Background

Environmental influences are translated by epigenetic mechanisms into expression changes of target genes, which may have significant roles in brain development. Administration of anesthetics also represents an environmental influence which can induce epigenetic changes, especially when exposed in the vulnerable phase of brain development. As an example, administration of ethanol, one of the oldest anesthetics known, during critical stages of brain development, results in significant chromatin remodelling [1, 2] in the promoters of several genes leading to long-term cognitive impairments, such as the brain-derived neurotrophic factor (BDNF) and c-Fos genes [1, 3]. Furthermore, general anesthesia can induce histone modification, resulting in suppression of, e.g., mouse Per2 expression in the suprachiasmatic nucleus [4] and c-Fos and BDNF histones within the immature rat hippocampus [5]. A further study revealed increased occupancy and methylation levels at the promoter region of the BDNF accompanied by suppressed BDNF expression and subsequent memory loss [6].

Together, these studies provide hints that epigenetic modulation may play a role in anesthesia related phenotypes. While isoflurane anesthesia can evoke neuroinflammation [7] and neuroapoptosis [8,9,10] in animal experiments, we investigated in this a priori in vitro analysis, whether isoflurane causes DNA methylation changes in an array of genes involved in apoptosis, cytokine production, and inflammatory response. We also assessed whether results obtained in the permanent hippocampal cell line HT22 are comparable with those in primary cultured hippocampal neurons. If results were comparable, further research, of cause, would be much easier than obtaining and maintaining in culture postnatal primary hippocampal neurons.

Therefore, we measured and compared methylation of an array of genes of HT22 cells and primary hippocampal neurons without and after isoflurane exposure. In case of changes in methylation of specific genes, mRNA analysis was performed to assess possible effects on gene expression.

Methods

Cell lines

HT22 is an immortalized mouse hippocampal cell line, which is a sub-line derived from parent HT4 cells that were originally immortalized from cultures of primary mouse hippocampal neurons [1]. The HT22 cells were a generous gift from the Salk Institute (La Jolla, CA, USA). HT22 cells were routinely maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, and in a 5% CO2 atmosphere at 37 °C.

Primary mouse hippocampal neurons were purchased from Lonza (M-HI-401, LOT numbers: 030807, 301,006, and 240,807, Basel, Switzerland). Mouse hippocampal neurons were isolated from E16–17, CD-1 mice which had been killed by decapitation, and then cryopreserved at primary passage. They stain positive for PGP and Tuj-1 and test negative for mycoplasma. Preparation of medium, coating of cell culture plates, thawing of cells / initiation of culture process, and maintenance of the cell culture was performed according to the manufacturer’s instructions. On day 1 of each experiment, thawing of cells and initiation of the culture process was started. Cells were seeded for experiments on day 4, 5, or 6 and underwent isoflurane or control exposures on the following day (day 5,6, or 7). The morphology of cultured cells was assessed once daily by microscopy.

Cell exposures

For isoflurane exposure, HT22 cells or primary hippocampal neurons were transferred into a Billups-Rothenburg chamber (MIC-101, Billups-Rothenburg, Del Mar, CA) [11]. An in-line calibrated anesthetic agent Vaporizer (Dräger, Lübeck, Germany) was used to deliver isoflurane to the gas phase of the culture wells. Control air (5% CO2, 21% O2, balance nitrogen) (Air Liquide, Düsseldorf, Germany) and 3% isoflurane were delivered into the chamber with 5 l/min over 20 min. Effluent isoflurane, and carbon dioxide concentrations were monitored during flushing the chamber via a sampling port connected to an anesthetic agent analyzer (DrägerVamos, Lübeck, Germany). The chamber was then placed immediately into an incubator and remained there for 4 h. Isoflurane concentrations (3%) were again checked after 30, 90, and 180 min after exposure while flushing the chamber for 5 min. Control experiments were performed in the same manner, except that no isoflurane was added.

Experimental protocol

Three hundred thousand HT22 cells were seeded in 6-well plates and underwent isoflurane exposure or control experiments without isoflurane. The experiments were repeated 6 times.

Primary hippocampal neurons were cultivated according to the manufacturer’s instructions. On culture days 4, 5, and 6 200.000 cells were seeded on a 24-well plate and underwent isoflurane or control exposures the following day (day 5, 6 or 7). For methylation analysis, this experiment was performed twice, resulting in 6 single experiments. Therefore, in the methylation analysis of primary hippocampal neurons, 2 experiments performed on day 5, 2 experiments performed on day 6, and 2 experiments performed on day 7 were included. For analysis of mRNA expression in primary hippocampal neurons, 3 independent experiments, in which cells underwent isoflurane or control exposures at day 5, 6 or 7 were included.

