Introduction

Upon implantation of biomaterials, an immune response is initiated. Thus, it is crucial to avoid a persistent and detrimental inflammation that could cause the failure of the implant.13 There has been an increasing interest in strategies to modulate the immune response to biomaterials, leading such response towards regeneration and remodeling.5,29 Despite the advancements, an unanimous consensus on the most efficient and reliable approach to trigger the appropriate immune responses to biomaterials is still lacking, mainly due to the absence of an in-depth understanding of the host inflammatory response to implants.

Among immune cells, macrophages play a crucial role on implants’ outcome. In fact, macrophages, ubiquitous in the body, are involved in tissue surveillance and homeostasis.15,23 Thus, they are one of the first immune cells to intervene after biomaterials implantation, and eventually initiate a foreign body response.23 This role of macrophages has attracted the attention of tissue engineers, who are studying macrophages’ early response to biomaterials, as a predictor of the in vivo outcome of the implant.4,6,13 It has been established that classically activated macrophages (also called M1) appear at the site of the injury at the early stage (day 1–5) (Fig. 1).27 Only after day 4, they are replaced by the alternatively activated macrophages (generally called M2). M1 macrophages have been reported to be pro-inflammatory, and produce high levels of oxidative metabolites and pro-inflammatory cytokines following an injury or pathogen infection; in vitro, this can be modeled through the administration of inflammatory mediators (e.g., INF-γ, LPS, TNF-α).16 On the contrary, M2 macrophages are activated by multiple cytokines, such as interleukin-4 (IL-4), -13 (IL-13), -10 (IL-10) and transforming growth factor-β (TGF-β).4,17,25 They support tissue repair by producing anti-inflammatory molecules such as IL-10 and TGF-β,10 triggering angiogenesis (by releasing platelet-derived growth factor and vascular endothelial growth factor) and matrix remodeling, while suppressing the inflammation.17 During the healing process, the time dependent switch to the M2 phenotype resolves the M1-mediated inflammatory reaction, and prompts tissue regeneration.6 The ability to control this polarizing switch can reduce scar tissue formation and favor the integration of the implant within host’s tissues.1 Thus, the administration of immunomodulating cytokines (such as IL-4) could be beneficial and effective in promoting the phenotypic switch of macrophages towards the alternatively activated lineage, if their release is timely controlled. It has been reported that macrophages, when treated with IL-4, assume an anti-inflammatory phenotype. IL-4 activates the transcription of genes typical of the M2 lineage, such as Il10, Mrc1, and Arg1.14,26 A controlled release of IL-4 could be beneficial to tune the early phase of inflammation, however the short half-life of IL-4 prevented its use.33

Figure 1
figure 1

Schematic representing the appearance of alternatively activated macrophages (blue) at the site of injury over time, replacing the classically activated macrophages (red) at the site of inflammation.

Full size image

In the past decade, collagen-based scaffolds have been successfully engineered with nanostructured delivery systems,21 for the local and controlled release of bioactive molecules, to reducing their side effects to the surrounding tissues, while presenting their bioactivity.7

Although delivery systems proved valuable tools in granting scaffolds the ability to trigger specific cellular mechanisms by the release of bioactive payloads, they also contribute to elicit an immune response towards the implant.11

Recently, we proposed a collagen-based scaffold, functionalized with poly(lactic-co-glycolic acid)—multistage porous silicon vector composite microspheres (PLGA-MSV).21 We demonstrated that concealing PLGA-MSV with a thin layer of type I collagen of the scaffold itself, it was possible to control their retention in the scaffold, as well as the release kinetic of their payload. Recently, this scaffold functionalization strategy proved to prevent early uptake of the delivery systems by macrophages, preserving the functionality of the payload and the desired release kinetic.18

In the present study, we investigated if the release of an anti-inflammatory cytokine from a collagen-based scaffold could induce the expression of M2-associated genes in macrophages, in vitro, and the recruitment of CD206+ macrophages in vivo. As a proof of concept the scaffold was functionalized with IL-4 and a controlled almost-zero order release was set up. Secondly, we confirmed the amount of IL-4 necessary for rat bone marrow derived macrophages (BMDM) to express Mrc1, Il10 and Arg1 genes. Finally, we tested the selected scaffold functionalized with IL-4-releasing PLGA-MSV (PLGA-MSV/IL-4 Scaffold), both in vitro with rat BMDM, and in vivo, in a rat subcutaneous pouch model. The subcutaneous pouch model was used as a pilot study model, to confirm in vivo the findings obtained in the in vitro 3D culture study.

