Effect of IL-1β on Survival and Energy Metabolism of R28 and RGC-5 Retinal Neurons

Abstract

purpose. Interleukin-(IL)1β expression is increased in the retina during a variety of diseases involving the death of retinal neurons and contributes to neurodegenerative processes through an unknown mechanism. This study was conducted to examine the effects of IL-1β on the metabolism and viability of RGC-5 and R28 retinal neuronal cells.

methods. Cellular reductive capacity was evaluated using WST-1 tetrazolium salt. Mitochondrial transmembrane potential was determined by JC-1 fluorescence. Cellular ATP levels were measured with a luciferase assay. Caspase-3/7 activation was detected with a DEVDase activity assay. Cell death and lysis was evaluated by measuring release of lactate dehydrogenase (LDH). Glycolysis was assessed by measuring glucose disappearance and lactate appearance in cell culture medium. Cellular respiration was followed polarographically.

results. IL-1β treatment caused a pronounced decrease in cellular reductive potential. IL-1β caused depletion of intracellular ATP, loss of mitochondrial transmembrane potential, caspase-3/7 activation, and LDH release. IL-1β treatment increased rates of glucose utilization and lactate production. The cells were partially protected from IL-1β toxicity by ample ambient glucose. However, glucose did not block the ability of IL-1β to cause a decline in mitochondrial transmembrane potential or ATP depletion. IL-1β decreased oxygen consumption of the R28 cells by nearly half, but did not lower cytochrome c oxidase activity.

conclusions. The present results suggest that IL-1β inhibits mitochondrial energy metabolism of these retinal neuronlike cells.

Interleukin (IL)-1 proteins (IL-1α and IL-1β) are proinflammatory and proapoptotic proteins that play a key role in neurodegenerative diseases. The expression of IL-1β is rapidly increased in response to central nervous system (CNS) infection, injury or neurotoxic insult and contributes to the resulting neuronal damage (for reviews, see Refs. ¹ , ² ). 

Several studies have demonstrated that loss or inhibition of IL-1 function protects against neuronal damage. Mice lacking the IL-1 type-I receptor (IL-1RI) exhibit drastically reduced neurodegeneration after hypoxic-ischemic insult, ³ and mice with deletions of both IL-1α and -1β genes (IL-1α/β dual knockouts) suffer much less neuronal damage after CNS injury. ^{4 5} Abrogating IL-1 action with IL-1 neutralizing antibodies and blocking IL-1 processing with caspase-1 inhibitors decreases neuronal damage after CNS injury. ^{6 7} Mice lacking IL-1 receptor antagonist (IL-1Ra), a natural inhibitor of IL-1 function, exhibit dramatically increased neuronal apoptosis after transient cerebral ischemia. ⁸ The protective effect of ibuprofen after transient forebrain ischemia has been attributed to its ability to increase the expression of IL-1Ra. ⁹ IL-1Ra also protects neurons from NMDA-induced apoptosis in hippocampal slice cultures containing activated microglia, ¹⁰ and protects cortical neurons from death induced by media conditioned by activated microglia. ¹¹  

Increased expression of IL-1β has been observed in several diseases and animal models involving apoptosis of retinal neurons. These include optic nerve ligation, ¹² diabetic retinopathy, ^{13 14 15 16 17 18 19} N-methyl-d-aspartate (NMDA) excitotoxicity, ^{20 21} endotoxin-induced uveitis (EIU), ²² experimental autoimmune uveoretinitis (EAU), ²³ and ischemia-reperfusion. ^{24 25 26} Blocking IL-1β function in the retina prevents neurodegeneration in several models. Intravitreous injection of IL-1Ra provides protection of ganglion cells after optic nerve ligation. ¹² siRNA targeting IL-1β prevents retinal ganglion cell loss after intravitreous NMDA injection. ²¹ Similarly, IL-1Ra inhibits thinning of the inner plexiform layer after NMDA injection. ²⁰ Finally, intravitreous injection of IL-1Ra or IL-1β neutralizing antibodies reduces ganglion cell loss after hypertension-induced retinal ischemia-reperfusion. ²⁵  

Whereas many in vivo experimental examples have indicated that IL-1β is an important mediator of neuronal death, IL-1β alone is seldom sufficient to cause apoptosis in vitro. ¹ In fact, studies with primary neuronal cultures have demonstrated protection by IL-1β. ^{27 28} Other studies have found that IL-1β exacerbates death caused by an additional toxic insult, such as NMDA-induced cytotoxicity. ²⁹ Numerous studies have been undertaken to examine the effect of IL-1β on apoptosis of pancreatic β cells. However, in almost all these studies a combination of IL-1β was used with tumor necrosis factor (TNF)-α and γ-interferon (IFNγ), ³⁰ all of which are produced by pancreas-invading T-cells during onset of type-1 diabetes and during progression of severe type-2 diabetes. ³¹ IL-1β alone is not capable of inducing apoptosis of pancreatic β cells. ³¹ However, IL-1β contributes to the toxicity of hyperglycemia to these ³² and bovine retinal endothelial ¹⁶ cells. 

Increased production of IL-1β in retinal neurodegenerative diseases, as well as the role of IL-1 proteins in neurodegenerative disorders led us to question whether IL-1β causes death of retinal neuronal cells. To study the effects of IL-1β on cell viability and death, we used the retinal neuronlike cell lines RGC-5 and R28, which have been used in numerous studies of retinal neuronal apoptosis. We found that IL-1β treatment caused caspase-3/7 activation and slight loss of cellular integrity, especially under culture conditions that coincided with the depletion of ambient glucose supplies. More dramatic was IL-1β’s effect on mitochondrial energy metabolism. IL-1β treatment diminished mitochondrial respiration, hence increasing the dependence on aerobic glycolysis. Thus, IL-1β may not be sufficient to kill, but may promote the death of retinal neuronal cells by inhibiting mitochondrial function. This effect may be particularly detrimental when glucose levels are not sufficient and exemplifies how inflammation and ischemia may combine to cause neurotoxicity. 

Materials and Methods

Media and cell culture supplements were purchased from Invitrogen (Carlsbad, CA). Tissue culture plastic ware was from Falcon (BD Biosciences, Bedford, MA). IL-1β was obtained from R&D Systems (Minneapolis, MN). 8-(4-Chlorophenylthio) 3′,5′-cyclic adenosine monophosphate (pCMT-cAMP), mannitol, 2-deoxyglucose, metabolic substrates, sodium azide, carbonyl cyanide m-chlorophenylhydrazone (CCCP), and miscellaneous chemicals were from Sigma-Aldrich (St. Louis, MO). 

