of the inflammatory phase of wound healing and expediting the inception of tissue regeneration. The abundant presence of M2 macrophages within the scaffold could be of two potential origins. Due to paracrine effects, initially seeded M2 macrophages may have attracted host M2 cells to the area.

The presence of M2 macrophages on GES alone implies that host cells are capable of infiltrating and synthesizing de novo tissue as the macroporous structure of the scaffold enabled facile cell seeding and suitable penetration of the cells. It is possible that cell-seeded group may have benefited from newly deposited connective tissue from both host M2 macrophages and allogeneic M2 cells. However, the experimental design of this investigation did not allow for distinguishing between the host and allogeneic cell sources. To address this limitation, future studies could consider implantation of M2 macrophages that stably express a label or distinguishing marker.

It is worth noting that M2 macrophages are key regulators of tissue repair.³⁹ However, when inflammation is not controlled, persistent activity of M2 macrophages may be deleterious to attaining tissue homeostasis. Excessive M2 macrophage activation can promote proliferation of myofibroblasts.^40,41 In this scenario, M2 exemplifies a breakpoint between wound healing and exacerbation of fibrotic process. Our data suggested diminished expression of α-SMA in the GES+M2 group, which is congruent with a reduced fibrous capsule thickness indicating controlled tissue repair by initial seeding with prohealing cells.

This study was not without limitations. The presence and efficacy of cell-seeded scaffolds needs to be evaluated at longer time points than used in this study (≥1 year). Additionally, macrophage subtype tracking in vivo was not utilized to determine their distribution, differentiation, and viability at the implant site. Furthermore, more samples will be required to overcome testing limitations present in this study for histological scoring.

In conclusion, our study demonstrates improvements in biomechanical and histological outcomes following the incorporation of macrophage subtypes into a genipin-cross-linked collagen biotextile for tissue repair and remodeling. Cell seeding with anti-inflammatory prohealing M2 macrophages resulted in an increase in interstitial collagen by host and decreased fibrous encapsulation and foreign body reaction at the interface of genipin-cross-linked collagen scaffolds.

Disclosure Statement

O.A. and A.H. disclose employment and equity ownership in association with CollaMedix Inc. O.A. is an inventor on a patent that is associated with electrochemical compaction of collagen as threads.

Funding Information

This study was supported by grant from the National Institutes of Health (NIH): R21HD095439 (O.A. and A.H.) and T32 AR007505 (P.M.).

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Figure S6

Supplementary Figure S7

Supplementary Figure S8