Intracellular albumin overload elicits endoplasmic reticulum stress and PKC-delta/p38 MAPK pathway activation to induce podocyte apoptosis

Podocyte injury is closely related to proteinuria and the progression of chronic kidney disease (CKD). Currently, there is no conclusive understanding about the mechanisms involved in albumin overload and podocyte apoptosis response. In this study, we sought to explore the ways by which intracellular albumin can mediate podocyte apoptosis. Here, immortalized mouse podocytes were treated with bovine serum albumin (BSA) at different times and concentrations, in the presence or absence of SB203580 (0.1 µM, inhibitor of mitogen-activated-protein kinase – p38MAPK). Using immunofluorescence images, flow cytometry and immunoblotting, we observed a time-dependent intracellular accumulation of fluorescent albumin-FITC-BSA, followed by concentration-and time-dependent effect of intracellular albumin overload on podocyte apoptosis, which was mediated by increased expression of the chaperone glucose-regulated-protein 78 (GRP 78) and phosphorylated inositol-requiring enzyme 1 alpha (pIRE1-α), as well as protein kinase C delta (PKC-δ), p38MAPK and cleaved caspase 12 expression. SB203580 prevented the cleavage of caspase 12 and the albumin-mediated podocyte apoptosis. These results suggest that intracellular albumin overload is associated with endoplasmic reticulum (ER) stress and upregulation of PKC-δ/p38MAPK/caspase 12 pathway, which may be a target for future therapeutic of albumin-induced podocyte apoptosis.

oxygen species (ROS) production are closely related to endoplasmic reticulum (ER) stress, PKC-δ [13][14][15] and p38 MAPK activation, which together are associated with apoptotic responses 13,[16][17][18] . In view of these findings, we sought to explore the hypothesis that under intracellular albumin overload, the ER stress/PKC-δ/p38MAPK/caspase 12 signaling pathway can play a critical role in podocyte apoptosis. Our data advance the understanding of the cell signaling pathways responsible for albumin-induced podocyte apoptosis and may contribute to the development of preventive and therapeutic strategies for albuminuria-associated chronic kidney diseases (CKD).

Results
Expression of podocin and synaptopodin. To validate our in vitro model, differentiated mouse podocytes were subjected to phalloidin immunofluorescence staining and podocin and synaptopodin protein expression (Fig. 1a).
Albumin overload induces podocyte apoptosis. Differentiated podocytes were cultured in BSA-free medium or with increasing concentrations of BSA (1, 5 or 10 mg/mL) at 37 °C for 24 hours. Under these conditions, podocyte apoptosis was documented by flow cytometry. The treatment with albumin 1 mg/mL induced significant increase of the early, late and consequently total podocyte apoptosis compared to control (untreated) cells. Albumin 5 mg/mL induced only early podocyte apoptosis, whereas albumin 10 mg/mL induced a significant increase in the early apoptosis but decreased the late apoptosis. Thus, in these conditions, total podocyte apoptosis did not change compared to the control group (Fig. 1b,c and Table 1).
Time and temperature-dependence of the FITC-BSA internalization in podocyte. We monitored the albumin internalization after stimulation with FITC-BSA (1 mg/mL in serum free medium) at 4 °C for 30 minutes and at 37 °C for 30 minutes, 1 and 3 hours. The signal was detected as green vesicles distributed in the cytosol. The nucleus was stained with DAPi (blue). Figure 2a,b shows that in 1 mg/mL FITC-BSA-treated cells for 30 minutes, the fluorescence intensity was significantly lower at 4 °C in comparison to 37 °C cells [(AU) 4 °C: 1.67 ± 0.10 (n = 9) versus 37 °C: 3.62 ± 0.27(n = 20), **p < 0.01], indicating that the albumin internalization podocytes under control conditions. Phalloidin images were captured on a Zeiss LSM 780 confocal microscope equipped with a 63× objective plan-apochromat zoom factor 1 and a laser excitation of 546 nm for phalloidin. Bar, 20 μm. Immunoblot analysis results are presented for the podocin and synaptopodin protein expression, as well as the internal control glyceraldehyde-3-phosphate dehydrogenase (GAPDH). (b) Representative flow cytometry data exhibiting podocyte apoptosis in the control or treated groups. The cells were treated with BSA (1, 5 and 10 mg/mL) for 24 hours, and apoptosis was evaluated by flow cytometry using FITC Annexin V/7-AAD. Q1, cells in necrosis; Q2, cells in late apoptosis; Q3, cells in early apoptosis and Q4, healthy cells. (c) The values in percentage (%) are expressed as mean ± SEM of 6-9 experiments in triplicate.

