Pre-Seeding of Simple Electrospun Scaffolds with a Combination of Endothelial Cells and Fibroblasts Strongly Promotes Angiogenesis

Abstract

Background:

Introduction of pro-angiogenic cells into tissue-engineered (TE) constructs (prevascularisation) is a promising approach to overcome delayed neovascularisation of such constructs post-implantation. Accordingly, in this study, we examined the contribution of human dermal microvascular endothelial cells (HDMECs) and human dermal fibroblasts (HDFs) alone and in combination on the formation of new blood vessels in ex-ovo chick chorioallantoic membrane (CAM) assay.

Methods:

Poly-3-hydroxybutyrate-co-3-hydroxyvalerate (PHBV) and polycaprolactone (PCL) were first examined in terms of their physical, mechanical, and biological performances. The effect of gelatin coating and co-culture conditions on enhancing endothelial cell viability and growth was then investigated. Finally, the angiogenic potential of HDMECs and HDFs were assessed macroscopically and histologically after seeding on simple electrospun PHBV scaffolds either in isolation or in indirect co-culture using an ex-ovo CAM assay.

Results:

The results demonstrated that PHBV was slightly more favourable than PCL for HDMECs in terms of cell metabolic activity. The gelatin coating of PHBV scaffolds and co-culture of HDMECs with HDFs both showed a positive impact on HDMECs viability and growth. Both cell types induced angiogenesis over 7 days in the CAM assay either in isolation or in co-culture. The introduction of HDMECs to the scaffolds resulted in the production of more blood vessels in the area of implantation than the introduction of HDFs, but the co-culture of HDMECs and HDFs gave the most significant angiogenic activity.

Conclusion:

Our findings showed that the in vitro prevascularisation of TE constructs with HDMECs and HDFs alone or in co-culture promotes angiogenesis in implantable TE constructs.

Introduction

Tissue and organ failure or losses are major problems that are seen in human health, and tissue engineering offers an opportunity to develop functional substitutes for damaged tissues. Most of the conventional tissue engineering approaches have mainly focused on attachment and proliferation of cells, and their formation of extracellular matrix (ECM) prior to implantation [1]. However, the lack of blood vessels in tissue-engineered (TE) constructs is one of the most critical challenges in the survival of engineered tissue substitutes [2, 3]. Although significant progress has been made over the last decades, the main problem with tissue engineering constructs still remains the same, and it is the slow formation of new blood vessels, also known as neovascularisation, post-implantation. When a TE substitute is implanted, nutrients and oxygen must be provided to enable the cells to survive in vivo, and the formation of a vascular network within TE substitutes can take weeks, which leads to the failure of the constructs [4]. For the diffusion of oxygen and nutrients from adjacent blood vessels to the new tissues, any distance more than 200 µm is viewed as very challenging [5].

The majority of the current strategies to circumvent delayed neovascularisation focus on the addition of pro-angiogenic factors to TE constructs (functionalisation strategy) [1], but the use of laboratory expanded pro-angiogenic cells such as endothelial cells (ECs), endothelial progenitor cells (EPCs), and stem cells to the tissue engineering systems (prevascularisation strategy) prior to implantation is also a well-established approach [6, 7]. The successful prevascularisation of a TE construct depends on three primary parameters: (1) cell type (single or co-culture of vessel-forming cells), (2) scaffold material selection (synthetic, natural or composite materials), and (3) culture conditions [8, 9].

Accordingly, in this study, we compared physical, mechanical, and biological performances of two widely-used polymers, Poly-3-hydroxybutyrate-co-3-hydroxyvalerate (PHBV) and polycaprolactone (PCL). PHBV is a biocompatible and biodegradable polyester, belongs to the polyhydroxyalkanoate family. One of its biodegradation products, 3-hydroxybutanoic acid, is also a natural product produced in the human body [10] which can be linked with its high biocompatibility. PCL is another biocompatible and bioresorbable synthetic polymer [11, 12], which does not produce an overly acidic environment in the degradation process and has been approved by United States Food and Drug Administration (FDA) as a biomaterial for the fabrication of several biomedical devices [13, 14]. Following the comparison of both polymers, we evaluated the effect of gelatin coating and the presence of human dermal fibroblasts (HDFs), as helper cells, on human dermal microvascular endothelial cell (HDMEC) growth and survival on simple electrospun PHBV scaffolds. Finally, we investigated the angiogenic activity of the PHBV scaffolds when cellularised with HDFs in isolation, HDMECs in isolation, and HDMECs in indirect co-culture with HDFs using an ex-ovo chick chorioallantoic membrane (CAM) assay.

