Cells

Primary rat ventricular cardiomyocytes (RVCM) were obtained from 3 to 5 day old Sprague Dawley (Scanbur, Sollentuna, Sweden) and cultured on 18 mm diameter coverslips for 5 days previously described [14] using a modified growth medium. Growth medium was either a 2:1 mixture of DMEM/F-12:PC-1, supplemented with 2.5% FBS and 0.05 pM of AngII or DMEM for primary cell isolation (Gibco), 1:1000 Cardiomyocyte Growth Supplement (Pierce), 10% FBS and 50 pM AngII. Rats were euthanized by rapid decapitation and the heart removed for generation of cardiomyocyte cultures. Quality of culture was determined using a cardiomyocyte characterization kit (Chemicon). Cardiomyocytes were cultured for 5 days prior to experiment and contracting clusters, seen with transmission light, were selected for recording. Expression of cardiomyocyte markers were confirmed using Troponin I (Chemicon) and Desmin (Chemicon) antibody staining according to manufacturer’s protocol. Rat aortic smooth muscle cells (ASMC, catalog number R6110, 3H Biomedical, ScienCell) were cultured according to manufacturer’s instruction. Briefly, cells were thawed and plated on poly-L-lysine coated coverslips in complete smooth muscle cell medium (SMCM, ScienCell) including 2% FBS and supplemented with 0.05 pM of AngII. Cells were cultured for three to 5 days before experiments, cells from initial plating and first passage were used. Cells were confirmed to express anti-α-smooth muscle actin by immunostaining (antibody from Abcam). Human Embryonic Kidney 293a (HEK, catalog number R70507, Thermo Fisher Scientific) cells were plated on 18 mm diameter coverslips 2 days prior to transfection with Exgen500 (Fermentas), experiments were performed 24 h after transfection. Manufacturers protocols were used for transfection using 1 μg DNA per transfection. HEK cells between passage 3 and 20 was used in experiments. RT-PCR was used to confirm RNA expression of Ca_V1.2 (Fig 2a).

All experiments were performed according to Karolinska Institutet regulations concerning care and use of laboratory animals, and were approved by the Stockholm North ethical evaluation board for animal research.

Reagents and constructs

All chemicals and drugs were obtained from Sigma-Aldrich, if not otherwise stated. Bradykinin was purchased from Abcam. 3,5-bis(trifluoromethyl)pyrazole (BTP2) a TRPC channel blocker was obtained from Santa Cruz. Rat angiotensin II type 1 receptor, AT₁R, with Venus fused at C-terminus was used for Ca²⁺ experiments [15]. The HA-tagged AT1R was made by insertion of PvuI restriction site through mutagenesis in the first extracellular loop between Pro331 and Phe332. Agilent QuikChangeII Site-Directed Mutagenesis kit together with the following primers were used: sense 5′ – GGTGATTGCCGAACGATCGGGGCCAGCGGTAC and antisense 5′ GTACCGCTGGCCCCGATCGTTCGGCAATCACC. The product was digested with PvuI (Thermo Scientific), and RSYPYDVPDYARS (Hemaglutinin flanked by PvuI sites) was inserted through ligation (Phusion Hot Start II, ThermoFisher). Structure of the construct was verified by using BigDye® Terminator V3.1 Cycle Sequencing Kit (Applied Biosystems). pGβ-2A-YFP-Gγ2-IRES-GαqmTq [16] was kindly provided by the lab of Th. W. J. Gadella and used for Gαq-Gβγ FRET (Förster Resonance Energy Transfer) measurements.

Membrane recruitment

Control and cells exposed to AngII were fixed using 2% paraformaldehyde for 10 min and stained using 1:1000 dilution of Alexa-594 fluorescence conjugated hemagglutinin (HA) antibody (Invitrogen). Cells were imaged using a Zeiss LSM780 in two channels for Venus and Alexa-594 fluorescence. Mean membrane and cytosol fluorescence intensities were calculated. Membrane fraction is reported as normalized to control ratio between mean membrane to cytosolic intensities.

PCR

Approximately 10⁷ HEK cells were used for RNA extraction using manufacturer’s protocols (RNeasy Plus, Qiagen). cDNA was obtained using iScript cDNA kit (BioRad) following manufacturers protocols. Primers for Ca_V1.1, Ca_V1.2 and Ca_V1.4 were designed and used for PCR amplification (Phusion High-Fidelity DNA polymerase, Thermo Scientific). PCR amplification was run using the manufacturer’s protocols with the following program changes: 40 cycles with 10s extension time. Annealing temperature was 56 °C for Ca_V1.4, 58 °C for Ca_V1.1 and Ca_V1.2. The product was run on a 2% Agarose gel (1:10,000 GelRed, Biotium) and imaged on a Li-Cor Oddyssey system. The following primer pairs were used at 0.5 μM:

CaV11_4689F CAAGAGGAGTATTATGGCTATCGG,

CaV11_5252R GGTCAGCAGTCCCTTCAGCA,

CaV12_6568F CTGCGACATGACCATAGAGG,

CaV12_7188R CAGCGAGGCTCGTACAGA,

CaV14_5045F CAAAGCCACGATGGTCTCCC and.

