Biofabrication of 3D-printed, pre-cross-linked alginate dialdehyde-gelatin (ADA-GEL) scaffolds for an in vivo metastatic arteriovenous loop tumor model

Abstract

Vascularized models mimicking tumor pathophysiology to investigate tumor characteristics are of high interest. The arteriovenous loop (AVL) model is an established method to vitalize bioengineered tissue grafts. In this model, an artificial vascular axis serves as the only connection between the living organism and the biomaterial. The objective of this study was to establish a three-dimensional (3D) printed, functional scaffold design for the AVL rodent model, in which human melanoma cells, derived from lymph node metastasis, are embedded in pre-cross-linked alginate dialdehyde-gelatin (ADA-GEL) and implanted in rats (N = 10) for 4 weeks. Bioink scaffolds were 3D-printed in two different shapes (n = 5), designed specifically for the AVL model's isolation chamber. Before implantation, the swelling behavior of the biofabricates was analyzed in vitro. The biocompatibility of the pre-cross-linked ADA-GEL and the impact of the scaffold-morphology were examined through macroscopic analysis and immunohistological stainings. The fluid uptake ratio of the hydrogel resulted in size extension, a finding which is highly relevant for the AVL model's closed system. Biofabricated scaffolds made of pre-cross-linked ADA-GEL remained stable in vivo and allowed for de novo fibrovascular tissue formation. The hypothesized biocompatibility of the analyzed hydrogel was confirmed. The two scaffold models exhibited differences regarding tumor growth and de novo fibrovascular tissue formation capacity. In both groups, metastatic cells were detected in the lymph nodes of rodents. The present study demonstrated that the AVL model is an excellent in vivo tool for melanoma research, combining biofabrication and vascularization with a high ability to replicate metastasis. At the same we conclude, that adapting the design of the biofabricated implants to the AVL model, depending specifically on the ink used, is of major importance.

Keywords: bioprinting; melanoma; metastasis; rat model; vascularization.

FIGURE 1

3D-printed bioink for AVL implantation: CAD models (designed using Tinkercad, Autodesk^©) of implanted constructs (A/D). 3D-printed implantation constructs using pre-cross-linked ADA–GEL after cross-linking and immersion in DMEM for 10 min (B/E). Schematic cross-sections of the implantation constructs: the construct for group A consisted of a 3D-printed, cell-laden bottom part and a cast, cell containing upper part (C); the construct for group B contained a 3D-printed acellular bottom part, a cellular cast middle layer, and an acellular cast upper layer (F). Scale bars = 1 cm.

FIGURE 2

Arteriovenous loop surgery: (A) operation field with exposed vessels and interposed venous graft (black arrow), (B) first end-to-end anastomosis (black arrow) between the artery and venous graft, (C) second end-to-end anastomosis (black arrow) between the venous graft and vein, (D) positioning the AVL on the 3D-printed half of the implantation construct; scale bars = 5 mm.

FIGURE 3

In vitro swelling assay of 3D-printed, pre-cross-linked ADA–GEL implantation scaffold constructs during 24 h; (A) calculation of the swelling ratio based on material weight extension of scaffold designs A and B (mean) and (B) calculation of the swelling ratio based on diameter extension of scaffold designs A and B (mean).

FIGURE 4

Macroscopic and microscopic appearance of tumor formation within the isolation chamber. ((A), left) Macroscopic appearance of the explant with tumor growth in group A; ((A), right) corresponding HE-stained histological cross-section. ((B), left) Macroscopic appearance of the explant with tumor growth in group B; ((B) right) corresponding HE-stained histological cross-section. Black arrows indicate the host vasculature of the AVL; scale bars = 500 µm.

FIGURE 5

HE-stained histological cross-sections of exemplary animals highlighting newly formed connective tissue areas for (A) group A and (B) group B; asterisks mark the remaining hydrogel, and dashed white lines frame the margins between newly formed tissue areas and hydrogel; scale bars = 200 μm. (C) Quantification of tissue ingrowth areas revealed a significant difference between constructs A and B (with the mean of four sections of biological replicates, n = 5) (*p ≤ 0.05, Mann–Whitney test).

FIGURE 6

HMB45-stained histological cross-sections of exemplary animals of (A) group A and (B) group B highlighting HMB45-positive tumor areas (brown color indicates HMB45-positive melanoma cells), where dashed black lines frame the margins between tumor areas and hydrogel, and asterisks mark the remaining hydrogel; scale bars = 200 μm. (C) Quantification of tumor growth areas in relation to total tissue showed significant differences between groups A and B (with the mean of 2 sections of biological replicates, n = 5) (*p ≤ 0.05, Mann–Whitney test).

FIGURE 7

Immunohistological visualization of the tumor microenvironment; (A) proliferation: Ki67-positive cells were brown-stained; scale bars = 20 μm. (B) Mesenchymal cells positive for vimentin are brown-stained; scale bars = 50 μm. (C) PAS histology of grown tumors; arrows indicate capillaries; scale bars = 50 µm. (D) Heterogeneous expression of SOX10, a melanoma plasticity marker, in tumor masses; scale bar = 50 μm (left: group A and right: group B).

FIGURE 8

CD68-stained histological cross-sections of representative animals showing macrophage-infiltrated areas of chamber explants (red color indicates CD68-positive cells) (A) group A and (B) group B; scale bars = 200 μm. (C) Quantification of CD68-positive immune cells per mm² did not differ significantly between the two experimental groups (with the mean of 2 sections of biological replicates, n = 5).

FIGURE 9

CD163-stained histological cross-sections of representative animals highlighting M2 macrophage-infiltrated areas of chamber explants (red color indicates CD163-positive cells) (A) group A and (B) group B; scale bars = 50 µm.

FIGURE 10

HE-stained histological cross-sections of representative animals of (A) group A and (B) group B, highlighting newly formed capillaries filled with Microfil^® (black arrows); scale bars = 200 µm.

FIGURE 11

α-smooth muscle actin-stained histological cross-sections of represented animals showing red-stained smooth muscle cells; (A) de novo-formed fibrovascular tissue area for group A, (B) de novo-formed fibrovascular tissue area for group B, (C) vascularized tumorous tissue for group A, and (D) vascularized tumorous tissue for group B; black arrows point exemplary blood vessels. Scale bars = 50 µm.

FIGURE 12

HMB45-stained histological cross-sections of representative animals of (A) group A and (B) group B, highlighting tissue areas of explanted lymph nodes infiltrated with metastatic melanoma cells (brown stained cells); scale bars = 200 μm. (C) Categorization of lymph node metastasis based on the presence of HMB45-positive cells, where category 0: no, category 1: few, category 2: moderate, and category 3: an abundant number of HMB45-positive cells were observed in explanted lymph nodes.