Intracellular albumin overload leads to ER stress and PKC-δ phosphorylation in podocytes.
We examined whether albumin overload triggers the ER stress response in podocytes. For these experiments, we used BSA without FITC. As shown in Fig. 3a,b and Table 2, the treatment with 1 mg/mL of BSA at 37 °C for 1 hour, but not for 30 minutes, 3 or 24 hours, induced significant increase in the expression of the chaperone GRP 78/BiP compared to the control group. The increased GRP 78 expression was parallel to the increase of phosphorylated IRE1-α expression, which was extended throughout the other experimental periods.
In addition to the ER stress marker, the treatment of podocytes with 1 mg/mL of BSA for 1, 3 or 24 hours, but not for 30 minutes, induced significant increase in the expression of phosphorylated PKC-δ compared to control group ( Fig. 3c and Table 2). p38MAPK and caspase-12 mediate albumin-induced podocyte apoptosis. Considering that both IRE1-α and PKC-δ can mediate apoptosis through activation of the mitogen-activated protein kinase pathway 19,20 , we next evaluated the expression of the phosphorylated p38MAPK. As shown in Fig. 4a and Table 2, podocytes treated with BSA (1 mg/mL) at 37 °C for 1, 3 and 24 hours, but not for 30 minutes, showed a significant increase of phosphorylated p38MAPK protein expression compared to control group. In addition, the stimulatory effect of albumin on the expression of phosphorylated p38MAPK for all treatment periods is closely related with the increase of cleaved caspase 12 expression ( Fig. 4b and Table 2).

