AA

KKD

a

ARRAA

KPENTD

1

riTksr2ao2iia

1h

The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943

Contents lists available at SciVerse ScienceDirect

The International Journal of Biochemistry& Cell Biology

journa l h o me page: www.elsev ier .com/ locate /b ioce l

TP-P2X4 signaling mediates NLRP3 inflammasome activation: novel pathway of diabetic nephropathy

ehong Chen1, Jianguo Zhang1, Weiwei Zhang, Jinhua Zhang, Jurong Yang,ailong Li, Yani He ∗

epartment of Nephrology, Daping Hospital, Research Institute of Surgery, Third Military Medical University, Chongqing 400042, PR China

r t i c l e i n f o

rticle history:eceived 27 November 2012eceived in revised form 6 February 2013ccepted 11 February 2013vailable online 19 February 2013

eywords:2X4xtracellular ATPLRP3 inflammasomeubulointerstitial inflammationiabetic nephropathy

a b s t r a c t

Tubulointerstitial inflammation plays a key role in the development of diabetic nephropathy (DN).Cytokines in the IL-1 family are the key pro-inflammatory cytokines of tubulointerstitial inflammation.Extracellular ATP can cause P2X receptors to activate the NOD-like receptor 3 (NLRP3) inflammasomeand cause IL-1� and IL-18 maturation and release. We investigated the role of ATP-P2X4 signaling inNLRP3 inflammasome activation and renal interstitial inflammation characteristic of DN. Ex vivo stud-ies, P2X4 showed increased expression in renal tubule epithelial cells in patients with nephropathy dueto type 2 diabetes compared to those in the control group. Linear correlation analysis shows that P2X4expression was positively related with urine IL-1� and IL-18 levels. Moreover, P2X4 expression was co-localized with NLRP3, IL-1�, and IL-18 expression. In vitro culture experiments showed NLRP3 proteinexpression, cleavage of caspase-1 and IL-1�, and release of IL-1�, IL-18 and ATP in HK-2 cells signifi-cantly increased after high glucose stimulation. However, apyrase, which consumes extracellular ATP,

completely blocked the changes caused by high glucose. The P2 receptor antagonist suramin, P2X recep-tor antagonist TNP-ATP, P2X4 selective antagonist 5-BDBD, and P2X4 gene silencing attenuated NLRP3expression, cleavage of caspase-1 and IL-1�, and release of IL-1� and IL-18 induced by high glucose.Taken together, these results suggest that ATP-P2X4 signaling mediates high glucose-induced activationof the NLRP3 inflammasome, regulates IL-1 family cytokine secretion, and causes the development oftubulointerstitial inflammation in DN.

. Introduction

Diabetic nephropathy (DN) is the leading cause of end-stageenal disease (Collins et al., 2012). Tubulointerstitial inflammations crucial in promoting the development and progression of DN.he NOD-like receptor 3 (NLRP3) inflammasome is activated inidney tissue of streptozotocin-induced diabetic rats, and suppres-ion of NLRP3 inflammasome activation can significantly reduceenal tissue inflammation and improve renal function (Wang et al.,012). NLRP3 knock-out mice are protected from renal tubule dam-ge and renal interstitial inflammation in kidney unilateral ureteralcclusion (UUO) and ischemia–reperfusion models (Shigeoka et al.,010; Vilaysane et al., 2010). These data show that the NLRP3

nflammasome plays a key role in the process of kidney sterile

nflammation. However, the mechanism of NLRP3 inflammasomectivation in DN is still unclear.

∗ Corresponding author. Tel.: +86 23 68757871; fax: +86 23 68893060.E-mail address: heynmail@yahoo.com (Y. He).

1 Both authors contributed equally to the study.

357-2725/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.biocel.2013.02.009

© 2013 Elsevier Ltd. All rights reserved.

A previous study reported that high glucose stimulated adeno-sine triphosphate (ATP) secretion by renal inherent cells (Soliniet al., 2005). Increased extracellular ATP (eATP) activates the P2receptors of immune cells (Trautmann, 2009) and initiates theimmune reaction (Vitiello et al., 2012). Activation of P2 receptorscauses the production of reactive oxygen species, chemokines andinflammatory markers, which consequently cause inflammation(Bours et al., 2011). It has been reported that eATP can activate theNLRP3 inflammasome and causes aseptic inflammation (Iyer et al.,2009).

Hyperglycemia can also cause retinal cell caspase-1 activationand interleukin (IL) -1� secretion in an eATP dependent man-ner (Trueblood et al., 2011). IL-1 family cytokines are pluripotentimmunomodulatory cytokines that play a key role in tubulointersti-tial inflammation in DN (Navarro-Gonzalez and Mora-Fernandez,2008). They have direct relationships with many chronic compli-cations of diabetes since they are able to induce production ofvarious inflammatory factors (chemokines, metal proteases, tumor

necrosis factor-� (TNF-�), etc.) and inflammatory cell infiltration(Devaraj et al., 2007; Vincent and Mohr, 2007). Pro-IL-1� and pro-IL-18, the precursor molecules of IL-1 family cytokines, rely oncaspase-1 shearing action to exert their biological effects (Weber

Bioch

etatdS

sHm(cioPesbrpit

2

2

rDmarapnPlfitMs

2

bb2cmwrw

2

bpraNaa

K. Chen et al. / The International Journal of

t al., 2010a). The NLRP3 inflammasome, the intracellular recep-or of danger-associated molecular pattern molecules (DAMPs), canctivate caspase-1, regulate IL-1� and IL-18 maturation and secre-ion, and trigger the inflammation of metabolic diseases such asiabetes, obesity, and gout (Liu-Bryan, 2010; Masters et al., 2010;tienstra et al., 2011).

