Down-Regulation of Integrin β1 and Focal Adhesion Kinase in Renal Glomeruli under Various Hemodynamic Conditions

Given that integrin β1 is an important component of the connection to maintain glomerular structural integrity, by binding with multiple extracellular matrix proteins and mediating intracellular signaling. Focal adhesion kinase (FAK) is the most essential intracellular integrator in the integrin β1-FAK signalling pathway. Here, we investigated the changes of the two molecules and visualized the possbile interaction between them under various hemodynamic conditions in podocytes. Mice kidney tissues were prepared using in vivo cryotechnique (IVCT) and then were stained and observed using light microscopy, confocal laser scanning microscopy and immunoelectron microscopy. The expression of these molecules were examined by western blot. Under the normal condition, integrin β1 stained continually and evenly at the membrane, and FAK was located in the cytoplasm and nuclei of the podocytes. There were significant colocalized plaques of two molecules. But under acute hypertensive and cardiac arrest conditions, integrin β1 decreased and stained intermittently. Similarly, FAK decreased and appeared uneven. Additionally, FAK translocated to the nuclei of the podocytes. As a result, the colocalization of integrin β1 and FAK reduced obviously under these conditions. Western blot assay showed a consistent result with the immunostaining. Collectively, the abnormal redistribution and decreased expressions of integrin β1 and FAK are important molecular events in regulating the functions of podocytes under abnormal hemodynamic conditions. IVCT could offer considerable advantages for morphological analysis when researching renal diseases.


Introduction
Glomerular podocytes are terminally differentiated cells that line the outer aspect of the glomerular basement membrane (GBM). The GBM forms the final barrier against protein loss, which explains why its dysfunction causes protein leakage into the urine, resulting in proteinuria [1]. Podocytes are injured in many types of human and experimental glomerular diseases, including hypertensive renal disease [2][3][4].
As an important component of the connection between podocytes and the GBM, the reduction of integrin a3b1 might represent one of the mechanisms of podocyturia in glomerular disease. Integrin a3b1 is a membrane glycoprotein consisting of two subunits, a larger a chain and a smaller b chain [5], and the alteration of subunit b1 can represents the change of integrina3b1. Under different stimuli, integrin a3b1 binds to multiple extracellular matrix (ECM) proteins, including laminin, collagen and fibronectin, which are the components of GBM and which transduce different intracellular signals [6]. At binding sites, they form focal contacts, which bring together cytoskeletal and signaling proteins during the processes of cell adhesion, spreading and migration. In many intracellular ''integrator'', FAK is the most essential one, which is a nonreceptor protein tyrosine kinase.
FAK appears to play a major role in mediating signals. Phosphorylation at Tyr397 might be the first step or an indispensable path in further signaling transduction. It has been demonstrated that FAK activation is significantly increased after podocyte injury. More recent studies have shown that inhibiting FAK activation reduces proteinuria and podocyte effacement [7]. It has been demonstrated in many experimental animals that FAK and FAK phosphorylation are increased in many pathological situations and that they are translocated in the nucleus [8,9], but the results was obtained in some chronic diseases or for a long period after administering treatment in vitro. Nevertheless, how these molecules change in the acute diseases or during the prophase of abnormal hemodynamic conditions remains to be examined.
It is well known that hemodynamic factors, such as blood flow and pressure, exert an important influence on the native structure and function [10][11][12], Acute hypertension, high pressure of glomerular blood capillaries impaired the size-selective barrier function of the slit diaphragm and glomerular basement mambrane, so that hyperfiltrated serum proteins are processed [13,14]. Cardiac arrest condition, ischemia and hypoxia, was reported to induce some changes in glomerular structures and also damages of renal microvascular cell-cell junctions, which then increased vascular permeability and local interstitial edema [15,16]. These influences appear in a split second, so we must maintain all of the components in situ to study the actual situation, which is impossible to achieve if we prepare the organs using conventional methods [17]. In contrast, IVCT is a technology that can arrest transiently dynamic structures in living animal organs. Moreover, IVCT has been used successfully to clarify the serum protein passing through the glomerular capillary loops under various hemodynamic conditions [18]. All of these experiments have indicated that IVCT is a sensitive and powerful fixation technique for visualizing the structural and functional changes that occur in living animal kidneys under various hemodynamic conditions.
In the present study, the alterations of integrin b1 and FAK in mouse kidneys under various hemodynamic conditions were visualized using IVCT in combination with freeze-substitution, and they were further analysed quantitatively by western blot. Our findings provide new insight into the mechanism of proteinuria during hemodynamic disorders.