CaV14_5545R TTTCGCCCACGATGATAGG (Cybergene).

[Ca²⁺]_i measurements

Intracellular Ca²⁺ concentration ([Ca²⁺]_i) measurements carried out in HEK cells were performed using the fluorescent Ca²⁺ indicator Fura Red (Invitrogen). Cells were loaded with Fura Red using a buffer consisting of 5 μM Fura Red and 1 μl of 20% Pluronic F-127 (Invitrogen) diluted in KREBs (110 mM NaCl, 4 mM KCl, 1 mM NaH₂PO₄xH₂O, 25 mM NaHCO₃, 1.5 mM CaCl₂x2H₂O, 1.25 mM MgCl₂x6H₂0, 10 mM Glucose and 20 mM HEPES, pH 7.40, sterile filtered). Cells grown on #1.5 18 mm diameter coverslips (Warner Instruments) were rinsed with KREBs buffer and loaded for 45 min with the loading buffer at 37 °C with 5% CO₂. After 45 min cells were washed again with KREBs prior to being mounted in a perfusion chamber (Chamlide). Through the use of heated optical table and in-line perfusion feed heater the cells were kept at 37 °C throughout experiments. Loading protocol was evaluated by testing different loading times, both at room temperature and in incubator. The minimal dye concentration usable was identified through Ca²⁺ calibration of serial dilution of dye.

Ca_V1.2 inhibitors, nifedipine and verapamil, were both mixed fresh for each experimental day. Solutions were kept in fridge and protected from light. We chose to use a high concentration of nifedipine, 100 μM, to compensate for loss of active drugs during the longer experiments. Degradation was unavoidable within each experiment with cells and solutions kept at 37 °C. UV and 488 nm laser excitation also contributed to some drug degradation throughout experiments.

[Ca²⁺] calibration was done by the following protocol, first a 3 min long washout of the KREBs buffer using a 0 [Ca²⁺] containing buffer (110 mM NaCl, 4 mM KCl, 1 mM NaH₂PO₄xH₂O, 25 mM NaHCO₃, 1.25 mM MgCl₂x6H₂0, 250 μM EGTA, 10 mM glucose and 20 mM HEPES, pH 7.40, sterile filtered) followed by 1 μM ionomycin in 0 [Ca²⁺] buffer for 4 min. Calibration ended with 7 min of 1.5 mM [Ca²⁺] KREBs buffer perfusion.

Ca²⁺ measurements of RVCM were performed using the fluorescent Ca²⁺ indicator Oregon Green BAPTA-1 (Invitrogen) with the same protocol as described for Fura Red. Oregon Green BAPTA1 was chosen as its weaker binding affinity compared to Fura Red made it insensitive to background Ca²⁺ sparklets in RVCM. Calibration was not possible for the RVCM Ca²⁺ experiments, thus the non-calibrated Ca²⁺ response for each cell was obtained by the mean intensity within each selected cell.

Nifedipine effects on bradykinin activation of bradykinin B2 receptors were evaluated through comparing signaling strength with or without pretreatment of 100 μM nifedipine on 100 nM bradykinin (Abcam) treatment. Experiments were performed on Oregon Green Bapta 1 loaded HEK cells. Cells were loaded in the same manner as for Fura Red but with 5 μM Oregon Green BAPTA-1. All other parameters were kept identical as for the 1 nM AngII experiments described above. Each experiment was calibrated as per the protocol described above.

Peak amplitudes were in all experiments calculated as peak above baseline. Baseline was defined as the mean intensity of a stable part of the trace prior to the peak. This calculation reflects the cytosolic increase in concentration rather than the absolute concentration.

Gαq dissociation

HEK cells expressing AT₁R and Gαq-Gβγ FRET sensor were exposed to 1 nM AngII with or without 3 min pretreatment of 100 μM nifedipine. FRET was measured using a Zeiss LSM 510 LIVE microscope with a 40×/1.2 water immersion objective and excitation of 405 nm. YFP signal was corrected for mTurquoise bleedthrough and used as FRET intensity. Transfecting HEK cells with GαqmTq allowed for determination of donor bleed through into FRET channel. Bleed through coefficient, b, was calculated as I_mTq/I_FRET. In experiments, FRET channel intensities were then corrected as: FRET_corrected = I_FRET – b*I_mTq. FRET ratios were finally calculated as I_mTurquoise / I_{FRETcorrected}. Increase in FRET ratio corresponds with increased G_q-protein dissociation.

