*3.2. In Vivo Experiment*

### 3.2.1. Histopathology

The skin of hamsters is constituted by the epidermis, dermis (papillary and reticular layers), hypodermis and a skeletal muscle layer beneath the hypodermis. After PLGA implantation, the inflammatory response induced by PLGA membranes was a well localized chronic inflammatory response in the implantation bed of the membranes, below the muscular layer, remaining during all period of observation. The histological features of the foreign body response induced by the PLGA membrane are represented in Figure 3A–L. The inflammatory response delimited the biomaterial, forming an interface with the adjacent connective tissue without presenting a fibrotic capsule in any timeline (Figure 3A–L), and was constituted by a foreign body reaction consisting mainly by mononuclear cells of the monocytic lineage including macrophages, epithelioid cells and multinucleate giant cells (Figure 3A–L). The membranes were located inside the inflammatory site, such as verified with polarizing microscopy, represented by semi-thin sections after 7 days (Figure 3A,B), and after 15 days (Figure 3D,E) post implantation (PI). Mononuclear cells (Figure 3C) and multinucleate giant cells (MNGCs) (Figure 3C yellow arrows) surrounded the membrane after 7 days PI; and a foreign body response (FBR) beneath the muscle layer was observed after 15 days PI (Figure 3F, arrow). Another histological image after 15 days PI is showing the FBR constituted Langhans-type multinucleate cells (LCs) (Figure 3G arrows), lymphocytes (Figure 3G square) and epithelioid cells (arrow) with an elongated shape in rows surrounding the membrane (Figure 3H arrows). After 30 days histological image shows the FBR formed as a well delimited lesion in the hypodermis (Figure 3I arrow), with bands of connective tissue (Figure 3J, thick arrows) and capillaries (Figure 3J, thin arrow) present inside the lesion. After 90 days PI, histological image shows persistence of the inflammation (Figure 3K, arrow) and presence of both MNGCs (Figure 3L, thin arrow) and LCs (Figure 3L, thick arrow).

The LCs (multiple nuclei disposed at the periphery of the cell) (Figure 4A, black arrows) increased in number from 7 to 30 days PI decreasing successively up to 90 days PI (Figure 4B). Associated to LCs, capillaries sections were present in high numbers at day 7 PI, decreasing along time (Figure 4C,D). A histological image of a region showing the presence of capillaries (Figure 4C, black arrows) along the evolution of the implanted PLGA (Figure 4C, asterisk) represents the PLGA-induced inflammatory reaction. The quantification of the number of capillaries is represented by the number of capillaries over time, showing that the number of capillaries decreased up to 90 PI (Figure 4D). Number of capillaries (transversal sections) were calculated on 20 photomicrographs randomly obtained from HE stained histological sections (objective lens 20×). Results were expressed as number of capillaries, /histological field ± standard error of the mean.

**Figure 2.** Primary hamster cheek pouch fibroblast-like cells cultured onto PLGA scaffolds coated with collagen I. (**A**): Phase-contrast image of primary fibroblast-like cells after 4 days. Scale bar: 100 μm; (**B**): CLSM image of collagen I-coated PLGA fibers. Scale bar: 20 μm; (**C**): CLSM image of fibroblast-like cells (3 × 10<sup>5</sup> ) onto PLGA fibers after 1 day (white arrows). Nuclei in Hoechst stain (blue) and cytoskeleton in conjugated Alexa 488 phalloidin stain (green) for actin. Scale bar: 30 μm. **Figure 2.** Primary hamster cheek pouch fibroblast-like cells cultured onto PLGA scaffolds coated with collagen I. (**A**): Phase-contrast image of primary fibroblast-like cells after 4 days. Scale bar: 100 µm; (**B**): CLSM image of collagen I-coated PLGA fibers. Scale bar: 20 µm; (**C**): CLSM image of fibroblast-like cells (3 <sup>×</sup> <sup>10</sup><sup>5</sup> ) onto PLGA fibers after 1 day (white arrows). Nuclei in Hoechst stain

(**D**): Cells adhered along different directions of PLGA fibers (white arrows). Scale bar: 30 μm. (**E**–**H**)