Cell Culture

The R28 cells were kindly provided by Gail Seigel (Ross Eye Institute, SUNY, Buffalo, NY). Cultures representing passages 50 to 60 were used for experiments. The R28 cell line originated from a mixed population of retinal cells (designated E1A-NR.3) immortalized with Psi2 to 12S-EIA replication-defective retroviral vector. The R28 clonal population was obtained by three rounds of cloning by limiting dilution while selecting for clones with neuronal morphologies. ³³ It has been shown that R28 cells grown with laminin and cAMP have a neuronlike phenotype. ³⁴ RGC-5 cells were kindly provided by Neeraj Agarwal (Department of Cell Biology and Genetics, University of North Texas Health Science Center, Dallas, TX). RGC-5 cells were also obtained from a mixed population of retinal cells immortalized with Psi2-12S-EIA replication-defective retroviral vector. ³⁵ RGC-5 cells are positive for several RGC markers and negative for several glial markers. ³⁵ All cells were passaged and maintained in low glucose formulation DME culture medium (Glutamax; Invitrogen) and 1 mM pyruvate, plus 10% fetal bovine serum (FBS), penicillin, and streptomycin. The final glucose concentration in the 10% FBS-supplemented medium was approximately 5.3 mM. For all experiments, the RGC-5 and R28 cells were plated onto tissue culture plastic coated with 1.0 μg/cm² laminin in the presence of 250 μM pCMT-cAMP and cultured for approximately 24 hours before treatment. Culture on laminin in the presence of cAMP analogue is a traditional method of stimulating neuronal differentiation of neuronal cell lines. ^{34 36 37}  

WST-1 Reduction Assay

Reductive capacity was measured with the WST-1 (4-[3-(4-Iodo-phenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) colorimetric assay (Quick Cell Proliferation Assay; BioVision, Mountain View, CA). The WST-1 reduction assay measures cellular reductive capacity through extracellular reduction of WST-1 tetrazolium salt by an electron-coupling reagent, resulting in a formation of a highly light-absorbent formazan product. ³⁸ WST-1 reduction provides an indirect measurement of cellular reductive capacity, correlating with NADH levels. ³⁹ Unless otherwise indicated, the cells were plated at 5 × 10⁴ cells/well (RGC-5) or 1 × 10⁵ (R28) cells/well in laminin-coated 96-well tissue culture plates (Microtest, Falcon; BD Biosciences) and in the presence of pCMT-cAMP and cultured for approximately 24 hours before treatment with IL-1β or glucose-free media. IL-1β and/or metabolic substrates were diluted to 10× final concentrations in the media before addition to cultures. After treatment, a one-tenth volume of 10× WST-1 reagent was added to the wells, and the cultures were incubated for 1 to 2 hours. WST-1 reduction was quantified by measuring the absorbance of the formazan product at 450 nm and reference absorbance at 650 nm with a plate reader (SpectraMax Plus; Molecular Devices, Sunnyvale, CA). 

Caspase-3/7 Activity Assay

Caspase-3/7 activity was measured (ApoONE Assay; Promega, Madison, WI). The cells were plated and treated as for WST-1 assay, except that black-walled 96-well tissue culture plates with clear bottoms (Optilux, Falcon; BD Biosciences) were used. After 24 hours of treatment, an equal volume of cell lysis buffer with substrate was added and caspase activity was measured, as per the manufacturer’s instructions, with a fluorescence plate reader (SpectraMax Gemini EM; Molecular Devices) with excitation at 485 nm and emission at 530 nm and a 515-nm cutoff. Relative caspase activities were then reported as fluorescence light units (FLU) as per the manufacturer’s instructions. To obtain protein-normalized caspase activities, we measured the FLUs of the DEVDase cleavage product (rhodamine-110) after 2, 3, and 4 hours of incubation. FLUs were converted to concentrations by comparison to a standard curve of FLUs for known concentrations of a rhodamine-110 reference standard (Invitrogen-Molecular Probes). Reaction rates were obtained from slopes (change in concentration per minute [Δconc/min]) of linear fits of data from each well. After the reactions, the cells were lysed by repeated trituration in 75μL/well of RIPA buffer followed by a 30-minute incubation, the plates were centrifuged, and the cellular protein concentrations in undiluted cleared lysates were measured with the BCA (bicinchoninic acid) protein assay (Pierce, Rockford, IL). Reaction rates were then normalized to protein content per well, and the means of these values (picomoles per minute per milligram protein) were reported. 

Glucose Utilization and Lactate Production Assays

Media glucose concentrations were assayed (Amplex Red glucose assay; Invitrogen) according to the manufacturer’s instructions. Media lactate concentrations were measured (Lactate Assay Kit; Biomedical Research Service Center, School of Medicine and Biomedical Sciences, SUNY, Buffalo, NY), according to the manufacturer’s instructions. 

Mitochondrial Transmembrane Electrical Potential Assay

Mitochondrial transmembrane electrical potential (ΔΨ_M) was measured using 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1; CBIC₂,(3); Molecular Probes, Invitrogen). JC-1 is a green fluorescent cationic carbocyanine dye that enters the negatively charged mitochondrial matrix and forms J-aggregates that exhibit red fluorescence. The ratio of red to green fluorescence is indicative of ΔΨ_M. ^{40 41} the cells were plated in black-walled, 96-well tissue culture plates with clear bottoms, as described for the caspase-3/7 assays. After 24 hours of treatment, 1.0 μg/mL final concentration of JC-1 was added, and the plates were incubated for 20 minutes at 37°C, followed by a gentle rinsing with phosphate-buffered saline (PBS). Green fluorescence (485 nm excitation, 530 nm emission, autocutoff) and red fluorescence (535 nm excitation, 590 nm emission, autocutoff) was read with a fluorescence plate reader (SpectraMax Gemini EM; Molecular Devices), and the ratio of red to green fluorescence was calculated for each well. The validity of the assay was tested by using the mitochondrial uncoupler CCCP (10 μM for 20 minutes) that reduced the red-to-green fluorescence ratio by 91% and 85% in the R28 and RGC-5 cells, respectively. 

Intracellular ATP Assay

To measure intracellular ATP contents, cells were plated in 96-well plates and treated as for the WST-1 assay. ATP was measured with a kit (ATP Bioluminescence Assay; Roche Applied Science, Indianapolis, IN). After treatment, the media were aspirated, and the cells were lysed by addition of 100 μL/well of cell lysis reagent, 10 minutes of shaking at room temperature, and repeated trituration. The samples were centrifuged at 1500g for 5 minutes at 4°C before analysis of ATP in supernatants, essentially per the kit manufacturer’s instructions, with a tube luminometer (Sirius; Berthold Detection Systems, Oak Ridge, TN). Sample protein contents were measured using the BCA protein assay (Pierce). 

