Inhibition of hepatocyte nuclear factor 1β contributes to cisplatin nephrotoxicity via regulation of nf‐κb pathway

Abstract Cisplatin nephrotoxicity has been considered as serious side effect caused by cisplatin‐based chemotherapy. Recent evidence indicates that renal tubular cell apoptosis and inflammation contribute to the progression of cisplatin‐induced acute kidney injury (AKI). Hepatocyte nuclear factor 1β (HNF1β) has been reported to regulate the development of kidney cystogenesis, diabetic nephrotoxicity, etc However, the regulatory mechanism of HNF1β in cisplatin nephrotoxicity is largely unknown. In the present study, we examined the effects of HNF1β deficiency on the development of cisplatin‐induced AKI in vitro and in vivo. HNF1β down‐regulation exacerbated cisplatin‐induced RPTC apoptosis by indirectly inducing NF‐κB p65 phosphorylation and nuclear translocation. HNF1β knockdown C57BL/6 mice were constructed by injecting intravenously with HNF1β‐interfering shRNA and PEI. The HNF1β scramble and knockdown mice were treated with 30 mg/kg cisplatin for 3 days to induce acute kidney injury. Cisplatin treatment caused increased caspase 3 cleavage and p65 phosphorylation, elevated serum urea nitrogen and creatinine, and obvious histological damage of kidney such as fractured tubules in control mice, which were enhanced in HNF1β knockdown mice. These results suggest that HNF1β may ameliorate cisplatin nephrotoxicity in vitro and in vivo, probably through regulating NF‐κB signalling pathway.

the NF-κB translocates to nucleus and transcriptionally activates target gene expression such as interleukin-1β (IL-1β) and tumour necrosis factor α (TNF-α). 1 Cisplatin has the ability to induce NF-κB phosphorylation in HEK293 cells and further decrease the secretion of pro-inflammatory cytokines such as IL-1β and TNF-α. 2 Cisplatin also can increase NF-κB expression in kidney tissues of rat. 3 During cisplatin-induced mouse renal injury, NF-κB activation, IκBα phosphorylation and p65 protein nuclear translocation were observed. Moreover, the P53 protein level, caspase 3/9 and the poly (ADP-ribose) polymerase (PARP) cleavage were also increased. 4,5 The transcriptional inhibition of NF-κB was reported to be able to ameliorate cisplatin-induced AKI. 6 However, the upstream genes regulating NF-κB transcriptional activation during cisplatin nephrotoxicity are still not fully understood.
The transcription factor HNF1β (hepatocyte nuclear factor 1 homeobox B) was found to be abundantly expressed in kidney and pancreas, 7 but not in brain and heart tissues. HNF1β was expressed higher in kidney than in lung and liver tissues. 8 Early reports have linked HNF1β regulation to kidney cystogenesis and development. 9 Patients with HNF1β mutations often complicated with diabetes, renal cysts and loss of kidney function. [10][11][12][13] Upon exposure to TNF-α and interferon-γ (IFN-γ), the proximal tubular cells exhibited inhibited transcriptional activity of HNF1β and dysfunction of mitochondria. 14 Thus, the transcriptional activity of HNF1β may be related to the inflammatory response of renal tubular cells.
HNF1β was involved in regulating various pathological processes. HNF1β knockdown in proximal tubular HK2 (human kidney 2) cells promoted the epithelial-to-mesenchymal transition. 15 HNF1β positively regulated the proliferation and tubulogenesis of normal kidney proximal epithelial cells (NRK-52E). 16 In our previous study, HNF1β was found to play protective role in cisplatin-induced tubular cell apoptosis in vitro. 17 However, in Wilms' tumour-derived G401 cells, HNF1β was reported to decrease cell proliferation and migratory abilities and increase apoptotic rate no matter through overexpression or small interfering RNA-silencing experiments. 18 Therefore, the role and mechanism of HNF1β in cisplatin-induced apoptosis still need further analysis.
In this study, the in vivo role of HNF1β in cisplatin nephrotoxicity was examined. Moreover, the regulatory function of HNF1β on NF-κB activation was analysed.
The construction of HNF1β knockdown mice was performed as described in previous studies with mild modifications. [19][20][21] Briefly, the HNF1β scramble or short hairpin RNA (shRNA) was mixed with PEI (Polyethyleneimine, 765090; Sigma) at a ratio of 1:2. C57BL/6 mice aged about 8 weeks were injected with 40 μg HNF1β scramble or shRNA plasmids by tail vein. Mice were randomly divided into three groups: Scramble group in which mice were treated with a non-specific shRNA for 24 hours (n = 6); and shRNA-HNF1β groups in which mice were treated with specific HNF1β shRNA for 24 hours (n = 8) or 72 hours (n = 8).
Mice in each group were intraperitoneally injected with 30 mg/kg cisplatin, and the cervical venous blood and kidneys of the mice were collected at 24, 48 and 72 hours after modelling, respectively.
To further evaluate the impact of HNF1β interference on cisplatin-induced nephrotoxicity, C57BL/6 male mice aged 6-8 weeks were randomly divided into four groups as follows: shRNA-scramble group, shRNA-HNF1β group, cisplatin with shRNA-scramble group and cisplatin with shRNA-HNF1β group. Mice were given tail injection of 40 μg HNF1β-interfering plasmids. An hour later, these mice were intraperitoneally injected with PBS or 30 mg/kg cisplatin for 72 hours. The mice were anaesthetized by intraperitoneal injection of 2% chloral hydrate (0.015 mL/g; Esite Biotechnology, China). The mouse kidneys were extracted after modelling for further analysis.
The mouse serum was separated from venous blood samples by centrifugation at 12 000 g at room temperature for 5 minutes. Renal function was assessed by measuring serum creatinine (0420-500; Stanbio) and blood urea nitrogen (BUN) (0580-250; Stanbio) following Jaffe's and GLDH reactions.
Then, Hoechst 33342 (10 μg/mL; Beyotime Biotechnology, China) was used for nuclear staining. The apoptotic cells were then visualized under a Nikon Inverted Microscope Eclipse Ti-E (Nikon, Japan) and counted from five randomly selected fields by ImageJ software.

