Primary human peritoneal mesothelial cells (HPMC) were obtained from an overnight (8 h) peritoneal lavage of seven patients admitted to the Hospital Clínico Universidad de Chile. The patients were under 50 years of age, presented neither systemic inflammatory diseases nor peritonitis and hemoperitoneum episodes, and were non-diabetic. HPMC were obtained from patients that had initiated PD within the first month prior to sampling. The peritoneal lavage bags were stored at 4°C for 1 h, and then 200 mL were extracted from the bottom of each bag. Cells were pelleted by centrifugation at 3000 rpm for 5 min and then cultured with Earle's M199 medium with 10% FBS, 2% Biogro-2, 50 UI penicillin, and 50 μg/mL streptomycin in a 5%:95% CO₂:air atmosphere at 37°C. Every 48 h, the culture medium was changed. After 10 days of culturing, a mesothelial cells monolayer was obtained for experimental use.

The cultures were rich in HPMC that contained no detectable levels of cell contaminants, such as leukocytes and fibroblasts. The initial culture confluences were 80–90%, and cultures were used up to five consecutive passages. Each culture was grown from individual effluents. Each experiment was performed using an individual culture, and the obtained results were included in respective graphs.

This study and respective protocol were approved by and carried out in accordance with recommendations from the Bioethics Commission of the Universidad de Chile, with written informed consent from all subjects. All subjects were required to provide written informed consent for study inclusion, in accordance with the Declaration of Helsinki.

HPMC monolayers were exposed to an isosmotic 300 mOsM solution (composed of 5.5 mM glucose and 1% BSA in a culture medium) and the following hyperosmotic solutions: 345 mOsM solution (composed of 45 mM glucose and 1% BSA in a culture medium), 395 mOsM solution (composed of 95 mM glucose and 1% BSA in a culture medium), and 485 mOsM solution (composed of 185 mM glucose and 1% BSA in a culture medium). Osmolarity was measured with a micro-osmometer (Advanced Instruments, Norwood, MA, USA).

Lactate dehydrogenase release: Cell death was determined by measuring lactate dehydrogenase activity released into the culture medium. After cells were incubated with the solutions described above for 24 h, lactate dehydrogenase activity in cell supernatants was determined by a colorimetric end-point kit according to the manufacturer's instructions (Roche, USA). Lactate dehydrogenase activity was measured at 590 nm. A calibration curve was produced to ensure linearity in the range studied. Various independent experiments were performed in triplicate. The background was subtracted, and data were expressed as the fraction of maximum release measured in the presence of 1% Triton X-100 (Sigma, USA).

Propidium iodide (PI)/Annexin V-FITC double labeling: Cells were harvested by centrifugation at 800 × g for 10 min, and the pellet was suspended in 100 μL PBS. Next, cells were incubated with PI (10 μg/mL, Sigma) and Annexin V-FITC (BD Pharmingen, San Diego, CA) according to the manufacturer's protocol for 20 min at room temperature in the dark. The cells were then washed and analyzed immediately by flow cytometry (FACSCanto, BD Biosciences, San Jose, CA) The PI and Annexin V excitation/emission wavelengths used were 488/>610 nm and 488/515–545 nm, respectively. A minimum of 30,000 cells/sample were analyzed. PI intensity analysis was performed using the FACSDiva software version 4.1.1 (BD Biosciences).

PI staining: Cells were harvested by centrifugation at 800 × g for 5 min, and the pellet was suspended in 200 μL PBS. Afterwards, cells were stained with PI (10 μg/mL, Sigma) for 20 min at room temperature in the dark. Then, cells were washed, and DNA content was analyzed with a flow cytometry system (FACSCanto, BD Biosciences, San Jose, CA). The PI excitation/emission wavelength used was 488/>610 nm. A minimum of 10,000 cells/sample were analyzed. PI intensity analysis was performed using the FACSDiva software version 4.1.1 (BD Biosciences; Simon and Fernández, 2009; Simon et al., 2010).

All solutions used in the flow cytometry experiments (i.e., washing and testing) were acquired from BD Biosciences and used according to the manufacturer's instructions.

