Physical interaction between heterologously expressed TRPP2 and STIM1 in HEK293 cells
Both TRPP2 and STIM1 are expressed in the ER membrane. To test if TRPP2 interacts with STIM1, C-terminalGFP-tagged TRPP2 (TRPP2-GFP) was co-expressed with N-terminal mCherry-tagged STIM1 (mCherry-STIM1) in HEK293 cells. Double fluorescent imaging showed that TRPP2-GFP colocalized very well with mCherry-STIM1 and ER-DsRed (an ER marker) in the transfected cells (Fig. 1a). Moreover, reciprocal co-immunoprecipitation assays further demonstrated that TRPP2 interacted with STIM1 in the transfected cells (Fig. 1b).
Next, we used sensitized emission FRET, a powerful tool to identify protein-protein interactions [27], to probe if TRPP2 and STIM1 directly interact with each other. Because both the N- and C-termini of TRPP2 as well as the C-terminus of STIM1 are all located at the cytoplasmic side of the ER membrane, we reasoned that if the two proteins interact, one of the cytoplasmic termini of TRPP2 should be in close proximity with the STIM1 C-terminus. To test this, mCherry-tagged TRPP2 with mCherry fused either at the N- or the C-terminus of TRPP2 (mCherry-TRPP2 and TRPP2-mCherry) were individually co-expressed with STIM1-GFP in HEK293 cells. Cells that expressed GFP-mCherry fusion protein were used as positive control, whereas those that co-expressed GFP and mCherry as separate proteins were used as negative control for the FRET experiments. Remarkably, the co-expression of mCherry-TRPP2 and STIM1-GFP led to a high FRET efficiency, to approximately a half of that achieved by the GFP-mCherry fusion protein. By contrast, the co-expression of TRPP2-mCherry and STIM1-GFP produced a very low FRET efficiency as that of GFP and mCherry (Fig. 2a,b). These results suggest that the N-terminus, but not C-terminus, of TRPP2 is closely associated with STIM1 C-terminus. To map the TRPP2 region mediating association with STIM1, we linked each of a series of TRPP2 N-terminus truncation derivatives with GFP tag (Fig. 2c) for Co-IP assay. As show in Fig. 2d, full-length TRPP2 and its delete N2 (112-221aa, ∆N2) truncated but not delete N1 (2-111aa, ∆N1) truncated derivatives bound strongly to STIM1, suggesting that the N1 domain of TRPP2 is responsible for the interaction.
Role of TRPP2-STIM1 complex in SOCE
It is well established that STIM1 senses ER Ca2+ store depletion to activate SOCE [15]. Since TRPP2 interacts with STIM1, it may also play a role in SOCE. To test this hypothesis, we first used small interface (si)RNA to knock down TRPP2 expression in HEK293 cells. As a control, we also used siRNA for STIM1. The TRPP2 and STIM1 specific siRNAs markedly suppressed the expression of endogenous TRPP2 and STIM1, respectively, in HEK293 cells (Supplementary Fig. 1). We then examined SOCE in the siRNA-transfected cells. Fluo-8AM loaded cells were treated with thapsigargin (TG, 2.5 μmol/L), an inhibitor of sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) that induces passive ER Ca2+ store depletion via both IP3R-dependent and independent pathways [33], in the absence of extracellular Ca2+ for 8 min and then Ca2+ (1 mmol/L) was reintroduced to the bath solution. Consistent with the specific role for STIM1 in Ca2+ entry, STIM1 siRNA strongly suppressed the increase in intracellular Ca2+ concentration ([Ca2+]i) upon reintroduction of extracellular Ca2+ (Ca2+ entry) but did not alter the TG-evoked [Ca2+]i rise in the absence of extracellular Ca2+ (Ca2+ release) (Fig. 3a,b). On the other hand, the overexpression of STIM1 moderately increased the TG-evoked Ca2+ entry without affecting Ca2+ release (Fig. 3a,b). The transfection of TRPP2 siRNA, however, caused significant decreases in both the Ca2+ release and Ca2+ entry induced by TG (Fig. 3c,d), with the effect on Ca2+ entry far less pronounced than that caused by STIM1 siRNA. The inhibitory effects of TRPP2 siRNA on Ca2+ release and entry were both rescued by co-expression of wild type TRPP2, but not its dominant negative mutant, D511V[34] (Fig. 3c,d). The overexpression of TRPP2, either in the absence or presence of STIM1 (endogenous or exogenous), did not alter the TG-induced Ca2+ release or Ca2+ entry (Fig. 3a-d). These results indicate that unlike STIM1, TRPP2 mainly exerts effect on Ca2+ release from the ER store, which then affects SOCE.
