Introduction
In the context of bone tissue engineering (BTE), timely vascularization of in vivo implanted constructs is critical for cell survival, especially in regions distant from the host vasculature, since diffusion of oxygen and nutrients is only limited to a distance of 150–200 µm (Jain et al., 2005; Nguyen et al., 2012). Angiogenesis is an essential component of the bone regeneration cascade and its insufficiency is a major limiting factor for the clinical translation of BTE strategies (Kanczler and Oreffo, 2008). Mesenchymal stromal cells (MSC) are increasingly being used for BTE (Pittenger et al., 2019; Shanbhag et al., 2019), and one strategy has been to coculture MSC with endothelial cells (EC), to create in vitro “pre-vascularized” constructs with a network of primitive vessels that functionally anastomose with the host vasculature when implanted in vivo (Levenberg et al., 2005; Rouwkema et al., 2006). MSC are reported to stabilize these networks by adopting a pericyte-like phenotype, thereby enhancing EC-mediated angiogenesis and in turn, bone regeneration (Keramaris et al., 2012; Loibl et al., 2014; Shanbhag et al., 2017a).
MSC derived from bone marrow (BMSC) are the most widely tested. However, progenitor cells from less-invasive sources, e.g., adipose and oral tissues, are being explored (Friedenstein et al., 1968; Pittenger et al., 2019). Oral tissues, such as dental pulp, periodontal ligament and gingiva, represent alternative sources of “MSC-like” progenitor cells (Sharpe, 2016). Gingiva, in particular, can be harvested with minimal morbidity and contains a subpopulation of multipotent progenitor cells (GPC), which demonstrate an MSC-like phenotype, immunomodulatory properties, and osteogenic potential both in vitro and in vivo (Fournier et al., 2010; Mitrano et al., 2010), thus representing promising alternatives to BMSC for BTE applications (Stefanska et al., 2020).
A critical aspect in the clinical translation of cell therapies is the use of safe and standardized culture conditions. Although commonly used for MSC expansion, several limitations of xenogeneic fetal bovine serum (FBS) supplementation have been highlighted, and current recommendations from health authorities advocate the use of “xeno-free” protocols whenever possible (Bieback et al., 2019a). Accordingly, xeno-free alternatives such as pooled human platelet lysate (HPL), have emerged and have been shown to be comparable, and often superior, to FBS for the proliferation and differentiation of various types of MSC (Fekete et al., 2012; Shanbhag et al., 2017b; Shanbhag et al., 2020a). We have recently reported that xeno-free culture of human GPC in HPL vs FBS media results in enhanced growth, gene expression and differentiation (Shanbhag et al., 2020b). Moreover, the proliferation and tube formation of EC is reported to be enhanced in HPL (Tasev et al., 2015) and other xeno-free media (Bauman et al., 2018).
Current BTE strategies rely mainly on monolayer expansion of MSC in plastic-adherent cultures (Rojewski et al., 2019). However, this two-dimensional (2D) culture system is not representative of the 3D in vivo microenvironment of MSC and may therefore alter their phenotype and diminish their regenerative and immunomodulatory potential (Banfi et al., 2000; Hoch and Leach, 2015; Ghazanfari et al., 2017). Similar observations have been reported in EC; single dissociated EC are reported to be more likely to undergo apoptosis (Korff and Augustin, 1998). In contrast, the self-assembly or spontaneous aggregation of cells into 3D spheroids is mediated by unique cell-cell and cell-extracellular matrix (ECM) interactions, biomechanical cues and signaling pathways, which more closely simulate the in vivo microenvironment (Sart et al., 2014; Cesarz and Tamama, 2016). In contrast to 2D monolayers, 3D spheroid culture has been reported to enhance survival, stemness, paracrine activity, immunomodulation and multi-lineage differentiation of MSC (Follin et al., 2016; Petrenko et al., 2017) (Kale et al., 2000; Chatterjea et al., 2017). In the context of BTE applications, we have observed particular advantages of spheroid vs monolayer culture via a strong upregulation of osteogenesis-related genes in BMSC and GPC (Shanbhag et al., 2020b).
