| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Severinghaus-Radiometer Research Laboratories, Department of Anesthesia and Perioperative Care, University of California at San Francisco, San Francisco, California.
Address correspondence and reprint requests to Philip E. Bickler, Sciences 255, Box 0542, University of California Medical Center, 513 Parnassus Avenue, San Francisco, CA 94143-0542. Address e-mail to bicklerp{at}anesthesia.ucsf.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The purpose of this study was to examine a number of Ca2+-sensitive and Ca2+-insensitve signaling pathways in hypoxic neurons to identify how isoflurane modulates neuroprotective signaling. We examined the regulation of intracellular Ca2+, the Ca2+-dependent phosphorylation of Pyk-2 and MEK-ERK1/2, the Ca2+-independent phosphorylation of the JNK-cJun and p38 pathways, and several regulators of apoptosis (Akt, GSK, Bad, Forkhead protein, and p90 RSK). We measured changes in the phosphorylation of these pathways and also the phosphorylation of the transcription factors associated with them. We found that isoflurane has complex effects in hypoxic neurons, augmenting the phosphorylation of hypoxia-sensitive MAP kinases in both Ca2+-dependent and independent pathways. In contrast to the increase in MAP kinase phosphorylation, isoflurane had mixed effects on the transcription factors and apoptosis regulators associated with these pathways.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cortical neurons for culture were obtained from BrainBits, Inc. (Springfield, IL) or prepared in our laboratory. Neurons were derived from gestational day 16–18 mouse embryos as described by Booher and Sensenbrenner (13) with minor modification. Neurons were plated at a density of 1 x 106 cells/mL on glass coverslips precoated with 0.05 mg/mL poly-d-lysine. Cells were allowed to adhere for 30 min, after which half of the media was replaced by fresh astrocyte conditioned medium (UCSF cell culture facility). Seventy-two hours later, the cells were treated with 7.5 µM cytosine arabinoside for 24 h to prevent glial division. The culture medium was then replaced with standard neurobasal serum-free medium containing B27 (Gibco/Invitrogen, Carlsbad, CA) and cells were studied after 3–7 days in culture. This same culture medium was used during exposure of the cultures to hypoxia and for recovery after hypoxia. To determine the number of astrocytes in neuronal cultures, sister cultures were fixed in 4% paraformaldehyde and stained with antibodies against glial fibiary acidic protein (1:1000) and neuron specific enolase (1:100), both from ICN (Costa Mesa, CA). The cultures contained 90%–95% neurons, 1%–3% glial cells and remainder non-staining cells.
We examined how the phosphorylation status of various MAP kinases and transcription factors at the end of 1 h of hypoxia correlated with survival 24 h later. The pathways and their components that were examined in this study are shown in Figure 1. Cultures were exposed to hypoxia by placing them into a 2-L airtight Billups-Rothenberg Modular Incubator Chamber (Del Mar, CA) through which 95% N2/5% CO2 gas, preheated to 37°C, was passed at 5–10 L/min. The temperature of the chamber was kept at 37° by heat lamp near the chamber and the temperature inside the chamber was monitored with a thermocouple thermometer. After 10 min of gas flow the chamber was sealed and placed in a 37°C incubator. The partial pressure of oxygen in the culture media was approx. 0.1–0.3 mm Hg, measured with a Clark oxygen electrode (Cameron Instruments, Port Aransas TX). After the hypoxia, the culture tray was removed from the chamber, briefly opened to restore oxygenation, and returned to the incubator. For cultures treated with isoflurane, gas flowed through a calibrated isoflurane vaporizer (Ohmeda Tec 3, Andover, MA) before entering the chamber. In experiments involving inhibitors of MAP kinases or other signals, the compounds were added to the culture media just before the start of hypoxia, and the media was replaced at the end of the hypoxia. Inhibitors of JNK (SP600125), an inhibitor of MEK1/2 (U0126), and PDK-1 (LY294002) were from Cell Signaling (Beverly, MA), while U73122 (phospholipase C inhibitor), calmidazolium (calmodulin inhibitor), and xestospongin-c (IP3-receptor antagonist) were obtained from Calbiochem (La Jolla, CA).
|
|
|
|
|
|
|
|
To characterize cell death, TUNEL staining of cultures was done 24 h after hypoxia with an In situ cell death detection kit from Roche Diagnostics (Indianapolis, IN).
