Skip to main content
NCBI home page
Research and Practice in Thrombosis and Haemostasis logoLink to Research and Practice in Thrombosis and Haemostasis
. 2025 May 21;9(4):102894. doi: 10.1016/j.rpth.2025.102894

Vascular-type heterogeneity is associated with differential gene expression profiles of endothelial cells under shear stress

Allaura A Cox 1, Christopher James Ng 1,∗
PMCID: PMC12268554  PMID: 40678357

Abstract

Background

Endothelial cell (EC) heterogeneity is an emerging area of research in EC biology. Vein, artery, and microvascular ECs have been shown to have heterogeneity in gene expression and function, termed “vascular-type” heterogeneity in this report. In addition to this innate heterogeneity, we hypothesized that different vascular-type ECs would also demonstrate heterogeneity in their response to shear stress.

Objectives

We interrogated whether vascular-type ECs would demonstrate variations in transcriptional expression patterns under shear stress.

Methods

Human umbilical vein ECs, human pulmonary arterial ECs, and human microvascular ECs were commercially purchased and subjected to 0, 1, 4, and 10 dynes/cm2 of laminar shear stress. After shear stress exposure, cellular alignment was analyzed, and RNA was extracted and evaluated via bulk RNA sequencing.

Results

All ECs demonstrated significant changes in alignment under shear stress. Shear stress significantly affected the transcriptomes of ECs, resulting in differential expression of genes and pathways. While several genes were differentially expressed by all 3 vascular EC types (44.2%), most differentially expressed genes were limited to 1 or 2 of the vascular types. Hemostatic and thrombotic genes were found to have differential expression patterns under conditions of shear stress, and von Willebrand factor demonstrated a pattern of vascular-type heterogeneity in response to shear stress.

Conclusion

Shear stress causes changes in cellular alignment and transcriptional patterns in ECs that are dependent upon underlying vascular type. Therefore, endothelial vascular-type heterogeneity can regulate response to shear stress, especially in hemostatic and thrombotic gene expression.

Keywords: endothelial cells, RNA, sequence analysis, von Willebrand factor

Essentials

  • •

    The effect of shear stress on gene expression in different types of endothelial cells (ECs) is unclear.

  • •

    We evaluated gene expression from different types of ECs.

  • •

    The magnitude of shear stress and type of ECs affected gene regulation.

  • •

    Interesting hemostatic and thrombotic gene expression patterns were identified.

1. Introduction

Endothelial cells (ECs) line the vascular system and play an important role in human health and disease. One of the unique aspects of ECs is their exposure to shear stress in vivo, and the fluid mechanics of endothelial biology suggest that the mechanobiology of fluid forces on ECs plays a large role in their regulation of basic biologic processes [1]. A wealth of data suggest that the application of shear stress has dramatic effects on EC cytoskeletal, transcriptional, and functional responses [1]. Responses to shear stress have been shown to be dependent on the type and magnitude of shear stress, and pathologic shear stress has been shown to be relevant in certain disease processes [[1], [2], [3]].

While historically thought of as a single, monoclonal cell type, recent reports highlight the multitude of transcriptomic, proteomic, and functional differences of different ECs, culminating in term “endothelial heterogeneity” [[4], [5], [6]]. While heterogeneity exists in terms of organ type, ie, brain vs cardiovascular, etc., a major determinant of endothelial heterogeneity is the “vascular type” of ECs, which can globally be considered to be arterial, venous, capillary, and lymphatic [4,6]. Even in these overarching groups, subtypes can exist, leading to further endothelial diversity, and there can be classification controversies. For example, microvascular ECs are sometimes considered to be capillary ECs and other times considered to be a separate subgroup.

There is clear evidence that shear stress affects hemostatic and thrombotic genes/proteins [[7], [8], [9], [10], [11], [12], [13]]. However, many of these previous reports have focused on specific individual genes or pathways and have not analyzed overall changes from the application of shear stress. Furthermore, these previous reports have not accounted for the concept of endothelial heterogeneity in the response of ECs to shear stress.

In this report, we sought to globally characterize the transcriptional response to shear stress in 3 different vascular-type ECs: human umbilical vein ECs (HUVECs), human pulmonary artery ECs (HPAECs), and human microvascular ECs (HMVECs). These 3 EC types were chosen to represent the different vascular types of ECs. We conducted bulk RNA sequencing (RNA-seq) analysis of these vascular-type ECs at different laminar shear stress magnitudes to determine global and specific transcriptional changes that may be reflective of the intersection of endothelial heterogeneity and laminar shear stress.

2. Methods

2.1. Cell lines

Primary HUVECs, HPAECs, and HMVECs were purchased from Lonza. All ECs used for experimentation were passaged no more than >7 times. ECs were sustained in EGM-2 MV Microvascular Endothelial Cell Growth Medium-2 BulletKit medium (Lonza) supplemented with 10% fetal bovine serum. Experiments were orchestrated to ensure similar passage across cell lines.

2.2. Cell culture under flow

Cells were harvested and reseeded into a μ-Slide I Luer (IBIDI 80166) at density of 60,000 cells/cm2 or 100,000 cells/cm2. After 24 hours, the cells were washed with phosphate-buffered saline, and a total of 120 μL of fresh medium was added to the cells. For control (static) samples, the slide was placed back in the incubator, and medium was changed every 24 hours until the conclusion of the experiment. For flow samples, end of each slide was connected to 100 cm of tubing: 0.5-mm diameter tubing was used to create 1 dyne/cm2 flow condition, 0.8-mm dimeter tubing to create 4 dynes/cm2 flow condition, and 1.6-mm diameter tubing to create 10 dynes/cm2 flow condition. One end of the tubing was then connected to a reservoir of medium that was kept inside the incubator with the slide, and the other end of the tubing was connected to a syringe outside of the incubator. The syringe was then attached to the Standard Infuse/Withdraw PHD ULTRA Syringe Pump (Harvard Apparatus 70-3007). The desired flow rate was input into the device and flow was started. Flow was applied to the cells for a total of 24 or 48 hours, and every 22 minutes, the direction of flow was changed from infuse to withdraw.