Traditional cell delivery methods involve direct seeding and attachment of cells on biomaterial scaffolds before in vivo transplantation. However, direct seeding may not be the optimal method for delivery of cell spheroids because the 3D structure, essential to maximize their in vivo effects, is lost. To preserve the 3D structure, encapsulation of spheroids in hydrogels represents an effective delivery system (Murphy et al., 2014; Murphy et al., 2015; Ho et al., 2018). Moreover, in the context of angiogenesis, when EC are cultured as spheroids in a hydrogel matrix, either alone or in coculture with MSC, 3D network formation occurs by closely mimicking in vivo sprouting angiogenesis (Korff and Augustin, 1999; Heiss et al., 2015). Since HPL is increasingly being used for clinical-grade MSC culture (Bieback et al., 2019b), extending its application as a hydrogel carrier represents a cost-effective strategy for tissue engineering. Furthermore, HPL gels may offer the added advantage of sustained cytokine release at regeneration sites (Robinson et al., 2016). Indeed, recent studies have demonstrated the potential of HPL hydrogels for encapsulating EC and MSC to create microvascular networks (Fortunato et al., 2016; Robinson et al., 2016).
Previous studies have investigated the capacity of MSC to support or enhance EC-mediated angiogenesis in monolayer cultures, most often in xenogeneic conditions (Steffens et al., 2009; Verseijden et al., 2010; Chen et al., 2013; Ucuzian et al., 2013; Strassburg et al., 2016). Others have studied angiogenesis-related outcomes in spheroid cocultures of MSC or fibroblasts with EC in xeno-free, i.e., human serum-supplemented, media (Eckermann et al., 2011; Bauman et al., 2018). In the former study, MSC-EC cocultures in xeno-free media (vs FBS) resulted in enhanced angiogenesis in an in vivo chick chorioallantoic membrane (CAM) assay. With this background, the objective of the present study was to test the in vitro and in vivo angiogenic potential of EC cocultured with GPC as 3D spheroids encapsulated in HPL hydrogels.
Materials and Methods
Cell Culture
The use of human cells and tissues was approved by the Regional Committees for Medical Research Ethics (REK) in Norway (2016–1266, REK sør-øst C). Monolayer cultures of primary human GPC isolated from healthy donors were established in 5% HPL (Bergenlys®, Bergen, Norway). Details of isolation and characterization of GPC have been reported elsewhere (Shanbhag et al., 2020b). Early passage human umbilical vein EC (HUVEC) were purchased and cultured in EGM-2 growth medium (both from Lonza Inc., Walkersville, United States) supplemented with either 2% FBS, as per the manufacturer’s recommendations, or with 5% HPL; all other media components were maintained. Cells were sub-cultured and expanded under humidified 5% CO2 at 37°C; passages 2-4 were used in experiments. Functionality of HPL cultured HUVEC was tested in an in vitro tube formation assay on matrigel (Corning, NY, United States), as previously described (Fujio et al., 2017). Phase contrast images (Nikon Eclipse TS100, Tokyo, Japan) were analyzed using ImageJ software (NIH, Bethesda, United States) and angiogenesis-related parameters (tube length, branching, segments and junctions) were automatically quantified using the Angiogenesis Analyzer plugin, as previously described (Carpentier et al., 2020).
3D Spheroid (co)Culture
3D aggregate spheroids of HUVEC were formed via guided self-assembly in microwell plates as recently described (Shanbhag et al., 2020b). Briefly, suspensions of dissociated monolayer HUVEC cultured in FBS or HPL, were seeded in microwell plates (Sphericalplate®, Kugelmeiers Ltd., Erlenbach, Switzerland) for 24 h to form spheroids of ∼1000 cells each. Cell viability in spheroids was assessed via the LIVE/DEAD® kit (Invitrogen). Sprout formation in FBS and HPL cultured HUVEC spheroids was assessed using phase and confocal microscopy: for the latter, immunofluorescence (IF) staining with CD31 was performed (Supplementary methods). For subsequent experiments, all cell culture was performed in HPL media. For coculture spheroids, microwells were seeded with suspensions of dissociated HUVEC and GPC in two different ratios, 1:1 and 5:1 (HUVEC:GPC), based on previous work (Ma et al., 2011; Pedersen et al., 2013). After 24 h, HUVEC and HUVEC-GPC spheroids were collected by gentle pipetting and encapsulated in HPL hydrogels (HPLG).
