- Email: [email protected]
Hyperglycaemia modifies energy metabolism and reactive oxygen species formation in endothelial cells in vitro
Archives of Biochemistry and Biophysics 542 (2014) 7–13
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Hyperglycaemia modifies energy metabolism and reactive oxygen species formation in endothelial cells in vitro Dorota Dymkowska ⇑, Beata Drabarek, Paulina Podszywałow-Bartnicka, Joanna Szczepanowska, Krzysztof Zabłocki Nencki Institute of Experimental Biology PAS, Warsaw, Poland
a r t i c l e
i n f o
Article history: Received 29 May 2013 and in revised form 6 November 2013 Available online 1 December 2013 Keywords: Endothelial cells Mitochondria Reactive oxygen species NADPH oxidase
a b s t r a c t There is significant evidence for an involvement of reactive oxygen species (ROS) in the pathogenesis of diabetic vascular complications through many metabolic and structural derangements. However, despite the advanced knowledge on the crucial role of ROS in cardiovascular damage, their intracellular source in endothelial cells exposed to high concentrations of glucose has not been precisely defined. Moreover, the molecular mechanism of action of elevated glucose on mitochondria has not been fully elucidated. The main aim of this study was to describe changes in the mitochondrial metabolism of human umbilical vein endothelial cells (HUVECs) treated with high glucose concentrations and to indicate the actual source of ROS in these cells. HUVECs exposed to 30 mM glucose exhibited an increased content of vascular adhesive molecule-1 (VCAM-1) and an excessive ROS production. Faster oxygen consumption and increased abundance of selected respiratory complexes coexist with slightly declined mitochondrial membrane potential and substantially elevated amount of uncoupling protein-2 (UCP2). Inhibition of NADPH oxidase (NOX) and modification of mitochondrial ROS generation with a mitochondrial uncoupler or respiratory chain inhibitors allowed concluding that the major source of ROS in HUVECs exposed to hyperglycaemic conditions is NOX. The mitochondrial respiratory chain seems not to participate in this phenomenon. Ó 2013 Elsevier Inc. All rights reserved.
Introduction Reactive oxygen species (ROS)1 generated during hyperglycaemia are postulated to be one of the most important contributors to the development and progression of diabetic vascular complications and often associated with endothelial dysfunction. Previous studies have indicated that not only prolonged exposure to hyperglycaemic conditions but also transient acute hyperglycaemia impairs endothelial function, as evidenced by an impaired ability to nitric oxide release [1]. Hyperglycaemia also alters the intracellular redox state which regulates the activity of several transcription factors, including nuclear factor jB (NF-jB), activator protein-1 (AP-1) and c-jun, implicated in the inducible expression of a wide variety of genes involved in oxidative stress and cellular stress response mechanisms [2]. For example, Bakkar et al. [3] showed that
⇑ Corresponding author. Address: Nencki Institute of Experimental Biology, 3 Pasteura Str. 02-093 Warsaw, Poland. Fax: +48 22 8225342. E-mail address: [email protected] (D. Dymkowska). 1 Abbreviations used: ROS, reactive oxygen species; HUVEC, human umbilical vein endothelial cells; VCAM-1, vascular adhesive molecule-1; ICAM-1, intracellular adhesive molecule-1; UCP2, uncoupling protein-2; eNOS, endothelial nitric oxide synthase; NOXNADPH oxidase, DPI diphenyleneiodonium - NADPH oxidase inhibitor. 0003-9861/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.abb.2013.11.008
the activation of NF-jB stimulated mitochondrial biogenesis and function, suggesting participation of so-called alternative signalling pathway in these processes. Activation of this pathway, mediated by NF-jB-inducing kinase, results in displacement of p52 protein to the nucleus and contributes to the mitochondrial mass increase. Moreover, an elevated serum glucose concentration decreases endothelial nitric oxide synthase (eNOS) mRNA and protein level in human endothelial cells by modulating mitochondrial ROS production. This, in turn, reduces nitric oxide (NO) availability [4]. Although several sources of reactive oxygen species may be involved in the response of endothelial cells to hyperglycaemia, NADPH oxidases (NOX) are postulated to be the most significant producers of ROS in the cardiovascular system. However, the mitochondrial electron transport needs also to be considered as an important source of reactive oxygen species. It should be noted that ROS, produced in a tightly controlled mode by NOX and the mitochondrial respiratory chain, could act as a physiological regulator of intracellular signalling pathways leading to cytokine secretion, vascular proliferation, hypertrophy and remodelling. On the other hand, an excessive ROS generation may be involved in atherosclerosis progression [5]. It also results in oxidative cell damage due to membrane lipid peroxidation, DNA strand breaks, and protein oxidation, all of which are
8
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
important processes contributing to endothelial dysfunction and cell death [6]. There is convincing evidence that hyperglycaemia leads to an activation of cellular stress pathways, and mitochondrial participation in this process is of particular interest. The high degree of endothelial mitochondria damage seen in atherosclerosis is one observation supporting such a correlation [7]. Mitochondrial ROS/RNS (reactive oxygen/nitrogen species)-evoked dysfunction results in a loss of control of metabolic function, activation of stress pathways and progression to vascular damage. However, still little is known about the changes in the metabolism of mitochondria in endothelial cells challenged by a high concentration of glucose. Some data indicate an increase in UCP2 level, which may protect endothelial cells from oxidative damage by slightly reducing the mitochondrial membrane potential and, consequently decreasing ROS formation. In fact, in primary aortic endothelial cells exposed to glucose at an extremely high concentration, UCP2 overexpression reduced intracellular ROS production as well as enhanced eNOS phosphorylation [8]. On the other hand, diabetes-associated metabolic disorders in the cardiovascular system may involve both mitochondrial dysfunction and the up-regulation of NADPH oxidase which may be the major source of ROS [9]. Thus, ROS may originate in endothelial cells mainly from the activity of NOX and the mitochondrial respiratory chain. For example, mitochondrial superoxide formation in bovine aortic endothelial cells (BAEC) was confirmed by chemical depletion of mitochondrial DNA (rho0 cells) or after treatment with mitochondria-targeted antioxidants [10]. Alternatively, increased ROS formation in human aortic endothelial cells was attributed to endothelial nitric oxide synthase activity [11]. In immortalized EA.hy926 cells hyperglycaemia-induced ROS appear to be produced by both NADPH oxidase and mitochondrial respiratory chain [12]. In this paper the mechanism of glucose-induced ROS formation in HUVEC cells is investigated. Selective inhibition of NOX and uncoupling of oxidative phosphorylation/inhibition of the respiratory chain allowed us to conclude that the increased ROS generation under hyperglycaemic conditions is dependent on NOX rather than on the mitochondrial electron transport. Moreover, the influence of hyperglycaemia on the mitochondrial energy metabolism and the proteins involved in oxidative phosphorylation and oxidative stress regulation was also investigated. Understanding of the mechanism(s) linking hyperglycaemia and energy metabolism of endothelial cells as well as a precise identification of the intracellular ROS source under such pathological conditions could help to design a strategy for cell protection against the deleterious effect of increased serum glucose concentration.
Methods Cell culture Human umbilical vein endothelial cells purchased from Lonza (Walkersville, MD USA) were grown in Endothelial Cell Growth Medium BulletKitÒ-2 (EGM-2 BulletKit, Lonza) at 37 °C in the atmosphere of 5% CO2 and 95% air. The culture medium was supplemented with conveniently packaged as single-use aliquots called SingleQuots, containing as follows: human recombinant epidermal growth factor, human fibroblast growth factor, vascular endothelial growth factor, ascorbic acid, hydrocortisone, human recombinant insulin-like growth factor, heparin, 2% foetal bovine serum and gentamicin with amphotericin (Lonza). Cells were passaged every 2 days. Passages 2 through 5 were used for all experiments. Confluent cells were grown in normal (5 mM) or high (30 mM) glucose conditions for 48 h.
