Isoflurane Pretreatment Inhibits Cytokine-induced Cell Death in Cultured Rat Smooth Muscle Cells and Human Endothelial Cells

Rat aorta VSMs were donated by Dr. L. Palmer, Ph.D. (Associate Professor, Departments of Anesthesiology and Pediatrics, University of Virginia Health System, Charlottesville, Virginia). The HUVECs were a gift from Dr. Tom G. Obrig, Ph.D. (Department of Internal Medicine and Immunology, University of Virginia Health System). The SMCs were grown in Dulbecco's minimal essential growth medium (Gibco, Rockville, MD) supplemented with 10% fetal bovine serum (FBS). HUVECs were grown in medium bought from the Roswell Park Memorial Institute (Buffalo, New York), supplemented with 50 μg/ml endothelial mitogen (Biomedical Technologies, Fishers, IN), 29.2 mg/ml l-glutamine, 76 μg/ml heparin (Sigma, St. Louis, MO), and 20% FBS. The cytokines TNF-α, IFN-γ, and IL-1β were purchased from R&D Systems (Minneapolis, MN) and diluted into phosphate-buffered saline (PBS; Gibco) containing 0.1% bovine serum albumin (BSA; Sigma). The cells, passages 5–9, were plated at a density of 4 × 10⁵/ml.

The initial experimental groups were pretreated with 1.5% isoflurane for 30 min in an airtight chamber before incubation with the cytokines at 37°C for periods of 24, 48, 72, or 96 h. Assessment was not possible for HUVECs at 96 h because of the need to change the medium for control cell survival. Isoflurane, 1.5%, was administered with an agent-specific vaporizer using a gas mixture of 95% O²and 5% CO²at a flow rate of 5 l/min. The isoflurane concentration in the chamber was measured with a gas analyzer (Datex Ohmeda, Finland). When the concentration remained constant at 1.5%, the chamber inlet and outlet were closed, and the chamber was placed in the incubator at 37°C for 30 min. The medium was then replaced with 1 ml fresh medium in which the cytokines were dissolved at the following concentrations: 0.1 ng/ml TNF-α, 5.0 ng/ml IFN-γ, and 5.0 ng/ml IL-1β. The cultures were incubated with no cytokines (negative control), one cytokine, or combinations of two or three cytokines. After the different incubation periods (24, 48, 72, or 96 h), the cells were washed twice with PBS (Gibco) and were then further treated according to the specific assay protocols

Dead cells are incapable of excluding trypan blue; therefore, the uptake of this dye was used as a marker of cell death. Trypan blue (10 μl) was mixed with 50 μl cell suspension. For each sample, at least 100 cells per field were counted with a hemocytometer under a light microscope. Cell survival is the number of total cells minus the number of dead cells.

Cell viability was determined by proliferation assay (Chemicon International, Temecula, CA) and is based on cleavage of the tetrazolium salt WST-1 to formazan by cellular mitochondrial dehydrogenase as described by Kcrzeniewski et al.  28Briefly, 100 μl of each sample was incubated with 10 μl of assay reagent in a 96-well plate for 2 h. The absorbance was measured with a spectrometer using a 450-nm filter. The viability of each sample was calculated as percentage of the absorbance of the sample cultured without cytokines.

An apoptag plus peroxidase detection kit (Intergren, Purchase, NY) was used to detect DNA fragmentation, by means of nick-end labeling (TUNEL) fragmented DNA with biotinylated deoxynucleotide triphosphate (dUTP) as previously described by Wang et al.  29Briefly, the cells were washed and treated with 20 μg/ml protein kinase K (Sigma). Endogenous peroxide was blocked by covering the sections with 3% H²O²(Sigma). The biotinylated dUTP was applied to the slides, which then were incubated in a humidified chamber at 37°C for 1 h. The labeled DNA fragments were then stained with an antidigoxigenin monoclonal antibody, and the cells were counterstained by immersion of the slides in 0.5% methyl green (Sigma). For each sample, two slides were made, and the total of stained cells were counted.