(blue) and cytoskeleton in conjugated Alexa 488 phalloidin stain (green) for actin. Scale bar: 30 µm. (**D**): Cells adhered along different directions of PLGA fibers (white arrows). Scale bar: 30 µm. (**E**–**H**) CLSM images of fibroblast-like cells nuclei exhibiting proliferation after 1, 2, 3, 6 and 12 days on sample surface; (**Ee**,**Ff**,**Gg**,**Hh**): Side view of the nuclei of cells grown onto the scaffold after 1 day (**Ee**), and the in depth migration of cells nuclei to the 3D directions of the scaffolds after 3 days (**Ff**, arrow), 6 days (**Gg**, arrows) and 12 days (**Hh**, arrows)—images of 3 different fields in the same sample exhibiting a great amount of cells through the bulk of the membrane, as depicted in images h1,2,3 (arrow heads); note cells going deeper in image h3; Scale bars: 20 µm. (**I**): Graph of the proliferation of cells onto collagen I-coated membranes seeded with 3 <sup>×</sup> <sup>10</sup><sup>5</sup> cells; (**J**): Graph of the expansion of cells in function of the quantity of inoculated cells: 10<sup>4</sup> (A) or 3 <sup>×</sup> <sup>10</sup><sup>5</sup> cells (B). \*\*\* *p* = 0.0001; \*\* *p* = 0.001 and \* *p* = 0.01; ns: not significant; (**K**): SEM image of PLGA fibers (white arrow) showing the network of collagen (white asterisk) in the interfibrillar spaces (scale bar: 5 µm). (**L**): SEM image of cells in contact with PLGA fibers (black arrows). Scale bar: 30 µm. (**M**): SEM image of elongated cells following PLGA fibers (white arrowheads), and cells in interfibrillar spaces showing different morphologies (white asterisks) (scale bar: 30 µm).

### 3.2.2. Immunohistochemistry and Morphometry

The inflammatory process was analyzed over time after PLGA membranes implantation. For this goal, the CD3 antigen was used to measure the total number of T cells, and the myeloperoxidase (MPO) reactivity for polymorphonuclear neutrophils (PMNs). Both immunohistochemistry assays quantified these cellular populations, which decreased up to 90 days, showing that the inflammatory process decreased along PLGA fragments degradation.

In the analysis, the total number of T cells (Figure 5A,C,E,G) showed variations over time. The number of T cells increased after 7 and 30 days PI and decreased after 15 and 90 days PI (*p* = 0.0002) (Figure 5A,C,E,G,I), while neutrophils (myeloperoxidase+ cells) levels were significantly elevated at the implantation bed after 15 days PI (*p* < 0.05), decreasing after 90 days PI (*p* = 0.004) (Figure 5B,D,F,H,J). Results are expressed as the number of CD3 or myeloperoxidase<sup>+</sup> cells/histological field ± SEM (standard error of the mean).

### 3.2.3. Transmission Electron Microscopy (TEM)

In this work we chose a high-resolution imaging technique of TEM to show detailed images of PLGA fragments in intimate contact with the cells over time. TEM images showed various populations of macrophages, epithelioid cells and multinucleated giant cells at the granulomatous inflammatory process induced by the PLGA membranes between 7 and 90 days PI (Figure 6A–H). The inflammatory process was observed in the periphery of PLGA, showing mononuclear cells forming a foreign body reaction (FBR) in contact with the extracellular matrix (ECM) 7 days PI (Figure 6A). In the periphery of PLGA, TEM image is showing PLGA fragments (Figure 6B, arrows) in contact with a multinucleate giant cell (Figure 6B(GC)) 7 days PI. Another TEM image shows PLGA fragments in contact with epithelioid cells (EP) (Figure 6C(EP)), presenting different morphologies, with prominent cytoplasmic extensions such as filopodia following the topography of PLGA fragments (Figure 6C, arrows) 7 days PI. PLGA fragments were found inside phagosomes (Figure 6D arrow), in the interior of the cell (Figure 6D, asterisk) or in the extracellular space (Figure 6E, arrow) 15 days PI. However, an interesting and rare light microscopy image shows LCs containing material inside (Figure 6F, arrows) 30 days PI, such as depicted by TEM images showing PLGA fragments in the interior of cells (Figure 6G, arrows) 30 days PI. Finally, TEM image 90 days PI show decreased fragments of PLGA between the cells (Figure 6H, arrows).