| Immunofluorescence
The RPTC HNF1β-negative control (NC) cells and knockdown (KD) cells were treated with or without 20 μmol/L cisplatin. Briefly, 1 × 10 5 cells were seeded into 35-mm dishes containing slides with a diameter of 1 cm and then were cultured until 80% confluence.  Thermo Fisher Scientific) for 10 minutes. The images were acquired by Nikon Inverted Microscope Eclipse Ti-E (Nikon, Japan), and the nuclear translocation was quantified using ImageJ software as previously described. 22

| TUNEL staining
The cells with a confluence of up to 80% were extracted from the cell incubator, namely RPTC HNF-1 NC cells and RPTC HNF-1 KD cells.

| Cytoplasm and nucleus protein extraction
RPTC HNF1β NC or KD cells were treated with or without 20 μmol/L cisplatin for 24 hours, and then, the cells were digested by SDS (S8010, Solarbio) lysis buffer and were centrifuged at 500 g for 5 minutes. The centrifuged precipitates were further used for isolation of cytoplasm and nucleus proteins by NE-PER nuclear and cytoplasmic extraction reagents (78833, Thermo Scientific).

| Western blot
RPTCs were treated with or without 20 μmol/L cisplatin for 24 hours.
Total cell proteins were extracted using SDS lysis buffer containing Millipore Company) was used for signal detection. The protein bands were quantified using ImageJ software.

| Haematoxylin and eosin (H&E) staining
Kidney cortex tissue sections were placed in a 60℃ oven for 4 hours, dewaxed with xylene in fume hood for 3 times, each time for 10 minutes, and then dehydrated with 100%, 96%, 90%, 80% and 70% concentrations of ethanol for 5 minutes, respectively. After washing with running tap water for 10 minutes, the sections were stained with haematoxylin for 30 seconds, decolorized with acid ethanol (1% HCl and 99% medical alcohol) for 1-2 seconds and then stained with eosin for 40 seconds. The sections were observed and photographed under a Nikon Inverted Microscope Eclipse Ti-E (Nikon, Japan).

| Immunohistochemical staining
The dewaxing and dehydrating processes were the same as HE staining. The tissues were blocked with 3% hydrogen peroxide for 20 minutes at room temperature in the dark. Then, the sections were sealed with 5% BSA for about 30 minutes at room temperature and were incubated with primary antibodies against cleaved cas-

| Statistical analysis
The results of this experiment were all repeated for 3 times, and Prism (GraphPad Software 7) was used for data analysis. t Test, oneway ANOVA and two-way ANOVA were used for data comparison between groups. The data were expressed as means ± SEM, and P < 0.05 was considered as statistically significant.