Cell Lysis and Cellular Monolayer Protein Content Assays

To determine the extent of cell death and lysis, cells were placed in 96-well plates, treated as for the WST-1 assay, and the LDH activity in the media was measured (CytoTox96 Nonradioactive Cytotoxicity Assay; Promega), as per the manufacturer’s instructions with 1:5 dilution of media samples. To measure protein content in attached cellular monolayers, we plated the cells in laminin-coated 48-well plates at 1.5 × 10⁵ cells/well (RGC-5) or 3 × 10⁵ (R28) cells/well and in the presence of pCMT-cAMP and cultured them for approximately 24 hours before treatment with IL-1β. The cells were cultured under the same conditions, including density and cell-to-medium ratio, as those used for the assays performed in 96-well plates. Attached cells were lysed by repeated trituration in 100 μL/well of RIPA buffer followed by a 30-minute incubation, and the cellular protein concentrations in undiluted cleared lysates were measured with the BCA protein assay. 

Assays of Total Respiratory and Cytochrome c Oxidase Activities

Respiratory activity was measured by polarographic monitoring of oxygen consumption using established methods. ^{42 43} The cells were plated at 4 × 10⁶ cells per 60-mm dish, with laminin coating and pCMT-cAMP and incubated for 24 hours before treatment. After treatment, the cells were harvested with trypsin-EDTA and suspended at 2 × 10⁶ cells/mL in fresh, complete culture medium containing 15 mM HEPES buffer. To determine total oxygen consumption rate in intact cells, cell suspensions were transferred to a thermostatically controlled chamber equipped with a magnetic stirring device, which was then sealed with a gas-tight stopper. The rate of oxygen consumption at 37°C was measured polarographically with a Clark-type oxygen electrode. The electrode was calibrated by using air-saturated water. Assuming an oxygen concentration of 434 nanoatoms/mL at saturation, the initial dissolved oxygen content in the 1.3 mL chamber was taken as 564 nanoatoms. The rate of oxygen consumption was determined twice for each cell suspension, and these rates were averaged. Mean and standard deviations were calculated from average rates obtained from three replicate cultures. Measurement of cytochrome c oxidase (respiratory complex IV) activity was accomplished in a similar fashion, except that cells were suspended in a respiratory medium containing 250 mM sucrose, 0.1% bovine serum albumin, 10 mM MgCl₂, 20 mM HEPES buffer, 5 mM KH₂PO₄ (pH 7.2), and 1 mM ADP. After measuring an initial O₂ consumption rate, 0.005% digitonin was measured, followed by 0.4 mM N,N,N′,N′-tetramethyl-p-phenylenediamine dihydrochloride (TMPD) and 4 mM ascorbic acid (vitamin C). 

Results

The effects of IL-1β on the reductive metabolism and viability of the RGC-5 and R28 retinal neuronal cells were tested by evaluating reductive potential, caspase activation, and cell lysis after 24-hour treatment with various concentrations (0, 0.01, 0.1, 1.0, and 10 ng/mL) of this cytokine (Fig. 1) . IL-1β concentrations above 0.01 ng/mL diminished WST-1 reduction of both cell types (Figs. 1A 1B) . WST-1 reduction was decreased by 50% to 59% in the RCG-5 cells (P < 0.001). The effect was particularly pronounced in the R28 cells, where IL-1β lowered reductive capacity by >90% (P < 0.001). This result did not coincide with a marked depletion of the number of attached cells, for although floating cells and debris appeared, both the RGC-5 and the R28 cultures maintained cell confluence throughout the experiment. To document the fact that IL-1β treatments did not coincide with the loss of confluence of the cell monolayer, we measured cellular protein in attached cells after treatment with IL-1β concentrations for 24 hours (Figs. 1C 1D) . IL-1β caused only slight decreases in protein levels, reaching 12% in RGC-5 cultures and 15% in R28 cultures. IL-1β treatment, however, caused caspase activation in these cells (Figs. 1E 1F) . Relative caspase-3/7 (DEVDase) activity was increased by 2.8-fold in RGC-5 cells treated with 0.1 ng/mL (P < 0.001) and 10-fold in the R28 cells treated with 1.0 ng/mL (P < 0.001). IL-1β treatment also caused slight increases in the amount of cell lysis occurring, as indicated by the release of LDH enzyme into the media (Figs. 1G 1H) . LDH release increased by 4.9-fold in RGC-5 cultures treated with 0.1 ng/mL (P < 0.001) and by 2.8-fold in R28 cultures treated with 1.0 ng/mL (P < 0.001) compared with untreated cells. Thus, physiologically relevant concentrations of IL-1β caused a marked decrease in the reductive potential of these cells that coincided with executioner caspase activation and some cell death; however, after 24 hours of treatment, cell death was not sufficient to cause an appreciable decrease in cell density. 

It was noted that the degree of inhibition of WST-1 reduction by IL-1β was dependent on cell plating density. This effect was illustrated by Figure 2A , which shows the effects of a 24-hour treatment with 10 ng/mL IL-1β on the WST-1 reduction by RGC-5 cells plated at an eightfold range of cell seeding densities. At the lowest plating density of 1.25 × 10⁴ cells/well of a 96-well plate, there was no effect of IL-1β. At double that density, 2.5 × 10⁴ cells/well, at which cells had nearly reached confluence by completion of the experiment, IL-1β caused a slight (19%) but significant (P < 0.05) decrease in WST-1 reduction. Of interest, the extent of WST-1 reduction by cells at the higher density was less than half that in the low-density cells, which is also not in keeping with this assay’s being representative of the number of viable cells present in these cultures. However, as cell plating densities further increased, WST-1 reduction in control cells rose accordingly. The effect of IL-1β was at its maximum at a density of 5.0 × 10⁴ cells/well, where it caused a 51% decrease in WST-1 reduction (P < 0.001). At higher densities the effects of IL-1β declined, with 42% and 27% decreases at 7.5 × 10⁴ and 1.0 × 10⁵ cells/well, respectively. Cell seeding density had similar effects on caspase activation (Fig. 2B) . IL-1β treatment caused a 2.7-fold increase (P < 0.001) in caspase-3/7 activity in cultures plated at 2.5 × 10⁴ cells/well, whereas caspase activity increased 5.4- (P < 0.01) and 6.3-fold (P < 0.001) in cultures plated at 3.75 × 10⁴ and 5.0 × 10⁴ cells/well, respectively. At the highest density of 1.0 × 10⁵ cells/well, caspase activation was only 3.2-fold (P < 0.001). Thus, the effects of IL-1β treatment on cellular reductive capacity and caspase activity were maximum in cultures seeded at the same relatively high seeding density. High culture densities alone were only slightly detrimental to these cells. In cultures not treated with IL-1β, protein-normalized caspase-3/7 activities increased 57% (P < 0.05) and 137% (P < 0.01) at the two highest seeding densities. 

