All mice were housed and handled in a pathogen-free facility of College of Pharmacy at Kyung Hee University in compliance with the Guide for the Care and Use of Laboratory Animals, fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. The mice were maintained on a 12-hour light/dark cycle at a constant temperature with free access to food (LabDiet 5L79, St. Louis, MO, USA) and water. Every effort was made to minimize the number of sacrificed animals and their suffering. Animals were anesthetized by intraperitoneal (i.p.) injection of tribromoethanol (18 ml of the working solution per kg body weight), of which the working solution was diluted 40-fold in 0.9% NaCl from the stock solution (10 g 2,2,2, tribromoethanol dissolved in 10 ml tertiary amyl alcohol) and euthanized by removing the heart. The experiment was approved by the local committee for the Care and Use of Laboratory Animals, College of Pharmacy Institutional Animal Care and Use Committee (license number: KHP-2010-04-06).

Female BALB/c mice (Japan SLC, Inc., Shizuoka, Japan) were used at age 7–9 weeks for experiment. Csk knockout heterozygote mice, B6.129S-Csk^tm1Sor/J, were purchased from The Jackson Laboratory (Bar Harbor, ME). Heterozygote and wild-type homozygote pairs were mated, and the progeny was genotyped by PCR with the primer set using the wild-type (5’-CGCAGTCTACGAGGTGATGA-3’), the mutant (5’-CCTTCTATCGCCTTCTTGACG-3’) and the common reverse primer (5’-GGGCTCAGTTCAAGTTCAGG-3’).

The selection and in vivo delivery of siRNA have been previously described [19–21]. In brief, more than 3 siRNAs per gene (Csk, Ulk3, and Cyp1a2) were synthesized by Genolution (Seoul, Korea), and only one of them was selected for in vivo injection. To determine their silencing efficacy, 20 nM of the siRNAs was transfected into B16F10 cells and NIH3T3 cells using G-Fectin (Genolution, Seoul, Korea) and Lipofectamine 2000 (Invitrogen, Carlsbad, CA) respectively according to their manufacturer’s instructions. TRP53 siRNA was used as a positive control for all in vitro transfection experiments. The target siRNAs and scrambled control siRNA sequences are shown in S4 Table.

For the in vivo delivery into mice, polyethylenimine called as in vivo-jetPEI™ (Polyplus, 201-10G, Illkirch-Graffenstaden, France) was used as the transfection reagent. According to the manufacturer’s instruction, 50 μg of siRNA and 6.5 μl of in vivo-jetPEI (N/P charge ratio of 6) were diluted with 50 μl of 10% glucose solution and 50 μl of sterile H₂O. The solution was vortexed gently and left for 15 min at room temperature. The mixture was injected into the tail veins of 7–9-week-old, female BALB/c mice, and 24 hours after the last injection, the treated mice were used for experiments (i.e., blood pressure measurement, collection of tissues for mRNA quantitation or for Western blotting analysis). The 1-, 2- or 3-time injection means once per day for 1, 2, or 3 days.

Total RNA was extracted from the tissues of siRNA-injected mice 24 hours after the last injection using TRIzol (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized from 500 ng of total RNA using the PrimeScript™ RT kit (TaKaRa, Shiga, Japan) according to the manufacturer’s protocol. Quantitative real-time PCR analysis was performed using SYBR Green I (TaKaRa, Shiga, Japan) on the ABI Step One Real-Time PCR system (Applied Biosystems, Foster, CA) using the following program: 45 cycles at 95°C for 10 s, 60°C for 15 s, and 72°C for 20 s. The primer sequences are shown in S5 Table.

To present the data as the fold-change of relative expression of “case” (e.g. target siRNA-treated, Csk^+/-) over “control” (e.g. control siRNA-treated, Csk^+/+), the 2^-∆∆Ct method was used as previously described [22]. First, the Gapdh Ct value was subtracted from each target gene Ct value to obtain the delta Ct (∆Ct) value for normalization and the ∆Ct of the “control” sample was averaged. Second, the averaged ∆Ct value of the “control” sample was subtracted from each ∆Ct value of the “case” sample to obtain the delta delta Ct (∆∆Ct = ∆Ct case—∆Ct control). Third, the ∆∆Ct values are converted to the linear form using the term 2^-∆∆Ct. The Standard Error of the Mean (SEM) is calculated from the final 2^-∆∆Ct values. Statistical significance was calculated for differences between the 2^-∆Ct values of the “control” and those of the “case” samples.

