Chondrogenic Predifferentiation Inhibits Vascular Endothelial Growth Factor Angiogenic Effect in Pericranium-Derived Spheroids

Abstract

Craniofacial reconstruction of critical bone defects typically requires a bone graft. As graft availability may be restricted by disease or comorbidities, tissue engineering approaches are actively sought. The pericranium could provide new bone graft material. During development and repair, bone transitions through a chondrogenic phase. However, with tissue engineering, pluripotent cells can differentiate directly into bone cells. Does ability to recapitulate bone formation in vitro affect osteogenesis and vascularization of pericranium grafts? To answer this, we obtained tissue from nine patients with preplanned craniotomy surgery and studied three-dimensional osteogenesis and angiogenesis of pericranium-derived spheroids. First, we established growth and differentiation conditions on Matrigel. For each spheroid sample, we investigated (i) continuous osteogenic differentiation (COD) and (ii) osteogenic differentiation preceded by chondrogenesis (CD → OD). The effect of vascular endothelial growth factor (VEGF) was compared to VEGF supplemented with fibroblast growth factor, interleukin (IL)-1, IL-6, platelet-derived growth factor, and tumor necrosis factor-α, a growth factor mix (GFM) with possible synergistic effects. In this limited sample, we observed no age- or sex-related differences in cell expansion. Similarly, no statistically significant differences in osteogenic or angiogenic scores between COD or CD → OD spheroids were noted with regular media. In COD, however, VEGF statistically significantly increased angiogenesis compared to control media (p = 0.007). Also, in COD, both VEGF and VEGF + GFM increased osteogenesis (p = 0.047 and p = 0.038, respectively). By contrast, in CD → OD, neither VEGF nor VEGF + GFM yielded statistically significant angiogenesis or osteogenesis scores compared to control media. To understand these results, we characterized spheroid protein expression by nanoliquid chromatography coupled to tandem mass spectrometry. Nine angiogenic proteins were either uniquely expressed or upregulated in COD compared to CD → OD: (i) endothelial markers JUP, PTGIS, PTGS2, and TYMP, (ii) tissue remodeling factors CHI3L1 and MMP14, and (iii) metabolic pathways modulators ANGPTL4, ITGA5, and WNT5A. ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A define a conserved angiogenic network and were >2-fold increased in VEGF compared to VEGF + GFM. Finally, we examined bone formation on printable poly-(propylene-fumarate) (PPF) scaffolds for individualized grafting. Under COD + VEGF conditions, PPF scaffolds loaded with pericranium-derived cells displayed hallmarks of spongiform-like bone formation. Thus, the human pericranium may be a potential repository for bone-generating cells with applications in craniofacial bone repair using tissue printing.

Impact statement

The work presented in this study establishes continuous osteogenic differentiation (COD), without a preliminary chondrogenic differentiation phase (CD → OD) as the preferred method to generate vascularized, osteogenic graft tissue using human pericranium-derived cells (PDCs). Using nanoliquid chromatography-tandem mass spectrometry analysis, we identified nine angiogenic proteins that were either uniquely expressed or upregulated in COD compared to CD → OD: (i) endothelial markers JUP, PTGIS, PTGS2, and TYMP, (ii) tissue remodeling factors CHI3L1 and MMP14, and (iii) metabolic pathway modulators ANGPTL4, ITGA5, and WNT5A. PDCs grown under these conditions generated trabecular bone on poly-(propylene-fumarate) scaffolds for individualized grafting.

Introduction

Craniofacial reconstruction surgery involves repair of bone and soft tissue of the head and neck secondary to trauma, removal of malignant or benign tumors or tissue necrosis. A typical limitation is the regenerative capacity of local tissue to close-off the original defect. Critical bone defects cannot heal on their own and require a bone graft. Of the ∼600,000 bone grafts performed each year in the United States over 30,000 are for craniofacial reconstruction.¹ Oftentimes, tissue availability depends on pre-existing conditions. For example, fibular free tissue transfer is the “gold standard” for mandibular reconstruction,² but cannot be implemented in many cases.³

Tissue engineering using effective combinations of scaffolds, stem cells, and cytokines may provide an alternative bone source. Useful scaffolds promote cellular adhesion, minimize immunogenicity, are biodegradable, and support differentiation.⁴ Recently, poly-(propylene-fumarate) (PPF) scaffolds with optimal interconnected porous structures and printing capability were developed for personalized grafting.^5,6 Stem cells impose a trade-off between pluripotent potential and unintended effects. For example, issues of ethics or teratoma formation limit use of embryonic and induced pluripotent stem cells, respectively.⁷ By contrast, adult mesenchymal stem cells (MSCs) are free of these issues, abundant, and easy to harvest, but their osteogenic potential varies from tissue to tissue. To date, bone marrow- and adipose-derived mesenchymal stem cells (ADSCs) have been the preferred source for bone engineering applications.

Periosteum represents another possible source of MSCs for bone applications because it has the ability to heal large defects in both long and flat bones.⁸ In the head and neck, the pericranium represents a large sheet of well vascularized tissue that is relatively easy to harvest with low morbidity during preplanned craniotomy.