In separate cultures, [Ca2+]i was measured before and after the period of hypoxia and/or isoflurane exposure with fura-2 (Molecular Probes) and a dual-excitation fluorescence spectrometer (Photon Technology International, South Brunswick, NJ) coupled to a Nikon Diaphot 200 inverted microscope. Cultures were incubated with 5–10 µM of fura-2 am for 15–30 min before measurements. Slit apertures in the emission light path were adjusted to allow only light from identified neurons to reach the photomultiplier tube. Relative [Ca2+]i levels were expressed as the ratio of 510 nm light intensity emitted from alternate 345 and 380 nm excitation light. Background fluorescence (i.e., fluorescence in the absence of fura) was subtracted from total fluorescence.
Western blots of proteins from culture homogenates were performed using standard methods. Five to 8 culture plates were pooled for each assay and each study was repeated 3–4 times. Protein content in each sample was measured (Bradford protein assay with Coomassie blue) and adjusted so that equal amounts of protein were applied to each lane. Protein bands were visualized after incubation with biotinylated secondary antibodies followed by an enhanced chemiluminescence assay and quantified by image analysis software (NIH Image). Immunostaining of various signaling proteins in situ was done on cultures fixed with 4% chilled paraformaldehyde. Relative protein levels in these in situ preparations were estimated with a microscope, digital camera, and NIH Image software, based on staining intensity (mean gray area). Antibodies to p-Akt (Ser 473 phosphorylation), MAP kinase p42/44 (Thr 202/204 phosphorylation), pRaf, p90RSK, pBad, pElk-1, pJNK, pc-Jun, p38, pMKK-3/6, pATF-2, pPDK-1, PFKHD, and pGSK were obtained from Cell Signaling Technology (Beverly, MA). Antibodies to pPyk-2, were purchased from Calbiochem, La Jolla CA.
Student's t-tests or analysis of variance with a Tukey-Kramer multiple comparison (StatView) were used to compare group means. Differences were considered significant for P < 0.05.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
One percent isoflurane reduced the percentage of dead neurons in cultures 24 h after a 1-h period of hypoxia (95% N2/5% CO2; Fig. 2 panels A and B). In cultures without isoflurane, 60 min hypoxia killed approximately 40% of neurons. One percent isoflurane reduced death to <8% of neurons.
|
To characterize cell death, we TUNEL stained cultures exposed to hypoxia, isoflurane, and a combination of the two. Isoflurane significantly reduced TUNEL positive cells present 24 h after hypoxia (Fig. 2 panels C and D). Of note, isoflurane alone (i.e., with normal levels of oxygen) caused a small but statistically significant increase in the percentage of TUNEL positive neurons from 1% to 2.5%.
Because isoflurane-induced increases in [Ca2+]i are important to isoflurane neuroprotection and preconditioning in hippocampal slice culture neurons (9), we measured [Ca2+]i in cortical neurons during and after hypoxia (Fig. 3). In normally oxygenated neurons the fura-2 fluorescence ratio was 0.85 ± 14 (n = 6). Isoflurane increased [Ca2+]i in normally oxygenated neurons by approximately 20% (change in fura-2 fluorescence ratio of 0.15–0.20). With washout of isoflurane, [Ca2+]i returned to baseline. In hypoxic neurons, the presence of isoflurane prevented the large increase in [Ca2+]i that occurs after 1 h of hypoxia. The changes in fura-2 fluorescence ratios in hypoxic neurons (Fig. 3) are consistent with increases in [Ca2+]i in the low micromolar range, a toxic level in most neurons (14).
|
The specific IP3 receptor antagonist, xestospongin C, substantially reduced the increase in [Ca2+]i seen in neurons exposed to isoflurane (Fig. 3), indicating that an important source of Ca2+ increase in neurons exposed to isoflurane is derived from the endoplasmic reticulum. Further, in cultures treated with xestospongin C, smaller increases in [Ca2+]i were observed after hypoxia or hypoxia combined with isoflurane. Thus, in our cultures, IP3 receptors are related to essentially all of the increase in [Ca2+]i elicited by isoflurane in normally oxygenated neurons and to a substantial amount of it in neurons during hypoxia.