Activated purinergic P2X receptors mediate endogenous DAMPsuch as ATP, biglycan, and amyloid protein (Babelova et al., 2009;alle et al., 2008; Riteau et al., 2010), activate the NLRP3 inflam-asome, and regulate aseptic inflammation in hypersensitivity

Weber et al., 2010b), acute pancreatitis (Hoque et al., 2011) andancer (Ghiringhelli et al., 2009). P2X receptors are ligand-gatedon channel receptors in the P2 family. There are seven subtypesf P2X receptors (P2X1-7). All are expressed in the kidney except2X3 (Turner et al., 2003). Renal tubule epithelial cells mainlyxpress P2X4 and P2X6. P2X receptor activation can promote renalodium/water discharge (Jankowski et al., 2011) and regulate renallood flow (Crawford et al., 2011) in physiological conditions. Theole of P2X4 in diabetic nephropathy is still undefined. The pur-ose of this study is to investigate the role of ATP-P2X4 signaling

n NLRP3 inflammasome activation and renal interstitial inflamma-ion of DN.

. Materials and methods

.1. Patients

A total of 45 patients with type 2 diabetic nephropathy wereecruited for this study from the Department of Nephrology inaping Hospital from January 1, 2011 to June 1, 2012. The enroll-ent criteria were as follows: patients were 40–70 years old with

history of type 2 diabetes; 24 h urine protein was above 150 mg;enal biopsy pathology led to a diagnosis of diabetic nephropathy;nd no fever with obvious infection lesions or high uric acid. Allatients used insulin to control blood glucose, angiotensin antago-ist and CCB to control blood pressure, and statins to control lipids.atients abstained from traditional Chinese medicine or sulfony-ureas for three months after renal biopsy. Normal kidney tissuesrom nephrectomies of renal hamartoma were collected to be usedn the control group. The protocol for this study was approved byhe Ethical and Protocol Review Committee of the Third Military

edical University, and informed consent was obtained from theubjects.

.2. Biochemical analysis

Blood and urine samples were collected 1 day prior to renaliopsy for biochemical analysis. Serum creatinine was measuredy the modified Jaffé rate-blanked alkaline picrate method. The4 h urinary protein excretion was measured by the benzethoniumhloride method. Serum uric acid was determined by an enzymaticethod. HbA1c was calculated from total glycated hemoglobin,hich was determined by the ion capture method; the normal

ange was <6.5%. The estimated glomerular filtration rate (eGFR)as calculated by the Cockcroft–Gault formula.

.3. Immunohistochemistry analysis

The expression of P2X4 and NLRP3 proteins was determinedy a 2-step immunohistochemical staining technique, as describedreviously (Liu et al., 2011). Specimens were deparaffinized andehydrated. After antigen retrieval, polyclonal primary anti-P2X4

ntibody (sc-28764, Santa Cruz Biotechnology, USA) and anti-LRP3 antibody (ab4207, Abcam, UK), rabbit IgG isotype controlntibody (ab27472, Abcam, UK) and goat IgG isotype controlntibody (sc-3887, Santa Cruz Biotechnology, USA) were added

emistry & Cell Biology 45 (2013) 932– 943 933

and specimens were incubated at 4 ◦C overnight. IgG-conjugatedhorseradish peroxidase (HRP) and 3,3-diaminobenzidine tetrahy-drochloride (ZLI-9032, Zhong Shan Golden Bridge BiologicalTechnology, Beijing, China) were employed to visualize antibodybinding.

Ten high power fields were randomly selected. Areas of brownnuclear staining for P2X4, NLRP3, IL-1� or IL-18 were counted andexpressed as the percentage of total renal tubular epithelial cells(RTECs). The stained areas were rated as described previously (Liuet al., 2011): 0, no staining or positive staining in <10% of RTECs; 1,weak positive staining in 10% to 35% of RTECs; 2, moderate positivestaining in 35–70% of RTECs; and 3, strong positive staining in >70%of RTECs. The counting of microscopic fields was performed by twoblinded pathologists.

2.4. Confocal fluorescence analysis

Tissue sections were blocked and incubated with anti-P2X4antibody or anti-NLRP3 antibody together with anti-IL-1� antibody(sc-52012, Santa Cruz Biotechnology, USA) or anti-IL-18 antibody(sc-133127, Santa Cruz Biotechnology, USA) at 4 ◦C overnight.After rinsing with PBS, the samples were stained with fluoresceinisothiocyanate conjugated (A0562 or A0568, Beyotime, China) orCy3-conjugated goat anti-mouse IgG (A0502 or A0521, Beyotime,China) for 60 min at 37 ◦C. Nuclei were stained with DAPI. The sam-ples were mounted with glycerol and visualized under a confocalscanning microscope (Lcssp-2, Leica, Germany).

2.5. Cell culture and siRNA transfection

HK-2 cells (CRL-2190, American Type Culture Collection) werecultured in Dulbecco’s modified Eagle’s medium with low glu-cose (HyClone, USA). The cells were seeded at 1.5 × 106 cells/10 cmdiameter dish for 2–3 days at 37 ◦C in a humidified atmospherecontaining 5% CO2. HK-2 cells were grown in media with nor-mal glucose concentration (5 mM), or high glucose concentration(15, 25, 35, and 50 mM), or high mannitol concentrations (5 mMglucose + 30 mM mannitol) for 48 h. HK-2 cells were incubated inmedia with high glucose (35 mM, 48 h) with and without 5 U/mlapyrase (Sigma–Aldrich, USA), 100 �M suramin (Sigma–Aldrich,USA), 10 �M TNP-ATP (Sigma–Aldrich, USA), or 2 �M 5-BDBD(Tocris Bioscience, USA).

Transfection of siRNA was performed as previously described(Wu et al., 2010). Transfection control cells were transfected withsiRNA for unrelated genes (fluorescein conjugated control siRNAs,sc-36869, Santa Cruz Biotechnology, USA) or control siRNA (sc-37007, Santa Cruz Biotechnology, USA), while experimental cellswere transfected with siRNA directed against P2X4 (sequence:GTA CTA CAG AGA CCT GGCT), with all transfections utilizingLipofectamine 2000 (Invitrogen, Carlsbad, CA) according to themanufacturer’s instructions. Cells incubated with Lipofectamine2000 were also used as a negative control. P2X4 knockdown wasconfirmed by Western blot and quantitative reverse transcription-PCR (qRT-PCR).