Materials and Methods

MSV Fabrication and Surface Modification

Discoidal MSV of 1 μm in diameter and 400 nm thick were fabricated by an optimized photolithography and electrochemical porosification method.8,31 MSV particles were oxidized and subsequently modified with (3-Aminopropyl)triethoxysilane (APTES), as described elsewhere.20

Loading of IL-4 in MSV

IL-4 (Peprotech) was loaded on APTES-modified MSV particles, through a loading procedure, previously optimized.20 Briefly, 30 µg of IL-4 was reconstituted in 300 µL of PBS (Gibco). APTES-modified MSV, previously dried in a vacuum oven (Thermo Scientific) over night, were dispersed in the IL-4 solution, and mixed at 300 rpm in a thermomixer (Thermo Scientific), for 2 h at 37 °C. Particles were then recovered, and lyophilized. The loading efficiency of IL-4 in MSV was calculated by measuring the IL-4 left in the loading solution, by enzyme-linked immunosorbent assay (ELISA) (R&D Systems). Three separate samples were prepared.

PLGA-MSV/IL-4 Fabrication

PLGA-MSV were loaded with IL-4 as follow (PLGA-MSV/IL-4): PLGA-MSV containing three different concentrations of IL-4 (0.5, 5 and 50 ng/mg of PLGA) were prepared through an optimized double emulsion method. Based on the loading efficiency of IL-4 in MSV particles (89.86%), the necessary amount of MSV was added to 500 µL of 50:50 PLGA (LACTEL) in dichloromethane (DCM) (Sigma-Aldrich) (50 mg/mL), and mixed thoroughly, as previously optimized.20 The DCM-based dispersion was dropped into 10 mL of a 1% polyvinyl alcohol (PVA) (88% hydrolyzed, ACROS Organics™) in aqueous solution, and homogenized at 3000 rpm for 5 min (IKA Eurostar). The emulsion was dropped into 40 mL of 1% (w/v) aqueous solution of PVA, and mixed at 750 rpm for 5 h, to allow for the evaporation of the solvent, at room temperature. The microspheres were spin down, washed three times with PBS by centrifugation (Legend X1R Centrifuge, Thermo Scientific) and lyophilized for 24 h. Empty PLGA-MSV were produced as control. The loading efficiency of IL-4 in PLGA-MSV particles was calculated by measuring the remaining IL-4 in the loading and washing solutions, by ELISA assay.

PLGA-MSV’s morphology and size (n = 100) was characterized by scanning electron microscopy (SEM) (FEI Nova NanoSEM 230). Microspheres were sputtered coated with platinum, by a Plasma Sciences CrC-150 Sputtering System (Cressington 208HR), and examined under a voltage of 7 kV. Also, PLGA-MSV were observed at the optical microscope (Eclipse Ti-E Nikon).

In Vitro Administration of IL-4 to BMDM

BMDM were extracted from rat bone marrow, as previously reported.18,30 Briefly, after sacrificing adult Lewis rats (n = 3), their femurs were cleaned of surrounding tissues and cut at both ends. The cavity was flushed with complete media, using a 5 mL syringe and a 25-gauge needle. Bone marrow cells were mechanically separated into single-cell suspension, filtered and cultured in media supplemented with macrophage colony-stimulating factor (10 ng/mL).

Once obtained, rat macrophages were seeded at the density of 40,000 cells/cm2 and maintained in High Glucose-Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum, 1% penicillin (100 UI/mL)-streptomycin (100 mg/mL), and 0.25 mg/mL amphotericin B at 37 °C in a humidified atmosphere (90%) with 5% CO2. Plated macrophages were exposed for 24 h to three different concentrations of IL-4 (0.5, 5 and 50 ng/mL), soluble IL-4 (20 ng/mL, positive control)2,25 or left untreated (CTRL). The experiment was performed in triplicate.

Confocal Laser Microscopy

To evaluate the expression of the mannose receptor, in response to administration of IL-4 in vitro, the treated macrophages were stained with anti-mannose antibody (ab64693, abcam), Wheat Germ Agglutinin (WGA, Life Technology), Alexa Fluor® 488 Conjugate and DAPI (Life Technology) according to manufacturers’ protocols. Samples were imaged with a confocal laser microscope (A1 Nikon Confocal Microscope, Nikon), and the images were analyzed through the NIS-Elements software (Nikon).

Subsequently, after scaffold fabrication (see following paragraph) also the PLGA-MSV/IL-4- functionalized scaffold was imaged by confocal laser microscopy and Z-stacks were acquired.