2.3. Cell orientation analysis

After the application of flow, 10 images of predetermined location on the slide were captured for each condition. Images were then analyzed in ImageJ using the Directionality Analysis Tool (Supplementary File 1). Flow direction during the withdrawal interval was defined as 0° or 180°. The orientation of the cells was estimated as the angle difference between the long axis of the cells and the direction of flow (Figure 1A, B).

Figure 1.

Figure 1

Open in a new tab

(A, B) Brightfield images of cell alignment analysis. 0° and 180° are parallel to the direction of flow while 90° is perpendicular to the direction of flow. (C–F) Histograms of cell alignment under static, 1 dyne/cm2, 4 dyne/cm2, and 10 dyne/cm2 conditions with angles grouped in cohorts every 30°. Statistical analysis was performed via analysis of variance with multiple condition correction. HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell. ∗P ≤ .05; ∗∗P ≤ .01; ∗∗∗P ≤ .001; ∗∗∗∗P ≤ .0001.

2.4. RNA-seq analysis

After relevant experimentation, RNA was extracted using miRNeasy Mini Kit (Qiagen 217084) after direct cellular lysis on the slide. RNA was quantified using the Qubit RNA BR Assay Kit (Invitrogen Q10210). The samples were sent for RNA-seq at the Genomics Core at the University of Colorado–Anschutz Medical Campus. Each sample had 10M reading pairs per 20M total reads. RNA-seq analysis was conducted using Salmon and edgeR-based quantification to generate raw counts and counts per million matrices, DESeq2 (see Supplementary File 2 for example analysis, full code available upon request) for differential analysis, and clusterProfiler for pathway analysis and visualization [14,15].

3. Results

3.1. Bidirectional laminar flow changes cellular orientation perpendicular to flow

Across all EC types, the application of flow appeared to increase the frequency of median angles (between 60° and 120°) perpendicular to the flow direction (Figure 1C vs D–F), and the increase in frequency of median angles appeared to correlate with the magnitude of shear stress. When comparing between different vascular-type ECs, there were differences in angles of HUVECs, HMVECs, and HPAECs in some conditions, suggesting that different vascular-type ECs respond to shear stress differently in terms of their cellular alignment, with HMVECs appearing to have the greatest magnitude of change in alignment (as compared with HUVECs and HPAECs). These data suggest that across all ECs tested, shear stress leads to cell alignment perpendicular to the direction of shear stress and that HMVECs demonstrate the most significant change in cell alignment.

3.2. Principal component analysis reveals transcriptional changes due to shear stress but minimal differences due to incubation time and seeding density

We analyzed whether differences in seeding density or shear stress exposure time would affect the overall transcriptional patterns of ECs. Seeding densities of 60,000 vs 100,000 cells/cm2 and shear stress exposure time of 24 hours vs 48 hours had little overall effect on the overall transcriptome of the ECs (Supplementary Figure 1). Therefore, for further analyses, samples seeded at both densities and exposure times were included.

We next evaluated global transcriptional differences by cell type and shear stress magnitude. There was a marked transcriptional change with the application of shear stress (highlighted by a rightward and downward direction on the principal component analysis [PCA] plot in Figure 2A). In addition, HMVECs appeared highly transcriptionally different than HPAECs and HUVECs, although all cell lines moved in the rightward/downward direction on the plot after application of shear stress (Figure 2A). Notably, we saw small differences in global transcription when comparing magnitudes of shear stress; eg, samples exposed to 10 dynes/cm2 (red samples) appeared to be “lower and to the right” as compared with samples exposed to 1 dyne/cm2 (dark blue) in the 2D PCA plot, although this difference was not striking.

Figure 2.

Figure 2

Open in a new tab

(A) RNA-seq PCA of all EC types. Dynes are demarcated by different colors and EC type by different shapes. (B) Volcano plot of differentially expressed genes. Log2fold change (FC) is expressed as the ratio of expression of shear stress samples relative to control samples. The horizontal dashed line demarcates P = 10e-20, and the vertical dashed line represents a FC cutoff of 1.5 or −1.5. (C) Venn diagram demonstrating DEGs when comparing shear stress vs static conditions in 3 EC types, HUVECs are shown in green, HPAECs in orange, and HMVECs in purple. (D) Venn diagram demonstrating DEGs when comparing different magnitudes of shear. DEGs identified in a 1 dynes/cm2 vs 0 dynes/cm2 analysis are shown in green, DEGs identified in a 4 dynes/cm2 vs 0 dynes/cm2 are shown in orange, and identified in a 10 dynes/cm2 vs 0 dynes/cm2 are shown in purple. DEG, differentially expressed gene; EC, endothelial cell; HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell; NS, not significant; PCA, principal component analysis; RNA-seq, RNA sequencing.

We next evaluated differentially expressed genes (DEGs) between static and shear stress conditions. As seen in Figure 2B, there were a high number of statistically significant DEGs between 2 conditions (a list of all DEGs is presented in Supplementary Table 1). As we found that there were transcriptional differences between different ECs, we next evaluated DEGs comparing static and shear stress conditions for each individual EC type. The Venn diagram in Figure 2C demonstrates the DEGs for each cell line. Across the 3 cell lines, 5265 (44.2% of the total DEGs in this analysis) were differentially expressed across all 3 EC types (see Supplementary Table 2). Genes such as KLF4 and NOS3 are included on this list, but interestingly, KLF2 was not.