The dependence on high cell density suggests that the effects of IL-1β either necessitate cell contact, are due to accumulation of a diffusible factor, or necessitate media depletion. Reduction of WST-1 and other tetrazolium salts is dependent on cellular production of reducing equivalents, primarily in the form of NADH, which may be compromised by glucose starvation. For example, reduction of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) declines as cells deplete media glucose and is dramatically decreased when cells are transferred to glucose-free media. ⁴⁴ Because RGC-5 cells were maintained and plated in the low-glucose DMEM formulation (5.3 mM glucose after serum supplementation), depletion of media glucose was hypothesized. Therefore, an experiment replicating that shown in Figure 2Awas performed, and the final concentrations of glucose in the cultures were measured (Fig. 2C) . Ambient glucose was nearly totally depleted in cultures plated at 2.5 × 10⁴ cells/well and higher. In cultures plated at 1.25 × 10⁴ cells/well, ambient glucose levels averaged 1.3 mM in control cultures and below 0.8 mM in IL-1β-treated cultures (P < 0.001). Thus, confluent RGC-5 cultures exhibited relatively rapid glucose utilization that seemed to be increased by IL-1β, and the effects of IL-1β on WST-1 reduction were observed most often in conditions that coincided with extensive glucose depletion of the media. 

IL-1β had a pronounced effect on WST-1 reduction by RGC-5 cells in glucose-depleted media. The effect of IL-1β on the ability of RGC-5 cells to adapt to glucose-deprivation was directly tested by abruptly switching the cells to nearly glucose-free medium (glucose-free DMEM containing approximately 0.3 mM glucose contributed by 10% serum) at the time of treatment with 10 ng/mL IL-1β and measuring WST-1 reduction 3, 6, 12, 24, 36, and 48 hours later (Fig. 2D) . WST-1 reduction rapidly declined in both control and IL-1β-treated cultures abruptly deprived of glucose, reaching essentially identical nadirs at 12 hours. At later times, WST-1 reduction values increased, but diverged significantly in control and IL-1β-treated cultures (P < 0.001). WST-1 reduction increased by 4.8-fold in control cultures, relative to the nadir. In IL-1β-treated cultures, WST-1 reduction increased by only 2.1-fold, to a mean value of only 48% of that in control cultures. These results suggested that IL-1β inhibited the ability of RGC-5 oxidative metabolism to adapt to glucose-free culture conditions. 

Analogous experiments were performed with R28 cells and qualitatively identical results were obtained (data not shown). Again, the effects of IL-1β were dependent on cell plating density, with the maximum effect in cultures plated at 0.75 to 1.0 × 10⁵ cells/well of a 96-well plate. This density coincided with media glucose depletion, and the R28 cells treated with IL-1β had lower final glucose concentrations than those that were not. However, because the R28 cells exhibited massive cell death within 24 hours of an abrupt change to glucose-free media conditions, the ability of the R28 cell reductive metabolism to adapt to acute glucose deprivation was not tested. 

Lower ultimate glucose concentration in cultures incubated with IL-1β suggested that this cytokine increased glucose utilization, perhaps by increasing glycolysis. Thus, the effects of IL-1β treatment on aerobic glycolysis were examined by measuring effects on the rates of glucose disappearance and lactate appearance in the RGC-5 and R28 cultures. Replicate confluent cultures were pretreated with or without 10 ng/mL IL-1β for 4 hours before feeding with fresh media containing approximately 5.3 mM glucose and with or without IL-1β. Media samples were taken 2, 4, 6, 8, 10, and 24 hours after feeding, and glucose and lactate concentrations were determined (Fig. 3) . Empirically, changes in glucose and lactate concentrations were linear between the 2- and 10-hour time points. Thus, glucose disappearance and lactate appearance rates for each culture (n = 6) were estimated by linear regression analysis of concentration data during this period, and the mean rates and significance of differences calculated from these. Control RGC-5 cells consumed glucose at a rate of 0.27 ± 0.006 mM/h. In the presence of IL-1β, this rate was significantly increased by 47% to 0.40 ± 0.01 mM/h (P < 0.001). On a per cell basis, glucose utilization rates were approximately 0.91 picomoles/cell/h and 1.3 picomoles/cell/h, respectively. By 24 hours, both the control and treated RGC-5 cultures had depleted media glucose to levels below 0.01 mM. Corresponding rates of lactate appearance in these cultures were 0.56 ± 0.05 and 0.86 ± 0.09 mM/h, respectively. Thus, IL-1β caused a similar 54% increase in lactate production by the RGC-5 cells (P < 0.01). The ratio of lactate appearance to glucose disappearance was slightly greater than 2.0 in the RGC-5 cultures under both conditions. IL-1β had a greater effect on the rate of glycolysis exhibited by the R28 cells. Whereas control R28 cultures exhibited a glucose consumption rate of 0.076 ± 0.015 mM/h, in the presence of IL-1β, the rate was increased by 2.7-fold, to 0.21 ± 0.01 mM/h (P < 0.001). These rates corresponded to cellular rates of approximately 0.13 picomoles/h/cell and 0.34 picomoles/h/cell, respectively. Thus, the R28 cells consumed only a fraction of the glucose used by the RGC-5 cells. The ultimate mean glucose concentrations in the R28 cultures at 24 hours were 3.4 and 0.5 mM in untreated and treated cultures, respectively. Lactate production by the R28 cells was increased 5.9-fold by IL-1β treatment, from 0.072 ± 0.018 to 0.43 ± 0.02 mM/h (P < 0.001). Thus, the ratio of lactate production to glucose utilization was essentially 1.0 in control cells and was increased to 2.0 by the addition of IL-1β. 