Animals
Adult C57BL/6J mice, weighing 20-30 g, were used. The experimental animal procedures were approved by the Animal Experimental Committee of China Medical University.

IVCT for Mouse Kidneys, Freeze-Substitution
Fixation and Paraffin-Embedding. The mice were classified into three groups, each containing five mice: a normotensive group; an acute hypertensive group; and a cardiac arrest group. The mice were anesthetized via intraperitoneal injection of sodium pentobarbital (50 mg/kg body weight). In the normotensive group, we resected the left kidney under normal blood circulation. In the acute hypertensive group, we prepared an animal model by ligating the abdominal aorta just below the branching renal arteries for 10 min [19], before removing the left kidney. The cardiac arrest animal model was achieved by injecting excessive amounts of the anesthetic and fixing the left kidney immediately after the heart stopped beating.
The renal tissues were then assessed with IVCT, which has been described in previous reports [10,14,19]. Briefly, a cryoknife precooled in liquid nitrogen (2196uC) was positioned over the left kidney of the mouse. The kidney was immediately cut with the cryoknife, and liquid isopentane-propane cryogen (2193uC) was simultaneously poured over it. The frozen kidneys were carefully trimmed with a dental electrical drill in liquid nitrogen. Some of the frozen specimens were transferred to freeze-substitution, as described below, while others were preserved in liquid nitrogen for biochemical examination.
The pieces were freeze-substituted in absolute acetone containing 2% paraformaldehyde (PFA) cooled in dry ice-acetone at 280uC for 48 h and then were gradually rewarmed to room temperature. They were washed in pure acetone twice, transferred into xylene, and then embedded in paraffin wax.
3.2. Immerse-Fixation of Resected Kidney Tissues. As a control group, five kidney specimens were routinely treated with conventional fixation procedures. Some pieces of the tissues were immersed in 4%PFA for 24 h at room temperature, followed by gradual dehydration in a series of alcohols and then transfer into xylene. Finally, the samples were embedded in paraffin wax. The remainder was maintained in 280uC for the biochemical examination.

Immunostaining on Deparaffinized Sections
The paraffin-embedded tissues were cut at a 5 mm thickness and were deparaffinized with xylene and a graded series of alcohols. Some of the sections were stained in common hematoxylin and eosin (HE) to observe them morphologically. Other sections were incubated with 3% hydrogen peroxidase (H 2 O 2 ) to block the nonspecific reactivity of endogenous peroxidase, followed by a wash with phosphate-buffered saline (PBS). The samples were repaired with sodium citrate buffer liquid at high pressure and were blocked with PBS, containing 5% normal bovine serum (Boster, Wuhan, China) for 1 h at 37uC. Some sections were subsequently incubated with the primary antibodies at 4uC overnight, Other sections were incubated with PBS without the primary antibody. After washing them three times with PBS, the sections were incubated with corresponding secondary antibodies at 37uC for 1 h. Thereafter, they were incubated with horseradish peroxidase (HRP)-conjugated avidin-biotin complex (ABC) for 20 min and were visualized with metal-enhanced 393-diaminobenzidine (DAB) (Boster, Wuhan, China) for 5 min (ABC-DAB method). To visualize a clearer structure, we stained them with hematoxylin for 1 min and then dehydrated them in a graded series of alcohols and xylene. Finally, the specimens were sealed with peucine and were photographed under a light microscope or a confocal laser scanning microscope (FV10-ASW2.1 Viewer). Simultaneously, the integral optical density (IOD) of the target protein was measured with Meta Morph (UIC, US), and the result was determined as the sum of the glomeruli. IOD was defined as the sum of the optical densities of all the positive pixels in the image, which represents the quality of the targeted protein.

Immunoelectron Microscopy
To detect the ultrastructural changes of glomeruli and the immunolocalization of podocyte proteins under various hemodynamic conditions, we stained some sections with 1% osmium tetroxide acid for 20 min after DAB dyeing, dehydrated them with a graded series of alcohols and acetone, and finally inversionembedded with Epon 812, polymerized at 60uC for 48 h. The sections were cut at a 70 nm thickness and stained with uranyl acetate, and then ultrastructural images were obtained with a transmission electron microscope.