Oxygen consumption Cellular respiration rate was measured polarographically as oxygen uptake using OROBOROS Oxygraph-2k (OROBOROSÒ INSTRUMENTS GmbH, Austria) at 37 °C. Confluent cells grown in 10 cm tissue culture dishes were trypsinized, centrifuged and resuspended in 2 ml PBS pre-warmed to 37 °C (approximate density 0.5 mg protein/ml). Prior to the assay calibration procedure and background correction was performed according to the manufacturer’s protocol. Oxygen consumption was measured in the presence of sequentially added respiratory substrates (1 mM pyruvate, 5 mM glucose, Sigma), oligomycin (0.1 lg/ml, Sigma) and CCCP (0.5 lM, Sigma). Respiration rate was normalized to the amount of protein in the assay. Mitochondrial membrane potential Mitochondrial membrane potential was measured fluorimetrically with 5,50 ,6,60 -tetrachloro-1,10 ,3,30 -tetraethylbenzimidazolcarbocyanine iodide (JC-1, Molecular Probes, Invitrogen) according to the method of Cossarizza et al. [13]. Confluent cells growing in 24-well culture plates were stained with JC-1 (5 lM final concentration in the culture medium) and incubated for 15 min at 37 °C in the dark. Then, the cells were rinsed three times with the culture medium followed by PBS. Finally, 0.5 ml PBS was added to each well. The cells with completely deenergized mitochondria due to addition of 2 lg/ml of valinomycin plus 5 lM CCCP were used as a control. Fluorescence was measured with a laser scanning cytometer iCYS equipped with argon laser (excitation 488 nm). The data were presented as the ratio of orange (energized) to green (deenergized) fluorescence. Reactive oxygen species ROS were measured with carboxy-20 ,70 -difluorodihydrofluorescein diacetate (DFFH2-DA, Molecular Probes, Invitrogen). Confluent cells grown in 12-well plates were stained with 10 lM DFFH2-DA in PBS for 30 min at 37 °C in the dark. Then, fluorescence was measured at 485 nm excitation and 520 nm emission wavelengths using Infinite 200 micro plate reader (Tecan). Total green fluorescence was normalized to the amount of protein in each well. Mitochondrial mass Mitochondrial mass was estimated fluorimetrically with Mitotracker Green (Molecular Probes, Invitrogen). Confluent cells growing in 24-well plates were stained with this dye at 100 nM concentration in PBS for 20 min at 37 °C. Then, the cells were rinsed three times with the culture medium and then with PBS. Finally, 0.5 ml PBS was added to each well. Fluorescence was measured with laser scanning cytometer iCYS equipped with argon laser (excitation 488 nm). Total green fluorescence was normalized to the amount of protein in each well. Cell lysis and Western blot analysis Cellular lysates were prepared as previously described [14]. Lysate aliquots containing 25–50 lg protein were loaded on 10% or 12% SDS–polyacrylamide gels, depending on the molecular mass of the protein to be investigated, electrophoresed and then transferred onto nitrocellulose membrane. The membranes were treated with specific primary antibodies listed below, as recommended by the manufacturer. All secondary antibodies conjugated with horseradish peroxidase were obtained from Abcam (1:5000). Western blots were developed with chemiluminescent substrate Luminata Classico (Millipore). The content of specific
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
9
proteins was determined densitometrically. The following primary antibodies were applied: VCAM-1 (Santa Cruz, 1:200); VDAC (Santa Cruz, 1:500); TOM20 (Santa Cruz, 1:2000); Hsp60 (Sigma, 1:400); UCP2 (Abcam, 1:1000); MitoProfile total OXPHOS human WB antibody cocktail (Mito Sciences, 1:250); b-actin (monoclonal anti-b-actin antibody peroxidase conjugated, 1:50,000, Sigma). RNA isolation and RT-PCR analysis To estimate UCP2 mRNA level in HUVEC cells, total RNA from cultured cells was isolated using TRIzol reagent (Sigma). RNA concentration and purity were measured spectrometrically as absorbance ratio 260 nm/280 nm and 1 lg of RNA was subjected to reverse transcription with M-MLV enzyme (Promega) and oligodT starters (Sigma). Real-time PCR was performed using 10% of the reverse transcription reaction volume and SYBR Green Master Mix (Applied Biosystems) on a 7500 ABI Prism Real-time PCR System (Applied Biosystems). The primers for UCP2 (NM_003355) were as follows: forward 50 -CAGTTCTACACCAAGGGCTC (within exon 2) and reverse 50 -GGGAGGTCATCTGTCATGAG (within exon 4); b-actin (NM_001101) was used as a reference gene, the primers were: forward 50 -CATGTACGTTGCTATCCAGGC and reverse 50 -CTCC TTAATGTCACGCACGAT. Cycling conditions were 2 min (50 °C), followed by 10 min (95 °C) and then 45 cycles of 15 s (95 °C) and 1 min (60 °C). Products of the PCR reaction were separated in a 2% agarose gel. After normalization to b-actin the 2DDCT method was used to determine the relative mRNA level [15]. Immunocytochemistry To visualise mitochondrial network and cytoskeleton architecture HUVEC cells grown on glass coverslips were stained with MitoTracker CMX Ros (Invitrogen, San Diego, CA, USA), DAPI and FITC-labelled phalloidin (Molecular Probes), as described previously [14]. Confocal fluorescence microscopy was carried out using a Leica TCS SP5 IL Spectral Confocal and Multiphoton Microscope. Expression of results Data are shown as means of ratios of treatment (experimental group) to control values ± SD, for the number of replicates indicated in the figure legends. One-sample t-test was performed on these ratios between analysed sample and control to test for deviation from unity and associated 95% confidence intervals (CIs) were computed. Since normality assumption cannot be strictly met and positive skewness may be expected, analogous computations were repeated for log-transformed values. The resulting p-values and confidence intervals were usually very similar for the both methods of calculation and we refer to the higher (conservative) p-values and wider confidence limits. Results and discussion Prolonged exposure of vascular endothelial cells to glucose at a concentration exceeding 10 mM is commonly applied as a procedure resembling hyperglycaemic conditions. Here, HUVEC cells challenged with 30 mM glucose for 48 h were employed to test the mitochondrial response to hyperglycaemia. Such treatment does not affect cellular viability, as determined with propidium iodide staining of the cells (not shown). As shown in Fig. 1A, incubation of HUVEC cells with 30 mM glucose for 48 h resulted in a substantially increased content of VCAM-1 protein (1.92 ± 0.98-fold). Similarly, the level of intracellular adhesive molecule-1 (ICAM-1) protein was also elevated (1.27 ± 0.34-fold, not shown). A shorter (for 24 h) exposure of the
Fig. 1. Effect of 30 mM glucose on vascular adhesive molecule 1 abundance and ROS formation in HUVEC cells. Cells were grown in standard medium supplemented with 30 mM glucose. (A) Western blot of VCAM-1 protein (one typical experiment out of twelve) for control cells (exposed to normal, 5 mM glucose) and cells treated with 30 mM glucose. Data on the right show the mean value ± SD for twelve experiments. VCAM-1 protein level in control cells was taken as 1. p < 0.008, 95% Cl: [1.27, 2.55]. (B) Effect of NADPH oxidase inhibitors (DPI and VAS2870) on ROS formation in cells treated with 30 mM glucose for 6 or 48 h. Data from ten independent experiments expressed as percentage of control (cells exposed to normal, 5 mM glucose). 95% Cl: [0.68, 0.86] for 30 mM glucose/DPI and [0.59, 0.96] for 30 mM glucose/VAS. (C) Effect of oxidative phosphorylation uncoupler (CCCP) on ROS generation in cells treated with 30 mM glucose. Data from five independent experiments. Confidence limit for 30 mM glucose/CCCP/DPI was [0.67, 0.88]. (D) Effects of rotenone and DPI on ROS generation in control cells. Data from five independent experiments. 95% Cl: [0.85, 0.99] for DPI, [1.01, 1.21] for rotenone, [0.83,1.20] for rotenone/DPI. p-values for data shown in panels B–D: ⁄p < 0.0002, ⁄⁄ p < 0.04 comparing to control; #p < 0.0004, ##p < 0.04 comparing to 30 mM glucose. Grey lines at the level of 1.00 define values for control cells without any additions. The first two bars on the left of graphs are identical for panels B and C.