The phospholipase C (PLC)-inositol phosphate system is a key regulator of Ca2+ release from the endoplasmic reticulum. Isoflurane increases the rate of membrane lipid hydrolysis by PLC with the subsequent formation of inositol phosphates such as IP3 (15). Notably, IP3-receptors are associated with the same Ca2+-dependent neuroprotective signaling as is isoflurane in hippocampal slice cultures (9). Therefore, we tested whether PLC could be an upstream modulator of survival signaling in cortical neurons. In hypoxic neurons, the PLC inhibitor U73122 (10 µM) partly reversed isoflurane's protection (Fig. 4). In previous studies with similar neuronal cultures, we found that this inhibitor did not significantly increase cell death after hypoxia (16).
|
Isoflurane increases signaling through the Ras-Raf-ERK MAP-kinase signaling pathway in hypoxic neurons and is essential for neuroprotection. The upstream ERK (p42/44) pathway component Pyk-2 is Ca2+-sensitive (17). Therefore, we determined if signaling in this pathway was increased in neurons exposed to a combination of hypoxia and isoflurane. In Figure 5, we show that ERK p44 levels were significantly increased by a combination of hypoxia and isoflurane, as detected by both Western blots and in situ immunostaining (Fig. 4 panels A–C). This immunostaining was prevented by a MEK1/2 inhibitor U0126 (an upstream regulator of ERK), by a PLC inhibitor (U73122), by a calmodulin inhibitor (calmidazolium), and by a IP3-receptor antagonist (xestospongin C) (Fig. 5, panels A and C), suggesting that p42/44 phosphorylation in hypoxic neurons is related to intracellular Ca2+ release mediated by PLC and IP3 receptors.
The sensitivity of MEK1/2 phosphorylation to changes in [Ca2+]i was investigated in the cultures by exposing them to a small concentration of the Ca2+-selective ionophore ionomycin (1 nM). Increased in situ immunostaining was observed in cultures fixed and stained after 1, 2, 5, 20, and 30 min after the addition of ionomycin to the cultures (P < 0.01 for each time point).
Isoflurane's protection of hypoxic neurons required MAP kinase ERK because U0126, which prevents phosphorylation of ERK in our neurons by blocking the upstream kinase MEK1/2, (9) abolished protection (Fig. 6). Because xestospongin C blocked phosphorylation of ERK during hypoxia in isoflurane-exposed neurons and ERK phosphorylation is required for protection, we determined if xestospongin C blocked protection as well. Xestospongin C partly reversed protection, suggesting that Ca2+ release from inside the cell is important in the protective mechanism involving the MAP kinase ERK pathway.
|
Levels of ERK-dependent phosphorylated proteins at other points in the ERK pathway were also examined (Fig. 7; cf. Fig. 1). Hypoxia and a combination of hypoxia and isoflurane administration increased pPyk-2 levels, measured both with Western blotting and with in situ immunostaining of fixed cultures. This Ca2+-dependent kinase is upstream of Raf. Consistent with this Ca2+-sensitivity, the increase in pPyk-2 in hypoxic neurons concomitantly exposed to isoflurane was prevented by a PLC inhibitor (U73122), a calmodulin inhibitor (calmidazolium) and the IP3 receptor inhibitor xestospongin C (Western blot in Fig. 7). In contrast to pPyk-2, pRaf levels increased with hypoxia or isoflurane but not in the presence of both (Western blots and in situ immunostaining in Fig. 7).
