Phosphorylation of β1-integrin in juxtaglomerular cells helps control blood pressure during the progression of diabetic nephropathy

The induction of a high blood pressure due to diabetic nephropathy depends on the increase in renin secretion from juxtaglomerular cells, but many aspects of how juxtaglomerular cells sense blood pressure changes in the afferent arteriole and consequently react remain unclear. In this study, we detected the juxtaglomerular cell-specic phosphorylation of the threonine-788/789 site of β1-integrin, and its expression was negatively correlated with renin production. This relationship was also observed in a culture system of a juxtaglomerular cell line, suggesting that β1-integrin is deeply involved in the regulation of renin production. The knockdown of β1-integrin in the culture system increased renin production, but the degree of the increase was comparable to the increase in renin production by knockdown of connexin-40, which is considered to be an important molecule that plays a role in the pressure sensing mechanism of juxtaglomerular cells. This suggests that the mechanism underlying the regulation of renin production by β1-integrin in juxtaglomerular cells may contribute to the pressure-sensing function of juxtaglomerular cells themselves. Threonine-788/789 phosphorylation of β1-integrin may be involved in the regulation of this pressoreceptor function. (176

Hypertension causes oxidative stress and in ammation in the kidney, leading to the development of DN [4]. Activation of RAAS [10], upregulation of endothelin-1 (ET-1) [11,12], upregulation of reactive oxygen species [13,14], and downregulation of nitric oxide (NO) [15,16] induced by a worsening renal function further elevate the blood pressure, leading to a vicious cycle of hypertension and glomerular damage. The rate-limiting enzyme involved in blood pressure regulation by the renin-angiotensin system is renin, and it is key to the renal functional deterioration and the vicious cycle of high blood pressure.
Studies concerning juxtaglomerular cells have revealed the following mechanisms underlying renin secretion: (a) stimulation of renin secretion through β-adrenergic receptors by the excitement of the kidney sympathetic nerves sensing a decrease in the blood ow quantity and blood pressure, (b) control of renin secretion through prostaglandin-E2/adenosine/ATP/NO released from the macula densa, which senses raw urinary NaCl levels; (c) feedback regulation of renin secretion by humoral factors, such as angiotensin II; (d) regulation by pressoreceptors (baroreceptor) of the juxtaglomerular cells, which sense the blood pressure [17,18,19]. Many intensive studies have clari ed the heteronomy-like mechanisms that receive regulatory factors secreted by organs other than juxtaglomerular cells (a-c); however, the independent mechanisms underlying how juxtaglomerular cells detect changes in blood pressure of the afferent arterioles by pressoreceptors and control renin production and secretion remains unclear.
Regulation of renin secretion through juxtaglomerular cells sensing the blood pressure or blood ow has been proven by perfusion experiments of isolated glomeruli [20,21]. The gap junction protein connexin-40 (Cx40) was recently shown to be essential for the pressoreceptor function of juxtaglomerular cells, as knockout of Cx40 speci cally in a mouse juxtaglomerular cell resulted in a high blood pressure due to renin hypersecretion congenitally, and no suppression of renin secretion was observed when perfusion pressure was applied to the juxtaglomerular apparatus isolated from the knockout mice [22]. β1-integrin is a molecule involved in cell-cell and cell-substrate attachment by dimerization with various subtypes of α-integrin. For example, α3β1-integrin has been shown to be expressed primarily in podocytes and is suspected of playing a critical role in the attachment of podocytes to the glomerular basement membrane. In addition, podocyte depletion is suspected to occur with the loss of the function of α3β1integrin [23,24]. As a regulatory mechanism of the β1-integrin function, phosphorylation/dephosphorylation of the intracellular domain alters the structure of integrin dimers. In particular, phosphorylation of the threonine-788/789 sites is known to regulate cell adhesion and cell motility [25,26,27].
We herein report the phosphorylation of β1-integrin speci c to juxtaglomerular cells and the relationship between the phosphorylated β1-integrin and renin production. Our ndings suggest the regulation of blood pressure by the phosphorylation of β1-integrin in juxtaglomerular cells.