The enzyme-linked immunosorbent assay kit (Roche Molecular Biochemicals, Mannhein, Germany) was used to characterize cell death by measurement of 5′-Bromo-2′-deoxy-Uridine (BrdU)–labeled DNA fragments in culture supernatants and cell lysate. Proliferating VSMs were incubated with BrdU, which is incorporated into the genomic DNA. The cells were exposed to all three cytokines or H²O²(1 mm) for periods of 2, 4, 6, and 8 h. Aliquots (100 μl) from the supernatant of each sample were transferred to an anti-BrdU antibody–coated 96-well plate. The remaining cells were lysed with a solution containing ethylenediaminetraacetic acid, BSA, and poloxyethylenesorbitan. The lysate was also transferred into a BrdU antibody–coated well plate. After incubation for 90 min at room temperature, the samples in both plates were heated by microwave irradiation to denature DNA. Anti-BrdU peroxidase conjugate was added followed by a 90-min incubation period at room temperature. The samples were incubated on a shaker in darkness with 3.3′-5,5′-tetra-methyl-benzidine until fluorescence developed. This reaction was stopped by 5.5% sulfuric acid, and the absorbance was measured with a spectrometer at 450 nm.

Cell counting was performed in a blinded manner. Comparisons between groups (n = 6) were made with an analysis of variance and a Tukey post hoc  test. Statistical analysis was performed with SigmaStat 2.0 (Jandel Scientific Software, San Rafael, CA). Data are presented as mean ± SD. P < 0.05 was considered significant.

The objective of the first experiment was to determine the effect of 30 min of pretreatment with 1.5% isoflurane on cell survival and viability. Isoflurane, 1.5%, was administered for 30 min in VSM and HUVEC cultures immediately prior to replacing the media with media in which TNF-α, IFN-γ, and IL-1β alone, in combinations of two, and all three together, were dissolved. Control cultures not exposed to isoflurane or cytokines showed less than 10% cell death and greater than 90% viability over 72 h for HUVECs and 96 h for VSMs. Isoflurane had no effect on cell viability or cell survival in cultures not exposed to cytokines. Cytokines decreased cell viability and cell survival with all combinations of cytokines and for all time periods. The effects of the cytokines were increased at longer time periods and in the presence of greater combinations of cytokines.

In VSMs, pretreatment with 1.5% isoflurane for 30 min immediately prior to cytokine increased cell survival and increased cell viability for all multiple combinations of cytokines at all time periods by 25–60%. In single cytokine experiments using VSMs, isoflurane pretreatment increased cell survival at all time points except for IL-1β at 24 h, TNF-α at 48 h, and IFN-γ at 72 h and increased cell viability for all time points except TNF-α alone and IL-1β alone at 24 h and TNF-α at 48 h. In HUVECs, isoflurane pretreatment increased cell survival in the presence of IFN-γ, IL-1β, and all three cytokines in combination at 24 h, for all cytokine combinations at 48 h, and all combinations except TNF-α alone at 72 h by 10–45%. Cell viability was increased for all combinations except IL-1β at 72 h. Isoflurane protection is illustrated at 96 h for VSMs (fig. 1) and 72 h for HUVECs (fig. 2).

Because of similar results with most parameters in the initial experiment, the remaining experiments were evaluated only in the presence of all three cytokines (TNF-α, IFN-γ, and IL-1β) and to the time periods at 24 and 72 h. To determine the dose–response effects of isoflurane pretreatment, VSMs and HUVECs were pretreated with 0.75, 1.5, or 2.5% isoflurane for 30 min prior to cytokines.

In VSMs, there was no significant difference between the effects of 0.75, 1.5, and 2.5% isoflurane pretreatment on cell viability and cell survival at either 24 or 72 h. In HUVECs, pretreatment with 0.75% isoflurane had significantly less effect on cell survival and viability compared with 1.5%; however, the effect of 0.75% was greater than no isoflurane pretreatment. The effects of 1.5% and 2.5% isoflurane were equal at both 24 and 72 h (fig. 3, survival only).

To determine whether the timing of 1.5% isoflurane administration either prior to or immediately after cytokines was critical, the time between the end of the isoflurane pretreatment and the cytokine exposure was extended to 1, 2, and 4 h in VSMs. It was also determined whether isoflurane treatment was effective if administered at the same time or when initiated 30 min after cytokine exposure.

In VSMs, cell survival and viability were the same whether 1.5% isoflurane pretreatment was for 30 min immediately prior to the cytokines or whether the time between the end of the 30 min of isoflurane pretreatment and the initiation of cytokine exposure was extended to 1 h. After a 2-h delay, the protective effect of isoflurane was less than after a 1-h delay but still significantly greater than no pretreatment. After a 4-h delay, the isoflurane pretreatment had no protective effect on cell survival or viability.

