Tuberous sclerosis complex 2 association with RelA/p65 is critical for NF-κB activation and endothelial cell inflammation

Abstract

Background

Endothelial cell (EC) inflammation is a key component of many inflammatory conditions including sepsis and acute lung injury (ALI). However, the role of Tuberous Sclerosis Complex 2 (TSC2) in activating NF-κB and inflammatory response in EC has not been addressed.

Methods

Human pulmonary artery endothelial cells (HPAEC) or human lung microvascular endothelial cells (HLMVEC) were transfected with siRNA targeting TSC2 and then challenged with thrombin, a procoagulant and proinflammatory mediator whose concentration is elevated in patients with ALI and sepsis, to address the role of TSC2 in activating NF-κB subunit RelA/p65 to cause inflammatory response in EC. In some experiments, lipopolysaccharide (LPS), a robust inducer of EC inflammation, was used to determine if TSC2 is common mediator of this response. The cells were evaluated for IκBα phosphorylation/degradation, RelA/p65 phosphorylation and nuclear accumulation by immunoblotting. DNA binding of nuclear RelA/p65 was determined using an ELISA-based assay kit. RelA/p65 transcriptional activity was determined by measuring NF-κB-luciferase reporter activity. TSC2 association with RelA/p65 was assessed by immunoprecipitation followed by immunoblotting.

Results

We found that RelA/p65 is constitutively associated with TSC2 (in addition to IκBα), and this association is reduced in thrombin-stimulated cells, suggesting a role of TSC2 in regulating RelA/p65 activation and EC inflammatory response. We examined this possibility by silencing TSC2 to disrupt its association with RelA/p65-IκBα complex. TSC2 silencing resulted in reduced IκBα phosphorylation/degradation, and subsequently, RelA/p65 nuclear translocation and DNA binding activity in response to thrombin. TSC2-silenced cells also showed reduced Ser⁵³⁶ phosphorylation of RelA/p65, a key modification required for its transcriptional function. Consistent with this, TSC2 silencing impaired NF-κB-dependent reporter activity and expression of proinflammatory mediators such as ICAM-1, VCAM-1 and IL-6 induced by thrombin. Similarly, TSC2 silencing was also effective in decreasing LPS-induced activation of RelA/p65 and expression of proinflammatory mediators, indicating that TSC2 is a common mediator of these responses. Notably, the proinflammatory action of TSC2 in EC appears to be independent of its ability to inhibit MTORC1.

Conclusions

Together, these results identify TSC2 as a critical component of RelA/p65-IκBα complex that aids in facilitating the activation of RelA/p65 to cause EC inflammatory response. Thus, the targeting of TSC2 may be a useful strategy for mitigating EC inflammation associated with intravascular coagulation and sepsis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-026-02922-z.

Introduction

Endothelial cells (EC) line the vasculature of the lungs where they create a blood-air interface and execute distinct functions such as preservation of barrier integrity, anti-coagulant, and anti-adhesive cell surface and thereby preventing vascular leakage, intravascular coagulation and inflammatory cell infiltration. During inflammatory disease states such as sepsis and acute lung injury (ALI), components of bacterial cell wall, such as lipopolysaccharides (LPS), and the release of procoagulant thrombin, lead to EC inflammation via activation of their respective receptors, Toll-like receptor (TLR) and proteinase-activated receptor 1 (PAR-1) [1–4]. Once an inflammatory response has been instigated, several proinflammatory mediators including adhesion molecules such as intercellular adhesion molecule 1 (ICAM-1) are induced. The interaction of ICAM-1 with its counter receptor β₂-integrins (CD11/CD18) on polymorphonuclear leukocytes (PMN) allows a firm and stable adhesion of PMN to EC [1, 5–7] and thus facilitates their transendothelial migration into the underlying interstitium [8–10].

We and others have shown that the transcription factor nuclear factor-kappa B (NF-κB) is an essential regulator of endothelial gene expression of ICAM-1 and other adhesion molecules (VCAM-1 and E-selectin) [1–4]. The prototypical NF-κB complex exists as a heterodimer of 50 kDa (p50) and 65 kDa (RelA/p65) subunits and is sequestered in the cytoplasm by inhibitor of κB (IκB) proteins the prototype of which is IκBα. Binding of IκBα masks the nuclear localization sequence (NLS) of NF-κB [11–13], and thus prevents its migration to the nucleus. Phosphorylation of IκBα (Ser³² and Ser³⁶) by a macromolecular IκB kinase (IKK) complex containing the catalytic subunits, IKKα and IKKβ, and the regulatory subunit, IKKγ leads to its rapid ubiquitination and proteasomal degradation. The released NF-κB migrates to the nucleus where it binds to the promoter of its target genes and activate their transcription [3]. Another important mechanism controlling NF-κB activity involves modulation of its transcriptional function by phosphorylation of RelA/p65 subunit [3, 11, 14]. Studies have shown that phosphorylation of RelA/p65 at Ser²⁷⁶, Ser³¹¹, Ser⁵²⁹, or Ser⁵³⁶ increases its transcriptional capacity of NF-κB in a cell- and stimulus-specific manner [3, 11, 14]. We have shown that thrombin enhances the transcriptional capacity of NF-κB by increasing the phosphorylation of RelA/p65 at Ser⁵³⁶ [3, 15, 16].

