Thioredoxin-interacting protein is a biomechanical regulator of Src activity: key role in endothelial cell stress fiber formation

Abstract

Rationale: Fluid shear stress differentially regulates endothelial cell stress fiber formation with decreased stress fibers in areas of disturbed flow compared with steady flow areas. Importantly, stress fibers are critical for several endothelial cell functions including cell shape, mechano-signal transduction, and endothelial cell-cell junction integrity. A key mediator of steady flow-induced stress fiber formation is Src that regulates downstream signaling mediators such as phosphorylation of cortactin, activity of focal adhesion kinase, and small GTPases.

Objective: Previously, we showed that thioredoxin-interacting protein (TXNIP, also VDUP1 [vitamin D upregulated protein 1] and TBP-2 [thioredoxin binding protein 2]) was regulated by fluid shear stress; TXNIP expression was increased in disturbed flow compared with steady flow areas. Although TXNIP was originally characterized for its role in redox and metabolic cellular functions, recent reports show important scaffold functions related to its α-arrestin structure. Based on these findings, we hypothesized that TXNIP acts as a biomechanical sensor that regulates Src kinase activity and stress fiber formation.

Methods and results: Using en face immunohistochemistry of the aorta and cultured endothelial cells, we show inverse relationship between TXNIP expression and Src activity. Specifically, steady flow increased Src activity and stress fiber formation, whereas it decreased TXNIP expression. In contrast, disturbed flow had opposite effects. We studied the role of TXNIP in regulating Src homology phosphatase-2 plasma membrane localization and vascular endothelial cadherin binding because Src homology phosphatase-2 indirectly regulates dephosphorylation of Src tyrosine 527 that inhibits Src activity. Using immunohistochemistry and immunoprecipitation, we found that TXNIP prevented Src homology phosphatase-2-vascular endothelial cadherin interaction.

Conclusions: In summary, these data characterize a fluid shear stress-mediated mechanism for stress fiber formation that involves a TXNIP-dependent vascular endothelial cadherin-Src homology phosphatase-2-Src pathway.

Keywords: TXNIP; disturbed flow; endothelial cells; laminar flow; stress fibers.

Figure 1. En face staining demonstrates differential regulation of Src pY416 and pY527 in KO animals compared to control

Aortas of control (A–C; H–J) or KO (D–F; K–M) were immuno-stained for VE-cadherin (A, D, H, K), Src pY416 (B, E), Src pY527 (I, L) or presented as merged images (C, F, J, M). (G, N) Quantification of the data using Image J show Src pY416 and pY527 intensity (* P<0.05 vs. control; n=4). White arrowheads denote formation of linear active Src patterns.

Figure 2. Western blot analysis demonstrates TXNIP expression

Src pY416 and pY527 are regulated by flow. (A) Total cell lysates of HUVEC exposed to s-flow or d-flow were immuno-blotted for pY416, pY527, Src and TXNIP. (B) Quantification of the data using Image J (* P<0.05 vs. s-flow; n=4).

Figure 3. VE-cadherin-SHP2 interaction is regulated by TXNIP

(A, B) Western blot analysis of immunoprecipitation samples from BAEC transfected with TX-WT-GFP or Y378A-GFP plasmids. (C) Western blot analysis of total cell lysates samples from BAEC transfected with control or TXNIP siRNA and TX-WT-GFP or Y378A-GFP mutant TXNIP. Quantification of the data using Image J (n=3).

Figure 4. CSK activity is regulated by flow

(A) Western blot analysis of HUVEC exposed to s-flow or d-flow and immuno-stained for active CSK, TXNIP and Actin. (B) Quantification of CSK activation in response to flow (* p<0.05 vs. s-flow; n=3).

Figure 5. TXNIP depletion results in spontaneous stress fiber formation

HUVEC transfected with control siRNA (A–C) or TXNIP siRNA (D–F) were immuno-stained for F-actin (A, D), Src pY416 (B, E) or presented as merged image (C, F). White arrows demonstrate sites of F-actin and Src pY416 colocalization. (G) Stress fiber area was measured and analyzed for signal intensity using Image J (* P<0.05 vs. control siRNA, n=4).

Figure 6. En face staining demonstrates flow regulation of TXNIP and stress fiber formation

s-flow (A–C) and d-flow (D–F) regions of animals demonstrate differential TXNIP expression (A, D), VE-cadherin (B, E) or presented as merged images (C, F). Aortic arch immuno-staining for TXNIP (G), F-actin (H) or presented as merged image (I) demonstrating both s-flow and d-flow regions as indicated by the broken white line.

Figure 7. Flow-mediated regulation of TXNIP expression and stress fiber formation

HUVEC were exposed to no flow (A–C), s-flow (D–F) or d-flow (G–I) conditions. Cells were fluorescently labeled for F-actin (A, D, G), TXNIP (B, E, H) or presented as merged (C, F, I).

Figure 8. Schematic model for TXNIP regulation of Src signaling and stress fiber formation

Under basal conditions, TXNIP prevents SHP2 to interact with VE-cadherin resulting in pY527 Src by CSK. Upon decreased TXNIP expression, SHP2 can be recruited to VE-cadherin, where it dephosphorylates CSK, relieving the inhibitory pY527 and allowing activation of Src via increased pY416. As a consequent of active Src, stress fibers formation is observed.