Use of Human Gingival Fibroblasts for Pre-Vascularization Strategies in Oral Tissue Engineering

Abstract

Background:

Cocultures of human gingival fibrobasts (hGF) and endothelial cells could enhance regeneration and repair models as well as improve vascularization limitations in tissue engineering. The aim of this study was to assess if hGF could support formation of stable vessel-like networks.

Methods:

Explant primary hGF were isolated from gum surgical wastes collected from healthy patients with no history of periodontitis. Human umbilical vein endothelial cells (HUVEC) were two-dimensional (2D) and three-dimensional (3D) cocultured in vitro with hGF at a cell ratio of 1:1 and medium of 1:1 of their respective media during at least 31 days. Vessel quantification of HUVEC networks was performed. In order to investigate the pericyte-like properties of hGF, the expression of perivascular markers α-SMA, NG2, CD146 and PDGFR-β was studied using immunocytochemistry and flow cytometry on 2D cultures.

Results:

hGF were able to support a long-lasting HUVEC network at least 31 days, even in the absence of a bioreactor with flow. As observed, HUVEC started to communicate with each other from day 7, constructing a network. Their interconnection increased significantly between day 2 and day 21 and lasted beyond the 31 days of observation. Moreover, we tried to explain the stability of the networks obtained and showed that a small population of hGF in close vicinity of HUVEC networks expressed perivascular markers.

Conclusion:

These findings highlight a new interesting property concerning hGF, accentuating their relevance in tissue engineering and periodontal regeneration. These promising results need to be confirmed using more 3D applications and in vivo testing.

Introduction

Vascularization of voluminous tissue-engineered constructs is a crucial issue for integration. Indeed, the challenge is to supply the inner cells of the transplant with nutrients and oxygen, as the slow passive diffusion may lead cells to quiescence or death [1]. These inner cells send stress signals to the host for angiogenesis to start, but vessel networks pace can take weeks before reaching them. Nowadays, one of the strategies for enhancing supply of nutrients and oxygen in oral transplant is to promote in vitro prevascularization before transplantation [2].

Along this way, the manufacturing of a high degree of vascularization implies an effective crosstalk between oral stromal and endothelial cells. Cocultures between mesenchymal stromal cells (MSC) and human umbilical vein endothelial cells (HUVEC) by direct contact showed a shift towards a higher expression of perivascular markers NG2, CD146 and PDGFR-β [3]. Orally, these communication skills were shown in periodontal ligament derived stem cells (PDLSC), that were proven to possess angiogenesis ability thanks to their paracrine function when cocultured with endothelial cells [4, 5]. Other studies even proved that PDLSC expressed pericyte markers and might ensure the functional role of perivascular cells when interacting with HUVEC [6, 7]. Thus, a strategy for gingival tissue engineering would consist in generating in vitro vascular networks by co-culturing gingival fibroblasts and endothelial cells [2]. However, these interactions have poorly been studied in the literature [8].

Human gingival fibroblasts (hGF) are the main cellular component of gingival connective tissue and renowned candidates for tissue engineering. From an anatomical point of view, these stromal cells belong to the lamina propia of the masticatory mucosa, that includes the attached, marginal and papillary mucosa as well as the hard palate [9]. hGF, among other oral cell types, have been considered as a potential cell source for tissue engineering for numerous reasons [10]. First, they can be harvested easily from gum biopsy or surgical waste and display a high proliferative potential when cultured. Indeed, these cells are easily accessible using minimally invasive surgeries without esthetic consequence [11]. Then, hGF possess scar-free healing properties that can be relevant for regenerative medicine therapies. For example, in comparison with skin repair, oral wounds have a faster and more scar-free healing with quick re-epithelialization as well as less inflammation [12]. These findings were recently confirmed by the use of hGF secretome to enhance skin wound healing, highlighting the anti-inflammatory and pro-angiogenic properties of these cells [13]. Moreover, some studies have found that a small population of hGF was multipotent and includes gingiva-derived mesenchymal stem cells (GMSCs) and gingival multipotent progenitor cells (GMPCs) [14–16]. Lastly, oral fibroblasts are generated from neural crest-derived ‘ectomesenchyme’ and not from mesoderm like the other fibroblasts in the body [17]. The embryonic origin of hGF is of tremendous importance as it could explain their specific properties, such as their remodeling capacities. In vivo evidence of the identity of the embryonic origin of the oral fibroblasts was provided in a recent study [18]. All these findings strongly support the use of gingival fibroblasts as an easily accessible source of stem-like cells and stem cells that enable optimal and fast healing.

