Generally, we focus on strategies from sample collection to organoid construction. According to the current situation in which no research has reported cell line construction, various digestion systems applied in tissue handling and adjustment in NICH organoids are essential for the efficiency and success rate. To construct the NICH organoid, we derived cells from tissue with diverse cell-derived systems. Initially, diverse collagenase subtypes and trypsinase with different periods were applied to separate NICH single cells. Then, the growth medium system of NICH organoids was based on multiple tissue organoid and tumor organoid induction methods16,17. Primarily, several collagenases reported in other organoid construction methods 18-20 were applied in NICH cell derivation (I, II, VI XI). However, only a single organoid was finally harvested (collagenase XI) (Figure 2A).
Based on the very limited NICH organoid number in the first trial, new cell derivation methods were imperatively required to boost the cell derviation rate. Then, NICH cell number was significantly enhanced by a human Millet Cell Separation Kit (130-095-929, USA) with NICH organoid medium (Figure 2B, C).
Organoids faithfully recapitulate NICH at the morphological and histological levels
To compare NICH organoid tissue characteristics to their corresponding tissues, we performed H&E staining (Figure 3A) (organoids were fixed with Matrigel to ensure that the organoid and Matrigel showed antibody enrichment conditions reflecting the real growth microenvironment) and evaluated the expression of the main protein biomarkers for NICH (Figure 3B), including PECAM-1, factor VIII, and GLUT-1. Of note, NICH organoids consist of the transformed endothelial cells of a tumor but do not include immune, vessel or connective tissue ingredients. Histological depiction of the primary NICH tissue used for organoid derivation uncovered major morphology if the organoid was captured or not from tissue. To boost the success rate, we used a new strategy to analyze NICH organoids with BME fixed 21,22 to reduce organoid loss and maintain the micro-environment when organoids grew and developed. Specifically, BME displayed advantages in revealing the real environment, including key protein concentrations. The results of H&E showed high similarities in structure and cell staining between tissue and its derived organoid. The major discrepancy between NICH and IH is the biomarker GLUT-1, in which IH manifested as positive while NICH was negative or weakly positive. According to the results, NICH PDOs consistently illustrated tissue characteristics with GLUT-1 negativity, while both CD31 and Factor VIII were positive. This is a crucial histological demonstration in which we faithfully constructed a new model to uncover this mysterious hemangioma without any cell line or animal model.
Transcriptomic characterization of NICH organoids.
Three biopsies were collected and employed with the NICH organoid medium system for further gene expression data analysis. To assess organoid gene expression profiles, we performed RNA sequencing (RNA-seq) on 3 NICH-derived organoids (named group A), correlative NICH tissue (named group B) and para-NICH tissue (named group C). The PCA manifested a high similarity among the samples in groups A or B (Figure 4A). However, the sample of group C3 was not included due to quality control. The sample of C2 in the PCA graph was so close to B2, indicating that the sampling distance may be too close for these biopsies during surgery. Then, C2 was excluded.
PCA revealed that three NICH PDOs generated from distinct NICH tissues exhibited a high degree of similarity, which clearly suggested that the method for constructing NICH organoids is reliable and consistent. Then, we specifically focused on NICH PDOs, NICH tissue and para-NICH tissue comparisons (Figure 4B, C, D). The analysis of heatmaps revealed that genes linked to keratinization (KRTAP4-3, KRTAP4-6, KRTAP7-1, KRTAP1-4 KRTAP5-9, KRTAP17-1, KRTAP10-2, KRTAP10-1, KRTAP5-4) were considerably downregulated in groups A and B in comparison to group C. The volcano plot depicts gene up- and downregulation trend pairwise comparisons in the three groups. 621 genes up 978 down in (group An in contrast with C(Figure 4E) and 152 genes up and 872 genes down group B compared with C (Figure 4F).. However, when compared to group B, group A exhibited variation expression diversities (1962 up and 2572 down)(Figure 4G) .This may be due to the reality that organoids still lack of skin tissue, blood and lymph vessels, in vivo real ECM environment, immnue system, all of which may lead to discrepancy in transcriptomes result.
Gene functional analysis of NICH organoids.
KEGG signaling pathway enrichment analysis indicated gene enrichment and correlation connections. In comparison to para-NICH tissue, the KEGG signaling pathway enrichment analysis indicated that chronic myeloid leukemia, vascular smooth muscle contraction, ECM receptor, focal adhesion, colorectal cancer, and small lung cancer signaling pathways were enriched in NICH tissue. KEGG function results between NICH PDOs and NICH tumors showed only slight enrichment in p53 and RNA polymerase in the organoid group, while NICH tumor focal adhesion and vascular smooth muscle contraction were in accordance with NICH characteristics (Figure S2 A B).
Specifically, related functional signal pathway KEGG enrichment was compared in the three groups. Tumor-related signaling pathways, including PI3K/AKT, MAPK, and RAS, were significantly enriched in both NICH tissues and NICH organoids (Figure S2 C, D).
Further characteristics and drug validation of NICH organoids.
When NICH organoids were seeded in 96-well plates, we surprisingly found that many cells split from the surrounding organoids (Figure 6A). On day 10, the development was intrinsically tissue-like, with numerous organoid clones encircled by a complex microenvironment. This characteristic reveals that this rare vascular tumor may be very dynamic during its formative period.
To identify these cells, components of this mysterious microenvironment around organoids, immunocytochemistry was performed with two key biomarkers, VE-Cadherin in green and Vimentin in red. These antibodies effectively differentiate cell types, whether endothelial cells or fibroblasts, and there is a pathway for ECs to shift to fibrosis by endothelial-to-mesenchymal transition23. The results demonstrate that the cells derived from NICH organoids were positive for VE-Cadherin (Figure 5C) but negative for vimentin (Figure 5D), indicating that the main cells may be EC cells.
Propranolol, a nonselective beta-adrenergic receptor blocker, can accelerate the regression of proliferating IHs. However, the effect of propranolol is controversial in NICH therapy. Some research has claimed that propranolol is effective in the early stage of NICH, while other investigators have claimed that propranolol is invalid for NICH 24 Interestingly, in this study, the results indicated that propranolol (Figure 6A) did not have a significant therapeutic effect on NICH organoids even at high dosages (160 μg/L). Moreover, neither sirolimus (Figure 6B) nor trametinib (Figure 6C) at different dosages had any appreciable effect on NICH organoid models, which indicates that more organoids need further drug screening.