Tuberous sclerosis complex 2 (TSC2), also known as tuberin, is a tumor suppressor that complexes with TSC1 (hamartin) to restrict cell growth [17]. TSC1/TSC2 complex exerts its tumor suppressor function by negatively regulating Mechanistic Target of Rapamycin Complex 1 (MTORC1) [18]. In this complex, TSC1 serves as a membrane-tethering anchor protein whereas TSC2 acts as a GTPase activating protein (GAP) to inactivate RheB, a small G-protein activator of MTORC1. Inactivation or loss of function mutations in either the TSC1 or TSC2 gene causes abnormal up-regulation of MTORC1 signaling and is associated with the pathogenesis of tuberous sclerosis and the proliferative lung disorder lymphangioleiomyomatosis [19]. While TSC2 primarily acts as an inhibitor of MTORC1, it can also exert its functions in an MTORC1-independent manner [20]. For example, loss of TSC2 results in increased expression of vascular endothelial growth factor (VEGF) via an MTORC1-independent mechanism that involves HIF-1α accumulation and chromatin remodeling [20]. Similarly, TSC2 modulation of Rho/Rac1 activities and thereby actin dynamics is also independent of its function as a negative regulator of MTORC1 signaling [21]. Studies have implicated TSC2/MTORC1 axis in both positive and negative regulation of IKK/NF-κB activity [22, 23]. In some contexts, TSC2 inhibits NF-κB activity, while in others, TSC2 promotes NF-κB activation and inflammation [22–24].

The specific effect of TSC2 on NF-κB depends on the cellular context and the presence of other signaling pathways [22–24]. However, the role of TSC2 in regulating NF-κB signaling in EC remains unclear, although a role for MTOR in restraining NF-κB activity to limit EC inflammation has been reported [25, 26]. In this study, we determined whether TSC2 plays a role in NF-κB signaling and thereby EC inflammation and whether TSC2 exerts these effects via MTORC1 or independently of it. Our data show that TSC2 is constitutively associated with RelA/p65, and this association is required for RelA/p65 activation to cause EC inflammation and that these proinflammatory actions of TSC2 appear to be MTORC1-independent.

Materials and methods

Reagents

Antibodies to phospho-IKKαβ (Ser^176/177) (2078 S), IKKβ (8943 S), phospho-IκBα (Ser^32/36) (9246 S), IκBα (4812 S), phospho-NF-κB/p65 (Ser⁵³⁶)(3033 S), phosphor-MTOR (Ser²⁴⁴⁸)(2971), MTOR (2983 S), phosphor-Raptor (Ser⁷⁹²) (2083), Raptor (2280 S), TATA Binding Protein (44059T), phosphor-p70s6k (Thr³⁸⁹) (9205 S), p70s6k (2708 S) and TSC2 (3612 S) were obtained from Cell Signaling Technologies. Antibodies to TSC2 (sc-271314), ICAM-1 (sc-8439), VCAM-1 (sc-13160), NF-κB/p65 (sc-8008), IκBα (sc-1643), β-Actin (sc-47778) and GAPDH (sc-32233) were from Santa Cruz Biotechnology. Antibody to α-Tubulin (T-6199) was from Sigma-Aldrich. Horseradish Peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (170–6516) and HRP-conjugated affinity purified goat anti-rabbit IgG antibody (170–6515) were from BioRad. Anti-mouse (500-M00-1MG) and anti-Rabbit (500-P00-1MG) IgG antibodies were from PeperoTech/Invitrogen. NF-κB (p65) transcription factor assay kit (10007889) was purchased from Cayman Chemical. Thrombin (HT1002a) was obtained from Enzyme Research Lab. LPS from Escherichia coli (O55:B5) (L2880) and cyclohexamide (01810-5G) was from Sigma-Aldrich. Protein A/G Magnetic beads (B23201) were from Selleck Chemicals.

Endothelial cell culture

Primary Human Pulmonary Artery Endothelial Cells (HPAEC) (CC-2530) and Human Lung Microvascular Endothelial Cells (HLMVEC) (CC-2527) were purchased from Lonza. Cells were cultured as described [26] in gelatin-coated flasks using endothelial basal medium 2 (EBM2) (Lonza) and supplemented with 10% fetal bovine serum (FBS) (Gibco) and bullet kit additives (Lonza). HPAEC and HLMVEC between passages 3 and 7 were grown in a humidified 37 °C incubator supplemented with 5% CO₂. For thrombin and LPS treatments, cells were incubated in 1% FBS containing MCDB (Corning) for 2 h prior to stimulation with Thrombin (5U/ml) and LPS (1 µg/ml).

Immunoprecipitation

HPAEC were grown on gelatin coated 100 mm dishes until they reached confluency. Cells were incubated in MCDB media supplemented with 1% FBS for 2 h and then treated with 5U/ml Thrombin for 1 h. Cells were washed with ice cold 1X phosphate buffered saline (PBS) (pH 7.2; Gibco) and then lysed in 500 µL of IP lysis buffer (50mM HEPES, 5mM EDTA, 50mM NaCl, 1% Triton-x supplemented with 1 tablet protease inhibitor cocktail (Roche)). Cell lysate was incubated with 30 µL magnetic protein A/G beads (Selleckchem) and 2–4 µg of appropriate antibody at 4 °C overnight with gentle shaking. Immunoprecipitated proteins-conjugated to magnetic beads were removed and washed 3 times with 300 µL of ice-cold IP lysis buffer. The immunoprecipitated proteins were extracted by boiling with SDS sample buffer for 7 min and analyzed by immunoblotting as described below.

siRNAs

The following siRNAs were used: Negative control: a non-targeting oligonucleotide; Human TSC2 SMART pool: 5’-GCAUUAAUCUCUUACCAUA-3’, 5’-CCAAUGUCCUCUUGUCUUU-3’, 5’-UCACCAGGCUCAUCAAGAA-3’, 5’-GGAAUGUGGCCUCAACAAU-3’ and Human RAPTOR SMART pool: 5’- GAAACCAUCGGUGCAAAUU-3’, 5’- AGAAGGGCAUUACGAGAUU-3’, 5’-UGGAGAAGCGUGUCAGAUA-3’, 5’-GCCCGUCGAUCUUCGUCUA-3’ (Horizon Discovery).