Cocultures between hGF and vascular endothelial cells could enhance regeneration and repair models as well as improve vascularization limitations in tissue engineering [19]. We therefore hypothesize that: (1) direct coculture between hGF and HUVEC enhances HUVEC organization in networks thus increasing vascularization, (2) when cocultured with HUVEC, a part of hGF could behave like pericytes and stabilize vascular networks throughout time.

The aim of this study was to assess if gingival fibroblasts could support formation of stable and functional vessel-like network in tissue engineering perspectives.

Materials and methods

Primary isolations and cell cultures

Regulatory authorizations

Samples were harvested according to the French legislation i.e. under the control of the declaration for conservation and preparation of human body elements for scientific research number DC 2008-412 (French ministry of higher education, research and innovation). Protocols were approved by the institutional committee for the protection of human subjects (Local Ethic Committee from academic hospital CHU de Bordeaux). According to Helsinki declaration of 1975, mothers provided written informed consent when umbilical cords were concerned and all patients gave their informed oral consent and denied having recently taken drugs that could affect connective tissue metabolism when biological wastes were concerned. All the samples were treated anonymously.

Human gingival fibroblasts

Human gingival fibroblasts (hGF) were isolated from patients who underwent surgery at the University Medical Hospital of Bordeaux (CHU de Bordeaux). For that purpose, gum surgical wastes were collected from healthy non-smoking patients with no history of periodontitis. Samples were kept in Dulbecco Modified Eagle's minimal essential Medium (DMEM) (Gibco, Thermofisher Scientific, Waltham, MA, USA) and immediately transferred to the laboratory. The epithelium was removed using a surgical scalpel (Swann-Morton, Sheffield, UK) and the remaining connective tissue was cut into small fragments, transferred and cultivated in cell culture flasks (Sarstedt, Nümbrecht, Germany) (37 °C, 5% CO₂). The culture medium consisted in DMEM completed with 10% fetal bovine serum (FBS) (PAN-Biotech, Aidenbach, Germany), 100UI/ml penicillin, 100 μg/ml streptomycin and 2.5 μg/ml amphotericin B (Gibco). Medium was changed every 3 days and cells were sub-cultured at 70–80% confluency. After the first passage, antibiotics and amphotericin were removed from the medium. For the experiments, hGF were grown when mentioned in either DMEM or a coculture medium (CoM) made of 1:1 DMEM/ complete endothelial cell growth medium-2 (EGM-2 MV) (Lonza, Walkersville, MD, USA) and used from passage 3 to 8.

Red fluorescent protein-labeled human umbilical vein endothelial cells

Human umbilical vein endothelial cells (HUVEC) were isolated from human umbilical cord of healthy newborns, as previously described [20]. A part of HUVEC at passage 1 was transduced with a vector containing the RFP gene under the control of the EF1a promoter following a procedure similar to the one previously described [21]. HUVEC and RFP-HUVEC were cultured in EGM-2 MV. Cells from passage 3 to 9 were used for the experiments.

Control cell types

Human brain vascular pericytes

Human brain vascular pericytes (hBVP) (ScienCell, Carlsbad, CA, USA) were purchased. Thawing and seeding protocol from manufacturer were strictly followed to ensure the phenotypic characteristics of pericytes. After the first passage, cells were cultured in CoM in order to keep the same medium for each condition and provide comparison within the experiments. Medium was changed every 3 days and cells were sub-cultured at 70–80% confluency. Cells were used from passage 3–5.

Stem cells from the apical papilla

Stem cells from the apical papilla (SCAP) were isolated from third molars germs obtained from young patients, as described previously [22]. The culture medium was composed of Minimum Essential Medium alpha (α-MEM) (Gibco) supplemented with 10% FBS (PAN-Biotech). Cells were used from passage 4 to 6.

Human dermal fibroblasts

Human dermal fibroblasts (hDF) were isolated from adult human skin as described previously [23]. The culture medium was DMEM/F12 (Gibco) supplemented with 10% FBS (PAN-Biotech). Cells were used from passage 4–6.

Two-dimension cocultures

In order to study modification in cells properties when in direct contact with each other, it was decided to use a 1:1 cell ratio for hGF/HUVEC coculture, as well as 1:1 hGF medium/HUVEC medium (CoM) as these are the most used coculture ratios in literature [2]. Cocultures were seeded either on culture flasks or on glass slides and cultured for at least seven days (10,000 cells in total/cm²). Medium was changed every three days.