Materials and methods

Materials

37% formaldehyde (FA) solution, 4′,6-diamidino-2-phenylindole (DAPI) solution, AlamarBlue^® Cell Metabolic Activity Assay, amphotericin B, dimethyl sulphoxide (DMSO), Dulbecco’s modified eagle’s medium (DMEM), fetal calf serum (FCS), hematoxylin solution, penicillin/streptomycin, phalloidin-fluorescein isothiocyanate (FITC), phalloidin-tetramethylrhodamine isothiocyanate (TRITC), polycaprolactone (PCL) (Mn: 80.000 g/mol), and trypsin EDTA were purchased from Sigma-Aldrich (St. Louis, MO, USA). Chloroform, dichloromethane (DCM), dimethylformamide (DMF), DPX mounting medium, industrial methylated spirit (IMS), methanol, and xylene were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Human dermal microvascular endothelial cells (HDMECs), EC GM MV supplement pack (For HDMECs), and endothelial cell growth medium MV (EC GM) (for HDMECs) were purchased from PromoCell (Heidelberg, Germany). CellTracker™ Green and CellTracker™ Red were purchased from ThermoFisher (San Jose, CA, USA). Poly-3-hydroxybutyrate-co-3-hydroxyvalerate (12%) (PHBV) was purchased from GoodFellow (London, United Kingdom). Optimum cutting temperature tissue freezing medium (OCT-TFM) was purchased from Leica Biosystems.

Methods

Comparison of PHBV and PCL

Comparing the mechanical and surface properties of the polymers

The polymer solutions were prepared before contact angle measurements. Firstly, PHBV (10% w/w) granules were dissolved in DCM:methanol (90:10 w/w) solvent blend in a fume hood. Then PCL (10% w/w) granules were dissolved in a chloroform:DMF (70:30 w/w) solvent blend. PHBV and PCL films from the solutions were then prepared using a spin-coater (Laurell WS-400B, North Wales, Pennsylvania, USA). Contact angle measurements of the PHBV and PCL were undertaken using a drop shape analyser (Krüss DSA100, Hamburg, Germany) under ambient laboratory conditions. In brief, a 5 µl droplet was dropped onto the polymer films, and the contact angles of the droplet on scaffolds were recorded from the software. The measurements of the three drops on three different substrates were measured for each polymer. Characterisation of PHBV and PCL electrospun scaffolds in terms of biomechanical and microstructural properties.

The DCM:methanol (90:10 w/w) solvent blend for electrospinning PHBV has previously been found to be the best composition in our lab [15, 16]. The solvent mixture for electrospinning PCL has previously been optimised elsewhere and a chloroform:DMF (70:30 w/w) blend was found to be best for electrospinning PCL nanofibres [17]. Thus, PHBV and PCL solutions were prepared using the solvent blends given above, and the polymer solutions were loaded into 5 ml syringes fitted with 0.6 mm inner diameter syringe tips. Syringes were then placed in a syringe pump (GenieTMPlus, KentScientific, Torrington, CT, USA). Aluminium foil was used as the collector and placed at a distance of 17 cm from the needle tips. The pump was set to 40 µl/min, and a 17 kV voltage was applied both to the collector and the tips. Electrospinning was conducted at room temperature until all the polymer solution was used.

Biomechanical testing samples were prepared cutting 20 mm × 10 mm pieces from dry electrospun scaffolds. The clamps of the device were positioned 10 mm away from each other, and the width and thickness of each scaffold were measured. Test samples were clamped with two grips in a tensiometer (BOSE Electroforce Test Instruments, Eden Prairie, MN, USA). Tensile tests were performed on each sample at a rate of 0.1 mm/s until the samples failed (n = 4). The raw data of these tests were used for drawing stress–strain and load–displacement graphs. Ultimate tensile strength (UTS) of the scaffolds was calculated from stress and stress curves of each sample, while stiffness was calculated from the load and displacement data.

The surface morphology of PHBV and PCL electrospun scaffolds were observed under SEM (Philips/FEI XL-20 SEM; Cambridge, UK). The samples were coated with gold using a gold sputter (Edwards sputter coater S150B, Crawley, England) prior to imaging. Fibre diameter and pore sizes were measured using ImageJ software.

Comparing the biocompatibility of the polymers in vitro and in vivo

Assessment of the metabolic activity of Human Dermal Microvascular Endothelial Cells (HDMECs) growing on PHBV and PCL

AlamarBlue^® Cell Viability Assay was used to measure the metabolic activities of HDMECs cultured on either PHBV or PCL over 11 days.

HDMECs were purchased from PromoCell and used between P2 and P4. Cells were thawed and cultured until reaching 80–90% confluency. Following the disinfection of the PHBV and PCL scaffolds with 70% ethanol for 45 min, HDMECs were trypsinized and counted. 4 × 10⁴ cells were resuspended in 0.25 ml of EC GM and then seeded onto the scaffolds in 12-well plates. Before adding culture medium, scaffolds were returned to the incubator for an hour to allow HDMECs to attach. Then, 4 ml of HDMECs culture medium was added to each well, and they were incubated at 37 °C overnight. Scaffolds were kept in culture for 11 days by changing the culture medium every 2–3 days.