2.6. RNA Extraction and real-time quantitative PCR

Total RNA was isolated using RNAout (TOYOBO, Japan) accord-ing to the manufacturer’s protocol, and P2X4 mRNA was evaluatedby quantitative real-time PCR. The primers for P2X4 weredesigned based on GenBank sequences (accession numbers:NM 001256796.1) and synthesized (Sagon Inc., Shanghai, China).

�-Actin was employed as an internal control, and its primer wasdesigned based on the GenBank sequence (accession number for�-actin: NM 001017992.2). cDNAs for real-time quantitative PCRwere synthesized using total RNAs from cell lysates. To avoid

9 Biochemistry & Cell Biology 45 (2013) 932– 943

cDsRGpB9fen

2

s2irtwcuiabwaUobt(R

2

wi

2

cbfliDa

twwctwcwscu

Table 1Clinical characteristics of patients with type 2 diabetic nephropathy and controls.

Controls Diabeticnephropathypatients

Number of patients (n) 11 45Sex (M/F) (n) 6/5 26/19Age (years) 47 ± 11 54 ± 10Duration of diabetes (years) – 7.3 ± 2.5BMI (kg/m2) 22.5 ± 2.7 25.4 ± 4.9Fasting blood glucose (mmol/l) 4.9 ± 0.7 6.1 ± 1.8HbA1c (%) 5.0 ± 0.7 7.8 ± 2.0**

Serum uric acid (�mol/l) 267.1 ± 48.8 279.1 ± 34.724 h urinary protein (g/24 h) <0.23 2.0 ± 0.8**

SBP (mmHg) 119.7 ± 6.9 144.4 ± 20.8**

DBP (mmHg) 78.5 ± 6.1 78.3 ± 11.5Estimated GFR (ml/min/1.73 m2) 107.3 ± 11.9 88.3 ± 28.8Urinary IL-1� (pg/ml) 65.1 ± 5.5 135.8 ± 22.7**

Urinary IL-18 (pg/ml) 58.6 ± 10.2 130.9 ± 21.5**

Data are mean ± SD. GFR, glomerular filtration rate; BMI, body mass index; HbA1c,

34 K. Chen et al. / The International Journal of

ontamination with genomic DNA, total RNAs were treated withNaseI (Invitrogen Corp., Carlsbad, CA, USA). cDNA was synthe-

ized with oligo (dT) primers in a 20 �l reaction with 2 �g of totalNA using the ImProm reverse transcription system (TaKaRa Bioroup, Japan) according to the manufacturer’s protocol. A 2 �l sam-le of cDNA was added to the SYBR Green PCR Master Mix (TaKaRaio Group, Japan) and subjected to PCR amplification (1 cycle at5 ◦C for 10 s, and 40 cycles at 95 ◦C for 5 s, 55 ◦C for 20 s, and 72 ◦Cor 30 s) in an iCycler system (Bio-Rad, Hercules, CA, USA). Relativexpression was calculated by using the 2−��Ct method with valuesormalized to the reference gene �-actin.

.7. Western blot analysis

Western blot analysis using whole-cell lysates and cell cultureupernatants was performed as described previously (Wu et al.,010; Hornung et al., 2008). Briefly, HK-2 cells (1 × 106) were lysed

n 200 �l of RIPA lysate buffer (Pierce, USA). Supernatants wereetained, assayed for protein content by the Bradford method, andhen subjected to SDS-PAGE. Cell culture supernatants (400 �l)ere precipitated by the addition of 400 �l methanol and 100 �l

hloroform, vortexed and centrifuged for 10 min at 20,000 × g. Thepper phase was discarded and 500 �l methanol was added to the

nterphase. This mixture was centrifuged for 10 min at 20,000 × gnd the protein pellet was dried at 55 ◦C, resuspended in Laemmliuffer and separated by 15% SDS-PAGE. After blocking, membranesere incubated with primary antibodies anti-NLRP3, anti-P2X4,

nti-IL-1�, anti-caspase-1 (sc-56036, Santa Cruz Biotechnology,SA), or anti-�-actin (sc-1616, Santa Cruz Biotechnology, USA)vernight at 4 ◦C. The membranes were then washed and incu-ated with secondary HRP-conjugated antibodies for 1 h at roomemperature. Specific bands were detected by using an ECL systemAmersham) and a Bio-Rad electrophoresis image analyzer (Bio-ad, Hemel Hampstead, UK).

.8. Enzymed-linked immunosorbent assay (ELISA)

IL-1�, IL-18 and IL-6 levels in cell culture supernatants and urineere determined using ELISA kits, according to the manufacturer’s

nstructions (Genmed Scientifics Inc., Shanghai, China).

.9. Intracellular K+ and Ca2+ measurement

After exposure to the defined experimental conditions, HK-2ells were incubated in 1.5 ml of media containing 5 �M potassium-inding benzofuran isophthalate-acetoxymethylester (PBFI-AM, auorescent K+ indicator) for 40 min at 37 ◦C. PBFI-AM fluorescent

mages of cells were acquired using an inverted microscope (LeicaMIRB, Leica Microsystems). The dye was excited at 340 nm. Imagenalysis was performed using Image J software (NIH, USA).

Fluo-3-acetoxymethyl ester (Fluo-3/AM; Invitrogen) was usedo estimate cytosolic free [Ca2+]. HK-2 cells cultured on coverslipsere incubated with Fluo-3/AM (5 �M) for 30 min at 37 ◦C. Cellsere washed with fresh DMEM and then scanned under a laser

onfocal microscope (Lcssp-2, Leica) with illumination by a kryp-on/argon laser at 488 nm excitation, and the emitted fluorescenceas captured at 526 nm. To ensure efficient quantum capture, the

ells were placed at the bottom of a recording chamber and images

ere recorded after 10–20 s (when fluorescence intensity became

table). During the scanning period of 120–180 s, the mean fluores-ence became constant and the average fluorescence intensity wassed to estimate the change of intracellular [Ca2+] in HK-2 cells.

glycosylated hemoglobin; SBP, systolic blood pressure; DBP, diastolic blood pres-sure.