When exposure to isoflurane and the cytokines were initiated simultaneously, the effects on cell survival and viability were the same as with the 30-min isoflurane treatment prior to cytokine exposure. In contrast, isoflurane did not affect cell survival or viability when it was initiated 30 min after cytokine exposure. Results for these timed experiments were the same for 24 and 72 h (fig. 4, 72-h survival only).

To determine whether halothane has similar effects as isoflurane, VSMs were exposed to 1.5% isoflurane or 1.2% halothane for 30 min prior to cytokine exposure. To determine whether isoflurane protection and halothane protection were similar in H²O²-induced cell death, pretreatment for 30 min was evaluated in VSMs containing H²O²(1 mm).

In cytokines, 1.2% halothane significantly increased VSM cell survival and increased cell viability to the same extent as did isoflurane at 24 and 72 h (data not shown). H²O²significantly decreased cell survival and viability. In H²O², 1.5% isoflurane and 1.2% halothane had the same protective effect on cell survival and viability at 24 and 72 h (fig. 5, 72-h survival only).

We investigated whether glibenclamide, which can inhibit K^ATPchannels, and staurosporine, which can inhibit PKCs, alter the effects of anesthetic pretreatment in VSMs and HUVECs. Pretreatment experiments with 1.5% isoflurane were performed in the presence of glibenclamide (10 μm) or staurosporine (4 nm) in cultures exposed to all three cytokines or H²O²(1 mm).

In VSMs and HUVECs, glibenclamide and staurosporine had no effect on cell survival or viability in cells not exposed to cytokines or H²O²in the presence or absence of isoflurane. Conversely, the pretreatment effects of isoflurane in VSM exposed to cytokines were attenuated by glibenclamide and staurosporine at 24 and 72 h. In HUVECs, glibenclamide and staurosporine completely abolished the pretreatment effect of isoflurane at 24 and 72 h (fig. 6, 72 h only) Likewise, the effects of isoflurane and halothane on cell survival and viability were abolished with glibenclamide and staurosporine in H²O²-induced cell death at both 24 and 72 h (fig. 5, 72-h survival only).

A TUNEL assay was used to evaluate the effects of 1.5% isoflurane pretreatment on DNA fragmentation in cytokine experiments. VSMs and HUVECs were evaluated after 72 h. Cytokines increased TUNEL staining by 78 ± 1% in HUVECs and 65 ± 1% in SMCs after 72 h. Isoflurane pretreatment almost completely abolished staining so that it only measured 8 ± 1% in SMCs and 7 ± 1% in HUVECs.

DNA fragmentation assay was measured every 2 h up to 8 h to determine whether there were differences between cytokines and H²O². In VSMs, cytokines increased lysate DNA fragmentation by 2.5-fold after 8 h but did not alter supernatant DNA fragmentation compared to controls, which is suggestive of apoptosis. In contrast, H²O²increased supernatant DNA fragmentation by 3.8-fold after 8 h but did not alter lysate DNA fragmentation, which is suggestive of necrosis.

Previous in vivo  studies indicated that anesthetic pretreatment protects the myocardium from ischemia–reperfusion injury, and in vitro  studies indicated that continuous volatile anesthetic administration attenuates apoptosis and cell death. 21–23Thus, we hypothesized that isoflurane pretreatment would protect VSMs and HUVECs from cytokine-induced cell death and increase cell viability. In this study, we showed that 30 min of pretreatment with 1.5% isoflurane increased cell survival and cell viability due to cytokine-induced injury for at least 96 h in VSMs and 72 h in HUVECs. Pretreatment with isoflurane and halothane also increased VSM and HUVEC cell survival and viability for at least 72 h after exposure to H²O².