Next, as the analyses in Figure 2B and C treated shear stress as a binary variable regardless of the magnitude of shear stress, we then identified DEGs under various magnitudes of shear stress (0 vs 1 dyne/cm2, 0 vs 4 dynes/cm2, and 0 vs 10 dynes/cm2). When comparing different magnitudes of shear stress, 52.6% of genes were shared, ie, responsive to any magnitude of shear stress, but as noted, different magnitudes of shear stress did cause unique genes to be up- or downregulated (Figure 2D).

3.3. Pathway analysis of DEGs identifies differentially expressed pathways

After identification of DEGs, we used them to identify differentially expressed pathways using canonical pathway analysis. Gene ontology biological processes (GO:BP) were identified when comparing static with shear stress in all EC types, and both activated and suppressed GO:BP pathways are displayed in dot-plot format in Figure 3A. In Figure 3B, the same GO:BP pathways are organized in a tree-plot format, demonstrating the similarities between different GO:BP pathways. The pathway analysis results suggest alterations in several global themes, such as “endocardial cushion development morphogenesis” and “immune other virus organism.” Some of the GO:BP pathways, such as “blood vessel development,” “circulatory system development,” and “vasculature development,” are activated in the shear stress condition rather than the static condition. In comparison, many of the immune-mediated GO:BP pathways, such as “defense response to other organism,” “innate immune response,” “response to other organism,” and “response to external biotic stimulus” are all suppressed in the shear stress condition.

Figure 3.

Figure 3

Open in a new tab

Pathway analysis of DEGs. All EC types are included in this analysis. (A) Gene ontology analysis demonstrates activated and suppressed GO:BP pathways when analyzing shear stress treated vs static ECs. The GeneRatio of the x-axis is defined as the number of core enrichment genes divided by the number of pathway genes. (B) A tree plot demonstrates the clustering of related GO:BP pathways implicated in the gene ontology analysis. In both panels, circle size denotes the number of genes and color shading represents the P value. DEG, differentially expressed gene; EC, endothelial cell; GO:BP gene ontology biological process.

3.4. Laminar flow alters endothelial gene expression

We next focused on specific gene sets that may be altered in the setting of shear stress. First, we investigated changes in endothelial-related genes (see Supplementary Table 3) when shear stress was applied to ECs. DEG analysis identified several endothelial-related genes that are primarily upregulated under conditions of shear stress (Figure 4A). In addition, known canonical shear stress genes such as KLF2, KLF4, and NOS3 (endothelial NOS) demonstrated a positive correlation with increasing shear stress (Figure 4B–D). Other genes are known to be associated with shear stress, including genes involved in cytoskeletal changes and relatively novel gene sets containing “atheroprone” and “atheroprotective” genes [1,[16], [17], [18]]. In a heatmap analysis of these genes, shear stress was associated with a general upregulation of “atheroprotective” and “shear” related genes and a relative downregulation of “cytoskeletal” and “atheroprone” genes (Figure 4E).

Figure 4.

Figure 4

Open in a new tab

(A) A custom gene set of endothelial genes (see Supplementary Table 3) was used for DEG analysis to focus on endothelial-related genes and is displayed as a volcano plot of DEGs. Log2 fold change (FC) is expressed as the ratio of expression of shear stress samples relative to control samples. The horizontal dashed line demarcates P = 10e-20, and the vertical dashed line represents a FC cutoff of 1.5 or −1.5. (B) Correlation plots and linear regression analysis examining the expression of known shear stress-dependent genes KLF2, KLF4, and NOS3 with shear stress demonstrate increased gene expression with increased shear stress across all EC types (HUVECs are shown in purple, HPAECs in green, and HMVECs in pink). Linear regression analysis was conducted and representative R2 values and P values are shown. (E) A custom set of endothelial genes organized in gene sets named “atheroprotective,” “atheroprone,” “cytoskeletal,” or “shear” was used to generate a heatmap analysis. DEG, differentially expressed gene; EC, endothelial cell; EndoMT, endothelial-to-mesenchymal transition; HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell; NS, not significant.

3.5. SMAD signaling pathway demonstrates changes in von Willebrand factor expression

As an example of an EC pathway that is known to be altered under shear stress, we analyzed the role of shear stress and endothelial heterogeneity in SMAD signaling [19]. Regarding SMAD pathway expression profiling, we demonstrated heterogeneous gene expression changes across different shear rates and different endothelial vascular types. For example, SMAD1/3/4/5 demonstrated upregulation in the static condition with relative pan-downregulation under any shear stress, while SMAD6/7/9 demonstrated significant upregulation with any shear stress (Supplementary Figure 2). Further analysis revealed that not only did shear stress affect SMAD pathway expression profiling, but that endothelial vascular type also contributed to heterogeneous expression. For example, HMVECs had high expression of SMAD1 under static conditions that decreased with shear stress, while HPAECs and HUVECs had relatively little change in SMAD1 expression under shear stress (Supplementary Figure 2). From a global perspective, the TGFβ pathway was highly upregulated under shear stress (Supplementary Figure 3).