** P < 0.01 vs. controls.

2.10. Statistical analysis

Statistical analyses were performed with SPSS 13.0 software.Data is usually presented as mean ± SD. The relationship betweentwo sets of variables was determined by linear correlation analysis.Other data from experiments were analyzed by t-test or ANOVAwherever appropriate. Difference was considered significant whenthe P value was less than 0.05.

3. Results

3.1. P2X4 expression related with renal interstitial inflammationin DN

45 patients with type 2 diabetes nephropathy and 11 non-diabetic cases with renal hamartoma were recruited in this study.Table 1 shows that HbA1c, 24 h urine protein, and systolic bloodpressure were higher in patients with type 2 diabetic nephropa-thy than those of the control group. Urine IL-1� and IL-18 levelswere also significantly higher in the DN group (P < 0.01, Table 1).Immunohistochemistry analysis revealed that NLRP3 expressionsignificantly increased in renal tubular epithelial cells in the DNgroup (Fig. 1A–C, and Supplementary Fig. 1A and B). Linear corre-lation analysis shows that NLRP3 expression score was positivelyrelated with urine IL-1� and IL-18 levels in the DN group (P < 0.001,Fig. 1D and E). Confocal microscopy analysis suggests that NLRP3expression was co-localized with IL-1� and IL-18 expression inrenal tubule epithelial cells of DN patients (Fig. 1F and G). Anti-NLRP3 antibody activity was antigen specific as demonstrated usingisotype control (Supplementary Figs. 1C and 2A).

In order to analyze the relationship between P2X4 expressionlevel and renal interstitial inflammation, immunohistochemistrywas used to detect P2X4 expression. P2X4 immunostaining scorewas higher in renal tubular epithelial cells of the DN group(Fig. 2A–C and Supplementary Fig. 1D and E). Moreover, P2X4immunostaining score was positively correlated with urine IL-1�and IL-18 levels (P < 0.001, Fig. 2D and E). P2X4 expression wasco-localized with NLRP3, IL-1�, and IL-18 expression in the DNgroup (Fig. 2F–H). Anti-P2X4 antibody activity was antigen spe-cific as demonstrated using isotype control (Supplementary Figs.

1F and 2B). These results suggest that P2X4 expression is signif-icantly increased in renal tubule cells of type 2 diabetic patientswith nephropathy, and is closely related to NLRP3 inflammsomeactivation and renal interstitial inflammation.

K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943 935

Fig. 1. NLRP3 expression is correlated with urine IL-1� and IL-18 in patients with type 2 diabetic nephropathy. (A and B) Immunohistochemistry detected NLRP3 expressionin renal tubule epithelial cells in control patients (A), and patients with type 2 diabetic nephropathy (B) (scale bars, 40 �m). (C) NLRP3 expression score. (D and E) Ther 2 nd (E) 2

d

3i

isaioIsIpt

elationship of NLRP3 expression with (D) urine IL-1� level (r = 0.673, P < 0.001), aetected co-localization of NLRP3 and IL-1� (F) and IL-18 (G) (scale bars, 40 �m).

.2. High glucose-induced NLRP3 expression and NLRP3nflammasome activation is extracellular ATP-dependent

NLRP3 inflammasome activation involves inflammasome prim-ng via upregulation of NLRP3 followed by triggering inflamma-ome formation, which activates caspase-1 leading to maturationnd secretion of IL-1� (Bauernfeind et al., 2009). In order tonvestigate the effect of high glucose on NLRP3 inflammasomef RTECs, NLRP3 protein expression, cleavage of caspase-1 andL-1�, and release of IL-1� and IL-18 were analyzed. NLRP3 expres-

ion, cleavage of caspase-1and IL-1�, and release of IL-1� andL-18 were enhanced in a dose dependent way (P < 0.05), with theeaks at 35 mM glucose stimulation (Fig. 3A–F). In cells that werereated with 35 mM glucose for 72 h, NLRP3 expression, cleavage

urine IL-18 level (r = 0.654, P < 0.001) was analyzed. (F and G) Cofocal microscopy

of caspase-1and IL-1�, and release of IL-1� and IL-18 changed ina time dependent way (P < 0.05), with peaks at 48 h (Fig. 3G–L).NLRP3 expression, cleavage of caspase-1and IL-1�, and release ofIL-1� and IL-18 were not affected by mannitol stimulation (P > 0.05,Supplementary Fig. 3).

ATP levels of cell culture supernatants gradually increased(P < 0.05), with peaks at 35 mM and 12 h, and were 5.8 times and5.4 times higher compared to those of control, respectively. How-ever, high glucose stimulation did not change intracellular ATPlevels (P > 0.05, Supplementary Fig. 4). As an isotonic comparison,

mannitol stimulation (5 mM glucose + 30 mM mannitol) did notchange the cell culture supernatant and intracellular ATP levels(P > 0.05, Supplementary Fig. 5). To further study whether highglucose stimulation requires eATP in order to activate the NLRP3

936 K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943

Fig. 2. P2X4 expression is correlated with NLRP3, IL-1�, and IL-18 expression in patients with type 2 diabetic nephropathy. (A and B) Immunohistochemistry detectedP2X4 expression in renal tubule epithelial cells in control patients (A), and patients with type 2 diabetic nephropathy (B) (scale bars, 40 �m). (C) P2X4 expression score.( .001)c

iAIea

D and E) The relation of P2X4 expression and (D) urine IL-1� level (r2 = 0.677, P < 0o-localization of P2X4 and NLRP3 (F), IL-1� (G), or IL-18 (H) (scale bars, 40 �m).