cells to 30 mM glucose resulted in a much less substantial inflammatory response. At the same time, a significant enhancement of ROS formation was observed (Fig. 1 B and C), as early as after 6 h of treatment with high concentration of glucose, thus prior to
10
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
any increase in the VCAM-1 protein level. These effects confirm the previously described response of endothelial cells to glucose at high concentration and are considered typical for endothelial cell dysfunction [16]. They also may participate in an impairment of the endothelial function leading to severe complications. It has previously been shown that an excessive ROS formation is responsible for the deleterious effect of hyperglycaemia in HUVECs both in the case of intermittent as well as constant elevation of glucose concentration [17]. Although the influence of ROS on endothelial cell functioning is unquestionable, the source of the excessive ROS formation within the cells challenged by hyperglycaemia has not been unequivocally established so far [18]. Data shown in Fig. 1 provide, to some extent, a solution to this question. Preincubation of HUVEC cells with 100 nM diphenyleneiodonium (DPI, a NOX inhibitor) for 1 h prior to ROS measurement led to a substantial reduction of ROS formation in cells exposed to 30 mM glucose for 6 or 48 h (Fig. 1B). To confirm this observation and avoid any risk of DPI-evoked side effects resulting from its limited specificity, similar experiments with the use of the more specific NADPH oxidase inhibitor VAS2870 [19] were undertaken. Indeed, this compound applied at 5 lM concentration for 1 h prior to ROS measurement significantly reduced the ROS generation in HUVEC cells exposed to 30 mM glucose for 48 h. This convincingly documents participation of NOX in the ROS formation induced by excessive glucose. On the other hand, a complete dissipation of the mitochondrial membrane potential with CCCP, which usually decreases mitochondrial ROS generation, resulted in only a slight and statistically insignificant reduction of ROS formation in the cells, independently of the glucose concentration (data not shown). However, an application of this uncoupler together with DPI in the cells exposed to 30 mM glucose brought down the ROS production to the level found in the control (normal glucose) cells (Fig. 1C). Furthermore, an inhibition of complex I of the respiratory chain by rotenone resulted in a reproducible enhancement of ROS generation in control cells (Fig. 1D). Similar effects were observed after addition of antimycin A, a complex III inhibitor (data not shown). The slight decrease in ROS formation in control cells treated with DPI alone may indicate some basal NOX-dependent ROS generation under control (normoglycaemic) conditions. On the other hand, an addition of DPI did not significantly prevent the rotenone action, indicating that DPI does not interfere with the respiratory chainoriginated ROS. Interestingly, rotenone and CCCP prevented an excessive ROS generation in BAEC cells exposed to 30 mM glucose [20]. This discrepancy probably reflects different metabolic properties of endothelial cells derived from various sources. In sum, the results presented hitherto suggest that NADPH oxidase is the major source of ROS generated by HUVEC cells exposed to 30 mM glucose. This is in agreement with the data obtained for bovine aortic endothelial cells, in which accelerated ROS formation was related to PKC-dependent activation of NOX [21]. Our preliminary results seem to indicate an increase of the proportion of NADPH oxidase isoform 4 (NOX4) in the purified mitochondrial fraction from HUVEC cells grown under hyperglycaemic conditions while the total amount of this enzyme found in a cellular homogenate was unchanged (data not shown). One may thus speculate that NOX4 may be recruited to mitochondria under hyperglycaemic conditions. Thus, in HUVEC cells exposed to glucose at high concentration ROS may also be formed in the mitochondria but without an involvement of the mitochondrial respiratory chain. However, the identification of the particular isoform of NOX responsible for the excessive ROS generation in HUVEC cells under hyperglycaemic conditions is beyond the scope of this article. Although the aforementioned data clearly show that the enhanced ROS generation in HUVEC cells grown in the presence of 30 mM glucose is due to NADPH oxidase activity, the substantial changes in the mitochondrial metabolism occurring concomitantly
cannot be ignored. In fact, as shown in Fig. 2A, the oxygen consumption by cells exposed to hyperglycaemic conditions is considerably enhanced. This is observed when cells utilize endogenous substrates (basal respiration), also after addition of oligomycin and a mitochondrial uncoupler (CCCP). The basal respiration rate did not change after addition of external respiratory substrates, indicating that intracellular substrates are sufficiently abundant to maintain oxygen consumption at the unchanged level. The similar degree of respiratory stimulation by high glucose treatment under all conditions tested (basal, inhibited ATP synthase, uncoupled oxidative phosphorylation) suggests an increase of the total respiratory capacity in these cells rather than partial dissipation of the mitochondrial membrane potential. This could be a consequence of an improved efficiency of the respiratory chain and/or an increased amount of mitochondria. The increased abundance of the components of the respiratory chain in cells exposed to 30 mM glucose shown in Fig. 2B supports the latter possibility. On the other hand, a lack of an effect of hyperglycaemia on the NRF-1 protein content (not shown) seems to exclude increased mitochondrial biogenesis under hyperglycaemic conditions. Therefore our results allowed assuming that glucose at high concentration triggers bioenergetic reserves as important factor in response to stress in HUVECs as it was indicated earlier [22]. The demonstrated respiratory stimulation by glucose treatment
Fig. 2. Effect of 30 mM glucose on oxygen consumption and the abundance of respiratory chain components in HUVEC cells. (A) Effect of hyperglycaemia on oxygen consumption. Absolute rate of oxygen consumption (pmol O2/s per mg protein) is shown in the right part of panel A. Values for control cells treated with the same inhibitor or uncoupler as cells exposed to 30 mM glucose were taken as 100%. Data for six independent experiments. ⁄p < 0.009 (lower confidence limits for basal 108%, for oligomycin 117%, for CCCP 106%). Substrates and inhibitors were added sequentially directly to the incubation chamber as indicated, at final concentrations: pyruvate 1 mM, glucose 5 mM, oligomycin 0.1 lg/ml, CCCP 0.5 lM. (B) Western blot analysis of selected respiratory chain components. Protein level of respiratory complexes II, III, IV and V was normalized to b-actin and is expressed as percentage of control. Data from three independent experiments. ⁄ p < 0.05 (upper confidence limits: for complex II-166%; for complex III-149%, for complex IV-175%, for complex V-117%). Grey lines at the level of 100% define values for control cells.
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
of HUVECs seems to resemble the response of BAEC cells to both ROS and RNS. Dranka and co-workers found that unaffected cells use only about 35% of their maximal respiratory capacity, a state intermediate between states 3 and 4, indicating the presence of a reserve respiratory capacity available for the cells when the bioenergetic demand is increased [22]. The hypothetical effect of hyperglycaemia on mitochondrial biogenesis was tested with the use of MitoTracker Green. As shown in Fig. 3A, the elevated glucose concentration does not alter the total mitochondrial mass, in line with the lack of an effect on the level of NRF-1 protein. This observation was also confirmed by a Western blot analysis of the abundance of the following marker proteins: heat shock protein 60 (HSP60), voltage-dependent anion channel (VDAC) and translocase of outer membrane 20 kDa subunit (TOM20). As shown in Fig. 3A, exposition of HUVECs to 30 mM did not change the level of these proteins. On the other
11
hand, hyperglycaemia affected the mitochondrial network architecture by increasing the proportion of fragmented vs. tubular organelles (Fig. 3B), which could correlate with an affected mitochondrial metabolism. Similar results have been observed in bovine retinal pericytes and rat retinal endothelial cells growing in the presence of glucose at high concentration. Trudeau et al. [23,24] observed that fragmentation of mitochondrial network was accompanied by a decreased mitochondrial membrane potential and reduced extracellular acidification. In retinocytes similar reorganization of the mitochondrial network was also associated with accelerated apoptosis and, consequently, was postulated as a cause of diabetic retinopathy. In addition, Makino and co-workers [25] suggested that superoxide anion radical might serve as a direct regulator of mitochondrial morphology in mouse coronary endothelial cells. Overproduction of ROS led to mitochondrial fragmentation and a treatment with antioxidative agents was
Fig. 3. Effect of 30 mM glucose on mitochondrial mass, and structure and abundance of mitochondrial markers. (A) Total mitochondrial mass estimated with MitoTracker Green and the amount of mitochondrial proteins HSP60, VDAC and TOM20. Data for MitoTracker Green staining for five independent experiments, 95% Cl: [0.89, 1.04]. Blots show results of two independent experiments out of six. C indicates control cells, HG – cells exposed to 30 mM glucose for 48 h. B, Mitochondrial organization: green, actin cytoskeleton stained with FITC-labelled phalloidin; red, mitochondria stained with MitoTracker CMX Ros; blue, nucleus stained with DAPI. All images are taken in the same confocal plane. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
12
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
proposed to help improve mitochondrial function. Reorganization of the mitochondrial network could reflect a change in cellular bioenergetics. Previously it has been shown that undisturbed mitochondrial membrane potential together with efficient ATP synthesis by oxidative phosphorylation and glycolysis, as well as unchanged ROS level, are required to maintain the balance between mitochondrial fission and fusion in HUVECs. In depolarized mitochondria, as well as after exposure to oxidative stress, fragmentation of the tubular network was observed [26]. Considering the above-mentioned link between the mitochondrial architecture and cellular metabolism, it seems likely that the changes in the mitochondrial network organization and the increased oxygen consumption in HUVEC cells exposed to 30 mM glucose could also be accompanied by a decrease of the mitochondrial membrane potential. In fact, we found a slight reduction of this parameter, with confidence interval between 0.57 and 1.08 relative to the control (Fig. 4A). In fact, the lack of linearity between the acceleration of oxygen consumption and the simultaneous reduction of mitochondrial membrane potential [27] speaks against a substantial mitochondrial deenergization in view of the relatively small (ca. 20%) stimulation of respiration observed here. This slight decrease of Dw under hyperglycaemic conditions shown here is in disagreement with the observations published previously by Rolo and Palmeira [28]. These authors pointed out that the hyperglycaemia-induced hyperpolarization drives the mitochondrial ROS formation which were decreased after mitochondrial uncoupling with CCCP. As has already been mentioned, CCCP did not influence ROS formation in HUVECs exposed to hyperglycaemic conditions (see Fig. 1C). Thus our results indicate the generation of ROS in the mitochondrial respiratory chain independent mechanism. A similar result was observed in retinal endothelial cells, where a heterogeneity of mitochondrial potential was
postulated and associated with increased cellular stress and appeared to reflect variations of mitochondrial responses. Several factors causing such a diversity of mitochondrial transmembrane potential were postulated, e.g., heterogeneous distribution of respiratory chain components, uneven access of metabolites and calcium, and/or changes in mitochondrial localization [23]. The slightly decreased mitochondrial membrane potential is in line with an increased amount of UCP2, as shown by Western blot data (Fig. 4B and C), probably responsible for dissipating the mitochondrial membrane potential [29]. The enhanced oxygen consumption (Fig. 2A) can also be partially explained by the increased UCP2 abundance. Giardina and co-workers [30] showed that mitochondria-originated oxidative stress may up-regulate the UCP2 content in macrophages, but oxidants acting outside mitochondria did not affect the amount of this protein. In macrophages treated with mitochondrial inhibitors (rotenone, antimycin A and diethyldithiocarbamate) to raise mitochondrial superoxide production, a rapid increase of UCP2 level was observed. Those authors found no such effect in response to superoxide produced outside mitochondria or in response to H2O2. Our results seem to be in disagreement with their data unless one of NOX isoforms is localised to mitochondria. We found that the increased UCP2 content is not accompanied by a parallel change of its mRNA level (Fig. 4B and C). This suggests that glucose at high concentration enhances the translational efficiency of the mRNA and possibly also stabilizes UCP2 in the inner mitochondrial membrane, as it was postulated earlier [30]. An increased abundance of UCP2 could be essential for reduction of oxidative stress and appears to depend on AMP-activated protein kinase. AMPK has been suggested to be a physiological regulator of ROS formation [31]. It can also be assumed that an increase in UCP2 level implies increased oxygen consumption. It also correlates with the observed slightly
Fig. 4. Effect of 30 mM glucose on mitochondrial membrane potential and UCP2 content. (A) Orange/green fluorescence ratio in cells loaded with JC-1. In control cells this is taken as 1. Data from seven independent experiments, 95%Cl: [0.57, 1.08]. (B) UCP2 protein level and mRNA content, both normalized to b-actin protein and transcript, respectively. Data expressed as percentage of control, derived from seven independent experiments for Western blot and three independent experiments for real time RTPCR. ⁄p < 0.008, 95%Cl: [1.20, 2.16]. (C) Western blot for one typical experiment out of seven and analysis of RT-PCR products of UCP2 mRNA for one typical experiment out of three are shown.
