Cross Study Bulk RNA Sequencing Analysis of Vascular Associated States in Engineered Vascular Systems

Abstract Background: Vascularization does not necessarily indicate a mature vascular wall. Endothelial cells or vessel-like networks show that vascular-associated structures are present, whereas a stable, functionally mature wall must be assessed using multiple features, including mural-cell coverage, basement membrane, contractile machinery, and functional responses. Because vascularization strategies and outcome measures differ among engineered models, this study evaluates transcriptomic evidence across separate biological dimensions rather than treating vascularization as a single endpoint. Methods: We performed an exploratory secondary analysis of three publicly available human bulk RNA-sequencing datasets with biological replicates (38 biological libraries in total). The datasets covered vascularized mesenchymal organoids, vascular-network organoids cultured under static or flow conditions, and vascular smooth muscle cells (VSMCs) cultured under two-dimensional or three-dimensional conditions. The core analysis used 11 complementary gene modules to characterize endothelial identity, transcriptional features associated with pericyte-like mural cells, contractile maturity, synthetic remodeling, and flow- and barrier-related programs. Statistical inference used the biological library as the unit, and each dataset retained its original experimental design. CAMERA and expression-matched random-module analyses were applied to the two organoid datasets; Gene Ontology enrichment was limited to three primary comparisons. GSE192986 served as an independent VSMC state reference, with a sensitivity analysis excluding ACTA2. Results: In the day-14 organoid-versus-spheroid comparison, vascularized mesenchymal organoids showed higher scores in all seven vascular-associated modules (FDR < 0.05 for each); CAMERA supported all seven, although the comparison also differed in scaffold, assembly method, and vascularization conditions. At day 60, several vascular-associated transcriptional signals were lower and hypoxia-related transcription was higher. In a separate vascular-network organoid dataset, high flow reduced hypoxia and inflammatory activation, did not significantly increase endothelial or flow-response module scores, and was associated with a lower contractile-maturity score (FDR = 0.00350). In the independent VSMC dataset, three-dimensional culture increased hypoxia- and inflammation-related expression and reduced proliferation-related expression; contractile maturity did not meet the FDR-adjusted significance threshold. Conclusions: In these datasets, vascularization-related conditions, culture duration, three-dimensional culture, and flow corresponded to distinct transcriptomic states. Their effects on endothelial, mural, contractile, barrier, and stress-response features should be assessed separately. The findings support a transcriptomic framework for organizing evidence about engineered vascular systems, but they do not by themselves establish vessel-wall structure, perfusion, permeability, or vasomotor function.