PSTPIP2 ameliorates aristolochic acid nephropathy by suppressing interleukin-19-mediated neutrophil extracellular trap formation

Aristolochic acid nephropathy (AAN) is a progressive kidney disease caused by herbal medicines. Previously, we found that proline–serine–threonine phosphatase interacting protein 2 (PSTPIP2) and neutrophil extracellular traps (NETs) play important roles in kidney injury and immune defense, respectively; however, the mechanism of AAN regulation by PSTPIP2 and NETs remains unclear. We found that renal tubular epithelial cell (RTEC) apoptosis, neutrophil infiltration, and inflammatory factor and NET production were increased in a mouse model of AAN, while PSTPIP2 expression was low. Conditional knock-in of PSTPIP2 in mouse kidneys inhibited cell apoptosis, reduced neutrophil infiltration, suppressed the production of inflammatory factors and NETs, and ameliorated renal dysfunction. In contrast, restoring normal PSTPIP2 expression promoted kidney injury. In vivo, the use of Ly6G-neutralizing antibody to remove neutrophils and peptidyl arginine deiminase 4 (PAD4) inhibitors to prevent NET formation reduced apoptosis, thereby alleviating kidney injury. In vitro, damaged RTECs released interleukin-19 (IL-19) via the PSTPIP2/nuclear factor (NF)-κB pathway and induced NET formation via the IL-20Rβ receptor. Concurrently, NETs promoted the apoptosis of damaged RTECs. PSTPIP2 affected NET formation by regulating IL-19 expression via inhibition of NF-κB pathway activation in RTECs, inhibiting their apoptosis and reducing kidney damage.