ERK phosphorylates a number of apoptosis regulating proteins including the transcription factors Elk-1, p90 RSK, and the pro-apoptotic protein Bad. We found that isoflurane, or a combination of isoflurane and hypoxia, decreased immunostaining of these phosphoproteins. In Figure 8, Western blots and in situ immunostaining images and intensity are presented for p-ELK and in situ immunostaining data for p90 RSK, pBAD (serine 155), and pBAD (serine 136). In the case of pElk-1, phosphorylation in hypoxic neurons exposed to isoflurane depended on MEK, PLC, and calmodulin but not on IP3 receptors. Isoflurane decreased the phosphorylation of p90 RSK in hypoxic neurons (Fig. 8, panel C.) Isoflurane and hypoxia also decreased the phosphorylation of the apoptosis regulator Bad at S155 but not at S136 (Fig. 8, panel D).
Levels of phospho-JNK increased in neuron cultures exposed to hypoxia, to isoflurane with normal levels of oxygen, and to a combination of hypoxia and isoflurane (Fig. 9). This was especially clear for the 54 kD p-JNK phosphoprotein. A JNK inhibitor (SP600125) prevented this increase in phosphorylation in hypoxic neurons exposed to isoflurane (Fig. 9 panel A). In addition, SP600125 completely reversed protection of hypoxic neurons by isoflurane (Fig. 9 panels B–D). In separate cultures, SP600125 did not increase the death of hypoxic neurons.
Phosphorylated c-Jun levels were increased by hypoxia, isoflurane, and a combination of the two (Fig. 10). This phosphorylation was opposed by antagonizing IP3 receptors with xestospongin C and by the JNK inhibitor SP600125. Taken together, the data in Figures 9 and 10 suggest that in hypoxic neurons, isoflurane increases the activity of the JNK to cJun pathway.
Hypoxia rapidly increased the intensity of p38 immunostaining in cultured neurons. Increased immunoreactivity was observed after 1 min of hypoxia, a change that was also observed after 5, 20, 30 and 60 min of hypoxia (each P < 0.05 compared with control; 60-min results shown in Fig. 11 panel A). Both hypoxia and isoflurane increased p38 levels detected by Western blotting, but a combination of the two produced a clear reduction in the p38 bands in the Western blots. Inhibition of PLC, calmodulin, and IP3 receptors did not prevent the decrease of p38 phosphorylation by isoflurane seen in Western blots during hypoxic conditions (data not shown). These finding were not surprising, considering that p38 levels do not increase after moderate increases in [Ca2+]i produced by application of a small concentration of calcium-selective ionophore (ionomycin, 1 nM, data not shown). Also consistent with an isoflurane-mediated activation of the p38 pathway was an increase in the level of pMKK-3, a kinase upstream of p38 (Fig. 11B).
We also examined the p38-dependent transcription factor ATF-2. Although this protein was little affected by hypoxia, it was increased by isoflurane (Western blots, Fig. 11C). However, with a combination of isoflurane and hypoxia an approximately 50% reduction in the levels of this protein compared with control was observed (Fig. 11, D).
Akt immunoreactivity (phosphoserine 495) in Western blots remained unchanged by 1 h of hypoxia, but the presence of isoflurane during hypoxia led to a clear increase in Akt (Fig. 12, A). The PI3-kinase inhibitor LY294002 decreased Akt immunoreactivity in hypoxic neurons treated with isoflurane. This inhibitor also reduced the capacity of isoflurane to protect hypoxic neurons (Fig. 12, B and C). LY294002 did not increase cell death in neurons treated only with hypoxia (data not shown).
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Specific novel findings of this study are that 1) isoflurane-mediated neuroprotection from hypoxia is associated with increases of MAP kinase phosphorylation in both Ca2+-dependent and Ca2+ independent pathways, 2) isoflurane increases the levels of the anti-apoptotic protein Akt and decreases pBad, and 3) isoflurane has selective effects on MAP kinase-dependent transcription factors, increasing some but decreasing others.