Western Blot Analysis
The glomeruli were isolated as previously described [19]. Briefly, cortical kidney tissue was finely minced and sieved through sequential sieves of 250, 150, ana 105 mm pore size. The purity of the glomeruli obtained was 92% to 96% with minor tubular contamination. We performed western blot with the enhanced chemiluminescence system. We prepared two equal amounts of glomeruli. One was homogenized in buffer (10 mmol/L Tris-HCI, 50 mmol/L NaCl, 5 mmol/L EDTA, and 1% TritonX-100, containing phosphatase and proteinase inhibitors) to obtain the total proteins (TPs). It was maintained at 4uC overnight. The other was homogenized in buffer (10 mmol/L HEPES, 10 mmol/L KCI, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, and 1 mmol/L DTT, containing phosphatase and proteinase inhibitors) to obtain cytoplasmic proteins (CPs) and nuclear proteins (NPs). Then, all of the homogenates were centrifuged, and the supernatants, which contained cytoplasmic proteins, were collected; the remainder was lysed again at 4uC overnight with the same buffer that was used in total protein extraction.
Proteins were electrophoresed on 10% SDS-polyacrylamide gels and were electrotransferred onto polyvinylidene difluoride membranes. After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were hybridized with specific primary antibody overnight at 4uC. The filters were then washed with PBS-0.1% Tween 20 and were incubated with corresponding HRP-conjugated secondary antibody for 2 h at room temperature. We detected immunoreactivity with the enhanced chemiluminescence system (Amersham Biosciences, Buckinghamshire, UK). The chemiluminescent signal was captured with a Fujifilm LAS-4000 luminescent image analyzer (Fujifilm, Tokyo, Japan). Equal protein loading was confirmed by GADPH and Lamin B western staining of the gel. The data shown are representative of at least three independent experiments with similar results.

Statistical Analysis
The values are expressed as the means 6SDs. For multiple comparisons with a single control, one-way analysis of variance (ANOVA), followed by Dunnett's test, was employed. The analyses were conducted using SPSS statistical software, version 17.0. P,0.05 was considered to be a statistically significant difference.

IVCT Exhibits a Clearer Morphological Alteration of the Glomeruli under Different Hemodynamic Conditions
To examine the native morphology in the mouse kidneys from different groups, we stained the deparaffinized sections with HE. For the sections prepared by IVCT, the capillary loops were smooth and plump under normotensive condition (Fig. 1A). However, they became wizened under acute hypertensive condition, while Bowman's space and the proximal tubules or distal tubules (Fig. 1B) were more widely opened than under normotensive or cardiac arrest conditions (Fig. 1C), as reported before [18,20]. In all of the above images, there were certain erythrocytes in the blood vessels. In contrast, the capillary loops in sections fixed by immersion were shrunken and there were hardly any erythrocytes in them (Fig. 1D). It was further confirmed quantitatively that the glomeruli from immersion fixation tissues were significantly contracted compared to the IVCT group [17].
Through IVCT, we obtained clearer images of the glomeruli, which visibly showed the instant morphological alterations that occurred under the different hemodynamic conditions.

Instantaneous Changes in the Distribution and Alteration of Integrin b1, FAK and pTyr-397 FAK under Various Hemodynamic Conditions Are Examined by Immunohistochemistry and Immunofluorescence Analysis
Immunolabeling showed integrin b1 staining along the glomerular capillary loops (GCLs) continually and evenly under normotensive condition ( Fig. 2A), while under acute hypertensive conditions, it decreased and stained intermittently. (Fig. 2B). In the cardiac arrest group, we could only observe weak immunolocalization (Fig. 2C). At the same time, FAK and pTyr-397FAK were distributed in the cytoplasm and nuclei of the podocytes under normotensive condition (Fig. 2J, 2O), but they decreased and became uneven under abnormal hemodynamic conditions ( Fig. 2K-2M, 2P-2R). The alterations of integrin b1 and FAK were further confirmed by double immunofluorescent staining (Fig. 3A-3I). Under abnormal hemodynamic conditions, the changes of integrin b1 and FAK were similar with that were observed by immunohistochemical method. In addition, we found the translocation of FAK to nuclei were very obvious under abnormal conditions compared with normal condition. Given integrin b1 and FAK were also immunolocalized in erythrocytes. To provide accurate localizations of two molecules we used WT1, a specific protein of podocytes, as a marker. In Fig. 3J-3R, we performed double immunofluorescence of integrin b1 and WT1. Under abnormal hymodynamic conditions, the colocalization of integrin b1 and WT1 decreased. Unexpectedly, in the immersionfixation group (Fig. 2D), the integrin b1 staining was the strongest among the four groups. In contrast, the immunolocalization of FAK and pTyr-397FAK was the weakest compared with the IVCT groups (Fig. 2M, 2R). Figs. 2E, 2N, and 2S show the IODs of each protein under various hemodynamic conditions. These results are expressed as the means 6SDs.