An AlamarBlue^® cell viability assay was performed at days 1, 4, 7 and 11. Briefly, a 0.1 mM AlamarBlue^® working solution was prepared by 10 × dilution of the 1 mM AlamarBlue^® stock solution with growth medium. Growth media were removed, and the scaffolds were washed with PBS. 1 ml of AlamarBlue^® working solution was added to each well and incubated at 37 °C for 4 h. After an incubation period, 200 µl of the solution was transferred into a 96-well plate, and the fluorescence readings were done at an excitation wavelength of 540 nm and an emission wavelength of 635 nm.

Assessment of the biocompatibility of the PHBV and PCL using an ex-ovo CAM assay

To evaluate the initial response of the CAM to both polymers, electrospun PHBV and PCL were implanted on the CAM. CAM assays were conducted under the guidelines of the Home Office, UK. The detailed protocol of the assay has been described in our previous studies [18–20]. In brief, fertilised chicken eggs were purchased from MedEggs (Norwich, UK) and cleaned with 20% industrial methylated spirit solution. Eggs were incubated at 37.5 °C for three days in a rocking egg incubator (RCOM King SURO, P&T Poultry, Powys, Wales). On day 3, the embryos were transferred gently into sterile petri dishes to start ex-ovo culture and incubated at 38 °C in a cell culture incubator (Binder, Tuttlingen, Germany) until day 7. On day 7, both polymers were disinfected with 70% ethanol and circles were cut to a diameter of 7 mm. They were then implanted onto the CAM until day 14. On day 14, the embryos were euthanised, and the response of CAM to the polymers was evaluated macroscopically and histologically. Survival rates of the embryos in electrospun PHBV and electrospun PCL implanted groups were analysed using the Kaplan–Meier method [21].

Haematoxylin and eosin (H&E) staining were performed for the histological evaluation of the response of CAM to both polymers. Briefly, samples were fixed in 3.7% formaldehyde for 30 min, and the fixed samples were embedded in optimal cutting temperature tissue freezing medium and frozen in liquid nitrogen for 3 min. Sections were cut 5–10 µm thick using a cryostat (Leica Biosystems, Nussloch, Germany) at − 20 °C. Sections were then stained with haematoxylin for 90 s and eosin for 5 min [18].

Assessing the effect of gelatin coating on the attachment of cells to electrospun PHBV nanofibres

Gelatin coating and cellularisation of the PHBV scaffolds with HDMECs and HDFs

Electrospun PHBV scaffolds were manufactured. PHBV scaffolds were cut into circular pieces and disinfected with 70% ethanol solution for 45 min.

For gelatin coating of the PHBV scaffolds, they were submerged in a sterile 0.2% gelatin solution (w/w in PBS) and incubated at 37 °C for 4 h prior to washing with phosphate buffered saline (PBS) three times.

For co-culture groups, the PHBV scaffolds were cellularised with HDFs in indirect contact with HDMECs. Briefly, HDFs were isolated from human skin grafts taken from patients as described previously [19]. Ethical approval for the use of skin excised in routine surgical operations and not needed for treatment of patients was granted by the local ethical approval committee of National Health Service Trust, Sheffield, UK (Ethics reference: 15/YH/0177). All patients provided written informed consent. HDFs were used between passage 3–6.

HDMECs were purchased from PromoCell and used between passage 2–4. Once both cells reached 80–90% confluency. 2 × 10⁴ HDFs were resuspended in 0.1 ml of EC GM (PromoCell Endothelial Cell Growth Medium MV basal medium supplemented with 2% FCS, 0.4% EC growth supplement, 10 ng/ml EGF, 90 µg/ml heparin, 1 µg/ml hydrocortisone) and then seeded onto the one side of the PHBV scaffolds. Before seeding HDMECs, scaffolds were returned to the incubator for 2 h to allow HDFs to attach. Then scaffolds were taken from the incubator and flipped over. 2 × 10⁴ HDMECs were resuspended in 0.1 ml of EC GM and then seeded onto the other surface of the PHBV scaffolds. Before adding culture medium, scaffolds were returned to the incubator for an hour to allow HDMECs to attach. Then, 2 ml of culture medium was added to each well, and they were returned to the incubator.

AlamarBlue^®Cell Viability Assay to evaluate the metabolic activities of HDMECs on PHBV scaffolds

AlamarBlue^® Cell Viability Assay was performed to evaluate the effect of gelatin coating and co-culture with HDFs on HDMECs growth at days 1, 4 and 7.