Endothelial cell transfection

Predesigned SMART pool short-interfering RNA (siRNA) duplexes specific for human TSC2 (siTSC2), RAPTOR (siRaptor) and a non-targeting siRNA control (siCon) (Dharmacon), were used to deplete TSC2 or RAPTOR in HPAEC or HLMVEC according to DharmaFect1 transfection protocol. Briefly, 100 nM siRNA and DharmaFect1 transfection reagent were mixed for 20 min and then incubated with cells that are 60–70% confluent for 18–24 h, after which fresh transfection media was removed and replaced with complete EBM2 media. Cells were utilized in experiments after 24–36 h [27]. In double knockdown experiments, siRNA specific for TSC2 (siTSC2), RAPTOR (siRaptor) and a non-targeting control siRNA (siCon) (Dharmacon) were used to transfect HPAEC. Briefly, 100 nM siCon or 50 nM siTSC2 and 50 nM siCon or 50 nM siRAPTOR and 50 nM siCon or 50 nM siTSC2 and 50 nM siRAPTOR were mixed with DharmaFect1 transfection reagent for 20 min and then incubated with cells that are 60–70% confluent for 18–24 h, after which fresh transfection media was removed and replaced with complete EBM2 media. Cells were utilized in experiments after 24–36 h [27].

Cell lysis and immunoblotting

After designated treatment time, HPAEC and HLMVEC were lysed in radioimmunoprecipitation (RIPA) buffer (150 mM NaCl, 5 mM NaF, 50 mM Tris-HCl, pH 7.4, 0.25 mM EDTA, pH 8.0, 1% deoxycholic acid, 1% Triton-X) (Thermo Fisher Scientific, Waltham, MA) supplemented with EDTA-free protease inhibitor cocktail tablet (Roche Diagnostics) and HALT phosphatase inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA) as described [28]. Cell lysates were resolved via SDS-PAGE and transferred onto nitrocellulose membranes and then blocked in 5% (w/v) nonfat dry milk in Tris-buffered saline + Tween 20 [10mM Tris (pH 8.0), 150 mM NaCl, and 0.05% Tween 20] (TBST) for non-phosphorylated antibodies and 5% bovine serum albumin (Cell Signaling Technology) in TBST for phosphorylated antibodies. Membranes were then incubated overnight at 4 °C in primary antibodies (1:1000) mixed in appropriate blocking buffer afterwhich they were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (1:10,000) (Bio-Rad Laboratories Inc) mixed in appropriate blocking buffer and visualized by enhanced chemiluminescence (ECL) detection system (SuperSignal West Femto Substrate (Thermo Fisher Scientific, Waltham, MA)) as described [29] and imaged using the Chemi-Doc Imaging System (Bio-Rad Laboratories Inc). Densitometric analysis of band intensity was analyzed using the “Measure” tool in Fiji ImageJ (https://imagej.net).

qRT-PCR

HPAEC were transfected with siControl or siTSC2 targeting siRNA. Cells were harvested and total RNA was isolated using RNeasy Mini Kit (Qiagen) and reverse transcription was performed using oligo (dT) primers (Integrated DNA Technologies) and superscript RT (Invitrogen) as described [25]. p70s6k and GAPDH were amplified using the primer sets: p70s6k: 5’-TAC TTC GGG TAC TTG GTA A-3’ (forward) and 5’-GAT GAA GGG ATG CTT TAC T-3’ (reverse), GAPDH: 5’-TATCGTGGAAGGACTCATGACC-3’ (forward) and 5’-TACATGGCAACTGTGAGGGG-3’ (reverse) (Integrated DNA Technologies). PCR consisted of 35 cycles of 30 s at 95 °C, 30 s at 55 °C, and 60 s at 72 °C using thermocycler (Bio-Rad Laboratories Inc).

Reporter gene construct and luciferase assay

The pNF-κB-LUC plasmid which contains five copies of consensus NF-κB sequences linked to a minimal E1B promoter-luciferase gene was obtained from Stratagene (La Jolla, CA). Normalizing transfection efficiencies was done using the pTKRLUC plasmid (Promega) containing the Renilla luciferase gene which is driven by the constitutively active thymidine kinase promoter. Transfection was performed using DEAE-dextran method as described [25]. Briefly, HPAEC were transfected with 100 nM siCon or siTSC2 and 24 h later, 5 µg of DNA was mixed with 50 µg/ml DEAE-dextran in serum-free MCDB and the mixture was added onto subconfluent cells. After 1 h, cells were incubated for 4 min with 10% DMSO in serum-free MCDB. The cells were then washed twice with EBM2 complete media and grown to confluency afterwhich they were challenged with thrombin (5U/ml) for 6 h. Cell extracts were prepared and assayed for Firefly and Renilla luciferase activities using a luminometer.

NF-κB (p65) nuclear translocation and NF-κB (p65) transcription factor DNA binding assay

HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with thrombin (5U/ml) for 1 h after which, HPAECs were washed twice with ice cold 1X PBS, scraped and resuspended in 400 µL of buffer A (10 mM HEPES, 10 mM KCl, 0.1 mM EGTA, 0.1 mM EDTA, 1 mM dithiothreitol, 1 tablet protease inhibitor cocktail (Roche)). After 20 min, 10% Nonidet P-40 was added. The samples were vortexed and centrifuged to collect the supernatants containing cytosolic proteins. The nuclear pellet was resuspended in 100 µL buffer B (20 mM HEPES (pH 7.9), 0.4 M KCl, 1 mM EGTA, 1 mM EDTA, 1 mM dithiothreitol, 1 tablet protease inhibitor cocktail (Roche)), homogenized and incubated for 30 min at 4 °C. The lysates were centrifuged and the supernatant containing nuclear proteins were transferred to new tubes. Protein concentration of the cytosolic and nuclear fractions was determined using a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA) and lysate was resolve via SDS-PAGE as described above. An ELISA-based NF-κB (p65) Transcription Factor DNA binding assay was performed according to kit protocol (Cayman Chemical, Ann Arbor, MI [Item No. 10007889]). Nuclear fraction cell extracts of siCon or siTSC2 transfected HPAEC were added to 96-well plate, provided by kit, and incubated overnight at 4 °C. Transcription factor NF-κB (p65) primary antibody was added to sample wells and 1 h later secondary antibody was added and incubated for 1 h. Transcription factor developing solution was added to each well and 30 min later the absorbance was read using a fluorescent plate reader.