HPMC were treated with the HGH solution and then loaded with different fluorescent dyes. For Ca²⁺ determinations: Either 5 μM Fluo-3 or 15 μM Fura-Red were used. Both of these dyes are Ca²⁺ specific and increase (Fluo-3) or decrease (Fura-Red) in fluorescent emission intensity upon binding Ca²⁺ (Nuñez-Villena et al., 2011). The Fluo-3 and Fura-Red excitation/emission wavelengths used were 488/533 and 488/>640 nm, respectively. For Na⁺ determinations: Dying procedures used 5 μM CoroNa Green-AM, a specific Na⁺ dye that increases in fluorescent emission intensity upon binding Na⁺ (Becerra et al., 2011). The CoroNa Green-AM excitation/emission wavelength used was 488/515–560 nm. For membrane-potential measurements cells: The cell-membrane depolarization indicator bis-(1,3-dibutylbarbituric acid) trimethine oxonol [DiBAC₄(3); (Becerra et al., 2011)] was used. A region of interest used to measure membrane potential was identified in the live cell subpopulation, in accordance with previous descriptions (Becerra et al., 2011). The DiBAC₄(3) excitation/emission wavelength used was 488/510 nm. For ROS measurements: Either 5 μM DCF (2′,7′-dichlorodihydrofluorescein diacetate [H₂DCFDA]) or 10 μM DHE (dihydroethidium) were used. These two florescent indicators exhibit increased fluorescent emission intensity upon binding ROS. The DCF and DHE excitation/emission wavelengths used were 488/520 and 488/>580 nm, respectively. All dyes were obtained from Invitrogen.

Dyes were added for 15–30 min at room temperature in the dark and then washed three times before measuring [except DiBAC₄(3); (Becerra et al., 2011)]. The labeled cells were then analyzed immediately by flow cytometry (FACSCanto, BD Biosciences, San José, CA). A minimum of 10,000 cells/sample were analyzed. Live cells were counted. Cellular dye intensity analysis was performed using the FACSDiva software v4.1.1 (BD Biosciences). All solutions used in the flow cytometry experiments (i.e., washing and testing) were acquired from BD Biosciences and used according to the manufacturer's instructions.

SiGENOME siRNA against NOX2 (siNOX2) and non-targeting siRNA (siCTRL) were used as control sequences (Dharmacon, Lafayette, CO). Transfections were performed using the DharmaFECT 4 transfection reagent (Dharmacon) according to manufacturer protocols.

Cells transfected with siNOX2 or siCTRL were lysed, and proteins were extracted. Whole cell extracts were subjected to 10% SDS-PAGE. Resolved proteins were transferred to a nitrocellulose membrane, blocked, and then incubated overnight with the anti-NOX2 antibody (Abcam). Tubulin was used as a loading control (Sigma). Protein content was determined by densitometric scanning of immunoreactive bands, and intensity values were obtained through the densitometry of individual bands as compared with tubulin and normalized against siCTRL.

The following were purchased from Calbiochem (USA): non-selective calcium channel blocker 2-aminoethoxydiphenyl borate (2-APB, 10 μM), L-type Ca²⁺ channel blocker nifedipine (1 μM), non-selective sodium channel blocker (Ambroxol, 1 μM), Na⁺ channel blocker amiloride (1 μM), cell-permeable calcium chelator BAPTA-AM (5 μM, pulse for 4 h), Catalase (10 μM), dipheniliodonium (10 μM), apocynin (1 mM), Rotenone (1 μM), Allopurinol (10 μM), 2-[1-(3-dimethylaminopropyl)-1H-indol-3-yl]-3-(1H-indol-3-yl)maleimide (BIM, 50 μM), Gö 6976 (1 μM), N-acetylcysteine (NAC, 5 mM), U73122 (10 μM), and gluthathione (GSH and GSSG, 10 mM). The following were purchased from Tocris (USA): L-NG-arginine methyl ester hydrochloride (L-NAME, 10 μM), the NO scavenger 2-Phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (PTIO, 10 μM), 2-(4-Morpholinyl)-8-phenyl-4H-1-benzopyran-4-one hydrochloride (Ly294002, 20 μM), 1,5-Dihydro-5-methyl-1-β-D-ribofuranosyl-1,4,5,6,8-pentaazaacenaphthylen-3-amine (API-2 10 μM), G–protein inhibitor (SCH 202676 hydrobromide, 10 μM), and tyrosine-kinase receptor inhibitor (AG 18, 50 μM). All reagents were added 1 h before HGH solution exposure and were maintained during the experiments. All inhibitors and blockers were tested to ensure efficiency before use.