**Figure 3.** FBR induced by the PLGA membrane: histological features. (**A**): Semi-thin section of a sample in Spurr resin in toluidine blue stain showing the inflammatory response surrounding the membrane (thick arrow) 7 days post implantation (PI). Scale bar: 100µm; (**B**): Polarization light microscopy of the same image in A. Scale bar: 200µm; (**C**): Semi-thin section in Spurr resin and toluidine blue stain showing mononuclear cells and MNGCs (thin yellow arrows) surrounding the membrane (thick arrow) 7 days PI. Scale bar: 100 µm; (**D**): Semi-thin section in Spurr resin in toluidine blue stain showing the FBR 15 days PI. Scale bar: 100 µm; (**E**): Polarization light microscopy of the same image in D. Scale bar: 100 µm; (**F**): Section in Masson's trichrome stain showing the FBR in the hypodermis 15 days PI (arrow). Scale bar: 1 mm; (**G**): Section in Masson's trichrome stain showing the FBR 15 days PI; note LCs (arrows) and lymphocytes (square). Scale bar: 200 µm; (**H**): Section in HE stain showing epithelioid cells (arrow) surrounding the membrane. Scale bar: 50 µm; (**I**): Section in HE stain showing the FBR in the hypodermis after 30 days PI (arrow). Scale bar: 500 µm; (**J**): Section in HE stain showing the inflammatory response 30 days PI; connective tissue (thick arrows) and capillaries (thin arrow). Scale bar: 100 µm; (**K**): Section in Masson's trichrome stains showing the FBR in the hypodermis after 90 days PI; note more cells (arrow). Scale bar: 1 mm; (**L**): Section in HE showing MNGC (thin arrow) and LC (thick arrow) in the inflammatory reaction 90 days PI. Scale bar: 50 µm. **Figure 3.** FBR induced by the PLGA membrane: histological features. (**A**): Semi-thin section of a sample in Spurr resin in toluidine blue stain showing the inflammatory response surrounding the membrane (thick arrow) 7 days post implantation (PI). Scale bar: 100 µm; (**B**): Polarization light microscopy of the same image in A. Scale bar: 200 µm; (**C**): Semi-thin section in Spurr resin and toluidine blue stain showing mononuclear cells and MNGCs (thin yellow arrows) surrounding the membrane (thick arrow) 7 days PI. Scale bar: 100 µm; (**D**): Semi-thin section in Spurr resin in toluidine blue stain showing the FBR 15 days PI. Scale bar: 100 µm; (**E**): Polarization light microscopy of the same image in D. Scale bar: 100 µm; (**F**): Section in Masson's trichrome stain showing the FBR in the hypodermis 15 days PI (arrow). Scale bar: 1 mm; (**G**): Section in Masson's trichrome stain showing the FBR 15 days PI; note LCs (arrows) and lymphocytes (square). Scale bar: 200 µm; (**H**): Section in HE stain showing epithelioid cells (arrow) surrounding the membrane. Scale bar: 50 µm; (**I**): Section in HE stain showing the FBR in the hypodermis after 30 days PI (arrow). Scale bar: 500 µm; (**J**): Section in HE stain showing the inflammatory response 30 days PI; connective tissue (thick arrows) and capillaries (thin arrow). Scale bar: 100 µm; (**K**): Section in Masson's trichrome stains showing the FBR in the hypodermis after 90 days PI; note more cells (arrow). Scale bar: 1 mm; (**L**): Section in HE showing MNGC (thin arrow) and LC (thick arrow) in the inflammatory reaction 90 days PI. Scale bar: 50 µm.

**Figure 4.** Histological sections of chronic inflammatory response induced by PLGA membranes. (**A**): Section showing the presence of LCs surrounding the chronic inflammation, 30 days PI. Scale bar: 50 µm; (**B**): Graphical representation of the number of LCs per histological field over time, showing decreased numbers up to 90 days PI. Number of MNGCs were calculated on 20 photomicrographs randomly obtained from HE stained sections (objective lens 20×). Results were expressed as number of MNGCs/histological field ± standard error of the mean; (**C**): Section showing the presence of capillaries (black arrows) along the evolution of the implanted PLGA (asterisk). Scale bar: 100 µm; (**D**): Graphical representation of the number of capillaries over time showing decreased number of capillaries up to 90 PI. Number of capillaries were calculated on 20 photomicrographs randomly obtained from HE stained sections (objective lens 20×). 3.2.2. Immunohistochemistry and Morphometry **Figure 4.** Histological sections of chronic inflammatory response induced by PLGA membranes. (**A**): Section showing the presence of LCs surrounding the chronic inflammation, 30 days PI. Scale bar: 50 µm; (**B**): Graphical representation of the number of LCs per histological field over time, showing decreased numbers up to 90 days PI. Number of MNGCs were calculated on 20 photomicrographs randomly obtained from HE stained sections (objective lens 20×). Results were expressed as number of MNGCs/histological field ± standard error of the mean; (**C**): Section showing the presence of capillaries (black arrows) along the evolution of the implanted PLGA (asterisk). Scale bar: 100 µm; (**D**): Graphical representation of the number of capillaries over time showing decreased number of capillaries up to 90 PI. Number of capillaries were calculated on 20 photomicrographs randomly obtained from HE stained sections (objective lens 20×).

