1. Introduction
Cell transplantation has potential for diabetes treatment and tissue regeneration. However, clinical translation faces a major obstacle: the absence of a functional vascular network at the transplantation site [
1,
2]. Although subcutaneous interstitial spaces offer the advantages of ease of handling and fewer complications, their inherently low vascular density creates a hypoxic and nutrient-poor microenvironment for the grafted cells, resulting in massive cell death [
3]. To overcome this limitation, strategies using bioengineering techniques and biomaterial-mediated angiogenesis are essential for subcutaneous cell transplantation [
4]. A key approach involves the subcutaneous implantation of biomaterials before cell delivery. These materials elicit a host of foreign body responses that induce vascularized bed formation. Subsequently, the cells are transplanted into this preformed, vascularized subcutaneous space, a strategy termed subcutaneous pre-vascularization. In islet transplantation, this pre-vascularization strategy enhances oxygen and nutrient delivery, enabling transplanted islets to respond rapidly to glycemic changes [
3,
4].
Another critical limitation compromising subcutaneous graft viability is the host immune response, which elicits fibrotic encapsulation of transplanted cells. This physical barrier restricts the diffusion of oxygen and other nutrients [
5]. Although immunoprotective devices physically isolate grafts from immune cells, they may not adequately address the effect of soluble immune mediators on cellular function. Consequently, the development of advanced biomaterials with dual angiogenic and immunomodulatory capabilities is essential to achieve long-term subcutaneous graft survival [
6]. Innate immune cells, particularly macrophages and neutrophils, are pivotal regulators of angiogenesis [
3,
7]. Pro-regenerative macrophages secrete vascular endothelial growth factor (VEGF), interleukin (IL)-10, and transforming growth factor-β (TGF-β) to stimulate endothelial proliferation and pericyte recruitment, while neutrophils release matrix metalloproteinase-9 (MMP-9) to remodel the extracellular matrix (ECM), facilitating vascular invasion [
8]. Macrophages and neutrophils establish a pro-angiogenic feedback loop via cytokine crosstalk (e.g., neutrophil-derived tumor necrosis factor-α (TNF-α) activating macrophages), amplifying neovascularization [
9]. Strategically designed subcutaneously implanted biomaterials can harness these immune mechanisms to generate mature, stable vascular networks capable of sustaining perfusion. This approach simultaneously constructs a microenvironment conducive to transplantation.
In this study, we implemented subcutaneous pre-vascularization by optimizing the material selection and geometry. Polyvinyl chloride (PVC) substrates were implanted to elicit a controlled innate immune response by selectively recruiting macrophages and neutrophils. This intervention initiates an angiogenic cascade by moderating inflammation, promoting macrophage polarization toward mannose receptor-positive (CD206
+) phenotypes, and effectively inhibiting fibrosis, while enhancing vessel maturation [
10]. The resulting microenvironment, enriched with IL-10 and TGF-β, supported the development of structurally mature and functionally stable vascular networks. This immune-driven pre-vascularized niche successfully supports the structural integration of vascular organoids (VOs) and long-term functional maintenance of pancreatic islets, providing a clinically translatable engineered platform for cell-based therapies.
2. Methods and materials
2.1. Subcutaneous pre-vascularized site establishment
All animal surgeries were performed at the Department of Medicine, Peking University, China, and the experimental protocols were approved by the Ethics Committee of the Department of Medicine, Peking University (BCJI0257) in compliance with relevant ethical regulations. Male BALB/c nude and C57BL/6J mice (10-week-old, n = 3 per group) were obtained from the university animal facility. Animals were maintained under standard conditions (24 °C, 12 h light/dark cycles) with ad libitum access to food and water.
PVC, fluorinated ethylene propylene (FEP), and polytetrafluoroethylene (PTFE) tubes (derived from medical-grade indwelling needles) were sterilized by immersion in 75% ethanol (15 min), then transferred to saline, and followed by ultraviolet (UV) irradiation (30 min). PVC tubes had an inner diameter (I.D.) of (1.0 ± 0.1) mm and outer diameter (O.D.) of (2.0 ± 0.1) mm; FEP tubes had an I.D. of (0.6 ± 0.1) mm and O.D. of (0.9 ± 0.1) mm; PTFE tubes had an I.D. of (0.9 ± 0.1) mm and O.D. of (1.9 ± 0.1) mm. Mice were anesthetized via continuous isoflurane inhalation (3%-4% induction, 1%-2% maintenance) and positioned supine. The thoracic region was shaved and disinfected with povidone-iodine solution. A 2-mm skin incision was made and subcutaneous pockets were created via blunt dissection using ophthalmic forceps. Sterilized tubes were implanted into the pockets using 5-0 nylon sutures before wound closure.