Fabrication of PLGA-MSV/IL-4 Scaffold

The collagen scaffold integrated with IL-4-loaded PLGA-MSV was fabricated as previously described.18,21 A 20 mg/mL of type I collagen (Nitta Casings Inc.) in an acetate buffer (pH 3.5) was prepared. The collagen suspension was buffered up to pH 5.5 with sodium hydroxide (0.1 M). The precipitated collagen was washed three times with DI water. The collagen was cross-linked through dispersion for 48 h in a 1,4-butanediol diglycidyl ether (BDDGE) (Sigma-Aldrich) aqueous solution (2.5 m M), setting up a BDDGE/collagen ratio of 1:100 w/w. The collagen was washed 3 times in DI water.

The cross-linked collagen was resuspended in DI water (10 mg/mL), and PLGA-MSV loaded with IL-4 (50 ng IL-4/mg of PLGA) were added to the collagen dispersion, for a final concentration of 50 ng of IL-4/scaffold. Collagen scaffolds with empty PLGA-MSV were used as controls (CTRL). The materials were freeze-dried with a controlled freezing ramp from 25 °C to −25 °C and a heating ramp from −25 to 25 °C in 50 min under vacuum conditions (P = 0.20 mbar). Discoidal collagen scaffold (7 mm × 1 mm) were produced. The evaluation of scaffolds’ morphology was performed through SEM and the measurement of its pores’ size by confocal laser microscopy, through the software NIS-Elements. The water contact angle of the scaffolds was measured to determine their hydrophilicity. The water contact angle was measured with a standard contact angle goniometer with DROPimage Standard system (Ramé-Hart). Each sample (n = 5) was tested in 5 different points, and the averaged reading was plotted over 5 min.

In Vitro Release of IL-4

The cumulative release of IL-4 from PLGA-MSV/IL-4 and from PLGA-MSV/IL-4 Scaffolds was assessed in vitro by ELISA, according to manufacturer’s protocol (R&D Systems). Three separate batches of PLGA-MSV/IL-4 were dispersed in PBS at 37 °C in mild agitation (100 rpm). At set time points particles were spun down, and 10% of the supernatant was collected. Similarly, CTRL and PLGA-MSV/IL-4 Scaffold with the highest concentration of IL-4 (50 ng/scaffold) were placed in scintillation vials, covered with PBS, and mixed at 100 rpm at 37 °C, pH 7 (physiological-like conditions). Every day, 10% of the supernatant was collected for analysis, up to two weeks.

BMDM 3D Culture on PLGA-MSV/IL-4 Scaffolds

Scaffolds (n = 3) were placed in a 24 well plate and seeded with 3 × 105 BMDM. Cells were deposited on the scaffold as a concentrated drop of 50 µL, and allowed to adhere to the scaffold for 20 min, before adding media (2 mL or complete media). At 24, 48 and 72 h scaffold were washed in PBS three times and processed for gene expression analysis and ELISA. The experiment was performed in triplicate.

Rat Subcutaneous Pouch

Adult Lewis rats (n = 12; Charles River Laboratories. Houston, TX, USA) were used for the in vivo study. The study was performed in conformity with the guidelines established by the American Association for Laboratory Animal Science. All procedures were approved by the Houston Methodist Institutional Animal Care and Use Committee (Protocol: AUP-1111-0058, approval date: 2/8/2012). Rats were administered analgesia and subsequently anesthetized using inhaled isoflurane gas. The dorsum of each animal was shaved. Under sterile conditions, four incisions were made (two per side of the dorsal midline), of each animal. Into each subcutaneous pocket one scaffold was placed (1 cm ×1 mm). Animals were euthanized at 24 and 72 h. At each time point, samples were collected for further analysis (histology, qPCR and flow cytometry). In particular, at each time point 6 animals were implanted with 4 scaffolds each. From each animal, 2 scaffold were used for histological evaluation, 1 for flow cytometry and 1 for q-PCR. At each time point collagen scaffolds were evaluated as CTRL.

Histology

Masson’s trichrome staining was performed to evaluate cell infiltration in the scaffolds implanted subcutaneously. Scaffolds and surrounding tissues were collected at the pre-defined time points, and embedded in optimum cutting temperature mounting media (Tissue Tek) and frozen. Frozen tissues were sliced using a cryostat at a thickness of 15 µm (Thermo Scientific) and washed twice in xylene and rehydrated sequentially with decreasing ethanol concentrations (100, 95, 90, 80, and 70%) and distilled water. Tissue slices were stained using a kit for Masson’s trichrome staining (Abcam; ab150686), according to manufacturer’s protocol. Stained slices were mounted with Cytoseal XYL (Thermo Scientific) mounting medium and then imaged with a histological microscope (ECLIPSE Ci-E, Nikon).