Preparation of cellularised electrospun PHBV scaffolds for implantation

Pre-labelling of cells prior to cell seeding

For imaging cells on the scaffolds prior to implantation, HDFs and HDMECs were pre-labelled with CellTracker™ Green and CellTracker™ Red prior to seeding. In brief, to label HDMECs, 50 µg of CellTracker™ Red dry powder was dissolved in 7.3 µl of dimethyl sulfoxide (DMSO) to prepare 10 mM stock solution. Then, 7 ml of serum-free EC GM was added to prepare a ~ 10 µM working dye solution. The pre-warmed dye solution was added gently to T75 flask, and HDMECs were incubated for 1 h under growth conditions. To label HDFs, 50 µg of CellTracker™ Green dry powder was dissolved in 10.75 µl of DMSO to prepare 10 mM stock solution. Then 10 ml of serum-free EC GM was added to prepare ~ 10 µM working dye solution. The pre-warmed dye solution was added gently to T75 flask, and HDFs were incubated for 1 h under growth conditions. Both flasks were washed three times with PBS prior to cell seeding onto PHBV scaffolds.

Cellularisation of the electrospun PHBV scaffolds

Gelatin coated electrospun PHBV scaffolds were manufactured. The PHBV scaffolds were disinfected by submerging them in 70% ethanol for 45 min prior to washing them with PBS three times. Following the disinfection, scaffolds were transferred to 24-well plates for cell culture experiments.

For cellularisation of the PHBV scaffolds, three different cell combinations were used: (1) HDFs, (2) HDMECs, and (3) HDMECs seeded on the lower surface of the PHBV fibres followed by HDFs seeded on the upper surfaces of the scaffolds.

Cellularisation of the scaffolds with HDFs in isolation

HDFs were used between passage 3–6 when they had reached 80–90% confluency. 0.5 × 10⁵ HDFs were resuspended in 0.1 ml of EC GM and then seeded onto the PHBV scaffolds. Before adding EC GM, scaffolds were returned to the incubator for an hour to allow HDFs to attach. Then, 2 ml of culture medium was added to each well, and they were incubated at 37 °C overnight prior to implantation.

Cellularisation of the scaffolds with HDMECs

HDMECs were used between passage 2–4 and were used at 80–90% confluency. 0.5 × 10⁵ HDMECs were resuspended in 0.1 ml of EC GM and then seeded onto the PHBV scaffolds. Before adding EC GM, scaffolds were returned to the incubator for an hour to allow HDMECs to attach. Then, 2 ml of culture medium was added to each well, and they were incubated at 37 °C overnight prior to implantation.

Cellularisation of the scaffolds with HDMECs in indirect contact with HDFs

HDMECs and HDFs were cultured as described above and at 80–90% confluency. 0.25 × 10⁵ HDFs were resuspended in 0.1 ml of EC GM and then seeded on one side of the PHBV scaffolds. Before seeding HDMECs, scaffolds were returned to the incubator for 2 h to allow HDFs to attach. Then scaffolds were taken out from the incubator and turned over. 0.25 × 10⁵ HDMECs were resuspended in 0.1 ml of EC GM and then seeded onto the other surface of the PHBV scaffolds. Before adding EC GM, scaffolds were returned to the incubator for an hour to allow HDMECs to attach. Then, 2 ml of culture medium was added to each well, and they were incubated at 37 °C overnight prior to implantation.

After incubation of the HDMECs and HDFs on the scaffolds overnight either in isolation or in co-culture, PHBV scaffolds were fixed in 3.7% formaldehyde for 30 min, and the fixed samples were embedded in optimal cutting temperature tissue freezing medium and frozen in liquid nitrogen for 3 min. Sections were cut 5–10 µm thick using a cryostat (Leica Biosystems, Nussloch, Germany) at − 20 °C. Sections were then directly imaged under a fluorescent microscope (Olympus IX3, Tokyo, Japan).

Evaluation of the angiogenic activity of the cellularised scaffolds using ex-ovo CAM assay

To evaluate the angiogenic activity of cellularised PHBV scaffolds, they were implanted on the CAM for 7 days. Briefly, circles of 7 mm diameter were cut from the scaffold and cellularised with HDMECs, HDFs and HDMECs in indirect culture with HDFs prior to implantation.

Fertilised chicken eggs were incubated at 37.5 °C for 3 days in a rocking egg incubator (RCOM King SURO, P&T Poultry, Powys, Wales). On day 3, the embryos were transferred into sterile petri dishes to start ex-ovo culture and incubated at 38 °C in a cell culture incubator (Binder, Tuttlingen, Germany) until day 7. On day 7, cellularised PHBV scaffolds were implanted onto the CAM and incubated for a further 7 days. The EC GM was added to the scaffolds twice a day. On day 14, the embryos were euthanised, and the angiogenesis was evaluated macroscopically and histologically.