2.2. Histological analysis
Following graft retrieval at designated time points, tissue specimens comprising the PVC/FEP tubes and surrounding skin were harvested, flattened, and photographed under a stereomicroscope to evaluate vascularization. Skin tissue was fixed in paraformaldehyde, dehydrated, and embedded in paraffin. Serial sections of 3-μm thickness were prepared for hematoxylin and eosin (H&E) staining (G1100; Solarbio, China), Masson’s trichrome staining (G1346; Solarbio), immunofluorescence (IF), and immunohistochemistry (IHC). IF was used to label human or murine vascular endothelial cells (mouse CD31 (mCD31) and human CD31 (hCD31)), vascular smooth muscle cells (α-smooth muscle actin (α-SMA)), pancreatic β-cells (insulin (INS)), macrophages (fusion 4/80 (F4/80) and CD206), and neutrophils (lymphocyte antigen 6G (Ly-6G)). After deparaffinization, rehydration, and antigen retrieval, slices were blocked with 10% donkey serum in phosphate-buffered saline (PBS; 30 min, room temperature (RT)). Primary antibodies diluted in blocking serum were applied: mouse anti-mCD31 (1:200; 3528; CST, USA), goat anti-hCD31 (1:200; AF806; R&D, USA), rabbit anti-α-SMA (1:200; ab5694; Abcam, UK), mouse anti-α-SMA (1:200; sc-32251; Santa Cruz, USA), rabbit anti-INS (1:100; A19066; Abclonal, China), rat anti-F4/80 (1:400; 123110; BioLegend, USA), rabbit anti-CD206 (1:200; ab64693; Abcam), and rat anti-Ly-6G (1:400; 127605; BioLegend), with incubation for 30 min at RT. Sections were rinsed three times in PBS (3 min per wash) before the application of species-matched secondary antibodies conjugated to Alexa Fluor dyes (1:400 in blocking serum, 30 min, RT). After three additional PBS washes (3 min each), the sections were mounted with 4′,6-diamidino-2-phenylindole (DAPI)-containing antifade medium (ab104139; Abcam) and cover-slipped. Fluorescent images were obtained using a Leica TCS SP8 system with Leica Application Suite X software (Leica Microsystems, Germany). Vessel diameter and density were quantified from maximum-intensity projections of CD31+ structures. ImageJ was used to calculate the proportion of F4/80, Ly-6G, and CD206+ cells in fluorescent images, vessel density in H&E-stained images, and collagen layer thickness in Masson’s trichrome-stained images.
IHC was used to detect IL-6, IL-10, TNF-α, and TGF-β expression. Paraffin sections were deparaffinized, rehydrated, and subjected to heat-mediated antigen retrieval. Endogenous peroxidase activity was blocked by incubation for 30 min with 3% hydrogen peroxide at RT, followed by three 3 min PBS washes. Non-specific binding was suppressed by incubation with 10% goat serum for 30 min. Sections were incubated with rabbit primary antibodies anti-IL-6 (1:100; A27935; Abclonal), IL-10 (1:100; A2171; Abclonal), TNF-α (1:100; A11534; Abclonal), and TGF-β (1:100; A25313; Abclonal) diluted in blocking serum for 30 min at RT. After three 3 min PBS washes, species-matched horseradish peroxidase (HRP)-conjugated secondary antibodies (1:400 in blocking serum) were applied for 60 min at RT. Following three additional PBS washes (3 min each), chromogenic development was initiated with 1% 3-3′ diaminobenzidine (DAB) solution under microscopic monitoring and terminated by rinsing with water for 20 min. Counterstaining was performed using hematoxylin (1 min) with subsequent 5 min water rinsing, immersion in 1% acid alcohol (5 s), and 60 min bluing under running water. Sections were dehydrated using graded ethanol, cleared in xylene, and mounted with a neutral resin under coverslips. Whole-slide images were acquired after complete drying and cytokine-positive areas were quantified using ImageJ threshold-based analysis.