Scanning Electron Microscopy

The scaffolds recovered from the rat subcutaneous pouches, were also prepared for SEM imaging. After recovery, the scaffolds were washed in PBS three times, and fixed with 2.5% glutaraldehyde (Electron Microscopy Science). Scaffolds were dehydrated in increasing concentrations of ethanol (up to 100%), vacuum dried, and coated with palladium (Hummer 6.2 Sputtering System, Anatech Ltd.). Samples were imaged at a voltage of 10 kV.

Flow Cytometry

At 24 and 72 h scaffolds (n = 6) were collected and digested in 2 mg/mL collagenase I (Worthington Biochemical Corp.), for 30 min at 37°C. Cells were recovered by spinning down at 500 g for 5 min, washed three times with PBS, and then fixed in 70% ethanol. After fixation, cells were washed with FACS buffer (BSA 0.1%). Cells were labeled with directly conjugated anti-macrophages PE (eBiosciences), CD3 BV605, CD45 PECy7, anti-CD206 FITC (Biorbyt). Cells were pooled, and a minimum of 10,000 events per sample was analyzed using a BD LSR Fortessa™ cell analyzer (BD Biosciences, San Jose, CA). Data analysis was performed using FlowJo (Tree Star, Ashland Inc., O).

Gene Expression Analysis

Scaffolds from the in vitro (at 24, 48, and 72 h) and ex vivo (72 h) experiments were lysed in 500 µL of Trizol reagent (Invitrogen). DNAse (Sigma) treatment followed. The cDNA was synthesized from 1 μg total RNA, using the iScript retrotranscription kit. Transcribed products were analyzed using commercially available mastermix, following appropriate target probes on an ABI 7500 Fast Sequence Detection System (Applied Biosystems, Foster City, CA) to evaluate the expression of the following genes: arginase (Arg1: Rn01469630_m1); mannose receptor (Mrc1: Rn01487342_m1); interleukin-10 (Il10: Rn01483988_g1); interleukin-6 (Il6: Rn01410330_m1); Tumor Necrosis Factor-alpha (Tnfα: Rn01525859_g1); interleukin 12-alpha (Il12b: Rn00575112_m1). Gene expression was normalized to the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (Gapdh; Rn01775763_g1). Values were normalized to those obtained from the negative control group (untreated cells, CTRL). The experiments were run in triplicate.

Enzyme-Linked Immunosorbent Assay (ELISA)

At each time point, the media was collected from all replicates and IL-10 produced by BMDM was quantified by ELISA kit (R&D Systems), according to manufacturer’s protocol.

Statistical Analysis

Statistical analysis was performed with the software GraphPad Prism, using a Two-Way ANOVA, and a post hoc Tukey’s multiple comparison test. For gene expression analysis one-way analysis of variance for multiple comparisons by the Student–Newman–Keuls multiple comparison test was used, as previously reported.9 All experiments were performed at least in triplicates. Data is presented as mean ± standard deviation. A value of p < 0.05 was considered statistically significant: *p < 0.05; **p < 0.01; ***p < 0.001; **** p < 0.0001.

Results

In Vitro BMDM Expression of M2-Associated Genes

To define the optimal concentration of IL-4-loaded PLGA-MSV to induce macrophages to express M2-associated markers, cells were treated with PLGA-MSV releasing different concentrations of IL-4, nominally 0.5, 5 and 50 ng/mL. As shown in Fig. 2, cells showed increased expression of the M2-associated markers tested, if treated with PLGA-MSV releasing higher concentration of IL-4, compared to untreated BMDM (CTRL). Among them, the particles releasing a concentration of 50 ng/mL of IL-4 were as effective as the soluble IL-4 in inducing a M2-associated phenotype in the macrophages. Following the exposure to PLGA-MSV/IL-4 releasing IL-4 at the highest concentration a 3.47 (± 0.8), 5.52 (± 0.9), and 15.31 (±1.3) fold increase was observed for Il10, Arg1 and Mrc1, respectively (Figs. 2a–2c). The expression of Mrc1 for the PLGA-MSV/IL-4 50 ng group was slightly significantly lower compared to its respective Free IL-4 group (*p < 0.05).