The inflammatory process was analyzed over time after PLGA membranes implantation. For this goal, the CD3 antigen was used to measure the total number of T cells, and the myeloperoxidase (MPO) reactivity for polymorphonuclear neutrophils (PMNs). Both immunohistochemistry assays quantified these cellular populations, which decreased up to 90 days, showing that the inflammatory process decreased along PLGA

In the analysis, the total number of T cells (Figure 5A,C,E,G) showed variations over time. The number of T cells increased after 7 and 30 days PI and decreased after 15 and 90 days PI (*p* = 0.0002) (Figure 5A,C,E,G,I), while neutrophils (myeloperoxidase+ cells) levels were significantly elevated at the implantation bed after 15 days PI (*p* < 0.05), decreasing after 90 days PI (*p* = 0.004) (Figure 5B,D,F,H,J). Results are expressed as the number of CD3 or myeloperoxidase<sup>+</sup> cells/histological field ± SEM (standard error of the mean).

fragments degradation.

*Polymers* **2022**, *14*, x FOR PEER REVIEW 15 of 25

**Figure 5.** Immunohistochemistry and quantification of CD3 and myeloperoxidase positive cells. Detail of histological sections of PLGA-induced inflammatory response containing CD3<sup>+</sup> cells. (**A**): 7 days PI (thin arrows); (**C**): 15 days PI (thin arrows); (**E**): 30 days PI (thin arrows); (**G**): 90 days PI (thin arrows). Detail of histological sections of PLGA-induced inflammatory response containing myeloperoxidase<sup>+</sup> cells: (**B**): 7 days PI (thick arrows); (**D**): 15 days PI (thick arrows); (**F**): 30 days PI (thick arrows); (**H**): 90 days PI (thick arrows). In (**G**), inset represents the positive control of the anti-CD3 antibody, depicting CD3<sup>+</sup> cells in human tonsil. Scale bars (**A**–**H**): 50 μm. (**I**): Graphical representation of the CD3<sup>+</sup> , showing decreased numbers at 15 and 90 days PI. (**J**): Graphical representation of myeloperoxidase<sup>+</sup> cells along time showing increased number of myeloperoxidase+ cells a 15 days PI, decreasing 90 days PI. \* *p <* 0.05, \*\* *p <* 0.01, \*\*\* *p <* 0.001, ns, indicates significant differences. 3.2.3. Transmission Electron Microscopy (TEM) **Figure 5.** Immunohistochemistry and quantification of CD3 and myeloperoxidase positive cells. Detail of histological sections of PLGA-induced inflammatory response containing CD3<sup>+</sup> cells. (**A**): 7 days PI (thin arrows); (**C**): 15 days PI (thin arrows); (**E**): 30 days PI (thin arrows); (**G**): 90 days PI (thin arrows). Detail of histological sections of PLGA-induced inflammatory response containing myeloperoxidase<sup>+</sup> cells: (**B**): 7 days PI (thick arrows); (**D**): 15 days PI (thick arrows); (**F**): 30 days PI (thick arrows); (**H**): 90 days PI (thick arrows). In (**G**), inset represents the positive control of the anti-CD3 antibody, depicting CD3<sup>+</sup> cells in human tonsil. Scale bars (**A**–**H**): 50 µm. (**I**): Graphical representation of the CD3<sup>+</sup> , showing decreased numbers at 15 and 90 days PI. (**J**): Graphical representation of myeloperoxidase<sup>+</sup> cells along time showing increased number of myeloperoxidase+ cells a 15 days PI, decreasing 90 days PI. \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, ns, indicates significant differences.

creased fragments of PLGA between the cells (Figure 6H, arrows).

**Figure 6.** Transmission electron microscopy (TEM) and histological images of the inflammatory response induced by PLGA membranes. (**A**): Image showing the periphery of the PLGA with mononuclear cells forming a FBR in contact with the extracellular matrix (ECM). Scale bar: 5 μm; (**B**): image showing PLGA fragments (arrows) in contact with a MNGC. Scale bar: 2 μm; (**C**): image showing PLGA fragments (arrows) in contact with EP cells taking different morphologies, exhibiting prominent cytoplasmic extensions such as filopodia following the topography of PLGA fragments (arrows). Scale bar: 2 μm; (**D**,**E**): 15 days PI. D: PLGA fragments inside phagosomes **Figure 6.** Transmission electron microscopy (TEM) and histological images of the inflammatory response induced by PLGA membranes. (**A**): Image showing the periphery of the PLGA with mononuclear cells forming a FBR in contact with the extracellular matrix (ECM). Scale bar: 5 µm; (**B**): image showing PLGA fragments (arrows) in contact with a MNGC. Scale bar: 2 µm; (**C**): image showing PLGA fragments (arrows) in contact with EP cells taking different morphologies, exhibiting prominent cytoplasmic extensions such as filopodia following the topography of PLGA fragments (arrows). Scale bar: 2 µm; (**D**,**E**): 15 days PI. D: PLGA fragments inside phagosomes (arrows in **D**) in the interior of the cell (asterisk in **D**) (Scale bar: 1 µm) or in the extracellular space (arrow in **E**). Scale bar: 2 µm; (**F**): Light microscopy image 30 days PI in Masson's trichrome staining showing LC containing material inside (arrows). Scale bar: 100 µm; (**G**): 30 days PI—PLGA fragments were swollen by cells (arrows). Scale bar 1 µm. (**H**): Note less amount of PLGA fragments between the cells 90 days PI (arrows). Scale bar: 2 µm.