Studies from other groups have shown that adenosine triphosphate (ATP) activates TRPP2 indirectly via IP3-induced Ca2+ release [9]. Thus, ATP (10 μmol/L) was used in place of TG to induce SOCE. Remarkably, the transfection of TRPP2 siRNA significantly suppressed ATP-induced Ca2+ release and the Ca2+ entry was diminished to a similar degree as that caused by STIM1 siRNA, although the latter exhibited no effect on ATP-induced Ca2+ release (Fig. 3a -d). Again, the co-expression of wild type TRPP2, but not the dominant negative D511V mutant, rescued the suppressed ATP-induced Ca2+ release and entry in the TRPP2 siRNA-transfected cells (Fig. 3c,d). The overexpression of TRPP2 in the absences of TRPP2 siRNA even enhanced the ATP-induced Ca2+ release and the subsequent Ca2+ entry (Fig. 3c,d). The enhancing effect on ATP-induced Ca2+ release was evident even with the overexpression of STIM1; however, for Ca2+ entry, the TRPP2-mediated increase was not obvious because STIM1 alone already had a similar effect.
Supporting the role of IP3Rs in ATP-induced Ca2+ release, IP3R inhibitor, 2-aminoethyl diphenylborinate (2APB) (100 μmol/L), abolished the ATP-induced Ca2+ release and Ca2+ entry under all conditions tested (Fig. 3a-d). Although 2APB may also inhibit SOCE and/or other nonspecific targets [35], the lack of any ATP-induced Ca2+ release in the presence and absence of TRPP2 is consistent with the idea that TRPP2 activation in the ER is triggered by IP3R-mediated Ca2+ release [9].
Next, we sought to suppress the TRPP2-STIM1 interaction in the ER. In above study, we have proved that TRPP2 ∆N1 has a weak interaction with STIM1. Therefore, we used TRPP2 ∆N1 as a dominant negative mutant for TRPP2-STIM1 interaction to identify the functional role of TRPP2-STIM1 complex in the SOCE using live Ca2+ fluorescence measurement. The results of [Ca2+]i measurement suggested that SOCE of TRPP2 ∆N1 and STIM1 co-transfected cells was remarkably reduced compared to TRPP2 ∆N2 and STIM1 or whole TRPP2 and STIM1 co-transfected HEK293 cells (Fig. 3e).
These results indicate that TRPP2 associates with STIM1 to crucially regulate SOCE process and TRPP2-STIM1 complex has a pivotal role in IP3-mediated Ca2+ signaling.
Functional role of TRPP2-STIM1 complex in SOCE in VSMCs
Both TRPP2 and STIM1 are expressed in VSMCs [36, 37]. Hypothesizing that endogenous TRPP2 physically interacts with STIM1 in VSMCs. Reciprocal co-immunoprecipitation assay showed that endogenous TRPP2 and STIM1 pulled down each other in the mouse aortic VSMCs (Fig. 4a,b). Additionally, to confirm the co-localization of TRPP2 and STIM1, we utilized a powerful method PLA, which detects proteins located within a radius of <40 nm. In the presence of both anti-TRPP2 and anti-STIM1 antibodies, red fluorescent dots indicated a positive signal of PLA in fixed fresh isolated VSMCs (Fig. 4c: (d)-(f)). A negative control incubating with anti-STIM1 antibody alone displayed a negligible number of fluorescent dots (Fig. 4c: (a)-(c)). These results suggest that TRPP2 indeed very colocalizes with STIM1 in mouse aortic VSMCs.
To identify the function of TRPP2-STIM1 complex in the homeostasis of [Ca2+]i in VSMCs, TRPP2 and STIM1 siRNAs were used to suppress TRPP2 and STIM1 expression in the primary cultured mouse aortic VSMCs (Supplementary Fig. 2a,b). TRPP2 siRNA did not affect Orai1, STIM1 and IP3R expression (Supplementary Fig. 2c). The [Ca2+]i measurement results showed that TRPP2, STIM1 or TRPP2+STIM1 siRNA transfection strongly suppressed the ATP-induced SOCE as compared to scrambled siRNA control in the primary cultured VSMCs (Fig. 5a,b). STIM1 siRNA with or without TRPP2 siRNA transfection markedly decreased the TG-induced SOCE but TRPP2 siRNA merely moderately suppressed the TG-induced SOCE as compared to scrambled siRNA control (Fig. 5b). The transfection with TRPP2 siRNA alone or together with STIM1 siRNA significantly suppressed the ATP- and TG-induced Ca2+ release (Supplementary Fig. 3). Furthermore, the ATP-induced SOCE and Ca2+ release were abolished by 2APB (Supplementary Fig. 3).