59
Aristolochic acid nephropathy (AAN) refers to the deterioration in kidney function 60 associated with toxic AAs (5). AAN is one of the most serious complications 61 associated with the use of traditional Chinese medicines, and cases of the disease have 62 been reported globally (6). Both mice and rats are susceptible to AA toxicity, allowing 63 in vivo studies on the mechanisms of AA-induced renal toxicity. Acute renal failure 64 with tubular necrosis can be induced in these model animals by a single high dose of 9 infiltration and cytokine production, which was reduced by renal PSTPIP2 183 overexpression. 184 We hypothesized that the transient depletion of neutrophils may provide therapeutic 185 benefits in AAI-induced kidney injury. We first examined whether treatment with a 186 specific Ly6G neutralizing antibody 1 d before and after AAI injection could deplete 187 their circulating and renal neutrophil levels. Control mice were treated with the same 188 dosage of rat IgG isotype control antibody ( Figure S3A). Compared to the control, we 189 found that the Ly6G antibody could specifically prevent the AAI-induced increase in Although increased neutrophil infiltration in the kidney is the major cause of AAN, 207 the mechanism of NET formation following AAI exposure remains unknown. To 208 evaluate NET formation, mice received a single IP injection of AAI and were 209 sacrificed after 3 d. Serum nucleosome, MPO, and dsDNA levels were significantly 210 increased after AAI treatment ( Figure 5A-C). Cit-H3 protein levels, another marker 211 of NETs, were also significantly increased after AAI treatment, as observed by 212 fluorescence microscopy and western blotting ( Figure 5D and E). We found that 213 serum nucleosome, MPO, and dsDNA levels ( Figure 5F-H), as well as Cit-H3 levels 214 ( Figure 5I and J), were significantly lower in AAI-treated PSTPIP2-cKI mice than 215 those in AAI-treated FF mice. Finally, we found that serum levels of nucleosomes, 216 dsDNA, and MPO, as well as the expression of Cit-H3, were significantly increased 217 in AAI-treated mice after PSTPIP2 expression was restored compared with those in 218 AAI-treated PSTPIP2-cKI mice ( Figure S5A-E). Our results demonstrate that 219 PSTPIP2 may play a vital role in the regulation of NET formation. 220 NET formation during programmed neutrophil cell death (NETosis) has recently been 221 shown to play a pro-injury role in renal inflammatory diseases (18). Accordingly, we 222 explored whether NETosis increases kidney injury induced by AAI. For this purpose, 223 we intravenously injected DNase I or GSK484, a potent inhibitor of NET formation, 224 into the AAI-treated mice ( Figure S6A). Serum nucleosome, MPO, and dsDNA levels 225 ( Figure S6B-D), as well as Cit-H3 levels ( Figure S6E and F), were significantly 226 11 decreased in mice treated with DNase I or GSK484. We tested the effects of DNase I 227 and GSK484 on the progression of AAN. Treatment with DNase I or GSK484 228 significantly reduced serum Cr and BUN levels ( Figure S7A and B) and attenuated 229 tissue damage ( Figure S7C) in AAI-treated mice. Moreover, IHC analysis showed 230 that the KIM-1 and TUNEL + cell levels were significantly decreased in DNase I-or 231 GSK484-treated mice ( Figure S7D and E). The activity of caspase-3 and level of 232 cleaved caspase-3 protein were also significantly downregulated in AAI-treated mice 233 treated with GSK484 or DNase I ( Figure S7F and G). Based on these observations, 234 we surmised that the capacity of neutrophils to cause or exacerbate apoptosis is linked 235 to their NET formation properties. Primary renal neutrophils were isolated from AAI-   249 To better characterize the specific role of PSTPIP2 in RTECs, overexpression of 250 PSTPIP2 was induced by transfection with the PSTPIP2 plasmid (pEX-2-PSTPIP2) in 251 mRTECs. We confirmed that the pEX-2-PSTPIP2 plasmid elevated PSTPIP2 252 expression ( Figure S9D-E). We stimulated RTECs with AAI for 20 h to explore 253 whether they secrete important cytokines ( Figure 6A). Real-time PCR analysis 254 showed that the expression of IL-19 was most significantly elevated compared with 255 other interleukins in AAI-treated mRTECs, and overexpression of PSTPIP2 256 significantly decreased IL-19 expression compared with that in control plasmid-257 transfected cells ( Figure 6B). We found that the level of IL-19 was significantly 258 increased in the supernatant of AAI-treated mRTECs ( Figure 6C), consistent with a 259 drastic increase in its mRNA and protein levels ( Figure 6D and E). Conversely, we 260 found that the expression of IL-19 was significantly reduced after overexpression of 261 PSTPIP2 in mRTECs (Figures 6K-L and S9F). Using immunofluorescence staining, 262 we observed that E-cadherin was colocalized with IL-19, suggesting that RTECs are 263 the major source of IL-19 after AAI-induced renal injury ( Figure S9A). Serum IL-19 264 levels have been shown to be increased with various kidney diseases, including 265 kidney transplants, hydronephrosis accompanied by ureteral stones, diabetes, and 266 polycystic kidney disease (Table S1). Notably, serum IL-19 and Cr levels were 267 positively correlated (R = 0.5472, P < 0.05; Figure 6F). The expression of IL-19 was 268 significantly lower in AAI-treated PSTPIP2-cKI mice than that in AAI-treated FF

272
IL-19 is a potent cytokine in neutropenia treatment (19). We assessed whether  is chemotactic to mouse bone marrow-derived neutrophils using a Transwell system 274 ( Figure 7A) and found that IL-19 stimulation was able to induce neutrophil migration, 275 especially at a concentration of 80 ng/mL ( Figure 7B). Although  shown to enhance neutrophil recruitment, it is unclear whether IL-19 induces NET 277 formation. We used a series of different concentrations of recombinant  278 to stimulate mouse bone marrow-derived neutrophils ( Figure S9G) for 4 h ( Figure 7C).

279
The protein levels of IL-20Rβ, the IL-19 membrane receptor, increased drastically 280 after stimulation with various rIL-19 concentrations ( Figure 7G and 7I), suggesting 281 that neutrophils express IL-20Rβ, which is upregulated in response to rIL-19 exposure. 282 We sought to determine whether IL-19 induces NET formation. Accordingly, we  Consistently, the activity of caspase-3 and level of cleaved caspase-3 protein were 293 significantly upregulated after stimulation with IL-19-derived NETs ( Figure S10B 294 and C).