In this work we chose a high-resolution imaging technique of TEM to show detailed images of PLGA fragments in intimate contact with the cells over time. TEM images showed various populations of macrophages, epithelioid cells and multinucleated giant cells at the granulomatous inflammatory process induced by the PLGA membranes between 7 and 90 days PI (Figure 6A–H). The inflammatory process was observed in the periphery of PLGA, showing mononuclear cells forming a foreign body reaction (FBR) in contact with the extracellular matrix (ECM) 7 days PI (Figure 6A). In the periphery of PLGA, TEM image is showing PLGA fragments (Figure 6B, arrows) in contact with a multinucleate giant cell (Figure 6B(GC)) 7 days PI. Another TEM image shows PLGA fragments in contact with epithelioid cells (EP) (Figure 6C(EP)), presenting different morphologies, with prominent cytoplasmic extensions such as filopodia following the topography of PLGA fragments (Figure 6C, arrows) 7 days PI. PLGA fragments were found inside phagosomes (Figure 6D arrow), in the interior of the cell (Figure 6D, asterisk) or in the extracellular space (Figure 6E, arrow) 15 days PI. However, an interesting and rare light microscopy image shows LCs containing material inside (Figure 6F, arrows) 30 days PI, such as depicted by TEM images showing PLGA fragments in the interior of cells (Figure 6G, arrows) 30 days PI. Finally, TEM image 90 days PI show de-

### **4. Discussion**

Poly (lactic-co-glycolic acid) (PLGA) is a synthetically manufactured linear copolymer constituted by different proportions of monomeric lactic acid (LA) and glycolic acid (GA). It is used in bioengineering as scaffolds for corneal and skin tissue engineering [30], as well as for bone regeneration [31–33] or for drug delivery [33,34]. PLGA presents a good biocompatibility, excellent mechanical properties, controllable biodegradability and electrospinnability. Herein, we are describing cell-interactions with electrospun PLGA (85:15) membranes in vitro (MDCK cell line or primary culture of hamster oral fibroblasts) or in vivo when implanted in the immunocompetent hamster subcutaneous tissue.

Recently, our group analyzed the physico-chemical properties of the electrospun PLGA membranes used in the present study [15]. The membranes were mostly constituted of fibers less than 1 µm in diameter [15]. In fact, biodegradable electrospun membranes constituted by thin fibers have been shown to promote cell adhesion, migration and proliferation [11,29,35]. In this sense, our team arouses the hypotheses that the fibronectincoated electrospun PLGA membranes used in the present study should contribute to cell proliferation due to its biomimetic properties, since it has been shown in a previous study that gingival fibroblasts can adhere and proliferate when seeded onto collagen or PLGA substrates [36].

To achieve better results, PLGA hybrid membranes containing collagen or fibronectin have been manufactured [37–40]. Indeed, human dermal fibroblasts cultured onto collagencoated electrospun PLGA substrates secreted extracellular matrix (ECM) on the substrate forming a fibrillar network in the PLGA interfiber spaces, promoting better cell–material interactions [38]. Our data showed similarities with the study of Sadeghi et al., [38] regarding collagen I-coated PLGA membranes, which promoted fibroblast-like cell proliferation over time and secreted ECM onto the surface of the scaffold. Inanç et al. [41] also showed that periodontal ligament cells seeded onto PLGA membranes exhibiting the same proportions of monomers as herein, produced collagen I and fibronectin which covered the PLGA fibers [40]. Considering the average length of a fibroblast, Lowery et al., [42] demonstrated that membranes possessing interfibrillar spaces above 6 µm and less than 20 µm are permissive to the fibroblast adhesion, locomotion and proliferation inside the scaffold, while this is not the case when the interfibrillar spaces are higher than 20 µm. Although in the present study the interfibrillar spaces were not measured, it can be hypothesized that the net of collagen I formed in the interfibrillar spaces after membrane bioactivation, can favor fibroblast-like cells proliferation over time.