The periosteum participates in bone fracture healing through timed production of a cartilaginous callus, tissue vascularization, and matrix mineralization followed by apoptotic chondrocyte death and osteoblast repopulation.⁹ During healing, the chondrogenic callus precedes bone formation. As new bone adjoining the fractured ends has greater strength than the original tissue. there is the prospect that induction of an initial chondrogenic state may be beneficial for in vitro bone formation.

Growth factors (GFs) are critical components of osteogenesis and angiogenesis.¹⁰ Vascular endothelial growth factor (VEGF) promotes new blood vessel formation¹¹ while platelet-derived growth factor (PDGF) increases vascularity and maturation of the vascular wall.¹² In endothelial cells, VEGF is sequestered in Weibel-Palade bodies (WPBs) as multimeric complexes with von Willebrand Factor (vWF), an endothelial cell biomarker, which also regulates tissue thrombosis.¹³ vWF is a biomarker of early angiogenesis, being exclusively produced by endothelial cells and megakaryocytes. As such, vWF is routinely used for histochemical identification of capillaries in tissue sections.^14,15 In addition, proinflammatory factors such as tumor necrosis factor (TNF)-α, interleukin (IL)-1, and IL-6 play critical roles in osteogenesis.^15–17

We used spheroids to study pericranium-derived cells (PDCs), bone formation, and angiogenesis as in vitro three-dimensional (3D) models are relevant to in vivo tissue architecture, cellular interconnectivity, and vascular network conditions,¹⁸ whereas two-dimensional (2D) models lack these features.

The goals of our study were twofold. First, we compared two models of PDC osteogenesis: continuous osteogenic differentiation (COD) and osteogenesis that is preceded by a chondrogenic phase (CD → OD). And second, we compared the effect of VEGF alone to that of VEGF supplemented with a growth factor mix (GFM) containing FGF, IL-1, IL-6, PDGF, and TNF-α, with possible synergistic effects.

In the absence of GFs, we find no difference in osteogenesis/angiogenesis scores between the two models. However, in COD, VEGF supplementation increased both osteogenesis and angiogenesis. An overexpressed, highly conserved, five-protein network consisting of ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A was identified by nanoliquid chromatography coupled to tandem mass spectrometry (nLC-MS/MS) analysis of spheroids grown in COD + VEGF media. Furthermore, electron micrographs of printable PPF scaffolds populated with PDCs and grown in COD + VEGF media showed hallmarks of spongiform-like bone formation. These results help establish protocols for bone tissue engineering using PDCs with downstream applications in personalized reconstructive craniofacial surgery.

Methods

Isolation of PDCs

Following Institutional Review Board approval for the study (IRB 18-000782), consent was obtained to harvest pericranium tissue during preplanned open craniotomy. With the pericranium exposed, roughly 1–3 cm² of pericranium was harvested. The tissue was placed in sterile saline, and immediately processed for PDC extraction using established protocols for MSC isolation.⁴ Buffers and media were from Life Technologies, unless otherwise stated. Briefly, pericranium tissue was minced with surgical scalpels in 3.0–5.0 mL of freshly prepared 1–1.5 mg/mL collagenase type 1 solution (C0130-1G; Sigma, St. Louis, MO) in advanced minimal essential media (AMEM, 12492013) containing 10% stem cell fetal bovine serum (12662-029) and 1% Glutamax/Pen–Strep (35050-061 and 15240-062, respectively). After 2 h of incubation at 37°C, the fat layer was separated by centrifugation at 500 g for 5 min and discarded. The pellet containing PDCs and other debris was washed, strained through a 70 μm sieve to remove large particulate material, and the cellular fraction was recovered by centrifugation at 500 g for 5 min. Cells were resuspended in AMEM and allowed to attach to polystyrene tissue culture flasks overnight in a 37°C, 5% CO₂, and 95% humidity atmosphere. The following day, media containing nonadherent cells were removed and fresh AMEM added. Fresh media were added every 2–3 days and cells passaged 1:2 at 60–80% confluency. Between 4 and 6 passages were needed to collect ∼10⁷ cells at which time aliquots were stored for cryopreservation and Matrigel incubation experiments started. Human adipose MSCs used in control experiments on PPF scaffolds were a gift from Al Dietz.

PDC differentiation on 3D scaffolds

For Matrigel experiments, 24-well plates were coated with 0.5 mL ice-cold Matrigel (Corning; 356234) and allowed to gel 30 min at 37°C. Then, 1 × 10⁶ PDCs in 0.5 mL of stem cell media (SCM, 12492-013) with 10% Matrigel were layered on top. After 2–3 days growth in SCM, spheroids were differentiated using Life Technologies differentiation kits A1007201 (COD) or A1007101 (CD → OD). Total differentiation time was 21 days (Fig. 2A). For CD → OD, spheroids were incubated 7 days with 100% chondrogenic media, 3 days with 50:50 mixture of chondrogenic and osteogenic media, and then 11 days with 100% osteogenic media. To study GF effects, 50 ng/mL final concentration each of VEGF, FGF, IL-1, IL-6, PDGF, and TNF-α was added in SCM media. For PPF scaffold incubations, 24-well plates were precoated with poly-HEMA to minimize cell adhesion to the well¹⁹ and maximize scaffold loading. Printed 1 × 1 × 1 cm, 44% porosity PPF scaffolds were sterilized with 95% ETOH and ultraviolet light, loaded with 2.5 × 10⁶ cells, and gently rocked overnight to enhance pore penetration. The following day, fresh media were added, and tissue culture was maintained for 21 days with medium changes every 3–5 days.