Three-dimensional cocultures

To obtain 3D cocultures, cells were embedded either in a composite hydrogel made of 2 mg/ml of type I collagen (Corning Incorporated, Tewksbury, MA, USA) at a concentration of 500,000 cells/ml or in a custom-made composite hydrogel of methacrylated collagen and hyaluronic acid synthetized by the ART BioPrint (INSERM, Bordeaux, France) at a concentration of 20.10⁶ cells/ml. This hydrogel was originally a biomaterial ink formulated to be used for extrusion-based bioprinting. 3D cocultures were cultivated in the CoM medium that was changed every three days. Network formation was observed under fluorescent binoculars (Leica Microsystems, Wetzlar, Germany) using the software LAS X (Leica Microsystems).

Vessel formation and quantification

An “Endothelial Tube Formation Assay” (ETFA) was performed in order to study the HUVEC branched structures when cocultured with hGF (and other cells when specified). RFP-HUVEC/hGF were prepared as described in paragraph 2.2. The “Angiogenesis Analyzer” tool for Image J software (National Institutes of Health, Bethesda, MD, USA) developed by Carpentier et al. was used and enabled the analysis of ETFA-derived images, providing different characteristics of the obtained networks [24]. For quantitative analyses, at least three random pictures were studied at each timepoint (Day 2–7–14–21–31).

Immunocytochemical staining of 2D cocultures

Double staining was performed on 2D cocultures. HUVEC were stained using von Willebrand Factor (vWF). Pericytes and potential peri-vascular cells obtained from hGF were stained using various markers (for complete markers names, see Table 1): α-SMA, NG2, CD 146 and PDGFR-β. The complete list of antibodies used is provided in Table 1. Briefly, cocultures were fixed using 100% methanol during 10 min at − 20 °C and washed twice with PBS 1X. After saturation using PBS 1X/Bovine Serum Albumin (BSA) 1% during 30 min at room temperature (RT), samples were incubated with primary antibodies (Table 1) for 2 h at RT. After 3 washes in PBS, secondary antibodies linked to fluorochromes (Table 1) were incubated with the samples for 1 h at RT. Finally, after 3 washes and 10 min staining with DAPI (Thermofisher Scientific), samples were analyzed with a confocal microscope (SPE microscope, Leica Microsystems).

Table 1.

List of antibodies used in the study

Antibody	Antigen Targeted	Supplier	Host isotype	Clone reference	Concentration (Stock solution)	Dilution
Antibodies for Immunofluorescent staining
vWF	von Willebrand Factor	Dako	Rabbit polyclonal	A0082	3.1 g/l	1/300
α-SMA	Alpha- Smooth Muscle Actin	Sigma-Aldrich	Mouse monoclonal	1A4	-	1/300
NG2	Neuron/Glial antigen 2 MCSP: Melanoma-associated chondroitin sulfate proteoglycan	Santa Cruz	Mouse monoclonal	LHM 2	200 μg/mL	1/300
CD146	MCAM: Melanoma cell adhesion molecule	Santa Cruz	Mouse monoclonal	P1H12	200 μg/mL	1/300
PDGFR-β	Platelet Derived Growth Factor Receptor beta	Invitrogen	Mouse monoclonal	18A2	1 mg/ml	1/100
Alexafluor 488	Mouse IgG (H + L)	Invitrogen	Goat polyclonal	A11001	2 μg/μL	1/500
Alexafluor 568	Rabbit IgG (H + L)	Invitrogen	Goat polyclonal	A11036	2 μg/μL	1/500
Antibodies for flow cytometry
Isotype PE	None	eBioscience	Mouse IgG1, κ	17–4714-42	0.1 μg/μL	1/100
CD31 PE	CD31 antigen	eBioscience	Mouse monoclonal	WM59	5 μL/Test	1/100
PDGFR-β PE	Platelet derived growth factor receptor beta	Biolegend	Mouse IgG1, κ	18A2 323605	400 μg/mL	1/100
NG2 PE	MCSP	Santa Cruz	Mouse monoclonal	LHM 2	200 μg/mL	1/100
MCAM PE	CD146	eBioscience	Mouse	P1H12	200 μg/mL	1/100

Cell sorting

Cell sorting was performed using anti-cluster of differenciation 31 (CD-31) antibody-coated magnetic beads (Dynabeads CD 31 endothelial cell, Invitrogen, Thermofisher Scientific) in order to remove HUVEC from the 2D coculture. The beads were used respecting the manufacturer’s protocol. First, 25 μl of beads were washed and resuspended using the same volume of Buffer 1 (PBS 1X with 0.1% BSA, pH7.4). Then, after coculture dissociation (using PBS with 0.25% trypsin and 1 mM EDTA solution) and centrifugation, cells were resuspended in 1 ml of Buffer 1. Cells and beads were mixed for 30 min at 4 °C under constant shaking, and finally split using magnet DynaMag2 (Invitrogen). Supernatant (containing hGF) was transferred into a new tube while beads were discarded.