3.6. Laminar shear stress induces angiogenic gene expression patterns

In addition to endothelial-related gene sets, which we curated manually, we also investigated whether shear stress would affect the expression of angiogenesis-related genes. By analyzing all expressed genes from the GO:BP GO:0001525 (termed “angiogenesis”) in a heatmap analysis, we determined that shear stress leads to significant alternations in the overall transcriptional landscape related to angiogenesis (Figure 5). Additionally, individual EC types demonstrated heterogeneity in their expression patterns of angiogenic genes under static and shear stress conditions (Figure 5). For example, HMVECs (pink columns in Figure 5) showed high expression of angiogenic genes (highlighted with a black box in the middle vertical portion of the heatmap), whereas HUVECs (purple columns) and HPAECs (green column) did not. While all 3 EC types were likely to downregulate these middle vertical genes under conditions of shear stress, HMVECs appeared to have the most significant downregulation. The top 10 most DEGs in HMVECs were PRKD2, ENG, HIF3A, JAM3, EPHB2, ADGRG1, NRP1, CCL2, and JAG1 (see Supplementary Table 4). As cellular proliferation/growth is an important aspect of angiogenesis, and cell cycle state has been shown to be affected by shear stress [20], we also evaluated the role of cell cycle signaling, as represented by cyclin dependent kinases, in our gene expression analysis. Overall, cyclin-related protein expression patterns were largely heterogenous, although a few families demonstrated some consistency; eg, CCN family genes were downregulated at higher shear stresses, and CCNE1/2 demonstrated relative upregulation with higher shear stress (Supplementary Figure 4). Furthermore, recent transcriptional analyses demonstrated a heavy influence of cell cycle genes; however, our analyses seemed to show a decreased emphasis of cell cycle-specific upregulation as compared with the previous data. There was a lack of upregulation of hallmark G2M checkpoints, although overarching proliferative pathways, such as TGFβ and WNT pathways, were upregulated (Supplementary Figure 3).

Figure 5.

Figure 5

Open in a new tab

Heatmap of angiogenic genes. Heatmap analysis of all genes associated with the gene ontology biological process GO:0001525 (angiogenesis) as of 2024 is displayed. A list of genes can be found at https://amigo.geneontology.org/amigo/term/GO:0001525. Overall, shear stress demonstrates a significant change in the expression of angiogenic genes. Similar to earlier heatmaps, there is some heterogeneity of up/downregulation of certain genes when comparing HMVECs vs HPAECs vs HUVECs. HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell.

3.7. Laminar shear stress induces changes in hemostatic genes

In addition to angiogenesis-related genes, we also investigated changes in hemostatic genes, with a particular focus on VWF. First, we analyzed changes in hemostatic genes via heatmap analysis. Heatmap analysis demonstrated shear stress-induced changes in hemostasis-related genes, with notable examples such as F2R, SERPINE1, ITGB3, and F3 (Figure 6A). To better visualize significant changes, a DEG analysis was conducted, and the statistically significant DEGs are displayed in a volcano plot (Figure 6B). As one of our focuses was on VWF, which appeared to be significantly downregulated under conditions of shear stress, we next generated a correlation plot of VWF expression with shear stress (Figure 6C). This analysis demonstrated an interesting finding: while all EC types demonstrated decreases in VWF expression with shear stress, HMVECs and HPAECs demonstrated a more significant decrease in VWF expression with increasing shear stress. To evaluate genes known to be associated with VWF changes in genome-wide association studies (GWASs), we conducted a heatmap analysis of genes identified in these GWASs (Figure 7) [21,22]. Some genes, such as STAB2 and KATA2, appeared to be positively correlated with VWF. To better quantify the relationship of von Willebrand factor (VWF)-associated GWAS genes with VWF expression, we conducted a Spearman correlation analysis according to EC vascular type. This demonstrated vascular-type EC heterogeneity in terms of VWF expression correlation with GWAS-related genes. Some genes, such as PXK and STXBP5, demonstrated correlation direction across all 3 vascular-type ECs while others, such as KAT2A, demonstrated differential positive/negative correlation based on vascular-type EC (Figure 6).

Figure 6.

Figure 6

Open in a new tab

Analysis of custom gene sets relating to hemostasis and von Willebrand factor (VWF). (A) A custom gene set of hemostatic endothelial genes was analyzed via heatmap analysis for transcriptional change associated with shear stress. Some genes, such as F2R and F3, appear to have significant changes under conditions of shear stress. (B) Volcano plot of statistically significant differentially expressed genes in the hemostatic gene set. (C) Correlation plot of VWF expression with shear stress magnitude. VWF appears to be downregulated under conditions of flow in 5B, and the correlation plot generated in 5C demonstrates an overall pattern of decreasing VWF expression with increasing shear stress. Linear regression analysis was conducted and representative R2 values and P values are shown. (D) Heatmap analysis of genes associated with VWF changes in GWASs. Based on the finding of decreased VWF expression with increasing shear stress, we evaluated a set of genes identified in VWF-related GWASs. EC, endothelial cell; FC, fold change; GWAS, genome-wide association study; HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell.

Figure 7.

Figure 7

Open in a new tab

Correlation coefficients of genome-wide association study-identified genes associated with von Willebrand factor levels. Spearman coefficients are displayed for respective vascular-type endothelial cells in the heatmap. The legend to the right displays the strength of the Spearman coefficient ranging from purple (low) to yellow (high). HMVEC, human microvascular endothelial cell; HPAEC, human pulmonary arterial endothelial cell; HUVEC, human umbilical vein endothelial cell.

4. Discussion

In this study, we analyzed the physiology and transcriptional response of different vascular-type ECs under conditions of shear stress.

We demonstrated that HUVECs, HMVECs, and HPAECs all change their cytoskeletal structure and cellular alignment under conditions of shear stress. However, there appears to be heterogeneity in the magnitude of cytoskeletal change based on the different vascular EC types. For example, HMVECs appear to demonstrate greater angle change than HUVECs and HPAECs. Interestingly, our data highlights a dichotomy in the field—at low shear stresses, cells appear to align perpendicular to the flow direction, while at higher shear stresses, cells align parallel to flow direction [23,24]. The ECs appeared to be elongated perpendicular to the direction of flow, a phenomenon that was seen across all vascular-type ECs. There are a few potential explanations for our findings. (1) The shear stress magnitudes may not have been sufficient to drive parallel alignment; however, the highest shear rate (10 dynes/cm2), demonstrated cell alignment in the direction of flow [[24], [25], [26]]. (2) The methodology we used in our system to achieve laminar flow, ie, a “back and forth” flow, could have caused the cells to align perpendicularly. (3) Finally, the cellular media we used contains 2 ng/mL of VEGF, and at that concentration range, Vion et al. [23] demonstrated enhanced perpendicular alignment of cells at low and high shear stress.