ELISA

HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with thrombin (5U/ml) or LPS (1ug/uL) for 6 h and cell media supernatant was collected. Interleukin-6 (IL-6) levels were determined using ELISA kit from R&D Systems (DY206).

Statistical analysis

Statistical analysis was performed using GraphPad prism 10.5.0 (GraphPad Software, San Diego, CA, USA). Data were analyzed by one-way ANOVA followed by Tukey’s post hoc test for comparison between multiple groups while two-group comparisons used 2-tailed unpaired Student’s t-test. Data are expressed as mean ± SD. A p-value < 0.05 was considered statistically significant.

Results

TSC2 is constitutively associated with RelA/p65 and this association is reduced in EC stimulated with thrombin

To address the role of TSC2 in NF-κB activation, we first evaluated its ability to interact with RelA/p65, an NF-κB subunit critically involved in EC inflammation [1, 30]. We tested this possibility by immunoprecipitating TSC2 from control and thrombin-challenged cells and then analyzed the TSC2 immunoprecipitates for the presence of RelA/p65 by immunoblotting. Because thrombin activation of RelA/p65 peaks at 1 h after stimulation, we used this timepoint for monitoring TSC2 interaction with RelA/p65 [28]. Results showed that TSC2 was constitutively associated with RelA/p65 and this association decreased when cells were stimulated with thrombin (Fig. 1A; upper panel). Immunoblotting of the input showed that thrombin had no effect on TSC2 level Fig. 1A; lower panel), excluding the possibility that reduced TSC2-RelA/p65 association in thrombin-stimulated cells was a result of decreased TSC2 levels. In reciprocal experiments, we analyzed RelA/p65 immunoprecipitates for the presence of TSC2. We found that RelA/p65 and TSC2 were associated in control untreated cells, but this interaction was disrupted upon thrombin stimulation (Fig. 1B; upper panel). As expected, RelA/p65 was constitutively associated with IκBα and this association was lost in thrombin-stimulated cells (Fig. 1B; upper panel) because of IκBα degradation (Fig. 1C; lower panel). Similarly, IκBα immunoprecipitates from control and thrombin-challenged cells were also evaluated for the presence of TSC2 and RelA/p65. A constitutive association was noted between IκBα and TSC2 which was disrupted in thrombin-stimulated cells (Fig. 1C; upper panel) because of IκBα degradation (Fig. 1C; lower panel). IκBα was constitutively associated with RelA/p65 as expected, and this association was lost because of IκBα degradation when cells were stimulated with thrombin (Fig. 1C; middle panel). TSC2, RelA/p65 or IκBα was barely detected in IgG immunoprecipitates (Fig. 1A-C), indicating the specificity of TSC2-RelA/p65-IκBα interaction. These results demonstrate that TSC2 is constitutively bound to RelA/p65-IκBα complex and this binding is markedly reduced upon stimulation cells with thrombin, suggesting a role of TSC2 in RelA/p65 activation.

Fig. 1.

TSC2 associates with RelA/p65 and this association is disrupted by thrombin. HPAECs were treated with thrombin (5 U/ml) for 1 h. Total cell lysates were subjected to immunoprecipitation (IP) using an anti-TSC2, anti-RelA, anti-IκBα or anti-IgG antibody. A TSC2 immunoprecipitates (upper panel) were immunoblotted (IB) with an anti-RelA or anti-TSC2 antibody. Total cell lysates (input; lower panel) were immunoblotted with an antibody to TSC2 or Actin. B RelA immunoprecipitates (upper panel) were immunoblotted with an anti-TSC2 or anti- IκBα antibody. Total cell lysates (input; lower panel) were immunoblotted with an antibody to RelA or Actin. C IκBα immunoprecipitates were immunoblotted with (upper panel) an anti-TSC2 or anti-IκBα antibody and (middle panel) an anti-RelA or anti-IκBα antibody. Total cell lysates (input; lower panel) were immunoblotted with an antibody to IκBα or Tubulin. MW, molecular weight ladder, TL, total lysate

TSC2 is required for IκBα phosphorylation/degradation-dependent nuclear translocation and DNA binding of RelA/p65

NF-κB is sequestered in the cytoplasm by IκBα which masks the NLS of RelA/p65 preventing its translocation to the nucleus [3, 11]. Activation of RelA/p65 requires it release from IκBα which can be accomplished primarily by phosphorylation (Ser³² and Ser³⁶)-dependent degradation of IκBα. Proinflammatory mediators such as thrombin and LPS activate the PAR-1 and TLR receptors respectively to initiate signaling cascades which converge on IκBα phosphorylation. Phosphorylation leads to degradation of IκBα and thereby release of RelA/p65 for its migration to the nucleus where it binds to the promoters of its target genes to activate their transcription. We examined if TSC2 association with RelA/p65-IκBα is required to drive these events to activate RelA/p65 signaling in EC. To examine this possibility, we used the knockdown approach to disrupt TSC2 association with RelA/p65-IκBα complex. Transfection with TSC2 siRNA resulted in an efficient depletion of TSC2 (Fig. 2A). Silencing of TSC2 by this approach markedly reduced IκBα phosphorylation and degradation caused by thrombin (Fig. 2B). We also performed a cycloheximide (CHX) chase experiment to track IκBα stability in the presence or absence of TSC2 knockdown and found that loss of TSC2 had no effect on IκBα stability (Supplemental Fig. 1). These results exclude the possibility that the reduced IκBα degradation in TSC2-depleted cells was due to changes in IκBα protein turnover.

Fig. 2.