2.3. VOs differentiation and transplantation
Induced pluripotent stem cell 18 (iPSC18) were cultured in NcTarget hPSC medium (RP01020; Shownin Biotechnologies, China) on Matrigel-coated Petri dishes (354277; Corning, USA), and passaged using TrypLE Select (12604013; Thermo Fisher, USA). Transfected iPSCs were maintained in NcTarget medium supplemented with 5 μmol·L−1 Y-27632 (T1870; TargetMol, USA) for 24 h post-transfection, followed by culture in Y-27632-free NcTarget medium with daily medium changes. To induce mesoderm progenitor cells, iPSCs were seeded on Matrigel-coated plates at 2 × 104 cells·cm−2 in LaSR medium containing 8 μmol·L−1 CHIR99021 (T2310; TargetMol). LaSR medium comprised advanced Dulbecco’s modified Eagle’s medium and Ham’s F-12 medium (DMEM/F12; 03.2001C; Eallbio, China) supplemented with 1× GlutaMAX (35050061; Thermo Fisher), 60 μg·mL−1 L-ascorbic acid (A8960; Sigma-Aldrich, USA), and 1% penicillin/streptomycin. After 24 h, the medium was replaced with fresh CHIR99021-containing LaSR medium for three days with daily changes. Mesoderm progenitor cells were then dissociated with TrypLE Select (37 °C, 1 min) and replated at 6.25 × 104 cells·cm−2 in differentiation medium LaSR supplemented with 30 ng·mL−1 VEGF-A (11066-HNAH; Sino Biological, China) and 30 ng·mL−1 fibroblast growth factor 2 (FGF2;10014-HNAE; Sino Biological) for three days with daily medium changes. iPSC18 lines carrying the CD31-tdTomato and α-SMA-green fluorescent protein (GFP) reporters were used for differentiation. Confocal microscopy revealed spherical VOs (100-200 μm diameter) composed of tdTomato+ endothelial and GFP+ smooth muscle cells at the differentiation endpoint. These organoids were used for subsequent flow characterization and transplantation. For all transplantation experiments, the “subcutaneous (SubQ) group” refers to the control group where cells were transplanted into an unmodified subcutaneous site. We resuspended 5000 VOs in 50 μL of gelatin solution (Collagen I:Matrigel = 3:1) and administered them via subcutaneous injection into pre-vascularized sites in mice.
2.4. RNA extraction and bulk RNA sequencing
Fresh tissues were snap-frozen in liquid nitrogen and total RNA was isolated using an RNA Extraction Kit (AC0202-B; Shandong Sparkjade Biotechnology, China). Lysis buffer (600 μL) was added to each sample, followed by thorough homogenization. The homogenate was centrifuged at 12 000g for 5 min and supernatant was transferred to a DNA adsorption column. After centrifugation at 12 000g for 1 min, 600 μL of 70% ethanol was added to the flow-through, mixed, and applied to an RNA adsorption column (centrifugation: 12 000g, 1 min). The column was treated with 700 μL deproteinization solution (12 000g, 1 min), followed by two washes with 500 μL rinse buffer (12 000g, 1 min per wash). Finally, RNA was eluted with 60 μL of RNase-free ddH2O (5 min RT incubation; 12 000g, 2 min). For RNA sequencing, the qualified RNA samples were used for library construction. Reads were aligned to mm39 using HISAT2 v.2.2.1, and gene expression was quantified using feature counts. DESeq2 was used to identify differentially expressed genes (|log2(fold change)| > 1, adjusted p < 0.05). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed using ClusterProfiler v.4.14.6; heatmaps were generated using pheatmap v.1.0.12.
2.5. Islet isolation and purification
Male Sprague-Dawley rats (8 weeks old) were euthanized under isoflurane anesthesia. After laparotomy, 10 mL of collagenase P solution (1 U·mL−1; 11213873001; Sigma-Aldrich) was retrograde-infused into the pancreas via the common bile duct. Excised pancreata were digested at 37 °C for 25 min and digestion was quenched by adding ice-cold Hanks’ balanced salt solution (HBSS; L201-500; BaiDi Biotechnology, China) containing 10% fetal bovine serum (FBS; 3023A; Umedium, China). The tissues were vigorously shaken for 15 s and filtered through a 450-μm sieve. After three washes with HBSS, the pellets were resuspended in a lymphocyte separation medium (LSM; 10 mL per rat; 25-072-CV; Corning). HBSS was layered carefully over the LSM suspension and centrifuged at 1000g (4 °C, 20 min). The intermediate layer containing the islets was collected, washed twice with HBSS, and resuspended. The islets were handpicked under a stereomicroscope to ensure > 90% purity. Purified islets were cultured overnight in Roswell Park Memorial Institute (RPMI) 1640 medium (PM150110; Procell, China) supplemented with 10% FBS and 1% penicillin/streptomycin (37 °C, 5% CO2).
2.6. Diabetic model induction and islet transplantation
One week before transplantation, the nude mice were fasted for 24 h (water ad libitum) and bedding was completely changed. Fasted mice received a single intraperitoneal injection of streptozotocin (STZ;160 mg·kg−1; 18883-66-4; AbMole, USA) dissolved in freshly prepared acetate-phosphate buffer (pH 4.5) before use. Diabetes status was confirmed when random blood glucose exceeded 16.7 mmol·L−1 over two consecutive days. For transplantation, 350 IEQ rat islets (overnight-cultured) were resuspended in 50 μL of gelatin solution (Collagen I:Matrigel = 3:1) and injected subcutaneously into pre-vascularized sites.