It has been well-documented that the Ca2+ store depletion will evoke STIM1 puncta formation which then activates Orai1 channels. The immunofluorescent experiments demonstrated that TRPP2 siRNA dramatically suppressed ATP-induced STIM1 puncta formation in the primary cultured VSMCs (Fig. 5c). 2APB abolished the ATP-induced STIM1 puncta formation in both groups (Fig. 5c). Meanwhile, no significant difference in TG-induced STIM1 puncta formation was found between scrambled siRNA control and TRPP2 siRNA transfection (Fig. 5c). The data indicated that the TRPP2-specific regulation of the STIM1 puncta formation was IP3R dependent.
Therefore, these results indicate that TRPP2 and STIM1 associate together in VSMCs involving agonist-induced SOCE.
Functional role of TRPP2-STIM1 association in agonist-induced contraction in endothelium denuded mouse aorta
The activation of G protein-coupled receptors (GPCRs) causes phospholipase Cβ to convert phosphatidylinositol 4,5-biphosphate into IP3 and diacylglycerol [38]. The IP3 activates TRPP2 indirectly via the Ca2+ release from IP3R [9]. Certain GPCRs are able to increase the [Ca2+]i via evoking the Ca2+ release and SOCE to contract VSMCs. Therefore, the TRPP2-STIM1 association is potentially important in the VSMCs contraction and blood vessel tone regulation. To identify the functional role of TRPP2-STIM1 complex, the tension of mouse aorta was measured. Unfortunately, no chemicals can specifically inhibit TRPP2 or STIM1. Thus, we used organ culture to transfect TRPP2 or STIM1 siRNA into isolated mouse aorta to suppress TRPP2 or STIM1 expression [11]. Immunofluorescence showed that TRPP2 and STIM1 specific siRNAs dramatically suppressed TRPP2 and STIM1 protein expression in the cultured mouse aorta (Supplementary Fig. 4).
In the tension measurement, the endothelial layer of mouse aorta was removed for specifically investigating the VSMC contraction. The results showed that phenylephrine (α receptor agonist, Phe, 10 μmol/L), endothelin 1 (endothelin receptor agonist, ET-1, 100 nmol/L) and TG (2.5 μmol/L) induced the vessel contraction in the Ca2+-free solution due to the Ca2+ release from the Ca2+ store (Fig. 6, Supplementary Fig. 5). Subsequent re-addition of 2.5 mmol/L Ca2+ into bath solution induced further contraction because of the SOCE (Fig. 6, Supplementary Fig. 5). Interestingly, TRPP2 siRNA markedly suppressed Phe- and ET-1-induced contraction but slightly inhibited TG-induced contraction as compared to scrambled siRNA controls in the Ca2+-free solution (Fig. 6a,b,f, Supplementary Fig. 5a,b). Moreover, TRPP2 siRNA markedly suppressed SOCE-induced contraction in Phe, ET-1 and TG treatments (Fig. 6a,c,g, Supplementary Fig. 5a,c). On the other hand, STIM1 siRNA significantly decreased the Phe-, ET-1- and TG-induced contraction in the Ca2+-free solution (Fig. 6d,h, Supplementary Fig. 5d) and the SOCE-induced contraction as compared to scrambled siRNA controls (Fig. 6e,i, Supplementary Fig. 5e).
To further investigate the involvement of the TRPP2-STIM1 association in the agonist-induced vessel contraction, Phe- and ET-1-induced dose-dependent contractions between TRPP2, STIM1 and scrambled siRNA transfection in mouse aorta were compared. The data indicated that TRPP2 or STIM1 siRNA separately suppressed the Phe- and ET-1-induced dose-dependent contraction in normal Krebs’ solution as compared to respective scrambled siRNA controls (Fig. 7a,c,e,g). More importantly, after the pretreatment of IP3R antagonist heparin (1 mg/mL), the differences between TRPP2 or STIM1 siRNA and respective controls were abolished (Fig. 7b,d,f,h), further supporting the notion that TRPP2 regulates the store Ca2+ release and SOCE necessary for the agonist-induced contraction, which needs IP3R activation.
Conditional knockout mouse was a very powerful tool for investigating the functional role of TRPP2 in agonist-induced vessel contraction. To create TRPP2 CKO mouse, two LoxP sites were inserted between PKD2 exon (Fig. 8a). Tagln-Cre mice were used to cross with PKD2 LoxP mice to specially delete PKD2 gene in smooth muscle and heart muscle tissues. The genotyping was used to identify TRPP2 CKO mice (Fig. 8b). Immunoblots data indicated that TRPP2 protein was not expressed in TRPP2 CKO mice aortic smooth muscle cells compared to Tagln-Cre control mice (Fig. 8c). Phe- and ET-1-induced vessel contractions were significantly decreased in TRPP2 CKO mice compared to Tagln-Cre control mice (Fig. 8d,e).
Taken together, these results very strongly supported our finding that TRPP2 was a crucial component participating in agonist-induced VSMCs contraction and blood vessel tone.