TSC2 depletion attenuates thrombin- and LPS-induced IκBα phosphorylation/degradation via inhibition of IKKα/β. A-C HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with thrombin (5 U/ml) for 1 h. A Total cell lysates were immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. B Total cell lysates were immunoblotted with an anti-phospho-IκBα (Ser³²/Ser³⁶) antibody. The blots were subsequently stripped and reprobed with an anti-IκBα antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. C Total cell lysates were immunoblotted with an anti-phospho-IKKα/IKKβ (Ser¹⁷⁶/Ser¹⁷⁷) antibody. The blots were subsequently stripped and reprobed with an anti-IKKβ antibody. Bars indicate mean ± SD (n = 3) for each condition. D-F HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with LPS (1 µg/ml) for 1 h. D Total cell lysates were immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. E Total cell lysates were immunoblotted with an anti-phospho-IκBα (Ser³²/Ser³⁶) antibody. The blots were subsequently stripped and reprobed with an anti-IκBα antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. F Total cell lysates were immunoblotted with an anti-phospho-IKKα/IKKβ (Ser¹⁷⁶/Ser¹⁷⁷) antibody. The blots were subsequently stripped and reprobed with an anti-IKKβ antibody. Bars indicate mean ± SD (n = 3) for each condition. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

The protective effect of TSC2 knockdown on IκBα phosphorylation led us to examine the status of IKK activation in TSC2-depleted cells. IKK complex, composed of IKKα and IKKβ kinase subunits and a regulatory IKKγ subunit, phosphorylates IκBα at Ser³² and Ser³⁶ [3]. Activation of IKK requires the phosphorylation of IKKα and IKKβ at Ser¹⁷⁶ and Ser¹⁷⁷ respectively [3]. We determined if TSC2 contributes to IκBα phosphorylation and degradation by influencing IKKα/β phosphorylation. Knockdown of TSC2 reduced IKKα/β phosphorylation at Ser¹⁷⁶ and Ser¹⁷⁷ caused by thrombin (Fig. 2C), consistent with the marked reduction in thrombin-induced IκBα phosphorylation and degradation in TSC2-depleted cells (Fig. 2B). In parallel experiments, we also determined the effect of TSC2 knockdown on IKK activation and thereby IκBα phosphorylation and degradation in response to LPS. Results showed that LPS-induced IκBα phosphorylation and degradation as well as IKK activation was significantly diminished in TSC2-depleted cells (Fig. 2D and F). Together, these findings show that TSC2 serves as a common mediator of IKK-dependent IκBα phosphorylation and degradation in the pathway leading to RelA/p65 activation.

Because IKK-dependent IκBα phosphorylation and degradation are necessary for nuclear translocation and DNA binding of the released RelA/p65, we investigated whether TSC2 silencing also impairs these events. To this end, control and TSC2-depleted cells were treated with or without thrombin, and nuclear extracts were immunoblotted with an anti-RelA/p65 antibody. Control cells showed increased levels of nuclear RelA/p65 after thrombin challenge, as expected. In contrast, this increase was decreased in TSC2-depleted cells (Fig. 3A). Similarly, thrombin-induced DNA binding activity of RelA/p65 was reduced in TSC2-depleted cells (Fig. 3B). Together, these data identify an important role of TSC2 in causing RelA/p65 activation.

Fig. 3.

TSC2 depletion inhibits thrombin-induced RelA/p65 nuclear translocation and DNA binding activity. HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with thrombin (5 U/ml) for 1 h. A Nuclear extracts were immunoblotted with an anti-RelA/p65 antibody. TATA binding protein (TBP) was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. B Nuclear extracts were analyzed for RelA/p65 DNA binding activity using an ELISA-based assay as described in Materials and Methods. Bars indicate mean ± SD (n = 5) for each condition. **p < 0.01, ***p < 0.001, and ****p < 0.0001

TSC2 promotes phosphorylation and transcriptional activity of RelA/p65 to cause EC inflammation

Given the impaired nuclear translocation and DNA binding of RelA/p65 in TSC2-depleted cells, we determined if TSC2 regulates phosphorylation of RelA/p65 at Ser⁵³⁶, a critical determinant of RelA/p65 transcriptional capacity [3, 14]. When stimulated with thrombin, control cells showed a robust increase in RelA/p65 phosphorylation, whereas this response was attenuated in TSC2-silenced cells (Fig. 4A). Similarly, TSC2 silencing diminished RelA/p65 phosphorylation at Ser⁵³⁶ induced by LPS (Fig. 4B). We next asked if the impaired phosphorylation, nuclear translocation, and DNA binding of RelA/p65 led to reduced transcriptional activity of NF-κB and thereby expression of its target genes in EC. As shown in Fig. 5C, thrombin-induced NF-κB-dependent reporter activity was markedly reduced in TSC2 knockdown cells.

Fig. 4.

TSC2 knockdown attenuates RelA/p65 phosphorylation and transcriptional activity. A&B HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with (A) thrombin (5 U/ml) or (B) LPS (1 µg/ml) for 1 h. Total cell lysates were immunoblotted with an anti-phospho-RelA/p65 (Ser⁵³⁶) antibody. The blots were subsequently stripped and reprobed with an anti-RelA/p65 antibody to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. C HPAEC were transfected with siCon or siTSC2. After 24 h, cells were transfected again with NF-κB-Firefly luciferase (NF-κBLUC) and Renilla-LUC constructs using diethylaminoethyl (DEAE)-dextran as described [28]. Twenty-four hours later, cells were challenged with thrombin (5 U/ml) for 6 h. Cell extracts were prepared and assayed for Firefly and Renilla luciferase activities. Renilla luciferase was used as an internal control for transfection efficiency. Bars indicate mean ± SD (n = 5–6) for each condition. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Fig. 5.