Statistical analysis

Statistical analysis was performed using Matlab, one way Anova analysis and subsequent multiple comparison tests based on studentized range distribution. Quantified data is shown as mean value +/− SEM.

AT₁R desensitization is dose-dependent

The majority of studies on AT₁R signaling have used AngII concentrations that are at least three orders of magnitude higher than the circulatory levels [1, 2, 17, 18], which range between 20 and 100 pM [5]. Repeated exposure to non-physiological AngII concentrations have consistently been shown to result in desensitization of the signaling response. Rat renal interstitial fluid has been reported to reach 3-5 nM AngII under pathological conditions [19–21]. Here we have compared the effect of doses of 1 nM and 100 nM AngII on the calcium signal from HEK cells expressing Venus tagged AT₁R and loaded with the Ca²⁺ sensitive dye Fura Red. Cells were exposed to three consecutive AngII doses, each 2 min long. The time delay between the 1st and 2nd exposure was 5 min, a common timescale for studies of desensitization [6, 17]. The 3rd exposure to AngII was performed 30 min after the 2nd exposure; a delay selected to test recovery of receptor signaling. Representative [Ca²⁺]_i responses to serial AngII exposure are shown in Fig. 1a. The 1st exposure to 100 nM AngII resulted in a seven-fold stronger [Ca²⁺]_i, response relative the 1st exposure to 1 nM AngII (Fig 1b). The 2nd exposure to 100 nM AngII shows a significant desensitization. In contrast, no desensitization was observed following the 2nd exposure to 1 nM AngII. A partial recovery of the [Ca²⁺]_i response to 100 nM AngII was observed following the 3rd exposure. The [Ca²⁺]_i response to the 3rd exposure to 1 nM AngII was significantly stronger than the [Ca²⁺]_i response following the 1st exposure (Fig 1b). Application of cyclopiazonic acid, an inhibitor of the sarcoendoplasmic reticulum calcium transport ATPase, completely abolished the [Ca²⁺]_i response to AngII, indicating that the increase in [Ca²⁺]_i is mainly due to release of Ca²⁺ from intracellular stores (data not shown).

Next we examined to which extent the dose-dependent [Ca²⁺]_i response to AngII could be attributed to effects on AT₁R plasma-membrane expression. For this purpose, we expressed AT₁R with an extracellular hemagglutinin (HA) tag in addition to a Venus tag at the intracellular C-terminus of HEK cells. Cells were fixed and labeled with an Alexa-594 conjugated anti-HA antibody 3 or 30 min after exposure to AngII. The AT₁R fraction inserted in the plasma membrane was calculated as the ratio between the fluorescence intensity of Alexa-594 (plasma membrane AT₁R) and Venus (total AT₁R) (Fig 1c). The effect of repeated exposure to 1 and 100 nM AngII on the fraction of AT₁R in the plasma membrane were subtle and did not parallel the changes in Ca²⁺ response.

AT₁R signaling is upregulated by nifedipine and verapamil

Drugs based on either the nifedipine or verapamil structure, are inhibitors of the L-type VGCC, Ca_V1.2 subtype [22], and are commonly used in the treatment of hypertension. Since there is emerging evidence of a cross-talk between the Ca_V1.2 and AT₁R [8, 23, 24], we tested if nifedipine and verapamil would affect the response to AngII in AT₁R expressing HEK cells. Ca_V1.2 is widely expressed, and the RNA expression of Ca_V1.2 in the HEK cells was confirmed using RT-PCR (Fig 2a). HEK cells were pretreated for 5 min with 1 μM nifedipine or 2 μM verapamil and then exposed to 1 nM AngII for 2 min (Fig. 2b-c). The concentration of nifedipine [25] and verapamil [26] were selected based on human plasma levels. Unexpectedly, we found that in the presence of either drug, there was a robust and highly significant increase of the calcium response to the initial dose of 1 nM AngII in both nifedipine and verapamil treated cells (Fig. 2b and c) To ensure that complete inhibition of Ca_V1.2 was achieved, as nifedipine is sensitive to heat and light (supplier’s information), we also tested a higher dose, 100 μM nifedipine (Fig 2b). No further increase in response was observed. To eliminate possible errors from drug degradation during the one-hour long protocol we elected to continue with the higher dose of nifedipine.

The effect of nifedipine on the [Ca²⁺]_i response to 1 nM AngII was studied using the same three-pulse protocol as described above with the addition of a pretreatment with 100 μM nifedipine (Fig. 2d and e). There was a subsequent reduction of the [Ca²⁺]_i response to the 2nd and 3rd dose of 1 nM AngII, suggesting a certain desensitization. There was no increase in the relative distribution of AT₁R inserted in the plasma membrane in nifedipine pretreated cells compared to control cells (data not shown).