nflammasome, apyrase (an enzyme that degrades extracellular

TP, 5 U/ml) was used to consume eATP, and then IL-1� and

L-18 levels were measured. Apyrase significantly inhibited NLRP3xpression, cleavage of caspase-1and IL-1�, and release of IL-1�nd IL-18 caused by high glucose (P < 0.05, Fig. 4).

and (E) urine IL-18 level (r2 = 0.627, P < 0.001). (F–H) Cofocal microscopy detected

3.3. High glucose caused ATP-P2X4 signaling activation

It has been reported eATP can activate P2 receptors (P2X andP2Y) on the cell membrane to cause increased intracellular calciumconcentration and decreased potassium concentration (Hattori

K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943 937

Fig. 3. Time- and dose-dependent effects of glucose on NLRP3 inflammasome activation in HK-2 cells. (A) Western blot analyses of NLRP3 protein expression in HK - 2 cellsexposed to increasing glucose concentrations for 48 h. (B) Western blot analyses of cleaved caspase-1 p10 and 17 kDa mature IL-1� in culture supernatants. (C and D) Relativequantitation of cleaved caspase-1 p10 (C) and mature IL-1� (D). (E and F) ELISA analysis of IL-1� (E) and IL-18 (F) levels in culture supernatants. (G) Western blot analysesof NLRP3 protein expression in HK-2 cells treated with 35 mM d-glucose medium at different time-points. (H) Western blot analyses of cleaved caspase-1 p10 and 17 kDamature IL-1� in cell culture supernatants. (I and J) Relative quantitation of cleaved caspase-1 (I) and mature IL-1� (J). (K and L) ELISA analysis of IL-1� (K) and IL-18 (L) levelsin cell culture supernatants. Values are mean ± SD of 3 independent experiments with triplicate dishes. *P < 0.05 vs. 5 mM or 0 h glucose; **P < 0.01 vs. 5 mM or 0 h glucose.

938 K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943

Fig. 4. Apyrase inhibits NLRP3 expression and NLRP3 inflammasome activation caused by high glucose. HK-2 cells were incubated in 5 mM (control) or 35 mM glucose (highglucose) for 48 h with or without apyrase (5 U/ml). (A) Western blot analysis of NLRP3 protein expression and its relative quantitation. (B) Western blot analysis of activatedcaspase-1 and mature IL-1� in cell culture supernatants. (C and D) Relative quantitation of cleaved caspase-1 p10 (C) and mature IL-1� (D). (E and F) ELISA analysis of IL-1�(E) and IL-18 (F) levels in cell culture supernatants. Values are mean ± SD of 3 independent experiments with triplicate dishes. **P < 0.01 vs. high glucose.

Fig. 5. Time-dependent effect of high glucose on intracellular K+ and Ca2+ concentrations in HK-2 cells. HK-2 cells were treated with 5 mM (control) or 35 mM glucose(high glucose) for different time periods (24–72 h). (A and B) PBFI-AM fluorescence (A) (scale bars, 100 �m) and estimation of intracellular potassium ion concentration byfluorescence intensity (B). (C and D) Fluo-3/AM fluorescence (C) (scale bars, 80 �m) and estimation of intracellular calcium ion concentration by fluorescence intensity (D).Values are mean ± SD of 3 independent experiments with triplicate dishes. **P < 0.01 vs. control.

K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943 939

Fig. 6. Effects of different antagonists on intracellular Ca2+ and K+ concentration changes in HK-2 cells under high glucose conditions. Cells were cultured in low glucose(5 mM) medium, high glucose (35 mM) medium, or high glucose medium with suramin (100 �M), TNP-ATP (10 �M) or 5-BDBD (2 �M). (A and B) PBFI-AM fluorescence (A,u tion by8 ntens*

awtwpm(ig

osPwc5tTtHc

vttgrs

pper) (scale bars, 100 �m) and estimation of intracellular potassium ion concentra0 �m) and estimation of intracellular calcium ion concentration by fluorescence i*P < 0.01 vs. high glucose.

nd Gouaux, 2012; Troadec et al., 2000). In order to determinehether P2 receptors were activated under high glucose condi-

ions, intracellular potassium and calcium concentration changesere estimated at different time points (24–72 h). Intracellularotassium concentration gradually declined and reached a mini-um at 48 h, which was about 70% of that in the control group

Fig. 5A and B). In contrast, intracellular calcium concentrationncreased to a peak at 48 h, which was 1.8 times that of the controlroup (Fig. 5C and D).

To further investigate what kind of P2 receptor subtypespened under high glucose conditions, P2 receptor antagonisturamin (100 �M), P2X receptor antagonist TNP-ATP (10 �M) and2X4 selective antagonist 5-BDBD (2 �M) were employed alongith high glucose, and the intracellular potassium and calcium

oncentration changes were determined. Suramin, TNP-ATP, and-BDBD significantly inhibited the potassium concentration reduc-ion caused by high glucose (P < 0.05, Fig. 6A and B). Moreover,NP-ATP and 5-BDBD restrained the intracellular calcium concen-ration elevation caused by high glucose (P < 0.05, Fig. 6A and C).owever, suramin had no effect on the increase in the calciumoncentration caused by high glucose (P > 0.05, Fig. 6C).

To explore whether high glucose requires eATP to acti-ate the P2X4 receptor, apyrase and high glucose were addedogether. Apyrase significantly inhibited the potassium concen-

ration decrease and calcium concentration increase under highlucose conditions (Fig. 7A–C), which suggests that P2X4 activationequires eATP. Western blotting found that P2X4 protein expres-ion significantly increased with high glucose stimulation to a level

fluorescence intensity (B). (A and C) Fluo-3/AM fluorescence (A, lower) (scale bars,ity (C). Values are mean ± SD of 3 independent experiments with triplicate dishes.

2.82 times of that in the control group and apyrase did not affectP2X4 protein expression level (Fig. 7D and E). These results indicatethat high glucose does not require eATP to induce P2X4 expression.