For histological evaluation, the CAMs were fixed in 3.7% formaldehyde for 30 min, and the fixed samples were sectioned to 5–10 µm thick using a cryostat as described previously. Sections were then stained with haematoxylin for 90 s and eosin for 5 min [18].

Quantification of angiogenesis

For the quantification of angiogenesis from macroscopic images, the images of the scaffolds implanted on CAM were taken using a digital microscope at day 14, and the number of blood vessels was quantified by counting all blood vessels growing towards the scaffolds in a spoke wheel pattern, as described previously [22].

For the quantification of angiogenesis from histological images, the total number of blood vessels adjacent to the scaffolds were quantified by counting blood vessels in H&E sections [18, 23]. Briefly, all discernible blood vessels adjacent to the scaffolds were counted by two independent researchers using two independent microscopes.

Statistical analysis

Statistical analysis was carried out with either one-way or two-way analysis of variance (ANOVA) using statistical analysis software (GraphPad Prism, CA, USA). Where relevant, n values are given in figure captions. Error bars indicate standard deviations in the graphs unless otherwise stated.

Results

Comparison of PHBV and PCL for their physical, mechanical, and biological performances

Contact angle measurements showed that the contact angle of the water droplet on PCL was slightly lower than the one on PHBV, indicating a less hydrophobic surface. SEM images of the electrospun PHBV and PCL showed that electrospinning of both polymers resulted in smooth and beadless nanofibre formation. Fibre diameters of PHBV and PCL were 0.67 ± 0.13 µm and 0.47 ± 0.1 µm, respectively. Mechanical characterisation of the polymers indicated that there was no statistically significant difference between the UTS of the scaffolds (3.62 ± 0.22 MPa and 3.94 ± 0.35 MPa, respectively for PHBV and PCL). However, the stiffness of the electrospun PHBV scaffolds was found to be approximately 4-times higher when compared with electrospun PCL scaffolds (0.49 ± 0.02 N/mm and 0.12 ± 0.01 N/mm, respectively for PHBV and PCL). The results of contact angle measurements, representative stress–strain curves, and SEM images of the PHBV and PCL scaffolds are given in Fig. 1.

Fig. 1.

Summary of the mechanical and physical properties of PHBV and PCL. A contact angle measurements, B representative stress–strain curves, C SEM images showing the ultrastructure of the electrospun PHBV and PCL scaffolds, D the results of mechanical testing of PHBV and PCL scaffolds

Metabolic activities of HDMECs on both polymers showed an increase over 11 days. By day 1, attachment of HDMECs was approximately 3.4-fold higher on PHBV fibres compared to PCL fibres. Moreover, no statistically significant difference was found between the metabolic activities of HDMECs on PHBV and tissue culture plastic (TCP) by day 1. At each time point, the metabolic activities of HDMECs on PHBV fibres were significantly greater when compared with the activities on PCL fibres. By day 11, the activity of cells on PHBV fibres was 1.1-fold higher than that measured on PCL fibres. Although both PHBV and PCL showed that they are both capable of providing a suitable environment for culturing HDMECs on them, PHBV gave slightly better results in terms of metabolic activity of cells (Fig. 2).

Fig. 2.

The metabolic activity of HDMECs cultured on PHBV and PCL over 11 days compared with cells cultured on TCP. ***p ≤ 0.001; *p ≤ 0.05; ns p ≥ 0.05, error bars indicate standard deviation (SD)

Following the in vitro assessment of the biological performances of the ECs on the PHBV and PCL, both polymers also showed good biocompatibility on the CAM with no adverse effects on embryo survival rates which were 78.8% and 76.5%, respectively for PHBV and PCL implanted groups. The histological evaluation of both implanted scaffolds showed similar changes in the structure of CAM with a small increase in cell density in the mesoderm layer in all scaffold groups. Both electrospun PHBV and PCL scaffolds showed good attachment to the CAM, and they showed similar degrees of cellular infiltration from the chick membrane (Fig. 3).

Fig. 3.

Macro images and histological images showing the response of ex-ovo CAM assay to electrospun membranes of PHBV and PCL. The lower graph shows the survival rate of the embryos when PHBV or PCL were implanted. Scale bars represent 3 mm and 200 µm for macro images and histological images, respectively

The effect of gelatin coating and co-culture with HDFs on HDMECs attachment and proliferation

The results of the AlamarBlue^® assay demonstrated that there was an increase in the activity of HDMECs from day 1 to 7 in all scaffold groups. TCP groups either coated or non-coated with gelatin showed higher activity when compared with the PHBV scaffold groups (Fig. 4).

Fig. 4.