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
decreased mitochondrial membrane potential. It cannot be excluded that UCP2 counteracts mitochondrial ROS generation. Elevated glucose concentration leads to endothelial cell dysfunction and affects mitochondrial metabolism. It seems that an increased abundance of UCP2 is crucial for the modulation of mitochondrial metabolism, as described earlier by Lowell et al. [32]. UCPs and particularly UCP2 have been shown as possible regulators of the production of mitochondrial oxidants. This protein acts by modulating the proton leak across the inner membrane, but the antioxidative action of UCP2 is independent of its effect on the mitochondrial membrane potential. It has been suggested that hyperglycaemia may increase cellular reactive oxygen species production and the abundance and activity of the inner mitochondrial membrane UCP2 [7]. In conclusion, our results indicate that exposition of HUVEC cells to a high concentration of glucose seriously affects mitochondrial energy metabolism, but these changes do not seem to be directly responsible for the observed accelerated ROS generation. This latter event is rather caused by an increased NADPH oxidase activity. Acknowledgments We thank Professor Lech Wojtczak for critical reading of the manuscript and Dr. Tomasz Wyszomirski for help in statistical data analysis. This work was supported by the National Science Centre grant N N301 291137 to D.D. References [1] M.A. Potenza, S. Gagliardi, C. Nacci, M.R. Carratu, M. Montagnani, Curr. Med. Chem. 16 (2009) 94–112. [2] C.K. Sen, L. Packer, FASEB J. 10 (1996) 709–720. [3] N. Bakkar, J. Wang, K.J. Ladner, H. Wang, J.M. Dahlman, M. Carathers, S. Acharyya, M.A. Rudnicki, A.D. Hollenbach, D.C. Guttridge, J. Cell Biol. 180 (2008) 787–802. [4] S. Srinivasan, M.E. Hatley, D.T. Bolick, L.A. Palmer, D. Edelstein, M. Brownlee, C.C. Hedrick, Diabetologia 47 (2004) 1727–1734.
13
[5] C.E. Murdoch, M. Zhang, A.C. Cave, A.M. Shah, Cardiovasc. Res. 71 (2006) 208– 215. [6] I. Al Ghouleh, N.K.H. Khoo, U.G. Knaus, K.K. Griendling, R.M. Touyz, V.J. Thannickal, A. Barchowsky, W.M. Nauseef, E.E. Kelley, P.M. Bauer, V. Darley-Usmar, S. Shiva, E. Cifuentes-Pagano, B.A. Freeman, M.T. Gladwin, P.J., Free Radical Biol. Med. 51 (2011) 1271–1288. [7] S.W. Ballinger, Free Radical Biol. Med. 38 (2005) 1278–1295. [8] X.Y. Tian, W.T. Wong, A. Xu, Y. Lu, Y. Zhang, L. Wang, W.S. Cheang, Y. Wang, X. Yao, Y. Huang, Circ. Res. 110 (2012) 1211–1216. [9] G.X. Shen, Can. J. Physiol. Pharmacol. 88 (2010) 241–248. [10] C. Quijano, L. Castro, G. Peluffo, V. Valez, R. Radi, Am. J. Physiol. Heart Circ. Physiol. 293 (2007) H3404–H3414. [11] S. Cai, J. Khoo, K.M. Channon, Cardiovasc. Res. 65 (2005) 823–831. [12] A. Koziel, A. Woyda-Ploszczyca, A. Kicinska, W. Jarmuszkiewicz, Pfluegers Arch. 464 (2012) 657–669. [13] A. Cossarizza, M. Baccarani-Contri, G. Kalashnikova, C. Franceschi, Biochem. Biophys. Res. Commun. 197 (1993) 40–45. [14] B. Drabarek, D. Dymkowska, J. Szczepanowska, K. Zabłocki, Int. J. Biochem. Cell Biol. 44 (2012) 1390–1397. [15] K.J. Livak, T.D. Schmittgen, Methods 25 (2001) 402–408. [16] D. Popov, Int. J. Diabetes Mellit. 2 (2010) 189–195. [17] L. Piconi, L. Quagliaro, R. Assaloni, R. Da Ros, A. Maier, G. Zuodar, A. Ceriello, Diabetes Metab. Res. Rev. 22 (2006) 198–203. [18] M.M. Spitaler, W.F. Graier, Diabetologia 45 (2002) 476–494. [19] S. Wind, K. Beuerlein, T. Eucker, H. Müller, P. Scheurer, M.E. Armitage, H. Ho, H.H.H.W. Schmidt, K. Wingler, Br. J. Pharmacol. 161 (2010) 885–898. [20] T. Nishikawa, D. Edelstein, X.L. Du, S. Yamagishi, T. Matsumura, Y. Kaneda, M.A. Yorek, D. Beebe, P.J. Oates, H.P. Hammes, I. Giardino, M. Brownlee, Nature 404 (2000) 787–790. [21] T. Inoguchi, P. Li, F. Umeda, H.Y. Yu, M. Kakimoto, M. Imamura, T. Aoki, T. Etoh, T. Hashimoto, M. Naruse, H. Sano, H. Utsumi, H. Nawata, Diabetes 49 (2000) 1939–1945. [22] B.P. Dranka, B.G. Hill, V.M. Darley-Usmar, Free Radical Biol. Med. 48 (2010) 905–914. [23] K. Trudeau, A.J.A. Molina, W. Guo, S. Roy, Am. J. Pathol. 177 (2010) 447–455. [24] K. Trudeau, A.J.A. Molina, S. Roy, Invest. Ophthalmol. Vis. Sci. 52 (2011) 8657– 8664. [25] A. Makino, B.T. Scott, W.H. Dillmann, Diabetologia 53 (2010) 1783–1794. [26] R.J. Giedt, D.R. Pfeiffer, A. Matzavinos, Ann. Biomed. Eng. 40 (2012) 1903–1916. [27] L. Wojtczak, K. Bogucka, J. Duszyn´ski, B. Zabłocka, A. Zółkiewska, Biochim. Biophys. Acta 1018 (1990) 177–181. [28] A.P. Rolo, C.M. Palmeira, Toxicol. Appl. Pharmacol. 212 (2006) 167–178. [29] B.D. Fink, Y.S. Hong, M.M. Mathahs, T.D. Scholz, J.S. Dillon, W.I. Sivitz, J. Biol. Chem. 277 (2002) 3918–3925. [30] T.M. Giardina, J.H. Steer, S.Z. Lo, D.A. Joyce, Biochim. Biophys. Acta 1777 (2008) 118–129. [31] Z. Xie, J. Zhang, J. Wu, B. Viollet, M.H. Zou, Diabetes 57 (2008) 3222–3230. [32] B.B. Lowell, G.I. Shulman, Science 307 (2005) 384–387.