3.4. P2X4 receptor mediated increased NLRP3, IL-1 ̌ and IL-18expression caused by high glucose

P2 receptor antagonists were used to investigate whether P2X4mediates NLRP3 inflammasome activation under high glucose con-ditions. Suramin, TNP-ATP and 5-BDBD reduced the increase inNLRP3 expression, cleavage of caspase-1 and IL-1�, and release ofIL-1� and IL-18 caused by high glucose stimulation (Fig. 8A–G).IL-6 in culture supernatants was not affected with mannitol stimu-lation (Fig. 8H). P2X4 siRNA transfection in HK-2 cells significantlydecreased P2X4 mRNA and protein expression levels (Fig. 9A andB, and Supplementary Fig. 6A). P2X4 gene silencing inhibited intra-cellular potassium efflux and extracellular calcium influx inducedby high glucose. The increases in NLRP3 expression, cleavage ofcaspase-1and IL-1�, and secretion of IL-1� and IL-18 were sim-ilarly blunted (Fig. 9C–I, and Supplementary Fig. 6B). The P2X4receptor may be the key receptor mediating NLRP3 inflammasomeactivation under high glucose conditions.

4. Discussion

We found that P2X4 expression was significantly increased inrenal tubule cells of type 2 diabetic patients with nephropathy,and is closely related to NLRP3 inflammsome activation and renal

940 K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943

Fig. 7. Extracellular apyrase inhibits P2X4 opening, but does not reduce increased P2X4 expression induced by high glucose. HK-2 cells were incubated in 5 mM (control)or 35 mM glucose (high glucose) for 48 h with or without apyrase (5 U/ml). (A and B) PBFI-AM fluorescence (A, upper) (scale bars, 100 �m) and estimation of intracellularpotassium ion concentration by fluorescence intensity (B). (A and C) Fluo-3/AM fluorescence (A, lower) (scale bars, 80 �m) and estimation of intracellular calcium ionc 2X4 pi

iNaMcctgcPgirskTn

oncentration by fluorescence intensity (C). (D and E) Western blot analyses of Pndependent experiments with triplicate dishes. **P < 0.01 vs. high glucose.

nterstitial inflammation. Further in vitro study showed that theLRP3 inflammasome was activated by high glucose stimulation,nd at the same time eATP levels were significantly increased.oreover, potassium concentration decreased and calcium con-

entration increased with high glucose stimulation. This effectan be blocked by the eATP-consuming enzyme apyrase andhe nonspecific P2 receptors antagonist suramin. Therefore, highlucose-induced changes in intracellular potassium and calciumoncentrations depend upon the activation of P2 receptors. The2X receptor antagonist TNP-ATP completely blocked the highlucose-induced potassium and calcium concentration changes,ndicating that eATP mainly activates P2X receptors rather than P2Yeceptors. Furthermore, the effect of 5-BDBD in preventing potas-

ium and calcium concentration changes indicates that P2X4 is theey receptor mediating intracellular ionic concentration changes.he mechanism of P2X4 expression induced by high glucose isot yet completely understood. It was reported that high glucose

rotein expression (D) and its relative quantitation (E); values are mean ± SD of 3

stimulation can activate JAK/STAT signaling pathways by increas-ing the activity of transcription factor STAT1 (Nareika et al., 2009),which would then increase P2X4 expression (Tang et al., 2008).

Copious evidence indicates that NLRP3 inflammasome activa-tion may regulate the inflammation process of diabetes and relatedcomplications (Masters et al., 2010; Wang et al., 2012). We foundthat apyrase could completely block NLRP3 inflammasome activa-tion caused by high glucose. Similarly, apyrase can block caspase-1activation in rat retinal cells caused by high glucose stimulation(Trueblood et al., 2011). Therefore, eATP might be the key moleculein high glucose-induced NLRP3 inflammasome activation.

The subtype of the specific P2 receptor subtypes involved in thisprocess has been controversial. While some studies found that P2X7

receptor opening is an important causal factor for NLRP3 inflam-masome activation (Niemi et al., 2011; Qu et al., 2009), others havedisagreed this notion based upon data from islet cells (Zhou et al.,2009), retinal cells (Trueblood et al., 2011) and fat cells (Sun et al.,

K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943 941

Fig. 8. NLRP3 expression and inflammasome activation in HK-2 cells cultured under high glucose conditions with different P2 receptor antagonists. HK-2 cells were incubatedin 5 mM (control) or 35 mM glucose (high glucose) with or without suramin (100 �M), TNP-ATP (10 �M) or 5-BDBD (2 �M). (A and B) Western blot analyses of NLRP3 proteine aspasc G) ande

2tbipc

t

xpression (A) and its relative quantitation (B). (C) Western blot analyses of cleaved cleaved caspase-1 (D) and mature IL-1� (E). (F-H) ELISA analysis of IL-1� (F), IL-18 (xperiments with triplicate dishes. **P < 0.01 vs. high glucose.

012). In the present study, we found that the specific P2X4 recep-or antagonist 5-BDBD as well as P2X4 receptor gene knockdownlocked NLRP3 inflammasome activation induced by high glucose,

ndicating that P2X4 may be the major ion channel facilitating

otassium outflow caused by high glucose in renal tubule epithelialells.

We found that the P2X4 receptor participates in the regula-ion of NLRP3 inflammasome activation, and also in the process of

e-1 and mature IL-1� in cell culture supernatants. (D and E) Relative quantitation of IL-6 (H) levels in cell culture supernatants. Values are mean ± SD of 3 independent

NLRP3 protein expression caused by high glucose. P2X4 receptoractivation causes cellular low potassium, activates many nucleartranscription factors such as c-Jun (Subramaniam et al., 2003), NF-�B (Tao et al., 2006), Sp1, Sp3 and CREB-1 (Wang et al., 2007), and

then increases NLRP3 gene expression. The details of the underlyingmechanism still need further research.