The MDCK cells are a model of canine kidney cell line used in drug screening and biomedical research [43,44]. Moreover, these cells have been widely used in a variety of research studies using synthetic polymers, regardless of polymers fabrication or formats, due to their ability to follow different topographies. In this sense, we appreciated the migration behavior of MDCK cells onto 3D electrospun PLGA scaffolds over time. Indeed, since the 3D electrospun PLGA scaffold used herein possesses randomly oriented fiber topography, conferring interconnected fiber-forming porous, it was considered an ideal architecture to appreciate the monolayer and epithelial barrier formation of these cells. Our in vitro data confirmed that MDCK cells also proliferated onto the PLGA membranes coated or not with fibronectin after three and twelve days, independently. Comparing the two conditions, cell proliferation showed significant better results on the fibronectin-coated PLGA membranes after 12 days. Accordingly, we showed a cluster of cells taking on a classical cuboidal morphology and forming a confluent monolayer with cell polarity after 12 days.

Our morphological study revealed that the implanted membrane promoted a highly cellular foreign body response up to 90 days PI, forming at all studied times a well-delimited structure in the site of implantation. The implantation of biomaterials in vivo frequently elicits a chronic inflammation at the tissue-implant interface followed by wound healing responses and tissue fibrosis [45–47]. Concomitant with the formation of macrophagederived foreign body multinucleated giant cells, constituting LCs or MNGCs, granulation tissue develops around the biomaterial. Factors secreted by macrophages adhering onto the surface of the biomaterial and the multinucleated giant cells formations attract fibroblasts and transform them into myofibroblasts [48,49]. These cells become aligned around the biomaterial and deposit great amount of collagen and other proteins creating a fibrous capsule around the implanted biomaterial. The fibrous encapsulation is considered as the end stage of the foreign body reaction and healing response to a biomaterial.

In the majority of cases, biomaterials promote the progression of such sterile inflammation to a foreign body response [47]. In the late 1980s and early 1990s, the occurrence of a typical nonspecific foreign-body reaction around internal fracture fixation implants made of pure polyglycolide or polylactide and even of glycolide-lactide copolymer was reported [50–54].

The foreign body response induced by our manufactured electrospun PLGA membranes was extremely cellular with a high amount of LCs derived from fusogenic macrophages. Macrophages are extremely versatile and plastic, being capable of adhering and recognizing foreign materials. At this time, they show typically a classically activated phenotype secreting inflammatory cytokines, ROS, protons release and degradative enzymes. Besides that, they display high phagocytic capacity, being able to phagocytose particles up to a size of 5 mm. In the presence of particles larger than 5 mm, macrophages coalesce to form foreign body multinucleated giant cells [55,56]. At this time, these cells display a reduced phagocytic activity, but an enhanced degradative capacity [56] at the expense of protons secretion, enzymes and ROS [57,58]. Kim [59] reported three steps on the biological pathways of local tissue responses to biomaterials: (i) organization of the inflammatory responses; (ii) monocytes migration to implantation site which differentiate into macrophages and fibrous capsule development and (iii) rapid degradation of the polymer and enhanced formation of fibrous tissues generated in the second step. These biological steps can guide the comprehension and analysis of the immunological responses in different hosts. In this sense, macrophages orchestrate the FBR that remain at the biomaterial/tissue interface when they do not succeed in phagocytizing the material [43,60]. We showed using histological and immunohistochemistry study that the electrospun PLGA membranes induced the recruitment of cells of the monocyte/macrophage lineage with the formation of epithelioid cells, LCs, as well as transitional cells which were similar to the cells seen after the implantation of fragments of MELINEX TM plastic in the subcutaneous tissue of rats [61,62]. Our ultrastructural study using TEM showed the difficulty of identifying these cell types since the transformation of monocytes to macrophages, macrophages into epithelioid cells or MNGCs were gradual, without a clear-cut distinction between the different cell types. Foreign body giant cells were recognized by the numerous nuclei inside cytoplasm, while macrophages, epithelioid cells and even giant cells, showed in the cell surface numerous invaginations and finger-like projections and numerous micropinocytotic vesicles. In addition, some cells mainly in the later stages (after 30 days PI) presented numerous filopodia, which surrounded the PLGA, or phagosomes, some of them containing fragments of the PLGA. The presence of phagosomes containing material similar to that of PLGA fibers could be correlated with the vacuoles seen in light microscopy inside some cells, suggesting that these cells have some capacity for phagocytosis. According to our knowledge, light microscopy images showing cells internalizing PLGA fragments were not found in the literature till now.