The T/G HA-VSMC (vascular smooth muscle cell) line (CRL-1999) was purchased from ATCC (American Type Culture Collection, Virginia, USA). This line of VSMC is a normal human cell line established from the normal aorta of an 11-month old child and one of the best-characterized cellular models for the analysis of vascular smooth muscle cell biology. Cells were cultured in F-12K Medium (ATCC, No. 30–2004) supplemented with 0.05 mg/ml ascorbic acid, 0.01 mg/ml bovine insulin, 0.01 mg/ml human transferrin, 10 ng/ml sodium selenite, 0.03 mg/ml Endothelial Cell Growth Supplement, 10 mM HEPES, 10 mM TES, and 10% fetal bovine serum under a 5% CO₂ in air-ventilated incubator. To knock-down Csk, VSMCs were transiently transfected with control or Csk siRNA using Lipofectamine 2000 and treated for 24 hours with DMSO or PP2 after 24 hours transfection and then proteins were extracted for Western blotting.

Total proteins were extracted from mouse tissues using PRO-PREP protein extraction solution (Intron Biotechnology, Gyeonggi-Do, Korea) according to the manufacturer's instruction. Protein concentrations were measured by Bradford assay [23].

Total proteins (10 to 40 μg) were separated by 8–10% SDS-PAGE and transferred to a nitrocellulose membrane (Pall, Ann Arbor, MI, USA). The membrane was blocked in 5% skim milk for 30 min at room temperature and subsequently incubated with an antibody overnight at 4°C; anti-Actin (sc-1616, goat polyclonal, 1:1000), and anti-Csk (sc-286, rabbit polyclonal, 1:1000) from Santa Cruz Biotechnology (Santa Cruz, CA, USA); anti-phospho-Src^Y416 (#2101, rabbit polyclonal, 1:1000), anti-phospho-Src^Y527 (#2105, rabbit polyclonal, 1:1000), and anti-Src (#2123, rabbit polyclonal, 1:2000) from Cell Signaling Technology (Danvers, MA, USA). The blot was incubated with a horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) for 1 hour at room temperature. Protein signals were detected using Luminol Reagent (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) and exposed to x-ray films (Agfa-Health Care NV, Mortsel, Belgium).

The protocol for blood pressure measurement has been previously described [19–21]. In brief, blood pressure was recorded intra-arterially with a computerized data acquisition system (AD Instruments, Bella Vista, Australia). First, mice (7-9-week old, female, 20–25 g) were anesthetized by intraperitoneal (i.p.) injection of tribromoethanol (18 ml of the working solution per kg body weight), of which the working solution was diluted 40-fold in 0.9% NaCl from the stock solution (10 g 2,2,2, tribromoethanol dissolved in 10 ml tertiary amyl alcohol). The intra-arterial catheter (a polyethylene tube; 0.2 mm I.D., 0.5 mm O.D.; Natsume, Tokyo, Japan) filled with 0.9% NaCl containing 100U/ml heparin was connected to the system. Second, while on anesthetics, a small incision was made to draw out the right carotid artery of the mouse. The catheter was inserted into the right carotid artery through a small puncture and tightly ligated with 4–0 silk suture thread around the artery. Third, blood pressure in the vessel was transmitted along the catheter to the transducer’s diaphragm (MLT0699 Disposable BP Transducer, AD Instruments, Bella Vista, Australia). The diaphragm signal was amplified through a bridge amplifier and recorded on a Power Lab system (Lab Chart 7.2, AD Instruments, Bella Vista, Australia).

Blood pressure was monitored for 2 hours after injection of the anesthetic and presented as the average value at each minute between 40 and 80 min (a linear graph) and as the average value during 40 minutes from 40 to 80 min (a bar graph) because blood pressure during the presented period was considered enough stable as evidenced by multiple repeated experiments.

To test the effect of PP2 (Merck Millipore, 529573, Darmstadt, Germany) or PP3, a negative control inhibitor (Merck Millipore, 529574, Darmstadt, Germany), blood pressure was measured in Csk^+/- and Csk^+/+ mice 24 hours after i.p. injection of PP2 or PP3 at doses of 2 or 10 μg per kg body weight.