In conclusion, we report that ATP-P2X4 signaling activationplays an important role in the inflammation of the kidney. High

942 K. Chen et al. / The International Journal of Biochemistry & Cell Biology 45 (2013) 932– 943

Fig. 9. P2X4 gene silencing suppresses NLRP3 expression and inflammasome activation induced by high glucose. Cells were transfected using LipofectamineTM 2000 withsiRNA control (50 nm), or P2X4 siRNA (50 nm). (A) QRT-PCR-detected P2X4 mRNA expression levels. (B) Detection of P2X4 protein expression by Western blot and its relativequantitation. **P < 0.01 vs. control. (C) Western blot analysis of NLRP3 protein expression and its relative quantitation. (D) Western blot analysis of cleaved caspase-1 andm d caspl ents w

gPssnn

ature IL-1b in cell culture supernatants. (E and F) Relative quantitation of cleaveevels in cell culture supernatants. Values are mean ± SD of 3 independent experim

lucose causes NLRP3 inflammasome activation through ATP-2X4 signaling, regulates IL-1 family cytokine maturation and

ecretion by renal tubular epithelial cells, and leads to renal inter-titial inflammation in diabetic nephropathy. This may provide aew endogenous anti-inflammatory target for delaying diabeticephropathy progression to end-stage renal disease.

ase-1 (E) and mature IL-1� (F). (H and I) ELISA analysis of IL-1� (H) and IL-18 (I)ith triplicate dishes. **P < 0.01 vs. high glucose.

Acknowledgments

This work was supported by grants from the National NaturalScience Foundation of China (No. 81070575). The authors acknowl-edge Dr. Lirong Lin, Dr. Jun Liu and Mr. Jun Zhan for their clinicaland technical help. The authors declare no conflict of interest.

Bioch

A

fj

R

B

B

B

C

C

D

G

H

H

H

H

I

J

L

L

M

N

N

K. Chen et al. / The International Journal of

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/.biocel.2013.02.009.

eferences

abelova A, Moreth K, Tsalastra-Greul W, Zeng-Brouwers J, Eickelberg O, YoungMF, et al. Biglycan, a danger signal that activates the NLRP3 inflammasome viatoll-like and P2X receptors. Journal of Biological Chemistry 2009;284:24035–48.

auernfeind FG, Horvath G, Stutz A, Alnemri ES, MacDonald K, Speert D, et al. Cuttingedge: NF-kappaB activating pattern recognition and cytokine receptors licenseNLRP3 inflammasome activation by regulating NLRP3 expression. Journal ofImmunology 2009;183:787–91.

ours MJ, Dagnelie PC, Giuliani AL, Wesselius A, Di Virgilio F. P2 receptors andextracellular ATP: a novel homeostatic pathway in inflammation. Frontiers inBioscience (Scholar Edition) 2011;3:1443–56.

ollins AJ, Foley RN, Chavers B, Gilbertson D, Herzog C, Johansen K, et al. UnitedStates Renal Data System 2011 Annual Data Report: Atlas of chronic kidney dis-ease & end-stage renal disease in the United States. American Journal of KidneyDiseases 2012;59(A7):e1–420.

rawford C, Kennedy-Lydon TM, Callaghan H, Sprott C, Simmons RL, Sawbridge L,et al. Extracellular nucleotides affect pericyte-mediated regulation of rat in situvasa recta diameter. Acta Physiologica (Oxford, England) 2011;202:241–51.

evaraj S, Cheung AT, Jialal I, Griffen SC, Nguyen D, Glaser N, et al. Evidenceof increased inflammation and microcirculatory abnormalities in patientswith type 1 diabetes and their role in microvascular complications. Diabetes2007;56:2790–6.

hiringhelli F, Apetoh L, Tesniere A, Aymeric L, Ma Y, Ortiz C, et al. Activation of theNLRP3 inflammasome in dendritic cells induces IL-1beta-dependent adaptiveimmunity against tumors. Nature Medicine 2009;15:1170–8.

alle A, Hornung V, Petzold GC, Stewart CR, Monks BG, Reinheckel T, et al. The NALP3inflammasome is involved in the innate immune response to amyloid-�. NatureImmunology 2008;9:857–65.

attori M, Gouaux E. Molecular mechanism of ATP binding and ion channel activa-tion in P2X receptors. Nature 2012;485:207–12.

oque R, Sohail M, Malik A, Sarwar S, Luo Y, Shah A, et al. TLR9 and the NLRP3inflammasome link acinar cell death with inflammation in acute pancreatitis.Gastroenterology 2011;141:358–69.

ornung V, Bauernfeind F, Halle A, Samstad EO, Kono H, Rock KL, et al. Silica crystalsand aluminum salts activate the NALP3 inflammasome through phagosomaldestabilization. Nature Immunology 2008;9:847–56.

yer SS, Pulskens WP, Sadler JJ, Butter LM, Teske GJ, Ulland TK, et al. Necroticcells trigger a sterile inflammatory response through the Nlrp3 inflammasome.Proceedings of the National Academy of Sciences of the United States of America2009;106:20388–93.

ankowski M, Szamocka E, Kowalski R, Angielski S, Szczepanska-Konkel M. Theeffects of P2X receptor agonists on renal sodium and water excretion in anaes-thetized rats. Acta Physiologica (Oxford, England) 2011;202:193–201.

iu-Bryan R. Intracellular innate immunity in gouty arthritis: role of NALP3 inflam-masome. Immunology and Cell Biology 2010;88:20–3.

iu J, Li K, He Y, Zhang J, Wang H, Yang J, et al. Anticubilin antisense RNA amelioratesadriamycin-induced tubulointerstitial injury in experimental rats. AmericanJournal of the Medical Sciences 2011;342:494–502.

asters SL, Dunne A, Subramanian SL, Hull RL, Tannahill GM, Sharp FA, et al.Activation of the NLRP3 inflammasome by islet amyloid polypeptide providesa mechanism for enhanced IL-1beta in type 2 diabetes. Nature Immunology2010;11:897–904.

areika A, Sundararaj KP, Im YB, Game BA, Lopes-Virella MF, Huang Y. High glu-cose and interferon gamma synergistically stimulate MMP-1 expression in U937

macrophages by increasing transcription factor STAT1 activity. Atherosclerosis2009;202:363–71.

avarro-Gonzalez JF, Mora-Fernandez C. The role of inflammatory cytokinesin diabetic nephropathy. Journal of the American Society of Nephrology2008;19:433–42.

emistry & Cell Biology 45 (2013) 932– 943 943

Niemi K, Teirila L, Lappalainen J, Rajamaki K, Baumann MH, Oorni K, et al. Serumamyloid A activates the NLRP3 inflammasome via P2X7 receptor and a cathepsinB-sensitive pathway. Journal of Immunology 2011;186:6119–28.