The protective effects of isoflurane pretreatment from cytokine-induced cell death were evaluated at three different concentrations ranging from 0.75 to 2.5%. There was no significant difference in the protective effect among all three concentrations in VSMs; however, in HUVECs 1.5% and 2.5% provided greater protection than 0.75%. Lui et al.  27demonstrated that pretreatment with concentrations as low as 0.2% isoflurane were sufficient to protect cultured chick myocytes from ischemia–reperfusion injury. While we did not evaluate isoflurane concentrations below 0.75%, it is possible that similarly low concentrations may be protective in our model with VSMs. On the other hand, the decreased effect of 0.75% in HUVECs indicates that low concentrations may not be protective in all cell types. Our study indicates that concentrations above 1.5% may not offer additional protective effects. In fact, Johnson et al.  22reported that supraclinical concentration of isoflurane (4%) and halothane (2.7%) afforded less protection than clinical concentration during H²O²-induced cell death. Clinically relevant concentrations of volatile anesthetics also are protective in animal models of myocardial ischemia–reperfusion injury, 24suggesting a close correlation between in vivo  and in vitro  results. Possible clinical relevance is further supported by evidence that clinical concentrations of isoflurane may decrease myocardial damage in humans undergoing coronary artery bypass surgery. 21 

Protection of VSMs by isoflurane (1.5%) and halothane (1.2%) pretreatment in cytokine- and H²O²-induced injury were the same. Similarly, Zaugg et al.  23showed that continuous administration of halothane and isoflurane inhibited norepinephrine-induced apoptosis to the same extent. In rat hearts, halothane and isoflurane have also been shown to have similar protective effects on myocardial injury. 30However, other studies have shown that the effects of volatile anesthetics are not similar. Johnson et al.  22indicated that halothane was more protective than isoflurane in H²O²-induced cell death. The opposite result was demonstrated by Roscoe et al. , 31who indicated that isoflurane but not halothane induced protection of cultured human cardiomyocytes. Whether or not there are significant differences in protection from isoflurane and halothane are undoubtedly model and injury dependent; however, our study indicates that pretreatment with each increases cell survival and viability to the same extent in the presence of cytokines and H²O²in VSMs and HUVECs.

The timing of the isoflurane administration was critical for increasing cell survival and viability in VSMs. We administered isoflurane either immediately prior to cytokines; 1, 2, and 4 h prior to cytokines; simultaneously with cytokines; or 30 min after cytokines. As long as the 30 min of isoflurane was completed within 1 h prior to the administration of cytokines or administered at the same time as the cytokines, the maximal protective effects were observed. The protective effect decreases when the time between the completion of isoflurane administration and cytokine exposure is extended to 2 h and is abolished at 4 h. When isoflurane treatment is initiated 30 min after the cytokine exposure, there is also no protective effect. These results indicate several important findings. First, isoflurane must be present at the time of or prior to the initiating events of cytokine-induced cell death. In other words, isoflurane is effective in inhibiting the initiation of the cytokine cascade but is ineffective in halting the cascade once initiated. Second, the mechanisms by which isoflurane prevents the initiation of the cytokine-induced cell death are lost after 2–4 h. Previous studies have not evaluated this time period; however, in dog hearts, anesthetics may be delivered as early as 30 min prior to ischemia. 26The apparent 2- to 4-h window that we demonstrated is particularly interesting because once the cytokine cascade is inhibited, the increase in cell survival and viability lasts at least 96 h. Because the effects of isoflurane appear to be lost after 2–4 h, it is likely that this increase in long-term survival is explained by inhibition of the initiation of the cytokine cascade. The observation that simultaneous administration of isoflurane with cytokine exposure also was effective indicates that the protective mechanisms have a very rapid onset.

Previous in vivo  studies have suggested that stimulation of K^ATPchannels or PKC is the protective mechanism for IPC and anesthetic preconditioning against ischemic injury. 32Blockade of isoflurane and halothane protection by glibenclamide and staurosporine argues for involvement of similar mechanisms against cytokine- and H²O²-induced cell death in cultured VSMs and HUVECs. Similar protective effects of volatile anesthetics and associated mechanisms in our cell culture studies compared to in vivo  studies may not be surprising since ischemic–reperfusion injury is partly mediated by cytokine and H²O²release. 3Importantly, because of the 4-h window, the effects of isoflurane and halothane on PKC and/or K^ATPchannels appear to be transient. The precise mechanisms by which PKC and/or K^ATPchannels may alter cell death are unclear, although the mechanisms appear to be effective against different cytokines alone and in combination and against H²O². Because isoflurane must be present at the initiation of cytokines, factors which are responsible for delayed protection, such as elevated heat shock proteins, are unlikely to be involved in our study. 12,13Whether these conclusions are also true for IPC or anesthetic preconditioning in vivo  remains to be studied.