Figure 2
figure 2

Rat BMDM’s in vitro expression of M2-associated genes (a–c), n = 3. Confocal laser microscopy images of CTRL (d) and BMDM treated with PLGA-MSV/IL-4 releasing the highest concentration of IL-4 (50 ng/mL) (e); scale bar: 25 µm. Nuclei are rendered in white, their membrane in red and the mannose receptor in blue. Quantification of the mannose receptor by fluorescence intensity of CTRL and of BMDM treated with PLGA-MSV/IL-4 50 ng (f). Values represent mean ± standard deviation. A value of p < 0.05 was considered statistically significant: *p < 0.05.

Full size image

Confocal laser microscopy revealed a higher expression of CD206 on the macrophages treated with the highest concentration of IL-4 (Figs. 2d, 2e), which was found to increase of approximately fourfolds compared to the CTRL (Fig. 2f).

Characterization of PLGA-MSV/IL-4 and In Vitro Release

Based on the expression of key M2-associated markers in response to different doses of IL-4, PLGA-MSV/IL-4 releasing 50 ng/mL was elected the formulation of choice, to treat macrophages. Thus, collagen scaffolds have been functionalized such that each scaffold (7 mm × 1 mm) would accomplish a concentration of 50 ng/mL, in vitro. In detail, since in the 3D BMDMs-seeded scaffolds macrophages were embedded in 2 mL of culturing media, each scaffold was functionalized with an amount of PLGA-MSV/IL-4 to accomplish an overall 50 ng/mL concentration of IL-4.

Towards this end, firstly, PLGA-MSV/IL-4 were fabricated. Their morphology and size distribution were evaluated by SEM and optical microscopy (Figs. 3a, 3b). As shown, microspheres were quite uniform in shape, and the MSV particles were fully embedded in the PLGA microspheres. They displayed a homogeneous size distribution, with a mean diameter of 8.22 (±0.4) (Fig. 3c).

Figure 3
figure 3

SEM micrograph of PLGA-MSV/IL-4 (a), and optical microscopy image of PLGA-MSV, displaying the fully embedded MSV particles in brown (b); scale bars: 10 and 3 µm, respectively. In figure c it is reported the size distribution of the resulting PLGA-MSV (n = 100). Values represent mean ± standard deviation.

Full size image

Subsequently, PLGA-MSV functionalized scaffolds were fully characterized. Scaffolds were initially imaged by SEM to evaluate their overall porosity and pores’ size. The scaffold was found highly porous and characterized by an anisotropic porosity (Fig. 4a). At higher magnification it was possible to observe the PLGA-MSV that were integrated in the collagen walls of the scaffold’s pores (Fig. 4b). Z-stacks of the scaffold were acquired by confocal laser microscopy. In Fig. 4c the 3D rendering of the scaffold (white) integrated with the PLGA-MSV/IL-4 (blue) is reported. Through the Z-stacks of the PLGA-MSV Scaffolds, they appeared highly porous and with an average pores’ size of 19000 ± 1700 µm2, approximately 7% smaller than the pores of collagen scaffolds not functionalized with the microspheres (21,000 ± 2900 µm2) (Fig. 4d). Figure 4e shows the water contact angle measurements of PLGA-MSV Scaffolds, compared to a collagen scaffold (CTRL). PLGA-MSV Scaffolds resulted slightly less hydrophilic than the CTRL. The contact angle of the water drop became 0° within 1 min, vs. the almost 5 min needed by the drop in the PLGA-MSV Scaffold, to be completely adsorbed.

Figure 4
figure 4

SEM micrographs of the collagen scaffolds, showing its overall structure and porosity (a), the PLGA-MSV fully embedded in the collagen matrix of the scaffold, at higher magnification (b). 3D rendering of the PLGA-MSV (blue) Scaffold (white), acquired by confocal laser microscopy (c); Scale bars: 500, 5 and 20 µm. Measurements of the area of the scaffolds’ pores, obtained by confocal microscopy (number of scaffolds: n = 3; number of pores measured/scaffold: n = 50) (d). Contact angle for the CTRL scaffold and the PLGA-MSV Scaffold (n = 5) (E). In vitro cumulative release of IL-4 from PLGA-MSV/IL-4 and PLGA-MSV/IL-4 Scaffold (n = 3) (f). Values are reported as mean ± standard deviation.

Full size image

The in vitro release of IL-4 is reported in Fig. 4f. As previously demonstrated the integration of the PLGA-MSV in the collagen matrix of the scaffold, enabled for a further level of control over the release, allowing for an almost-zero order release kinetic as assessed by linear regression interpolation of the release data (Figure S1, supplementary information).