Some shear-stress–related cytoskeletal gene expression changes in our analysis are consistent with those of other reports [27], such as increases in CFL1 and GSN and decreases in ILK, VIM, LOXL2, and P4HA1. However, some genes showed opposite expression pattern changes as compared with Chu et al. [27]; eg, they found increased expression in CNN2, TGM2, and FLNA whereas we found these genes to be downregulated. Similarly, Chu et al. [27] found FBN1, IGTAV, and ITGAE to be downregulated, whereas we found increased expression.Click or tap here to enter text. As compared to other reports, we found VCL to be downregulated, whereas Dekker et al. [10] found slightly increased expression.

Regarding RNA-seq analysis, we demonstrated that different vascular-type ECs undergo significant changes to their transcriptome after the application of laminar shear stress. Importantly, we noted that canonical shear stress-responsive genes in ECs, such as KLF2, KLF4, and NOS3, all demonstrated increases in gene expression under conditions of flow, and the degree of upregulation was positively correlated with the magnitude of shear stress [[8], [9], [10],17,28]. These results suggest that our experiments are an appropriate model to evaluate the effects of shear stress on different vascular-type ECs. Furthermore, the overall transcriptional changes of the different vascular-type ECs in our analysis were congruent with emerging transcriptional gene sets of atheroprone and atheroprotective patterns, whereby the application of shear stress leads to upregulation of atheroprotective genes and downregulation of atheroprone genes [17]. Our data aligns with other previously published reports on the effects of shear stress on particular endothelial gene sets, such as SMAD signaling and cell cycle changes [19,20]. We found TGFβ signaling (involved in SMAD signaling) to be upregulated, similar to previous findings [19,20]. However, we did note some differences; eg, we did not find as significant of a change in cell cycle-related pathway gene expression as did previous analyses. We posit that this is likely because they focused on pulsatile vs laminar shear stress [20], whereas our analyses were limited to laminar shear stress.

Our comparative analysis of the statistically significant up- and downregulated genes across different vascular-type ECs demonstrates that a significant number are common to all 3 types of ECs. This would suggest that ECs as a whole share inherent pathways that are common in their response to shear stress, which may reflect conservation of cellular functions and upregulation of “high level” shear stress regulators, such as KLF4 and NOS3. Interestingly, KLF2 was not included as a “common” gene, but this is likely due to statistical cutoffs as KLF2 demonstrated significant upregulation with shear stress in our analyses. However, despite this shared response, the majority of DEGs (55.8%) were not shared among all ECs. There is likely a discrepancy between statistical and biological significance here; due to sample sizes, some genes may not be statistically significant but may strongly trend or have been previously implicated in shear stress-dependent regulation. However, we also note that we may be highlighting aspects of endothelial heterogeneity that have not been well evaluated previously, and therefore, further research into individual pathways and mechanisms are warranted to understand the role of endothelial heterogeneity in this context. These data suggest that vascular-type heterogeneity may play a significant role in the biological response of ECs to laminar flow and may have implications in the pathologic contributions of ECs under conditions of altered flow, such as in stenotic vessels or vascular malformations.

The analysis of specific up- and downregulated genes identified a significant number of DEGs, and pathway analysis suggests some broad “themes” of gene set changes, notably immunologic, chemotactic, and blood vessel/endocardial development. These genes set changes that are consistent with other analyses of the effects of shear stress on ECs. Maurya et al. [17] demonstrated that different types of flow, oscillatory and pulsatile flow, led to changes in “inflammation”-related genes such as SELE, VCAM, and IL8. In our analysis, these are grouped under the atheroprone and atheroprotective gene sets, but it seems likely that changes in inflammation-related genes would lead to gene set changes that are grouped under the “immunologic” heading of gene ontology analysis. Similarly, “blood vessel/endocardial development” GO:BP headings would seem to overlap with “angiogenic” themes, and as noted in Figure 5, the application of shear stress appears to lead to significant gene expression changes of angiogenic genes.

Regarding our focus on hemostatic/thrombotic genes, we saw significant changes relating to these genes under conditions of shear stress. Notably, this up/downregulation was not uniform across all endothelial types, suggesting that vascular-type heterogeneity may be an important variable in the hemostatic/thrombotic response to shear stress. Specifically, we saw similar directionality as in previously published reports in genes such as thrombomodulin (upregulation), F2R, also known as PAR1 (upregulation), and F3, also known as tissue factor (upregulation) [[10], [11], [12],29,30]. On the other hand, we observed some incongruent expression patterns in our data compared to earlier reports. For example, F8 was upregulated in our dataset but had minimal or decreased expression in Hough et al. [31]. SERPINE1 was downregulated in our dataset but upregulated in Bergh et al. [11]. PLAT was downregulated in Bergh et al. and TFPI was upregulated in Hough et al., and yet both were not differentially expressed in our dataset [11,31].

We further focused on the evaluation of VWF. Previous data have been conflicting in the context of shear stress and VWF regulation. Initial reports suggest that VWF release is increased under shear stress, but at the mRNA level, there were no significant changes in VWF mRNA content [32]. Other reports using a variety of different ECs demonstrated no significant changes to VWF mRNA when exposed to shear stress [12,33]. However, other studies using gene arrays, VWF-promoter plasmids, quantitative polymerase chain reaction in HUVECs, and single-cell RNA-seq demonstrated VWF increases [10,[34], [35], [36]]. The shear stress regulation of VWF may also be dependent on the shear stress magnitude as HUVECs and endothelial colony-forming cells demonstrated a decrease in VWF expression after exposure to arterial shear stress (10-70 dynes/cm2) but an increase in VWF expression under venous shear stress (1-6 dynes/cm2) [31,37]. Our results demonstrate an interesting finding; under conditions of shear stress, arterial and microvascular ECs demonstrate decreased VWF expression with increasing shear stress magnitude, while VWF expression in HUVECs is largely unchanged. To better understand drivers of this discrepancy in VWF expression under shear stress, our analysis of genes associated with VWF levels identified by GWASs [21,22] demonstrated heterogeneity in correlation depending on vascular-type EC. While some genes, such as TNPO1, demonstrated consistent correlational directionality (ie, all positively or negatively correlated), others, such as KAT2A, demonstrated positive correlation in HPAECs and HMVECs but negative correlation in HUVECs. This may suggest that shear stress modulates VWF differentially based on the vascular type of ECs, perhaps though histone or other epigenetic changes. One caveat of this analysis is that these GWAS-related genes were identified by association with VWF plasma levels, while our analysis focused on VWF expression at the mRNA level. Some of the GWAS-related genes are thought to be associated with VWF-related clearance (STAB2, STX2) or other mechanisms that would be unlikely to have a strong effect on VWF mRNA levels.