295
The transcription factor NF-κB has been reported to regulate IL-19 gene transcription 296 (19), but it is unclear whether PSTPIP2 regulates IL-19 expression by activating or 297 inhibiting the NF-κB pathway. In addition, to verify the correlation between PSTPIP2 298 and NF-κB p65, we used mRTEC extract for co-immunoprecipitation (Co-IP) 299 experiments using anti-NF-κB p65 and anti-PSTPIP2 antibodies. The Co-IP results 300 confirmed an interaction between PSTPIP2 and NF-κB p65 in mRTECs ( Figure 8A).

301
As expected, NF-κB was strongly activated in AAI-treated mRTECs, as evidenced by 302 the markedly increased phosphorylation (S536) and nuclear localization of NF-κB 303 p65, accompanied by enhanced phosphorylation of IκBα (S32/36), and 304 overexpression of PSTPIP2 inhibited AAI-induced activation of NF-κB signaling 305 ( Figure 8B and C). In addition, single IF staining showed that the overexpression of 306 PSTPIP2 significantly inhibited the nuclear transfer of p65 ( Figure 8D). The NF-κB 307 pathway was inhibited after a 2-h pretreatment with the NF-κB inhibitor PDTC (25 308 μM; Figure 8F), immediately after which, we removed the cell supernatant and 309 retreated the cells with AAI for 20 h ( Figure 8E). Real-time PCR analysis showed that 310 IL-19 was visibly upregulated in the PDTC+AAI-treated group compared to the AAI-311 treated group ( Figure 8G). Collectively, these results indicate that PSTPIP2 inhibits 312 NF-κB signaling to induce IL-19 release from the renal tubular epithelial cells. This study provides a new perspective on the role of PSTPIP2 in acute AAN and 316 identifies a potential mechanism of PSTPIP2 in inhibiting epithelial apoptosis to 317 alleviate AAI-induced nephrotoxicity. We confirmed that PSTPIP2 was significantly 318 downregulated in both AAI-induced acute AAN mouse models and AAI-exposed 319 mRTECs. We also found that the conditional knock-in of PSTPIP2 in mouse kidneys 320 was sufficient to ameliorate renal dysfunction, histological injury, cell apoptosis, and 321 inflammatory responses induced by AAI in acute AAN mouse models. Accordingly, 322 the restoration of normal PSTPIP2 expression in PSTPIP2-cKI mice promoted renal 323 injury and cell apoptosis, confirming the critical role of PSTPIP2 in acute AAN.

324
Mechanistically, we found that PSTPIP2 served as a negative regulator of acute AAN, 325 at least in part, via reducing apoptosis of RTECs by regulating the IL-19-mediated 326 formation of NETs. Our results therefore support the potential of PSTPIP2-targeted 327 therapy in patients with acute AAN.

328
First, we examined the role of PSTPIP2 in renal injury and inflammation. PSTPIP2 329 plays a crucial role in autoimmune diseases and is involved in macrophage activation, 330 neutrophil migration, and cytokine production (20 NETs. In our current study, the levels of NET markers, such as nucleosome, MPO, 367 dsDNA, and Cit-H3, were significantly increased in AAI-exposed mice, and 368 conditional knock-in of PSTPIP2 expression in the kidney reduced NET formation.

369
However, reducing PSTPIP2 expression in mouse kidneys significantly increased 370 NET formation. Indeed, a previous study showed that AAN kidneys were protected by 371 treatment with the inhibitor of NET formation GSK484 or DNase I (37), which is 372 consistent with the findings in glomerular disease (38).

373
PSTPIP2 plays an important role in cytokine production. Notably, the expression of 374 IL-19 was the most significantly elevated among the interleukins in AAI-treated 375 mRTECs. Overexpression of PSTPIP2 significantly decreased IL-19 mRNA 376 expression. In the current study, we performed a complementary retrospective 377 association study using serum samples from patients with various kidney diseases.