Qu Y, Ramachandra L, Mohr S, Franchi L, Harding CV, Nunez G, et al. P2X7receptor-stimulated secretion of MHC class II-containing exosomes requires theASC/NLRP3 inflammasome but is independent of caspase-1. Journal of Immunol-ogy 2009;182:5052–62.

Riteau N, Gasse P, Fauconnier L, Gombault A, Couegnat M, Fick L, et al. ExtracellularATP is a danger signal activating P2X7 receptor in lung inflammation and fibrosis.American Journal of Respiratory and Critical Care Medicine 2010;182:774–83.

Shigeoka AA, Mueller JL, Kambo A, Mathison JC, King AJ, Hall WF, et al.An inflammasome-independent role for epithelial-expressed Nlrp3 in renalischemia-reperfusion injury. Journal of Immunology 2010;185:6277–85.

Solini A, Iacobini C, Ricci C, Chiozzi P, Amadio L, Pricci F, et al. Purinergic modu-lation of mesangial extracellular matrix production: role in diabetic and otherglomerular diseases. Kidney International 2005;67:875–85.

Stienstra R, van Diepen JA, Tack CJ, Zaki MH, van de Veerdonk FL, Perera D, et al.Inflammasome is a central player in the induction of obesity and insulin resis-tance. Proceedings of the National Academy of Sciences of the United States ofAmerica 2011;108:15324–9.

Subramaniam S, Strelau J, Unsicker K. Growth differentiation factor-15 preventslow potassium-induced cell death of cerebellar granule neurons by differ-ential regulation of Akt and ERK pathways. Journal of Biological Chemistry2003;278:8904–12.

Sun S, Xia S, Ji Y, Kersten S, Qi L. The ATP-P2X7 signaling axis is dispensablefor obesity-associated inflammasome activation in adipose tissue. Diabetes2012;61:1471–8.

Tang Y, Matsuoka I, Ono T, Inoue K, Kimura J. Selective up-regulation of P2X4-receptor gene expression by interferon-gamma in vascular endothelial cells.Journal of Pharmacological Sciences 2008;107:419–27.

Tao Y, Yan D, Yang Q, Zeng R, Wang Y. Low K+ promotes NF-kappaB/DNA bind-ing in neuronal apoptosis induced by K+ loss. Molecular and Cellular Biology2006;26:1038–50.

Trautmann A. Extracellular ATP in the immune system: more than just a dangersignal. Science Signaling 2009;2:pe6.

Troadec JD, Thirion S, Petturiti D, Poujeol P. Potassium efflux triggered by P2Ypurinoceptor activation in cultured pituicytes. Pflügers Archiv European Journalof Physiology 2000;440:770–7.

Trueblood KE, Mohr S, Dubyak GR. Purinergic regulation of high-glucose-inducedcaspase-1 activation in the rat retinal Muller cell line rMC-1. American Journalof Physiology – Cell Physiology 2011;301:C1213–23.

Turner CM, Vonend O, Chan C, Burnstock G, Unwin RJ. The pattern of distribu-tion of selected ATP-sensitive P2 receptor subtypes in normal rat kidney: animmunohistological study. Cells Tissues Organs 2003;175:105–17.

Vilaysane A, Chun J, Seamone ME, Wang W, Chin R, Hirota S, et al. The NLRP3 inflam-masome promotes renal inflammation and contributes to CKD. Journal of theAmerican Society of Nephrology 2010;21:1732–44.

Vincent JA, Mohr S. Inhibition of caspase-1/interleukin-1beta signaling preventsdegeneration of retinal capillaries in diabetes and galactosemia. Diabetes2007;56:224–30.

Vitiello L, Gorini S, Rosano G, la Sala A. Immunoregulation through extracellularnucleotides. Blood 2012;120:511–8.

Wang C, Pan Y, Zhang QY, Wang FM, Kong LD. Quercetin and allopurinol amelioratekidney injury in STZ-treated rats with regulation of renal NLRP3 inflammasomeactivation and lipid accumulation. PLoS ONE 2012;7:e38285.

Wang G, Kawakami K, Gick G. Regulation of Na K-ATPase alpha1 subunit gene tran-scription in response to low K(+): role of CRE/ATF- and GC box-binding proteins.Journal of Cellular Physiology 2007;213:167–76.

Weber A, Wasiliew P, Kracht M. Interleukin-1beta (IL-1beta) processing pathway.Science Signaling 2010a;19, 3:cm2.

Weber FC, Esser PR, Muller T, Ganesan J, Pellegatti P, Simon MM, et al. Lack ofthe purinergic receptor P2X(7) results in resistance to contact hypersensitivity.Journal of Experimental Medicine 2010b;207:2609–19.

Wu X, He Y, Jing Y, Li K, Zhang J. Albumin overload induces apoptosis in renal tubularepithelial cells through a CHOP-dependent pathway. OMICS 2010;14:61–73.

Zhou R, Tardivel A, Thorens B, Choi I, Tschopp J. Thioredoxin-interacting pro-tein links oxidative stress to inflammasome activation. Nature Immunology2009;11:136–40.