2.7. Intraperitoneal glucose tolerance test (IPGTT)
On post-transplantation day 50, the mice were fasted for 8 h (water ad libitum) with bedding replacement. After fasting, 2 g·kg−1 glucose in saline was administered intraperitoneally. Blood glucose levels were measured at 0, 15, 30, 60, 90, and 120 min post-injection. The glucose area under the curve (AUC) was calculated using the trapezoidal rule.
2.8. Flow cytometry analysis
On day 3 of differentiation, VOs were dissociated using 0.025% trypsin-ethylenediaminetetraacetic acid (EDTA; 37 °C, 3 min). The digestion was terminated by adding 5 mL of DMEM/F12, followed by centrifugation (300g, 1 min). After the cell pellets were washed with PBS, cells were analyzed on a CytoFlex flow cytometer (Beckman, USA). Singlets were gated using forward scatter-height (FSC-H)/forward scatter-width (FSC-W) parameters before assessing CD31+ and α-SMA+ populations.
2.9. Angiogenesis assay
VOs from dual-reporter iPSCs (CD31-tdTomato/α-SMA-GFP) were embedded in three dimensional (3D) hydrogels for live imaging. A base hydrogel layer was formed by polymerizing 100 μL of Collagen I and Matrigel (211251; NEST Biotechnology, China) in confocal dishes (37 °C, 3 h), and 500 VOs suspended in 100 μL of gel solution were overlaid onto the base and repolymerized. After adding 2 mL of differentiation medium with 15% FBS (FBS210500; Jet Biofil, China), angiogenic sprouts were imaged by confocal microscopy (Leica Microsystems) on days 1, 3, and 7 post-encapsulation using Z-stacks for 3D reconstruction.
2.10. Bioluminescence imaging tracing
For non-invasive in vivo tracing, iPSC18s were transduced with a vector expressing firefly luciferase and tdTomato under the control of the CAG promoter. The VOs were subcutaneously transplanted into a mixture of Collagen I and Matrigel (3:1). Mice were anesthetized via isoflurane inhalation and intraperitoneally injected with D-luciferin (D12505; Beijing LABLEAD, China). Bioluminescence imaging was performed to assess cell survival and localization using the IVIS Spectrum System (PerkinElmer, USA). Signals were acquired and analyzed using Living Image software.
2.11. Statistical analysis
All statistical analyses were performed using GraphPad Prism software, version 9.0. Data are presented as mean ± standard error of the mean (SEM). Normality was assessed using the D′Agostino-Pearson omnibus and Shapiro-Wilk tests. For comparisons between two groups, unpaired Student’s t-tests (equal variance), Welch’s t-tests (unequal variance), or Mann-Whitney U tests (non-normal data) were applied. For three or more groups, one-way analysis of variance (ANOVA) with Tukey’s post-hoc test (normally distributed data with homogeneous variance) or Kruskal-Wallis test with Bonferroni correction (non-normal data) was used. Statistical significance was set at p < 0.05.
3. Results
3.1. PVC tube grafts promote subcutaneous vessel formation
Subcutaneously implanted PVC tubes provide a vascularized microenvironment that enhances cell engraftment and survival. To evaluate material-dependent effects on angiogenesis, FEP, PTFE, and PVC tubes were implanted subcutaneously in mice for 14 days (
Fig. 1(a)). PVC tubes induced significantly greater neovascularization than that of FEP or PTFE tubes (Fig. S1 in Appendix A), with new vessels forming both around and within the lumen. This enhanced vascularization was observed in immunocompetent C57BL/6J mice and maintained in immunodeficient nude and nonobese diabetic (NOD)/severe combined immunodeficient (SCID)/interleukin 2 receptor gamma chain (IL-2Rγ) (null) (NSG) mice, whereas FEP tubes exhibited minimal angiogenesis and were consequently designated as the negative control group. Subsequent evaluation of tube length revealed that 5-mm PVC tubes generated superior pre-vascularization versus that of 10-mm PVC tubes (Fig. S2 in Appendix A). Histological quantification demonstrated a significantly higher vessel density in 5-mm PVC grafts than that in 10-mm PVC or FEP tubes across C57BL/6J, nude, and NSG mice (Fig. S3 in Appendix A), establishing 5-mm PVC and FEP tubes for subsequent experiments (
Fig. 1(b)).