FIG. 2.

Osteogenesis and angiogenesis of PDC spheroids. (A) Experiment schematic. (B) Osteogenic differentiation of patient PDCs; 1 × 10⁶ cells were incubated with OM for 21 days, fixed, and stained with Alizarin Red to demonstrate osteogenic potential. 10 × optical microscope images shown for each patient sample. (C) Osteogenic score in COD and CD → OD spheroids. Spheroids were incubated as shown in (A) with media alone or media containing VEGF or VEGF + GFM (fibroblast growth factor, IL-1, IL-6, platelet-derived growth factor, and tumor necrosis factor-α), fixed in 4% paraformaldehyde, paraffin embedded, and prepared into 5 μm section slides for staining with Alizarin Red. Scores were calculated using pixel-by-pixel image analysis with IHC-Profiler plugin for ImageJ and averaged from five slides. Values are expressed as % of highest score taken as 100%. Statistical significance (p-values) between groups (bars) was calculated with two-tailed student's t test. (D). Angiogenic scores. Spheroid incubations, tissue processing, and image analysis were performed as in (C). Staining for angiogenesis was with Alexa 488-conjugated rabbit anti-human von Willebrand Factor monoclonal antibody. CD → OD, chondrogenic predifferentiation followed by osteogenic differentiation; COD, continuous osteogenic differentiation. CM, chondrogenic media; OM, osteogenic media; GFM, growth factor mix; VEGF, vascular endothelial growth factor; IL, interleukin. Color images are available online.

Tissue processing, staining, and image analysis

After 21 days, spheroids were fixed in 4% paraformaldehyde, paraffin embedded, and 5 μm tissue sections were prepared. For staining, slides were deparaffinized in xylene or xylene substitute (2 × 3 min), rehydrated with sequential 100%, 90%, and 70% ETOH (2 × 2 min each), and washed with distilled water. Staining for osteogenic differentiation with Alizarin Red (Sigma; 5533-25G) was performed according to the manufacturer's protocols. vWF staining was performed with Alexa 488-conjugated rabbit anti-human vWF monoclonal antibody EPSISR15 from Abcam.²⁰ Image capture and analysis were performed with IHC Profiler plugin for ImageJ.²¹ In all cases, scores for five slides were averaged. The result was a digital osteogenic and angiogenic score for each sample and incubation condition. To capture differences between samples, final scores were expressed as a percentage of the maximum score (taken as 100%) for each condition.

Scanning electron microscopy imaging

PPF scaffolds with cells fixed in 4% paraformaldehyde were washed in Trump's fixative²² for at least 1 h at 4°C, washed in phosphate buffer, rinsed in water, and dehydrated through graded ethanol solutions before critical point drying with carbon dioxide. Finally, scaffolds were mounted on an aluminum stub and sputter coated for 60 s with gold-palladium. Imaging was performed with a Hitachi S-4700 cold field emission scanning electron microscope (SEM) at 5 kV accelerating voltage.

nLC-MS/MS analysis

Protein recovery from formalin-fixed paraffin-embedded tissue slides and processing for tandem mass spectrometry analysis were performed as previously described.²³ After Matrigel digestion with MatriSperse,²⁴ equal cell counts recovered from M2, F3, and F6 were pooled and duplicate analyses were performed. Database searching and criteria for protein inclusion were previously described.²³ The full sample preparation protocol is available online²⁵ Only proteins with Log₂ ion peaks >20 and two or more unique peptides were compared across groups.²⁶

Statistical analysis

One-way analysis of variance (ANOVA) for treatment differences, regression analysis for age- and sex-related correlations, Fisher's exact test for sex-related effects, and student's t tests for group mean differences were performed in Excel or R. Differences in “Angiogenesis” gene ontology (GO) term enrichment were calculated with z scores. In all cases, statistical significance was assigned at p < 0.05.

Results

Preparation of PDCs

A total of 19 patients undergoing craniotomy and specimen collection were enrolled in the study. Four preparations become contaminated and were discarded. Six specimens were cryopreserved and not further processed. Clinical information on the nine patients who provided experimental data is shown in Table 1. Patients ranged in age from 20 to 81 years, 67% were females and 33% males. A typical starting material is shown in Figure 1A. With one exception (M1), all PDC preparations reached 1 × 10⁷ cells within ∼40 days, requiring 4–6 passages, except F6 that required six passages (Fig. 1B). No age or sex correlations with osteogenic or angiogenic scores were observed (Supplementary Fig. S1 and Supplementary Table S1). These initial data show that human PDCs from both sexes and various ages expand in tissue culture in sufficient numbers to provide starting material for bone grafting.