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Hyperglycaemia modifies energy metabolism and reactive oxygen species formation in endothelial cells in vitro Dorota Dymkowska ⇑, Beata Drabarek, Paulina Podszywałow-Bartnicka, Joanna Szczepanowska, Krzysztof Zabłocki Nencki Institute of Experimental Biology PAS, Warsaw, Poland
a r t i c l e
i n f o
Article history: Received 29 May 2013 and in revised form 6 November 2013 Available online 1 December 2013 Keywords: Endothelial cells Mitochondria Reactive oxygen species NADPH oxidase
a b s t r a c t There is significant evidence for an involvement of reactive oxygen species (ROS) in the pathogenesis of diabetic vascular complications through many metabolic and structural derangements. However, despite the advanced knowledge on the crucial role of ROS in cardiovascular damage, their intracellular source in endothelial cells exposed to high concentrations of glucose has not been precisely defined. Moreover, the molecular mechanism of action of elevated glucose on mitochondria has not been fully elucidated. The main aim of this study was to describe changes in the mitochondrial metabolism of human umbilical vein endothelial cells (HUVECs) treated with high glucose concentrations and to indicate the actual source of ROS in these cells. HUVECs exposed to 30 mM glucose exhibited an increased content of vascular adhesive molecule-1 (VCAM-1) and an excessive ROS production. Faster oxygen consumption and increased abundance of selected respiratory complexes coexist with slightly declined mitochondrial membrane potential and substantially elevated amount of uncoupling protein-2 (UCP2). Inhibition of NADPH oxidase (NOX) and modification of mitochondrial ROS generation with a mitochondrial uncoupler or respiratory chain inhibitors allowed concluding that the major source of ROS in HUVECs exposed to hyperglycaemic conditions is NOX. The mitochondrial respiratory chain seems not to participate in this phenomenon. Ó 2013 Elsevier Inc. All rights reserved.
Introduction Reactive oxygen species (ROS)1 generated during hyperglycaemia are postulated to be one of the most important contributors to the development and progression of diabetic vascular complications and often associated with endothelial dysfunction. Previous studies have indicated that not only prolonged exposure to hyperglycaemic conditions but also transient acute hyperglycaemia impairs endothelial function, as evidenced by an impaired ability to nitric oxide release [1]. Hyperglycaemia also alters the intracellular redox state which regulates the activity of several transcription factors, including nuclear factor jB (NF-jB), activator protein-1 (AP-1) and c-jun, implicated in the inducible expression of a wide variety of genes involved in oxidative stress and cellular stress response mechanisms [2]. For example, Bakkar et al. [3] showed that
⇑ Corresponding author. Address: Nencki Institute of Experimental Biology, 3 Pasteura Str. 02-093 Warsaw, Poland. Fax: +48 22 8225342. E-mail address: [email protected] (D. Dymkowska). 1 Abbreviations used: ROS, reactive oxygen species; HUVEC, human umbilical vein endothelial cells; VCAM-1, vascular adhesive molecule-1; ICAM-1, intracellular adhesive molecule-1; UCP2, uncoupling protein-2; eNOS, endothelial nitric oxide synthase; NOXNADPH oxidase, DPI diphenyleneiodonium - NADPH oxidase inhibitor. 0003-9861/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.abb.2013.11.008
the activation of NF-jB stimulated mitochondrial biogenesis and function, suggesting participation of so-called alternative signalling pathway in these processes. Activation of this pathway, mediated by NF-jB-inducing kinase, results in displacement of p52 protein to the nucleus and contributes to the mitochondrial mass increase. Moreover, an elevated serum glucose concentration decreases endothelial nitric oxide synthase (eNOS) mRNA and protein level in human endothelial cells by modulating mitochondrial ROS production. This, in turn, reduces nitric oxide (NO) availability [4]. Although several sources of reactive oxygen species may be involved in the response of endothelial cells to hyperglycaemia, NADPH oxidases (NOX) are postulated to be the most significant producers of ROS in the cardiovascular system. However, the mitochondrial electron transport needs also to be considered as an important source of reactive oxygen species. It should be noted that ROS, produced in a tightly controlled mode by NOX and the mitochondrial respiratory chain, could act as a physiological regulator of intracellular signalling pathways leading to cytokine secretion, vascular proliferation, hypertrophy and remodelling. On the other hand, an excessive ROS generation may be involved in atherosclerosis progression [5]. It also results in oxidative cell damage due to membrane lipid peroxidation, DNA strand breaks, and protein oxidation, all of which are
8
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
important processes contributing to endothelial dysfunction and cell death [6]. There is convincing evidence that hyperglycaemia leads to an activation of cellular stress pathways, and mitochondrial participation in this process is of particular interest. The high degree of endothelial mitochondria damage seen in atherosclerosis is one observation supporting such a correlation [7]. Mitochondrial ROS/RNS (reactive oxygen/nitrogen species)-evoked dysfunction results in a loss of control of metabolic function, activation of stress pathways and progression to vascular damage. However, still little is known about the changes in the metabolism of mitochondria in endothelial cells challenged by a high concentration of glucose. Some data indicate an increase in UCP2 level, which may protect endothelial cells from oxidative damage by slightly reducing the mitochondrial membrane potential and, consequently decreasing ROS formation. In fact, in primary aortic endothelial cells exposed to glucose at an extremely high concentration, UCP2 overexpression reduced intracellular ROS production as well as enhanced eNOS phosphorylation [8]. On the other hand, diabetes-associated metabolic disorders in the cardiovascular system may involve both mitochondrial dysfunction and the up-regulation of NADPH oxidase which may be the major source of ROS [9]. Thus, ROS may originate in endothelial cells mainly from the activity of NOX and the mitochondrial respiratory chain. For example, mitochondrial superoxide formation in bovine aortic endothelial cells (BAEC) was confirmed by chemical depletion of mitochondrial DNA (rho0 cells) or after treatment with mitochondria-targeted antioxidants [10]. Alternatively, increased ROS formation in human aortic endothelial cells was attributed to endothelial nitric oxide synthase activity [11]. In immortalized EA.hy926 cells hyperglycaemia-induced ROS appear to be produced by both NADPH oxidase and mitochondrial respiratory chain [12]. In this paper the mechanism of glucose-induced ROS formation in HUVEC cells is investigated. Selective inhibition of NOX and uncoupling of oxidative phosphorylation/inhibition of the respiratory chain allowed us to conclude that the increased ROS generation under hyperglycaemic conditions is dependent on NOX rather than on the mitochondrial electron transport. Moreover, the influence of hyperglycaemia on the mitochondrial energy metabolism and the proteins involved in oxidative phosphorylation and oxidative stress regulation was also investigated. Understanding of the mechanism(s) linking hyperglycaemia and energy metabolism of endothelial cells as well as a precise identification of the intracellular ROS source under such pathological conditions could help to design a strategy for cell protection against the deleterious effect of increased serum glucose concentration.
Methods Cell culture Human umbilical vein endothelial cells purchased from Lonza (Walkersville, MD USA) were grown in Endothelial Cell Growth Medium BulletKitÒ-2 (EGM-2 BulletKit, Lonza) at 37 °C in the atmosphere of 5% CO2 and 95% air. The culture medium was supplemented with conveniently packaged as single-use aliquots called SingleQuots, containing as follows: human recombinant epidermal growth factor, human fibroblast growth factor, vascular endothelial growth factor, ascorbic acid, hydrocortisone, human recombinant insulin-like growth factor, heparin, 2% foetal bovine serum and gentamicin with amphotericin (Lonza). Cells were passaged every 2 days. Passages 2 through 5 were used for all experiments. Confluent cells were grown in normal (5 mM) or high (30 mM) glucose conditions for 48 h.