In Vitro BMDM 3D Culture: Gene Expression Analysis

The expression of classically and alternatively activated macrophages’ markers was evaluated in vitro, through 3D culture of BMDM on PLGA-MSV/IL-4 Scaffolds, at 24, 48 and 72 h.

Figure 5a shows a clear increasing trend in the expression of the maker associated to alternatively activated macrophages. A significant fold increase compared to untreated macrophages was assessed as 308.61 ± 32, 21.35 ± 1.8, 0.85 ± 0.06 for Il6, 0.29 ± 0.05, 0.14 ± 0.03, 5.71 ± 0.8 for Tnfα and 39.65 ± 5.2, 34.60 ± 6.8, 0.65 ± 0.06 for Il12b, at 24, 48 and 72 h respectively (***p < 0.001; ****p < 0.0001). On the contrary the expression of the markers associated to the alternatively activated macrophages decreased over time, and resulted 24.53 ± 15, 29.15 ± 5.9, 92.32 ± 11 for Il10, 0.09 ± 0.03, 0.05 ± 0.01, 107.54 ± 0.3 for Mrc1 and 8.92 ± 5.2, 14.73 ± 2.9, 16.59 ± 5.9 for Arg1 at 24, 48 and 72 h respectively (*p < 0.05, **p < 0.01,*** p < 0.001; ****p < 0.0001).

Figure 5
figure 5

BMDM’s in vitro gene expression analysis of the M1-associated genes Il6, Tnfα and Il12b, and of the M2-associated genes Il10, Mrc1 and Arg1, at 24, 48 and 72 h (n = 3) (a). Summary of relative-fold gene expression (b). IL-10 quantification at 24, 48 and 72 h, quantified by ELISA, n = 3 (c). Values are reported as mean ± standard deviation. A value of p < 0.05 was considered statistically significant: *p < 0.05; **p < 0.01; *** p < 0.001; ****p < 0.0001.

Full size image

In Vitro IL-10 Quantification

Similarly to the expression of the il-10 gene, also the production of the cytokine IL-10 exponentially increased over time, reaching 253.62 ± 17 at 24 h, 819.33 ± 11 at 48 h, and peaking at 1041.24 ± 41 pg/mL at 72 h, as summarized in Fig. 5b (*p < 0.05, **p < 0.01; ***p < 0.001). IL-10 was found to be produced in response to the IL-4 treatment, and it was quantified by ELISA (Fig. 5c).

Evaluation of Cell Infiltration by Histology and SEM

The IL-4 releasing scaffolds were tested in vivo, in a rat subcutaneous pouch model. Scaffolds were recovered at 24 and 72 h and the scaffolds were processed for histological evaluation and SEM imaging. Tissue slices were prepared and stained by Masson’s trichrome stain. The connective tissue stained in blue, nuclei were stained in dark purple, and cytoplasm was stained pink. In Fig. 6, the yellow dotted lines delimit the scaffold. Based on the histological evaluation, we found that cells infiltrated and migrated almost in the entirety of the scaffold within 72 h after implantation. The cell infiltration appeared comparable in both CTRL and PLGA-MSV/IL-4 scaffold groups, at 72 h. However, at 24 h it was observed a slightly higher amount of infiltrating cells in the PLGA-MSV/IL-4 Scaffold, compared to the CTRL, in correspondence of the edges of the scaffold. This was also confirmed at higher magnification through SEM. Overall, the presence of giant body cells was not observed, and the surrounding tissues did not show significant alterations.

Figure 6
figure 6

Masson’s trichrome staining of sections of PLGA-MSV/IL-4 Scaffolds recovered from a rat subcutaneous pouch at 24 h and 72 h. The interface between the scaffold and the surrounding rat tissues is indicated by a yellow dotted line; Scale bar: 500 µm. Insets show the cell infiltration in the scaffolds, and the microspheres are visible in the scaffold sections (indicated by the black arrow). In black and white, SEM images show the morphology of the infiltrating cells and their interaction with the matrix of the scaffold; Scale bar: 20 µm.