378
The results indicated that IL-19 was positively correlated with serum Cr level ( Figure   379 6F), a classic indicator of kidney injury. IL-19 is a member of the IL-10 family of 380 18 cytokines, which has indispensable functions in many inflammatory processes. It is 381 also a member of the IL-20 subfamily and binds to the IL-20 receptor to regulate 382 various processes, such as antimicrobial activity, wound healing, and tissue 383 remodeling (39). Cellular sources have been characterized as having myeloid and 384 epithelial origins, which correspond with the significant levels found in the 385 RPTEC/TERT1 renal epithelial cell system. IL-19 participates in several animal and 386 human diseases, including psoriasis, inflammatory bowel disease, atherogenesis, 387 endotoxic shock, rheumatoid arthritis, and cancer (39) protected mice from ischemic AKI. We previously found that apoptosis of RTECs 409 contributed to cisplatin nephrotoxicity (45). We also showed for the first time that controversial. In the current study, TUNEL + cell numbers and caspase-3 activity were 416 significantly increased in the kidneys of mice with AAN, whereas they were 417 significantly decreased in the kidneys of PSTPIP2 conditional knock-in mice.

418
Conversely, the level of apoptosis was markedly increased with a reduction in renal 419 PSTPIP2 expression. We investigated whether NETs also developed during AAI-420 induced apoptosis and found that the kidneys of neutrophil-depleted mice treated with 421 the Ly6G antibody and NET-depleted mice treated with GSK484 (a PAD4 inhibitor)     DNase I (10 U/100 μL, 0.9% NaCl) or GSK484 (4 mg/kg) was IP injected daily for 3 461 d prior to AAI treatment, and six mice per group were separately sacrificed at 3-d    The primary renal neutrophils of mice subjected to IP injections of AAI and DNase I

491
To quantify serum NET levels, we used mouse MPO ELISA kits to detect serum MPO 492 levels. Serum nucleosome quantification was performed using the Cell Death kit.

493
DNA levels in free serum and cell supernatants were quantified using the PicoGreen 494 assay kit according to the manufacturer's instructions.   Data are expressed as the mean ± standard error of the mean (SEM) from at least three 508 independent experiments. Differences between two groups were compared using a 509 two-tailed Student's t-test. Differences between multiple groups were compared using 510 a one-way analysis of variance (ANOVA) followed by Tukey's post hoc test.

870
Data are presented as the mean ± SEM of six biological replicates per condition. Each 871 dot represents a sample. Statistically significant differences were determined by an 872 independent sample t-test (A-D) and one-way ANOVA followed by Tukey's post hoc 873 test (F-I). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: non-significant.  Data are presented as the mean ± SEM of three biological replicates per condition.

924
Each dot represents a sample. Statistically significant differences were determined by 925 an independent sample t-test and one-way ANOVA followed by Tukey's post hoc test 926 (D-G). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: non-significant.  Statistically significant differences were determined by an independent sample t-test 948 and one-way ANOVA followed by Tukey's post hoc test (B,C,F,G). *P < 0.05, **P < 949 0.01, ***P < 0.001, ****P < 0.0001, ns: non-significant. Mouse bone marrow-derived neutrophils were extracted as follows: Femurs and 50 tibias were aseptically extracted after the animals were sacrificed, and the red marrow 51 cavities were exposed after removing the cartilage at both ends. A 1-mL sterile 52 syringe was used to aspirate a small volume of dilution containing 10% standard fetal 53 bovine serum or serum-containing medium, the marrow cavity was rinsed to obtain 54 marrow, and a single-cell suspension was prepared. Mature neutrophil extraction and 55 purification were performed using the mouse bone marrow neutrophil isolation kit, 56 according to the manufacturer's instructions.

57
To isolate and purify mouse renal tissue-derived primary neutrophils, a single-cell 58 suspension was prepared using a gentleMACS™ Dissociator and the Multi Tissue

208
Total RNA was extracted from kidney tissues using TRIzol reagent (Invitrogen).