Our findings show that PLC is required for the protection of hypoxic neurons by isoflurane. This may be related to the stimulation of PLC-dependent membrane lipid hydrolysis by isoflurane (15), an increase in IP3, and subsequent release of Ca2+ from the endoplasmic reticulum. PLC is involved in the isoflurane-mediated activation of the focal adhesion kinase pp125, which plays a role in protecting neurons from hypoxic death (18). Also of note, the PLC activation that occurs with the activation of neuroprotective growth factor receptors stimulates intracellular signaling that activates PI3-kinase/Akt and several MAP kinase pathways such as ERK (19). This was also observed with isoflurane (Figs. 5, 12).
This study adds to mounting evidence that moderate, not large, increases in [Ca2+]i in hypoxic neurons are associated with neuroprotection derived from volatile anesthetics and growth factors. We showed that increases in [Ca2+]i are linked with the activation of MAP kinase ERK and c-JUN because the phosphorylation of these signals is prevented by antagonizing IP3 receptors, which represent the major source of isoflurane-induced increase in [Ca2+]i (Figs. 2, 6, 10) (8,9). Further, preventing an increase in [Ca2+]i decreases the phosphorylation of the ERK-dependent transcription factor Elk-1 (Fig. 8).
We found that the presence of isoflurane enables neurons to avoid large and potentially toxic increases in [Ca2+]i during hypoxia. This was also seen in hippocampal neurons in organotypic cultures (9). Our data suggest that a moderate increase in [Ca2+]i is required for survival, because when increases in [Ca2+]i during hypoxia are prevented by xestospongin C (Fig. 3), death increases (Fig. 6). Although large increase in [Ca2+]i can clearly contribute to excitotoxic neuronal death after anoxia or ischemia (20–22), we suggest that more moderate increases in [Ca2+]I facilitate neuroprotective signaling processes. This is consistent with a substantial body of evidence showing that moderately increased [Ca2+]i may be a critical survival signal in a variety of different contexts (23–25). Thus, one aspect of isoflurane's neuroprotective actions may be summarized as keeping [Ca2+]i not only within survivable limits during hypoxia, but at a slightly increased level associated with active survival signaling.
Neurons undergo apoptosis unless they receive sustained trophic support with appropriate levels of intracellular Ca2+ and/or growth factor receptor stimulation (24). The MAP kinase p42/44 pathway is key to this Ca2+-dependent trophic support system (26) and was found to be required for isoflurane's protection (Fig. 6). The MAP kinase p42/44 couples growth factor receptors for NGF and BDNF to neuroprotective effects such as N-methyl-d-aspartate receptor modulation and subsequent reduction of glutamate excitotoxicity (10). In addition, p42/44 phosphorylates HIF-1
, a transcription factor required for the adaptation of cells to acute and chronic hypoxia (27).
MAP kinase ERK-related transcription factors may be very important in mediating survival by both Ca2+-dependent and Ca2+-independent mechanisms. We found that the ERK-dependent transcription factors Elk-1, p90RSK, and p-Bad are all modulated by isoflurane. Elk-1 is survival-associated transcription factor phosphorylated by p42/44 that functions in a neuroprotective role against glutamate toxicity in neurons (28). Hughes et al. (29) showed that sustained activation of the ERK pathway and the ERK-dependent p90 RSK family of kinases are involved in excitotoxic death. Therefore, a decrease in the level of phosphorylated p90RSK is probably protective against death after hypoxia.
Although isoflurane substantially reduced cell death and TUNEL staining in neurons after hypoxia, some of its actions on apoptosis-regulating proteins appear to be consistent with favoring apoptosis. For example, isoflurane reduced levels of p-Bad phosphorylation at Serine-155 but had no effect at Serine-136. Bad facilitates the final common pathway of the intrinsic apoptosis cascade, and its activity is regulated by phosphorylation at serine residues 112, 136, and 155. The S-112/155 residues are phosphorylated by p90RSK and PKA, whereas Akt phosphorylates S-136. Phosphorylation decreases Bad's proapoptotic actions, by blocking the dimerization of Bad to Bcl-xL. No change S-136 phosphorylation and a decrease in S-155 phosphorylation (Fig. 8), therefore, seem inconsistent with neuroprotection. However, isoflurane acts against apoptosis via phosphorylation of Akt at S-473 and phosphorylation of GSK-3 at Serine-9, which, in terms of preventing cell death, must be the more important effect. Activated Akt inhibits apoptosis in neurons (30).
