| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology and Neurological Surgery, State University of New York at Stony Brook, HSC University Hospital, Stony Brook, New York
Address correspondence and reprint requests to Ira J. Rampil, MS, MD, Department of Anesthesiology and Neurological Surgery, State University of New York at Stony Brook, L4-060 HSC University Hospital, 100 East Loop Road, Stony Brook, NY 11794-8480. Address e-mail to ira.rampil{at}sunysb.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The current prevailing theory of anesthetic action involves lipophilic sites within membrane-sited proteins (1). The most extensively studied of these proteins are the ligand-gated ion channels, many of which have their electrical properties altered by clinically relevant concentrations of volatile anesthetics (2,3).
Relatively unexplored, however, is the possibility of other effects on cell function by anesthetics. Cells, including neurons, have many other transducers of the local environment. Some of these are within the cell membrane, others are deep within the cytoplasm. These transducers initiate amplification cascades that, in turn, among many functions, may alter genetic control of the cell. Genetic control may be modulated in several broad areas: chromatin remodeling, DNA transcription into RNA, maturation and translation of RNA into protein, post-transcriptional modification of protein, and, finally, degradation of protein. Chromatin remodeling is the process of unwinding the tightly wrapped DNA into a structure that can bind promoters and transcription factors that target a specific gene or genes and allows the helix to be "unzipped." Transcription uses a large complex of many protein subunits to make an RNA copy of the target DNA. Transcribed RNA must then be processed to have noncoding sections spliced out and an adenylated tail added, so that another large multi-unit complex can translate the messenger RNA (mRNA) into a protein. Once synthesized, a protein may be modified by many processes, including cleavage, phosphorylation, and glycosolation. Finally, protein concentration in the cell is governed by the balance of production and degradation.
Using GeneChip technology, we performed an unbiased survey of the changes in gene expression across the entire known transcribed genome of Rattus norvegicus (common laboratory rat) from the basolateral region of the amygdala. In particular, we examined the changes at 6 h in expression of protein-coding mRNA after a brief dose of isoflurane, compared with similarly handled, but unexposed, control rats. Delineating the complete response of a cell to an anesthetic may help to understand the etiology of unwanted side effects and improve the safety and efficacy of general anesthesia.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Genechip Methods
Approximately 20 µg of total mRNA was processed for the gene chip analysis using the Enzo BioArray RNA Transcript Labeling Kit (T7) and the Affymetrix (Santa Clara, CA) Rat Expression Set 230-A GeneChip. The isolated amygdalar mRNA was reverse transcribed with elongated DNA primers yielding cDNA templates with T7 RNA polymerase promoters. The cDNA templates were then transcribed to cRNA with incorporation of biotinylated nucleotides. The labeled cRNA was then hybridized to the Affymetrix RAE 230A GeneChips and their biotin labels were linked to a fluorophore conjugated streptavidin, which served as a reporter for the final UV scanning of the chips. No amplification was used in this protocol.
The design of Affymetrix GeneChips includes both external and internal controls. These microarrays measure gene expression by hybridizing unique 25 base long nucleotide probes to matching portions of cellular RNA. These 25-mer probes are called expressed sequence tags (ESTs). The external controls consist of several bacterial and other RNA ESTs added to the final mix. Internal control was achieved by replicating the measurement of each EST in multiple pairs (usually 11 pairs per EST). Each pair consisted of a spot with a perfect nucleotide match to a section of the EST and a spot with a single mismatched base in the middle of the sequence. Using paired Student's t-tests and a variety of other criteria, the analysis system was able to exclude (i.e., marked ABSENT) from each GeneChip a number of ESTs whose measured expression was marginal or noisy. Exclusive of external control values, the 230A GeneChip provided normalized expression values for 15,866 ESTs, which we transformed using log base 2. To minimize statistical noise, we created an ad hoc rule that for each EST, further analysis was performed only if there were at least 7 of 9 valid (i.e., PRESENT per Affymetrix software) measurements in the baseline group and 4 of 5 in the experimental group being compared. The mean expression, sd, and coefficient of variation, were assessed for each expressed EST in each experimental group of rats. Data were compared using unpaired, heteroscedastic two-tailed Student's t-test. Because these comparisons were performed for the purpose of hypothesis generation rather than hypothesis testing, Bonferroni correction was not performed and P
0.05 was accepted as significant. Only genes linked to biological function per the PubMed Gene database (5) and/or the Gene Ontology Database (curated and maintained by the European Bioinformatics Institute, http://www.ebi.ac.uk/GOA/) as of August 2005 were analyzed. Those functional annotations, as well as other information regarding specific genes, can be found within the Gene database at www.pubmed.gov. We subdivided genes into functional families using the Gene Ontology hierarchy. When citing a specific gene we use the following convention: Symbol (standard gene abbreviated name) for a gene with accepted curation and annotation available through PubMed Gene, Symbol_pred for genes with incomplete curation, but likely to be a rat homolog of the predicted gene, and "Symbol for genes for which our group found sufficient sequence homology or other data to suggest a probable, but not certain, identity with a similar gene from a different species, or a similar rat gene.