FIG. 1.

Extraction, expansion, and differentiation of PDCs. (A) Typical size of harvested tissue swaths. (B) PDT. Cells were counted and passaged at 60–80% confluency. PDT was measured at passage 2 and 3 according to PDT = CT × log₂/[log(Nh) − log(Ns)], where Ns and Nh are the number of seeded and harvested cells, respectively, and CT is the cell culture time (hours). Numbers above bars represent total passages required to yield 1 × 10⁷ cells. p = passage number. (C) Chondrogenic and (D) osteogenic differentiation. Differentiation media and staining were performed as described in section “Methods.” 10 × Magnification. PDC, pericranium-derived cell; PDT, population doubling time. Color images are available online.

Table 1.

Clinical Data for Tissue Donations That Yielded Experimental Data

ID	Age	Sex	Medical condition
F1	63	Female	Left frontal convexity meningioma
M1	69	Male	Right temporal bone encephalocele and CSF fluid leak
F2	81	Female	Large right frontal meningioma
M2	29	Male	Camurati-Engelmann Syndrome, stenosis of left internal auditory canal, left-sided hearing loss, and facial nerve neuropathy
F3	46	Female	Temporal encephalocele, CSF leak
M3	61	Male	Progressive Graves ophthalmopathy, bilateral ptosis and vision loss
F4	45	Female	Esthesioneuroblastoma
F5	20	Female	CPA petroclival epidermoid

CFS, cerebrospinal fluid; CPA, cerebellopontine angle.

COD and CD → OD differentiation of PDCs

Following bone fracture, a cartilaginous callus appears at the reduction site that later progresses to new bone tissue with greater strength than the original bone.⁹ Experiments were performed to study the effect of a chondrogenic state preceding osteogenesis compared to uninterrupted osteogenesis on new bone formation and angiogenesis using Matrigel spheroid cell culture. We compared the effect of VEGF²⁷ alone or supplemented with FGF, IL-1, IL-6, PDGF, and TNF-α. The differentiation scheme is shown in Figure 2A. To demonstrate osteogenic differentiation, duplicate incubations of 1 × 10⁶ PDCs per patient were performed in osteogenic media followed by staining with Alizarin Red. Representative tissue sections are shown in Figure 2B. Similar incubations in chondrogenic media were performed to demonstrate chondrogenic differentiation (not shown). Osteogenic and angiogenic scores were calculated with the IHC Profiler²¹ and are shown in Supplementary Tables S3, S4, S5. Osteogenic and angiogenic scores across all six groups were initially compared by one-way ANOVA, resulting in p = 0.016 for angiogenesis and p = 0.217 for osteogenesis (Supplementary Data). Further analysis was performed for COD and CD → OD separately. Under COD conditions, VEGF alone (p = 0.047) and VEGF + GFM (p = 0.038) statistically significantly increased osteogenic scores compared to control media (Fig. 2C). In addition, VEGF alone (p = 0.007), but not VEGF + GFM (p = 0.051), also statistically significantly increased angiogenic scores (Fig. 2D). By contrast, In CD → OD, neither VEGF nor the VEGF + GFM was effective compared to control media (Fig. 2C, D). These results suggest that factor(s) in the GFM may oppose VEGF angiogenesis, but not its osteogenic effect.

Tandem mass spectrometry analysis of protein expression

For nLC-MS/MS analysis, spheroids from three patients were pooled and duplicates analyses were performed. Protein peaks common between duplicate runs were filtered to include only those with associated Gene Names and two or more unique peptides.²⁸ For the COD group, 662 control (CT), 673 (VEGF), and 725 (VEGF + GFM) proteins were further considered, and for the CD → OD group, 569 (CT), 700 (VEGF), and 600 (VEGF + GFM) were considered. Next, a search for angiogenic proteins was performed using the amigo.geneontology.org search engine with the term “Angiogenesis.” Of the 4267 gene products identified, 598 are expressed in Homo sapiens, and 490 are nonredundant protein terms (Supplementary Data). To determine if any of the treatments produced an enrichment in angiogenic proteins, we performed Venn analysis (Fig. 3A, B) followed by z-score statistical analysis of “Angiogenesis” GO enrichment (Fig. 3C). Slightly different enrichment scores were observed, but none reached statistical significance. Next, we analyzed expression levels of 44 angiogenic proteins representing the most inclusive pool across all samples (Table 2). We identified nine proteins that were either uniquely expressed or upregulated in COD compared to CD → OD: the endothelial markers JUP, PTGIS, PTGS2, and TYMP, the tissue remodeling factors CHI3L1 and MMP14, and the activators/inhibitors of metabolic pathways ANGPTL4, ITGA5, and WNT5A (Table 2). Of these proteins, ANGPTL4, CHI3L1, JUP, and TYMP were >2-fold increased in COD + VEGF compared to COD + VEGF + GFM. Using STITCH network linking analysis,²⁹ we identified a highly conserved five-protein network (ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A) that is overexpressed in the COD + VEGF (Fig. 3D). The protein–protein interaction enrichment p-value was 0.000162, indicating more interactions among themselves than what would be expected for a random set of proteins of similar size, drawn from the genome. Hence, the proteins are at least partially biologically connected as a group.