The results suggest that the JNK pathway is important in isoflurane's neuroprotection because a JNK inhibitor blocked isoflurane's neuroprotection (Fig. 9). Although this is the first report showing that JNK is required for neuroprotection of hypoxic neurons by isoflurane, it is consistent with the protective role of JNK in cerebral and cardiac preconditioning (12,31,32). Although JNK signaling has both apoptotic and anti-apoptotic effects, its activation in response to stress can play a role in mediating cell survival (33). The proapoptotic effects of JNK (34) are inhibited when the Akt pathway and the ERK pathways are both activated (35), a coincidence observed in our neurons (Figs. 5 and 12). It therefore appears that signaling in both the ERK and JNK pathways are simultaneously required for isoflurane protection from hypoxia. It should be noted that the protein kinase inhibitors used in these studies, although the most specific available, probably inhibit multiple kinases to some degree.
Activation of the MAP kinase p38 pathway is generally associated with stress signaling and events that culminate in cell death. Inhibitors of p38 decrease cell death after cerebral ischemia (36), although p38 can be activated by mild and survivable hypoxia or preconditioning in neurons. In our study, hypoxia by itself produced sustained increases in p38 levels, and this was attenuated by isoflurane. It is not clear if this represents a protective feature of isoflurane's effects on intracellular signaling in hypoxic neurons or if it contributes to injury. Of interest, preconditioning with isoflurane increases p38 expression in cortex and hippocampus, and may be necessary for the preconditioning protection observed in intact rats (11).
This study cannot predict that isoflurane has the same actions in the brain of intact animals as it does on tissue cultures. Cultures differ in many ways from intact animals, including less sensitivity to hypoxia, and intrinsically different structure and physiology compared with the intact brain. Another limitation of this and similar studies is that we examined changes in signaling factors only 24 hours after hypoxia. It is clear that hypoxic or ischemic injury may evolve over long periods and that anesthetic neuroprotection presenting early may fade with time.
We found that isoflurane and hypoxia interact to produce a complex series of neuroprotective signals in cultured hippocampal neurons, resulting in a substantial reduction in cell death and TUNEL staining after hypoxia. The neuroprotective actions of isoflurane include an increase in the phosphorylation of hypoxia-sensitive MAP kinases in both Ca2+-dependent and independent pathways. In contrast to the increase in MAP kinase phosphorylation, isoflurane had mixed effects on the transcription factors and apoptosis regulators associated with these pathways, pointing to MAP kinases as crucial to the neuroprotective actions of this anesthetic.
|
ACKNOWLEDGMENT |
|---|
|
Footnotes |
|---|
Supported by grant RO1 GM 52212 from the National Institutes of Health to PEB.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Zhang, Y. Dong, B. Zhang, F. Ichinose, X. Wu, D. J. Culley, G. Crosby, R. E. Tanzi, and Z. Xie Isoflurane-Induced Caspase-3 Activation Is Dependent on Cytosolic Calcium and Can Be Attenuated by Memantine J. Neurosci., April 23, 2008; 28(17): 4551 - 4560. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Liang, Q. Wang, Y. Li, B. Kang, M. F. Eckenhoff, R. G. Eckenhoff, and H. Wei A Presenilin-1 Mutation Renders Neurons Vulnerable to Isoflurane Toxicity Anesth. Analg., February 1, 2008; 106(2): 492 - 500. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
significant difference compared with the hypoxia group.
compared with non-xestospongin c treated group. Bars show means ± sd for 6 cultures.