The effect of gelatin coating and indirect co-culture with HDFs on HDMECs metabolic activity assessed using AlamarBlue^® Assay over 7 days

Gelatin coating of TCP did not significantly affect the growth of HDMECs at any of the time points. However, gelatin coating significantly improved the activity on HDMECs seeded on PHBV scaffolds at day 1 and 4. The increases in the metabolic activity with the gelatin coating were 1.9-fold and 1.4-fold at days 1 and 4, respectively. Although the difference was not statistically significant at day 7, gelatin coating improved the activity of HDMECs on PHBV scaffolds 1.2-fold when compared to control PHBV scaffolds.

The presence of HDFs in indirect contact with HDMECs also significantly improved the metabolic activity of HDMECs at all time points. The activity of HDMECs was increased 2.7-fold, 2.1-fold, and 1.6-fold at day 1, 4, and 7, respectively.

To evaluate the settlement of the cells on them, sections of the cellularised scaffolds with CellTracker™ labelled HDFs and HDMECs were directly investigated under a fluorescent microscope prior to implantation to CAMs. The results showed that both types of cells were attached to the surfaces of the scaffolds and formed a monolayer when seeded either in isolation or in indirect co-culture. Fluorescent images of the scaffold sections are shown in Fig. 5.

Fig. 5.

Cross-sections of the cellularised PHBV scaffolds with HDFs (labelled with CellTracker™ Green) and HDMECs (labelled with CellTracker™ Red) either in isolation or in co-culture prior to implantation to CAM for the evaluation of angiogenic activities. Scale bars represent 100 µm

Evaluation of the angiogenic activity of cellularised PHBV scaffolds in an ex-ovo CAM assay

The ex-ovo CAM assay results revealed that the presence of both HDFs and HDMECs either in isolation or when co-cultured together significantly increased the angiogenic activity in the area of implantation. None of the implanted scaffolds affected the embryo survival rate, which was over 77% for all groups. The presence of HDFs and HDMECs increased the angiogenic activity 1.7-fold and 2.3-fold, respectively. The results showed that when HDMECs were seeded on PHBV scaffolds, they stimulated angiogenesis significantly greater than when HDFs were used on their own. However, the most significant angiogenic response was observed when scaffolds were seeded with both HDMECs and HDMECs. PHBV scaffolds with HDMECs and HDFs increased the number of blood vessels from 17.5 ± 1.9 to 46.0 ± 3.4 compared to PHBV controls. The macroscopic and histologic evaluations of the CAM assay results are summarised in Fig. 6.

Fig. 6.

Representative images demonstrating the angiogenic potential of PHBV scaffolds cellularised with HDMECs or HDFs or HDMECs in indirect culture with HDFs. The graphs below show the quantified results from the macroscopic and histological analysis of the scaffolds. Scale bars represent 3 mm and 250 µm for macroimages and histological images, respectively. Black and green arrows indicate PHBV scaffolds and blood vessels, respectively. ***p ≤ 0.001; **p ≤ 0.01; p ≤ 0.05, n = 6±SD

Discussion

PHBV and PCL are biocompatible polymers widely used in tissue engineering applications [11, 17, 24, 25], and both polymers are commonly electrospun for the fabrication of tissue engineering scaffolds [26, 27]. In this study, we compared both polymers in terms of their physical, mechanical and biological performances in order to select one of them for further prevascularisation studies.

Our results showed that PCL was slightly less hydrophobic than PHBV, and SEM images of the electrospun PHBV and PCL showed that electrospinning of both polymers resulted in smooth and beadless nanofibre formation. In line with our findings, the high hydrophobicity of PHBV has previously been reported by several groups [24, 28]. However, despite the more hydrophobic characteristics of it, the biological evaluation of the polymers showed that the attachment of HDMECs onto PHBV nanofibres was approximately 3.4-fold higher compared to cell attachment to PCL fibres. This might be likely due to the surface stiffness of the PHBV being approximately four times greater, and the average fibre diameter of PHBV was approximately 1.4-fold higher than those of PCL. Jalali et al. have previously reported that substrate stiffness has an effect on the adhesive behaviour of vascular ECs. Their results indicated that the adherence of ECs and their exhibition of dense actin structures were higher when cultured on stiff surfaces than on soft surfaces, although the stiffness did not affect the EC viability [29]. Similarly, Ataollahi et al. [30] showed that the attachment and proliferation of ECs in stiff substrates were higher than those on soft substrates. In addition to the surface stiffness, the average fibre diameter of electrospun fibres has also been shown to have an effect on EC attachment. Rüder et al. reported that EC adhesion was facilitated by increasing fibre diameters [31]. Ko et al. [32] demonstrated that ECs proliferated better and homogeneously on electrospun scaffolds with higher than 600 nm fibre diameter than on the scaffolds with smaller diameters.