FIG. 3.

Angiogenic protein expression in PDC spheroids. (A) COD and (B) CD → OD analysis of spheroid proteins by nanoliquid chromatography coupled to tandem mass spectrometry. Only proteins with log₂ values >20 and 2 or more unique peptides were considered (see section “Methods”). Angiogenic protein content is shown relative to 490 nonredundant “Angiogenesis” gene ontology protein terms retrieved from amigo.geneontology.org (Supplementary Table S2). (C) Relative enrichment scores showed no statistically significant differences. (D) Network interaction analysis of ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A using STITCH. These proteins were >2-fold overexpressed in the COD + VEGF. Color images are available online.

Table 2.

Angiogenic Proteins in Pericranium-Derived Cell Spheroids

Genes	COD CT	COD VEGF	COD MIX
ANGPTL4	ANGPTL4	ANGPTL4 ↑	ANGPTL4
ANPEP	ANPEP	ANPEP	ANPEP
ANXA1	ANXA1	ANXA1	ANXA1
ANXA2	ANXA2	ANXA2	ANXA2
APOD	APOD	APOD
C3	C3	C3	C3
CAV1	CAV1	CAV1	CAV1
CHI3L1		CHI3L1
CLIC4	CLIC4	CLIC4	CLIC4
COL18A1
DCN	DCN	DCN	DCN
EMILIN1	EMILIN1	EMILIN1	EMILIN1
FAP	FAP		FAP
FN1	FN1	FN1	FN1
GPNMB	GPNMB		GPNMB
GREM1			GREM1
HSPB1	HSPB1	HSPB1	HSPB1
HSPG2
ITGA5			ITGA5
ITGAV	ITGAV	ITGAV	ITGAV
ITGB1	ITGB1		ITGB1
JUP	JUP	JUP↑
KRT1	KRT1	KRT1	KRT1
LOXL2	LOXL2	LOXL2	LOXL2
MFGE8	MFGE8	MFGE8	MFGE8
MMP14	MMP14		MMP14
MMP2	MMP2	MMP2	MMP2
MYH9	MYH9	MYH9	MYH9
PGK1	PGK1	PGK1	PGK1
PKM	PKM	PKM	PKM
PTGIS	PTGIS	PTGIS↑	PTGIS
PTGS2		PTGS2
RNH1	RNH1	RNH1	RNH1
S100A7	S100A7	S100A7
SARS	SARS	SARS	SARS
SERPINE1	SERPINE1	SERPINE1	SERPINE1
SERPINF1	SERPINF1	SERPINF1
TGFBI	TGFBI	TGFBI	TGFBI
THBS1	THBS1	THBS1	THBS1
THBS2	THBS2	THBS2	THBS2
THY1			THY1
TYMP		TYMP
WNT5A		WNT5A↑	WNT5A

Spheroids were prepared as described in section “Methods” and analyzed by nLc-MS/MS. For simplicity, only data for the COD family are shown. Group to group (e.g., COD CT vs. CD → OD CT) and within family (e.g., COD CT vs. COD VEGF) changes in protein levels indicated. Group to group: bold, ≥2-fold upregulated, italic, ≥2-fold downregulated. Within family: underlined, uniquely expressed, ↑, ≥2-fold upregulated compared to COD CT.

nLc-Ms/MS, nanoliquid chromatography coupled to tandem mass spectrometry; COD, continuous osteogenic differentiation; CD, chondrogenic differentiation; VEGF, vascular endothelial growth factor; CT, control.

SEM analysis of osteogenic differentiation on PPF scaffolds

Biodegradable PPF scaffolds with adjustable porosity were previously described.^5,6 These scaffolds are printable, thus amenable for personalized grafting. To investigate osteogenic differentiation, 2.5 × 10⁶ PDCs were applied to 1 × 1 × 1 cm, 44% porosity PPF scaffolds (Fig. 4A) and incubated with osteogenic media (COD CT) or osteogenic media supplemented with VEGF (COD + VEGF). For comparison, 2.5 × 10⁶ ADSCs were also incubated with control osteogenic media. Using SEM, minimal osteogenesis was observed in the ADSC control group at 21 days, consistent with reported osteogenic differentiation potential for these cells (Fig. 4B). By contrast, enhanced osteogenesis with striations reminiscent of trabecular bone formation was observed with PDCs (Fig. 4C). Finally, addition of VEGF further enhanced bone formation, with long, organized bone trabeculae (Fig. 4D). The results demonstrate enhanced bone formation with VEGF supplementation on PPF scaffolds.

FIG. 4.

Trabecular bone formation on PPF scaffolds. (A). Scaffold incubation in 24-well plates. Printed, sterile 1 × 1 × 1 cm PPF scaffolds were seeded with 1 × 10⁷ adipose-derived mesenchymal stem cells (B) or PDCs (C, D) and incubated for 21 days in basic osteogenic media (B, C) or osteogenic media supplemented with VEGF (D). Scanning electron micrographs of surface bone formation shown. PPF, poly-(propylene-fumarate). Color images are available online.