Because the effects of IL-1β coincided with media glucose depletion, the effects of providing additional glucose and other potential metabolic substrates to the RGC-5 and R28 cultures were examined. To accomplish this, 10 mM concentrations of several potential energy substrates, including glucose, glutamine, lactate, and pyruvate, were added simultaneously with 10 ng/mL IL-1β and WST-1 reduction was assayed after a 24-hour incubation period (Figs. 4A 4B) . The effects 10 mM mannitol were also examined to control for effects of increased osmolarity. IL-1β inhibited WST-1 reduction in the control cultures of the RGC-5 and R28 cells by 47% (P < 0.001) and 70% (P < 0.001), respectively. Nearly identical percentage reductions were observed in the presence of mannitol, glutamine, lactate, and pyruvate. Only 10 mM glucose negated, or reversed, the effects of IL-1β on WST-1 reduction. For the RGC-5 cells, WST-1 reduction in the 10 mM glucose-fed cultures was 84% higher than in the control cells. These glucose-fed RGC-5 cultures exhibited only a 9% (insignificant) decrease in WST-1 reduction in the presence of IL-1β. Glutamine and lactate caused slight, but significant (P < 0.05) increases in WST-1 reduction, but these substrates did not block the effects of IL-1β. Glucose alone did not increase WST-1 reduction in the R28 cultures. However, IL-1β actually increased WST-1 reduction by 65% (P < 0.01) when glucose was added to the R28 cell cultures. Thus, provision of ample glucose, but not other substrates, had a profound effect on inhibition of WST-1 reduction by IL-1β. 

The effects of glucose were compared to two other glycolytic substrates—fructose and mannose—as well as an inhibitor of glycolysis, 2-deoxyglucose. A final concentration of 20 mM of these sugars was added at the time of treatment with 10 ng/mL IL-1β, and WST-1 reduction was assayed after 24 hours of incubation (Figs. 4C 4D) . Glucose at 20 mM exhibited effects nearly identical with those observed in the previous experiment, and mannose exhibited effects nearly indistinguishable from glucose. In the RGC-5 cells, WST-1 reduction was increased by glucose and mannose, by 75% and 55%, respectively. IL-1β had essentially no effect on WST-1 reduction in the presence of these sugars. The results in the presence of fructose were intermediate. WST-1 reduction was increased 55% in the absence of IL-1β, but IL-1β treatment decreased WST-1 reduction. However, the decrease was only 28%, compared to 40% in the control cultures with no added sugar. Addition of 2-deoxyglucose inhibited WST-1 reduction by 54%, with no significant additional effect of IL-1β. In the R28 cells, glucose or fructose alone did not increase WST-1 reduction. However, IL-1β increased WST-1 reduction in the presence of glucose and mannose by 88% (P < 0.01) and 77% (P < 0.001), respectively. Addition of fructose to the R28 cultures had essentially no effect on WST-1 reduction. Like RGC-5 cells, 2-deoxyglucose inhibited WST-1 reduction by 55% in the R28 cells, with no further decrease caused by IL-1β. These results suggest that glycolytic substrates, especially glucose and mannose, are able to negate or reverse the effects of IL-1β on WST-1 reduction and that blocking glycolysis with 2-deoxyglucose greatly inhibits WST-1 reduction. 

It was reasoned that IL-1β could cause mitochondrial dysfunction, leading to energy depletion and eventual cell death, but only when sufficient amounts of glycolytic substrates were unavailable. To test this theory, the effects of providing 20 mM glucose on several metabolic and apoptotic parameters were measured in the RGC-5 and R28 cells treated with 10 ng/mL IL-1β (Fig. 5) . JC-1 was used to evaluate the effect of IL-1β on the mitochondrial transmembrane electrical potential (ΔΨ_M). Loss of ΔΨ_M can be triggered by opening of the mitochondrial permeability transition pore (MPTP), by action of a mitochondrial uncoupling agent, or by inhibition of the electron transport chain. IL-1β treatment also decreased the ratio of red to green fluorescence, corresponding to a reduction in JC-1 aggregation due lowering of ΔΨ_M (Figs. 5A 5B) . In the control RGC-5 cells, without extra glucose added, this ratio decreased 32%, from 19.4 to 13.0 (P < 0.001). With ample glucose added IL-1β had a slightly greater effect, decreasing the ratio by 44%, from 16.7 to 9.4 (P < 0.001). For comparison, complete loss of ΔΨ_M by decoupling of RGC-5 cells with 10 μM CCCP for 20 minutes caused an 85% decrease in the JC-1 red-to-green fluorescence ratio, and blocking the electron transport chain by treatment with 1 mM sodium azide for 24 hours caused a 92% decrease in the JC-1 fluorescence ratio (data not shown). In the R28 cells, the effects on ΔΨ_M were similar. IL-1β treatment caused the JC-1 red-to-green fluorescence ratio to decrease by 30%, from 25.1 to 17.5 (P < 0.001) in control cells. In glucose-fed cells, the decrease was 40%, from 25.9 to 15.5 (P < 0.001). For comparison, in the R28 cells mitochondrial decoupling with 10 μM CCCP for 20 minutes caused a 91% decrease in the JC-1 red-to-green fluorescence ratio, and treatment with 1 mM sodium azide for 24 hours caused an 85% decrease in the ratio (data not shown). Thus, IL-1β treatment caused similar fractional loss of ΔΨ_M in both cell lines, and this effect was accentuated in the presence of ample ambient glucose. 

The effect of IL-1β on caspase-3/7 activity in control cultures and cultures fed with 20 mM glucose was compared (Figs. 5C 5D) . In the control RGC-5 cultures, IL-1β caused a 5.1-fold increase in caspase activity (P < 0.001). With glucose feeding, IL-1β had no significant effect on caspase activity. In the R28 cultures, IL-1β caused a 9.2-fold increase in caspase activity (P < 0.001) when no additional glucose was added. With ample glucose present, IL-1β caused a 5.1-fold rise in caspase activity. Thus, prevention of glucose deprivation blocked caspase activation in the IL-1β-treated RGC-5 cells and greatly diminished caspase activation in the IL-1β-treated R28 cells. 