On day 1, the degree of attachment of HDMECs on PCL was dramatically lower than that on PHBV. However, the metabolic activity of HDMECs on PCL was quite close to that measured on PHBV on day 11, but it was still statistically significantly lower. This is likely because of the contact inhibition of the cell growth [33, 34] on PHBV after a certain time point as the surface area of the scaffolds that cells can attach and grow were quite similar, and the cells on PHBV became confluent earlier than PCL. After cells become confluent, the metabolic activity of HDMECs on PHBV might be slowed down and remained stable. Accordingly, the difference between PHBV and PCL on day 11 might be reduced due to the contact inhibition of the metabolic activity of HDMECs associated with reduced cell proliferation and migration [35]. Throughout the experiments, the metabolic activities of HDMECs on TCP were significantly higher at all time points. This is an expected outcome as TCP is an optimised material for cell culture, which is positively charged to improve the attachment and consequently, the proliferation of cells on it [36, 37]. We have previously observed this outcome in our previous studies where the attachment and growth on TCP were both significantly higher than those on polymeric scaffolds [12, 17, 19].

Further evaluations of PHBV and PCL on CAM indicated that both polymers had similar degrees of biocompatibilities without any adverse effects on embryo survival (78.8% and 76.5%, respectively for PHBV and PCL implanted groups) and caused similar changes in the histological structures of CAMs. We have previously reported the average survival rates of the chick embryos when cultured ex-ovo (shell-less) as 25%, 68%, and 83% by inexperienced, intermediate, and experienced users, respectively [20]. The survival rates of the chicks were between 73 and 75% when we implanted biphasic PCL barrier membranes [17] and 75% when we implanted drug releasing PHBV scaffolds [18]. The ex-ovo CAM assay has many advantages over the in-ovo CAM assay such as accessibility of the CAM during the implantation period, monitoring of angiogenesis, and quality of imaging of angiogenesis. However, the major drawback of shell-less culture is the comparably lower survival rates than in-ovo-CAM assay [20].

Electrospinning is a method that enables one to produce TE scaffolds with a wide range of properties in terms of material composition, fibre diameter, thickness, porosity, and degradation rates [38–41]. Electrospun nanofibres have been shown to provide better surface properties for ECs to adhere and proliferate on compared to microfibres [42–44]. This is likely due to the nanofibres being structurally similar to the ECM of natural tissue with their submicron-scale topography and highly packed morphology [42, 45]. Accordingly, we successfully fabricated PHBV electrospun fibres with a diameter of ~ 0.67 µm via electrospinning. Our laboratory has previously shown that PHBV nanofibrous scaffolds allow the diffusion of oxygen and nutrients while supporting cell attachment and growth [15, 46]. Although PHBV nanofibres provide a relatively favourable environment to ECs, in their natural environment, ECs are in contact with collagen, laminin and fibronectin rich ECM. [47, 48]. It is known that ECM plays a key role also in EC migration, morphogenesis, survival, vessel stabilisation [49]. Several studies have been reported on how ECs can be grown efficiently in vitro. These culture systems are mostly focused on surface coating with fibronectin, collagen and gelatin, and providing essential growth factors and the presence of stromal cells in contact with ECs [50, 51]. Communication of ECs with surrounding stromal cells such as smooth muscle cells (SMCs), fibroblasts or pericytes has also been proven to have significant importance for the angiogenic process [52, 53].

Our group has previously reported the positive influence of fibroblasts on improving the survival and growth of ECs in polymeric scaffolds when cultured in indirect contact [46]. Recently, we have demonstrated how non-cross-linked gelatin coating can be used to improve the attachment and growth of ECs to biomaterials that have relatively weak biological properties [19]. In line with our results, Ma et al. [54] reported that surface modification of electrospun PCL scaffolds with gelatin coating enhances the EC spreading and proliferation. The positive impact of either non-cross-linked or cross-linked gelatin coating on attachment and proliferation of human Schwann cells has been shown by Vleggeert-Lankamp et al. [55]. Accordingly, in this study, we showed that gelatin coating and co-culture with HDFs are both practical approaches to improve the viability of HDMECs in synthetic nanofibrous channels with an ultimate aim of stimulating angiogenesis as a prevascularisation approach. Our results demonstrated that gelatin coating did not affect the attachment and proliferation of HDMECs on TCP, whereas the metabolic activity of HDMECs grown on gelatin coated PHBV fibres at days 1 and 4 was significantly higher when compared to controls. Similarly, the presence of HDFs significantly improved the activity of HDMECs at all time points.

Our group has been working on methods to improve the culture of ECs on tissue engineering scaffolds, and most of these studies aimed to create in vitro platforms to study angiogenesis in vitro. However, none of these studies had attempted to develop a prevascularised construct to promote angiogenesis. Thus, in this study, we assessed the efficiency of gelatin coating and the presence of HDFs to increase the survival and growth of HDMECs on simple electrospun scaffolds to be used to induce angiogenesis in ex-ovo CAM assay.