| Inhibition of HNF1β exacerbated cisplatininduced RPTC apoptosis
We previously determined the protective role of HNF1β in cisplatininduced RPTC apoptosis; however, the exact mechanism of HNF1β regulating cisplatin nephrotoxicity is still unclear. In this study, the HNF1β-NF-κB signalling pathway during cisplatin-induced apoptosis was examined. RPTC HNF1β scramble (negative control, NC) and HNF1β shRNA (knockdown, KD) cells were treated with 20 μmol/L cisplatin for 24 hours, and total proteins were extracted for Western blot analysis. HNF1β expression level was down-regulated to about 65.8% after stimulation of cisplatin ( Figure 1A,B). Hoechst staining of RPTC nuclei indicated the cisplatin-induced apoptosis of HNF1β knockdown cells was significantly increased in comparison with that of RPTC HNF1β NC cells (data not shown). The cytosolic protein GAPDH and nucleic protein lamin B1 were used as internal control for Western blot analysis ( Figure 2C). The translocation of phosphor-P65 and total P65 from nucleus to cytoplasm was quantified using ImageJ software and summarized as Figure 2D.

| Inhibition of HNF1β-induced NF-κB signalling activity during cisplatin nephrotoxicity
Cisplatin promoted the phosphor-P65 and total P65 translocation by 2.69 and 3.39 times, respectively. The phosphor-P65 and total P65 translocation in HNF1β KD cells was up-regulated by 2.33 and 2.05 times, compared with HNF1β NC cells.

| Suppression of p65 phosphorylation rescued RPTC apoptosis induced by HNF1β down-regulation during cisplatin treatment
To further identify the impact of NF-κB p65 signals on cisplatin-induced RPTC HNF1β KD cell apoptosis, the p65 nuclear translocation inhibitor TPCA-1 was used for stimulation of RPTCs together with cisplatin. The apoptotic morphological changes in RPTCs were observed through the Hoechst staining assay ( Figure 3A). HNF1β interference increased the fluorescence intensity of apoptotic cells induced by cisplatin, which were inhibited by TPCA-1 ( Figure 3B). In addition, Western blot analy-

| HNF1β was not colocalized with NF-κB p65 during cisplatin-induced RPTC apoptosis
As it was indicated that HNF1β was protective against cisplatin-induced RPTC apoptosis, the nuclear translocation of HNF1β was analysed using immunofluorescent staining. As is shown in Figure 4A, cisplatin obviously promoted colocalization of HNF1β (green signals) and cell nucleus (blue signals). The quantification of fluorescence intensities of green and blue signals confirmed the results ( Figure 4B).
To further analyse whether HNF1β as a nuclear transcriptional regulatory factor colocalized with P65 protein, the cytoplasm and nucleus proteins of RPTCs treated with or without cisplatin were isolated and used for co-immunoprecipitation (IP) assay. To our surprise, whether or not treated with cisplatin in RPTCs, HNF1β did not colocalize with P65 neither in cytoplasm nor in nucleus ( Figure 4C).

| HNF1β knockdown C57BL/6 mice were successfully constructed
To further analyse the function of HNF1β in cisplatin nephrotoxicity in vivo, HNF1β knockdown C57BL/6 mice were constructed by injecting intravenously with 40 μg HNF1β scramble shRNA and interfering shRNA. The kidney cortex after injection of scramble shRNA or interfering shRNA for 1 and 3 days was used for haematoxylin and eosin (H&E) and immunohistochemical staining ( Figure 5A) and F I G U R E 4 HNF1β nuclear translocation in RPTCs affected by cisplatin treatment. RPTCs were treated with or without 20 μmol/L cisplatin for 24 h and then were immunostained with anti-HNF1β antibody and FITClabelled secondary antibody. The cell nucleus was stained with DAPI. (A) Fluorescence images were obtained using laser confocal microscope. Scale bars are 10 μm. (B) The fluorescence intensities of green and blue signals in Figure 4A were quantified using ImageJ software. (C) Co-immunoprecipitation of HNF1β and P65 proteins in cytoplasm and nucleus of RPTCs treated with or without cisplatin. Lamin B1 and GAPDH were used as internal controls of nuclear and cytoplasmic proteins, respectively Western blot analysis ( Figure 5B). As shown in Figure 5A, the tubules and glomeruli of mice were relatively intact and neatly arranged after the injection of interfering plasmid for 1 and 3 days, indicating that injection of interfering plasmid has no obvious toxic effects on renal tissue. About 44.4% and 75.1% of inhibition of HNF1β signalling were observed in renal cortex at 1 or 3 days after injection with HNF1β knockdown plasmid compared with scramble plasmid ( Figure 5C). Furthermore, Western blot assay confirmed about 44.7% and 87% reduction in HNF1β expression in renal cortex at 1 or 3 days after injection with HNF1β knockdown plasmid compared with scramble plasmid ( Figure 5D).