To discern the effects of IL-1β treatment and glucose depletion on energy levels, cellular ATP contents were measured (Figs. 5E 5F) . In the RGC-5 cells lacking additional glucose, IL-1β treatment caused a 36% decrease in ATP level, from 17.6 to 11.2 ng/mg protein (P < 0.001). Supplementation with 20 mM glucose at the time of IL-1β addition increased the ATP levels relative to that of the control cultures, with ATP contents of 25.1 ng/mg protein without IL-1β and 16.2 ng/mg protein with IL-1β treatment. Thus, IL-1β caused a similar 35% decrease in ATP content (P < 0.001) when ample glucose was present, but the IL-1β-treated cells maintained levels that were comparable to those of the control cultures without additional glucose or IL-1β treatment. The R28 cells exhibited a qualitatively similar phenomenon, but with a much greater increase in ATP levels caused by glucose supplementation. In the R28 cells lacking additional glucose, IL-1β caused a 46% depletion of ATP (P < 0.001) from 17.4 to 9.42 ng/mg protein. In the presence of added glucose, ATP levels were 38.2 and 25.7 ng/mg protein, without and with IL-1β treatment, respectively. These levels corresponded to a 33% decrease in ATP (P < 0.001) caused by IL-1β. However, even with IL-1β treatment, the ATP levels were higher than control cells without additional glucose. Thus, IL-1β treatment caused a significant depletion of cellular energy stores represented by ATP. Not surprisingly, energy depletion is most severe when cells are treated with IL-1β in the absence of ample ambient glucose. 

The effect of glucose on the release of LDH enzyme into the media was evaluated (Figs. 5G 5H) . LDH release was increased 65% by IL-1β treatment of the control RGC-5 cells (P < 0.01). When these cells were fed with 20 mM glucose at the time of IL-1β treatment, IL-1β had no significant effect on LDH release. IL-1β treatment of the R28 cells caused a 3.0-fold increase in LDH release (P < 0.001). With glucose feeding, IL-1β caused a 2.1-fold increase in LDH release (P < 0.001). Comparing LDH levels in the glucose-fed and control R28 cells demonstrated that glucose significantly diminished LDH release in response to IL-1β by 50% (P < 0.001). Thus, glucose feeding prevented caspase activation and cell lysis in RGC-5 cells and partially inhibited these effects in the R28 cells. 

The increase in aerobic glycolysis, decline of ΔΨ_M, and ATP depletion are each consistent with diminished oxidative phosphorylation, specifically mitochondrial electron transport and the reduction of oxygen. To determine whether IL-1β affected mitochondrial respiration, the effect of IL-1β treatment on oxygen consumption of the R28 cells was measured. The R28 cells were examined because they exhibited a much greater rate of oxygen consumption than did the RGC-5 cells (data not shown). The R28 cultures were fed with fresh medium containing 5.3 mM glucose and treated with 10 ng/mL IL-1β for 24 hours or with 0.5 mM sodium azide for 18 hours, before polarographic measurement of cellular oxygen consumption (Fig. 6) . IL-1β treatment caused a significant 54% decrease in oxygen consumption of intact cells (Fig. 6A) , from 16.5 nanoatoms/min × 10⁶ cells in untreated cells to 7.7 nanoatoms/min ×10⁶ cells (P < 0.05). For comparison, treatment for 18 hours with 0.5 mM sodium azide, which inhibits cytochrome oxidase (complex IV of the electron transport chain), caused a similar significant 56% decrease in oxygen consumption to 7.4 nanoatoms/min × 10⁶ cells (P < 0.05). However, IL-1β treatment had no effect when complex IV activity was evaluated via measuring oxygen consumption by permeabilized cells provided with ADP and electron donors (TMPD/ascorbate) that supply electrons directly to cytochrome c (Fig. 6B) . Complex IV-mediated oxygen consumption was 34 nanoatoms/min × 10⁶ cells in control, and nearly identical in IL-1β-treated cells. As a positive control, treatment for 18 hours with 0.5 mM sodium azide significantly lowered complex IV activity by 30% (P < 0.05). Thus, the effect of IL-1β on respiratory activity was not due to diminished cytochrome c oxidase activity. 

Discussion

The present study found that IL-1β alone was able to cause the metabolic dysfunction of the RGC-5 and R28 cells, particularly when glycolytic substrates were lacking. The effects of IL-1β were much more pronounced in the R28 cells. The results suggest that IL-1β may contribute to neuronal apoptosis by causing energy depletion. It has been well established that neuronal cells, including retinal neurons, are relatively intolerant to hypoglycemia. ^{45 46} Glucose deprivation and inhibition of glycolysis exacerbate the damage caused by neurotoxic agents, including glutamate. ^{47 48 49 50} Impairment of brain energy metabolism increases neuronal damage caused by cerebral ischemia, injury, and neurodegenerative processes. ^{51 52 53 54} Energy depletion and IL-1β have been linked in other models of neuronal apoptosis, particularly models of ischemia employing glucose deprivation. ⁵⁵ Thus, in retinal diseases characterized by decreased perfusion, localized expression of IL-1β could contribute to neuronal death by compounding the effects of ischemia on neuronal energy depletion. 

IL-1β inhibited the reduction of WST-1 tetrazolium salt to a formazan product by both the RGC-5 and R28 cells. The extent of this effect was clearly much greater than the extent of cell death observed and provides a cautionary example for the use of tetrazolium salts. Reduction of WST-1 and other tetrazolium salts are commonly used to measure the relative number of cells for proliferation and toxicity assays. However, this use assumes that the amount of formazan product produced by each cell is not affected by culture conditions or treatments. As it has become apparent that reduction of tetrazolium salts depends on the reductive capacity of the cells, they have been referred to as indicators of mitochondrial dehydrogenase activity and of “mitochondrial viability.” ⁵⁶ the effect of IL-1β on WST-1 reduction is consistent with a loss of mitochondrial NADH production or diminished shuttling of NADH between the cytoplasm and mitochondria. ³⁸ Glycolysis can be upregulated in response to hypoxia and agents that inhibit mitochondrial oxidative phosphorylation. The phenomenon of increased glycolysis in response to inhibition of respiration is often referred to as the Pasteur effect, ⁵⁷ and its extent is even taken as a measure of mitochondrial oxidation. ⁵⁸ In cultures with depleted ambient glucose, the cells would be forced to adapt energy metabolism to maintain viability in the absence of glycolysis. The effect of IL-1β on WST-1 reduction suggests that this adaptation is inhibited by this cytokine. Such a process was illustrated when the RGC-5 cells were suddenly subjected to nearly glucose-free medium. The cells exhibited a sharp decline in WST-1 reduction capacity, which then recovered more fully if IL-1β was not present. We speculate that the recovery of WST-1 reduction occurred as the cells adapted to energy production in the absence of glycolysis, using an alternative source of acetyl-CoA generation, and that this process was impaired by IL-1β. 