Discussion

In craniofacial reconstruction surgery, free tissue flaps are used to repair critical defects, but grafting material is often restricted by comorbidities or associated diseases. Therefore, tissue engineering solutions are actively investigated. In this study, we take advantage of the reported potential of the periosteum to fill large bone defects.⁸ We used Matrigel to develop methods for 3D osteogenesis and vascularization for translation to PPF scaffolds that can be tailor-made to patient specifications. PDCs extracted from tissue samples from male and female patients 20–81 years old demonstrated ability to expand and differentiate into bone cells irrespective of sex and age. In the absence of GFs, no change in osteogenesis and angiogenesis scores was observed in COD compared to CD → OD. However, addition of VEGF uniquely increased angiogenesis in COD. Both VEGF and VEGF + GFM also increased osteogenesis in COD. By contrast, in CD → OD, addition of GFs was ineffective on these measures. nLC-MS/MS analysis of spheroid tissue sections revealed that factor(s) in GFM oppose the proangiogenic effect of VEGF, resulting in downregulation of a highly conserved network involved in angiogenesis containing ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A.

To optimize conditions for cell growth, we first studied PDC spheroids on Matrigel. Matrigel is a basement membrane-derived preparation extracted from Engelbreth–Holm–Swarm mouse sarcoma that provides an extracellular matrix-like 3D environment for cell anchorage and growth. It has been widely used to study spheroid and organoid biology with better inferences for translational work than 2D tissue culture.³⁰ Non-GF reduced Matrigel (“Matrigel for Organoid Culture”; Corning) was used throughout these experiments motivated by the observation that osteogenic properties are accelerated in MSC spheroids cultured on complete Matrigel.³¹ This preparation contains over 1300 different proteins (of which ∼63% are also present in GF reduced Matrigel) when analyzed by mass spectrometry,³⁰ including transforming growth factor (TGF)-β, epidermal growth factor, insulin-like growth factor 1, FGF, and PDGF,³² which were shown to promote osteoblast proliferation in vitro.³³ In addition, β-FGF, TGF-β, and PDGF in serum-free medium were necessary for in vitro osteogenic differentiation of human bone marrow-derived MSCs.³⁴

We hypothesized that prior induction of a chondrogenic state in PDC spheroids may affect 3D osteogenesis and vascularization. This was predicated on observations that, in bone fractures, periosteum initiates a cartilaginous callus with nascent blood vessels.³⁵ As cartilage-surrounding extracellular cell matrix becomes mineralized, chondrocytes undergo apoptosis, and osteogenesis takes place with the end results that bone with better strength is produced.⁹ However, in the absence of GFs, osteogenic and angiogenic scores for COD and CD → OD were similar even though exposure to osteogenic media in the CD → OD group was approximately half of that in the COD group (11 days compared to 21 days). The data do not allow us to distinguish if chondrogenic and osteogenic media exposure induced similar differentiation processes or short exposure to osteogenic media is sufficient to activate them.

vWF is a multidomain protein-binding glycoprotein present in plasma and endothelial cells and an early biomarker of angiogenesis. VWF is essential for WPB formation that regulates angiogenin-2 (Ang-2) levels by controlling storage and synthesis. Ang-2 released from WPB synergizes with VEGFR2 signaling to destabilize blood vessels and promote angiogenesis.^13,36 Our data show that in COD spheroids, VEGF uniquely increased vWF levels relative to control media and VEGF + GFM (Fig. 2D). Recent observations that vWF promotes angiogenesis through its heparin-binding domain³⁷ raise the possibility that proteins that compete for binding may sequester vWF into nonproductive complexes. To further understand protein production by GFs, we performed nLc-MS/MS analysis of spheroids. Among the sets of proteins considered for analysis, there was no statistically significant enrichment in angiogenic proteins, suggesting that differences in angiogenesis scores may be caused by relative protein expression levels rather than a larger pool of angiogenic proteins. We identified five proteins that were more than twofold downregulated in VEGF + GFM compared to VEGF alone: ANGPTL4, ITGA5, PTGIS, PTGS2, and WNT5A. Given these proteins' interaction, as a highly conserved proangiogenic network, the results indicate that a factor or factors in the GFM opposed VEGF on angiogenic protein synthesis. Unfortunately, our design does not allow us to identify factor(s) involved. Furthermore, the effects may be tissue and development stage specific. For example, IL-1α and TNF-α reportedly increased WNT5A in keratinocytes,³⁸ but in adipocytes, IL-1α upregulation is associated with downregulation of ANGPTL4.³⁹ Further experiments where the effect of FGF, IL-1, IL-6, PDGF, and TNF-α is separately investigated are required.

Finally, when knowledge gained from growing PDC spheroids on Matrigel was applied to PPF scaffolds, we noted a substantial bone-like growth upon COD + VEGF incubation (Fig. 4D).