It is well accepted that K^ATPchannels are involved in the mechanistic pathway of both ischemic and anesthetic preconditioning in in vivo  models of myocardial protection. 24,26Volatile anesthetics offer myocardial protection in coronary ligation models, an effect which is inhibited by blocking the K^ATPchannels. 20Similarly, K^ATPchannels have been shown to be involved in protection of coronary endothelium during ischemia–reperfusion in vivo . 33Initial studies focused on the role of sarcolemmal K^ATPchannel activity in IPC, but since the isolation of a mitochondrial K^ATPchannel, studies have shown that the mitochondrial channel is crucial to the protection elicited by IPC. 18,19,34Ockaili et al.  17reported that direct stimulation of the mitochondrial K^ATPchannel produced protection from ischemic injury in rats while blockade of this channel eliminated the protective effect. We evaluated only the effects of nonspecific K^ATPchannel blockade with glibenclamide, and, although it is likely that mitochondrial K^ATPare more importantly involved, further studies will be required to determine whether the same applies to isoflurane and halothane protection during cytokine- and H²O²-induced cell death.

Activation of PKC appears to be a critical step in preconditioning because the phenomenon can be induced by PKC activators or blocked by PKC inhibitors. Wang et al.  10have shown that stimulation of PKC provides protection against Ca²⁺overload injury in isolated rat hearts. Zhao et al.  35demonstrated that transvected, basally active PKC-δ conferred protection from simulated ischemic injury in cell culture. Blockade of PKC abolishes the protective effects on various forms of preconditioning, including Ca²⁺preconditioning and direct activation of mitochondrial K^ATPchannels in isolated rat hearts. 10There are numerous PKC isozymes, of which the ε isozyme has recently been shown to be critical in IPC. 37Staurosporine is a nonspecific antagonist of a PKC, and further evaluation will be required to determine the role of specific PKC isoenzymes in the VSM and HUVEC cell cultures.

It has been hypothesized that PKC is part of a linear cascade with the mitochondrial K^ATPchannel as the end effector. Sato et al.  16demonstrated that increased mitochondrial K^ATPchannel opening is protective during an ischemic injury and can be elicited through priming the channel with PKC activity or by direct stimulation of the channel. This linear cascade, however, is challenged by the findings both in vitro  and in vivo  that blockade of PKC activity with simultaneous direct stimulation of mitochondrial K^ATPchannels does not lead to any protective effect. 10,36This is supported by our study because if isoflurane increases K^ATPchannel activity, blockade of PKC by staurosporine should not have abolished the protective effect. Based on these findings, it appears that PKC and mitochondrial K^ATPchannels are involved in a network more complicated than a simple linear cascade. While previous studies have indicated that these mechanisms apply to ischemic–reperfusion injury, our study suggests that these mechanisms also may be involved in cellular protection from cytokines and H²O².

The observation in our study that isoflurane and halothane inhibit cell death in both cytokine- and H²O²-induced injury may be important. Evaluation of DNA fragments in lysate versus  supernatant suggests that cytokines induce apoptosis while H²O²causes necrosis. Differentiation of apoptosis and necrosis is complex and is not conclusive by DNA fragmentation assays alone; however, our finding is consistent with other studies that have used additional methodologies. 4,23These results are also consistent with other studies that indicate that volatile anesthetics inhibit both apoptosis and necrosis. 22,23The role of PKC in inhibiting both cytokine-induced apoptosis and H²O²-induced necrosis is supported by a study by Lui et al. , 37who showed that IPC activation of PKCε attenuates both apoptosis and necrosis in cardiomyocytes. Furthermore, evidence of the role of isoflurane in inhibition of apoptosis and/or necrosis is suggested by our results indicating staining of DNA fragmentation by TUNEL assay is greatly attenuated in VSMs and HUVECs.

In conclusion, this study demonstrates that 30 min of pretreatment with 1.5% isoflurane decreases cell death due to cytokine-induced injury for at least 96 h in VSMs and 72 h in HUVECs. The timing of isoflurane administration is critical as isoflurane inhibits the initiation of the cytokine cascade but does not appear to halt it. These initial inhibitor studies suggest involvement of PKC and K^ATPchannels in isoflurane and halothane protection against both cytokine- and H²O²-induced cell death of VSMs and HUVECs.

The authors thank Tom G. Obrig, Ph.D. (Professor of Internal Medicine and Immunology, University of Virginia Health System, Charlottesville, Virginia), and Julianne J. Sando, Ph.D. (Professor of Anesthesiology), for their gift of cells, their help, and their advice.