Flow cytometry

Flow cytometry was performed to study cell surface markers. Briefly, cells were detached (PBS with 0.25% trypsin and 1 mM EDTA), counted and fixed in 1% paraformaldehyde in PBS for 20 min at 4 °C. Cells (500,000 cells per sample) were washed twice with PBS and incubated with PBS containing 1% BSA for 30 min at room temperature. Cells were stained with each antibody for an hour at RT (for antibodies, see Table 1). Flow cytometry was performed on a fluorescence-activated cell sorter (BD Accuri C6, Accuri Cytometers, Ann Arbor, MI, USA), gating and selecting 10,000 events per sample. Each experiment was performed at least three times. Data was analyzed using Cflow software (Accuri Cytometers).

Statistical analyses

Data were expressed as mean ± standard deviation (SD). For comparison of experimental groups, analysis of variance (ANOVA) was performed using XLstat® software (Addinsoft, Paris, France). A p value < 0.05 was considered to be statistically significant.

Results

Morphology and phenotype of primary human gingival fibroblasts

Primary hGF used in this study displayed a classic spindle-shaped morphology when observed with phase contrast microscopy as shown on Fig. 1A. The use of coculture medium increased cell proliferation and appeared to reduce the size of cells, explaining the higher confluence among hGF cultured in coculture medium (Fig. 1A, B). hGF cell surface antigen expression was investigated using flow cytometry. Cells were positive for mesenchymal stromal cell markers (CD 90, CD 73, CD 105) and remained negative for hematopoietic marker CD 34 and endothelial marker CD 31 (Fig. 1C) whatever the medium used.

Fig. 1.

Characteristics of primary hGF used in this study. A Phase contrast microscopy at passage 6, day 5. Representative pictures from more than three independent experiments are shown. Scale bars: 100 μm (×10 magnification). B Growth curve. C Expression of cell surface markers (using flow cytometry)

Two-dimension coculture with endothelial cells and quantification of obtained pseudo-vessel networks

2D coculture between hGF and RFP-HUVEC was studied during 31 days. RFP-HUVEC enabled to visualize their organization throughout time using fluorescent microscopy (Fig. 2A). As observed, endothelial cells started to communicate with each other from day 7, constructing a network (Fig. 2A). Their interconnection increased significantly between day 2 and day 21 and lasted beyond the 31 days of observation (Fig. 2B). Network characteristics were studied using the vessel quantification tool “Angiogenesis Analyzer” for ImageJ (Fig. 2B).

Fig. 2.

Two-dimension in vitro coculture of hGF and RFP-HUVEC. A Fluorescent microscopy. Representative pictures from more than three independent experiments are shown. Scale bars: 100 μm (×10 magnification). B Vessel quantification: total branching length, number of junctions and number of meshes using the “Angiogenesis Analyzer” tool for ImageJ (N = 3). * p < 0.05 was considered statistically significant

While the total branching length remained statistically stable at day 31 we noticed a slight thinning of the branched network in comparison to previous time points. This could be explained by either the lack of perfusion which is necessary to remodel vessel structures or by the progressive loss of the RFP-fluorescent signal induced by long-term culture as observed in the picture. However, Day 31 displayed a better interconnection of the network.

A potential cause of this long-term organization and stability of HUVEC networks was investigated by trying to find the presence of perivascular-like cells among fibroblasts.

Enhanced expression of peri-vascular markers among gingival fibroblasts when cocultured with endothelial cells

Immunofluorescent staining protocols were performed on 14 day-2D cocultures in order to identify possible perivascular cells. HUVEC were stained using vWF and potential perivascular cells using CD146, α-SMA, NG2 and PDGFR-β markers.

As shown in the Fig. 3, these pericyte markers were particularly expressed by some hGF standing next to HUVEC nodes and junctions. Moreover, while CD146 was expressed by HUVEC and co-localized with vWF staining, some other cells close to HUVEC also expressed CD146 as shown with white arrows (Fig. 3A). α-SMA was detected mainly among hGF which were close to the HUVEC network (Fig. 3B). PDGFR-β and vWF also colocalized, suggesting the presence of PDGF receptors on HUVEC (Fig. 3C).