Encapsulation in Hydrogels
Since HPL was used to establish xeno-free cultures of GPC and HUVEC, its application as a hydrogel scaffold was also investigated. Initially, HPLG were produced via addition of thrombin solution [1 IU/ml human thrombin and one TIU/ml aprotinin in 40 mM CaCl2 solution (all from Sigma-Aldrich)] to sterile-filtered HPL followed by incubation at 37°C for 15 min. The resulting hydrogel was referred to as “unmodified” HPLG (0F). For encapsulation, HUVEC or coculture spheroids were suspended in HPL solution, quickly mixed with the thrombin solution, and added to culture plates with gentle shaking to ensure uniform distribution of the spheroids. The plates were transferred to the incubator for 15 min to ensure complete gelation and thereafter supplemented with EGM-2 growth medium for the indicated culture periods.
Subsequently, to improve the hydrogels mechanical properties, HPL was supplemented with fibrinogen (Sigma-Aldrich) in concentrations of 1.25 (1.25F), 2.5 (2.5F), 6.25, 12.5, and 25 mg/ml. Gelation and spheroid encapsulation was performed using the same thrombin solution as described above. These hydrogels were referred to as “modified” HPLG. Rheological properties of modified HPLG were assessed as described in the Supplementary methods. Only 0, 1.25 and 2.5F HPLG were used in subsequent experiments (see Hydrogel Properties Influence HUVEC Sprouting).
Sprouting Angiogenesis Assay
The in vitro angiogenic potential of mono- and coculture spheroids was tested in a sprout assay, as previously described (Nakatsu and Hughes, 2008). Briefly, HUVEC or HUVEC-GPC spheroids were encapsulated in HPLG and cultured for 72 h in EGM-2 medium to observe sprout formation. In selective experiments, HUVEC spheroids (encapsulated in 0F HPLG) were cultured with a monolayer of GPC on top of the gel, i.e., “indirect” cocultures – to test whether paracrine factors from GPC influenced HIUVEC sprouting. In “direct” cocultures, prior to spheroid formation, dissociated GPC and HUVEC were labeled with fluorescent green (DiO, 5 μL/ml) and red (Dil, 5 μL/ml) dyes (Vybrant® cell-labelling solution, Invitrogen), respectively. HUVEC-only spheroids (only red-labelled cells) were formed as controls. Spheroids of HUVEC or HUVEC-GPC (1:1 or 5:1 HUVEC:GPC) were encapsulated in modified HPLG (0, 1.25 or 2.5F), and cultured in EGM-2 for up to 72 h in 8-well μ-slides® (ibidi, Munich, Germany). Subsequently, the constructs were fixed in 4% paraformaldehyde (PFA) and permeabilized using 0.2% Triton X-100 (Sigma-Aldrich). Prior to imaging, nuclei were stained using 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich).
Confocal Microscopy
Whole mount imaging of HPLG-encapsulated spheroids was performed using an Andor Dragonfly 5050 high-speed confocal microscope and Fusion software (both from Oxford Instruments, Abingdon, United Kingdom). Z-stacks were acquired from the top of each gel, with steps of 4 μm to a depth of up to 200 μm. Each image was captured with a high-speed iXon 888 Life EMCCD camera with 1024 × 1024 resolution at 100–200 × magnification. Green (GPC) and red (HUVEC) stained cells/sprouts, and their nuclei (DAPI), were scanned in the corresponding channels using 546, 647 and 405 lasers, respectively. Images were processed using the Imaris software (Oxford Instruments) and transferred to ImageJ (NIH) for analysis. Using the Sprout Morphology plugin, segmentation and thresholding of images was performed to separate GPC, HUVEC and nuclei. Images were calibrated using scale bars and sprout lengths (in μm) were automatically or manually calculated using ImageJ, as described elsewhere (Eglinger et al., 2017), on segmented images showing only HUVEC in the red-channel.