Immunoelectron microscopy shows ultrastructural changes of integrin b1 and FAK expression in the hypertensive condition
To visualize the topographical migration of integrin b1 and FAK under hypertensive condition, we detected them using immunoelectron microscopy (Fig. 4). In the normotensive group (Fig. 4D), we observed that the foot processes tightly approached each other and became flatter, compared to the immersion fixation group (Fig. 4F). In contrast, the foot processes appeared different degrees of fusion in the hypertensive group (Fig. 4E). Integrin b1 was distributed on the basolateral membrane of podocytes under normotensive conditions (Fig. 4D), whereas FAK was located in the cytoplasm and nuclei of the podocytes (Fig. 4A). In contrast, under hypertensive condition, integrin b1 assembled along the basal membrane only, while FAK was strikingly gathered in the nuclei, indicating the rapid translocation of integrin b1 and FAK during the prophase of the aorta ligation process (Fig. 4B, 4E).

Acute Hypertensive and Cardiac Arrest Conditions Decrease the Expression of Integrin b1 and FAK and Translocate FAK under Abnormal Hemodynamic Conditions
To further evaluate the expressions of integrin b1 and FAK quantitatively, we examined the two podocyte protein levels in glomeruli. From Fig. 5A, we can see two bands at 90 kDa and 125 kDa. Evidently, the band at 90 kDa represents integrin b1,and the other band corresponds to FAK. The total protein levels of integrin b1 and FAK were consistently decreased under abnormal hemodynamic conditions compared to normotensive condition. To further confirm the nuclear translocation of FAK under abnormal conditions, we extracted the nuclear proteins from glomeruli and performed western blot. We found that FAK in cytoplasmic protein was gradually reduced, but it apparently increased in nuclear protein. These results suggested that FAK was translocated to the nuclei under abnormal hemodynamic conditions.
All of the results were consistent with the immunostaining results and could be attributed to the alteration of renal hemodynamic conditions.