Strengths of our report are a comprehensive and a priori evaluation of gene expression in ECs under shear stress. We studied different vascular-type ECs and various shear types for a comparative analysis of both vascular type and shear stress magnitude. Our evaluation accounts for overall transcriptional changes as well as limited gene sets/themes to highlight changes. Limitations of our analysis are that we used “healthy” control ECs, and the extrapolation of this to diseased EC phenotypes may be limited. Furthermore, our analyses are limited to RNA-seq and imaging, and we do not have other functional analyses, which certainly would be a desired approach in future studies based on this work. We also used a single type of flow, laminar flow, in a “back and forth” mechanism that is not well represented in vivo. However, our data suggest that we observed similar up/downregulation of shear stress-related genes as other published reports. Finally, although we studied multiple shear stress magnitudes, we did not approach “high” arterial shear rates, such as 20 dynes/cm2, although the ranges of various arterial vs venous shear rates vary widely in the published literature.

In summary, we present a comparative analysis of transcriptional changes in different vascular-type ECs under shear stress. We demonstrate significant heterogeneity in the shear stress-related transcriptional responses of venous vs arterial vs microvascular ECs and highlight specific changes in hemostatic/thrombotic genes as well as in an EC-related gene, VWF.

Acknowledgments

Funding

This work was supported by grants/support from the Health Resources & Services Administration–Maternal and Child Health (5H30MC00008-20-00 to C.J.N.), the National Institutes of Health (P01HL44457 to C.J.N.), and the Children’s Hospital Foundation–Tanabe Bobrow Fund (C.J.N.). The funder provided support in the form of salaries for authors but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. There was no additional external funding received for this study.

Author contributions

A.A.C. made substantial contributions to concept and design, analysis and/or interpretation of data; critical writing or revising the intellectual content, and final approval of the version to be published. C.J.N. made substantial contributions to concept and design, analysis and/or interpretation of data; critical writing or revising the intellectual content, and final approval of the version to be published.

Relationship disclosure

C.J.N. has served as a consultant for Takeda and CSL Behring in the past. Takeda and CSL Behring did not have any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. A.A.C. has no conflicts of interest to disclose.

Data availability

All RNA-seq data has been deposited in GEO–accession number GSE294621.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used ChatGPT and/or Microsoft Copilot to review the text for readability and conciseness. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Footnotes

Handling Editor: Dr Robert A. Campbell

The online version contains supplementary material available at https://doi.org/10.1016/j.rpth.2025.102894

Supplementary material

Supplementary Fig S1.

Supplementary Fig S1

Open in a new tab

Principle Component Analysis (PCA) of ECs subjected to shear stress. Dots represent an individual sample. Seeding cells at 60,000 or 100,000 cells/cm2 did not have a significant effect on overall transcriptional differences (A). Similarly, shear stress exposure time of 24hrs or 48hrs did not have a significant effect on overall transcriptional differences (B).

Supplementary Fig S2.

Supplementary Fig S2

Open in a new tab

Heatmap analysis of SMAD related genes referenced in the gene ontology gene set GO:0060395.

Supplementary Fig S3.

Supplementary Fig S3

Open in a new tab

Pathway analysis of Molecular Signatures Database’s Hallmark pathways. Shear stress treated samples are compared to static samples. Blue denotes pathways that are downregulated under shear stress and red denotes pathways that are uregulated. The normalized enrichment score (NES) is shown at bottom, and the size of the circles represents the number of genes found in the pathway (count).

Supplementary Fig S4.

Supplementary Fig S4

Open in a new tab

Heatmap analysis of cyclin-related genes.

Supplementary File 1
mmc1.docx (13.5KB, docx)
Supplementary File 2
mmc2.zip (1.1KB, zip)
Supplementary Table 1

DESEQ2 output of list of differentially expressed genes of shear stress (any magnitude) vs static conditions.

mmc3.xlsx (1.1MB, xlsx)
Supplementary Table 2

A list of all genes found to be differentially expressed in HUVECs, HMVECs, and HPAECs exposed to shear stress, i.e. a gene that were universally differentially expressed regardless of cell type.

mmc4.xlsx (76.3KB, xlsx)
Supplementary Table 3

Curated list of custom endothelial gene sets based on previous reports. Some genes can be associated with multiple gene sets, denoted by a comma between respective gene sets.

mmc5.xlsx (12.2KB, xlsx)
Supplementary Table 4

DESEQ2 output of list of differentially expressed angiogenic genes (as defined by GO:0001525, see Figure 5) in HMVECs.

mmc6.xlsx (27.3KB, xlsx)