For the confirmation of the microarray results quantitative PCR analysis was performed using the SYBR Green method (6), integrated in the one-step reverse transcription (RT)-PCR kit from Qiagen (Valencia, CA; catalog#204243) following the manufacturer's protocol. The real-time PCR was performed in the thermocycler DNA Engine Opticon 2 (MJ Research, Waltham, MA) and the results were evaluated with the integrated analysis software (Opticon Monitor 2.02). Two total RNA isolations were used in triplicate measurements for the BC group and the ISO group. These RNA samples were tested for the genes Jun (which was constitutively expressed in all of our micro array results) and Gabrb3 (which showed decreased expression in our micro array ISO group). The following oligonucleotide primers were used: Jun (forward) 5'-ggtctacgccaacctcagcaacttc-3' and (reverse) 5'-gatccgctcctgagactccatgtc-3'. Gabrb3 (forward) 5'-gctgtacgggctcaggatcacc-3' and (reverse) 5'-ggtaggcacctgtggcgaagac-3'. To assure the specificity of the oligonucleotide primers an RT-PCR with melting curve analysis was performed showing only single peaks for each of the primer pairs used. As negative controls, nontemplate controls were included in the experiment, where no RNA was added to the reaction mix and where no PCR products were detected after 40 cycles. Approximately 0.2 µg of total RNA was used as template in a total reaction volume of 20 µL (for further details see the protocol of the Qiagen kit). The parameters for the thermocycler DNA Engine Opticon 2 were: One cycle of incubation at 50°C for 30 min for RT followed by one cycle of 95°C for 15 min for activation of hot start TaqDNA polymerase. This was followed by 40 cycles of: 94°C for 30 s double stranded DNA melting, 64°C for 30 s primer annealing, 72°C for 20 s elongation, and 78°C for 10 s measuring of fluorescence of SYBR Green. At the end of the program a melting curve was measured for all samples from 65°C to 95°C to confirm the specific amplification of only one pure PCR product for each reaction. With the Opticon Monitor software 2.02 average cycle threshold (CT) values for Gabrb3 and Jun were obtained after background subtraction with the negative controls and baseline determination at a CT value of 0.004. These values were used in relative quantification by subtracting the average BC values from the average ISO values, yielding the difference CT values for both genes in the ISO group.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The coefficient of variation (CV), illustrated in Figure 1, was similar in both groups, with 95% of ESTs having a CV <7.0% demonstrating reproducibility in these populations and adequacy of the sample sizes.
|
Administration of 15 min of isoflurane 2.0 Vol% resulted in the altered expression of 424 ESTs compared with the baseline control expression profile. After removal of ESTs with annotation inadequate to specify function, and cases of degeneracy where a single Locus Link assignment had multiple Affymetrix Probe IDs there were 269 unique, annotated genes.
Genes with altered expression after exposure to isoflurane were grouped into broad functional groups based on their annotation from PubMed Gene (7). Figure 2 conveys the differential changes in expression by group and a complete enumeration of the genes with altered expression is contained in the Appendix (available at www.anesthesia-analgesia.org).
|
The real-time kinetic PCR study confirmed stability of mRNA expression in a housekeeping gene that was constitutively expressed (Jun) and the amplification of mRNA for a gene (Gabrb3) found increased by microarray analysis.