Experiments were performed in Earle's M199 medium solution with 1% FBS, 2% Biogro-2, 50 UI penicillin, and 50 μg/mL streptomycin containing the following (in mM): NaCl 130, KCl 2.7, Na₂HPO₄ 10, KH₂PO₄ 1.8, CaCl₂ 2.5, MgCl₂ 1, HEPES 5. Experiments were performed at a pH 7.4, as adjusted with HCl/NaOH in a 5%:95% CO₂:air atmosphere at 37°C. For the Ca²⁺-free solution, CaCl₂ was not added and was replaced with MgCl₂. For the Na⁺-free solution, NaCl was replaced with NMDG-Cl.

All results are presented as the means ± SD. Statistical differences were assessed by a student's t-test (Mann–Whitney), one-way analysis of variance (ANOVA; or the non-parametric Kruskal–Wallis method) followed by Dunn's post-hoc test, or two-way ANOVA followed by Tukey's post-hoc test, as respectively indicated in the figure legends. Differences were considered significant at p < 0.05.

The HPMC had a round, short-spindle morphology with a cobblestone appearance and formed a confluent monolayer (Figure 1A). In accordance with previous descriptions for wild-type mesothelial cells, the HPMC expressed the proteins pancitokeratin and GLUT-1 (Figure 1B). Furthermore, HPMC monolayer viability was preliminarily evaluated through incorporation of the vital dye calcein (not shown).

Following this, the total HPMC death induced by HGH exposure was determined. The HPMC monolayer was exposed to either an isosmotic solution (300 mOsM), as a control, or to HGH solutions of 345, 395, and 485 mOsM for 24 h, after which lactate dehydrogenase release was measured. HPMC incubated with the 485 mOsm HGH solution exhibited significant cell death, measured as lactate dehydrogenase release, whereas incubation with the 345 and 395 mOsM HGH solutions showed no differences compared to the control isosmotic solution (Figure 1C). Additionally, experiments were performed that challenged HPMC to either isosmotic or HGH solutions while simultaneously evaluating PI and annexin V-FITC staining. Cells exposed to HGH incorporated PI and bound annexin V-FITC, indicating that the HGH solution induced extensive HPMC death (Figures 1D,E). Similar results were obtained when measuring HGH-induced cell death via the DNA content detected by PI incorporation. HPMC exposed to HGH showed strongly increased PI incorporation in the death population, along with decreased PI incorporation in the viable HPMC region (Figures 1F,G). These results suggest that at 485 mOsm, the hyperosmotic solution efficiently generates extensive HPMC cell death.

To test the relevance of glucose in the hypertonic solution, experiments using sorbitol as the hypertonic compound were performed in the absence of any glucose. The results demonstrated that the high-sorbitol hypertonic solution did not induce HPMC death, suggesting that HGH-induced HPMC death is dependent on glucose (Figure 1H).

Considering that intracellular ion levels can modulate cell death in almost all cell types, focus was placed on determining if HGH-induced mesothelial cell death was mediated by external ions. Calcium levels were independently determined with two different calcium-sensitive fluorescent dyes, one that increases (Fluo-3) and another that decreases (Fura-Red) in fluorescence upon calcium binding. Additionally, intracellular sodium changes were measured using the Na⁺-sensitive fluorescence probe CoroNa. HPMC exposed to the 485 mOsm HGH solution exhibited a significant Fluo-3 fluorescence increase and Fura-Red decrease, whereas at 345 and 395 mOsM HGH, no differences were observed as compared to the control isosmotic solution, indicating the occurrence of an intracellular Ca²⁺ increase (Figures 2A,B). Similarly, HGH-treated HPMC showed an increase in CoroNa fluorescence, indicating an increase in intracellular sodium (Figure 2C).

To test if intracellular Ca²⁺ and Na⁺ influxes were facilitated by an ion channel-mediated influx, experiments were performed using calcium and sodium channels blockers.

A non-selective Ca²⁺ channel blocker, 2-APB, effectively inhibited the fluorescence change of Fluo-3 and Fura-Red (Figures 2D,E, gray bars). Similar results were obtained using the L-type Ca²⁺ channel blocker nifedipine (Figures 2D,E, filled bars). Likewise, the sodium channel blockers, ambroxol and amiloride, abolished the CoroNa fluorescence increase (Figure 2F, gray and filled bars, respectively).