Discussion
In the present in vitro study, we demonstrated that albumin in a concentration-and time-dependent manner induces podocyte injury and apoptosis. These changes are associated with intracellular albumin overload, ER stress, upregulation of PKC-δ, p38MAPK and caspase 12.
Micropuncture studies in rats demonstrated that under physiological conditions, the albumin concentration in Bowman's space is approximately 22.9 µg/mL 6,21 and that this value can increase approximately 40-fold in nephropathy states 22 . On the other hand, in vitro studies revealed that in proximal tubular cells albumin endocytosis (range of 100 µg/mL to 1 mg/mL) was effective within 30 minutes 23 . However, the exposition of human podocytes to extracellular high concentrations of albumin (range of 5 to 10 mg/mL) for 24 and 48 hours led to increased cell death 24,25 . In the current study, using flow cytometry, we observed that prolonged exposure (24 hours) of immortalized mouse podocytes to albumin 1 mg/mL resulted in increased apoptosis. However, 5 and 10 mg/mL albumin failed to induce apoptosis. In part, our results differ from those obtained by Yoshida et al. 25 , possibly due to differences in our experimental design, including time of stimulation and evaluation method, in addition to the mechanism of saturation kinetics 26 .
The albumin internalization mechanisms in podocytes have been the main target of several studies, especially for the relative contribution of the multiple pathways involved in endocytosis process, which remain to be clarified. Endocytosis process can occur through non-specific binding of the molecule to the cell membrane and its subsequent internalization, or may be mediated by specific receptors 27 . Although the albumin endocytosis pathway has not been the focus of the current study, in a previous experimental trial, we observed that mouse podocyte express protein such as clathrin, caveolin-1, V-ATPase E-subunit, megalin and cubilin, which have been demonstrated as mediators of albumin endocytosis in kidney cells [28][29][30][31] . However, for the immunofluorescence analysis, we used an accurate method described by Bitsikas et al. 32 , which count overlap of different staining patterns, and we did not observe significant overlap of FITC-BSA and each protein investigated (data not shown),  suggesting that in our experimental model the albumin internalization process can be mediated by other ways, including free fatty acids (FFAs) bound to albumin, G-protein-coupled receptors and Ras-related C3 botulinum toxin substrate 1 (RAC1) 33 .
Given the apoptotic effect of albumin, we focused our attention on the intracellular albumin overload-mediated podocyte apoptosis signaling in a time-dependent manner. Incubating podocytes with FITC-BSA (1 mg/mL), we observed that albumin-containing vesicles were evidenced in 30 minutes, 1 h and 3 h, but not after 24 hours of treatment. The time-dependent albumin overload occurred in parallel with apoptosis.
It is known that part of the albumin-containing vesicles in podocyte migrate through the cell body to lysosomal degradation 3 . However, persistent intracellular albumin overload may perturb cellular functions through reactive oxygen species generation and ER stress 34 . ER is the site of synthesis and folding of the secreted, membrane-bound and organelle-targeted proteins. Thus, the accumulation of unfolded and/or misfolded proteins in the ER can be referred to as ER stress 35 . In this condition, ER membrane protein kinase R (PKR)-like ER kinase (PERK), inositol-requiring enzyme 1 (IRE1) and activating transcription factor 6 (ATF6) activate the unfolded protein response (UPR). Under physiological conditions, the ER stress sensor proteins such as PERK, IRE1 and ATF6 interact with the chaperone GRP 78, which suppresses their activity. However, in response to ER stress, GRP 78 is recruited from ER stress sensor proteins to promote protein folding. In this condition, the ER stress sensor proteins are auto-activated and phosphorylate their target 36 . Using HEK293 cells, Zhu et al. 15 reported that in response to ER stress, activated IRE1-α recruits the adaptor molecule TNF-receptor-associated factor 2 (TRAF2) to the ER membrane and activates the c-Jun N-terminal kinase (JNK). In addition, it is known that in in vivo and in vitro studies, the PKC-δ participates in ER-stress-induced apoptosis, primarily activating the JNK pathway 20,[37][38][39] . In the current study, the treatment of podocytes with BSA 1 mg/mL for 1 hour resulted in increased GRP 78, phosphorylated IRE1-α and PKC-δ expression. Together, these results indicate that the GRP 78/IRE1-α/PKC-δ signaling pathway contributed to ER stress and induction of albumin-overload-dependent podocyte injury. Additionally, we observed that with 1 mg/mL BSA, the increase of phosphorylated IRE1-α expression (30 min) was prior to the increase of phosphorylated PKC-δ expression, which was significantly increased in 1 hour, and both proteins sustained the phosphorylated states during the apoptotic process, confirming their pivotal role in the cell death. Given that the phosphorylation of PKC-δ is associated with ER stress response in podocytes treated with albumin, it is important to understand the key mechanisms by which PKC-δ initiates and maintains the apoptotic process.
We have recently demonstrated that both PKC-δ and p38MAPK are involved in the angiotensin-II-mediated podocyte apoptosis process 40 . In addition, PKC-δ has been shown to interact with several members of the mitogen-activated protein kinase (MAPK) family, including p38MAPK 20 and NH2 terminal kinase (JNK) 20,39 . Other in vitro studies reported that podocytes from patients with nephrotic-range proteinuria or podocytes from mice exposed to albumin, resulted in p38MAPK-mediated apoptosis 25,41 . However, the pathway linking p38MAPK and podocyte injury was not demonstrated. Our data revealed that albumin-mediated expression of phosphorylated p38MAPK and cleaved caspase 12, as well as podocyte apoptosis, was prevented by SB203580. Together, these results reinforce the role of the p38MAPK/caspase 12 interaction on albumin-induced podocyte apoptosis. Thus, our data provide new evidences that albumin-mediated podocyte apoptosis involves complex intracellular pathways including GRP 78/IRE1-α/PKC-δ/p38MAPK/caspase 12 activation (Fig. 6).

Conclusion
ER stress conditions have been observed in numerous pathological events including cardiovascular and renal diseases. Despite its importance in pathological conditions, several complex questions remain regarding ER-stress-induced apoptosis. Given that ER stress mediators are well defined and specific, they could be useful targets for therapy. Therefore, our current study demonstrated a crosstalk between albumin overload/ER stress/ podocyte apoptosis and identified new targets for the prevention of podocyte injury in nephropathies associated with albuminuria.

Albumin internalization.
Statistical analysis. The data are reported as the mean ± SEM. For comparisons among the groups, a one-way ANOVA followed by Bonferroni's test (GraphPad Prism 6 Software, San Diego, USA) was performed.
The differences with p < 0.05 were considered statistically significant.
Ethics approval. All of the experimental protocols were conducted in accordance with the ethical standards approved by the Institutional Animal Care and Use Committee of the University of Sao Paulo (Protocols no. 792/2015).

Data Availability Statement
All data generated or analyzed during this study are included in this published article (and its Supplementary  Information Files).