A cartoon synapse in Figure 3 provides an overview of the expression changes relevant to specialized neuronal function. Of the few affected genes producing ionophore protein subunits, most were downregulated. This included a gamma-aminobutyric acid subunit (Gabrb3) and an N-methyl-d-aspartic acid (NMDA) glutamate (Grinl1a) subunit as well as two potassium channels (Task5, Kctd2). Among calcium ionophores, Cacna1, a voltage sensitive channel subunit is upregulated, whereas Vdac1, a mitochondrial voltage-gated anion pore involved in apoptosis as well as fear conditioning (8), is suppressed. Gja1, a major component of myocardial (and probably central nervous system) gap junctions is also increased. A particularly interesting finding was the upregulation of Kcr1, a regulator of the Herg protein (9,10), the voltage-gated potassium channel apparently responsible for many examples of long QT syndrome (11,12). Synaptic ionophores and receptors are structurally positioned and modulated by proteins within the postsynaptic density (13). Isoflurane alters expression of several of these proteins, for example, Dlgap1, Cask, and Gda, which interact with glutamate receptors. AMPA and NMDA glutamate receptors modulate their density at the synapse by endocytosis and exocytosis of receptors attached to vesicles held in storage (14,15). Nsg1 has been newly assigned as an important factor in this recycling of receptors (16,17), but it is constitutively expressed with no difference between experimental groups. Rab8 is a small GTPase that functions as a vesicular transport protein important to delivery of AMPA receptors to the postsynaptic membrane and the creation of long term potentiation (18). Both Rab8a and Rab8b are increased by isoflurane. Epb4.1L3, an isoflurane downregulated protein which binds spectrin to actin, may act as the molecular conveyors of these vesicles to and from the cell surface. The expression of other actin-related genes is also modified by isoflurane, including upregulation of components of the Arc 2/3 complex, which creates branch points in the actin cytoskeleton. Isoflurane caused an increase in expression of extracellular adhesion molecules in the immunoglobulin (Igsf4a1, Opcml) and integrin (Cib1) families, as well as a collagen type I receptor (Cd36L2). Opcml may also act as an accessory to the µ-opioid receptor and play a role in synaptic machinery (19). Finally, the expression of several modulators of presynaptic neurotransmitter release and other vesicular traffic were altered. These included the aforementioned Rab proteins (20), as well as Vamp2 (otherwise known as synaptobrevin 2) which is thought to participate in neurotransmitter release at a step between docking and fusion, Surf4 which may be a vesicle packaging factor, and Frequenin, a calcium binding protein implicated in the neurotransmitter release process.
|
Generic intracellular vesicular traffic was probably diminished by the downregulation of several other RAB GTPase related proteins, including Rabep1 and Arfgap1. Copb1, a constituent of non-clathrin coated vesicles, was also downregulated. Curiously, the gene for the Rhesus blood group antigen, Rhced, which is actually an ammonia transporter molecule, is upregulated by isoflurane.
Isoflurane altered the expression of total of 35 genes associated with signaling subsystems defined here as non-neuron-specific ionophores, internal signaling cascade members, and proteins associated with G-protein coupled receptors. Among the signaling pathways altered was the nitric oxide system with downregulated expression of dimethylarginine dimethylaminohydrolase (Ddah1). Expression of one cytokine (Bmp3) decreased, whereas Bmp4 and Cklfsf7 increased. Isoflurane downregulated a subunit of the progesterone hormone receptor (Pgrmc2) and a regulator of the adrenergic receptor (Adrbk1) and upregulated the parathyroid hormone receptor (Pthr1), an orphan receptor (‘GPCR172b), as well as part of the cAMP receptor (Adcyap1r1). G protein subunits Gng7 and Gng8 themselves were downregulated and Gkap1, a modulator, increased. A kinase associated with the JNK (jun N-terminal kinase) signaling cascade (Mapk10) was upregulated. Inducible cyclooxygenase (Ptgs2, aka Cox2), a key enzyme in prostaglandin biosynthesis, is overexpressed after isoflurane as well. One may hypothesize that some of the alterations in inflammation seen after general anesthesia (21,22) may result from this altered expression of Ptgs2.