Discussion
It is well known that podocytes are the targets of many forms of injury, and hypertensive injury might be one of the major causes, however, the precise mechanism has not been fully elucidated. In the present study, we visualized the alteration and rearrangement of integrin b1 and FAK under different hemodynamic conditions and as evidenced by IVCT with immunohistochemistry, immunofluorescence and western blot analysis. The results exhibited the reduction and transposition of integrin b1 and FAK under abnormal hemodynamic conditions.
It has been generally accepted that kidney morphology and function are dependent on the maintenance of normal blood pressure [10]. Additionally, hemodynamic factors change instantaneously [20]. To study these changes accurately, we must identify a method that can capture all of the biological components in situ. IVCT has been confirmed as having these capabilities [10,20]. This technique overcomes the technical problems of conventional fixation processes by cryofixing target organs in situ under living status, which could clarify the native morphological features of kidney tissues. In addition, IVCT can capture more molecular translocation events instantly in situ, which is much closer to the living state. Meanwhile, we can detect the erythrocytes in capillary vessels. We stained the nuclear with DAPI, ensuring the expression changes of integrin b1 and FAK in podocytes, excluding the expression of them in erythrocytes ( Fig  S1).
Under normotensive condition, integrin b1 was continuous and linear-like along the glomerular capillary loop, which is consistent with the previous report [21]. Reduced expression of integrin a3b1 in podocytes has been demonstrated in humans with FSGS [22], in diabetic nephropathy [5], and in PAN model rats [23]. However, how it changes under abnormal hemodynamic conditions in vivo has not been studied to date. Our study revealed the downregulation of integrin b1 in the prophase of acute hypertension, cardiac arrest. Furthermore, we found integrin b1 attached to the basal membrane of the podocytes, and the foot processes appeared with different degrees of fusion under hypertensive condition, which might strengthen the adhesion force between podocytes and GBM, while under normotensive condition, they were distributed in the basolateral membrane. This ultrastructural appearance was similar to a previous study [24]. Consistent with our observation, Cecile Dessapt exposed podocytes to the FX3000 strain unit (Flex-cell Int, USA), which can mimic glomerular hypertension, demonstrating that integrin b1 was downregulated under hypertensive condition [25]. Many studies have reported that integrin b1 downregulation is causally related to the loss of podocytes, which was supported by our recent observations (Wang, Li et al. unpublished data), in which we found podocytes in the urinary sediment of chronic hypertension patients, accompanied by a parallel reduction of podocytes in renal biopsy tissue. Additionally, another study suggested that a-actinin-4, which can regulate integrin b1 activation, is indispensable for maintaining strong podocyte adhesion to GBM [26]. FAK, as a signal molecule, has been demonstrated to be a key factor in mediating cell adhesion, and FAK activation was highly correlated with cell adhesion force [27]. In the present study, FAK and pTyr-397 FAK were visualized in the cytoplasm and nuclei of the podocytes in normotensive group, whereas the total expression of the two decreased under hypertensive condition. In contrast, the previous study suggested that pTyr397-FAK increased under pathological conditions, such as low-dosage LPS injections and rabbit anti-GBM-induced podocyte injury [7]. One possible explanation for these paradoxical findings is that abnormal hemodynamic injury is underlain by a different mechanism than other injuries. Another possible explanation for the discrepancy is the observation time. We observed the change immediately after ligating the aorta, which could be realized by IVCT. In comparison, they were observed several hours after treatment. Herein, we can speculate that the FAK and pTyr397-FAK experience a transient reduction during the prophase of podocyte injury. However, the precise underlying mechanisms of adhesion are not yet clear. A previous study suggested that FAK upregulated integrin activation to enhance integrin binding [28]. Conversely, another study reported that FAK was the downstream reactor with integrin, augmenting osteoblast adhesion [29]. However, which is the upstream of the signalling pathway under abnormal hemodynamic conditions has been a matter of debate until now.
Additionally, ischemia and hypoxia are important factors for podocytes. Several studies have documented that ischemia and hypoxia are widely associated with podocytes, except for in tubular and endothelial damage. Mark C. Wagner found that podocytes became swollen, accompanied by narrowed filtration slits, after stopping blood flow for 45 minutes [30]. Similarly, another report showed that podocytes were flattened and that processes widened after ischemia [31]. Furthermore, podocytes were demonstrated to be susceptible to apoptosis after exposure to low oxygen [32]. However, the molecular events that induce podocyte injury during ischemia are unknown. In this study, it was shown that the total proteins levels of integrin b1 and FAK decreased under cardiac arrest conditions compared to normotensive condition. In comparison, the level of FAK in the nuclei increased. Thus, we speculate that the accumulation of FAK in the nuclei is likely to participate in the regulation of ishchemic injury. In the immersion fixation group, the expression of integrin b1 and FAK was theoretically the least in all of the groups, because that group underwent the longest duration of ischemia. Unexpectedly, the expression of integrin b1 appeared the highest by immunohistochemistry. A possible explanation is that many superficial proteins of integrin b1 were redistributed or even washed off during the immersion fixation steps. Thus, what we observed was unture. This conclusion could be further confirmed by western blot analysis.
In summary, we visualized the alteration of integrin b1 and FAK during the prophase of the abnormal hymodynamic conditions using IVCT. Integrin b1 and FAK are two essential factors in the regulation of podocyte changes under abnormal hemodynamic conditions. Although the detailed molecular mechanism is not yet clear, pTyr397-FAK is absolutely necessary for the  regulation of podocyte changes. Above all, IVCT can maintain all of the substances in situ, providing accurate and clear images. Our next step is to investigate the associated molecules in the integrin-FAK signaling pathway, which should provide new insights into the regulation mechanisms during hemodynamic disorders.