Neutrophils are present within the first days (up to day 2) after the biomaterial implantation [25,63]. Host derived chemoattractants released from activated platelets, endothelial cells, mast cells and injured cells direct neutrophils to the site of implantation. Neutrophils trigger a phagocytic response, degranulation and secrete ROS and proteolytic enzymes. The latter can corrode material surface and the degradation remnants could also trigger waves of neutrophils arrival and prolong the inflammatory response (reviewed in [63]. This event could also be occurring in our implanted PLGA membrane since it is susceptible to degradation when implanted in the body or in vitro [15,63,64]. Indeed, neutrophils were present in the foreign body reaction during all timelines of the study. In addition, the persistence of neutrophils in the reaction is a source of chemokines and activations factors for monocytes, macrophages, immature dendritic cells and lymphocytes [65,66]. Effectively a progressive influx of lymphocytes was verified in the late stages. Furthermore, activated lymphocytes at the implantation site could be the source of IL4 and IL13, both known to favor the macrophage fusion on biomaterials [67].

FBR are present within the implantation bed of different biomaterials including collagen-based biomaterials [68–70]. As already described, macrophage fusion is dependent on the presence of the fusogenic molecules on their surface but also on the environmental signals such as the quality and quantity of the adsorbed proteins in the provisional matrix on the biomaterial, on the surface itself and on the topography of the biomaterial surface [48,71]. Few reports elucidate the direct effects of the biomaterial surface on foreign body giant cells formation [72]. It was shown that decrease in monocyte adhesion and foreign body giant cell formation occurs in hydrophilic, anionic and nonionic polyacrylamide/polyacrylic acid surfaces compared to hydrophilic and hydrophobic, cationic surfaces [73]. Besides that, smooth flat surfaces induce considerably more foreign body giant cell formation than rough surfaces [27], while other reports relate that larger PLGA microspheres of 30 mm induce more foreign body giant cell formation than smaller 6 mm microspheres [74]. The implantation of porous materials constituted by high surface-tovolume is prone to show higher ratios of macrophages and foreign body giant cells than smooth-surface implants [75].

Albeit the observed long-standing foreign body response induced by the electrospun PLGA membranes, a fibrous capsule was never seen during all time points studied. Absence of fibrous encapsulation was previously found after the implantation of non-resorbable biomaterials [70] and was also reported in the subcutaneous tissue of rats implanted with PLGA [20,76,77]. Mitragotri and Lahann [75] reported that greater numbers of macrophages and foreign body giant cells on the surface of implants develop more fibrosis and encapsulation of the biomaterials. In addition, Whitaker [78] discussed that high levels of protein adsorption leads to increased cell adhesion and therefore, increased fibrous encapsulation. Considering the aforementioned concepts regarding fibrosis reaction, we can hypothesize that the 3D electrospun porous scaffolds used herein for implantation may have not had high levels of protein adsorption, and may have had decreased cell adhesion, resulting in no fibrous encapsulation. On the other hand, Al-Maawi [79] reported that the accumulation of macrophages and foreign body giant cells on the surface of a biomaterial is correlated with the formation of a highly vascularized granulation tissue and neovessels in the connective tissues. Moreover, Madden [80] reported that, nonporous scaffolds or those with 20 µm pores led to a significant increase in the fibrous capsule thickness, and pores of approximately 30–40 µm reduced the fibrous capsule thickness and the number of M1 macrophages. Several morphological variants of foreign body giant cells are already described but only the LCs and MNGCs participate on the foreign body response to biomaterials. The induction of granulomas around glass implanted in the dorsal subcutaneous tissue of rats allowed to demonstrate that LCs are precursors of MNGCs, and that the epithelioid cells derived from mature macrophages [62]. In view of these arguments, our study suggests that the tissue response induced by the electrospun PLGA membranes could reflect the constant arrival of monocytes to the implantation bed with LCs formation associated to the persistence of fusogenic molecules originated from activated macrophages without fibrous capsule formation.

In the recent years, researchers have become more focused on designing surfacemodified PLGA membranes to minimize immune responses. Research studies from Kim et al. [20], Huang et al. [81] and Lee et al. [82], who have implanted PLGA 75:25; 50:50; 75:25, in rats, reported foreign body giant cells formation in intimate contact with PLGA, and decreased inflammatory cells in PLGA-coated membranes in comparison with PLGA alone. In the present study, foreign body giant cells were formed among the inflammatory cells over time. We hypothesized that the degradation products of PLGA 85:15 produced herein induced an inflammatory response in the site of implantation.