Fig. 3.

Expression of perivascular markers by immunocytochemistry in a 14 day-2D coculture. Representative pictures from more than three independent experiments are shown. Scale bars: 25 μm (× 40 magnification). A CD146. White arrows show CD146 + cells that are not HUVEC. B α-SMA. C PDGFR-β. D NG2

After observing the expression of perivascular markers by a small population of fibroblasts, the aim was to understand whether perivascular properties were displayed before the coculture with HUVEC or enhanced by the coculture.

Expression of perivascular markers in gingival fibroblasts

In order to investigate the pericyte-like properties of hGF, the expression of pericyte markers NG2, CD146 and PDGFR-β was studied using flow cytometry. Three distinct conditions were tested as summarized in the Fig. 4. As expected, the positive control, pericytes (hBVP), exhibited a higher expression of all pericyte markers (Fig. 5). Interestingly, hGF which were separated from HUVEC expressed significantly more positive cells to CD146 and PDGFR-β than hGF in monoculture. This result suggested the need for a crosstalk between cells to induce a pericyte-like phenotype among some fibroblasts (Fig. 5). When comparing the groups hGF monoculture and hGF separated, results were statistically significant for CD 146 and PDGFR-β (p < 0.01) but not for NG2 (p = 0.3) (Fig. 5).

Fig. 4.

Explanation of the different cell conditions studied by flow cytometry. Pericyte marker expression in hBVP (positive control), hGF (negative control) and hGF separated from HUVEC after at least 10 days-2D coculture, in coculture medium. Results of flow cytometry are presented in Fig. 5

Fig. 5.

Flow cytometry of pericyte markers. Conditions are described in Fig. 4. Statistical analyses were made using ANOVA. *p < 0.05 was considered to be statistically significant

Endothelial cell organization in 3D coculture with gingival fibroblasts

This long-lasting prevascularization network was intended for a tissue engineering application in 3D. First, a comparison was made between hGF, hDF and SCAP in 3D cocultures in a collagen hydrogel (Fig. 6A). It showed that among the three cell types, HUVEC were able to organize better when cocultured with hGF. Indeed, networks were denser and more interconnected (Fig. 6A, B). Then, similarly to the 2D coculture, a 3D coculture system hGF/RFP-HUVEC was then performed by embedding the cells in a composite hydrogel made of collagen and hyaluronic acid, which is a material ink that was made to be used as a bioink for extrusion-based 3D-bioprinting. Fluorescent microscopy observation was performed to follow the network formation of RFP-HUVEC (Fig. 6C). As previously observed, HUVEC organized themselves into 3D networks that remained for at least 35 days in vitro and displayed even denser networks when compared to 2D cocultures (Fig. 6C).

Fig. 6.

Three dimension in vitro cocultures with HUVEC. A Evolution of HUVEC network when 3D cocultured with hGF, hDF and SCAP. Representative pictures from more than three independent experiments are shown. Scale bars: 200 μm (×8 magnification). B Total branching lengths (per field) of HUVEC networks at day 2, 7 and 14. Statistical analyses using ANOVA. *p < 0.05 was considered to be statistically significant. C HUVEC and hGF in vitro 3D coculture using a tissue engineering approach (coculture embedded in a hydrogel usable for 3D extrusion-based bioprinting). Representative pictures from more than three independent experiments are shown. Scale bars: 200 μm (×8 magnification)

Discussion

This study proved the relevance of hGF in coculture with endothelial cells to improve vascularization in tissue engineering. hGF could replace other cell types that are mainly used in coculture techniques such as human dermal fibroblasts (hDF) and mesenchymal stromal/stem cells from oral or non-oral origin [25, 26]. Indeed, fibroblasts play an essential role in the angiogenic process through their production of extracellular matrix molecules [19]. In addition to easy and uneventful harvesting, the use of gingival fibroblasts instead of hDF could be justified by a difference between skin and oral mucosa in terms of healing [12, 13, 27].

Our results are in accordance with the sole study using hGF and HUVEC coculture for tissue engineering [8]. Cheung et al. proved that the use of the co-culture in perfused constructs made of specific polyurethane hydrogels enhanced cell growth and increased vascular endothelial growth factor (VEGF) synthesis [8]. Our study showed sustained vascular networks using natural hydrogels (collagen and hyaluronic acid) and without any perfusion. Clinical applications seem achievable as short preincubation times (from 7 to 21 days) are enough and the networks remaining at least three weeks give time to ensure connection to the recipient site.