After a 14-day subcutaneous implantation in C57BL/6J and nude mice, 5-mm PVC tubes exhibited robust neovascularization, as confirmed by stereomicroscopy (
Fig. 1(c)), H&E staining, and vessel density quantification (
Figs. 1(d)-(g)). Masson’s trichrome staining further revealed extensive collagen deposition and fibrotic encapsulation around FEP tubes (
Fig. 1(h)), with significantly thicker collagen layers than in the PVC implants (
Fig. 1(i)), indicating impaired nutrient diffusion [
11]. IF analysis confirmed mature vascular structures in PVC grafts, featuring α-SMA
+ muscle cells surrounding mCD31
+ endothelial cells. In contrast, FEP grafts contained sparse, small-diameter vessels lacking endothelial organization (
Fig. 1(j)).
To validate these findings, RNA sequencing was performed on subcutaneous tissues from nude mice harvested 14 days after PVC or FEP tube implantation along with untreated skin. Gene set enrichment analysis (GSEA) comparing PVC grafts to the untreated control group revealed significant enrichment of pro-angiogenic pathways, vascular development, and neovascularization alongside upregulated collagen catabolism and ECM remodeling pathways (
Fig. 1(k)). Similarly, PVC versus FEP comparisons demonstrated comparable activation of vascular neogenesis and ECM degradation pathways (
Fig. 1(l)), although some signatures differed from those observed in the control comparisons.
3.2. PVC tubes recruit immune cells to enhance subcutaneous vascularization
IF analysis revealed a significantly higher microvessel density in the PVC grafts than that in the FEP tubes (
Figs. 2(a)-(c)). Although there was no significant difference in the average vascular diameter, we observed mature vascular structures with a maximum diameter exceeding 80 μm in the PVC group (Fig. S4 in Appendix A), while all vessels in the FEP group were relatively small (< 40 μm).
Neutrophils and macrophages cooperatively regulated angiogenesis in the subcutaneous niche [
12], [
13], [
14]. As primary responders to hypoxia or inflammation, neutrophils initiate angiogenesis through TNF-α-mediated macrophage activation and polarization toward CD206
+ reparative phenotypes. Activated macrophages secrete neutrophil-specific chemokines, amplifying neutrophil recruitment and establishing a self-sustaining pro-angiogenic loop. IF analysis confirmed robust neovascularization-immune cell colocalization in PVC grafts, with nascent vessels consistently accompanied by F4/80
+ macrophages and Ly-6G
+ neutrophils (
Figs. 2(a) and
(b)). Quantitative assessment revealed significantly increased macrophage and neutrophil densities in the PVC versus those in the FEP groups (
Figs. 2(d) and
(e)). Furthermore, PVC grafts exhibited increased numbers of CD206
+ reparative macrophages (
Figs. 2(f) and
(g)), demonstrating enhanced pro-angiogenic immunomodulation.
Principal component analysis (PCA) of RNA sequencing revealed different transcriptional profiles separating PVC grafts from FEP and untreated control groups, with FEP samples clustering near the control tissues (
Fig. 2(h)). GSEA comparing PVC to control grafts demonstrated significant upregulation of inflammatory response pathways, macrophage/neutrophil activation signatures, IL-10/TGF-β signaling, and tissue remodeling pathways indicative of macrophage reparative polarization. Comparative PVC versus FEP analysis corroborated the IF observations, showing enrichment in acute inflammation, myeloid cell activation, and macrophage phenotypic switching pathways (
Fig. 2(i)). The differentially expressed genes included regulators of angiogenesis, ECM remodeling, and cytokine signaling (
Fig. 2(j)). Notably, significant upregulation of
Mrc1 (a CD206
+ macrophage marker) validated the enhanced reparative responses and neovascularization in PVC grafts.
3.3. Implanted PVC tubes elicit pro-angiogenic factors
When implants come in contact with host blood, exudates, or other bodily fluids, an immediate inflammatory response is triggered. Host cells involved in wound healing release cytokines, chemokines, reactive oxygen species, and other enzymatic products that recruit tissue-resident macrophages and undifferentiated monocytes to the injury site. Once activated, macrophages secrete a range of signaling molecules (e.g., IL-1, IL-6, IL-10, IL-12, TNF-α, and TGF-β) that attract fibroblasts; these fibroblasts then secrete collagen during cell proliferation and neovascularization. To explore the expression patterns of the relevant cytokines, we conducted immunohistochemical analyses of subcutaneous tissues 14 days after transplantation (
Fig. 3(a)). The expression levels of IL-6, IL-10, TNF-α, and TGF-β in the PVC group were significantly higher than those in the FEP group (
Fig. 3(b)). These cytokines mediate neovascularization. Specifically, IL-6 and IL-10 can initiate inflammatory responses and promote angiogenesis by inducing pro-angiogenic factors, such as VEGF. In addition, TNF-α activates extracellular regulated kinase-specific pathways in fibroblasts, resulting in increased TGF-β expression.