Statistical analysis was performed using SPSS (PASW Statistics 22.0), and all data between case and control groups were analyzed by Mann-Whitney U-test since the U-test was in general considered more powerful than the t-test [24]. All data were expressed as mean ± SEM. P < 0.05 was considered statistically significant. * P<0.05, ** P<0.01 versus respective controls.

Fig 1A shows the locations of nearby genes including CYP1A2, CSK, and ULK3 on the human 15q24 locus with the comparable gene locations on mouse chromosome 9 in the opposite direction. The lead SNP (rs1378942 designated as *) is located in the first intron of CSK. To test the effect of candidate gene variation on blood pressure, we used silencing of candidate genes by siRNA injection into mice in vivo. Both mouse homologs Csk and Ulk3 were selected as comparably strong candidate genes among nearby genes on the 15q24 locus based on eQTL analysis studies (S1 Table). Cyp1a2 was tested as a negative control because there has been reported neither significant association with blood pressure in GWAS or significant link to the lead SNP (rs1378942) in eQTL resources (S1 and S3 Tables). Regarding its functional role, CYP1A2 is associated with habitual caffeine consumption, and mediates drug metabolism and the synthesis of cholesterol, steroids, and lipids (S3 Table) [25, 26].

Three siRNAs per gene were synthesized and tested for their silencing efficacies in NIH3T3 and B16F10 cells, and the most effective siRNA was selected for in vivo injection (S4 Table). The selected siRNAs caused Csk, Ulk3, and Cyp1a2 levels to decline by 40.1%, 74.5%, and 44.9% in cells, respectively. The siRNAs were mixed with polyethylenimine and injected into mouse tail veins. Then, to estimate the systemic change by siRNA injection, we measured target genes mRNA levels 24 hours later by quantitative real-time PCR, comparing them between candidate gene siRNA- and control siRNA-injected mouse tissues.

After 1-time injection, Csk mRNA levels decreased significantly in heart by 37% (P = 0.038), in kidney by 20% (P = 0.038), and in lung by 29% (P = 0.017), respectively, and Ulk3 mRNA levels fell in lung by 32% (P = 0.001) (Fig 1B and 1C). Cyp1a2 mRNA levels fell in liver by 49% (P = 0.036) and in kidney 21% (P = 0.032) (Fig 1D).

To study the functional effect of siRNA injection, blood pressure was measured 24 hours after the last injection of Csk, Ulk3, or Cyp1a2 (negative control) siRNAs.

After 2-time injection, blood pressure did not change in Cyp1a2 siRNA-injected group (74.3 mmHg, N = 6) compared with control groups (75.2 mmHg, N = 6) (P = 1.000) (Fig 2G and 2H). However, Csk siRNA-injected mice (88.2 mmHg, N = 7) showed significantly increased blood pressure compared with control groups (74.9 mmHg, N = 7) (P<0.001) (Fig 2A and 2B). Interestingly Ulk3 siRNA-injected mice (74.3 mmHg, N = 7) did not show significant changes in blood pressure versus control mice (73.9 mmHg, N = 8) (P = 1.000) (Fig 2D and 2E).

To confirm reduction of the target gene mRNA, Csk, Ulk3, and Cyp1a2 mRNA levels were measured in mice tissues after 2-time injection of each siRNA. Csk mRNA levels decreased significantly in heart by 31% (P = 0.001) and in liver by 19% (P = 0.01), respectively, and Ulk3 mRNA levels also decreased in liver by 34% (P = 0.001), in spleen by 37% (P = 0.001), and in lung by 44% (P = 0.003), respectively (Fig 2C and 2F). Cyp1a2 transcripts also fell in liver by 32% (P = 0.045) after 2-time injection (Fig 2I). The results showed that only Csk gene knock-down induced a significant change in blood pressure although all three genes (Csk, Ulk3, and Cyp1a2) were knock-downed by siRNA injections in various tissues.

To test the lasting effect of siRNA silencing, Csk siRNAs were injected into mice once per day for 3 days (3-time injection). Blood pressure significantly differed between Csk siRNA-treated mice (87.4 mmHg, N = 7) and control mice (80.6 mmHg, N = 7) (P = 0.017) even after 3-time injection (Fig 3A and 3B).