Interestingly, the reduction of either external Ca²⁺ or Na⁺ effectively prevented HGH-induced HPMC death. An initial 4 h pulse of extracellular Ca²⁺ absence effectively decreased HPMC death as induced by the HGH solution (Figure 2G). Similarly, a 4 h pulse of intracellular Ca²⁺ chelating was also able to reduce HGH-induced HPMC death (Figure 2H). Furthermore, a culture medium depleted of Na⁺ was prepared, thus keeping osmolarity and tonicity constant, with Na⁺ replaced with the non-permeant cation NMDG⁺. HPMC cultured in a Na⁺-free medium were resistant to HGH-induced HPMC death (Figure 2I). Considering that Na⁺ influx-dependent cell death is often induced by plasma membrane depolarization, assessments were also performed to determine if the HGH solution modified HPMC membrane potential. The HGH solution increased the fluorescence of the membrane potential fluorescent indicator DIBAC₄(3), suggesting that the HGH solution induced HPMC membrane depolarization (Figure 2J).

Next, tests were performed to evaluate if HGH solution-induced HPMC death was dependent on ion channel activity. For this, HPMC were exposed to either Ca²⁺ or Na⁺ channel blockers. Preincubation with 2-APB, a non-specific calcium channel blocker, prevented HPMC death as induced by HGH exposure (Figure 2K, open bars). Furthermore, HPMC incubated in the presence of nifedipine, an L-type Ca²⁺ channel blocker, and exposed to HGH solution did not exhibit significant cell death (Figure 2K, full bars). Besides this, cells incubated with a non-specific sodium channel blocker, ambroxol, effectively resisted HGH solution-induced HPMC death (Figure 2L, open bars). In addition to this, cells pre-incubated with a more specific Na⁺ channel blocker, amiloride, were also resistant to the HGH solution challenge (Figure 2L, full bars).

Taking into consideration that ROS family members are involved in several processes of cell death, the involvement of these reactive molecules in HGH-induced HPMC death was analyzed. HPMC exposed to 485 mOsM HGH showed a strong increase in oxidative stress, which was measured through the fluorescence of two ROS-sensitive fluorescent probes, DCF, which is selective to H₂O₂, and DHE, which is more selective to O2•-. In turn, HPMC treated with the 300 mOsM solution and with the 345 and 395 mOsM HGH solutions did not exhibit any changes in DCF or DHE fluorescence (Figures 3A,B, respectively). Interestingly, HGH solution-induced ROS generation was significantly decreased when cells were treated with Ca²⁺ or Na⁺ channel blockers (Figure 3).

To test the participation of ROS in HPMC death induced by HGH exposure, dose-response studies were performed. HPMC exposed to H₂O₂ alone resulted in cell death comparable to that detected using the HGH solution (Figure 3C). Interestingly, the oxidized form of the endogenous molecule glutathione (GSSG) generated HPMC death that was similar to that produced using the HGH solution (Figure 3D). Following these results, tests were carried out to assess if inhibiting the ROS burst produced by the HGH solution challenge would decrease HGH-induced HPMC death. For this, HGH-treated HPMC were pre-incubated with a catalase enzyme to reduce H₂O₂ contents, and then cell death was measured (Dröge, 2002; Varela et al., 2004). The catalase pre-treatment significantly decreased HGH solution-induced HPMC death, suggesting the participation of H₂O₂ in this process (Figure 3E). Similar results were observed using both the antioxidant agent NAC (Figure 3F) and the reduced form of the endogenous reducing molecule glutathione (GSH; Figure 3G).

Since it is well-accepted that cell death is often induced by an imbalance in the sources of intracellular oxidative stress, experiments were performed to investigate if the ROS generating enzyme NOX was involved in HGH-induced HPMC death. HPMC were pre-incubated with the non-selective NOX inhibitor dipheniliodonium (DPI), and cells were challenged with HGH. The DPI treatment abolished HGH-induced HPMC death (Figure 4A). Likewise, the NOX inhibitor, apocynin, also decreased HPMC death as induced by the HGH solution (Figure 4B), suggesting the participation of NOX in HGH-induced HPMC death.

It has been reported that NO production is stimulated by exposure to high glucose concentrations (Liao et al., 2010; Hua et al., 2012; Zhai et al., 2013). Additionally, NO is involved in necrosis and apoptosis for a number of cell types, likely due to the contributions of NO to ROS formation (Azuara et al., 2003; Borutaite and Brown, 2003; Marriott et al., 2004). Therefore, assessments were carried out to determine if NO generation participated in the HPMC death induced by HGH conditions. HPMC were pre-incubated with L-NAME, an inhibitor of NOS, and cells were then exposed to the HGH solution to determine cell death. L-NAME-treated HPMC were resistant to HGH-induced HPMC death (Figure 4C). Similarly, the use of a NO scavenger (PTIO) also evidenced protective effects to the HGH challenge (Figure 4D), indicating that HGH-induced HPMC death is dependent on the NO generated from NOS activity.