Consistent with these observations, GSEA of bulk RNA sequencing data indicated that cytokine-related pathways were significantly upregulated in the PVC group compared with both in the FEP and control groups, with positive regulatory trends in IL-1, IL-6, IL-10, TNF-α, and TGF-β (
Fig. 3(c)). KEGG and GO analyses of the upregulated genes showed enrichment in the PI3K-Akt pathway, ECM-receptor interaction, and MAPK pathway, all of which are closely associated with angiogenesis and tissue remodeling. Moreover, the upregulation of the Toll-like receptor pathway, efferocytosis, complement and coagulation cascades, neutrophil extracellular trap formation, as well as TNF and TGF-β pathways, suggested increased inflammation, activation of innate immunity, and enhanced cytokine secretion (
Figs. 3(d) and
(e)).
3.4. Successful engraftment of VOs in engineered subcutaneous sites
Although subcutaneous sites offer convenient manipulation and a low complication risk [
15,
16], their inherently low vascular density can lead to hypoxia-induced necrosis of transplanted cells, limiting clinical application [
1,
17]. Developing strategies for vascular reconstruction and neovascularization is crucial [
18] for poorly perfused regions. Establishing functional blood supply and inducing angiogenesis are essential for successful engraftment and long-term tissue survival [
3,
7].
To assess the efficacy of PVC tube-modified subcutaneous sites, iPSC-derived VOs constructs were orthotopically transplanted subcutaneously (
Figs. 4(a) and
(b)). For
in vitro visualization of VOs formation and functional verification, we generated a dual-reporter cell line expressing CD31-dTomato and α-SMA-GFP, which enabled the rapid morphological assessment of vascular networks in 3D cultures [
19,
20]. Upon completion of VOs differentiation, confocal microscopy revealed dTomato-labeled endothelial cells surrounded by GFP-labeled smooth muscle cells, forming spherical structures with diameters ranging from 100 to 200 μm (
Fig. 4(c)). We quantified the percentage of endothelial and smooth muscle cells in the VOs via flow cytometry (
Fig. 4(d)), embedded the dual-reporter VOs in a mixed gel solution (Collagen I:Matrigel = 3:1), and monitored their growth in 3D culture using confocal microscopy. In angiogenesis assays, endothelial cells were observed to extend outward first, with smooth muscle cells subsequently migrating along endothelial trajectories, consistent with physiological vascular development (
Fig. 4(e)).
Subsequently, we constructed an iPSC18 line overexpressing dTomato and luciferase, which were differentiated into VOs and transplanted subcutaneously, and observed the colonization and survival of the cells using bioluminescence imaging experiments. Fourteen days after subcutaneous transplantation of PVC tubes, we injected VOs and mixed gel solution into the pre-vascularized site (
Fig. 4(f)). VOs survived for more than 14 days in the pre-vascularized subcutaneous PVC tubes, whereas VOs in the unmodified subcutaneous interstitial space died in large quantities at 7-day post-transplantation, suggesting that the vascularization strategy of PVC tubes could effectively ameliorate the poor cellular hypoxic and ischemic subcutaneous survival environment (
Fig. 4(g)). Robust neovascularization was observed in the PVC group via stereoscopy and H&E staining at 14-day post-VO injection (
Fig. 4(h)), demonstrating functional vessel formation. In contrast, VOs injection alone failed to establish a perfused vasculature. IF images revealed VO-derived vessels through sequential detection of α-SMA
+ vascular smooth muscle and hCD31
+ endothelial cells. Magnified views confirmed canonical vessel architecture, with smooth muscle cells circumferentially aligned around the endothelial layers (
Fig. 4(i)).