We next asked if the nifedipine potentiation of the response to 1 nM AngII could be attributed to increased G-protein activity. To address this question we examined the dissociation of G_αq and G_βγ subunits induced by AT₁R activation using a Gαq-Gβγ FRET sensor [16] (Fig 3a). Increased FRET ratio indicates increased G_q-protein dissociation. The FRET ratio between GαqmTq (donor) and YFP-Gγ2 (acceptor) in response to 1 nM AngII was studied in the presence or absence of nifedipine. No significant difference in FRET ratio was observed between nifedipine treated and control cells (Fig. 3b and c). This suggests that up-regulation of AT₁R in nifedipine pretreated cells is not due to increased G-protein activation.

AT₁R signaling is upregulated by inhibition of TRPC channels

We examined whether the potentiation of the AngII signal observed following Ca_V1.2 inhibition may also occur following the inhibition of other calcium entry pathways. The TRPC channels are ubiquitously expressed ion channels with high calcium permeability. It was recently reported that the response to AngII was enhanced in arterial myocytes from TRP deficient mice [27]. Here we used BTP2 to inhibit the activity of TRPC channels. BTP2 blocks the calcium influx via TRPC3 and TRPC5 (but not transient receptor potential cation channel subfamily V6) with an IC50 of ~0.1-0.3 μM, and is generally considered to have a many-fold greater potency than other available BTP2 inhibitors [28, 29]. HEK cells were first exposed to 1 nM AngII. This was followed by 5 min treatment with either vehicle or 10 μM BTP2. The cells were then exposed to a second dose of 1 nM AngII (Fig 4a). The relative change in response between first and second AngII exposure was calculated, and showed a significant potentiation of the AngII following BTP2 treatment (Fig 4b).

To test whether the sensitivity for calcium influx observed for the AT₁R cells is a general phenomenon for GqPCR, we also tested if nifedipine exposure would potentiate [Ca²⁺]_i signaling from the bradykinin receptor B2 (B2 receptors), which are constitutively expressed in HEK cells. Exposure of cells to 100 nM bradykinin resulted in similar [Ca²⁺]_i response as for 100 nM AngII while in the presence of nifedipine a weaker [Ca²⁺]_i response was observed suggesting that not all GqPCRs are potentiated by nifedipine.

AT₁R signaling in cardiovascular cells is upregulated by VGCC inhibition

The finding that AT₁R signaling is enhanced following inhibition of Ca²⁺ influx may have implications for AngII regulation of cardiovascular function, since AT₁R inhibitors are commonly used to treat cardiovascular disease [30, 31] and Ca_V1.2 is abundantly expressed throughout arterial and cardiac tissue [32]. We therefore tested if our findings in HEK cells could be reproduced in rat derived primary cultures of both aortic smooth muscle cells (ASMC) and ventricular cardiomyocytes (RVCM).

ASMC at three to 5 days in vitro expressing alpha-smooth muscle actin were used for these experiments. Cells were loaded with the Ca²⁺ sensitive dye Oregon Green BAPTA1 and exposed to two consecutive physiological doses of AngII thirty minutes apart. Typical [Ca²⁺]_i response to 1 nM AngII exposures are shown in Fig. 5a (upper trace). A weaker response was observed at the 2nd exposure, in contrast to what had been found in HEK cells (Fig. 1a-b). We next tested if blocking of Ca_V1.2 by nifedipine or verapamil would affect the AngII response in ASMC. Nifedipine and verapamil doses were selected based on blood plasma levels reported in patients [33, 34]. ASMC were exposed to 1 nM AngII for 2 min and then allowed 15 min recovery followed by 15 min treatment with vehicle, nifedipine or verapamil and finally exposed to 1 nM AngII for 2 min. Example traces, of the [Ca²⁺]_i response in presence of nifedipine and verapamil are shown in Fig. 5a (middle and lower trace). The ratio between the second and first AngII response was calculated. Both nifedipine and verapamil treatment were found to strongly (2-fold) enhance the [Ca²⁺]_i response to AngII (Fig 5b) in ASMC.

The same protocol was used to study the effects of VGCC inhibition on AT₁R signaling in RVCM. RVCM at 5 days in vitro expressing the cardiomyocyte marker Troponin I and displaying contractile activity were used for these experiments. A representative [Ca²⁺]_i response trace to 1 nM AngII exposure is shown in Fig. 5c (upper trace). A stronger response was observed at the 2nd exposure, similar to what had been found in HEK cells (see Fig. 1a-b) but in contrast with ASMC response (see Fig 5b). Blocking of Ca_V1.2 by nifedipine (Fig. 5c middle trace) or verapamil (Fig. 5c bottom trace) were both found to strongly enhance the response to AngII (Fig 5d). Notably, the enhancement by verapamil was significantly stronger than for nifedipine treatment in RVCM but not in ASMC (Fig 5b).