As illustrated in Tables 1A and 1B, isoflurane impacts a host of genes whose products control the cell's use of DNA and RNA. Altered cell cycle and DNA replication-related and repair-related ESTs/genes are represented by cyclin-dependent kinases (Cdkl1, Cdk4) and the specific inhibitor, Cdkn2c, all of which are increased in expression by isoflurane. Increased also was Lztfl1 (involved in chromosome segregation), DNA primase Prim1, DNA helicase Recql, and DNA telomerase inhibitor ‘Pinx1. DNA exonuclease Isg20 was inhibited.
|
|
The first subgroup contains histone-related genes, including the downregulated histone deacetylating genes Tada3L, Thap7, and "p66. These genes all function in the same gene silencing mechanism involving the nucleosome remodeling deacetylase complex (NuRD) (23,24). Normally, these proteins participate in the very first step of transcriptional regulation by condensation of chromatin (winding DNA into a compact, inaccessible structure), resulting in a very strong suppression of transcription.
The downregulation of these NuRD genes, together with the upregulation of the ASF1 anti-silencing function 1 homolog (Asf1a) by isoflurane, implies that certain chromatin areas (and the genes they contain) are kept accessible longer than under normal conditions, which, in turn, would allow for upregulated transcription from those areas. Other members of this functional gene subgroup are the histone acetyl transferase, Myst2, "Set involved in nucleosome assembly, and the histone 2a subunit Th2a, with their respective directions of change in expression listed in Tables 1A and 1B.
The second subgroup of genes belongs to two prominent gene regulatory pathways. One pathway shows strong repression of regulatory genes from the steroid-thyroid hormone receptor superfamily, represented by the nuclear retinoic acid receptor family of transcription factors (Rxrb, "Rap80, both suppressed) which initiate regulatory cascades involved in growth (morphogenic effects) and metabolism. The other pathway is strongly promoted by isoflurane and is associated with the early growth response family (Egr2, Madh3) and its downstream JUN pathways, which are involved in long term memory and neuronal plasticity (25,26). One of their upregulated target genes is Bteb1, a basic transcription factor that, when knocked out, impairs contextual fear-conditioning related synaptic plasticity (27).
The 17 post-transcriptional and pre-translational genes are involved in the maturation processes of RNA, e.g., RNA splicing, post-transcriptional RNA modifications, and recruiting and transport of RNA. They include the neuron specific genes adenosine deaminase (Adar), essential for axonal growth, and survival of motor neuron1 (Smn1) found in axons and growth cones of neurons.
Data obtained in this study suggest that exposure to isoflurane globally depresses protein synthesis from mRNA. Significant changes are seen in subunits of the protein complex (the eukaryotic Initiation Factor, or eIF) that controls the translation initiation process. These two subunits, eIF2b and eIF4, are recognized as major regulatory controls that the cell uses to globally control synthesis (28). The first factor, eIF2, has three subunits, and its known role is to provide an energy charge to a transfer RNA-amino acid complex and deliver it to the small ribosome and the start codon of the mRNA being translated. Phosphorylation of eIF2 stops this process by preventing exchange for a new energy charge in its ß subunit. Such signals as starvation, heat shock, and viral infection act via this phosphorylation to decrease global protein synthesis. After isoflurane, the 
chain of the ß subunit increases in mRNA expression, which may alter the balance of this regulatory point. At the same time, the expression of a subunit of the transcription initiation promoter complex eIF4 is reduced. Phosphorylation of eIF4 is a mechanism by which hormones, growth factors, and cytokines increase the rate of synthesis; therefore, reduction in the available pool of the complete complex will reduce synthesis. Other genes involved in translation whose expression is altered after isoflurane are enumerated in Table 1B. The complete Affymetrix expression data for this experiment has been archived on the NCBI Gene Expression Omnibus database as GSE1779.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The methodology used in this study, global mRNA expression analysis via GeneChipTM technology, is powerful but in some respects not mature. At present, only this methodology allows an unbiased survey or full picture snapshot of DNA transcription, whereas other means of detecting transcriptional activity examine only a few genes at a time. Unfortunately, the ESTs used to identify each gene are not necessarily identical across all commercial providers of the GeneChips, leading to potentially discordant results across vendors when comparing results in the literature. This study is further limited by the limits of interpretation of mRNA expression, which, although a potent factor in the control of protein production, is not the only factor influencing the final protein concentration. Indeed, isoflurane itself seems here to be a regulator of protein translation rates by its action on the initiation complex.