3.5. Islet transplantation in pre-vascularized subcutaneous site reverses diabetes
Islet isolation damages native vasculature and ECM. However, the pancreatic islets rely on a functional vascular network to receive oxygen and nutrients, regulate INS secretion, and remove metabolic waste [
21,
22]. Without a viable vascular supply, islets experience hypoxia and nutrient deficiency, which rapidly compromise their survival and function [
23,
24]. Our previous experiments showed that subcutaneous PVC tube implantation induced angiogenesis, creating a vascularized site within 14 days. A two-step approach was used to assess their potential as islet transplantation sites. First, diabetes was induced in mice by intraperitoneal (i.p.) STZ injection two weeks after subcutaneous tube insertion. Subsequently, 350 islet equivalents (IEQs) of rat islets were transplanted into the prevascularized subcutaneous area (
Fig. 5(a)). A 63-day post-transplantation long-term analysis confirmed successful islet engraftment. Stereomicroscopy revealed that the encapsulated islet gel in PVC tubes was encircled by a dense, perfused vascular network (
Fig. 5(b)). Serial blood glucose monitoring demonstrated that transplantation of rat islets into the PVC-vascularized subcutaneous space reversed diabetes. In the PVC group, two mice normalized blood glucose (< 11.1 mmol·L
−1) by day 24 post-transplantation, with the remaining mice achieving an 89% recovery rate by day 52. In contrast, the FEP and SubQ group of mice sustained hyperglycemia (> 11.1 mmol·L
−1), developed polydipsia, polyphagia, polyuria, and weight loss, and exhibited higher mortality rates (
Figs. 5(c)-(e), Fig. S5 in Appendix A). On day 52 post-transplantation, IPGTT was performed to assess glucose homeostasis. The PVC group demonstrated significantly decreased post-glucose blood glucose fluctuations and improved glycemic regulation, as reflected by the substantially lower AUC calculated from IPGTT results compared with that of the other groups (
Fig. 5(f)). Histological analysis of islet-transplanted tissues using H&E and Masson’s trichrome staining revealed extensive neovascular infiltration into the gel matrix. Erythrocyte-containing vessels were observed around and within the islet grafts, confirming successful revascularization (
Fig. 5(g)). Immunofluorescent staining for β cells (INS
+) and smooth muscle cells (α-SMA
+) further showed a well-formed vascular network adjacent to the transplanted islets, consistent with other histological findings (
Fig. 5(h)).
4. Discussion
The material and length of subcutaneously implanted tubes are critical factors that influence vascularization. In an experiment comparing tubes of identical lengths but different materials (Fig. S1), the PVC tubes demonstrated significantly greater efficacy in promoting vascularization than that of FEP and PTFE tubes. When PVC and PTFE tubes had comparable diameters (approximately (1.0 ± 0.1) mm), PVC consistently outperformed PTFE, underscoring the pivotal role of material selection in vascularization outcomes. Regarding tube length, a comparative analysis of C57BL/6J, nude, and NSG mice revealed that 5-mm PVC tubes exhibited a significantly higher vessel density than their 10-mm counterparts (Figs. S2 and S3). Therefore, the shorter 5-mm PVC tubes may promote vascular migration toward the center of the tube lumen by reducing diffusion distance constraints.
Subcutaneous cell transplantation is challenging due to poor oxygenation and blood supply. Previous research has shown that silicone or nylon catheter implantation in mice triggered neovascularization via host immune responses, thereby benefiting islet transplantation [
25]. Our study revealed that PVC tubes can form functional vascular networks subcutaneously through autoimmune reactions, thereby solving nutrient delivery issues in cell transplantation. Notably, this process remains effective in immunodeficient nude and NSG mice, proving that adaptive immunity (T/B cells) is not required and the innate immune system is sufficient for vascularization. During the foreign body response, neutrophils and macrophages, which are crucial innate immune cells, are quickly recruited to the implantation site.
This study demonstrated that PVC implants activate the macrophage-neutrophil axis. Ly-6G
+ neutrophils infiltrate rapidly post-implantation, initiating neovascularization. Macrophages polarize to the reparative phenotype (CD206
+) and upregulate TNF-α, IL-6, IL-10, and TGF-β. This coordinated response organizes α-SMA
+ cells around mCD31
+ endothelial cells, forming mature blood vessels > 80 μm in diameter. Mechanistically, neutrophils secrete pro-angiogenic factors, such as VEGF and MMP-9, which play important roles in initiating angiogenesis [
12]. Macrophage-expressing hypoxia-inducible factors further enhance this process. These cells form a bidirectional activation loop; pro-inflammatory mediators amplify the foreign body response, whereas anti-inflammatory signals maintain immunological balance. This regulatory network creates the ideal inflammatory environment for vascular development [
13,
26]. Neutrophils and macrophages release pro-inflammatory cytokines, such as TNF-α and IL-6, to potentiate the foreign body reaction, which engage in a reciprocal activation and regulatory feedback loop. However, macrophages also exhibit immunomodulatory functions by secreting IL-10 and TGF-β [
27]. This bidirectional communication between neutrophils and macrophages orchestrates the dynamic equilibrium of the inflammatory cascade, culminating in the establishment of a permissive microenvironment conducive to neovascularization.