CAM Assay
The in vivo angiogenic potential of mono- and coculture spheroids was tested in an ex ovo CAM assay in developing chick embryos, in accordance with the Norwegian Animal Research Authority (Mattilsynet), where an experimental period < 14 days did not require formal ethical approval. Briefly, fertilized chicken eggs were incubated at 37°C for 72 h with intermittent rotation. On embryonic day 3, the eggs were carefully opened, their contents transferred into petri dishes and incubated in humidified air at 37°C. On day 7, HUVEC or HUVEC-GPC spheroids encapsulated in 1.25F HPLG (50 spheroids in 50 µL gel; 1:1 HPL:EGM-2) were implanted on the CAMs avoiding pre-existing blood vessels. To maintain their positions on the CAMs, the gels were contained within silicone O-rings (⌀ 10 mm). During the incubation period, some embryos were terminated as a result of embryonic death unrelated to the implants. Implants from these terminated embryos were harvested for live cell-staining using Calcein AM (Invitrogen). On day 14, the regions within the O-rings in the remaining embryos were recorded using a digital stereomicroscope (Leica Biosystems, Heerbrugg, Switzerland). Subsequently, the CAMs were fixed in 4% PFA and regions around the O-rings were harvested, embedded in paraffin and analyzed histologically following hematoxylin and eosin staining. Quantification of angiogenesis-related parameters (vessel density, vessel length, segments and branching points) in CAM images was performed using the Wimasis® automated image analysis software (Onimagin Technologies, Cordoba, Spain) (Montali et al., 2017).
Statistical Analysis
Statistical analysis was performed using the Prism 9.0 software (GraphPad Software, San Diego, CA, United States). Data are presented as means ± SD, unless specified. Normality testing was performed via the Shapiro-Wilk test. The student t test, Mann-Whitney U test and one-way analysis of variance (ANOVA), followed by post-hoc Tukey’s or Dunn’s test for multiple comparisons, were applied when appropriate and p < 0.05 was considered statistically significant.
Discussion
3D cocultures of MSC and EC can promote angiogenesis and potentially overcome the challenges of in vivo vascularization in BTE constructs (Rouwkema et al., 2006; Nguyen et al., 2012). The aim of this study was to test whether GPC, as an alternative to BMSC, supported EC-mediated sprouting angiogenesis in xeno-free HPL cultures.
HUVEC represent a feasible and frequently used model to study EC behavior in experimental settings (Morin and Tranquillo, 2013). Consistent with previous reports, tube formation and sprouting by HUVEC was improved in HPL vs FBS. Sprouting angiogenesis by 3D-cultured EC in vitro is considered to be a close representation of the in vivo angiogenic cascade, recapitulating all the key events during which quiescent EC become activated to proteolytically degrade their surrounding ECM, e.g., hydrogels, directionally migrate towards the angiogenic stimulus, proliferate, and organize into new 3D capillary networks (Korff and Augustin, 1999; Chappell et al., 2012; Nowak-Sliwinska et al., 2018). Moreover, these sprout-networks have revealed functional lumenized capillaries, which anastomose with host vasculatures when implanted in vivo (Alajati et al., 2008; Finkenzeller et al., 2009; Morin and Tranquillo, 2011). A recent study reported superior sprouting of EC in human serum vs FBS supplemented media (Bauman et al., 2018). Consistently, a trend for superior sprout formation by HPL vs FBS cultured HUVEC spheroids was observed herein. Thus, HPL appears to be a feasible xeno-free alternative for HUVEC culture.