The use of pro-angiogenic agents is a well-established approach to promote neovascularisation [56, 57], and vascular endothelial growth factor (VEGF) is recognised as the most effective stimulator of angiogenesis [58, 59] by taking key roles in the angiogenic cascade [60]. However, the exogenous use of VEGF has also been shown to cause leaky [61], permeable [62], and haemorrhagic [63] blood vessels, which are observed in tumorigenesis [64]. In addition, the exogenous use of pro-angiogenic agents is expensive, highly dose-dependent, and the delivery of them requires a very-well controlled system for the administration of these poorly stable agents at effective doses [4]. Thus, seeking alternative approaches such as the use of pro-angiogenic cells for the prevascularisation of the tissue engineering scaffolds to promote angiogenesis is worth exploring to circumvent the delayed neovascularisation of these constructs [65]. Prevascularisation is based on shortening the time required for neovascularisation in the post-implantation period by the inosculation of the pre-formed vasculature with the existing vessels of host tissue [66, 67]. To date, the use of several cell types has been reported in prevascularisation studies. Nor et al. [68] prevascularised poly-L-lactic acid (PLLA) scaffolds with HDMECs and showed the anastomosis of them with the natural vasculature of mice. Similarly, Unger et al. [69] reported that co-culture of HDMECs with osteoblasts induced the migration of the host’s blood vessels into developed TE construct. Duttenhoefer et al. [7] showed the formation of tubular structures by co-culturing EPCs and mesenchymal stem cells (MSCs) in polyurethane scaffolds for 7 days. Hadjizadeh et al. [70] demonstrated the formation of a network between two neighbouring fibres when PLLA fibres prevascularised with Human Umbilical Vein Endothelial Cells (HUVECs) were embedded in a fibrin gel and revealed that the presence of fibroblasts was seen on top of the fibrin gel.

The CAM assay is a rapid, cost-effective, and widely-accepted in vivo bioassay to evaluate the angiogenic activity of pro-angiogenic agents and cells as well as the initial response to biomaterials. It is easier to perform than most in vivo assays and allows direct visualisation of blood vessel formation in response to the test materials when performed shell-less (ex-ovo) [18, 20, 71]. We have previously demonstrated the efficiency of this ex-ovo CAM assay to measure the direct angiogenic response to biomaterials [17–20]. Following the determination of the optimal culture conditions for HDMECs, three different cell culture systems were investigated in ex-ovo CAM assay in terms of improving the angiogenic activity, as a prevascularisation approach. The results of the CAM assay revealed that both HDFs and HDMECs either in isolation or when co-cultured together significantly increased the angiogenic activity in the area of implantation. HDMECs were found to be more effective for stimulating angiogenesis. However, the most significant angiogenic response was observed when HDFs were also present in indirect contact with HDMECs. This increase in the angiogenic activity is more likely due to the growth factors that are released from cultured ECs in the scaffolds. ECs has previously been shown to secrete VEGF [72, 73] and basic fibroblast growth factor (bFGF) [74] even under normoxic culture conditions. In addition, co-culture of ECs with stromal cells has been shown to increase the secretion of VEGF when compared to mono-cultured cells [75]. Similarly, fibroblasts have previously been reported to play a crucial role in the angiogenic process by generating ECM molecules such as collagen and fibronectin [76, 77], growth factors, and pro-angiogenic factors [78, 79]. They have been shown to secrete VEGF, and fibroblast conditioned media has been reported to enhance the capillary development of ECs [80]. In vitro generated ECM has also been demonstrated to increase the neovascularisation of the implanted constructs in a rat animal model, and researchers have hypothesised that this is more likely due to the angiogenic factors that are released to and stored in the generated ECM [81]. Similarly, murine cells and murine cell-derived ECM have been shown to possess angiogenic properties in ex-ovo CAM assay possibly due to the growth factors, angiogenic factors, cytokines, and trace elements stored in the ECM [12].

A critical issue in the translation of tissue engineering substitutes into the clinic is the neovascularisation post-implantation. Although future investigations will be necessary to explore how cells influence angiogenesis at the molecular level in terms of growth factors and cytokines, the current study reveals essential information about how simple electrospun scaffolds can be functionalised with different types of pro-angiogenic cells to stimulate angiogenesis in vivo. To summarise, the observations made in this study suggested that gelatin coating and co-culture of HDFs both showed a positive impact on HDMECs viability and growth and the physical presence of HDMECs and HDFs either in isolation or in co-culture strongly induced angiogenesis in ex-ovo CAM assay. The results of the CAM assay demonstrated that the presence of HDMECs showed a stronger angiogenic reaction than the introduction of HDFs alone, but the use of HDMECs and HDFs together gave the most significant angiogenic activity.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical statement

Ethical approval for the tissue acquisition was granted by the National Research Ethics Service (NRES) Committee Yorkshire and The Humber–Sheffield (REC Ref.: 15/YH/0177, REC opinion date: 03/06/2015).