References

  • 1.Davis M.J., Earley S., Li Y.S., Chien S. Vascular mechanotransduction. Physiol Rev. 2023;103:1247–1421. doi: 10.1152/physrev.00053.2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Helmlinger G., Geiger R.V., Schreck S., Nerem R.M. Effects of pulsatile flow on cultured vascular endothelial cell morphology. J Biomech Eng. 1991;113:123–131. doi: 10.1115/1.2891226. [DOI] [PubMed] [Google Scholar]
  • 3.Guo D., Chien S., Shyy J.Y.J. Regulation of endothelial cell cycle by laminar versus oscillatory flow: distinct modes of interactions of AMP-activated protein kinase and Akt pathways. Circ Res. 2007;100:564–571. doi: 10.1161/01.RES.0000259561.23876.c5. [DOI] [PubMed] [Google Scholar]
  • 4.Pasut A., Becker L.M., Cuypers A., Carmeliet P. Endothelial cell plasticity at the single-cell level. Angiogenesis. 2021;24:311–326. doi: 10.1007/s10456-021-09797-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kalucka J., de Rooij L.P.M.H., Goveia J., Rohlenova K., Dumas S.J., Meta E., et al. Single-cell transcriptome atlas of murine endothelial cells. Cell. 2020;180:764–779.e20. doi: 10.1016/j.cell.2020.01.015. [DOI] [PubMed] [Google Scholar]
  • 6.Becker L.M., Pasut A., Cuypers A., Carmeliet P. In: The vasculome: from many, one. Galis Z.S., editor. Academic Press; New York: 2022. Endothelial cell heterogeneity and plasticity in health and disease—new insights from single-cell studies; pp. 17–28. [Google Scholar]
  • 7.Hough C., Cameron Cl, Notley C.R.P., Brown C., O’Brien L., Keightley A.M., et al. Influence of a GT repeat element on shear stress responsiveness of the VWF gene promoter. J Thromb Haemost. 2008;6:1183–1190. doi: 10.1111/j.1538-7836.2008.03011.x. [DOI] [PubMed] [Google Scholar]
  • 8.Dekker R.J., Boon R.A., Rondaij M.G., Kragt A., Volger O.L., Elderkamp Y.W., et al. KLF2 provokes a gene expression pattern that establishes functional quiescent differentiation of the endothelium. Blood. 2006;107:4354–4363. doi: 10.1182/blood-2005-08-3465. [DOI] [PubMed] [Google Scholar]
  • 9.Dekker R.J., van Thienen J.V., Rohlena J., de Jager S.C., Elderkamp Y.W., Seppen J., et al. Endothelial KLF2 links local arterial shear stress levels to the expression of vascular tone-regulating genes. Am J Pathol. 2005;167:609–618. doi: 10.1016/S0002-9440(10)63002-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Dekker R.J., van Soest S., Fontijn R.D., Salamanca S., de Groot P.G., VanBavel E., et al. Prolonged fluid shear stress induces a distinct set of endothelial cell genes, most specifically lung Krüppel-like factor (KLF2) Blood. 2002;100:1689–1698. doi: 10.1182/blood-2002-01-0046. [DOI] [PubMed] [Google Scholar]
  • 11.Bergh N., Larsson P., Ulfhammer E., Jern S. Effect of shear stress, statins and TNF-α on hemostatic genes in human endothelial cells. Biochem Biophys Res Commun. 2012;420:166–171. doi: 10.1016/j.bbrc.2012.02.136. [DOI] [PubMed] [Google Scholar]
  • 12.Ensley A.E., Nerem R.M., Anderson D.E.J., Hanson S.R., Hinds M.T. Fluid shear stress alters the hemostatic properties of endothelial outgrowth cells. Tissue Eng Part A. 2012;18:127–136. doi: 10.1089/ten.tea.2010.0290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Uzarski J.S., Scott E.W., McFetridge P.S. Adaptation of endothelial cells to physiologically-modeled, variable shear stress. PLOS One. 2013;8 doi: 10.1371/journal.pone.0057004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Yu G., Wang L.G., Han Y., He Q.Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–287. doi: 10.1089/omi.2011.0118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Malek A.M., Alper S.L., Izumo S. Hemodynamic shear stress and its role in atherosclerosis. JAMA. 1999;282:2035–2042. doi: 10.1001/jama.282.21.2035. [DOI] [PubMed] [Google Scholar]
  • 17.Maurya M.R., Gupta S., Li J.Y.S., Ajami N.E., Chen Z.B., Shyy J.Y.J., et al. Longitudinal shear stress response in human endothelial cells to atheroprone and atheroprotective conditions. Proc Natl Acad Sci U S A. 2021;118 doi: 10.1073/pnas.2023236118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Chiu J.J., Chien S. Effects of disturbed flow on vascular endothelium: pathophysiological basis and clinical perspectives. Physiol Rev. 2011;91:327–387. doi: 10.1152/physrev.00047.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Deng H., Min E., Baeyens N., Coon B.G., Hu R., Zhuang Z.W., et al. Activation of Smad2/3 signaling by low fluid shear stress mediates artery inward remodeling. Proc Natl Acad Sci U S A. 2021;118 doi: 10.1073/pnas.2105339118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Deng H., Rukhlenko O.S., Joshi D., Hu X., Junk P., Tuliakova A., et al. cSTAR analysis identifies endothelial cell cycle as a key regulator of flow-dependent artery remodeling. Sci Adv. 2025;11 doi: 10.1126/sciadv.ado9970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sabater–Lleal M., Huffman J.E., de Vries P.S., Marten J., Mastrangelo M.A., Song C., et al. Genome-wide association transethnic meta-analyses identifies novel associations regulating coagulation factor VIII and von Willebrand factor plasma levels. Circulation. 2019;139:620–635. doi: 10.1161/CIRCULATIONAHA.118.034532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Smith N.L., Chen M.H., Dehghan A., Strachan D.P., Basu S., Soranzo N., et al. Novel associations of multiple genetic loci with plasma levels of factor VII, factor VIII, and von Willebrand factor: the CHARGE (Cohorts for Heart and Aging Research in Genome Epidemiology) Consortium. Circulation. 