The effects of IL-1β on WST-1 reduction suggests that NADH production by the TCA cycle is impaired. When ample amounts of glycolytic substrates are available, upregulation of glycolysis would provide an alternative means of producing NADH for the reduction of WST-1, which would explain why IL-1β has no effect on WST-1 reduction by RGC-5 cells provided with ample glucose or mannose. Because IL-1β greatly upregulated glycolysis by the R28 cells, this may also explain why IL-1β treatment actually increased R28 WST-1 reduction in the presence of ample ambient glucose or mannose. Provision of nonglycolytic sources of acetyl-CoA, such as pyruvate and lactate, did not affect WST-1 reduction. In addition, glutamine, which can supply the TCA cycle through conversion to glutamate and then α-ketoglutarate, had little effect on WST-1 reduction, further suggesting that cells treated with IL-1β are highly dependent on glycolysis. If IL-1β or other inflammatory mediators inhibit neuronal mitochondrial energy metabolism, then provision of alternative sources of acetyl groups, such as pyruvate or ketone bodies, would not represent a viable means of preventing neuronal energy depletion in situations involving microglial activation and/or inflammation. 

RGC-5 cells exhibited a relatively high rate of aerobic glycolysis, even in the absence of IL-1β. The rate of lactate production by RGC-5 cells was, within experimental deviation, twice that of glucose utilization, both in the presence and absence of IL-1β. However, IL-1β increased the rates of both glucose utilization and lactate production by approximately 50%. In the presence of IL-1β, lactate production increased so that these cells produced final lactate concentrations that approached those expected if all media glucose and pyruvate had been stoichiometrically converted to lactate (data not shown). In contrast, R28 cells could not be considered highly glycolytic. Under control conditions, the cells exhibited a specific rate of lactate production that was approximately 7% that of the RGC-5 cells. The R28 cells only produced one lactate molecule for each glucose molecule consumed, suggesting that considerable glucose oxidation occurred. In response to IL-1β treatment, the R28 cells increased their rates of glucose utilization by nearly threefold and lactate production by sixfold, bringing them to a state of stoichiometric 1:2 conversion rate of glucose to lactate. 

In a previous study, lactate production by retinal neuronal cell lines, including RGC-5 cells, was examined. Winkler et al. ⁵⁹ measured lactate production by several retinal cell lines, including low-passage human retinal pigmented epithelial cells (hRPE), a rat Müller cell line (rMC-1), a mouse photoreceptor cell line (661W), and RGC-5 cells. All exhibited a relatively high rate of aerobic glycolysis that increased when the mitochondrial electron transport chain was blocked with antimycin A (chemical hypoxia). The rate of lactate production by 661W cells was 0.9 picomoles/cell/h, which was more than doubled by antimycin A. The rate of lactate production by RGC-5 cells reported by Winkler et al. was ∼0.56 picomoles/cell/h. They also found that antimycin A treatment nearly tripled lactate production by RGC-5 cells, to ∼1.5 picomoles/cell/h. In contrast, we found our RGC-5 cultures to be normally highly glycolytic, leaving little leeway for the Pasteur effect. 

The increased basal rate of aerobic glycolysis observed in our RGC-5 cultures could represent dedifferentiation away from a neuronal cell phenotype. It has been suggested that neurons obtain most of their energy via mitochondrial oxidative phosphorylation, not glycolysis, whereas astrocytes perform aerobic glycolysis and provide lactate for the consumption of neurons. ⁶⁰ This symbiosis has been termed the “astrocyte-neuron lactate shuttle.” ⁶¹ In support of this hypothesis, neurons in culture preferred lactate over glucose, ⁶² and differentiation of neural stem cells into astrocytic cells was accompanied by increased aerobic glycolysis. ⁶³ Thus, the relatively high amounts of aerobic glycolysis observed in RGC-5 cells may represent an astrocytic, rather than neuronal, characteristic. 

A similar increase in aerobic glycolysis has been observed in astrocytes treated with IL-1α, ^{64 65} as well as ovarian cells and cardiac myocytes treated with IL-1β. ^{66 67} Similarly, a mixture of IL-1β, TNFα, and IFNγ increased the rate of glycolysis in enterocytes. ⁶⁸ Bolanos et al. ⁶⁹ demonstrated that inflammatory stimuli inhibited mitochondrial function in both neurons and astrocytes largely through inhibition of cytochrome c oxidase (complex IV in the electron transport chain) by nitric oxide (NO·). ⁶⁹ the decrease in oxygen consumption by the R28 cells caused by IL-1β suggests that the electron transport chain is compromised. Lack of sufficient respiration and mitochondrial ATP production and subsequent energy depletion would trigger the acceleration of glycolysis, increasing glucose utilization and lactate production. However, IL-1β treatment did not decrease cytochrome c oxidase activity in the R28 cells. The mechanism by which IL-1β inhibits oxidative phosphorylation by the R28 cells is now under investigation. 

Whereas inhibition of mitochondrial function greatly increased glycolysis in astrocytes ⁷⁰ and retinal Müller cells, ⁷¹ this response was relatively minor in neurons. ^{72 73 74} Consequently, the astrocytes were able to maintain ATP levels when mitochondrial function was impaired, whereas the neurons were not. ⁷⁵ Thus, the ability of the RGC-5 and R28 cells to increase their rates of aerobic glycolysis may represent another astrocytic characteristic. However, it should be noted that the theorized inability of neuronal cells to use glycolysis for maintenance of ATP levels is not a certainty. Thorn et al. ⁷⁶ recently provided evidence that lactate accumulation during focal cerebral ischemia is probably due to impaired pyruvate oxidation and increased lactate production by neurons, rather than an imbalance in astrocyte lactate production and neuron lactate utilization. In addition, Winkler et al. ⁷⁷ observed that retinal neurons and photoreceptors rely on glycolysis in the absence of mitochondrial respiration. They also found that both 661W photoreceptors and RGC-5 cells preferentially use glucose even in the presence of 10 mM lactate. ⁵⁹ A very recent hypothesis suggests that the ultimate product of neuronal glycolysis is lactate, which is transported into mitochondria and converted to pyruvate via a mitochondrial form of LDH. ^{78 79}  

In the presence of ample glucose, IL-1β treatment decreased oxygen consumption, decreased mitochondrial transmembrane potential, decreased cellular ATP levels, and increased aerobic glycolysis. All these effects are consistent with impairment of the electron transport chain and/or the TCA cycle and induction of the Pasteur effect. When ample glycolytic substrates are present IL-1β-treated cells can maintain reductive metabolism. In the absence of ample glycolytic substrates, as is the case in ischemic tissue, severe energy and NADH depletion could result. Thus, inhibition of mitochondrial function by IL-1β may have important implications for neuronal cells in retinal regions experiencing both inflammatory and ischemic insults and may help to explain how IL-1β contributes to neuronal cell death during ischemia. 