Immediately after 4 h of isoflurane exposure, DNA and RNA were extracted from the respective cells.

Methylation analysis

DNA was extracted with the DNeasy Blood & Tissue Kits (Qiagen, Hilden, Germany) and DNA concentration was measured by a spectrometer (Eppendorf, Hamburg, Germany).

Methylation of genes was analyzed with EpiTect Methyl II PCR Array Systems (Qiagen, Hilden, Germany), which examines the methylation status across a CpG-rich sequence in gene promoters. The method is based on the detection of remaining input DNA after cleavage with a methylation-sensitive and/or a methylation-dependent restriction enzyme (EpiTect Methyl II DNA Restriction Kit, Qiagen, Hilden, Germany). These enzymes digest unmethylated and methylated DNA, respectively. Following digestion, the remaining DNA is quantified by real-time PCR (RT2 SYBR® Green qPCR Mastermix, Qiagen, Hilden, Germany) in each individual enzyme reaction using primers that flank a promoter (gene) region of interest. We analyzed the DNA methylation status of 22 genes of the pathways focused EpiTect Methyl II PCR Array Systems (Qiagen, Hilden, Germany), i.e., apoptosis (EAMM-121Z, Table 1), cytokine production (EAMM-541Z, Table 2), and inflammatory response and autoimmunity (EAMM-521Z, Table 3). The relative fractions of methylated and unmethylated DNA are subsequently determined by comparing the amount in each digest with that of a mock (no enzymes added) digest using the ΔCt method as recommended by the manufactor (Qiagen, Hilden, Germany).

Table 1 Apoptosis Array
Full size table
Table 2 Cytokine Array
Full size table
Table 3 Inflammatory Response & Autoimmunity Array
Full size table

Quantitative real-time PCR

RNA was extracted with peqGOLD MircoSpin Total RNA Kits (Peqlab Biotechnology, Erlangen, Germany) at the indicated time point. RNA concentration was measured by a spectrometer (Eppendorf, Hamburg, Germany). First-strand cDNA was synthesized from 1 μg of total RNA with oligo-dT primers (Roche, Mannheim, Germany) by using Superscript II reverse transcriptase, as recommended by the supplier (Invitrogen, Carlsbad, CA). Relative quantification of mRNA expression was carried out by using primers for those genes whose methylation level showed a difference greater than 20% among in the primary hippocampal neurons with and without isoflurane exposure (Cxcl12, Cxcl14, Gata3, Il11, Il13, and Il4ra,; for primer details see supplementary Table 1, for gene details see Table 3), taking several considerations of reference gene selection into account [11]. Contamination with genomic DNA was circumvented by DNase treatment of RNA and by using RNA as template. The PCR reaction mix was prepared using the Quantitect SYBR Green Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. PCR was performed in the Step One Plus (Applied Biosystems, Foster City, USA). Real-Time PCR experiments were performed three times.

Statistical analysis

Data are presented as means ± standard deviation (SD). Data were analyzed using Graph Pad Prism 6 (Graph Pad Software, La Jolla, CA). Methylation analysis was done according to the manufacturer’s instructions. A change in methylation of > 20% was considered as significant, according to the instructions of the manufacturer and comparable to other studies using the EpiTect Methyl II PCR Array System [12, 13]. Here, DNA methylation was compared between apoptosis-related genes in pleomorphic adenomas of the salivary glands [13], respectively in genes involved in inflammation and autoimmunity in inflammatory bowel disease [12]. mRNA expression was done in genes with significant methylation changes, according to references [12, 13], and analyzed with an unpaired t-test. This An a priori alpha error p of less than 0.05 was considered to indicate statistical significance.

Results

Methylation of HT22 cells and primary hippocampal cells without isoflurane exposure