Results
Phosphorylation of threonine-788/789 sites of β1-integrin speci c to juxtaglomerular cells To investigate β1-integrin phosphorylation during DN progression, speci c antibodies against phosphorylated threonine-788 and threonine-789 of β1-integrin were applied to kidney tissue sections from various stages of streptozotocin-treated rats (STZ rat) to study the distribution and phosphorylation of β1-integrin. Both antibodies showed intensive reactions at the juxtaglomerular apparatus in the kidney of STZ rats and controls (Fig. 1a). Since stained cells seemed to exist around afferent arterioles, double staining with anti-renin antibody was performed, and these cells were con rmed to be renin-secreting juxtaglomerular cells (Fig. 1b).
A negative correlation between the number of cells containing threonine-788/789 phosphorylated β1integrin and the renin expression The changes in the phosphorylation of threonine-788/789 of β1-integrin speci c to juxtaglomerular cells during DN progression were investigated in the kidneys of STZ rats, along with their relationship with the renin production. The production of renin gradually increased after the onset of DN, peaking in the second month but then decreasing and returning to the original level in the sixth month. In contrast, the number of juxtaglomerular cells in which threonine-788/789 phosphorylation of β1-integrin was detected gradually decreased after the onset of DN, peaking in the second month and then increasing further before ultimately returning to the original level (Fig. 2). These results showed that there was a negative correlation between the renin production and threonine phosphorylation of β1-integrin during DN progression in STZ rats (correlation coe cient = 0.61).
Activation/inhibition of protein kinase Cε reproduced the negative correlation between threonine-788/789-phosphorylated β1-integrin and the renin expression in vitro As a negative correlation between the renin production and threonine-788/789 phosphorylation of β1integrin in juxtaglomerular cells was observed in kidneys of STZ rats, the relationship between them was examined in cultured cells in vitro. As4.1 cells, a line of mouse juxtaglomerular cells, were used to examine the changes in renin production after the addition of PMA/BIM-1, an activator/inhibitor of PKCε, which is a subtype of PKC that phosphorylates threonine-788/789 of β1-integrin [27]. As TGF-β1 also reportedly promotes the threonine-788/789 phosphorylation of β1-integrin speci cally [26], we also investigated the effect of TGF-β1 addition to the culture.
The results showed that renin production in As4.1 cells after 24 h of drug addition was inhibited by PKC activation, and conversely, renin production was enhanced when PKC was inhibited, con rming a negative relationship between these two, similar to that in vivo ( Figure. 3a). TGF-β1 also inhibited renin production and supported this relationship. When we monitored the changes in renin production up to 24 h after PMA/BIM-1 treatment, we found that the renin production showed severe changes, including a rapid rise and fall in the early period up to three hours after treatment, but thereafter there was a gradual suppression and steady increase respectively until the end of the experiment (Fig. 3b). Western blotting for phosphorylated threonine-788/789 of β1-integrin showed results compatible with the above ndings, wherein the acceleration of phosphorylation by PMA occurred 3 h after drug addition and was sustained for 24 h. BIM-1 suppressed the phosphorylation of the sites, and TGF-β1 promoted the phosphorylation but not as much as PMA (Fig. 3c).
Knockdown of the β1-integrin expression results in out-of-control renin expression, as does knockdown of the connexin-40 expression Since PMA and BIM are effective across a broad spectrum of PKC subtypes, we cannot ignore the possibility that the changes in renin production seen in Fig. 3 are a combination of effects from other phosphorylation pathways in addition to the phosphorylation of β1-integrin by PKCε. In order to clarify the effect of β1-integrin on renin production, we performed knockdown of β1-integrin in As4.1 cells and examined its effect on renin production. The results showed that knockdown of β1-integrin enhanced renin production in As4.1 cells to about twice that in controls. Knockdown of Cx40, a putative pressoreceptor, also increased the renin production to the same extent, suggesting that these renin production levels represent an uncontrollable state of renin production of juxtaglomerular cells (Fig. 4).

Discussion
In this study, we demonstrated for the rst time the phosphorylation of the threonine-788/789 of β1integrin speci c to juxtaglomerular cells and also found a negative correlation between the phosphorylation of this site of β1-integrin and renin production, suggesting that the phosphorylation/dephosphorylation of this site of β1-integrin may be part of the blood pressure regulation mechanism in the process of DN.
Various reports have described the effects of phosphorylation of threonine-788/789 of β1-integrin on cell adhesion and motility [25,26,27], and a recent report noted that this site of β1-integrin functions as a phospho-switch to activate/deactivate integrin dimers and regulate integrin functions, such as the strength of cell adhesion [30]. In juxtaglomerular cells, this mechanism may contribute to the production and secretion of renin, which is a speci c function of this type of cell.
Everett et al. reported that renin production is reduced in advanced diabetes [31]. They counted the number of renin-positive cells per juxtaglomerular apparatus in a strain of rats that spontaneously develop diabetes and reported that the index peaked at two months after the onset of diabetes, which is consistent with our ndings in this study. In our observation of STZ rats, the blood pressure continued to increase after the onset of diabetes, but the renin production began to decrease at two months after the onset of diabetes, and the blood pressure gradually decreased accordingly. During this process, the level of threonine-788/789 phosphorylation of β1-integrin was negatively correlated with the change in renin production, but the causal relationship between the phosphorylation of β1-integrin and renin production was unknown. To con rm this, studies using a juxtaglomerular cell line showed that the activation of integrin dimers by phospho-switching is required to suppress renin production. In STZ rats, the phosphoswitch is inactivated after the onset of DN, which may prevent the suppression of renin production in response to elevated blood pressure. Thereafter, some unknown mechanism reactivates the phosphoswitch, thereby restoring the suppression of renin production. There may be a temporal change in the predominance of PKCε and phosphatase activities during DN.
In addition, the mechanism by which activated integrin dimers suppress renin production may involve regulating the sensing of pressure by juxtaglomerular cells. This is because renin production in cultured cells is not regulated by a heteronomous mechanism that receives regulatory factors secreted from other organs and cells but is controlled by the juxtaglomerular cells themselves and thus may be involved in the pressoreceptor function, the only independent regulatory mechanism of renin production in juxtaglomerular cells.
Cx40 is essential for the pressoreceptor function in juxtaglomerular cells, and knockout mice with this molecule are known to be unable to suppress renin secretion [22]. Cx is a component of the intercellular adhesion apparatus and is expected to be well related to integrins, which are also cell adhesion molecules on the cell membrane. In fact, direct/indirect interaction between β1-integrin and Cx has been reported in several studies [32,33,34]. Knockdown of Cx40 in the present study resulted in an unregulated increase in renin production, as in Cx40 knockout mice, while the knockdown of β1-integrin also increased it to the same extent, strongly suggesting that these molecules may be involved in the same renin production mechanism. This suggests that β1-integrin may be involved in the pressure sensing of Cx40 in juxtaglomerular cells. For example, there may be a mechanism that controls the sensitivity of Cx40 pressure sensing on the cell membrane by altering the cell-substrate adhesion strength. These results strongly suggest that the ne regulation of pressure sensing by the synergistic interaction of integrins and Cx40 controls the renin production of juxtaglomerular cells.
A further analysis of the mechanism of blood pressure sensing by Cx40 and its regulation by integrins including mechanism of phosphorylation/dephosphorylation of β1-integrin during DN may lead to the elucidation of the pressoreceptor-mediated regulation of blood pressure and the development of new DN therapies targeting this mechanism.