Oxygen consumption Cellular respiration rate was measured polarographically as oxygen uptake using OROBOROS Oxygraph-2k (OROBOROSÒ INSTRUMENTS GmbH, Austria) at 37 °C. Confluent cells grown in 10 cm tissue culture dishes were trypsinized, centrifuged and resuspended in 2 ml PBS pre-warmed to 37 °C (approximate density 0.5 mg protein/ml). Prior to the assay calibration procedure and background correction was performed according to the manufacturer’s protocol. Oxygen consumption was measured in the presence of sequentially added respiratory substrates (1 mM pyruvate, 5 mM glucose, Sigma), oligomycin (0.1 lg/ml, Sigma) and CCCP (0.5 lM, Sigma). Respiration rate was normalized to the amount of protein in the assay. Mitochondrial membrane potential Mitochondrial membrane potential was measured fluorimetrically with 5,50 ,6,60 -tetrachloro-1,10 ,3,30 -tetraethylbenzimidazolcarbocyanine iodide (JC-1, Molecular Probes, Invitrogen) according to the method of Cossarizza et al. [13]. Confluent cells growing in 24-well culture plates were stained with JC-1 (5 lM final concentration in the culture medium) and incubated for 15 min at 37 °C in the dark. Then, the cells were rinsed three times with the culture medium followed by PBS. Finally, 0.5 ml PBS was added to each well. The cells with completely deenergized mitochondria due to addition of 2 lg/ml of valinomycin plus 5 lM CCCP were used as a control. Fluorescence was measured with a laser scanning cytometer iCYS equipped with argon laser (excitation 488 nm). The data were presented as the ratio of orange (energized) to green (deenergized) fluorescence. Reactive oxygen species ROS were measured with carboxy-20 ,70 -difluorodihydrofluorescein diacetate (DFFH2-DA, Molecular Probes, Invitrogen). Confluent cells grown in 12-well plates were stained with 10 lM DFFH2-DA in PBS for 30 min at 37 °C in the dark. Then, fluorescence was measured at 485 nm excitation and 520 nm emission wavelengths using Infinite 200 micro plate reader (Tecan). Total green fluorescence was normalized to the amount of protein in each well. Mitochondrial mass Mitochondrial mass was estimated fluorimetrically with Mitotracker Green (Molecular Probes, Invitrogen). Confluent cells growing in 24-well plates were stained with this dye at 100 nM concentration in PBS for 20 min at 37 °C. Then, the cells were rinsed three times with the culture medium and then with PBS. Finally, 0.5 ml PBS was added to each well. Fluorescence was measured with laser scanning cytometer iCYS equipped with argon laser (excitation 488 nm). Total green fluorescence was normalized to the amount of protein in each well. Cell lysis and Western blot analysis Cellular lysates were prepared as previously described [14]. Lysate aliquots containing 25–50 lg protein were loaded on 10% or 12% SDS–polyacrylamide gels, depending on the molecular mass of the protein to be investigated, electrophoresed and then transferred onto nitrocellulose membrane. The membranes were treated with specific primary antibodies listed below, as recommended by the manufacturer. All secondary antibodies conjugated with horseradish peroxidase were obtained from Abcam (1:5000). Western blots were developed with chemiluminescent substrate Luminata Classico (Millipore). The content of specific
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
9
proteins was determined densitometrically. The following primary antibodies were applied: VCAM-1 (Santa Cruz, 1:200); VDAC (Santa Cruz, 1:500); TOM20 (Santa Cruz, 1:2000); Hsp60 (Sigma, 1:400); UCP2 (Abcam, 1:1000); MitoProfile total OXPHOS human WB antibody cocktail (Mito Sciences, 1:250); b-actin (monoclonal anti-b-actin antibody peroxidase conjugated, 1:50,000, Sigma). RNA isolation and RT-PCR analysis To estimate UCP2 mRNA level in HUVEC cells, total RNA from cultured cells was isolated using TRIzol reagent (Sigma). RNA concentration and purity were measured spectrometrically as absorbance ratio 260 nm/280 nm and 1 lg of RNA was subjected to reverse transcription with M-MLV enzyme (Promega) and oligodT starters (Sigma). Real-time PCR was performed using 10% of the reverse transcription reaction volume and SYBR Green Master Mix (Applied Biosystems) on a 7500 ABI Prism Real-time PCR System (Applied Biosystems). The primers for UCP2 (NM_003355) were as follows: forward 50 -CAGTTCTACACCAAGGGCTC (within exon 2) and reverse 50 -GGGAGGTCATCTGTCATGAG (within exon 4); b-actin (NM_001101) was used as a reference gene, the primers were: forward 50 -CATGTACGTTGCTATCCAGGC and reverse 50 -CTCC TTAATGTCACGCACGAT. Cycling conditions were 2 min (50 °C), followed by 10 min (95 °C) and then 45 cycles of 15 s (95 °C) and 1 min (60 °C). Products of the PCR reaction were separated in a 2% agarose gel. After normalization to b-actin the 2DDCT method was used to determine the relative mRNA level [15]. Immunocytochemistry To visualise mitochondrial network and cytoskeleton architecture HUVEC cells grown on glass coverslips were stained with MitoTracker CMX Ros (Invitrogen, San Diego, CA, USA), DAPI and FITC-labelled phalloidin (Molecular Probes), as described previously [14]. Confocal fluorescence microscopy was carried out using a Leica TCS SP5 IL Spectral Confocal and Multiphoton Microscope. Expression of results Data are shown as means of ratios of treatment (experimental group) to control values ± SD, for the number of replicates indicated in the figure legends. One-sample t-test was performed on these ratios between analysed sample and control to test for deviation from unity and associated 95% confidence intervals (CIs) were computed. Since normality assumption cannot be strictly met and positive skewness may be expected, analogous computations were repeated for log-transformed values. The resulting p-values and confidence intervals were usually very similar for the both methods of calculation and we refer to the higher (conservative) p-values and wider confidence limits. Results and discussion Prolonged exposure of vascular endothelial cells to glucose at a concentration exceeding 10 mM is commonly applied as a procedure resembling hyperglycaemic conditions. Here, HUVEC cells challenged with 30 mM glucose for 48 h were employed to test the mitochondrial response to hyperglycaemia. Such treatment does not affect cellular viability, as determined with propidium iodide staining of the cells (not shown). As shown in Fig. 1A, incubation of HUVEC cells with 30 mM glucose for 48 h resulted in a substantially increased content of VCAM-1 protein (1.92 ± 0.98-fold). Similarly, the level of intracellular adhesive molecule-1 (ICAM-1) protein was also elevated (1.27 ± 0.34-fold, not shown). A shorter (for 24 h) exposure of the
Fig. 1. Effect of 30 mM glucose on vascular adhesive molecule 1 abundance and ROS formation in HUVEC cells. Cells were grown in standard medium supplemented with 30 mM glucose. (A) Western blot of VCAM-1 protein (one typical experiment out of twelve) for control cells (exposed to normal, 5 mM glucose) and cells treated with 30 mM glucose. Data on the right show the mean value ± SD for twelve experiments. VCAM-1 protein level in control cells was taken as 1. p < 0.008, 95% Cl: [1.27, 2.55]. (B) Effect of NADPH oxidase inhibitors (DPI and VAS2870) on ROS formation in cells treated with 30 mM glucose for 6 or 48 h. Data from ten independent experiments expressed as percentage of control (cells exposed to normal, 5 mM glucose). 95% Cl: [0.68, 0.86] for 30 mM glucose/DPI and [0.59, 0.96] for 30 mM glucose/VAS. (C) Effect of oxidative phosphorylation uncoupler (CCCP) on ROS generation in cells treated with 30 mM glucose. Data from five independent experiments. Confidence limit for 30 mM glucose/CCCP/DPI was [0.67, 0.88]. (D) Effects of rotenone and DPI on ROS generation in control cells. Data from five independent experiments. 95% Cl: [0.85, 0.99] for DPI, [1.01, 1.21] for rotenone, [0.83,1.20] for rotenone/DPI. p-values for data shown in panels B–D: ⁄p < 0.0002, ⁄⁄ p < 0.04 comparing to control; #p < 0.0004, ##p < 0.04 comparing to 30 mM glucose. Grey lines at the level of 1.00 define values for control cells without any additions. The first two bars on the left of graphs are identical for panels B and C.