 Supported by a grant from the Juvenile Diabetes Research Foundation Diabetic Retinopathy Center (PI, TWG).

 Submitted for publication August 9, 2007; revised January 21 and June 7, 2008; accepted October 21, 2008.

 Disclosure: S.F. Abcouwer, None; S. Shanmugam, None; P.F. Gomez, None; S. Shushanov, None; A.J. Barber, None; K.F. Lanoue, None; P.G. Quinn, None; M. Kester, None; T.W. Gardner, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Effect of IL-1β on WST-1 reductive potential, caspase activity, cell lysis, and attached cellular protein content in RGC-5 and R28 cultures. RGC-5 (A, C, E, G) and R28 (B, D, F, H) cultures were treated with 0, 0.01, 0.1, 1.0 or 10 ng/mL IL-1β. After 24 hours of treatment, reduction of WST-1 tetrazolium salt (A, B, n = 6), cellular protein in attached monolayers (C, D, n = 3), relative caspase-3/7 activity (E, F, n = 5), and cell lysis, as indicated by release into the media of LDH (G, H, n = 6), were measured. WST-1 reductive potentials are indicated by formazan product optical densities (OD, 450 nm with reference absorbance at 650 nm). Relative caspase-3/7 activities are indicated by Z-DEVD-rhodamine-110 product FLUs (485 nm excitation and 530 nm emission). LDH release values are indicated by optical densities (OD, 490 nm) of the formazan salt reaction end product. Protein contents were measured with the BCA assay after cell lysis in RIPA buffer. Comparisons between control and treated groups were calculated by using Student’s t-test: *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 2.

 

Effect of plating density and media glucose on mitochondrial reductive potential and caspase activity of RGC-5 cells. (A) RGC-5 cells plated at the indicated seeding densities (n = 5) were treated with 10 ng/mL IL-1β for 24 hours before WST-1 reduction was assessed. (B) RGC-5 cells plated at the seeding densities indicated (n = 4) were treated with 10 ng/mL IL-1β for 24 hours before protein-normalized caspase-3/7 was measured. (C) RGC-5 cells plated at the seeding densities indicated (n = 5) were treated with 10 ng/mL IL-1β for 24 hours before measurement of the glucose content of the media. (D) RGC-5 cultures were fed with glucose-free medium, with or without 10 ng/mL IL-1β (n = 6), and WST-1 reduction was assayed at various times thereafter, as indicated. The reduction of WST-1 tetrazolium salt (OD, 450 nm with reference absorbance at 650 nm), caspase-3/7 activity (rates of Z-DEVD-rhodamine-110 product formation per cellular protein), and glucose concentrations were assayed. The mean and standard deviation are indicated. Probabilities were calculated with two-tailed Student’s t-test. Comparisons between control and IL-1β-treated groups within each plating density: *P < 0.05, **P < 0.01, and ***P < 0.001. For caspase activities, comparisons between control cultures at the lowest seeding density and at higher seeding densities: #P < 0.05 and ##P < 0.01.

Figure 3.

 

Effect of IL-1β on aerobic glycolysis. RGC-5 (A, C) and R28 (B, D) cultures were pretreated with or without 10 ng/mL IL-1β for 4 hours and then fed with fresh media with and without IL-1β. Media were removed at the times indicated, and glucose (A, B) and lactate (C, D) concentrations were determined. The mean and standard deviation for concentrations of each group of cultures (n = 6) are shown. Dashed lines: linear regression of mean concentrations for each time course. Slopes represent the mean and standard deviations of utilization or production rates obtained for each culture by separate linear regression analyses. Comparisons between control and IL-1β-treated groups at each time point were calculated using the two-tailed Student’s t-test: *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 4.

 

Effect of glycolytic and nonglycolytic energy substrates on WST-1 reductive potential of the RGC-5 and R28 cells. Metabolic substrates were added to cultures at the same time as cultures were treated with or without IL-1β, and 24 hours later the reduction of WST-1 tetrazolium salt was assayed. Mean values and standard deviations of optical densities (OD, 450 nm with reference absorbance at 650 nm) are shown. Effects of adding 10-mM concentrations of glycolytic (glucose), nonglycolytic (glutamine, lactate, and pyruvate), and a nonmetabolizable osmolite control (mannitol) on WST-1 reduction (n = 5) by RGC-5 (A) and R28 (B) cells. Effects of adding 20-mM concentrations of various glycolytic substrates (glucose, fructose, and mannose) and a competitive inhibitor of hexokinase, 2-deoxyglucose (2-dGlucose) on WST-1 reduction (n = 6) by RGC-5 (C) and R28 (D) cells. Control cultures received no additional metabolic substrate or inhibitor at the time of IL-1β treatment. Comparisons between control and IL-1β-treated groups were calculated with the two-tailed Student’s t-test: *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 5.

 

Effect of glucose and IL-1β on mitochondrial electronegativity, caspase activation, energy levels, and cell death. RGC-5 (A, C, E, G) and R28 (B, D, F, H) cultures were treated with or without 10 ng/mL IL-1β, with or without the addition of 20 mM glucose at the time of addition of IL-1β. After 24 hours of treatment, mitochondrial transmembrane electrical potential, as indicated by the ratio of JC-1 red-to-green fluorescence (A, B, n = 15), caspase-3/7 activity (C, D, n = 10), ATP content (E, F, n = 10), and cell lysis, as indicated by release into the media of LDH (G, H, n = 6), were measured. JC-1 ratios represent the ratio of red fluorescence (535 nm excitation, 590 nm emission) to green fluorescence (485 nm excitation, 530 nm emission). The mean and standard deviation for results obtained for each group of cultures are shown. P for comparisons between control and IL-1β-treated groups at each condition were calculated using two-tailed Student’s t-tes: *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 6.

 

Effect of IL-1β on R28 cell oxygen consumption and cytochrome c oxidase activity. The R28 cells were fed with fresh media, with or without (control) 10 ng/mL IL-1β, and oxygen consumption was determined polarographically 24 hours later. As a positive control, the cells were treated with 0.5 mM sodium azide for 18 hours before oxygen consumption was measured. (A) Oxygen consumption of intact cells suspended in fresh medium, representative of total respiratory activity. (B) Oxygen consumption of digitonin-permeabilized cells incubated with ADP and TMPD/ascorbate (an electron donor to cytochrome c) representative of cytochrome c oxidase (complex IV) activity. Oxygen consumption was measured twice in each culture and averaged. Results are the mean and standard deviation of average values obtained in three replicate cultures. P for comparisons between control and treated groups were calculated with two-tailed Student’s t-test: *P < 0.05.