TSC2 depletion attenuates thrombin- and LPS-induced expression of proinflammatory mediators. A&B HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with thrombin (5 U/ml) for 6 h. ATotal cell lysates were immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. B Total cell lysates were immunoblotted with an anti-VCAM-1 antibody. Blot was stripped and then immunoblotted with an anti-ICAM-1 antibody. Actin levels were used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. C&D HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with LPS (1 µg/ml) for 6 h. C Total cell lysates were immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3–4) for each condition. D Total cell lysates were immunoblotted with an anti-VCAM-1 antibody. Actin levels were used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. E&F HLMVEC were transfected with siCon or siTSC2 for 48 h and then challenged with LPS (1 µg/ml) for 6 h. E Total cell lysates were immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. F Total cell lysates were immunoblotted with an anti-VCAM-1 antibody. Actin levels were used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. G&H HPAEC were transfected with siCon or siTSC2 for 48 h and then challenged with (G) thrombin (5 U/ml) or H LPS (1 µg/ml) for 6 h. Cells were harvested and analyzed by ELISA to assess the levels of the proinflammatory mediator IL-6. Bars indicate mean ± SD (n = 3) for each condition. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Since the expression of proinflammatory genes, ICAM-1 and VCAM-1, relies on RelA/p65 activation [1, 30], we examined if loss of TSC2 influences the expression of these molecules. Indeed, TSC2-depleted cells displayed reduced expression of ICAM-1 and VCAM-1 in response to both thrombin and LPS (Fig. 5A-D). Importantly, TSC2 silencing was also effective in inhibiting LPS-induced VCAM-1 expression not only in HPAEC but also in HLMVEC (Fig. 5E-F). Thus, the role of TSC2 in EC inflammation appears to be consistent in both macrovascular and microvascular EC in the lung. The production of interleukin-6 (IL-6) is a clinical feature of NF-κB activation in the vasculature and plasma levels of IL-6 are elevated in septic patients with ALI [31, 32]. We found that thrombin- and LPS-induced IL-6 levels were reduced in TSC2-depleted cells (Fig. 5G&H). Collectively, these results show that TSC2 promotes EC inflammation by inducing the release of RelA/p65 from IκBα in the cytoplasm, facilitating its phosphorylation at Ser⁵³⁶, nuclear translocation, DNA binding of RelA/p65, and transcriptional activation of proinflammatory genes.

TSC2 mediates EC inflammation through an MTORC1-independent mechanism

The ability of TSC2 to associate with RelA/p65-IκBα complex raised the possibility that it might activate RelA/p65 independently of MTORC1. To test this, we first determined the status of MTORC1 activation in TSC2-depleted cells. TSC2 knockdown (Fig. 6A) resulted in increased activating phosphorylation of MTOR at Ser²⁴⁴⁸ (Fig. 6C) and decreased inhibitory phosphorylation of RAPTOR at Ser⁷⁹² (Fig. 6D). These data indicate that loss of TSC2 results in MTORC1 activation, as expected. We also determined if MTORC1 activation in TSC2-depleted cells results in increased level/activity of its downstream target p70s6k. Results showed that loss of TSC2 caused an increase in the level but not the activating phosphorylation of p70s6k (Thr³⁸⁹) in both HPAEC (Fig. 6A&G) and HLMVEC (Fig. 6B&H). To confirm that this increased levels of p70s6k was not due to an off-target effect of the siRNA, a second TSC2- specific siRNA was used which also resulted in decreased TSC2 and increased p70s6k levels (Fig. 6I). qRT-PCR analysis of TSC2-depleted cells revealed that the increased p70s6k protein level was a result of increased p70s6k mRNA expression (6J). These data indicate that TSC2 depletion in EC results in MTORC1 activation and downstream signaling via increased p70s6k levels. These findings prompted us to investigate if TSC2 knockdown suppresses EC inflammation through MTORC1 activation. To this end, we disrupted MTORC1 signaling by silencing RAPTOR in TSC2-depleted cells and evaluated the status of EC inflammation in response to thrombin (Fig. 7A&B). TSC2 and RAPTOR siRNAs effectively and selectively depleted TSC2 and RAPTOR respectively both in control (untreated) and thrombin-stimulated cells (Fig. 7A&B). Impairing MTORC1 activation by this approach (via RAPTOR depletion) did not restore VCAM-1 expression in TSC2-depleted cells (Fig. 7C). Surprisingly, depleting RAPTOR alone (thereby preventing MTORC1 activation) also inhibited thrombin-induced VCAM-1 expression (Fig. 7C). These findings suggest that TSC2 regulates EC inflammatory response through an MTORC1-independent pathway.

Fig. 6.

TSC2 depletion augments MTORC1 signaling. A HPAEC were transfected with siCon or siTSC2 for 48 h. Total cell lysates were prepared and immunoblotted with an anti-TSC2 antibody. Tubulin was used to monitor loading. Bars indicate mean ± SD (n = 5–6) for each condition. B HLMVEC were transfected with siCon or siTSC2 for 48 h. Total cell lysates were prepared and immunoblotted with an anti-TSC2 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 6–7) for each condition. C HPAEC were transfected with siCon or siTSC2 for 48 h. Total cell lysates were prepared and immunoblotted with an anti-phospho-MTOR (Ser²⁴⁴⁸) antibody. Blot was stripped and reprobed with an anti-MTOR antibody to monitor loading. Bars indicate mean ± SD (n = 6) for each condition. D HPAEC were transfected with siCon or siTSC2 for 48 h. Total cell lysates were prepared and immunoblotted with an anti-phospho-RAPTOR (Ser⁷⁹²) antibody. Blot was stripped and reprobed with an anti-RAPTOR antibody to monitor loading. Bars indicate mean ± SD (n = 6) for each condition. E HPAEC total cell lysate was immunoblotted with an anti-phospho-p70s6k (Thr³⁸⁹) antibody. The blots were subsequently stripped and probed with an anti-p70s6k antibody. Actin was used as a loading control. Bars indicate mean ± SD (n = 6) for each condition. F TSC2-depleted HLMVEC total cell lysate was immunoblotted with an anti-phospho-p70s6k (Thr³⁸⁹) antibody. The blots were subsequently stripped and probed with an anti-p70s6k antibody. Actin was used as a loading control. Bars indicate mean ± SD (n = 6–7) for each condition. G To rule out that the blot in Fig. 6E was not completely stripped a fresh gel was run and HPAEC total cell lysates were immunoblotted with an anti-p70s6k antibody. Actin was used as a loading control. Bars indicate mean ± SD (n = 6) for each condition. H To rule out that the blot in Fig. 6F was not completely stripped a fresh gel was run and HLMVEC total cell lysate was immunoblotted with an anti-p70s6k antibody. Actin was used as a loading control. Bars indicate mean ± SD (n = 6) for each condition. I HPAEC were transfected with siCon or a different siTSC2 (Santacruz). Total cell lysates were prepared and immunoblotted with anti-TSC2 and anti-p70s6k antibodies. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. J HPAEC were transfected with siCon or siTSC2 for 48 h. RNA was isolated and RT-qPCR was performed using p70s6k specific RNA primer and mRNA level was quantified. GAPDH was used as a control. Bars indicate mean ± SD (n = 5–6) for each condition. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p 0.0001

Fig. 7.