cells to 30 mM glucose resulted in a much less substantial inflammatory response. At the same time, a significant enhancement of ROS formation was observed (Fig. 1 B and C), as early as after 6 h of treatment with high concentration of glucose, thus prior to
10
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
any increase in the VCAM-1 protein level. These effects confirm the previously described response of endothelial cells to glucose at high concentration and are considered typical for endothelial cell dysfunction [16]. They also may participate in an impairment of the endothelial function leading to severe complications. It has previously been shown that an excessive ROS formation is responsible for the deleterious effect of hyperglycaemia in HUVECs both in the case of intermittent as well as constant elevation of glucose concentration [17]. Although the influence of ROS on endothelial cell functioning is unquestionable, the source of the excessive ROS formation within the cells challenged by hyperglycaemia has not been unequivocally established so far [18]. Data shown in Fig. 1 provide, to some extent, a solution to this question. Preincubation of HUVEC cells with 100 nM diphenyleneiodonium (DPI, a NOX inhibitor) for 1 h prior to ROS measurement led to a substantial reduction of ROS formation in cells exposed to 30 mM glucose for 6 or 48 h (Fig. 1B). To confirm this observation and avoid any risk of DPI-evoked side effects resulting from its limited specificity, similar experiments with the use of the more specific NADPH oxidase inhibitor VAS2870 [19] were undertaken. Indeed, this compound applied at 5 lM concentration for 1 h prior to ROS measurement significantly reduced the ROS generation in HUVEC cells exposed to 30 mM glucose for 48 h. This convincingly documents participation of NOX in the ROS formation induced by excessive glucose. On the other hand, a complete dissipation of the mitochondrial membrane potential with CCCP, which usually decreases mitochondrial ROS generation, resulted in only a slight and statistically insignificant reduction of ROS formation in the cells, independently of the glucose concentration (data not shown). However, an application of this uncoupler together with DPI in the cells exposed to 30 mM glucose brought down the ROS production to the level found in the control (normal glucose) cells (Fig. 1C). Furthermore, an inhibition of complex I of the respiratory chain by rotenone resulted in a reproducible enhancement of ROS generation in control cells (Fig. 1D). Similar effects were observed after addition of antimycin A, a complex III inhibitor (data not shown). The slight decrease in ROS formation in control cells treated with DPI alone may indicate some basal NOX-dependent ROS generation under control (normoglycaemic) conditions. On the other hand, an addition of DPI did not significantly prevent the rotenone action, indicating that DPI does not interfere with the respiratory chainoriginated ROS. Interestingly, rotenone and CCCP prevented an excessive ROS generation in BAEC cells exposed to 30 mM glucose [20]. This discrepancy probably reflects different metabolic properties of endothelial cells derived from various sources. In sum, the results presented hitherto suggest that NADPH oxidase is the major source of ROS generated by HUVEC cells exposed to 30 mM glucose. This is in agreement with the data obtained for bovine aortic endothelial cells, in which accelerated ROS formation was related to PKC-dependent activation of NOX [21]. Our preliminary results seem to indicate an increase of the proportion of NADPH oxidase isoform 4 (NOX4) in the purified mitochondrial fraction from HUVEC cells grown under hyperglycaemic conditions while the total amount of this enzyme found in a cellular homogenate was unchanged (data not shown). One may thus speculate that NOX4 may be recruited to mitochondria under hyperglycaemic conditions. Thus, in HUVEC cells exposed to glucose at high concentration ROS may also be formed in the mitochondria but without an involvement of the mitochondrial respiratory chain. However, the identification of the particular isoform of NOX responsible for the excessive ROS generation in HUVEC cells under hyperglycaemic conditions is beyond the scope of this article. Although the aforementioned data clearly show that the enhanced ROS generation in HUVEC cells grown in the presence of 30 mM glucose is due to NADPH oxidase activity, the substantial changes in the mitochondrial metabolism occurring concomitantly
cannot be ignored. In fact, as shown in Fig. 2A, the oxygen consumption by cells exposed to hyperglycaemic conditions is considerably enhanced. This is observed when cells utilize endogenous substrates (basal respiration), also after addition of oligomycin and a mitochondrial uncoupler (CCCP). The basal respiration rate did not change after addition of external respiratory substrates, indicating that intracellular substrates are sufficiently abundant to maintain oxygen consumption at the unchanged level. The similar degree of respiratory stimulation by high glucose treatment under all conditions tested (basal, inhibited ATP synthase, uncoupled oxidative phosphorylation) suggests an increase of the total respiratory capacity in these cells rather than partial dissipation of the mitochondrial membrane potential. This could be a consequence of an improved efficiency of the respiratory chain and/or an increased amount of mitochondria. The increased abundance of the components of the respiratory chain in cells exposed to 30 mM glucose shown in Fig. 2B supports the latter possibility. On the other hand, a lack of an effect of hyperglycaemia on the NRF-1 protein content (not shown) seems to exclude increased mitochondrial biogenesis under hyperglycaemic conditions. Therefore our results allowed assuming that glucose at high concentration triggers bioenergetic reserves as important factor in response to stress in HUVECs as it was indicated earlier [22]. The demonstrated respiratory stimulation by glucose treatment
Fig. 2. Effect of 30 mM glucose on oxygen consumption and the abundance of respiratory chain components in HUVEC cells. (A) Effect of hyperglycaemia on oxygen consumption. Absolute rate of oxygen consumption (pmol O2/s per mg protein) is shown in the right part of panel A. Values for control cells treated with the same inhibitor or uncoupler as cells exposed to 30 mM glucose were taken as 100%. Data for six independent experiments. ⁄p < 0.009 (lower confidence limits for basal 108%, for oligomycin 117%, for CCCP 106%). Substrates and inhibitors were added sequentially directly to the incubation chamber as indicated, at final concentrations: pyruvate 1 mM, glucose 5 mM, oligomycin 0.1 lg/ml, CCCP 0.5 lM. (B) Western blot analysis of selected respiratory chain components. Protein level of respiratory complexes II, III, IV and V was normalized to b-actin and is expressed as percentage of control. Data from three independent experiments. ⁄ p < 0.05 (upper confidence limits: for complex II-166%; for complex III-149%, for complex IV-175%, for complex V-117%). Grey lines at the level of 100% define values for control cells.
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
of HUVECs seems to resemble the response of BAEC cells to both ROS and RNS. Dranka and co-workers found that unaffected cells use only about 35% of their maximal respiratory capacity, a state intermediate between states 3 and 4, indicating the presence of a reserve respiratory capacity available for the cells when the bioenergetic demand is increased [22]. The hypothetical effect of hyperglycaemia on mitochondrial biogenesis was tested with the use of MitoTracker Green. As shown in Fig. 3A, the elevated glucose concentration does not alter the total mitochondrial mass, in line with the lack of an effect on the level of NRF-1 protein. This observation was also confirmed by a Western blot analysis of the abundance of the following marker proteins: heat shock protein 60 (HSP60), voltage-dependent anion channel (VDAC) and translocase of outer membrane 20 kDa subunit (TOM20). As shown in Fig. 3A, exposition of HUVECs to 30 mM did not change the level of these proteins. On the other
11
hand, hyperglycaemia affected the mitochondrial network architecture by increasing the proportion of fragmented vs. tubular organelles (Fig. 3B), which could correlate with an affected mitochondrial metabolism. Similar results have been observed in bovine retinal pericytes and rat retinal endothelial cells growing in the presence of glucose at high concentration. Trudeau et al. [23,24] observed that fragmentation of mitochondrial network was accompanied by a decreased mitochondrial membrane potential and reduced extracellular acidification. In retinocytes similar reorganization of the mitochondrial network was also associated with accelerated apoptosis and, consequently, was postulated as a cause of diabetic retinopathy. In addition, Makino and co-workers [25] suggested that superoxide anion radical might serve as a direct regulator of mitochondrial morphology in mouse coronary endothelial cells. Overproduction of ROS led to mitochondrial fragmentation and a treatment with antioxidative agents was
Fig. 3. Effect of 30 mM glucose on mitochondrial mass, and structure and abundance of mitochondrial markers. (A) Total mitochondrial mass estimated with MitoTracker Green and the amount of mitochondrial proteins HSP60, VDAC and TOM20. Data for MitoTracker Green staining for five independent experiments, 95% Cl: [0.89, 1.04]. Blots show results of two independent experiments out of six. C indicates control cells, HG – cells exposed to 30 mM glucose for 48 h. B, Mitochondrial organization: green, actin cytoskeleton stained with FITC-labelled phalloidin; red, mitochondria stained with MitoTracker CMX Ros; blue, nucleus stained with DAPI. All images are taken in the same confocal plane. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
12
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
proposed to help improve mitochondrial function. Reorganization of the mitochondrial network could reflect a change in cellular bioenergetics. Previously it has been shown that undisturbed mitochondrial membrane potential together with efficient ATP synthesis by oxidative phosphorylation and glycolysis, as well as unchanged ROS level, are required to maintain the balance between mitochondrial fission and fusion in HUVECs. In depolarized mitochondria, as well as after exposure to oxidative stress, fragmentation of the tubular network was observed [26]. Considering the above-mentioned link between the mitochondrial architecture and cellular metabolism, it seems likely that the changes in the mitochondrial network organization and the increased oxygen consumption in HUVEC cells exposed to 30 mM glucose could also be accompanied by a decrease of the mitochondrial membrane potential. In fact, we found a slight reduction of this parameter, with confidence interval between 0.57 and 1.08 relative to the control (Fig. 4A). In fact, the lack of linearity between the acceleration of oxygen consumption and the simultaneous reduction of mitochondrial membrane potential [27] speaks against a substantial mitochondrial deenergization in view of the relatively small (ca. 20%) stimulation of respiration observed here. This slight decrease of Dw under hyperglycaemic conditions shown here is in disagreement with the observations published previously by Rolo and Palmeira [28]. These authors pointed out that the hyperglycaemia-induced hyperpolarization drives the mitochondrial ROS formation which were decreased after mitochondrial uncoupling with CCCP. As has already been mentioned, CCCP did not influence ROS formation in HUVECs exposed to hyperglycaemic conditions (see Fig. 1C). Thus our results indicate the generation of ROS in the mitochondrial respiratory chain independent mechanism. A similar result was observed in retinal endothelial cells, where a heterogeneity of mitochondrial potential was
postulated and associated with increased cellular stress and appeared to reflect variations of mitochondrial responses. Several factors causing such a diversity of mitochondrial transmembrane potential were postulated, e.g., heterogeneous distribution of respiratory chain components, uneven access of metabolites and calcium, and/or changes in mitochondrial localization [23]. The slightly decreased mitochondrial membrane potential is in line with an increased amount of UCP2, as shown by Western blot data (Fig. 4B and C), probably responsible for dissipating the mitochondrial membrane potential [29]. The enhanced oxygen consumption (Fig. 2A) can also be partially explained by the increased UCP2 abundance. Giardina and co-workers [30] showed that mitochondria-originated oxidative stress may up-regulate the UCP2 content in macrophages, but oxidants acting outside mitochondria did not affect the amount of this protein. In macrophages treated with mitochondrial inhibitors (rotenone, antimycin A and diethyldithiocarbamate) to raise mitochondrial superoxide production, a rapid increase of UCP2 level was observed. Those authors found no such effect in response to superoxide produced outside mitochondria or in response to H2O2. Our results seem to be in disagreement with their data unless one of NOX isoforms is localised to mitochondria. We found that the increased UCP2 content is not accompanied by a parallel change of its mRNA level (Fig. 4B and C). This suggests that glucose at high concentration enhances the translational efficiency of the mRNA and possibly also stabilizes UCP2 in the inner mitochondrial membrane, as it was postulated earlier [30]. An increased abundance of UCP2 could be essential for reduction of oxidative stress and appears to depend on AMP-activated protein kinase. AMPK has been suggested to be a physiological regulator of ROS formation [31]. It can also be assumed that an increase in UCP2 level implies increased oxygen consumption. It also correlates with the observed slightly
Fig. 4. Effect of 30 mM glucose on mitochondrial membrane potential and UCP2 content. (A) Orange/green fluorescence ratio in cells loaded with JC-1. In control cells this is taken as 1. Data from seven independent experiments, 95%Cl: [0.57, 1.08]. (B) UCP2 protein level and mRNA content, both normalized to b-actin protein and transcript, respectively. Data expressed as percentage of control, derived from seven independent experiments for Western blot and three independent experiments for real time RTPCR. ⁄p < 0.008, 95%Cl: [1.20, 2.16]. (C) Western blot for one typical experiment out of seven and analysis of RT-PCR products of UCP2 mRNA for one typical experiment out of three are shown.