| HNF1β ameliorated cisplatin-induced nephrotoxicity in vivo
The expression of HNF1β was further analysed in C57BL/6 mice after cisplatin treatment for 3 days. Western blot results indicated the expression level of HNF1β was reduced after cisplatin treatment for 2 and 3 days ( Figure 6A,B). The extent of caspase 3 cleavage was also enhanced after cisplatin treatment for 2 and 3 days ( Figure 6C,D). The immunohistochemical staining of cleaved caspase 3 further confirmed the cisplatin-induced nephrotoxicity in day 2 and day 3 ( Figure 6E,F).
To further explore the functional effect of HNF1β, the HNF1β scramble and knockdown mice were treated with cisplatin for 3 days to induce acute kidney injury. HE staining indicated that the tubules in the sham groups injected with scramble or HNF1β shRNA were intact and neatly arranged. Cisplatin treatment caused obvious histological damage of kidney such as fractured tubules in control mice, which were enhanced in HNF1β knockdown mice ( Figure 7A,B).

| D ISCUSS I ON
In this study, we found that HNF1β protects from cisplatin-induced kidney injury both in vivo and in vitro. In addition, HNF1β may play the protective role by negatively regulate the NF-κB signalling pathway. Cisplatin treatment promoted the nuclear translocation of HNF1β and p65 NF-κB; however, no obvious interaction between these two proteins was observed in RPTCs either with or without cisplatin treatment.
It was surprising for us to observe the down-regulation of  Figure 5B 1 to 24 hours after hypoxia stimulation in kidney proximal tubular HK2 (human kidney 2) cells and was down-regulated under prolonged 1% oxygen treatment [15]. The expression of HNF1β was also down-regulated in cystic kidneys [18]. Taken together, the results demonstrated that HNF1β expression was time-dependent and quite possible to be reduced after long-term stimulation. In addition, the transcriptional activation of HNF1β may largely depend on the nuclear translocation.
NF-κB activation triggers the release of cytochrome c from mitochondria, and stimulates the intrinsic apoptotic pathway, whereas P53 mainly activates extrinsic apoptotic pathway. 23,24 Upon cisplatin stimulation, HNF1β reduced the phosphorylation of p65 NF-κB at serine-536 but not p53, indicating that HNF1β may affect cisplatin nephrotoxicity through regulating the intrinsic apoptotic pathway.
HNF1α is the other member of HNF1 transcription factor family.
HNF1α nuclear translocation was inhibited during C2-ceramide-induced hepatocyte injury. 25 In addition, HNF1α increased the p65 NF-κB expression and nuclear accumulation, leading to NF-κB signalling activation. 26 However, in our study, HNF1β nuclear translocation was enhanced after cisplatin treatment and the NF-κB signalling was increased after HNF1β interference, suggesting that HNF1β may regulate cell apoptosis in the opposite manner of

HNF1α.
Moreover, in addition to NF-κB nuclear translocation, the NFAT5 (nuclear factor of activated T cells-5) was also reported to translocate from cytoplasm to nucleus during ultraviolet B irradiation-stimulating human lens epithelial cells. The interaction between NFAT5 and NF-κB p65 subunit was also increased. 27 Thus, NFAT5 may also be the potential downstream target of HNF1β. In this study, we did not observe the colocalization of HNF1β and NF-κB p65 subunit neither in cytoplasm nor in nucleus, suggesting the HNF1β may indirectly modulate NF-κB P65 signalling pathway.
In our previous study, blockade of NF-κB decreased cisplatin-induced microRNA-375 expression and further increased HNF1β activity, 17 suggesting NF-κB can negatively regulate HNF1β activity. In this study, we confirmed that HNF1β can influence NF-κB signalling pathway in turn, indicating that there is a mechanism of negative feedback regulation of apoptosis pathway mediated by HNF1β/ NF-κB.
In conclusion, the inhibition of HNF1β induced by cisplatin contributes to nephrotoxicity either in vitro or in vivo. HNF1β and NF-κB signalling pathway can indirectly regulate each other and play important roles in cisplatin-induced acute kidney injury.

F I G U R E 6
Protein expression in renal cortex tissues after treatment with cisplatin. C57BL/6 mice were injected intraperitoneally with 30 mg/kg cisplatin for 0, 1, 2 and 3 d. (A, B) Western blot analysis of HNF1β expression and caspase 3 cleavage in renal cortex. GAPDH was used as internal loading control. (C, D) The ratios of HNF1β and cleaved caspase 3 in relation to GAPDH in Figure 6A,B were quantified using ImageJ software. (E) IHC analysis of cleaved caspase 3 expression in renal cortex tissues. (F) The mean integral optical density (IOD/area) was assessed on each section from six random fields using Image-Pro Plus 6.0 software

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.