In a first step, we measured methylation of genes in HT22 and in primary hippocampal neurons in control experiments without exposure to isoflurane (Fig. 1a-c). Interestingly, the methylation status between the immortalized cell line HT22 and the primary hippocampal neuron cultures was markedly different. In general, methylation of genes in HT22 cells was greater than in the primary cells. Eight of 22 genes of the apoptosis array (Bid, Bik, Cideb, Dapk1, Hrk, Ltbr, Tnfrsf21, and Tnfrsf25) were significantly greater (> 20%) methylated in HT22 cells compared to primary hippocampal neurons (Fig. 1a). For example, methylation of Bik was 99.4% in HT22 cells, whereas it was 0.2% in primary hippocampal neurons (Fig. 1A). In the cytokine array, genes Bcl10, Bcl3, Cd114, Furin, Gata3, Irf1, Ltb, Myd88, and Stat5a showed > 20% methylation changes in the HT22 cells compared to the primary hippocampal neurons (e.g. Ltb: 99.6% vs. 0.4%) (Fig. 1b). Most differences were found in genes of the inflammatory array. Here, 12 of 22 genes (Cx3cl1, Cxcl12, Cxcl14, Gata3, Il10ra, Il11, Il13, Il13ra1, Il17ra, Il7, Ltb, and Tgfb1) were significantly greater (> 20%) methylated in HT22 cells compared to primary hippocampal neurons (Fig. 1c). Methylation of Cxcl12 was 99.7% in HT22 cells, whereas it was 0.6% in primary hippocampal neurons (Fig. 1c). Only one gene (Il18) was nearly completely methylated in the primary neurons (> 99.8%).

Fig. 1
figure 1

Methylation of genes in the functional gene panels of HT22 cells (black columns) and primary cultured hippocampal neurons (grey columns). Data represent means ± standard deviation (SD) and were obtained from 6 single experiments. Gene details are given in Tables 1, 2, and 3. a Apoptosis gene panel. b Cytokine gene panel. c Inflammatory gene panel

Full size image

Influence on isoflurane exposure on methylation in primary hippocampal cells

We measured whether isoflurane exposure alters methylation of genes in primary hippocampal neurons (Fig. 2). However, no changes were seen for genes of the apoptosis array (Fig. 2a) and the cytokine array (Fig. 2b). There was a greater than 20% methylation change for several genes in the inflammatory array after isoflurane exposure: Cxcl12, Cxcl14, Gata3, Il11, Il13, and Il4ra (Fig. 2c). Methylation of Cxcl12 was 0.6% without, but 27.5% with isoflurane exposure. For all genes, variation between the single experiments was seen (Fig. 2c). In HT22 cells, neither methylation of the genes in the apoptosis array (A), nor those in the cytokine (B) or in the inflammatory arrays (C) was altered by isoflurane (Supplementary Fig. 1).

Fig. 2
figure 2

a-c Methylation of genes in various functional gene panels in primary hippocampal neurons without isoflurane exposure (black columns) and with isoflurane exposure (grey columns). Data represent means ± standard deviation (SD) and were obtained from 6 single experiments. Gene details are given in Tables 1, 2, and 3. a Apoptosis gene panel. b Cytokine gene panel. c Inflammatory gene panel. dCxcl12 mRNA expression analysis in primary hippocampal neurons with and without exposure to isoflurane. Data are mean ± standard deviation (SD), obtained from 3 single experiments, and analyzed by students unpaired t-test.

Full size image

mRNA expression analysis of inflammatory genes in primary cells

To elucidate whether greater methylation of several genes in the inflammatory array with isoflurane exposure altered mRNA expression of these genes, RT primer were designed for genes, in which methylation was > 20% different after isoflurane exposure (Cxcl12, Cxcl14, Gata3, Il11, Il13, and Il4ra, for primer sequences see supplementary Table 1). mRNA expression was then measured in cDNA samples of cells exposed to isoflurane and in cells under control conditions. Expression of Cxcl12 was significantly lower under isoflurane exposure (p = 0.023, Fig. 2d), whereas no differences were found for the genes Cxcl14, and Gata3 (data not shown). No RT product could be obtained for genes Il11, Il13, and Il14ra.

Discussion

In our study, we demonstrate wide differences in the gene methylation status between the HT22 cell line and primary hippocampal neurons in the absence of isoflurane exposure, but greater methylation of several genes involved in inflammation upon isoflurane exposure and significant downregulation of Cxcl12 mRNA expression in primary hippocampal neurons. This could represent further epigenetic mechanisms of anesthesia related phenotype in the immature brain. Our results suggest that further studies investigating epigenetic effects of isoflurane or other volatile anesthetics should not be performed on HT22 cells but on primary hippocampal neurons, albeit technically more difficult.

A greater level of methylation was seen in the cell line HT22 than in primary hippocampal neurons. There are only few data on DNA methylation in immortalized cell lines compared to respective primary cells [14,15,16]. In general, there are cell-type specific methylation signatures preserved in primary cell lines and tissues as well as methylation differences induced by cell culture [16]. For example, the Forkhead box protein L2 promoter is hypermethylated in immortalized but not in primary cells [14]. This argues against a use of HT22 cells to study epigenetic changes, e.g., induced by volatile anesthetics.

In the primary hippocampal neurons, no differences were seen in genes in pathways of cytokine production and apoptosis, but there was a greater methylation of several genes in inflammatory pathways.