The formation and stability of in vitro and in vivo capillary-like networks by EC can be enhanced via coculture with MSC (Pedersen et al., 2012; Pedersen et al., 2013; Ma et al., 2014). We have previously reported that GPC demonstrate MSC-like phenotype and properties in xeno-free cultures (Shanbhag et al., 2020b). Accordingly, 3D cocultures of HUVEC and GPC were established herein. To test whether GPC promoted HUVEC sprouting via cell-to-cell contact or paracrine mechanisms, direct and indirect cocultures were established, respectively. While indirect coculture with GPC revealed a trend for greater HUVEC sprouting, direct coculture with GPC in a 5:1 ratio significantly improved HUVEC sprouting. These results are consistent with previous studies of HUVEC spheroids cocultured with BMSC (Hsu et al., 2014; Robinson et al., 2016; Bauman et al., 2018), and studies highlighting the importance of direct cell-to-cell contacts, rather than paracrine interactions, in coculture settings (Ball et al., 2004; Liang et al., 2017).
To optimize the 3D cocultures, two different coculture ratios were tested. While a 1:1 ratio of MSC and EC is most frequently reported (Ma et al., 2011; Shanbhag et al., 2017a), previous studies from our group and others have suggested that higher proportions of EC may improve angiogenesis in cocultures (Verseijden et al., 2010; Pedersen et al., 2012; Pedersen et al., 2013). However, no significant differences in HUVEC sprouting were observed between high (5:1) and low (1:1) coculture ratios herein. Notably, only the 5:1 cocultures showed significantly greater sprouting vs HUVEC only spheroids. Considerably greater spreading or migration of GPC was observed in spheroids with relatively more GPC, i.e., in 1:1 spheroids. Spreading preceded HUVEC sprouting and may have provided a substrate for sprout growth and elongation. Indeed, MSC are reported to show signs of pericytic differentiation, e.g., via expression of smooth-muscle markers, in EC cocultures (Lozito et al., 2009). Similar patterns of spreading by MSC have been reported in 3D cocultures embedded in collagen gels (Shah and Kang, 2018). This is in contrast to non-embedded 3D cocultures, where MSC do not spread, and EC, in the absence of an ECM, organize into internal networks within the spheroids rather than external sprouts (Rouwkema et al., 2006; Verseijden et al., 2010; Eckermann et al., 2011; Marshall et al., 2018). In the present study, an “embedded” spheroid model was selected to recapitulate angiogenic sprouting by using HPL hydrogels as ECM scaffolds to deliver the “pre-vascularized” constructs in vivo (Robinson et al., 2016).
Recent studies have demonstrated the benefits of HPLG for EC-mediated angiogenesis (Fortunato et al., 2016; Robinson et al., 2016). HPLG are produced by simulating the in vivo coagulation cascade, i.e., via addition of thrombin and/or CaCl2 to convert fibrinogen to fibrin, and thus represent highly biomimetic scaffolds for tissue engineering applications. Together with cell culture in HPL supplemented media, this would represent a fully xeno-free coculture system with a high potential for clinical translation. Although HPLG can support capillary-like network formation by EC, their mechanical properties may be considered insufficient for in vivo implantation, especially in non-contained bone defects. Thus, the HPLG were supplemented with fibrinogen for more predictable in vivo delivery. Fibrin gels are routinely used as scaffolds in a range of applications including BTE (Soffer et al., 2003). Moreover, fibrin gels have been extensively used to study EC sprouting angiogenesis (Morin and Tranquillo, 2013). Notably (unmodified) HPLG have been shown to be superior to fibrin gels in this regard (Robinson et al., 2016). However, the mechanical properties of unmodified HPLG may only allow injectable delivery due to their highly liquid nature. In the present study, it was hypothesized that supplementation of HPLG with fibrinogen would enhance the mechanical properties of the gels, while retaining the biological activity of HPL. Although the addition of fibrinogen seemingly improved the mechanical properties of HPLG, the biological activity (HUVEC viability and sprouting) declined beyond a concentration of 2.5 mg/ml. Interestingly, HUVEC sprouting in 1.25F gels was slightly enhanced vs unmodified HPLG and significantly enhanced vs 2.5F HPLG. This contrasted with a previous study comparing unmodified HPLG and 1.25 or 2.5 mg/ml fibrin gels (Robinson et al., 2016). In the context of BTE, hydrogel stiffness is also reported to influence MSC fate-determination and osteogenic differentiation (Hwang et al., 2015; Sun et al., 2018; Zigon-Branc et al., 2019). Our findings, together with previous reports (Rao et al., 2012; Hsu et al., 2014; Robinson et al., 2016), highlight the importance of ECM/scaffold properties on EC-mediated angiogenesis within tissue engineered constructs.