The present study is also limited by only a brief exposure to a single anesthetic and only a single time point of analysis 6 hours after the exposure. Adding further complexity to analysis, tissue samples such as those used in this study are composed of mixed populations of cells, including neurons, glia, and vascular cells.
The full genomic response to a stimulus consists of sequential cascades of expression of genes, many of which serve as transcription factors that influence subsequent waves of expression. Although protein assays specifically and directly quantify intracellular protein, an important step in validating the hypotheses generated by RNA microarrays, the microarrays have some advantage in providing a broad overview and some mechanistic detail. As tools for proteomic enquires mature, it will become possible to measure and contrast the changes in actual protein content of thousands of proteins against the mRNA expression data, thus allowing investigation of regulation of translation and modulation of cell function by general anesthesia. In this study, the process of seeking biologically clustered networks of genes with altered expression mitigates the multiple comparison issue and allows generation of hypotheses (below) which may now be tested at either the RNA, protein, or functional level.
This study has exposed a complex and prolonged network of genomic changes in neurons after exposure to isoflurane. The function of a few proteins have been known, or anticipated to change, in response to anesthesia, but most of the changes in expression seen here have not. Although much of the prior research into general anesthetic action has examined the direct effects on ionophore function, isoflurane appears to further alter ionophore function by altering expression of regulatory and interacting structural proteins. Isoflurane also appears to influence synaptic exocytosis and other forms of cellular transport. As noted above, there is also a potent impact on proteins that control genomic processes. Awareness of these changes raises a number of interesting questions about the extent of anesthetic action and the potent impacts on patient outcome. Indeed, a number of significant new hypotheses regarding the action of isoflurane follow biological analysis of the observed expression changes. For example, long-term potentiation and long-term memory are currently thought to be initiated by synaptically mediated increases in calcium ion in the postsynaptic terminal (29). This increase is mediated by glutamate opening NMDA channels, which directly allow entry of calcium plus depolarization of the terminal by AMPA channels, which, in turn, open voltage-gated calcium channels. The excess calcium triggers a number of needed intracellular signaling systems, including calmodulin. Exposure to a brief period of isoflurane reduces expression of AMPA channels (Grinl1a) and calmodulin (camk2d) while increasing the expression of voltage-gated calcium channels (Cacna1c).
Two other proteins thought to be vital to the memory process, Vdac1 (8) and Dlgap1 (30), are also dysregulated by isoflurane. One may hypothesize that these changes may act in concert to impair the cellular process of learning. The expression data also suggest that global protein synthesis is downregulated by the changes by the reduction in Eif4a and the increase in Eif2b. This would negatively impact memory formation as well as many other critical functions of the brain. Finally, we note the possible mediation of a serious side effect of anesthesia by the genomic impact of isoflurane on a recently described pathway. Isoflurane upregulates Kcr1, a protein that modulates Herg, the voltage-gated potassium channel thought responsible for many cases of the prolonged QT syndrome. If specific genetic pathways altered by exposure to general anesthetics prove to be correlated with specific side effects, it may be possible to directly target and therapeutically alter the cell's undesirable response and thus contribute to the development of safer, less debilitating anesthetics for future patients.
|
Footnotes |
|---|
Supported, in part, by the SUNY Medical School and departmental sources.
Supplemental data available at www.anesthesia-analgesia.org
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Kalenka, J. Hinkelbein, R. E. Feldmann Jr, W. Kuschinsky, K. F. Waschke, and M. H. Maurer The Effects of Sevoflurane Anesthesia on Rat Brain Proteins: A Proteomic Time-Course Analysis Anesth. Analg., May 1, 2007; 104(5): 1129 - 1135. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Z. Pan, R. G. Eckenhoff, and M. F. Eckenhoff Limitations of Microarray Studies Anesth. Analg., May 1, 2007; 104(5): 1300 - 1301. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