To study Csk protein levels in Csk siRNA-injected mice, Western blotting for Csk was performed using mouse lung tissues. Csk protein levels declined after all 1-, 2- and 3-time injections in lung tissues (Fig 3C). After 2-time injection of Csk siRNA, Csk protein levels fell in lung by 53% (P = 0.016) (Fig 3C and 3D). These data imply that Csk protein levels may stay decreased systematically after repeated injections of siRNAs despite of rebounding mRNA levels.

To confirm the blood pressure change in Csk siRNA-injected mice, we examined Csk knockout mice. Because the Csk knockout homozygote (Csk^-/-) was embryonic-lethal, the Csk heterozygote (Csk^+/-) mice were compared with the wild-type (Csk^+/+) mice. The Csk mRNA level in Csk heterozygotes decreased in most tissues by half (Fig 4C). Consistent with the results in Csk siRNA-injected mice, blood pressure in Csk^+/- mice was significantly higher than in the wild-type (Csk^+/-, 86 mmHg, N = 12; Csk^+/+, 80 mmHg, N = 15; P = 0.021) (Fig 4A and 4B).

Since Csk is a negative regulator of Src protein [14] and Src plays a role in contraction of vascular smooth muscle cells [27], Western blotting analysis was performed for Csk and phosphorylated Src proteins in the aorta of Csk^+/- and Csk^+/+ mice. Csk protein levels were reduced by 52% (P = 0.002) in the Csk^+/- aorta (Fig 4D and 4E). While the total Src protein was not changed in Csk^+/- aorta, phosphorylation of Src^Y416 (active) significantly increased by 38% (P = 0.046) and that of Src^Y527 (inactive) decreased by 25% (P = 0.01), suggesting Src as a target of Csk in the aorta (Fig 4D and 4E).

To study whether Csk involves Src in modulating blood pressure, we examined the effect of PP2, a Src inhibitor in Csk^+/- mice. High blood pressure in Csk^+/- mice (93.9 mmHg, N = 8) was significantly reduced 24 hours after i.p. injection of 10 μg PP2 per kg body weight (83.5 mmHg, N = 8; P = 0.038) (5A and 5B). Lesser dose of PP2 seemed to decrease blood pressure without significance in Csk^+/- mice (2 μg/kg PP2, 87.2 mmHg, N = 5, P = 0.284) while PP2 did not affect the basal blood pressure in the wild-type (0μg/kg, 80.2 mmHg, N = 8; 2 μg/kg, 82.1 mmHg, N = 8; 10 μg/kg, 80.2 mmHg, N = 10) (Fig 5B). We studied the effect of PP3, a negative control inhibitor of PP2 and did not see any significant change of blood pressure by its treatment (Csk^+/+, 80 mmHg, N = 18; Csk^+/-, 89 mmHg, N = 8; Csk^+/- + 10 μg/kg PP3, 89 mmHg, N = 5), indicating that PP2 was specific in its effect on blood pressure (S1 Fig).

To confirm inhibition of Src by PP2, we performed Western blotting analysis for phosphorylated Src proteins in the aorta of Csk^+/- and Csk^+/+ mice treated with 10 μg/kg PP2. As shown in Fig 5C and 5D, phosphorylated Src^Y416 (active) was significantly reduced in the PP2-treated Csk^+/- aorta compared to vehicle-treated aorta and phosphorylated Src^Y527 (inactive) was also slightly decreased.

Further to study regulation of Src by Csk in functionally relevant cell types, we cultured human vascular smooth muscle cells (T/G HA-VSMC, ATCC), transfected cells with Csk siRNA, and then treated with 10, 25, 50 μM PP2 for 24 hours. The Western blotting analysis showed that phosphorylation of Src^Y416 was conspicuously increased by Csk siRNAs transfection in VSMCs consistent with the result from the Csk^+/- mice aorta (Fig 5E). Moreover, PP2 treatment dose-dependently decreased phosphorylation of Src^Y416 in Csk siRNA-transfected cells (Fig 5E). Phosphorylation of Src^Y527 did not significantly change by Csk siRNAs transfection, but seemed to decrease by PP2 treatment in VSMCs (Fig 5E).

These results suggest that Src is specifically involved in blood pressure regulation by Csk.