D. Dymkowska et al. / Archives of Biochemistry and Biophysics 542 (2014) 7–13
decreased mitochondrial membrane potential. It cannot be excluded that UCP2 counteracts mitochondrial ROS generation. Elevated glucose concentration leads to endothelial cell dysfunction and affects mitochondrial metabolism. It seems that an increased abundance of UCP2 is crucial for the modulation of mitochondrial metabolism, as described earlier by Lowell et al. [32]. UCPs and particularly UCP2 have been shown as possible regulators of the production of mitochondrial oxidants. This protein acts by modulating the proton leak across the inner membrane, but the antioxidative action of UCP2 is independent of its effect on the mitochondrial membrane potential. It has been suggested that hyperglycaemia may increase cellular reactive oxygen species production and the abundance and activity of the inner mitochondrial membrane UCP2 [7]. In conclusion, our results indicate that exposition of HUVEC cells to a high concentration of glucose seriously affects mitochondrial energy metabolism, but these changes do not seem to be directly responsible for the observed accelerated ROS generation. This latter event is rather caused by an increased NADPH oxidase activity. Acknowledgments We thank Professor Lech Wojtczak for critical reading of the manuscript and Dr. Tomasz Wyszomirski for help in statistical data analysis. This work was supported by the National Science Centre grant N N301 291137 to D.D. References [1] M.A. Potenza, S. Gagliardi, C. Nacci, M.R. Carratu, M. Montagnani, Curr. Med. Chem. 16 (2009) 94–112. [2] C.K. Sen, L. Packer, FASEB J. 10 (1996) 709–720. [3] N. Bakkar, J. Wang, K.J. Ladner, H. Wang, J.M. Dahlman, M. Carathers, S. Acharyya, M.A. Rudnicki, A.D. Hollenbach, D.C. Guttridge, J. Cell Biol. 180 (2008) 787–802. [4] S. Srinivasan, M.E. Hatley, D.T. Bolick, L.A. Palmer, D. Edelstein, M. Brownlee, C.C. Hedrick, Diabetologia 47 (2004) 1727–1734.
13
[5] C.E. Murdoch, M. Zhang, A.C. Cave, A.M. Shah, Cardiovasc. Res. 71 (2006) 208– 215. [6] I. Al Ghouleh, N.K.H. Khoo, U.G. Knaus, K.K. Griendling, R.M. Touyz, V.J. Thannickal, A. Barchowsky, W.M. Nauseef, E.E. Kelley, P.M. Bauer, V. Darley-Usmar, S. Shiva, E. Cifuentes-Pagano, B.A. Freeman, M.T. Gladwin, P.J., Free Radical Biol. Med. 51 (2011) 1271–1288. [7] S.W. Ballinger, Free Radical Biol. Med. 38 (2005) 1278–1295. [8] X.Y. Tian, W.T. Wong, A. Xu, Y. Lu, Y. Zhang, L. Wang, W.S. Cheang, Y. Wang, X. Yao, Y. Huang, Circ. Res. 110 (2012) 1211–1216. [9] G.X. Shen, Can. J. Physiol. Pharmacol. 88 (2010) 241–248. [10] C. Quijano, L. Castro, G. Peluffo, V. Valez, R. Radi, Am. J. Physiol. Heart Circ. Physiol. 293 (2007) H3404–H3414. [11] S. Cai, J. Khoo, K.M. Channon, Cardiovasc. Res. 65 (2005) 823–831. [12] A. Koziel, A. Woyda-Ploszczyca, A. Kicinska, W. Jarmuszkiewicz, Pfluegers Arch. 464 (2012) 657–669. [13] A. Cossarizza, M. Baccarani-Contri, G. Kalashnikova, C. Franceschi, Biochem. Biophys. Res. Commun. 197 (1993) 40–45. [14] B. Drabarek, D. Dymkowska, J. Szczepanowska, K. Zabłocki, Int. J. Biochem. Cell Biol. 44 (2012) 1390–1397. [15] K.J. Livak, T.D. Schmittgen, Methods 25 (2001) 402–408. [16] D. Popov, Int. J. Diabetes Mellit. 2 (2010) 189–195. [17] L. Piconi, L. Quagliaro, R. Assaloni, R. Da Ros, A. Maier, G. Zuodar, A. Ceriello, Diabetes Metab. Res. Rev. 22 (2006) 198–203. [18] M.M. Spitaler, W.F. Graier, Diabetologia 45 (2002) 476–494. [19] S. Wind, K. Beuerlein, T. Eucker, H. Müller, P. Scheurer, M.E. Armitage, H. Ho, H.H.H.W. Schmidt, K. Wingler, Br. J. Pharmacol. 161 (2010) 885–898. [20] T. Nishikawa, D. Edelstein, X.L. Du, S. Yamagishi, T. Matsumura, Y. Kaneda, M.A. Yorek, D. Beebe, P.J. Oates, H.P. Hammes, I. Giardino, M. Brownlee, Nature 404 (2000) 787–790. [21] T. Inoguchi, P. Li, F. Umeda, H.Y. Yu, M. Kakimoto, M. Imamura, T. Aoki, T. Etoh, T. Hashimoto, M. Naruse, H. Sano, H. Utsumi, H. Nawata, Diabetes 49 (2000) 1939–1945. [22] B.P. Dranka, B.G. Hill, V.M. Darley-Usmar, Free Radical Biol. Med. 48 (2010) 905–914. [23] K. Trudeau, A.J.A. Molina, W. Guo, S. Roy, Am. J. Pathol. 177 (2010) 447–455. [24] K. Trudeau, A.J.A. Molina, S. Roy, Invest. Ophthalmol. Vis. Sci. 52 (2011) 8657– 8664. [25] A. Makino, B.T. Scott, W.H. Dillmann, Diabetologia 53 (2010) 1783–1794. [26] R.J. Giedt, D.R. Pfeiffer, A. Matzavinos, Ann. Biomed. Eng. 40 (2012) 1903–1916. [27] L. Wojtczak, K. Bogucka, J. Duszyn´ski, B. Zabłocka, A. Zółkiewska, Biochim. Biophys. Acta 1018 (1990) 177–181. [28] A.P. Rolo, C.M. Palmeira, Toxicol. Appl. Pharmacol. 212 (2006) 167–178. [29] B.D. Fink, Y.S. Hong, M.M. Mathahs, T.D. Scholz, J.S. Dillon, W.I. Sivitz, J. Biol. Chem. 277 (2002) 3918–3925. [30] T.M. Giardina, J.H. Steer, S.Z. Lo, D.A. Joyce, Biochim. Biophys. Acta 1777 (2008) 118–129. [31] Z. Xie, J. Zhang, J. Wu, B. Viollet, M.H. Zou, Diabetes 57 (2008) 3222–3230. [32] B.B. Lowell, G.I. Shulman, Science 307 (2005) 384–387.