Recently, five separate mechanisms of bone formation have been recognized⁴⁰: (i) endochondral ossification, in which MSCs develop separate chondrogenic and osteogenic lineages, (ii) intramembranous ossification, in which osteoblast development occurs in the absence of a chondrocyte template, (iii) chondrocyte to osteogenic precursor, in which immature chondrocytes give rise to osteogenic precursor cells, (iv) chondrocyte de-differentiation into precursor bone cells that proceed to differentiate into bone cells, and (v) direct transdifferentiation of chondrocytes into osteoblasts. Further development of these models complete with identification of biomarkers and pathways involved is greatly anticipated.

In summary, our results show that human PDCs can expand in cell culture irrespective of age and sex. Furthermore, PDCs form spheroids with higher osteogenic and angiogenic scores on Matrigel after supplementation with GFM. GFM supplementation also increased new bone formation on PPF scaffolds. Finally, beginning with a chondrocytic medium to mimic callus formation and then switching to osteogenic media did not augment bone formation or angiogenesis. These observations will catalyze future research into bone substitutes for the repair of critical craniofacial defects.

Footnotes

Acknowledgment

The authors wish to acknowledge departmental funding support for C.M.P.

Disclosure Statement

No competing financial interests exist.

Funding Information

No external funding was received.

Supplementary Material

References

Zuk

P.A.

Tissue engineering craniofacial defects with adult stem cells? Are we ready yet? Pediatr Res, 63, 478, 2008.

Chim

, Salgado

C.J.

, Mardini

, and Chen

H.C.

Reconstruction of mandibular defects. Semin Plast Surg, 24, 188, 2010.

Hundepool

A.C.

, Dumans

A.G.

, Hofer

S.O.

, et al. Rehabilitation after mandibular reconstruction with fibula free-flap: clinical outcome and quality of life assessment. Int J Oral Maxillofac Surg, 37, 1009, 2008.

San-Marina

, Sharma

, Voss

S.G.

, Janus

J.R.

, and Hamilton

G.S.

, 3rd. Assessment of scaffolding properties for chondrogenic differentiation of adipose-derived mesenchymal stem cells in nasal reconstruction. JAMA Facial Plast Surg, 19, 108, 2017.

Liu

, Miller

A.L.

, 2nd, Waletzki

B.E.

, Yaszemski

M.J.

, and Lu

Novel biodegradable poly(propylene fumarate)-co-poly(l-lactic acid) porous scaffolds fabricated by phase separation for tissue engineering applications. RSC Adv, 5, 21301, 2015.

Parry

J.A.

, Olthof

M.G.

, Shogren

K.L.

, et al. Three-dimension-printed porous poly(propylene fumarate) scaffolds with delayed rhBMP-2 release for anterior cruciate ligament graft fixation. Tissue Eng Part A, 23, 359, 2017.

Rohban

, and Pieber

T.R.

Mesenchymal stem and progenitor cells in regeneration: tissue specificity and regenerative potential. Stem Cells Int, 2017, 5173732, 2017.

Chang

, and Knothe Tate

M.L.

Concise review: the periosteum: tapping into a reservoir of clinically useful progenitor cells. Stem Cells Transl Med, 1, 480, 2012.

Einhorn

T.A.

, and Gerstenfeld

L.C.

Fracture healing: mechanisms and interventions. Nat Rev Rheumatol, 11, 45, 2015.

10.

Hyzy

S.L.

, Kajan

, Wilson

D.S.

, et al. Inhibition of angiogenesis impairs bone healing in an in vivo murine rapid resynostosis model. J Biomed Mater Res A, 105, 2742, 2017.

11.

Raines

A.L.

, Berger

M.B.

, Patel

, Hyzy

S.L.

, Boyan

B.D.

, and Schwartz

VEGF-A regulates angiogenesis during osseointegration of Ti implants via paracrine/autocrine regulation of osteoblast response to hierarchical microstructure of the surface. J Biomed Mater Res A, 107, 423, 2019.

12.

Raica

, and Cimpean

A.M.

Platelet-derived growth factor (PDGF)/PDGF receptors (PDGFR) axis as target for antitumor and antiangiogenic therapy. Pharmaceuticals (Basel), 3, 572, 2010.

13.

Randi

A.M.

, and Laffan

M.A.

Von Willebrand factor and angiogenesis: basic and applied issues. J Thromb Haemost, 15, 13, 2017.

14.

Zanetta

, Marcus

S.G.

, Vasile

, et al. Expression of von Willebrand factor, an endothelial cell marker, is up-regulated by angiogenesis factors: a potential method for objective assessment of tumor angiogenesis. Int J Cancer, 85, 281, 2000.

15.

T.N.

, Kasper

F.K.

, and Mikos

A.G.

Strategies for controlled delivery of growth factors and cells for bone regeneration. Adv Drug Deliv Rev, 64, 1292, 2012.

16.

, Pompili

V.J.

, and Das

Neovascularization and hematopoietic stem cells. Cell Biochem Biophys, 67, 235, 2013.

17.

Osta

, Benedetti

, and Miossec

Classical and paradoxical effects of TNF-alpha on bone homeostasis. Front Immunol, 5, 48, 2014.

18.

Ravi

, Paramesh

, Kaviya

S.R.

, Anuradha

, and Solomon

F.D.