2010;121:1382–1392. doi: 10.1161/CIRCULATIONAHA.109.869156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Vion A.C., Perovic T., Petit C., Hollfinger I., Bartels–Klein E., Frampton E., et al. Endothelial cell orientation and polarity are controlled by shear stress and VEGF through distinct signaling pathways. Front Physiol. 2021;11 doi: 10.3389/fphys.2020.623769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chandurkar M.K., Mittal N., Royer-Weeden S.P., Lehmann S.D., Michels E.B., Haarman S.E., et al. Transient low shear-stress preconditioning influences long-term endothelial traction and alignment under high shear flow. Am J Physiol Heart Circ Physiol. 2024;326:H1180–H1192. doi: 10.1152/ajpheart.00067.2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Baeyens N., Nicoli S., Coon B.G., Ross T.D., Van den Dries K., Han J., et al. Vascular remodeling is governed by a VEGFR3-dependent fluid shear stress set point. eLife. 2015;4 doi: 10.7554/eLife.04645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Choublier N., Taghi M., Menet M.C., Le Gall M., Bruce J., Chafey P., et al. Exposure of human cerebral microvascular endothelial cells hCMEC/D3 to laminar shear stress induces vascular protective responses. Fluids Barriers CNS. 2022;19:41. doi: 10.1186/s12987-022-00344-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Chu T.J., Peters D.G. Serial analysis of the vascular endothelial transcriptome under static and shear stress conditions. Physiol Genomics. 2008;34:185–192. doi: 10.1152/physiolgenomics.90201.2008. [DOI] [PubMed] [Google Scholar]
  • 28.Nayak L., Lin Z., Jain M.K. “Go with the flow”: how Krüppel-like factor 2 regulates the vasoprotective effects of shear stress. Antioxid Redox Signal. 2011;15:1449–1461. doi: 10.1089/ars.2010.3647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Peghaire C., Dufton N.P., Lang M., Salles-Crawley I.I., Ahnström J., Kalna V., et al. The transcription factor ERG regulates a low shear stress-induced anti-thrombotic pathway in the microvasculature. Nat Commun. 2019;10:5014. doi: 10.1038/s41467-019-12897-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Lin M.C., Almus–Jacobs F., Chen H.H., Parry G.C., Mackman N., Shyy J.Y., et al. Shear stress induction of the tissue factor gene. J Clin Invest. 1997;99:737–744. doi: 10.1172/JCI119219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hough C., Notley C., Mo A., Videl B., Lillicrap D. Heterogeneity and reciprocity of FVIII and VWF expression, and the response to shear stress in cultured human endothelial cells. J Thromb Haemost. 2022;20:2507–2518. doi: 10.1111/jth.15841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Galbusera M., Zoja C., Donadelli R., Paris S., Morigi M., Benigni A., et al. Fluid shear stress modulates von Willebrand factor release from human vascular endothelium. Blood. 1997;90:1558–1564. [PubMed] [Google Scholar]
  • 33.Egorova A.D., DeRuiter M.C., de Boer H.C., van de Pas S., Gittenberger-de Groot A.C., van Zonneveld A.J., et al. Endothelial colony-forming cells show a mature transcriptional response to shear stress. In Vitro Cell Dev Biol Anim. 2012;48:21–29. doi: 10.1007/s11626-011-9470-z. [DOI] [PubMed] [Google Scholar]
  • 34.Shoajei S., Tafazzoli–Shahdpour M., Shokrgozar M.A., Haghighipour N. Alteration of human umbilical vein endothelial cell gene expression in different biomechanical environments. Cell Biol Int. 2014;38:577–581. doi: 10.1002/cbin.10237. [DOI] [PubMed] [Google Scholar]
  • 35.Liu Z., Ruter D.L., Quigley K., Tanke N.T., Jiang Y., Bautch V.L. Single-cell RNA sequencing reveals endothelial cell transcriptome heterogeneity under homeostatic laminar flow. Arterioscler Thromb Vasc Biol. 2021;41:2575–2584. doi: 10.1161/ATVBAHA.121.316797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hough C., Cuthbert C.D., Notley C., Brown C., Hegadorn C., Berber E., et al. Cell type–specific regulation of von Willebrand factor expression by the E4BP4 transcriptional repressor. Blood. 2005;105:1531–1539. doi: 10.1182/blood-2002-10-3093. [DOI] [PubMed] [Google Scholar]
  • 37.Tsaryk R., Yucel N., Leonard E.V., Diaz N., Bondareva O., Odenthal-Schnittler M., et al. Shear stress switches the association of endothelial enhancers from ETV/ETS to KLF transcription factor binding sites. Sci Rep. 2022;12:4795. doi: 10.1038/s41598-022-08645-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File 1
mmc1.docx (13.5KB, docx)
Supplementary File 2
mmc2.zip (1.1KB, zip)
Supplementary Table 1

DESEQ2 output of list of differentially expressed genes of shear stress (any magnitude) vs static conditions.

mmc3.xlsx (1.1MB, xlsx)
Supplementary Table 2

A list of all genes found to be differentially expressed in HUVECs, HMVECs, and HPAECs exposed to shear stress, i.e. a gene that were universally differentially expressed regardless of cell type.

mmc4.xlsx (76.3KB, xlsx)
Supplementary Table 3

Curated list of custom endothelial gene sets based on previous reports. Some genes can be associated with multiple gene sets, denoted by a comma between respective gene sets.

mmc5.xlsx (12.2KB, xlsx)
Supplementary Table 4

DESEQ2 output of list of differentially expressed angiogenic genes (as defined by GO:0001525, see Figure 5) in HMVECs.

mmc6.xlsx (27.3KB, xlsx)

Data Availability Statement

All RNA-seq data has been deposited in GEO–accession number GSE294621.

Articles from Research and Practice in Thrombosis and Haemostasis are provided here courtesy of Elsevier

RESOURCES