Isoflurane anesthesia can evoke neuroinflammation [7] and methylation of genes in the inflammatory pathway may be one molecular mechanism leading to neuroinflammation. Epigenetic mechanisms for isoflurane induced neuroinflammation have already been reported for the aged brain: isoflurane exposure led to declines in chromatin histone acetylation with resulting downregulation of the BDNF- tyrosine kinase receptor B signaling pathway and with increased levels of inflammatory Interleukin 1 beta and Interleukin 6 in aged rats [17]. Our study, therefore, may first indicate that epigenetic modulation of inflammatory genes may also play a role in the volatile anesthetic induced neuroinflammation in the immature brain and this should be investigated in further studies.

Special regard may be given to the C-X-C motif chemokine 12 (CXCL12), whose mRNA was significantly downregulated, while no differences were found for the genes Cxcl14 and Gata3 and no Real-Time PCR product could be obtained for genes Il11, Il13, and Il14ra. It might be speculated that expression of these genes is too low to become detective.

CXCL12 is a 68-amino-acid CXC chemokine with an essential role in several important processes such as vascular and neuronal development and hematopoiesis [18]. The CXCL12 response takes place at a very early stage of embryonic development and seems to be broadly operative whenever cell migration is required [19]. Mice lacking CXCL12 die prenatally and exhibit defects in vascularization, neuronal development, and hematopoiesis [20]. CXCL12 could be a key chemokine regulating the entrance of lymphocytes into the CNS [21] and immune factors, such as TNFα and interleukins might influence CXCL12 expression [22,23,24,25]. In further studies, a promoter characterization of the genes of interest, such as the Cxcl12 promoter, with special regard to the CpG islands, would be of great importance. The EpiTect Methyl II PCR Array System examines the methylation status across a CpG-rich sequence in gene promoter, however, the exact length of the amplified CpG island/promoter and the physical location of the annealing sites of the PCR primers are not disclosed.

Limitations of the study should be mentioned. One important item is the relevant dose and duration of isoflurane administration necessary to evoke DNA methylation changes, if any, and previous data are missing. Previous animal studies are difficult to interpret since long durations and supraanesthetic doses of isoflurane were administered to young animals [26]. While such intense exposures are maybe needed to investigate possible neuropathological and other pathological changes, these conditions hardly relate to human clinical practice. We chose an isoflurane concentration of 3 vol% inspired but a MAC value cannot be obtained for cell lines. Since the primary hippocampal neurons were exposed to isoflurane on culture days 5, 6, and 7 (mouse hippocampal neurons were isolated from E16–17 mice), therefore, one might speculate that the 3% isoflurane corresponds to 1 MAC [27], which is a usual clinical dose.

To this end, it also needs to be questioned whether the observed effect of isoflurane is transferable to other volatile anesthetics such as sevoflurane. In a landmark paper, neuroapoptotic properties of equipotent anesthetic concentrations of desflurane, isoflurane, or sevoflurane were compared in neonatal mice and similar neurotoxic profiles were found [27]. However, whether this finding holds true only for only this condition or has to be tested as research hypothesis needs to be clarified.

Furthermore, it needs to be questioned, whether exposure of the primary hippocampal neurons on days 5, 6, and 7 was plausible. We believe that this time frame represents a vulnerable phase of brain development. Exposure on three following days was done due to practical and cost reasons, however, analysis of neurons on always the same day may result in different conclusions. Exposure of neurons on three following days may although partly explain the variation of methylation in the experiments, which was only seen in the primary hippocampal neurons. It is important to note, that RNA for mRNA expression analysis was obtained from three independent experiments, indicating that our experimental setup was reliable.

For interpreting the results of the EpiTect Methyl II PCR Array System, we considered a methylation change of > 20% as significant. According to the manufacturer, the significance of the level of methylated DNA must be defined by the researcher. Methylated DNA may have biological significance if such methylation status is associated with a specific phenotype and measuring the corresponding gene expression between samples is recommended. While we did not investigate phenotype changes, we observed significant differences in mRNA expression.

Conclusions

In conclusion, we demonstrate wide differences in methylation status between HT22 cells and primary cultured hippocampal neurons, with increased methylation in HT22 cells, suggesting that further epigenetic investigations on effects of isoflurane should not be performed in HT22 cells. Furthermore, upon isoflurane exposure in hippocampal neurons we found a greater DNA methylation from several genes involved in inflammation with significant downregulation of Cxcl12 mRNA expression. Based on this a priori analysis, further investigations addressing possible DNA methylation changes in inflammatory genes due to exposure to anesthetics should be performed.

Full name of all genes is given in Tables 1, 2, and 3.