Adequate oxygen and nutrient supplies are critical for maintaining the functionality of cellular grafts, particularly during the early post-transplantation phase [
28]. For instance, the development of a functional vascular network in pancreatic islet grafts typically requires 7-14 days [
29]. Consequently, the islets remain in an avascular environment during the initial post-transplantation period, which is detrimental to their survival [
30]. This avascularity triggers multiple cellular pathways leading to islet apoptosis, a challenge that can be addressed through subcutaneous prevascularization strategies implemented before cell transplantation [
17,
31]. Herein, we validated the functional efficacy of PVC by transplanting VOs and pancreatic islets into subcutaneous prevascularized sites. PVC-constructed vascularized subcutaneous niches significantly enhanced the survival of the transplanted cells. Specifically, VOs survived for up to 14 days at pre-vascularized sites compared with only 7 days in the control group. Notably, VOs-derived blood vessels (hCD31
+) formed structural integration with the host vasculature, indicating functional anastomosis. With islet transplantation, an 89% diabetes reversal rate was achieved, accompanied by normal glucose tolerance. Histological analysis confirmed that blood vessels infiltrated deep into the islet grafts, supporting β-cell INS secretion. In contrast, FEP tubes induced severe fibrosis (
Figs. 1(h) and
(i)) and vascular attenuation, resulting in graft failure. This stark contrast underscores the critical importance of a precisely balanced immune response, as excessive inflammation or fibrosis disrupts vascular maturation.
The subcutaneous pre-vascularization strategy proposed in this study offers a novel approach for the cellular therapy of diabetes and demonstrates broad application prospects in the fields of tissue engineering and regenerative medicine. Our PVC-based pre-vascularization strategy provides two major advantages. First, it revolutionized islet transplantation by offering a safer alternative. Unlike conventional intrahepatic transplantation, which has risks associated with portal vein infusion, this method uses a PVC-generated site for in situ vascular network regeneration, converting the procedure into a minimally invasive local injection and reducing perioperative risks. Second, its versatile design supports the transplantation of various oxygen-sensitive cells, such as dopaminergic neurons for Parkinson’s disease and hepatocytes in liver failure. By pre-establishing a vascularized microenvironment, this strategy significantly enhances the survival rate and functional integration efficiency of the transplanted cells.
With the rapid development of biomaterials, the properties of pre-vascularization materials can be further optimized. Our SEM results indicated that PVC, FEP, and PTFE showed no significant differences in surface morphology or roughness, suggesting that their differing vascularization effects likely originated from the material’s chemical properties, such as surface energy, functional group distribution, or hydrophilic-hydrophobic balance, rather than the physical structure. Notably, the high chemical inertness and hydrophobicity of FEP may lead to insufficient protein adsorption and poor host integration, thereby eliciting a weak and ineffective immune response that fails to initiate an effective angiogenic program. In contrast, PVC induces a more moderate and balanced inflammatory response, promoting immune cell synergy and formation of functional vascular networks. Therefore, future research should focus on the chemical attributes of these materials and their regulatory mechanisms in the immune microenvironment in vivo. For example, the development of biodegradable polymer materials, such as poly lactic-co-glycolic acid (PLGA) or gelatin hydrogels, can enable automatic degradation after vascular induction, avoiding the risk of long-term foreign body retention. Alternatively, surface functionalization modifications, such as grafting VEGF, stromal cell-derived factor-1α (SDF-1α), or other chemokines, could precisely regulate immune cell recruitment and polarization behavior, enhancing the maturity and stability of vascular networks. Concurrently, minimally invasive implantation strategies for pre-vascularization devices, such as the use of endoscopes or interventional catheters for precise subcutaneous placement, should be explored to further improve clinical operability. This platform also has potential for integration with emerging technologies, such as using 3D printing to construct personalized vascular stents or combining gene editing technologies to endow transplanted cells with anti-inflammatory and pro-angiogenic properties, creating innovative “material-cell” combination therapy models. In the long term, the subcutaneous pre-vascularization strategy could not only serve as a universal platform for cell transplantation, but also act as a bridge for in vitro tissue construction and in vivo implantation, promoting the clinical translation of regenerative medicine.
Although the present study demonstrated the marked efficacy of the PVC tube-mediated subcutaneous pre-vascularization strategy in supporting transplanted cell survival and function, several limitations should be acknowledged. First, all experiments were conducted in small rodent models (mice), whose subcutaneous tissue thickness (0.5-1 mm) differed substantially from that of humans (typically > 5 mm), which limited the clinical translatability of our findings. Furthermore, although our study revealed the crucial roles of macrophages and neutrophils in angiogenesis, the causal necessity and detailed downstream signaling mechanisms have not been explicitly verified through cell-specific depletion or activation experiments. However, PVC, a nondegradable synthetic polymer, requires further systematic evaluation in large animal models to assess its long-term biocompatibility and potential fibrotic risks. Finally, the current strategy relies on the pre-implantation of a foreign body to initiate vascularization; this additional step may increase the complexity of the clinical application and reduce patient acceptance. Future studies should focus on developing biodegradable pre-vascularization materials and validating their safety and functional sustainability in large animal models.
PDF
(8042KB)