TSC2 regulates EC inflammation independent of MTORC1. A-C HPAEC were transfected with siCon, siTSC2, siRAPTOR or both siTSC2 and siRAPTOR for 48 h and then challenged with thrombin (5 U/ml) for 6 h. A&B Total cell lysates from untreated (A) or thrombin-treated (B) were immunoblotted with an anti-TSC2 or anti-RAPTOR antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 3) for each condition. C Total cell lysates were immunoblotted with an anti-VCAM-1 antibody. Actin was used to monitor loading. Bars indicate mean ± SD (n = 5) for each condition. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Discussion

The major findings of this study are that TSC2, a well-established inhibitor of MTORC1 [33], is an important component of RelA/p65-IκBα complex that plays a critical role in facilitating RelA/p65 activation to cause EC inflammation. We present evidence that TSC2 is constitutively associated with RelA/p65-IκBα complex, and that this association is required for RelA/p65 activation. Disrupting this association via knockdown of TSC2 resulted in impaired IκBα phosphorylation/degradation and thereby reduced nuclear translocation and DNA binding of RelA/p65 in response to thrombin. TSC2 knockdown also reduced thrombin-induced NF-κB-dependent reporter activity and expression of ICAM-1, VCAM-1, and IL-6 by attenuating Ser⁵³⁶ phosphorylation of RelA/p65. The proinflammatory action of TSC2 in EC appears to be MTORC1-independent. Importantly, TSC2 knockdown was also effective in decreasing LPS-induced RelA/p65 activation and EC inflammation, indicating that TSC2 serves as a common mediator of these responses.

The relationship between TSC2 and NF-κB is complex and appears to be context-dependent. In some contexts, TSC2 acts as a positive regulator of NF-kB activation, while in others it can inhibit NF-κB activity [22–24]. For example, loss of TSC2 attenuated NF-κB signaling in mouse embryonic fibroblasts and monocytes but increased NF-κB activity in PTEN-deficient prostate cancer cells [22–24]. Thus, the specific role of TSC2 in NF-κB regulation depends on the cellular context and the presence of other signaling pathways [22–24]. However, the role of TSC2 in regulating NF-κB signaling in EC has not been studied. We addressed this question by asking if TSC2 directly interacts with RelA/p65-IκBα complex and whether this association plays a role in aiding RelA activation. Analysis of TSC2, RelA/p65 or IκBα immunoprecipitates each revealed that TSC2 is constitutively associated with RelA/p65-IκBα complex and this association is reduced upon stimulation with thrombin. These findings suggest that binding of TSC2 may facilitate the degradation of IκBα for efficient release of RelA/p65 for its translocation and DNA binding in the nucleus. Our data that disrupting TSC2 association with RelA/p65-IκBα via its depletion inhibited thrombin- or LPS-induced IκBα phosphorylation and degradation support this possibility. The impaired nuclear localization and DNA binding of RelA/p65 in TSC2-depleted cells further indicated a role of TSC2 association in promoting IκBα degradation-dependent RelA/p65 migration to the nucleus. Our studies, however, do not establish whether the binding between TSC2 and RelA/p65-IκBα is direct. Identifying the binding domains of TSC2, RelA/p65, and IκBα required for their interaction may lead to development of inhibitory peptides to disrupt the interaction between these proteins, and thereby suppress RelA/p65 activation and EC inflammation. Addressing these important questions requires another comprehensive study.

Another essential event controlling RelA/p65 activation involves phosphorylation of Ser⁵³⁶ in its transactivation domain which increases the transcriptional capacity of RelA/p65 [16, 34]. The RelA/p65 phosphorylation occurs in parallel with IκBα phosphorylation as both phosphorylation events are catalyzed by IKKα/β [14,34]. The signals transmitted from different proinflammatory stimuli binding to their respective receptors converge on IKK and activate it via phosphorylation of IKKα and IKKβ subunits at Ser¹⁷⁶ and Ser¹⁷⁷ respectively [3]. IKK phosphorylation of RelA/p65 not only increases the transactivation potential, but also lowers the affinity of RelA/p65 for IκBα and thus in concert with IκBα phosphorylation/degradation promotes the nuclear translocation and DNA binding of RelA/p65 [13]. Because TSC2 is required for IκBα phosphorylation/degradation, we investigated if TSC2 also regulates IKKα/β phosphorylation. We found that like IκBα phosphorylation, TSC2 silencing was effective in decreasing IKKα/β phosphorylation caused by thrombin or LPS. Similarly, RelA/p65 Ser⁵³⁶ phosphorylation by thrombin or LPS was also reduced in TSC2-depleted cells. These findings support a role of TSC2 in promoting IκBα and RelA/p65 phosphorylation via activation of IKKα/β.