Moreover, our data suggest that hGF were able to support a long-lasting HUVEC network (at least 31 days), and this may be thanks to a part of hGF that exhibited pericyte-like capacities, even in the absence of a bioreactor with flow. Indeed, similarly to studies with PDLSC and hDF, direct contact between hGF and HUVEC could explain the expression of perivascular markers by a small population of fibroblasts. In native tissues, pericytes are fibroblast-like cells covering between 22 and 99% of the endothelial cell surface and needed to stabilize vessel networks [28]. Various markers including alpha SMA, NG2, CD140 a, CD140b and CD146 are used together to confirm their phenotype [29, 30]. NG2 plays a role in pericyte localization next to the endothelial layer and interaction with endothelial cells [31]. CD146 is used as a marker for endothelial cell lineage and is highly expressed by cellular components of the blood vessel wall including endothelial cells, smooth muscle cells and pericytes [32]. Moreover, CD146 functions as a coreceptor of PDGF receptor-β (PDGFR- β or CD140b) to mediate pericyte recruitment to cerebrovascular endothelial cells [33]. In this study, different experiments showed that a small population of hGF in close vicinity of HUVEC networks displayed pericyte-like behavior (α-actin production and PDGFR-β expression, Fig. 3). The fibroblasts used upregulated the expression of PDGFR-β when in contact with HUVEC but not alone (even in the coculture medium) as shown by flow cytometry of hGF monocultures in the coculture medium (Fig. 5). Indeed, when comparing the groups hGF monoculture and hGF separated from coculture, results were statistically significant for CD146 and PDGFR-β (p < 0.01) but not for NG2 (p = 0.3). Two hypotheses could explain these results with NG2: (1) hGF originally express NG2 as cells derived from the neural crest do express physiologically this marker, (2) NG2 seems to be a pericyte marker exclusively associated to the arterial system, explaining its low expression around the HUVEC networks [30].

Other studies with other endothelial cell types displayed a different cell growth and gene expression when compared to HUVEC [34], microvascular cells being closer to native oral tissues than HUVEC. In order to keep an autogenous approach, an interesting source of endothelial cells could be obtained from hGF. Indeed, Liu et al. were able to induce and differentiate hGF into vascular endothelial-like cells using VEGF₁₆₅ [35].

Questions remain about the different fibroblast subpopulations residing in the oral mucosa and their possibly various properties [27]. The phenotype of fibroblasts depends on the tissue extraction method and culture conditions. Here, cells from explant primary culture were used and displayed classic markers in flow cytometry. One of the limitations of the study was the limited expression of perivascular markers by the positive control which were commercial human pericytes extracted from the brain (hBVP). This could be explained by the fact that the cells were grown in the coculture medium instead of their commercial medium. Moreover, characteristics depend on the type of vessel studied explaining that markers vary depending on the type of vessel [30]. Pericytes from the brain were used as control because of the same embryogenic provenance as gingival fibroblasts, the neural crest, as well as their increased pro-angiogenic activity compared to other pericytes [36]. Since hGF come from the neural crest, the isolation of these hGF becoming pericytes could provide useful alternatives to brain pericytes to study cell therapeutic approaches and vascularization enhancement.

To conclude, hGF are efficient in supporting the creation of prevascularized tissue-engineered constructs since HUVEC/hGF in vitro cocultures display the formation of a vascular network more rapidly than hDF and SCAPS and within a few days. Obtained networks evolved to a highly branched organization and remained stable in vitro for at least 3 weeks. This sustained network organization for at least three weeks in vitro gives new possibilities to enhance vascularization in tissue engineering applications using prevascularization concepts [2]. Our study showed that this process takes place through the appearance of an hGF subset that expressed pericytes markers. These findings highlight a new interesting property concerning human gingival fibroblasts characteristics, accentuating their relevance in tissue engineering. These promising results need to be confirmed using more 3D applications and in vivo testing.

Funding

Fondation de l'Avenir pour la Recherche Médicale Appliquée, AP-RM-20-025, Adrien Naveau.

Declarations

Conflict of interest

The authors report no conflicts of interest.

Ethical statement

Samples were harvested according to the French legislation i.e. under the control of the declaration for conservation and preparation of human body elements for scientific research number DC 2008-412 (French ministry of higher education, research and innovation). Protocols were approved by the institutional committee for the protection of human subjects (Local Ethic Committee from academic hospital CHU de Bordeaux).