Multimodal profiling reveals distinct endothelial activation pathways regulated by flow and heparan sulfate

This dataset contains the raw and processed quantitative outputs supporting the manuscript “Multimodal Profiling Reveals Distinct Endothelial Activation Pathways Regulated by Flow and Heparan Sulfate” (Harding, O’Hare, et al., 2025). The data quantify endothelial inflammatory and oxidative stress responses under four experimental conditions: (1) static, (2) uniform flow (12 dyn/cm²), (3) static + heparinase III, and (4) flow + heparinase III. The spreadsheets compile results from confocal immunofluorescence imaging, Western blot densitometry, reactive oxygen species (ROS) assays, and RNA sequencing. Quantified variables include mean fluorescence intensity (MFI) for protein markers (KLF2, ICAM-1, E-selectin, Nrf2, and Nf-κB p65), nuclear localization coefficients (Pearson’s coefficient for Nrf2 and DAPI), normalized band intensity ratios for Nox4, ICAM-1, and β-actin, and fluorescence-based ROS signals (H₂DCFDA and DHE assays). The RNA-seq dataset provides normalized counts, fold changes, and gene-set enrichment scores for oxidant, antioxidant, pro-inflammatory, and anti-inflammatory gene groups. Each data table corresponds to averaged biological replicates (n = 3–7 per condition) with normalization to static controls. Together, these data provide a quantitative framework linking glycocalyx integrity and flow-dependent mechanotransduction to endothelial inflammation and redox regulation. The files can be reused for comparative analyses of mechanosensitive gene programs, image quantification benchmarking, or integration with other omics datasets examining vascular dysfunction.

Methods
Overview

All numeric values included in the spreadsheets derive from image-based fluorescence analysis, Western blot densitometry, and RNA-seq differential-expression outputs. Each sample was classified by condition (Static, Flow, Static + Hep III, Flow + Hep III) and replicate number. No data were averaged or filtered beyond standard normalization steps, allowing full reuse for independent statistical analysis.

1. Image Quantification (Immunocytochemistry and ROS Assays)

Microscopy acquisition: Confocal (Zeiss LSM 710/800/880) images were collected using identical laser power, gain, and exposure settings within each experiment.

Mean Fluorescence Intensity (MFI): For each marker (KLF2, ICAM-1, E-selectin, Nrf2, Nf-κB), ImageJ/Fiji was used to select cell-level regions of interest. Background fluorescence was subtracted, and the integrated density was divided by area to yield MFI values.

Colocalization (Nrf2): Pearson’s correlation coefficient between Nrf2 and DAPI channels was calculated using the Coloc2 plug-in (ImageJ/Fiji) for ≥ 3 images per slide; averaged per sample to represent nuclear activation.

ROS quantification: For H₂DCFDA (total ROS), live cell imaging was obtained using a Zeiss AxioObserver widefield microscope equipped with an incubator set at 37°C. Fluorescence images were converted to grayscale and measured for mean intensity. Nine images per sample were averaged to produce one replicate data point.

Normalization: MFI values were normalized to the static mean (set = 1.0) to compare fold changes across conditions.

2. Western Blot Densitometry

Blots were imaged using Bio-Rad ChemiDoc. Band intensities for target proteins (ICAM-1 and Nox4) and β-actin were measured with ImageJ/Fiji.
Background intensity was subtracted for each lane.
Ratios of target / β-actin were calculated and normalized to the static control lane (set = 1.0).
Each value represents one biological replicate; group averages and standard deviations are listed in the spreadsheet.

3. RNA-seq Quantification and Gene-Set Scoring

Normalized read counts were generated using DESeq2 (v1.26.0).
Columns in the spreadsheet include: raw counts, baseMean, log₂ fold-change (Flow vs Static, Flow + Hep III vs Static + Hep III, etc.), Wald p-values, and adjusted p-values (Benjamini–Hochberg FDR).
Gene-set enrichment scores (normalized enrichment score = NES) for oxidant, antioxidant, pro-inflammatory, and anti-inflammatory categories were computed via GSEA v2.2.1.
Each gene was annotated by its directional contribution (up/down-regulated) based on the Wald statistic ≥ 1 or ≤ –1.