3D cell culture systems: advantages and applications. J Cell Physiol, 230, 16, 2015.

19.

Al Habyan

, Kalos

, Szymborski

, and McCaffrey

Multicellular detachment generates metastatic spheroids during intra-abdominal dissemination in epithelial ovarian cancer. Oncogene, 37, 5127, 2018.

20.

Gao

, Yourick

J.J.

, and Sprando

R.L.

Comparative transcriptomic analysis of endothelial progenitor cells derived from umbilical cord blood and adult peripheral blood: implications for the generation of induced pluripotent stem cells. Stem Cell Res, 25, 202, 2017.

21.

Varghese

, Bukhari

A.B.

, Malhotra

, and De

IHC profiler: an open source plugin for the quantitative evaluation and automated scoring of immunohistochemistry images of human tissue samples. PLoS One, 9, e96801, 2014.

22.

McDowell

E.M.

, and Trump

B.F.

Histologic fixatives suitable for diagnostic light and electron microscopy. Arch Pathol Lab Med, 100, 405, 1976.

23.

Voss

, San-Marina

, Oldenburg

M.S.

, et al. Histone variants as stem cell biomarkers for long-term injection medialization laryngoplasty. Laryngoscope, 128, E402, 2018.

24.

Gonneaud

, Jones

, Turgeon

, et al. A SILAC-based method for quantitative proteomic analysis of intestinal organoids. Sci Rep, 6, 38195, 2016.

25.

Janus

J.R.

, Voss

S.G.

, Madden

B.J.

, et al. Time-course mass spectrometry data of adipose mesenchymal stem cells acquiring chondrogenic phenotype. BMC Res Notes, 11, 666, 2018.

26.

Omenn

G.S.

, States

D.J.

, Adamski

, et al. Overview of the HUPO plasma proteome project: results from the pilot phase with 35 collaborating laboratories and multiple analytical groups, generating a core dataset of 3020 proteins and a publicly-available database. Proteomics, 5, 3226, 2005.

27.

Street

, Bao

, deGuzman

, et al. Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. Proc Natl Acad Sci U S A, 99, 9656, 2002.

28.

Ong

S.E.

, and Mann

Mass spectrometry-based proteomics turns quantitative. Nat Chem Biol, 1, 252, 2005.

29.

Kuhn

, von Mering

, Campillos

, Jensen

L.J.

, and Bork

STITCH: interaction networks of chemicals and proteins. Nucleic Acids Res, 36(Database issue), D684, 2008.

30.

Hughes

C.S.

, Postovit

L.M.

, and Lajoie

G.A.

Matrigel: a complex protein mixture required for optimal growth of cell culture. Proteomics, 10, 1886, 2010.

31.

Yamaguchi

, Ohno

, Sato

, Kido

, and Fukushima

Mesenchymal stem cell spheroids exhibit enhanced in-vitro and in-vivo osteoregenerative potential. BMC Biotechnol, 14, 105, 2014.

32.

Vukicevic

, Kleinman

H.K.

, Luyten

F.P.

, Roberts

A.B.

, Roche

N.S.

, and Reddi

A.H.

Identification of multiple active growth factors in basement membrane Matrigel suggests caution in interpretation of cellular activity related to extracellular matrix components. Exp Cell Res, 202, 1, 1992.

33.

Lind

Growth factor stimulation of bone healing. Effects on osteoblasts, osteomies, and implants fixation. Acta Orthop Scand Suppl, 283, 2, 1998.

34.

, Boucher

, Koh

, et al. PDGF, TGF-beta, and FGF signaling is important for differentiation and growth of mesenchymal stem cells (MSCs): transcriptional profiling can identify markers and signaling pathways important in differentiation of MSCs into adipogenic, chondrogenic, and osteogenic lineages. Blood, 112, 295, 2008.

35.

Lin

, Fateh

, Salem

D.M.

, and Intini

Periosteum: biology and applications in craniofacial bone regeneration. J Dent Res, 93, 109, 2014.

36.

Starke

R.D.

, Ferraro

, Paschalaki

K.E.

, et al. Endothelial von Willebrand factor regulates angiogenesis. Blood, 117, 1071, 2011.

37.

Ishihara

, Ishihara

, Starke

R.D.

, et al. The heparin binding domain of von Willebrand factor binds to growth factors and promotes angiogenesis in wound healing. Blood, 133, 2559, 2019.

38.

Kim

J.E.

, Bang

S.H.

, Choi

J.H.

, et al. Interaction of Wnt5a with notch1 is critical for the pathogenesis of psoriasis. Ann Dermatol, 28, 45, 2016.

39.

Quintero

, Gonzalez-Muniesa

, Garcia-Diaz

D.F.

, and Martinez

J.A.

Effects of hyperoxia exposure on metabolic markers and gene expression in 3T3–L1 adipocytes. J Physiol Biochem, 68, 663, 2012.

40.

Aghajanian

, and Mohan

The art of building bone: emerging role of chondrocyte-to-osteoblast transdifferentiation in endochondral ossification. Bone Res, 6, 19, 2018.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.20 MB

0.02 MB

0.03 MB

0.04 MB

0.00 MB