Considering that other intracellular sources of ROS, such as mitochondria and xanthine oxidase, are involved in cell death (Dröge, 2002), experiments were performed using a mitochondrial uncoupler, rotenone, and a xanthine oxidase inhibitor, allopurinol. Rotenone treatment was modestly effective in inhibiting cell death induced by the HGH solution challenge (Figure 4E). Furthermore, incubation with allopurinol had no effect on inhibiting HPMC death as induced by the HGH solution (Figure 4F).

Several isoforms of NOX have been reported, such as NOX1 to NOX5. The NOX2 isoform is primarily linked to pathological processes. Although the results obtained using DPI and apocynin suggested the potential contribution of NOX2 in HGH-induced HPMC death, this result was far from being conclusive. Therefore, to more fully test the participation of NOX2, HPMC was transfected with a specific siRNA against NOX2 expression (siNOX2). NOX2 expression was severely decreased by siNOX2 transfection, demonstrating the efficiency of this siRNA (Figures 4G,H). Importantly, non-targeting siRNA (siCTRL) and a non-related siRNA (siRNA against NOX5) induced no changes in NOX2 expression (Figures 4G,H). Of note, HGH-induced HPMC death decreased when siNOX2 was transfected, indicating that this NOX isoform is crucial for ROS-mediated HGH-induced HPMC death (Figure 4I).

NOX activity is triggered by several phosphorylations in serine residues, as mediated by PKC (el Benna et al., 1994; Park and Babior, 1997; Dang et al., 1999; Lopes et al., 1999; Babior, 2002, 2004; Simon and Stutzin, 2008). Therefore, tests were performed to determine if PKC activity participated in HGH-induced HPMC death. For this, HPMC were pre-incubated with a non-selective PKC inhibitor, BIM, and cells were then exposed to the HGH solution. BIM treatment decreased the HPMC death generated by HGH exposure (Figure 5A). Conventional PKC (cPKC), composed of PKCα, PKCβ, and PKCγ, is a PKC isoform that phosphorylates NOX (el Benna et al., 1994; Park and Babior, 1997; Lopes et al., 1999; Babior, 2002, 2004; Simon and Stutzin, 2008). Considering this, Gö 6976 was used to selectively inhibit cPKC, thus allowing for determinations of if this PKC isoform is involved in HGH-induced HPMC death. In accordance with the results observed using BIM, Gö 6976-treated cells were resistant to the HPMC death induced by HGH exposure (Figure 5B). cPKC are activated by DAG and Ca²⁺, and HGH exposure increased intracellular Ca²⁺ content (Figures 2A,B). On the other hand, DAG is generated by the actions of PLC. Therefore, the PLC inhibitor U73122 was used to assess the involvement of PLC activity. U73122 inhibited HGH-induced HPMC death (Figure 5C). Generally, PLC is activated by G-protein actions. HPMC incubated with a non-selective G-protein inhibitor were resistant to death as induced by the HGH solution (Figure 5D). These results suggest that the G-protein/PLC/cPKC pathway participates in the HPMC death induced by the HGH condition.

Previous studies have demonstrated that hyperglycemia induces PI3-K/Akt pathway activation to mediate several cellular processes (Hao et al., 2011; Zhao et al., 2012). Additionally, the PI3-K/Akt pathway promotes cell death in numerous cell types, often through mechanisms mediated by ROS overproduction (Zhang and Yang, 2013). Considering this information, assessments were carried out to determine if the PI3-K/Akt pathway participates in HGH-induced HPMC death. For this, HPMC were pre-incubated with a non-selective PI3-K inhibitor, Ly294002, and cells were then subjected to the HGH condition. The Ly294002-treated mesothelial cells showed decreased HGH-induced HPMC death (Figure 5E). Additionally, mesothelial cells were incubated with a selective inhibitor of Akt, API-2, and the effect of HGH exposure was evaluated. The results showed that Akt inhibition provoked a partial, but significant, inhibition of HPMC death induced by the HGH condition (Figure 5F). These results are in agreement with previously showed in apoptotic rat PMC (Kaifu et al., 2009). PI3-K/Akt signaling is triggered by the activation of the tyrosine-kinase receptor, and HPMC cultured with a non-selective inhibitor of the tyrosine-kinase receptor were resistant to cell death induced by the HGH solution (Figure 5G). These findings suggest that the tyrosine-kinase receptor/PI3-K/Akt pathway is involved in HGH-induced HPMC death.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.