The ability of TSC2 to regulate IκBα phosphorylation/degradation led us to examine the status of RelA/p65 nuclear translocation and DNA binding in TSC2-depleted cells. We noted that the loss of TSC2 was associated with impaired nuclear translocation of RelA/p65 after thrombin challenge. Similarly, thrombin-induced DNA binding of nuclear RelA/p65 was also diminished in the absence of TSC2. These results are consistent with a role of TSC2 in IKK-mediated IκBα phosphorylation/degradation-dependent migration and DNA binding of RelA/p65 in the nucleus. Given that TSC2 is also required for RelA/p65 phosphorylation that confers transcriptional competency to DNA-bound RelA/p65, we evaluated the role of TSC2 in controlling the transcriptional activity of RelA/p65. To this end, we monitored the effect of TSC2 knockdown on NF-κB-dependent reporter activity. A substantial decrease in thrombin-induced NF-κB-dependent reporter activity in TSC2-depleted cells confirmed the involvement of TSC2 in regulating RelA/p65 transcriptional activity. To further corroborate these results, we determined the effect of TSC2 knockdown on endogenous NF-κB-dependent gene expression. ICAM-1 and VCAM-1 represent two important genes whose expression is crucial for PMN adhesion and migration across endothelial barrier [3]. We found that depletion of TSC2 caused a significant decrease in ICAM-1 and VCAM-1 levels induced by thrombin or LPS both in HPAEC and HLMVEC. Similarly, thrombin- or LPS-induced production of IL-6, an important proinflammatory cytokine whose levels are elevated in septic patients with ALI [31, 32], was also reduced in TSC2-depleted cells. These findings indicate that TSC2 is a common mediator of inflammation both in macrovascular and microvascular EC in the lung.

Notably, TSC2 knockdown was more effective in inhibiting IL-6 production compared to VCAM-1 and ICAM-1 expression (Fig. 5). The differential effects of TSC2 knockdown could be the result of a more dominant role of NF-κB in IL-6 compared to VCAM-1 and ICAM-1 expression. Indeed, while NF-κB is essential for expression of these molecules, the maximal VCAM-1 or ICAM-1 expression requires participation of other transcription factors including AP-1, GATA, and Sp1 [35–38]. It is possible that TSC2 regulates only NF-κB but not AP-1, GATA, and Sp1 transcription factors and consequently partial inhibition of VCAM-1 or ICAM-1 expression in TSC-depleted cells.

The established role of TSC2 in suppressing MTORC1 activation and relaying majority of its effects via this mechanism [39] and our previous findings that MTOR limits thrombin-induced EC inflammation by restraining RelA/p65 activation [25] prompted us to determine if the proinflammatory action of TSC2 derives from its ability to inhibit MTORC1. Activation of MTORC1 results in the phosphorylation of its downstream substrate p70s6k [40]. Surprisingly, we found that depletion of TSC2 caused an increase in p70s6k both at mRNA and protein levels. However, it remains to be determined whether the increased p70s6k level contributes to the anti-inflammatory effect in TSC2-depleted cells. We found that loss of TSC2 resulted in MTORC1 activation as evidenced by an increase in activating phosphorylation of MTOR and a decrease in inhibitory phosphorylation of RAPTOR. However, silencing RAPTOR to prevent MTORC1 activation failed to relieve the suppressive effect of TSC2 knockdown on thrombin-induced VCAM-1 expression. These data argue against involvement of MTORC1 in TSC2 regulation of EC inflammation. Intriguingly, however, silencing RAPTOR alone produced a similar inhibitory effect on VCAM-1 expression as TSC2 knockdown in control cells. This is counterintuitive given the reported effect of MTOR knockdown in augmenting thrombin-induced EC inflammation [26]. The reasons for this discrepancy are not clear. One possibility could be that silencing RAPTOR alone may not fully inactivate MTORC1 and the residual activity may be sufficient to suppress EC inflammation. Another possibility is that RAPTOR-deficient MTORC1 may gain a new function that is proinflammatory in nature. Additional comprehensive studies are required to test these possibilities in a rigorous manner.

In summary, the present study identifies TSC2 as a critical component of RelA/p65-IκBα complex that serves to promote EC inflammation by facilitating the phosphorylation, nuclear translocation, and DNA binding activity of RelA/p65 secondary to its release from IκBα. The action of TSC2 in causing EC inflammation appears to be MTORC1-independent. Thus, the specific targeting of TSC2 may be a useful strategy for preventing EC inflammation associated with intravascular coagulation and sepsis.

Abbreviations

ALI

Acute Lung Injury

CST

Cell Signaling Technologies

DEAE

Diethylethanolamine

DMSO

Dimethyl Sulfoxide

EC

Endothelial Cell

EDTA

Ethylenediaminetetraacetic acid

EBM2

Endothelial Basal Medium 2

EGTA

Ethylene glycol-bis(β-aminoethyl ether)-N, N,N′,N′-tetraacetic acid

GAP

GTPase Activating Protein

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

HLMVEC

Human Lung Microvascular Endothelial Cell

HPAEC

Human Pulmonary Artery Endothelial Cell

HIF-1α

Hypoxia-Inducible Factor 1-alpha

IB

Immunoblotting

ICAM-1

Intercellular Adhesion Molecule-1

IKK

IκB kinase

IκBα

Inhibitor of κB-alpha

IP

Immunoprecipitation

IgG

Immunoglobulin G

kDa

Kilodalton

LPS

Lipopolysaccharides

MTOR

Mechanistic Target of Rapamycin

MTORC1

Mechanistic Target of Rapamycin Complex 1

MW

Molecular Weight

NLS

Nuclear Localization Signal

NF-κB

Nuclear Factor-Kappa B

PAR1

Proteinase-Activated Receptor 1

PMN

Polymorphonuclear Leukocytes

RAPTOR

Regulatory Associated Protein of mTOR

RIPA

Radioimmunoprecipitation

SDS-PAGE

Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis

Ser

Serine

siCon

siControl

TBST

Tris-buffered saline + Tween 20

TBP

TATA Binding Protein

TL

Total Lysate

TSC1

Tuberous Sclerosis Complex 1

TSC2

Tuberous Sclerosis Complex 2

TLR

Toll-like receptor

VCAM-1

Vascular Cell Adhesion Molecule-1

VEGF

Vascular Endothelial Growth Factor

Authors’ contributions

AR and IT conceived the study. AR and IT designed the research studies. IT and RAN conducted the experiments. IT and AR wrote the manuscript. All authors read and approved the submission of this manuscript.

Funding

This work was supported by NHLBI grants HL168872, HL148695 and NIGMS grant GM130463. This work was also supported by T32 Pulmonology Training grant T32HL171029 and NIH Center Grant P30ES001247.

Data availability

The data supporting this study’s findings are available within the main text.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.