Full size image

Flow Cytometry

The overall amount of macrophages found in the scaffolds recovered from the rat subcutaneous pouch was quantified by flow cytometry, at 24 and 72 h (Fig. 7). The content of macrophages decreased between 24 and 72 h in both the CTRL and PLGA-MSV/IL-4 Scaffold. In the CTRL group the overall amount of macrophages decreased from 71% to 50%; while in the PLGA-MSV/IL-4 scaffold (Fig. 7a), the amount of macrophages remained almost unvaried, slightly decreasing from 46% to 38%. Thus, at both time points, the overall percentage of macrophages in the PLGA-MSV/IL-4 Scaffold groups resulted lower than in the CTRL. However, as early as 24 h, it was found that 84% of such macrophages were CD206+, compared to only the 2% in the CTRL (Fig. 7b). The percentage of macrophages expressing the markers associated to the alternatively activated macrophages remained almost unvaried at 72 h (3% for the CTRL, and 73% for the PLGA-MSV/IL-4 Scaffold).

Figure 7
figure 7

Quantification of the amount of macrophages among the infiltrating cells recovered from the scaffolds implanted in the rat subcutaneous pouch, by flow cytometry (n = 3) (a). Quantification of the CD206+ macrophages by flow cytometry (b). Expression analysis of the in vivo samples of the selected M1- and M2- associated genes, at 72 h (n = 3). Values are reported as mean ± standard deviation. A value of p < 0.05 was considered statistically significant: ** p < 0.01; ****p < 0.0001.

Full size image

Gene Expression Analysis of In Vivo Samples

In Fig. 7c, the results of gene expression analysis of Il6, Tnfα, Il12b and Il10, Mrc1, Arg1 at 72 h are reported, normalized to the CTRL. The expression of such genes was as follow: 29.52 ± 9.6 for Il6, 7.05 ± 0.02 for Tnfα, 9.49 ± 0.9 for Il12b, 80.11 ± 10 for Il10, 65.37 ± 2.7 for Mrc1 and 5.27 ± 0.4 for Arg1 (**p < 0.01; ****p < 0.0001).

Discussion

The classical pathway of macrophage activation is a well disclosed mechanism of immunity against pathogens and foreign bodies.10 Macrophages are on the first line of defense against foreign bodies, due to their phagocytic activity.15 For this reason, they play a crucial role also in the fate of grafts for tissue repair. Modulating their response to biomaterials currently represents the Holy Grail of tissue engineering. Currently, several alternative routes to enhance tissue restoration are been explored. The inhibition of leukocyte recruitment at the site of implantation, in order to shorten the inflammatory phase, has been difficult to achieve and proved largely unsuccessful.24 Moreover, the transition between an inflammatory to an anti-inflammatory environment seems to be a crucial step in tissue healing. Thus, it appears that a confined and regulated inflammatory reaction is instrumental to reestablish homeostasis.

Among immune cells, M2 macrophages have been found to actively participate to tissue regeneration, by secreting anti-inflammatory cytokines and growth factors.28 Among the modulating molecules that are produced in the context of tissue repair and homeostasis, IL-4 has attracted increasing interest for its role in macrophages’ alternative pathway of activation. For this reason the release of IL-4 at the surface of a biomaterial could favor its in vivo success. Some pioneer research groups have highlighted the benefit of using components of the extracellular matrix for the design of biomaterials for tissue engineering, for their ability to induce a favorable immune response.3,4 Among all of the components of the extracellular matrix, collagen is a material of choice, and has been successfully used for both soft and hard tissue repair.19,21

In this study, we aimed at shedding some lights on the possibility to favorably modulate the response of macrophages to collagen based scaffolds, by releasing immunomodulatory mediators. We developed a collagen scaffold functionalized with IL-4 loaded PLGA-MSV microspheres (PLGA-MSV/IL-4 Scaffold) in the attempt to promote the expression of M2-associated genes such as Il10, Mrc1 and Arg1 by macrophages. Towards this end, it was fabricated a PLGA-MSV/IL-4 Scaffold functionalized with an amount of IL-4 such as to support the in vitro polarization of BMDM.