To test the in vivo angiogenic potential of spheroid-HPLG constructs, a CAM assay in the developing chick embryo was used. This offers a relatively rapid and cost-effective model for in vivo biomaterial/xenograft testing, particularly for angiogenesis, in a naturally immunocompromised host with a rapidly developing vascular bed (Moreno-Jimenez et al., 2016; Ribatti, 2016). Cell viability and sprouting of both HUVEC and HUVEC-GPC spheroids was confirmed after 24 h in excised CAMs. Interestingly, in the absence of GPC, HUVEC appeared to dissociate from spheroids and organize into tube-like networks as observed in monolayer cultures. Seven days after implantation, a dense network of capillaries was observed macroscopically in the CAM-regions implanted with both HUVEC and HUVEC-GPC spheroids. This is consistent with previous studies reporting angiogenesis in CAMs implanted with xenogeneic (Steffens et al., 2009; Strassburg et al., 2016) or xeno-free coculture spheroids (Bauman et al., 2018). In the latter study, the integration of sprouts with the CAM vasculature was confirmed via immunohistochemistry (Bauman et al., 2018). However, no significant advantage of HUVEC-GPC coculture was observed in the CAM assay herein, and therefore, the benefits of coculture observed for in vitro sprouting were not translated in vivo.
Some limitations of our study must be acknowledged. While most previous studies have reported the in ovo “eggshell window” method for the CAM assay (Steffens et al., 2009; Liu et al., 2012; Strassburg et al., 2016; Bauman et al., 2018), a complete ex ovo method was used in our study. In the former, the construct is placed on the CAM through an opening in the eggshell; retention of the embryo within the egg and coverage of the window during the experimental period are advantageous in terms of hydration and reduced risk of contamination. Exposure of the CAMs in our method contributed to dehydration and faster resorption of the HPLG, which may have compromised existing sprout-networks and precluded the formation of new sprouts. Moreover, a longer observation period was used herein (7 days) compared to previous reports (3 days) (Liu et al., 2012; Bauman et al., 2018), which may also have masked any “early differences” between the groups; a single “end-point” was selected herein to minimize disturbance and exposure of the CAMs. An in ovo model with shorter/multiple observation periods may offer a more reliable picture in future studies. Moreover, ectopic implantation of the constructs in more relevant animal models, e.g., immunocompromised mice, may provide further clues regarding hydrogel degradation and vascular anastomosis.
It has been reported that in the absence of supporting cells, EC networks are stable for a shorter duration in vitro (Pedersen et al., 2012; Pill et al., 2018). When implanted in vivo, the engineered vessels must remain stable long enough to anastomose with the native vasculature and sustain the implanted cells (Pedersen et al., 2013). In the present study, GPC were found to be organized in close contact with HUVEC sprouts and appeared to provide a “substrate” for sprout formation/elongation. Thus, it may be hypothesized that GPC could help to stabilize EC networks in more challenging in vivo conditions (Zhang et al., 2020). In the context of BTE, it is unclear whether MSC/GPC in cocultures serve dual functions of supporting angiogenesis and promoting osteogenesis, i.e., osteogenic differentiation and/or paracrine stimulation. In a meta-analysis of MSC-EC co-transplantation studies in vivo, we observed a significant benefit of coculture for bone, but not vessel, regeneration (Shanbhag et al., 2017a). Further research is needed to clarify whether MSC, and other supporting cells, adopt a pericyte- and/or osteoblast-like phenotype when cocultured with EC. Finally, further optimization of culture conditions, e.g., cell ratios, media, ECM/scaffolds, etc., to promote both osteogenesis and angiogenesis, and not one or the other, is needed prior to clinical application.