Fluid shear stress modulation of the CD44 and hyaluronic acid axis of the basal endothelial glycocalyx - Dataset

Physiological FSS (10 dynes/cm²) exposure resulted in a significant and sustained increase in HA across both HLMVEC surfaces while low (0.5 dynes/cm²) and high (30 dynes/cm²) rates of FSS displayed a transient increase in HA, returning to static levels with prolonged exposure (24 hours). CD44 was found to be highly enriched in the basal HLMVEC GCX compared to the apical GCX under all FSS conditions.

Disrupting apical HA or actin remodeling was found to disrupt the normal flow-adaptive phenotype, providing evidence for mechanical coupling without yet proving direct apical-to-basal HA dependence.</p><p dir="ltr"><b>Confocal Z-stack Data and Processing:</b></p><p dir="ltr">Hyaluronic acid (HA), CD44, F-actin (FA), and E-selectin (ES) data presented in this manuscript were captured and quantified from confocal z-stacks collected on a Leica STELLARIS 8 laser scanning confocal microscope using a plan-apochromat 63 x 1.40 NA oil objective with a z-stack slice thickness of 0.1 um for the HA and CD44 images and 0.2 um for the FA and ES images.

All images saved here are in their raw .tif format.</p><p dir="ltr">Each image was collected with nuclei signal in channel 1 with the secondary tag (HA, CD44, or FA) in channel 2. The ES data is an outlier here as ES data was collected through co-staining with HA signature as well. Therefore the ES images have 3 channels with channel 1 as the nuclei, channel 2 as HA, and channel 3 as ES signal.

HA channel was not removed out of transparency, while only the HA images in the HA folder were quantified and presented in the manuscript.</p><p dir="ltr">All confocal images are separated here by their secondary tag (HA, CD44, FA, and ES) then by their fluid shear stress exposure magnitude (static (0 dynes/cm2), 10 dynes/cm2, 0.5 dynes/cm2, and 30 dynes/cm2). A negative control folder is also presented for each folder (NegCTRL).

Within the static folder raw .tifs can be found along with the raw data output from the quantification pipelines also published here. Within the 0.5 dynes/cm2 and 30 dynes/cm2 folders nested folders separate the images and a raw data sheet by exposure time (0.5 hours, 12 hours, 24 hours). Within the 10 dynes/cm2 folders for each tag there are folders denoting the dosing condition (CTRL (no treatment), HYAL (hyaluronidase dosed), and Cyto-D (cytochalasin-D dosed) prior to exposure time folders.

Finally images within the final nested folder are named by their (exposure condition: CTRL vs HYAL, ect)_(shear rate: 10D (10 dynes/cm2), 0.5D, ect)_(exposure time: 0.5H, 12H, 24H)-(replicate number), for example cytochalasin-D dosed sample number 4 exposed to 10 dynes/cm2 of shear stress for 24 hours would be labeled CYTOD_10D_24H - 4.</p><p dir="ltr">The HA, CD44, and FA data presented here were separated into apical and basal sum projections using our previously published ImageJ macro that leverages the nuclei signal and channel to divide the stack into an apical and basal stack then produce a sum projection.

The DOI for this publication is: doi.org/10.1371/journal.pone.0339318 and the DOI for the data repository is: 10.5061/dryad.c59zw3rn1.</p><p dir="ltr">The CellProfiler pipelines used to quantify the HA and CD44 images are presented here titled "HA_AB_COV_II" and "CD44_AB_COV_II" respectively. In the HA and CD44 pipelines, pairs of apical and basal images must be uploaded. For the ES images sum projections of each raw z-stack were created and integrated intensity measured manually in ImageJ.

Within the ES folder there is one combined datasheet containing all raw integrated intensity measurements made with column titles corresponding to the folder names as previously explained.</p><p dir="ltr">The format within the raw data files in each of the final nested folders depend on the tag used. For the HA and CD44 data, the CellProfiler pipeline was configured to report the coverage or "AreaOccupied_AreaOccupied_Comb" and the integrated intensity Mean_Comb_XXX_Intensity_IntegratedIntensity" for both the apical (_GCX_A) and basal (_GCX_B) image output image for each original z-stack.

Only integrated intensity data was normalized around the apical static value for each tag (HA/CD44) and included in the publication.</p><p dir="ltr">The F-actin (FA) data was also split into apical and basal signal before processing. Marcotti et al.’s method (DOI:10.3389/fcomp.2021.745831) for order parameter quantification was leveraged here as a measure of FA alignment. Within the raw data files for FA data the apical (Apical_OrderParameter) and basal (Basal_OrderParameter) order parameter are reported.</p><p dir="ltr"><b>Cancer Cell Attachment Data and Quantification:</b></p><p dir="ltr">Within the MDA-MB-231_AttachmentImages folder all images and data corresponding to the cancer cell attachment data are presented.

There is one excel sheet with the total cancer cell attachment count for each replicate with column names corresponding to folder names as explained previously. For each cancer cell attachment data point or replicate, 3 images taken along the flow path were summed as explained in the methodology section, therefore each final folder contains at least 3 images, some with 6 as the corresponding bright field image is also reported.

The GFP tagged cancer cell channel and bright field image of the HLMVEC monolayer are differentiated by "_GFP" to denote the cancer cell channel or "_BF" for brightfield.</p>