Materials And Methods
Animal experimental protocols were approved by the Institutional Animal Care and Use Committee (Permission number: 194019) and carried out according to the Tokai University Animal Experimentation Regulations. All animal studies were also carried out in compliance with the ARRIVE guidelines [35]. All methods were carried out in accordance with relevant guidelines and regulations.

Animals
Five-week-old male SD-rats (CLEA Japan Inc., Tokyo, Japan) underwent intravenous injection of streptozotocin (STZ; 60 mg/kg). After one week, individuals with more than 400 mg/dl blood glucose were used subsequent experiments as an animal model of DN.
Quantitative real-time polymerase chain reaction (qRT-PCR) Total RNA was recovered from kidney tissues or cultured cells with an RNaqueous RNA puri cation kit (ThermoFisher) and converted to cDNA using a SuperScript III First-Strand Synthesis System (ThermoFisher). qRT-PCR was performed with the cDNA as previously described [28]. The reaction mixture was prepared as described in the kit's instruction manual (TaqMan Gene Expression Assays; ThermoFisher), containing primers and probes for rat or mouse renin (Assay ID: Rn00561847_m1, Mm02342887_mH) and endogenous controls (18S ribosomal RNA or β-actin). PCR was performed on an ABI StepOne Plus (ThermoFisher). Data were analyzed by the comparative Ct method, and the amounts of renin mRNA were expressed relative to that of endogenous controls.

Western blotting
Cells were lysed by xTractor buffer (Takara Bio, Shiga, Japan) with ProteoGuard protease inhibitor (Takara Bio) and Cryonase nuclease (Takara Bio). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed with NuPAGE 4-12% Bis-Tris gel (ThermoFisher) and proteins were transferred to a PVDF membrane with an iBlot dry blotting system (ThermoFisher). After blocking with 5% cow-albumin in Tris-buffered saline, the PVDF membrane was treated with the rabbit polyclonal antibodies for Thr788/789-phospho-integrin β1 mentioned above, washed with TBS containing 0.05% Tween 20 and treated with horseradish peroxidase-conjugated anti-rabbit IgG antibody. After washing, the reactions were visualized with DAB.

Statistical analyses
Results are represented as the mean (± standard deviation) of at least four samples. Comparisons between two groups were done using Student's t-test. Spearman's correlation coe cient analysis was used to examine the relationship between parameters. P < 0.05 was considered statistically signi cant. Figure 1 Detection of threonine-788/789 phosphorylation of β1-integrin in juxtaglomerular cells. a. The reaction of antibodies for Thr788/789-phospho-integrin β1 on sections of normal rat kidney was visualized by DAB and counterstained by hematoxylin. Bar: 20 μm. b. Double staining with uorescent antibodies for renin (green) and Thr788/789-phospho-integrin β1 (red). Auto uorescence of red blood cells was detected as weak green signals. The merged gure with DAPI nuclear staining (blue) showed cells co-expressing renin and phosphorylated integrin β1. Bars: 20 μm.

Figure 2
Negative relationship between the renin expression and number of cells positive for Thr788/789phophorylation of integrin β1. The renal renin expression at various points in STZ rats was determined by qRT-PCR and represented as bars (left axis). The number of cells positive for Thr788/789-phosphointegrin β1 and number of glomeruli were counted in sections of kidneys of STZ rats, and their ratios were represented as a line (right axis).