While it is difficult to determine the specific role of any individual factor, it is important to highlight that surface topography, surface chemistry and surface energy may influence protein adsorption, platelet activation, cell growth and biocompatibility [83]. In particular, Adabi et al. reported that for some applications, hydrophobic surface could be preferred, since hydrophobic patterns showed more adsorption on intestinal mucus [83]. Since, our future challenges for the use of this membrane as a dressing material aim to accelerate the healing process of oral ulcerations in hamsters' model, we chose a hydrophobic membrane for better adherence on the oral mucus of the animals. Another important issue refers to biocompatibility. Biocompatibility is not only polymer's intrinsic property-dependent, but also biological environment-dependent [4]. For this reason, the intensity and length of specific polymer-tissue interactions can vary greatly in different organs, tissues and

species [4,19]. The device used herein was implanted in hamsters to appreciate the immune responses due to our future objective of working with this animal model to develop new future dressing for chemotherapy-induced oral mucositis ulcerations. Indeed, hamster models have been extensively studied to observe pathological effects of radiation exposure or chemotherapy induced oral mucositis and help in the development of effective treatments [84–87]. Furthermore, Syrian golden hamsters seem to be an ideal animal model due to their low cost, small size, easiness to handle and ability to accurately reflect disease progression in humans [88]. Therefore, we used hamsters to verify the biocompatibility of the material. However, there remains a lack of available reagents for studying hamster immune responses which remains an issue to better characterize the type of immune response generated, critical for understanding protection from disease.

It is worthy to mention that, regarding our future challenges of applying the 3D electrospun PLGA (85:15) membrane with the addition of primary fibroblasts as a cell delivery dressing, we performed in vivo tests using a device with a higher surface area to volume which may lead to a higher degradation of the matrix [19].

Moreover, it is also important to mention that the accumulation of PLGA degradative products could cause significant host inflammatory response, a microenvironment favoring tissue fibrosis that is mainly mediated by M1 subtype macrophage [82]. However, considering our in vivo experiments in the present study, the scaffolds became well vascularized and there were no evidence of necrosis or encapsulation, which would have been strong contraindications for future clinical applications [72].

### **5. Conclusions**

The present study appreciated cellular and tissue responses regarding PLGA (85:15) membranes produced by the 3D electrospinning method, for future applications in regenerative medicine. For this goal, both in vitro experiments were performed with primary fibroblast-like cells or MDCK cells that showed proliferation over PLGA electrospun membranes along time. In the in vivo experiment, cell-free irradiated PLGA membranes resulted in a chronic granulomatous inflammatory response in all time points after implantation, mainly constituted of epithelioid cells and LCs. Lymphocytes, myeloperoxidase**<sup>+</sup>** cells, LCs and capillaries decreased after 90 days post implantation. Light microscopy revealed foreign body giant cells showing internalized materials, without fibrous tissue formation. TEM analysis also showed cells exhibiting internalized PLGA fragments decreasing over time. Accordingly, we can conclude that MNGCs constituted a granulomatous reaction around the polymer, which resolved over time, probably preventing a fibrous capsule formation. We expect that the results of this work will boost other studies for translational applications of electrospun fiber membranes as cutaneous or oral dressings.

**Author Contributions:** Investigation and writing—original draft preparation, A.C.; Investigation and writing—original draft preparation C.M.T.; writing-review and editing M.L.D.; formal analysis, R.P.G.; investigation, T.P.; formal analysis, J.C.; Conceptualization and writing-review and editing, M.F.; writing-review and editing, L.R.d.A.; Conceptualization and writing-review and editing K.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors thank the Brazilian Agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; Grant No.310917/2014-0, Marcos Lopes Dias.; Grant No.308287/2016-9, Marcos Farina.) PVE fellowship program (Grant No.#406407/2013-4, Karine Anselme.), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro, FAPERJ (Grant No. E-26/201.304/2014, Marcos Lopes Dias.; Grant No. E-26/203.028/2017, Marcos Farina.) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), for financial support. The authors also thank Centre National de la Recherche Scientifique, CNRS (PICS project Biointerfaces Grant No.#272116, Karine Anselme), for financial support.

**Institutional Review Board Statement:** The ethical committee for animal use in scientific experiments (Comissão de Ética no Uso de Animais—CEUA) at Health Science Center at Federal University of Rio de Janeiro, Brazil, registered in the National Council of Animal Experimental Control (Conselho Nacional de Controle de Experimentação Animal—CONCEA), process number 01200.001568/201387, certified the use of hamsters in this study (protocol No. 003/15, on 04/15/2015), acquired from Oswaldo Cruz Foundation (Fundação Oswaldo Cruz—Fiocruz).

**Data Availability Statement:** The raw/processed data required to reproduce these findings can be shared upon request.

**Acknowledgments:** We would like to thank the electron microscopy and confocal microscopy platforms of Institut de Science des Matériaux de Mulhouse (IS2M), Unidade de Multiusuário Padrón-Lins UNIMICRO/UFRJ for the electron microscopy facilities, the electron microscopy and confocal microscopy platforms of Centro Nacional de Biologia Estrutural e Bioimagem (CENABIO), and INCT-465656/2014-5.

**Conflicts of Interest:** The authors declare no conflict of interest.