Firstly, we encapsulated IL-4 into PLGA-MSV. They displayed a core composed of MSV particles fully encapsulated into the PLGA outer shell. PLGA-MSV proved effective in loading and retaining high amount of proteins and to control their release.20 Three different formulations of IL-4 loaded PLGA-MSV were tested in vitro with rat BMDM.10,15 This experiment had the purpose to determine the formulation to most efficiently induce the expression of the selected M2-associated genes in BMDM as reported in the literature.22 Furthermore, this initial experiment assessed the preservation of IL-4 bioactivity after encapsulation in PLGA-MSV. The PLGA-MSV of choice (PLGA-MSV/IL-4 50 ng) was integrated into a 3D porous collagen scaffold, through a method previously developed.18,21 The resulting PLGA-MSV/IL-4 Scaffold was fully characterized; our scaffold displayed slightly smaller pores and lower hydrophilicity than the CTRL (collagen scaffold), that can both be attributed to the presence of the hydrophobic PLGA microspheres,32 but not significantly to impact cell infiltration. The in vitro release of the payload was characterized over 2 weeks. This was a satisfying accomplishment, as we aimed to obtain an in vivo release of IL-4 for the first week after scaffolds’ implantation, during which macrophage polarization is observed (schematic 1).27 Subsequently, we established 3D cultures of rat BMDM on the PLGA-MSV/IL-4 Scaffold of choice. It was found that in vitro, the release of IL-4 did induce the upregulation of the M2-associated genes of interest, including Mrc1, which is a key marker for the alternative activation of macrophages.14 It was also observed that the BMDM over-expressed Il10 gene, vs. Il12, whose expression presented an evident decreasing trend over 72 h (Fig. 5). This correlated with the high and increasing amount of IL-10 that was found to be produced after 72 h by BMDM exposed to PLGA-MSV/IL-4 Scaffold. This data was found in agreement with that reported in the literature about IL-4-primed cells, that produce high levels of IL-10.12,22

It is currently widely recognized that alternatively activated macrophages appear at the site of inflammation within 3–4 days.27 24 h after implantation of the PLGA-MSV/IL-4 Scaffold in the rat subcutaneous pouch model we observed a higher cellular infiltration compared to the CTRL, but no alterations were found in the surrounding tissues, in neither groups tested. In both CTRL and treated groups, the majority of infiltrating cells were macrophages (anti-macrophage+ and CD45+). In CTRL more macrophages were found than in the PLGA-MSV/IL-4 Scaffold group. However, about 80% of the macrophages were CD206+ at as early as 24 h, in discrepancy with what reported in the literature.14 This finding opens to the need for further elucidation of the in vivo mechanism mediated by PLGA-MSV/IL-4 Scaffold. We also observed a significantly higher expression of the M2-associated genes Il10 and Mrc1 compared to the CTRL at 72 h.

Altogether, this data demonstrated that the release of IL-4 in vivo allowed for the recruitment of a significant higher amount of CD206+ macrophages in the scaffold, and only 24 h after implantation. However, it has not been possible to find a consistent pattern of gene expression at 24 h (data not shown). Macrophages are extremely plastic cells, and their phenotype can mutate very quickly in response to changes in the local biochemical milieu (e.g., production of pro- and anti-inflammatory mediators).16 Thus, it could be hypothesized that the inability to identify a clear trend of expression in the six genes we studied is due to the fact that, within the first 24 h, macrophages are exposed to dramatic stimuli changes, which could have led to the generation of different subsets of CD206+ macrophages. It has been proposed that at least three different M2 subsets exist (M2-a, -b and –c), however several authors have theorized the presence of macrophages presenting intermediate phenotypes, and thus that even more subsets could exist.23 Given the complexity of the matter, further investigation is needed for a more conclusive interpretation of macrophages’ response to engineered biomaterials releasing immunomodulating cytokines.

Facilitating the resolution of the inflammation through the controlled release of immunomodulating cytokines, such as IL-4, is an appealing strategy. Currently an increasing number of scientists are leaning towards this innovative strategy. As the underlying mechanism to reach tissues’ functional recovery is complex and includes multiple cellular cross-talks, further studies are needed to develop materials with immunomodulating abilities.

Conclusions

It has become increasingly clear that the microenvironmental cues presented by scaffolds (e.g., chemical properties, stiffness, topography or ligand presentation) are critical to elicit and modulate favorable cell responses. Despite the advances in understanding these cellular effects with both somatic and stem cells, the mechanisms of such cues on immune cells remains widely unknown. Although a simplified model of the complex interplay between biomaterials and macrophages, this study helped at shedding some light on part of the complex mechanisms regulating macrophages’ response to tissue engineering scaffolds. In particular, in vitro it was found that BMDM expressed the M2-associated genes Il10, Mrc1 and Arg1 when cultured in the PLGA-MSV/IL-4 Scaffolds; it was also observed the production of IL-10 over 72 h. In a rat subcutaneous model PLGA-MSV/IL-4 Scaffold promoted the recruitment of macrophages, the majority of which were found to be CD206+ and expressing the M2-associated genes Il10, Mrc1 and Arg1.

Additional studies are necessary to clarify the efficacy of the presented approach, however, the potential benefit of this strategy to the current biomaterials-driven healing approaches is suggested.