Csk was identified as a tyrosine kinase of Src, a non-receptor tyrosine kinase [14]. Csk inactivates Src by increasing phosphorylation of Src^Y527, acting as a negative regulator of Src (Fig 6) [31]. Phosphorylation of Src^Y416 is increased in the Csk^+/- mouse aorta and in Csk knock-down cells than in the wild-type (Fig 5C, 5D and 5E). Furthermore, inhibition of Src decreases high blood pressure in Csk^+/- mice (Fig 5A and 5B), strongly suggesting that hypertension occurs through Src.

Src is one member of eight Src family tyrosine kinases (c-Src, c-Yes, Fyn, c-Fgr, Lyn, Hck, Lck, and Blk) [31]. Among them, the activity of Src is important in Ang II-induced vasoconstriction of vascular smooth muscle cells (VSMCs). Ang II induces phosphorylation of Src^Y416 via angiotensin type 1 receptor (AT1R) in rat VSMCs [32]. Ang II-induced intracellular Ca²⁺ mobilization is blunted in Src^-/- VSMC and vascular contraction is inhibited by PP2, a Src inhibitor [27]. Interestingly low activation of Csk is associated with increased Ang II-mediated Src signaling in VSMCs of Spontaneously Hypertensive Rats (SHR) [16]. These reports and our findings suggest that the Csk-Src signal may be a key regulatory point in Ang II-mediated vasoconstriction and blood pressure regulation (Fig 6).

Downstream of the Csk-Src signal, extracellular signal-regulated kinases (ERK1/2) may play an important role in vasoconstriction. In SHRs, inhibition of ERK signaling by U0126 reduces hypertension, vascular thickening, and smooth muscle myosin light chain kinase expression [33]. Moreover, Ang II-mediated ERK1/2 activation requires Src in rat VSMCs [16, 34]. Therefore, we speculate that Csk negatively modulates Ang II-mediated Src/ERK signaling in vasoconstriction of VSMCs.

Ang II is a major effector hormone of the renin-angiotensin system and has various effects that are mediated by AT1R to raise blood pressure, including direct vasoconstriction in the vascular smooth muscle, indirect vasoconstriction by noradrenaline release from the sympathetic nerve system, sodium reabsorption in the kidney, sodium and fluid retention by aldosterone secretion from the adrenal gland, and vascular growth stimulation [35].

In addition to already mentioned role of Src in VSMCs, Src activation is required for Ang II-mediated functions in the kidney and in the brain. In the renal proximal tubule-like cells, PP2 and Src siRNA inhibits Ang II-induced epithelial-to-mesenchymal transition (EMT) associated with progressive kidney damage [36]. Also Src is activated in Ang II-induced fluid reabsorption via apical Na⁺/H⁺ exchanger (NHE) and basolateral Na⁺-HCO₃^- cotransporter (NBC) and Csk overexpression inhibits activation of both NHE and NBC [37, 38]. In the distal tubule of kidney, Src is required for inhibition of the renal outer medullary potassium (ROMK) channel during hypovolemia (low blood volume) by low potassium diet [39–42]. Therefore, hypertension by low potassium diet may be associated with the inhibitory effect of Src on the ROMK channel.

In the hypothalamus of hypertensive rats induced by two-kidney one-clip (2K1C), Src activation is higher than sham-operated rats and is more increased by Ang II [43]. Two specific Src inhibitors (PP2, SU6656) abolish increase in sympathetic nerve activity and blood pressure of 2K1C rats, suggesting that Src in the brain mediates sympathetic activation by renovascular hypertension.

Considering that Src plays regulatory roles in diverse tissues, hypertension in Csk^+/- mice may rather be the combined result of differential effects by chronic Src activation in multiple tissues (e.g., vasculature, kidney, or/and brain) than the result of a defect in one particular tissue. Therefore, to fully understand the mechanism of hypertension in Csk^+/- mice, changes in Src-associated signaling need to be further studied in each tissue.

In summary, we have identified CSK a causative gene in the blood pressure GWAS locus 15q24 by selecting candidate genes using the eQTL analysis studies, by verifying the functional effect of genes using in vivo siRNA delivery in mice, and by confirming our findings in knockout heterozygote mice. Furthermore, we showed that Csk involves Src by negatively regulating its activity to modulate blood pressure. Our result that Csk regulates blood pressure through Src suggests a novel pathway for the development of hypertension and provides a therapeutic target for the treatment of high blood pressure.